(Advances in The Study of Behavior 27) Anders Pape Møller, Manfred Milinski and Peter J.B. Slater (Eds.) - Stress and Behavior-Elsevier, Academic Press (1998)

Advances in
THE STUDY OF BEHAVIOR

VOLUME 27
Advances in
THE STUDY OF
BEHAVIOR
Edited by
J. B. SLATER
PETER
JAY S. ROSENBLATT
T. SNOWDON
CHARLES
MANFRED
MILINSKI
Stress and Behavior
A Volume in
Advances in
THE STUDY OF BEHAVIOR
VOLUME 27
Edited by
PAPEMOLLER
ANDERS
Lahoratoire d’Ecologie
UniversitP Pierre et Marie Curie
Paris, France
MANFRED
MILINSKI
Zoologisches Institut
Ahteilung Verhalten-sokologie Universitat Bern
Hinterkappelen, Switzerland
J . B. SLATER
PETER
School of Environmental and Evolutionary Biology
University of St. Andrews
Fife, United Kingdom
ACADEMIC PRESS
San Diego London Boston New York Sydney Tokyo Toronto
T h k book ir, p r i n t e d on acid-free paper. @
Copyright 0 1998 by A C A D E M I C PRESS
A l l Rights Reserved.
No part o f this publication may be reproduced or transmitted in any form or by any
iiicans. electronic or mechanical. including photocopy. recording, or any information
\torage and retrieval system. without permission i n writing from the Publisher.
The appearance of the code at the bottom o f the first page of a chapter in this book
indicates the Publisher's consent that copies of the chapter may be made for
personal or internal use of specific clients. This consent i s given on the condition.
however. that the copier pdy the stated per copy fee through the Copyright Clearance
Center. Inc. ( 2 2 2 Rosewood Drive. Danvers. Massachusetts 0 1923). for copying
heyond that permitted by Sections 107 or 108 o f the U . S . Copyright Law. This consent
does not extend to other kinds o f copying. such 3 s copying for general distribution. for
advertising or promotional purpoes. for creating new collective works. or for resale.
Copy fee\ for pre-199X chapters are as shown on the t i t l e pages. If no fee code
appears on the title page, the copy fee i s the hame as for current chapters.
0065-3454/9X $25.00
Academic Press
(I dii,i.\;mi Brui.i. & C'onilitrii?
o/'Iioi.c~orrrt
525 B Street. Suite 1900. San Diep, California 97_101-4495.USA
http:/Jwww.apnet.com
Acadeniic Press Limited

24-2X Oval Road, London N W I 7DX. U K
http://www.Iibuk.co.uk/ap/
International Standard Book Number: 0-1 2-004527-3
PRINTED IN THE UNITED STATES OF AMERICA

9X 99 00 01 01 03 Q W 9 8 7 6 5 4 3 2 I
Contents
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...
Xlll
The Concept of Stress and Its Relevance for Animal Behavior

DIETRICH VON HOLST
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
I1. The Concept of Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
I11. Social Stress in Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . 42
IV . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Stress and Immune Response

VICTOR APANIUS
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
I1. The Nature of Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
I11. The Nature of Immunocompetence . . . . . . . . . . . . . . . . . . 136
IV. Neurological Linking of Stress and Immunocompetence . . 140
V. Endocrine Linkage of Stress and Immunocompetence . . . . 142
Vl . Why Stress Alters Immunocompetence . . . . . . . . . . . . . . . 145
VII . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Behavioral Variability and Limits to Evolutionary Adaptation

under Stress
P. A . PARSONS
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
I1. Energy Limits to Adaptation . . . . . . . . . . . . . . . . . . . . . . . 158
I11. Variability and the Survival of Variants . . . . . . . . . . . . . . . 164
IV . Extending the Limits of Adaptation . . . . . . . . . . . . . . . . . . 165
V. From Stress-Resistance Genotypes to a
Connected Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
VI . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
V
vi CONTENTS
Developmental Instability as a General Measure of Stress

ANDERS PAPE M@LLER
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
I1. Genetic and Environmental Determinants of
Developmental Instability . . . . . . . . . . . . . . . . . . . . . . . . . . 186
111. Directional Selection and Developmental Instability . . . . . 190
IV . Fitness Correlates of Developmental Instability . . . . . . . . . 192
V . Practical Uses of Developmental Instability . . . . . . . . . . . . 193
VI . Conclusions and Prospects for Future Studies . . . . . . . . . . 206
VII . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
Stress and Decision-Making under the Risk of Predation:

Recent Developments from Behavioral. Reproductive. and
Ecological Perspectives
STEVEN L . LIMA
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
I1. Behavior of Feeding Animals: Classical Motivations . . . . . 217
111. Patterns of Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
IV . After an Encounter with a Predator . . . . . . . . . . . . . . . . . . 235
V . Social Situations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
VI . Reproduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
VII . Long-Term Consequences of Decision Making . . . . . . . . . 245
VIII . Ecological Influences and Implications . . . . . . . . . . . . . . . . 248
IX. Additional Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 261
X . Conclusions and Summary . . . . . . . . . . . . . . . . . . . . . . . . . 264
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Parasitic Stress and Self-Medication in Wild Animals

G. A . LOZANO
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
I1. Self-Medication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
111. Prophylactic Se1f.Medication . . . . . . . . . . . . . . . . . . . . . . . . 294
IV. Therapeutic Self-Medication . . . . . . . . . . . . . . . . . . . . . . . . 298
V . Skepticism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
VI . Behavioral Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
CONTENTS vii
VII . Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

VIII . Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . 310
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
Stress and Human Behavior: Attractiveness. Women’s Sexual

Development. Postpartum Depression. and Baby’s Cry
RANDY THORNHILL AND F. BRYANT FURLOW
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
I1. Human Attraction and Attractiveness . . . . . . . . . . . . . . . . 321
I11. Parent-Daughter Relations and Women’s Sexual
Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
IV . Postpartum Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
V . Infant Crying as a Signal of Phenotypic Quality . . . . . . . . . 352
VI . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
Welfare. Stress. and the Evolution of Feelings

DONALD M . BROOM
I . Feelings. Their Role and Their Evolution . . . . . . . . . . . . . 371
I1. Welfare. Stress. and Feelings . . . . . . . . . . . . . . . . . . . . . . . . 394
I11. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
Biological Conservation and Stress

HERIBERT HOFER AND MARION L . EAST
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
I1. Stress in a Conservation Biology Context . . . . . . . . . . . . . . 407
I11. Designing a Conservation Study to Measure Stress and
Its Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
IV. The Natural History of Stress . . . . . . . . . . . . . . . . . . . . . . . 428
V . Effects of Anthropogenic Stressors . . . . . . . . . . . . . . . . . . . 452
VI . Conservation Research and Management Activities as
Stressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
VII . The Equivalence of Natural and Anthropogenic Stressors 486
...
Vlll CONTENTS
VIII . Minimizing Occurrence and Impact of Stress in

Conservation Research and Management . . . . . . . . . . . . . . 488
IX . Conclusions: How Important Is Stress in Biological
Conservation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
X . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
Contents of Previous Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . 549

Contributors
Numbers in parentheses indicate the pages on which the authors’ contributions begin.
VICTOR APANIUS (133),Department of Biological Sciences, Florida International

University, University Park, Miami, Florida 33199
DONALD M. BROOM (371), Department of Clinical Veterinary Medicine, Univer-

sity of Cambridge, Madingley Road, Cambridge CB3 OES, United Kingdom
MARION L. EAST (405), Max -Planck -1nstitut f i r Verhaltensphysiologie, 0-82319

Seewiesen Post Starnberg, Germany
BRYANT FURLOW (319). Department of Biology, University of New Mexico,

Albuquerque, New Mexico 87131
HERIBERT HOFER (405). Max -P/anck -1nstitut far Verhaltensphysiologie, D -

82319 Seewiesen Post Starnberg, Germany
DIETRICH VON HOLST (1). Department of Animal Physiology, University of

Bayreuth, 95440 Bayreuth, Germany
STEVEN L. LIMA (215), Department of Life Sciences, Indiana State University,

Terre Haute, Indiana 47809
G. A. LOZANO (291), Department o,f Biology, University of California, Riverside,

Calljornia 92522
ANDERS PAPE M0LLER (181), Laboratoire d’Ecologie, UniversitP Pierre et

Marie Curie, Paris Cedex 5, France
P. A. PARSONS (155), School of Genetics and Human Variation, La Trobe Univer-

sity, Bundoora, Victoria 3083, Australia
RANDY THORNHILL (31Y), Department of Biology, University of New Mexico,

Albuquerque, New Mexico 87131
ix
This Page Intentionally Left Blank
Preface
The aim of the Advances in the Study of Behavior series remains as it

has been since the series began: to serve the increasing number of scientists
who are engaged in the study of animal behavior by presenting their theoret-
ical ideas and research to their colleagues and to those in neighboring fields.
We hope that the series will continue its “contribution to the development
of cooperation and communication among scientists in our field,” as its
intended role was phrased in the preface to the first volume in 1965. Since
that time, traditional areas of animal behavior research have been given
new vigor through the links they have formed with related fields and by
the closer relationship that now exists between those studying animal and
human subjects.
Beginning with Volume 25, which was on the special topic of Parental
Care, we departed from the previous policy of publishing articles on varied
subjects in every volume. This volume, titled “Stress and Behavior,” is the
second thematic volume. The next volume will again be a broad-ranging
one, as was the last, and it is our intention to continue the series with this
mixture of wide-ranging volumes of eclectic interest and occasional volumes
focusing on particular themes that appear timely to us.
Although volumes such as this do represent a new initiative, they do not,
we believe, violate the basic principles underlying the series. The specific
theme of Stress and Behavior was chosen because it is an especially exciting
and active area of research at present, and one to which researchers with
a wide range of approaches and backgrounds are making important contri-
butions. We have invited as contributors leading experts across this range,
thus giving a truly multidisciplinary perspective on the topic.
For this volume we have been fortunate to be joined by Dr. Anders
Pape Mflller as guest editor, and his expertise in the area has been of
immense help to us. Sadly, this will also be the last volume for which Dr.
Manfred Milinski will act as an associate editor: he has made a valuable
contribution to the series, for which we are very grateful. His place will be
taken by Dr. Tim Roper, of the University of Sussex, and we look forward
to benefiting from his broad interests and well-honed editorial skills in
future volumes.
P. J. B. Slater
xi
Stress & Behavior
Introduction
All organisms suffer from a deficiency of one or more resources during

their lifetime, and conditions for development, growth, survival, and repro-
duction are rarely, if ever, optimal. This was realized by Charles Darwin,
who used competition among conspecifics for limiting resources as a corner-
stone of his theory of natural selection. Although suboptimal living condi-
tions are widespread, the relationship between environmental conditions
and behavior has not attracted much attention from scientists (for excep-
tions, see Hoffmann and Parsons, 1989; Maller and Swaddle, 1997).
The conditions under which organisms live are frequently suboptimal,
and the difference between suboptimal and optimal conditions is often
perceived by individuals as causing a change in their state. Stress is an
appealing but illusive concept in biology, with definitions being almost as
numerous as the different fields of research. Although stress can be defined
explicitly in terms of concentrations of biochemicals involved in metabolism
(Ivanovics and Wiebe, 1982), a general but still operational definition is
that provided by Hoffmann and Parsons (1989): the state caused by any
agent that results in suboptimal performance and potentially causes perma-
nent damage to an individual.
Life has evolved under stressful conditions, although exceptions exist,
such as the relatively constant environments experienced by organisms
living in caves and hot springs. The fact that organisms generally live
and reproduce under adverse environmental conditions has not been well
appreciated by the majority of the community of biologists. Theoretical
evolutionary biology has not considered stress to be of overriding impor-
tance. For example, Sewall Wright’s famous shifting balance theory is based
on the concept of a fitness landscape with multiple valleys and peaks. The
unappreciated fact that different phenotypes may be expressed under poor
and optimal conditions may have enormous effects on the ability of species
to evolve from one peak to another. Empirical evolutionary biologists have
performed most of their studies on fruit flies, mice, and rats under benign
laboratory conditions where stress is much reduced compared to natural
situations. Interestingly, phenotypic and genetic parameters are not congru-
ent or even comparable in such situations, as revealed by recent studies
of animals reared under more and less benign conditions (Bijlsma and
Loeschcke, 1997; Maller and Swaddle, 1997).
This discrepancy between conditions under which animals are studied
and conditions under which they live also applies to behavioral research.
...
Xlll
xiv INTRODUCTION
Laboratory studies are generally performed under benign conditions with

ad libitum availability of food and a virtual absence of predators and para-
sites. Field studies are predominantly conducted in optimal habitats with
high population densities, whereas much less is known about behavior of
animals in low-quality, marginal habitats where stress is predominant.
Behavior under adverse environmental conditions is a topic of general
interest for two reasons. First, such studies may provide us with a better
understanding of the conditions under which most evolution has taken
place. Second, studies of behavior under stressful conditions may give us
a better understanding of the consequences of global change, as well as
provide us with important information on conservation biology.
In the first chapter, von Holst describes how social relationships, espe-
cially aggressive ones, can influence the physiological state of individuals
in many positive and negative ways. This chapter introduces the neuroendo-
crinology of the stress response. Stressed organisms are easy targets of
infectious diseases. The complex strategic use of the immune system some-
times affords suppression of immune defenses under stress as discussed by
Apanius in the second chapter. In the third chapter, Parsons presents an
energetic approach to the fitness of organisms that are challenged by biotic
and abiotic stress. Under stressed free-living conditions, his environmental
model suggests that favored “good genotypes” are likely to be stress resis-
tant and heterozygous. In the fourth chapter, Mgller discusses develop-
mental stability as a reflection of the ability of organisms to buffer their
developmental trajectories against disturbance. The inability to fulfill this
goal can be assessed in terms of fluctuating asymmetry, i.e., random devia-
tions from perfect symmetry, and used for studies of, for example, environ-
mental monitoring, animal welfare, and human medicine. In the fifth chap-
ter, Lima discusses the many ways in which behavioral decision-making
alters the nature of predator-induced stress. In the sixth chapter, Lozano
discusses the growing evidence that wild animals use self-medication in
response to parasitic stress. In chapter 7, Thornhill and Furlow examine
how stress interacts with human behavior in relation to physical attractive-
ness, the development of women’s sexuality, and parental investment in
babies. The biological function of feelings such as pain and fear may be to
affect an organism’s behavior in such a way that it maximizes the chances
that good things will happen and minimizes the chances that bad things
will happen. Broom discusses welfare, stress, and the evolution of feelings
in chapter 8. The last chapter, by Hofer and East, reviews why stress has
important implications for biological conservation and considers practical
ways in which conservationists can identify and tackle problems due to
stress.
INTRODUCTION xv
We believe that a concerted research effort concerning behavior of ani-

mals living under adverse environmental conditions will add considerably
to our understanding not only of the role that behavior plays in the general
coping strategies of animals, but also of the flexibility of behavioral strate-
gies under variable environmental conditions. The authors of the chapters
of this thematic volume of Advances in the Study of Behavior have shown
some ways in which this may be achieved.
A. P. M@ller
M. Milinski
References
Biljsma, R. and Loeschcke, V. (eds.) (1997). “Stress. Adaptation. and Evolution.” Birk-
hauser. Basle.
Hoffmann, A. A. and Parsons, P. A. (1989). “Evolutionary Genetics and Environmental
Stress.” Oxford University Press, Oxford.
Ivanovics, A. M. and Wiebe, W. J. (1982). Toward a working definition of stress: A review
and critique. In “Stress Effects on Natural Ecosystems” (G. W. Barrett and R. Rosenberg,
eds.), pp. 13-27. Wiley, New York.
M@ller,A. P. and Swaddle, J . P. (1997). “Asymmetry, Developmental Stability and Evolution.”
Oxford University Press, Oxford.
ADVANCES IN THE STUDY OF BEHAVIOR. VOL 21
The Concept of Stress and Its Relevance for

Animal Behavior
VON HOLST
DIETRICH
DEPARTMENT OF ANIMAL PHYSIOLOGY
UNIVERSITY OF BAYREUTH
95440 BAYREUTH, GERMANY
1. INTRODUCTION
Mammals live in social systems, which differ from species to species but
are relatively constant for any species, although some variation as a function
of the ecological situation is possible. These social systems are maintained
by constant contact between the animals, which not only affects the behavior
of the individuals, but may also positively or negatively influence their
fertility and health. The negative consequences of social interactions are
usually explained by the stress concept as shown in a particularly impressive
way in the Australian dasyurid marsupials of the genus Antechinus.
This genus is widely distributed in Australia and feeds mainly on insects
and small vertebrates. All species examined so far exhibit an extremely
synchronous life cycle: At the end of September-during the Australian
spring-the females give birth to their young, which are weaned in January,
but continue to live in harmony with their mothers for a few more months.
At the end of May, the young leave their birthplace and spread out within
their habitat. The short reproductive season commences in August, during
the Australian winter. During the search for females, the males roam their
territory and are continually involved in vehement fights with other males.
Following the 2- to 3-week reproductive season and before the end of the
first year of their life, virtually all the males “die off.” The females survive
and after a 1-month gestation period they give birth to their young. A new
cycle ensues (Woolley, 1966). The death of the males is due to typical stress
reactions characterized by a tenfold increase in the plasma levels of free
glucocorticosteroids and a simultaneous breakdown of the immune and
inflammatory responses. As a consequence, gastrointestinal hemorrhaging
associated with gastroduodenal ulcers, bacterially induced hepatic necrosis,
heavy parasitic diseases, and other infections cause the death of all males
1
Copyright 0 1998 by Academic Press
All rights of reproduction in any form reserved
0065-3454198 $ZS.MI
2 DIETRICH VON HOLST
within a short period of time (Barnett, 1964; Bradley et al., 1980; McDonald
et nl., 1981, 1986).
The physiological changes causing death are based mainly on the in-
creased levels of aggression between the males. Accordingly, if males are
captured before the breeding seasons and housed singly, they may live to
about 2 years of age as do females under natural conditions. This means
that the males die of stress mainly due to their enhanced aggression and
persistent sexual activity.
In this chapter, the significance of the stress concept in gaining a better
understanding of social mechanisms in nonhuman mammals will be exam-
ined. In the second section the development of this concept during the last
50 years and the resulting current understanding of different stress reactions
are described. The triggers of stress reactions are mainly psychical processes
resulting from the assessment of a situation by an individual. Dependent on
its coping behavior, these processes lead to different physiological response
patterns, which can result in a number of pathophysiological effects. In the
third section the most important currently applied methods in assessing
stress levels in animals are introduced. Particular attention is paid to meth-
odological problems as well as to the limits of interpretation. Focal points
are the sympathetico-adrenomedullary and pituitary-adrenocortical sys-
tems, the pituitary-gonadal axis, and the immune system. In the fourth
section an overview is provided of the relationships between social situa-
tions and stress responses, in which I concentrate mainly on our research
on the monogamous and territorial tree shrews and the polygamous and
territorial European wild rabbits. In these cases the social rank of an individ-
ual, as well as its sociopositive interactions with conspecifics, and the stability
of the social system are determinants in the effects of a social situation on
the individual’s vitality and fertility.
11. THECONCEPT
OF STRESS
A. INTRODUCTION
Few biomedical terms are as popular as stress. However, its definition
is as inconsistent as the research strategies of the scientists from a variety
of disciplines (biomedicine, psychology, or sociology) working on stress-
related topics (Lazarus and Folkman, 1984;Levine and Ursin, 1991; Weiner,
1991). Although it is probably impossible to find a definition that the
majority of researchers will agree upon, and some authors even suggest
that the concept is meaningless (e.g., Engel, 1985), the concept of stress
has a long history that goes back to the ancient Greeks. As early as the
STRESS AND ITS RELEVANCE FOR ANIMAL BEHAVIOR 3
year 400 B.c., Hippocrates ascribed the causes of disease to disturbing forces
of nature and referred to the adaptive responses of the body as the “healing
power of nature.” One hundred years later psychological stress was men-
tioned by Epicurus, who suggested that coping with emotional challenges
is a way of improving the quality of life (cited from Chrousos et al., 1988).
All recent stress concepts deal with the daily social and nonsocial stimuli
that are challenging or threatening to the survival, health, and reproductive
success of animals and that are, therefore, an essential part of natural se-
lection.
OF THE STRESS
B. DEVELOPMENT CONCEPT
Modern biomedical stress research is based in particular on the work of
the American physiologist Walter B. Cannon and his colleagues, and the
work of the Canadian physician Hans Selye.
1. Cannon’s Fight or Flight Syndrome

In 1929, Cannon published an important monograph entitled “Bodily
Changes in Pain, Hunger, Fear and Rage,” in which he summarized the
results of decades of research into the effects of emotional challenges on
physiological processes. Cannon did not regard emotions as purely subjec-
tive sensations, but as all-encompassing phenomena that also embrace ob-
jective physiological and ethological components and could, therefore, be
analyzed scientifically. This opinion is still held today (Buck, 1988a).
Cannon found a multifarious mosaic of changes in bodily functions in
both animals and humans in emotionally stimulating situations: a reduction
in gastric and intestinal function; an increase in heart rate, blood pressure,
and breathing rate as well as in the number of red blood cells and the sugar
content in the blood; and accelerated blood clotting. All these effects were
attributed by Cannon to the increased activity of the sympathetic nervous
system (Fig. 1).
Cannon not only concentrated on the explanation of these causal control-
ling mechanisms, but also questioned the adaptive value of this variety of
reactions. His conclusion was the following: All these effects increase the
capability of an individual to react actively to critical situations in its envi-
ronment-to prepare it for fight or flight.
However, Cannon also realized that not every emotional process results
in the activation of the organism. A difficult situation that cannot be changed
by action can trigger apathetic, inactive behavior and lead to, among other
things, a reduction in pulse rate and blood pressure. An early and detailed
description of this reaction is given by Charles Darwin in 1872 in his book
Release of
Decreased
Lipolysis Glycolysis
clotting time
-
I I I
increased
ventilation
1
Increased blood flow to

brain, heart, and skeletal muscles
I
I Increased plasma levels of I
FIG. 1. Cannon’s fight or flight response: Activation of the sympathetic nervous system and release of the adrenomedullary hormones
epinephrine and norepinephrine. Their effector organs and the effects on the whole organism are shown (see also Fig. 11).
STRESS A N D ITS RELEVANCE FOR ANIMAL BEHAVIOR 5
“The Expression of the Emotions in Man and Animals,” in which he spoke

of the feeling of despair or grief.
2. Selye’s General Adaptation Syndrome
Cannon and associates were concerned mainly with the acute responses
of an individual to potentially dangerous stimuli, while recognizing that
repeated exposure to such stimuli results in adaptive changes in the organ-
ism that make it more resistent to challenge.
The adaptation of an organism to chronic challenges, on the other hand,
was the main interest in the research by Hans Selye, who also introduced
the term stress into biomedical research (Selye, 1950). Contrary to the use
of this term in everyday language and in other scientific disciplines, he
designated stress as the response of an organism to any strong and poten-
tially damaging stimulus, while for the damaging stimulus he introduced
the term stressor.
a. General Adaptation Syndrome. In 1936 Selye published a short paper
“A Syndrome Produced by Diverse Nocuous Agents,” describing for the
first time a pattern of physiological reactions in response to various damag-
ing agents or critical situations, such as injuries, cold, infections, intoxica-
tions, burns, o r strong muscular exercise.
An organism responds to these different stressors with stimulus-specific
responses, such as with immunological responses to infections or with in-
creased erythrocyte numbers to oxygen deficits. However, no matter how
variable the nature of these stressors, according to Selye, they always elicit
the same pattern of physiological responses, which seem to represent a
generalized effort of the organism to adapt itself to the new situation. The
response of the organism to stressors is accomplished in three stages, which
Selye called the general adaptation or stress syndrome.
1. Alarm Reaction. The initial responses to physiological changes in-
duced by a stressor are thymolymphatic involution, gastrointestinal ulcer-
ation, and loss of cortical lipids and medullary chromaffin substances from
the adrenals, indicating an activation of the sympathetico-adrenomedullary
and pituitary-adrenocortical systems. If the stressor is too strong (severe
burns, extreme temperatures), death may result within a few hours.
However, if the stressor is not too strong and has only a brief effect,
then it usually has no further consequences for the organism, which quickly
regains its original state.
2. Stage of Resistance. If the challenge persists, the body adapts itself
to tolerable stressors, such as very low temperatures or unavoidable physical
exertion, by changing its entire physiological state. According to Selye, the
increased activity of the adrenal cortex during this stage of defense or
resistance is of particular importance. The adrenal cortex adapts to the

increased production and secretion of its hormones by markedly increasing
its size. Concomitant with this, those functions unnecessary to coping with
the stressor, such as growth, gonadal activity, and immunological resistance
are suppressed.
3. Stage of Exhaustion. If the stressor is sufficiently severe and prolonged,
the adaptation mechanisms will finally fail and lead to the death of the indi-
vidual.
Long-term, tolerable stress therefore impairs fertility and vitality in ani-
mals. Simultaneously, the initial advantageous physical adaptive reactions
(particularly the increased production of adrenocortical hormones) were
thought by Selye to lead to a number of diseases (referred to as “diseases
of adaptation”), ranging from high blood pressure and gastric ulcers to
diabetes and cancer (Selye, 1950, 1976, 1981).
This concept of stress had a lasting effect on research. Ever since the
1950s, hundreds of scientific publications with the term stress in their titles
have been published each year. Due to the central role of the adrenocortical
system in the Selyean concept, research on stress centered to a large extent
around the adrenal cortex and its hormones, while other endocrine re-
sponses or systems such as the gastrointestinal or the adrenomedullary
system were largely neglected, even though changes in these systems were
clearly recognized. As a consequence, it became common practice to equate
stress with adrenocortical activity: Increased serum levels or excretion rates
of glucocorticosteroids, such as cortisol and/or corticosterone, or other
indications of heightened adrenocortical activity were used as an index of
the adaptation of an organism to a stressful situation or to the intensity of
a stressor. Although, even by today’s standards, this approach may appear
attractive methodologically, it is important not to equate stress with adreno-
cortical function, as the responses of an organism to new and sudden
demands comprise almost all physiological systems. Heightened adrenocor-
tical activity constitutes only one part of this response pattern and is in no
way sufficient to characterize the stress state of an animal, especially because
adaptive responses to stressful situations may occur without any heightened
adrenocortical activity (discussed later).
b. Physical versus Psychical Stress. While originally the adrenocortical
activity was assessed by changes in adrenal gland weight and morphology,
developments in the late 1950s yielded the first biochemical methods en-
abling determination of adrenocortical activity by measurement of their
hormone levels in plasma or urine. This led to a growing interest of psychol-
ogists and physiologists in emotionally induced adrenocortical activation.
By 1956, Mason and Brady had demonstrated for the first time increased
17-hydroxycorticosteroid plasma levels in rhesus monkeys in an emotionally

distressing situation (“conditioned anxiety”), and at the same time pre-
viously impossible studies on humans began. In the following years, count-
less studies on mammals (including humans) demonstrated strong adreno-
cortical activation not only during acute emotional arousal, but also in long-
lasting emotionally disturbing situations.
Nowadays, emotional “loads” are cited as the most common reasons for
stress in humans and, as pointed out by Ursin and Olff in a recent review
(1993), emotional processes are also the most commonly used stressors in
animal research. These emotional processes must be considered even when
the experimenter assumes he or she is dealing with physical stressors. It is
necessary to bear in mind, however, that an activation of the adrenocortical
system can also be induced without any concomitant emotional arousal
(such as during surgery under deep anesthesia or during infections and the
resulting release of mediators by the immune system).
A most important contribution to the modern stress concept is the work
of Mason and associates on the effects of psychological influences on the
general endocrine response pattern (Mason, 1968a,c). His own work, as
well as the results of the relevant literature, led Mason to the conclusion
that situations of novelty, uncertainty, or unpredictability are especially
potent in inducing heightened adrenocortical activity. Today it is generally
accepted that unpredictability is most effective in stimulating adrenocortical
activity in a variety of situations. Correspondingly, if an individual is given
information about the occurrence of an adverse stimulus, its predictability
leads to a reduction of the adrenocortical response. One illustration of this
is the study by Dess and associates (1983) on dogs that were subjected
to a series of either predictable or unpredictable electric shocks. In the
predictable condition the animals were presented with a tone prior to the
onset of shock, while in the unpredictable condition, no tone was presented.
Dogs that did not have the signal preceding the shock showed an adrenocor-
tical response two to three times that observed in animals with the predict-
able shock experience.
Furthermore, as shown by Mason, even subtle everyday changes in the
environment, usually not considered as stressful, such as presence or ab-
sence of familiar persons in a room in which monkeys were kept in cages,
can result in measurable changes of adrenocortical activity. These results
suggest “that the central nervous system exerts a constant ‘tonicity’ upon
this endocrine system, in much the same fashion as has been previously
demonstrated for the autonomic and skeletal muscular effector systems”
(Mason, 1968b).
c. UnspeciJicity of the Stress Response. Mason also questioned the basic
premise of the Selyean stress concept of a nonspecific response by an
organism to many different stimuli or agents (stressors). Instead, he consid-

ered the Selyean stress response to be a specific physiological response to
its corresponding psychological reaction, which is probably induced by the
different Selyean stressors. Whether an animal is immobilized, subjected
to unavoidable electric shocks or extreme temperatures, or whether it is
forced to swim to exhaustion, it is always in a hopeless situation that is out
of its control and that may be responsible for the adrenocortical activation
(Mason, 1968b). The same opinion was held by Bush (1962, p. 321), who
stated in a review:
It is probable that very severe burns, and large doses of certain agents such as bacterial
pyrogens, histamine, and peptones. cause a brisk release of ACTH that is independent
of any emotional concomitants; but . . . severe exercise, cold, and fasting produce little
or no effect on the secretion and metabolism of cortisol in man unless they are part of
a situation that provokes emotion. (1962, p. 321)
d. Predictability and Control. As mentioned above, the typical Selyean

stress response occurs in those situations that are characterized by uncer-
tainty or unpredictability. Prolonged stress responses can incur a high bio-
logical cost, leading to a number of immunological, gastrointestinal, and
cardiovascular changes that may reduce the vitality of the animals. There-
fore, mechanisms have evolved whereby the animals can reduce excessive
adrenocortical activation. The most important factor involved in reducing
hormonal responses to adverse stimuli is control. Control can be defined
as the capacity of an animal to produce active responses during the presence
of an adverse stimulus. These responses may allow the animal to avoid or
escape from the stimulus, but they may also provide the animal with the
opportunity to change from one set of stimulus conditions to another, rather
than to escape the adverse situation entirely. In both cases control reduces
an animal’s physiological stress response.
Particularly impressive support for this is provided by the research con-
ducted by Jay M. Weiss on the development of gastric ulceration in labora-
tory rats (Weiss, 1972). In one of the earliest experiments, two rats were
restrained in an apparatus for 21 hours with identical electrodes on their
tails attached to the same shock-delivering device. Every minute a tone
was presented to the rats for 10 s, which was followed by a light electric
shock. One of the animals (“avoidance-escape rat”) was given the possibility
of avoiding the shock by touching a panel with its nose during the presenta-
tion of the signal, or of terminating it after the beginning of the shock; the
other rat had no possibility of influencing the shock outcome (“yoked rat”).
Every time the avoidance-escape rat received a shock the helpless yoked
rat was given exactly the same shock. Thus, the two animals received
exactly the same physical stressor, but they differed in their control over
the situation. The two most important results of this study were (1) Simply
receiving shock itself is not in particular responsible for the production of
ulcers, but rather whether or not the animal is able to control the shock.
The helpless yoked rats developed much more ulceration than did their
partners; and (2) The more often an animal terminated the signal andlor
the shock by its behavior, the less ulceration developed. That is, animals
that can exercise control over a stressful situation do receive the relevant
feedback when they respond by getting the information that they are “doing
the right thing.” This is never the case in the helpless yoked animals. These
results led Weiss to the conclusion that the most important aspect of an
animal coping in a stressful situation is whether or not it can predict the
consequences of its behavior. This conclusion was elegantly confirmed by
experiments in which avoidance-escape animals with control over shock
were given a brief electric shock every time they performed the previously
correct response. Thus, each avoidance-escape response now produced
precisely the wrong kind of feedback stimulus, a shock. In this “negative”
feedback situation, the animals developed even more ulceration than did
their helpless yoked partners (Fig. 2).
The results of this research have been confirmed many times over by
experiments designed to embrace endocrinological parameters and carried
out on other species: Animals that are allowed to control the stimulus or
I t-7 Nonshock
p < .05
Signal Signal + punishment
FIG.2. Length of gastric lesions (medians) of nonshock, avoidance-escape, and yoked

groups of rats exposed to shock pulses that were preceded by a warning signal (lefi) and of
groups that perceived a shock pulse whenever they performed an avoidance-escape response
(right). Significant differences between the two shock groups and nonshock groups are indi-
cated: *p < .05; **p < .01; ***p < .001. For details see text. Adapted from Weiss (1971).
with kind permission from American Psychological Association, Washington, D.C.
situation show less (and in some cases no) physiological stress responses
(e.g., glucocorticosteroid levels not different from those of undisturbed
controls), whereas their yoked counterparts exhibit extremely high levels
of glucocorticosteroids and other signs of stress (e.g., Davis et al., 1977;
Hanson et al., 1976; Seligman, 1975; Weiss, 1984). Hence, current opinion
links Selye’s stress response or the activation of the pituitary-adrenocortical
system to psychological processes, resulting from uncertainty to loss of
control and helplessness.
3. Active and Passive Stress Responses
An important modification of the original Selyean stress concept was
made in 1977 by James P. Henry and Patricia M. Stephens. In their mono-
graph “Stress, Health, and the Social Environment” they summarized the
results of zoological, psychological, sociological, and medical research into
stress and concluded that two independent chronic stress reactions needed
to be distinguished from each other: active and passive stress.
The central theme of this concept of two different stress responses is the
relationship between styles of coping; limbic (emotional) processes and
neuroendocrine stress responses. Every threatening stimulus or challenge
to control immediately induces Cannon’s fight or flight response, followed
within a few minutes by adrenocortical activation as the animal makes a
behavioral effort to ensure that control over a conspecific or a situation is
retained. If control is not possible, different types of coping are seen in
nonhuman animals and humans alike, which clearly differ behaviorally and
physiologically (Henry, 1986, 1992).
a. Active Chronic Stress. If an animal reacts with a style of coping charac-
terized by active attempts to control the situation, for example, by fighting
to maintain or defend a social position or a territory or by fleeing to avoid the
situation, Cannon’s sympathetico-adrenomedullary system is chronically
activated; the activity of the adrenocortical axis can, but may not necessarily,
be increased in this response. According to this concept, this active stress
response is characterized by subjective feelings of anger or fear, depend-
ing on the context. Chronic active stress or the constantly heightened
sympathetico-adrenomedullary activity may lead to arteriosclerosis and
cardiovascular diseases. Recent studies have even shown a distinct re-
sponse pattern, activated by the brain in differing emotional states within
the sympathetico-adrenomedullary system. The neurosympathetic outflow
of norepinephrine, the “fight hormone,” can be independently activated
by the “flight hormone,” epinephrine, that is released from the adrenal
medulla (Hucklebridge et al., 1981; de Boer et al., 1990).
b. Passive Chronic Stress. When active coping (e.g., by flight) is not
feasible, a state of helplessness emerges, characterized mostly by immobility
and symptoms indicative of depression. This passive stress response is

characterized by greatly enhanced activity of the pituitary-adrenocortical
system, while the activity of the sympathetico-adrenomedullary system re-
mains more or less unchanged (Fig. 3).
Perceived
stimulus
Threat to cont ss of control
Active behavior
@-, Passive behavior
(Fight -flight) (Nonaggressive)
Behavioral arousal with Inhibition of spatially organized

challenge to status behaviors and status
Defense reaction Defeat reaction

Territorial or status control Loss of territorial or
with mobility, display, and status control,
low sex drives
I I
Control Striving Loss of control

YAnger") ("Fear") ("Depression")
Norepinephrine t t tc
Epinephrine tc
t cf
Glucocorticosteroids cc t t
Sex hormones t 4 4
FIG. 3. Schematic diagram contrasting the active and passive stress responses. The
sympathetico-adrenomedullary system is divided into two branches: one of fight, anger, and
norepinephrine; another of flight, fear, and epinephrine. Adapted from Henry et al. (1995).
with kind permission from Lippincott-Raven Publishers, Philadelphia.
c. Coping Behavior and Appraisal. These two styles of coping depend

on the appraisal of the situation by the animal and on the quality and/or
duration of the stressor. They thus may represent alternative strategies to
the solution of a problem. For example, to avoid the constant attacks of a
dominant rival, active escape may in the short term have the same result
as hiding quietly in the corner of a cage, especially if escape is impossible.
Furthermore, van Oortmerssen and associates (1985) demonstrated in a
study of wild house mice (Mus rnusculus) that the two coping patterns may
play different roles depending on the social structures and dynamics of
populations: Aggressive mice do better in settled stable demes, whereas
nonaggressive mice fare better in growing colonies.
Appraisal of a stimulus or situation as well as the resulting coping behav-
ior are basically psychological processes. There are, therefore, no clear
relationships between stimuli imposed on individuals and their physiological
responses. It is the behavioral, psychological, and thus the physiological
responses of individuals to stimuli that differ depending on their genetics,
prenatal influences, and especially postnatal learning processes (e.g., Fok-
kema et al., 1988; Henry et al., 1993).
A striking example for the significance of social experiences on stress
responses is provided by the work of Sachser and associates on guinea pigs
(Cavia aperea f porcellus). Male and female guinea pigs can be kept in
large groups without any behavioral or physiological signs of stress. When
two adult males from different colonies are confronted in an experimental
arena in the presence of an unfamiliar female, they arrange themselves in
a dominance order within a short time in the absence of any serious fighting
(Sachser, 1986; Sachser and Lick, 1991). However, if two males, each reared
with only a single female, are confronted in the same way, both display
continuously very high levels of aggressive behavior and extreme stress
responses and die within a few days unless separated (Sachser and Lick,
1989). The ability to come to a peaceful arrangement with conspecifics is
dependent on social experiences with male conspecifics around and shortly
after puberty (Sachser, 1993). A few low-key confrontations with an unfa-
miliar male, introduced into their enclosure 5 times for 10 min between
the age of 90 and 138days, were sufficient to reduce the fights with unknown
males in later chronic confrontations to the same low levels of animals
raised in colonies. That is, only 50 min of aggressive experience around
puberty is required to enable adult male guinea pigs to come to a stress-
free arrangement with conspecifics (Sachser et al., 1994).
The crucial role of social experiences for behavior and stress responses
was confirmed in a further approach (Sachser and Renninger, 1993).
Colony- and individually reared males were singly introduced into unfamil-
iar colonies of conspecifics for a period of 30 days. Colony-reared males
easily adjusted to the new social situation: On the first day they only
explored the new environment but did not court any female, thereby avoid-
ing attacks from the male residents. In the course of the following days
they gradually integrated into the social network of the established colonies.
Changes could not be determined in either their body weights or their
plasma concentrations of cortisol, androgens, or catecholamines. In con-
trast, individually reared males were involved from the beginning in court-
ship behavior, threat displays, and fighting. As a consequence, they re-
sponded to the new situation with substantial body weight loss as well as
with extreme increases in plasma cortisol levels (Fig. 4).
These data from guinea pigs clearly demonstrate the causal relationship
between social experience around puberty, behavior of the individuals as
adults, and the degree of their stress in unfamiliar social situations.
d. General Physiological Response Pattern, Physiological studies of re-
cent decades have revealed that many, if not all, neuroendocrine systems
respond to stressors. In his comprehensive treatise on motivation and
emotions, Buck (1988b) differentiates between behavior involved in self-
preservation (offensive and defensive behavior) and behavior concerned
with reproduction. These are accompanied by the arousal of different parts
of the limbic system and the hypothalamus as well as by patterns of neuroen-
docrine response, each peculiar to the particular emotion involved. Apart
from modifications in the sympathetic nervous system and the adrenal
700 1 I I Males raised

-f in colonies
m -
s
-
m
-0 individually
-
.500
a
E
c
0
100
-1 0 -5 0 5 10 15 20
Days before and after introduction into an unfamiliar colony
FIG. 4. Plasma cortisol levels ( M 2 SEM) of 6 colony- and 6 individually reared male
guinea pigs before and after transfer into an unfamiliar colony. Significant differences between
the two groups: **p < .01; ***p < ,001. Adapted from Sachser and Renninger (1993), with
kind permission from II Sedicesimo, Florence, Italy.
14 DIETRICH VON HOLS’I
glands, examples are to be found in the modification of neuroendocrine

systems that are involved in the regulation of reproduction (e.g., follicle-
stimulating hormone [FSH], luteinizing hormone [LH], testosterone, estro-
gen, prolactin), in metabolism (e.g., growth hormone, thyroid-stimulating
hormone [TSH], thyroxine, insulin), in osmoregulation and regulation of
blood pressure (e.g., aldosterone, vasopressin, renin), and in immune re-
sponse.
In accordance with the Selyean hypothesis, it appears that gonadal activity
is always inhibited by passive stress, whereas, dependent on the context,
active stress can have an inhibiting or activating effect (see Section 111,BJ).
However, the immune system appears, at least in the long term, to be
more or less inhibited by all stress reactions. Divergent hormonal patterns
probably have differential effects on the function of the immune system.
Current knowledge is insufficient regarding the other systems and prevents
any general statements on their participation in a given stress reaction or
on their long-term effects.
Future findings will most certainly lead to further differentiation of the
Henry-Stephens concept, particularly regarding the participation of the
immune system in stress reactions. However, this concept of two indepen-
dent stress axes has proved durable in zoology as well as in medicine and
psychology over the past 20 years (e.g., Bohus et al., 1987; Henry and
Meehan, 1981; von Holst, 1986a,b; Lemaire et al., 1993; Lundberg and
Frankenhaeuser, 1980; Mormkde et al., 1990).
4. Stress-A Useful Concept for Behavioral Research
In a very general form, Selye (1952) defined stress as “a non-specific
response of the body to any demand made on it.” It is only in this sense
that the term stress can be employed usefully today. In contrast to the
original Selyean assumption, “nonspecific” must be interpreted as those
reactions triggered within the body that are not a result of peripheral
changes (e.g., a drop in blood sugar content or blood pressure) and, there-
fore, represent correction mechanisms of homeostatic processes. These are
neuroendocrine response patterns induced by the central nervous system,
which change the organism’s physiological state and thereby generally lead
to its activation. These neuroendocrine stress reactions differ depending
on the situation as well as the behavior of the animals and concomitant
emotional processes.
This definition of stress in no way implicates definitive reaction patterns
or the participation of specific endocrine systems. However, it does assume
physiological reactions that could be detrimental to the individual if they
reached sufficient intensity or were of long duration. Social stress or psycho-
social stress describes the state of an animal, in which interactions with
conspecifics trigger central nervous processes, which themselves induce

physiological reactions that can lead to detrimental effects on the animal’s
vitality in the long term. Reduced fertility is not automatically also a conse-
quence of these stress reactions.
As the underlying factor in stress reactions is usually to be found in
emotional processes, a given situation or stimulus can vary in its effect
from one individual to the next by acting as an extremely damaging stressor,
a harmless influence, or even a positive trigger. Hence, the recording of
physiological stress reactions is-independent of the relevance of these
reactions to diseases-a methodological basis used to rate the appraisal of
a given situation by an individual and, therefore, to evaluate the biological
significance of social interactions and situations as well as that of nonsocial
influences (such as climatic factors or housing conditions).
C. ASSESSMENT
OF STRESS
1. General Methodological Considerations

Basically the effects of stressors can be measured on two levels, which,
of course, are not exclusive.
a. Epidemiological Approach. This approach is often taken in medicoso-
ciologically oriented research; the actual aim of research is to clarify the
role of stressors in the development of malfunctions and diseases, from
cardiovascular, gastrointestinal, or renal diseases to tumor growth and infer-
tility. These mainly epidemiological studies usually focus on a few variables
relevant for the respective organ or system, without paying much attention
to the underlying physiological (regulatory) mechanisms (e.g., Ader et al.,
1991; Adler et al., 1986; Dohrenwend and Dohrenwend, 1974; Friedman
and Rosenman, 1974; Levi, 1971; Price, 1982).
b. Physiological Approach. This approach mainly investigates those
structures of the central nervous system involved in stress reactions and
their effects on peripheral neuroendocrine and immunological parameters.
It is not the aim of this chapter to discuss those central nervous structures
involved in controlling neuroendocrine processes, but rather to focus on
changes in peripheral parameters. These are of particular importance, as
they not only indicate the presence of a stressful situation, but also allow
limited statements on possible pathophysiological consequences of the situ-
ation for the individual. Hence, this is the preferred approach in research
into stress (for details, see Section 111,B).
A multitude of very different physiological parameters must be assessed,
as individuals can react to the same stimuli in very different ways. However,
even today this is not the case in most studies. Measures are usually selected
on the basis of methodological constraints rather than based on present
knowledge, which makes interpretation of apparently contradictory results

often difficult or impossible. Although the determination of many different
hormones, immunological mediators, and other clinically relevant parame-
ters present few problems today, the interpretation of data is complicated
by numerous possibilities for methodological errors. These are described
briefly in the following section (for details, see textbooks of endocrinology).
2. Methodological Problems
a. Animal Housing and Handling. Every organism responds to chal-
lenges with arousal responses of varying intensity (acute stress response).
Moving an animal that is accustomed to a specific laboratory environment
into a new cage or an unfamiliar room is sufficient disruption to act as a
strong stressor for hours (e.g.. laboratory rats: Schuurman, 1981) or even
days depending on the species. This can result in a total masking of actually
interesting social situations or interactions. Thus, in tree shrews, transfer
into a new cage within an experimental room results in increased serum
levels of glucocorticosteroids and catecholamines for about 1week. Interest-
ingly, serum levels of testosterone follow a biphasic pattern; after an initial
decrease over a few days, the testosterone levels increase significantly above
initial levels as the animals become habituated to the new situation (Fig.
5). This same biphasic testosterone response was found already in 1968 by
Mason and co-workers (1968a) in rhesus monkeys during and after a 3-
day avoidance session.
All handling of animals (e.g., weighing or the taking of blood samples)
also functions as a stressor that acts on the corresponding variables within
150 12 males per

-a,
u)
sampling point
-
-a,
Q
.- 125
*-
0
..
* * -0- COrtlSOl
--c - Norepinephrine
75
-: .-m. Testosterone
0 5 10 15 20
Days after transfer into a new room
FIG. 5. Some endocrine responses of male tree shrews after a transfer into an unfamiliar
room. Significant differences to initial values: * p < .05.
seconds (catecholamines), minutes (glucocorticosteroids, thyroid and go-

nadal hormones), or a few hours (some immunological parameters); the
repeated taking of blood samples on consecutive days can therefore lead
to extensive stress reactions. After withdrawal of larger quantities of
blood, a physiological stimulus is added to the psychological one, leading
among other things to a sustained stimulation of the sympathetico-
adrenomedullary system of variable duration. Finally, even in laboratory
conditions, changes in climatic conditions or in food and introduced (nonvir-
ulent) germs are capable of influencing many different endocrine as well
as other parameters that indicate the presence or absence of stress. These
problems can largely be avoided by standardized laboratory conditions.
b. Diurnal and Other Variations of Physiological Parameters. There are
a number of other problems associated with the measurement of hormone
levels. In the majority of hormones secretion is not continuous, but occurs
in a pulsatile fashion. The pulse amplitude is usually highest at the beginning
of the activity period and lowest at the end, resulting in a marked diurnal
rhythmicity of the hormone output (Fig. 6). Due to this pulsatile secretion,
the hormone concentrations in the blood can change by a factor of 10
or more within minutes. Hence, baseline values exhibit much intra- and
interindividual variation, even if great care is taken to exclude all potential
interfering factors. This prevents interpretation on the level of the individual
of most endocrine parameters based on single blood samples. Although it
is possible to obtain blood samples from larger laboratory animals, over
several hours and up to a few days, by insertion of cannulas into the blood
vessels, this method is generally stressful to the animals, inhibits freedom
of movement, and is therefore only of limited application in laboratory
experiments (Fagin et al., 1983; Schuurman, 1981). Furthermore, many
physiological parameters also show annual rhythms or other periodicities,
which may influence hormone values.
It is possible to circumvent the general influence of such rhythms in
controlled laboratory conditions by, for example, always taking blood sam-
ples for endocrinological investigations at the same time of day. Neverthe-
less, because of the pulsatile secretory pattern of hormone release, marked
individual variations will remain. Furthermore, no information is available
on the influence of changing day length and other naturally occurring factors
on such daily rhythms and, therefore, how comparable such values are even
if they are collected at the same time each day. It is also more or less
unknown whether stressful situations lead always to the same changes in
the levels of different parameters during the day or not. A case in point
are our investigations carried out on tree shrews with the aid of telemetry,
which reveal a particular increase in heart rate during the night (sleep
Body temperature ("C) Heart rate (beatshin)

41 360
39 280
37 20 0
35 120
Triiodothyronine (nglrnl serum) Glucose (mg/100 ml blood)

06 140
05 120
04 100
03 80
Cortisol (ng/ml serum) Testosterone (ng/ml serum)

12 24
0 18
4 12
0 6
0 4 8 12 16 20 24 0 4 8 12 16 20 24
Time Time
FIG. 6. Diurnal rhythms of some physiological parameters in male tree shrews. Night
periods are characterized by dark color. Depending on the blood parameter, each point
represents the mean (t SEM) of 20-80 males: heart rate and body temperature are hourly
means ( 2 SEM) measured with implanted radio transmitters from 12 males and females.
periods) if the animals are in a stressful situation, even though they appear
to be sleeping normally (e.g., Figs. 20 and 40).
In summary, in order to gain relatively reliable data on endocrine and
other physiological processes on the individual level, it is necessary to
collect these data only on individuals that are kept in a controlled laboratory
environment; in natural conditions in the field it is usually possible to detect
only strong effects.
3. Physiological Markers of Stress
In this section, I shall briefly discuss those methods that appear to me
to be the most important or those that are most commonly employed in
assessing the level of stress in an individual, as well as their application
and power. To begin with, those systems will be briefly introduced that are
necessary to understand the choice of variables that are measured.
a. Pituitary-Adrenocortical System. Each of the paired adrenal glands
of higher vertebrates is composed of two distinct and functionally different
tissues-the adrenal cortex and the adrenal medulla. The cortex forms the
outer part of the adrenal gland and consists of three zones: the outer
glomerulosa; the zona fasciculata; and the inner zona reticularis, all of
which produce large numbers of different steroid hormones.
Glucocorticosteroids. For the present discussion, the hormones of the
zona fasciculata are of special relevance, as they are released immediately
in stressful situations. The most important and biologically relevant hor-
mones are cortisol and corticosterone, the presence of which varies from
species to species. For example, it is corticosterone almost exclusively that
is found in the blood of rats and mice; in primates and guinea pigs cortisol
is predominant; whereas other species (such as rabbits, hamsters, or tree
shrews) exhibit both hormones, although they are liable to differential
change during stress (e.g., rabbit: Kass et af., 1954; Krum and Glenn, 1965;
hamster: Ottenweller etal, 1985). Because of their strong effects on carbohy-
drate and protein metabolism, all hormones of the zona fasciculata are
grouped together as glucocorticosteroids: They increase the production of
glucose from protein resources, and this is then stored in the liver as
glycogen (a process referred to as gluconeogenesis), thus increasing the
available glucose necessary for energetic processes during stress. Further-
more, they inhibit inflammatory processes and suppress many immunologi-
cal responses by directly acting on receptors of the thymus and blood cells.
Finally, glucocorticosteroids are required for the action of catecholamines
such as for the induction of vasoconstriction by norepinephrine (e.g., Beato
and Doenecke, 1980; Munck et af., 1984).
Long-term increased glucocorticosteroid levels selectively reduce gluco-
corticosteroid receptors in the hippocampus (Brooke et af., 1994). Further-
more, high levels of glucocorticosteroids, such as are found in individuals
suffering from chronic stress, are known to cause severe dendritic atrophy.
This atrophy is particularly notable in hippocampal neurons and may con-
tribute to the cognitive impairment found in persistently challenged individ-
uals (e.g., Aus der Muhlen and Ockenfels, 1969; Magarinos et af., 1996;
McEwen et al., 199.5; Uno et af., 1994; Sapolsky, 1991, 1992) (Table I).
The synthesis and release of glucocorticosteroids are controlled by the
pituitary hormone ACTH (adrenocorticotrophic hormone), which itself is
controlled by the hypothalamic corticotrophin-releasing hormone (CRH).
In emotionally induced stress reactions the release of corticosteroids ap-
pears to be largely controlled by CRH, whereas physical pressures can
result in an increase in ACTH and, therefore, also in an increase in glucocor-
TABLE I
ACUTEA N D POTENTIAL EFFECTS
LONG-TERM OF GLUCOCORTICOSTEROIDS
Elevated levels of glucocorticosteroid hormones

Acute effects Chronic effects
Mobilization of energy (Gluconeogenesis) Loss of muscle mass, fatigue, steroid

diabetes
Lipolysis (synergistic with catecholamines) Arteriosclerosis
Raised muscle contractibility (permissive to Hypertension
catecholamines)
Sodium retention and diuresis Hypertension
Release of calcium from bones Osteoporosis
Elevated release of hydrochloric acid and pepsinogen Ulcerat ion
in stomach
Antiinflammatory and immunosuppressive actions Decreased wound healing, increased
disease susceptibility
Suppression of gonadal activity Decreased sexual behavior, sterility
? Dendritic atrophy (especially of
hippocampal neurons)
Neural responses, including altered cognition and Psychoses and depression
sensory threshold
ticosteroids through other mechanisms (such as through direct action of

interleukin 1 during infections or through vasopressin during disturbances
of the electrolyte balance: e.g., Aguilera et al., 1992; Berkenbosch et al.,
1992; Brown, 1991; Dallman, 1991; Dempsher and Gann, 1983; Lilly et al.,
1983; Rivier, 1991; Smelik and Vermes, 1980) (Fig. 7).
While the release of glucocorticosteroids is usually controlled by ACTH,
an additional possibility for the modification of the adrenocortical activity
that has so far largely been ignored, is by its innervation. Henry and associ-
ates (1976) discussed the morphological and physiological evidence and
presented their own data indicating that the activation of the adrenal cortex
in dominant and aggressive fighting mice is due to direct sympathetic ner-
vous stimulation, while in subordinate and repeatedly defeated mice the
normal hormonal ACTH pathway is involved (see also Hucklebridge et
al., 1981).
Apart from its effect on the adrenal cortex, ACTH acts directly on the
central nervous system. This indicates that ACTH may play an important
role in the establishment and maintenance of social hierarchies, as was
pointed out by Brain and Poole in 1974: Injection of ACTH suppresses
defensive fighting behavior in mice pitted against trained fighters. In addi-
tion, acquisition of both actively and passively conditioned avoidance re-
sponses is enhanced by ACTH and the disappearance of these responses
-u=- Stressor
Interleukin 1 I
Epinephrine ]
Vasopressin I
I Adrenal cortex
I
I Effector cells I
FIG. 7. Schematic diagram of the hypothalamo-pituitary-adrenocorticalaxis. Stimulating
(+) and inhibiting (-) influences are indicated.
is delayed. On the basis of these results, Brain and Poole proposed that
subordination in a dominance hierarchy may be a form of conditioned
avoidance response, which causes subordinates to avoid further attacks by
dominants either by fleeing or by signaling subordination.
Corticosterone treatment has no apparent effect on offensive, aggressive
behavior, but increases submissiveness in mice. Evidence for this is found
in the occurrence of “the rigid upright posture” and the failure to defend
themselves when attacked by an opponent (Leshner, 1981; Leshner and
Politch, 1979; Leshner et al., 1980; Politch and Leshner, 1977). The authors
conclude that, whereas ACTH may be important in the regulation of aggres-

sion, corticosterone regulates submission.
As already mentioned, the influence of stressors induces an increased
production and release of glucocorticosteroids. Long-term stress can there-
fore lead to ACTH-induced hypertrophy and hyperplasia in cells, resulting
in a substantial enlarging of the zona fasciculata and hence of the entire
adrenal gland.
Minerafocorticosteroids. The second group of adrenocortical hormones
are produced in the zona glomerulosa and are called mineralocorticostero-
ids after their function. The only physiologically relevant hormone is aldo-
sterone, which affects sodium reabsorption in the distal tubuli of the kidneys
and is hence involved in water and electrolyte metabolism. Its secretion is
regulated by several factors (mainly by the concentrations of potassium
and/or angiotensin I1 in the serum). Although a participation of aldosterone
in stress responses (“conditioned anxiety”) has been demonstrated in rhesus
monkeys, the direction of the initial change varies between the animals
(Mason et al., 1968b). Because such studies on aldosterone and stress are
few and far between, it will not be considered here.
Sex steroids. The third group of adrenocortical hormones are sex steroids,
particularly androgens such as dehydroepiandrosterone and androstenedi-
one, which are normally released in considerable amounts by ACTH. How-
ever, in certain states (e.g., puberty, aging, and stress) there is a divergence
between the stimulation of cortisol release on one hand and adrenal andro-
gens on the other, which indicates the additional release of adrenal andro-
gens by other, probably pituitary, factors (for details, e.g., Labhart, 1986).
Compared to the biologically relevant testicular androgen testosterone, the
biological effectiveness of adrenal androgens is very weak and little is
known about their physiological role under normal conditions. However,
the possibility cannot be ruled out that female reproduction (inclusive of
fetal development) may be impaired by increased androgen concentrations
in stressful situations (for a recent review, see Collaer and Hines, 1995).
In contrast to the mineralocorticoids, glucocorticosteroids and sex hor-
mones in the blood are mainly bound to transport proteins (corticoid-
binding protein (CBP) and albumin) and free and bound fractions are at
equilibrium. Only the free fractions exhibit biological activity. Concentra-
tion of these transport proteins is variable (e.g., increase during pregnancy:
decrease during starvation).
Assessment of adrenocortical activity
In the simplest case, changes in
A D R E N A L G L A N D WEIGHT A N D HISTOLOGY.
the size and weight of the adrenals can be useful to infer changes in activity.
Adrenal weight and histology were the first parameters used to asess the
extent of adrenocortical activity by Selye and other scientists up to the end

of the 1950s. Its usefulness is restricted, as the animals have to be killed
to gain access to the organs and it is not possible to follow responses of
the adrenocortical system in animals on the individual level. Even so,
weighing is still often the only way of assessing adrenal activity in smaller
animals in the wild, particularly as the weight of the adrenals is not affected
by the capture and killing of the animal.
Adrenal weights do not, however, provide information either on current hor-
mone levels or on short-term changes in adrenal activity, as changes in size
require several days of continual stress. Therefore, weights are more indicative
of the adaptive state of the adrenal cortex. In addition, they provide little more
than semiquantitative indications of current hormone concentrations.
In many cases of less recent laboratory and field research on small rodents
no absolute adrenal weights were given, and only results of calculations
relative to the body weights of the animals were supplied. These relative
adrenal weights are aimed at highlighting developmental differences be-
tween individuals and should compensate for differences in the size of their
organs. In my opinion, these values are not, however, satisfactory measures
of the adrenal activity in individuals, as body weights are particularly prone
to rapid change if animals are stressed. Although relative adrenal weights
do not allow any conclusions as to the adrenocortical activities or serum
glucocorticosteroid concentrations in animals, increased relative adrenal
weights do indicate stress.
Chemical, histological, or histochemical studies of adrenal glands, as used
to determine adrenocortical function in the initial research into stress, have
since lost all relevance because of the development of direct methods in
the determination of hormone concentrations.
HORMONE MEASUREMENTS. The direct measurement of glucocorticosteroid
concentrations in blood samples (serum or plasma) by radioimmunoassays
and other methods is quite simple. Since, however, glucocorticosteroid
concentrations increase after only 3 min subsequent to the beginning of
the blood sampling procedure, “true” baseline levels can usually be ob-
tained only under laboratory conditions.
In the controlled laboratory environment, this methodology is applicable
to assessing the effects of social and other stimuli on adrenal activity in
individuals, as the necessary blood sample size is so small ( 4 0 ~ 1 that, )
even in small animals with body weights far below 100 g, blood sampling
at 1- to 2-day intervals over long time spans is possible without detrimental
effects due to blood loss. As mentioned previously, though, depending on
the species and its emotional reaction and resulting psychological processes,
an insufficient time lapse between each blood sampling procedure may
result in typical stress responses with heightened glucocorticosteroid levels
in the serum. For example, regular blood sample collection at 7-day intervals
over several months induces no quantifiable physiological changes in tree
shrews, whereas sampling at 4-day intervals o r less induces clear stress
reactions after only two to three blood sample collections.
One largely neglected aspect that may be particularly relevant to stress
research is the relationship between free and protein-bound hormone levels.
In the majority of studies only the total amount of the hormones (bound
and unbound) is determined. As mentioned earlier, the biologically active
fraction of the glucocorticosteroids are the free hormones: They affect
tissues and regulate the release of glucocorticosteroids from the adrenal
glands by their negative feedback effects on hypothalamic and hypophyseal
structures. The concentration in the blood of these proteins, that bind to
and transport the hormones, is usually restricted and can be saturated by
increased hormone concentrations. Dependent on the concentration of
transport proteins in the blood, this means that a small increase in total
hormone concentration can lead to a substantial increase in the concentra-
tion of biologically active free hormones, as shown in laboratory mice
(Bronson and ElefthCriou, 1965a). However, there appear to be substantial
interspecific differences: Serum concentrations of both cortisol and cortico-
sterone in tree shrews in acute stressful situations can rise by factors of
4-5 within 30 min, without affecting the ratio of free to bound hormones
(correlation between initial values and stress values of free and protein-
bound glucocorticosteroids is always >.90).
On the other hand, a decrease in concentration of transport proteins, as
the consequence of a glucocorticosteroid-induced general protein mobiliza-
tion, body weight loss during stress, and/or as the consequence of increased
testosterone levels, can increase the free hormone levels, although the
total concentration of hormones remains the same or even decreases (e.g.,
Blanchard et al., 1993; Bradley et af., 1980).
In mammals, free (non-protein-bound) hormones and their metabolites
are largely excreted with the urine. The determination of the excretion
rates of glucocorticosteroids (as well as those of other hormones) should
therefore be especially appropriate in making statements on hormonal
changes in mammals in stressful situations. The main limitation associated
with the measurement of hormone levels in the urine is the considerable
time lag between the appearance of the hormones in the blood and their
excretion with the urine. Furthermore, the concentration of hormones in
the urine varies according to the amount of urine produced, and both urine
production and the drinking behavior of animals are influenced by stressors.
Already in 1859, Claude Bernard reported the occurrence of oliguria in
association with pain or emotional reactions, and the antidiuretic effect of
emotional stimuli has repeatedly been confirmed by numerous subsequent
studies on mammals including human beings. There are, however, also

reports of diuretic responses to psychological stimuli (Mason et al., 1968b).
It is not yet known what situations or specific stress responses are associated
with these divergent effects on urine excretion. Conclusions on the activity
of the adrenocortical (or other) endocrine systems, drawn from the concen-
trations of specific hormones or their metabolites in individual urine samples
(e.g., collected at the beginning of the activity period of an animal), are
therefore not particularly reliable. This problem is aggravated by the fact
that reliable internal standards correcting for changes in urine concentration
and/or loss of urine are not available. The creatinine concentration in urine
samples is often used as an internal standard, but is subject to substantial
change under stress conditions, as are all other urinary parameters, thus
making it unsuitable as an internal standard. This means that statements
on the activity of endocrine systems are usually possible only if the urine
is collected quantitatively, throughout the entire day, in either one or several
samples. With the exception of humans, this is possible only in laboratory
conditions and requires the isolation of the animals. Very often the animals
have to be kept in specialized urine collection or metabolic cages (or, in
psychological research on monkeys, often in restraint chairs), which results
in extreme restriction of freedom of movement and hence usually in stress
reactions or the requirement of long habituation periods of the animals to
the situation. This usually precludes any research into the effects of social
interactions or factors on endocrine processes. In many species that can be
kept in cages on lattices, the influence of specific social stimuli (e.g., sight
or smell of rivals or sex partners or separation of mother and its infant)
can be investigated by collection of the urine in a basin beneath the cage.
As steroid hormones (such as the glucocorticosteroids and sex hormones)
are not destroyed by delayed collection or by drying out, the urine can be
collected at fixed time intervals (e.g., 24 hours), dissolved in distilled water
and the hormone concentrations can be determined (Fenske, 1989). In this
manner acute changes in hormone secretion due to stressors, as well as
chronic effects of social input, can be assessed (e.g., Fig. 8).
In field studies, some groups have also used concentrations of steroid
hormones in samples of feces as indications of adrenocortical activity (e.g.,
Miller et al., 1991). Although the quantitative collection of feces of individu-
als is sometimes possible, even under natural conditions usually only single
samples are collected (e.g., morning feces). The interpretation of these
fecal hormone concentrations is subject to the same methodological con-
straints as those of single urinary samples, even when sampled at predeter-
mined times of the day.
For the past few years salivary steroid hormone levels have been used
to analyze the stressful effects of different social situations. The hormone
c
0
2
m
c
.-c
e3
L
15
10
5
-
Daily confrontations
and visual contact
Male 475
dominant
P
0)
s o
.-
c
-6 -4 -2 C C C 2 4 6
Days before resp. after confrontation
FIG. 8. Daily urinary cortisol excretion of two male tree shrews that lived in a cage separated
by a nontransparent partition. After habituation to the new cage the partition between the
animals was removed on 3 days daily for 10 min, which resulted in slight fights and the
establishment of a dominance order (C). For the rest of the days both animals were separated
by a wire mesh partition to allow visual contact between the rivals. As evident from the figure,
cortisol excretion increased in the subordinate male during the period of visual contact with
its rival and returned to initial values after separation by a nontransparent partition, while
the opposite was found in the dominant individual. Horizontal dashed line: Mean cortisol
excretion during the 6 days before the confrontation.
concentrations in the saliva correspond in most species to those in the

blood and are independent of saliva production. This method has many
advantages compared to blood sampling procedures, as it is noninvasive,
very fast (about 1 ml of saliva is needed), and it measures only the biologi-
cally relevant free (non-protein-bound) fraction of the hormones (Riad-
Fahmy et af., 1982; Vining and McGinley, 1986; Wade, 1991). For these
reasons, this method is widely used in psychological studies on humans and
in some animal welfare studies on larger mammals, such as dogs or pigs
(e.g., Beerda et al., 1996; De Jonge et al., 1996; van Eck et al., 1996; Ekkel
et al., 1996; Kirschbaum and Hellhammer, 1989); recently Fenske (1996)
showed that this method is also applicable to small mammals. In studies
with guinea pigs he found a good correlation between saliva and plasma
concentrations of cortisol, but, in contrast to studies in humans, not of
testosterone. The reason for this discrepancy cannot be explained so far.
CHALLENGE TESTS. Due to the time required to catch and handle the animals
for blood sampling, in many cases it is impossible, even in the laboratory,
to obtain “real” baseline hormonal levels; therefore, some authors have

used challenge tests to gain information on the state of the adrenocortical
system. The most common challenge used in studies on small rodents in
the laboratory is the “open field test.” Animals transferred to an open field
respond to this situation with acute stress reactions, for example, an increase
of glucocorticosteroid concentrations in their serum, which can be deter-
mined by taking a blood sample after a predetermined time (e.g., 15 rnin).
As many studied have shown, animals living in a stressful situation show
higher glucocorticosteroid levels in subsequent challenge tests than do
control animals.
The transfer of laboratory animals to an open field is not always a
sufficiently strong stressor to elicit maximal release of glucocorticosteroids.
Thus, following maximal stimulation of their release by high amounts of
ACTH, serum cortisol values in male guinea pigs are about three times
higher after 240 min than they are 240 rnin after the transfer to an open field
(Sachser, 1994a). Nevertheless, there is a very good correlation between the
cortisol values of the individuals in the open field and those in the ACTH
test. This indicates that challenge tests, such as transfer to a novel room
or cage, give a suitable measurement of an animal’s adrenocortical activity
or secretory capacity.
In the field, the procedure of catching the animals or of the anesthesia
necessary for blood sampling in larger species has been successfully used
as a standard challenge to determine the adrenocortical capacity of individu-
als. A particularly simple challenge for the determination of the secretory
capacity of the adrenal cortex in tree shrews in the laboratory, which so
far has not been used by other researchers, is the blood sampling procedure
itself (von Holst, 1986b). To this aim the animals were brought in their
sleeping boxes to a laboratory and blood samples were taken 1, 5, 15, and
30 min after the room had first been entered. Between sampling the animals
were returned to their sleeping boxes, but remained in the laboratory for
the entire period.
This challenge test is a strong stressor for all tree shrews: Their heart
rates are elevated for the entire experimental period and the levels of
catecholamines, glucocorticosteroids, and glucose in the blood increase
greatly. Shortly after the test and transfer to the home cage, heart rate and
all other parameters return to the initial levels. On average, the cortisol
values of control animals increase within 30 min from less than 10 ng/ml
serum to approximately 60 ng/ml serum, with substantial differences being
observed between the individuals (for examples, see Fig. 9). As long as the
animals live under constant conditions, repetitions of these challenge tests
after periods ranging from 1 week to several months result in almost identi-
cal (and individually different) values for all animals ( r > .92; p < .001;
150
120
.-C
30
Male 21
d
0 ' ,
0 10 20 30
Minutes after first disturbance
FIG. 9. Blood sampling challenge tests (BSCT): Adrenocortical responses of 3 males to 3

challenge tests separated by 3-5 months.
based on data from several experiments with more than 150 animals; see also
Fig. 9). The blood sampling procedure elicits a maximal glucocorticosteroid
release in tree shrews within the first 15 min, which cannot be further
increased by injection of higher doses of ACTH. Accordingly, in vitro
superfusion analyses of the adrenals from controls and stressed animals
show an extremely good correlation between the in vitro cortisol production
of the adrenals after maximal stimulation of their secretion through addition
of ACTH to the superfusion medium, and the serum cortisol values ob-
tained from the individuals earlier in a challenge test (Fig. 10). The individ-
ual differences of cortisol challenge values are therefore due to correspond-
ing differences in the adrenal capacities of the individuals to synthesize and
release glucocorticosteroids after stimulation. Chronic stress (e.g., transfer
to a new room or a confrontation with a dominant rival) always leads to
an increase in these challenge values by up to 200%. O n the other hand,
the opposite is found when males habituate to a new room or become
dominant in a confrontation (see also Fig. 22).
It must be emphasized here, that an alteration of a challenge value must
not be taken as an indication of an equivalent alteration of serum baseline
levels of the glucocorticosteroid hormones. This is due to the fact that the
secretory activity of the adrenal glands is dependent on the nominal value
120 1
Correlation coefficient
r = .91; p<.oo1
v = 12.13 + 6 . 3 7 ~
20 7 , , , , , , , , , , I
0 3 6 9 12 15 18
In vitro cortisol secretion of adrenals (nglmin)
FIG. 10. Correlation between the serum cortisol values of 12 males 15 min after beginning
of a BSCT and the in vitro cortisol release of their adrenals 15 min after ACTH has been added
to the superfusion medium. The in v i m tests were performed 8 days after the challenge tests.
of the hormone levels as well as on the metabolism and clearance of the

hormones, all of which are factors that may be influenced by stressors.
Therefore, increased secretory capacities of the adrenal glands in animals
are found especially during active stress, without simultaneously increased
baseline adrenocortical hormone levels.
In summary, while an increased adrenocortical capacity (as measured
with a challenge test) does not necessarily have to be associated with
increased levels of glucocorticosteroids in the blood, it is a sensitive measure
of stress, which can also be used in field studies.
b. Sympathetico-Adrenomedullary System. In contrast to the cortex, the
adrenal medulla arises embryologically from neural tissue, and remains a
functional part of the autonomic (vegetative) nervous system. It may in
fact be considered as a specialized sympathetic ganglion that, on activation
of the sympathetic nervous system, discharges epinephrine and norepineph-
rine directly into the blood (Fig. 11). Epinephrine in the periphery is derived
primarily or wholly from the adrenal medulla; norepinephrine, however,
may be secreted from the adrenal medulla, or its presence may be due
to the overflow of neurotransmitters from the noradrenergic sympathetic
nerves into the circulation. In rats and cats, for example, approximately
60-70% of the peripheral serum norepinephrine is derived from sympa-
thetic nerves (Kvetnansky et af., 1979; Stoddard, 1991). Both the activation
of the sympathetic nervous system, as well as the release of adrenomedullary
Autonomic nervous system
Sympathetic system Parasympathetic syster
nal
rd
Cervical
Hair follicle muscle

Blood vessel
Thoracic
Lumbar
Sacral
FIG. 11. Schematic diagram of the autonomic nervous system with the effector organs of
its two subdivisions-the sympathetic and the parasympathetic nervous system.
hormones, therefore act upon different bodily functions in very similar

ways and generally increase the capabilities of the organism (Cannon’s
fight or flight response): Hence the sympathetic nervous system and
adrenal medulla are usually summarily referred to as the sympathetico-
adrenomedullary system. Apart from its catabolic effects on peripheral
organs, peripherally released epinephrine (but not norepinephrine) also
affects the central nervous system by eliciting general arousal and subjective
feelings ranging from restlessness to anxiety. Furthermore, there is also
evidence that the release of epinephrine in situations eliciting fear enhances
the learning of avoidance behavior, while high doses apparently produce

the reverse effects (McGaugh and Gold, 1989).
The biosynthesis of the catecholamines epinephrine (E) and norepineph-
rine (NE) from tyrosine progresses in several steps in both the noradrenergic
neurons of the sympathetic nervous system and in the adrenal medulla;
the first step-the conversion of tyrosine to dopa-is catalyzed by the
enzyme tyrosine hydroxylase (TH), which is the pacemaking step of the
catecholamine biosynthesis. For the subsequent two steps-the formation
of dopamine and norepinephrine-enzyme levels are not usually limited.
The final conversion of norepinephrine to epinephrine is catalyzed by the
enzyme phenylethanolamine-N-methyltransferase(PNMT). PNMT activi-
ties in sympathetic neurons are very low, thus norepinephrine is the main
end product of catecholamine biosynthesis (the transmitter), while in adre-
nal medullary tissue PNMT activities are high, resulting in epinephrine as
the main adrenomedullary hormone.
The TH levels in the medulla are controlled mainly by neuronal influences
from the sympathetic nervous system (Thoenen et al., 1969; Ungar and
Phillips, 1983). Repeated stimulation of catecholamine release in stressful
situations leads to an adaptive increase of the TH levels in the adrenal
medullary tissue, which in turn increases the capacity of the adrenal gland
to synthesize and release its hormones. In laboratory rats, for example,
there is a doubling of adrenal T H activities within less than one week of
daily immobilization stress (Fukuhara et al., 1992; Kopin, 1980; Kopin et
al., 1988). PNMT levels, on the other hand, are usually not influenced by
sympathetic (splanchnic) nerves, but by the glucocorticosteroid hormones
of the adrenal cortex, leading to a 50% increase under persistent stress
(Ciaranello, 1978; Kopin, 1980; Wurtman and Axelrod, 1966). There are,
however, differences in the regulation of PNMT activities even between
different strains of rats, as has been demonstrated recently by Lemaire
and colleagues (1993). According to these authors, the increase of PNMT
activity in male Wistar rats, kept in unstable mixed-sex groups, is dependent
on the activation of the sympathetic nervous system, as it can be completely
abolished by denervation of the adrenals.
Assessment of sympathetico-adrenomedullary activity. Each capture and
handling of an animal for the collection of blood samples immediately
activates the sympathetico-adrenomedullary system and triggers the release
of catecholamines into the blood, thus making the collection of baseline
values impossible without prior introduction of cannulas into the blood
vessels (Stoddard, 1991). Therefore, with one exception (Fokkema, 1985),
I am not aware of investigations dealing with catecholamine baseline values
in animals in social situations.
32 DIETRICH VON nouT
CHALLENGE TESTS. It appears that challenge tests are useful techniques to

gain information on the adaptive state of the sympathetico-adrenomedul-
lary system. As already mentioned, the adrenal medulla adapts to an in-
creased production and secretion of its hormones by increasing the activity
of its pacemaking enzyme, tyrosine hydroxylase. Accordingly, the stimu-
lated adrenal gland increases its secretion of the two hormones epinephrine
and norepinephrine.
In our research on tree shrews, we use their emotional response to
handling during blood sampling as the challenge. We found that catechola-
mine levels in the blood of tree shrews are maximally raised as rapidly as
20 s after the first disturbance. When we take several blood samples from
an individual within a period of 1-15 min, serum catecholamine levels
remain more or less the same, while differences of several hundred percent
are found between different individuals. The mean value of the norepineph-
rine and epinephrine concentrations in several blood samples from a given
individual taken during a challenge test is therefore considered to be an
indication of the adaptation state of its sympathetico-adrenomedullary sys-
tem. This suggestion is supported by the following results: Between repeated
challenge tests separated by at least 1 week, the individually different values
of norepinephrine (as well as of epinephrine) are highly correlated (Fig.
12). Stressful situations, such as transfer to a new cage or a confrontation
with a rival, increase the norepinephrine and epinephrine values by up to
12
0
55 10
z
0
-
.
E
F a
f
.-C
a,
r = .85 r = .91
2
2 4 6 8 1 0 1 2 2 3 4 5 6 7
NE (nglml serum) TH (nmollh adrenals)
FIG. 12. Correlations between BSCT values of serum norepinephrine (NE) of male tree
shrews at two sampling points separated by 8 days (left) as well as between serum norepineph-
rine values and adrenal tyrosine hydroxylase activities (TH) of animals. Unpublished data:
after Hutzelmeyer (1987).
100% within 1-3 days. After habituation to the new situation (room or
cage) or removal of the rival, challenge values approximate the initial
individual levels (see also Fig. 5). Finally, a very good correlation is found
between the norepinephrine and epinephrine challenge values of tree
shrews and their adrenal tyrosine hydroxylase activities (Fig. 12).
Thus, challenge levels of catecholamines are a relevant measure of the
activity or adaptive state of the sympathetico-adrenomedullary system,
which can be used in the laboratory as well as under more natural conditions
in large enclosures, as has been shown in studies on tree shrews, guinea
pigs, and wild rabbits. In guinea pigs, for example, catecholamine challenge
values are clearly related to individual behavioral strategies in situations
of social conflict: Plasma norepinephrine levels were significantly higher
weeks before as well as directly after a 10-min agonistic encounter in
offensive “fighters” compared to nonaggressive “nonfighters” (Sachser,
1987). Thus, norepinephrine responsiveness is of predictive value for the
behavior of guinea pigs in contest situations. Similar results are also known
from tree shrews (e.g., Fig. 16).
Excretion rates of catecholamines and their metabolites are frequently
used in psychosocial stress studies in humans, and less frequently in psycho-
physiological studies in monkeys, but not in behaviorally oriented studies
on other animals. This has many reasons, some of which have been already
mentioned (see discussion on glucocorticosteroid excretion). Even more
important is, however, the fact that catecholamines and their metabolites
are rapidly metabolized and/or destroyed after urination. To avoid degrada-
tion, the urine must therefore be strongly acidified, which makes its collec-
tion more or less impossible, except from animals kept in a restraint chair.
ENZYME ACTIVITIES. As mentioned above, TH and PNMT levels in the adre-
nal medulla adapt to stimulation with an increase in their activities. The
determination of these two enzymes as a measure of the activity of the
sympathetico-adrenomedullary system was introduced by Henry and col-
leagues (1972; Kvetnansky et al., 1970) in their work on social stress in
mice. In the meantime, this method is widely established and is being
applied to many different animal species by other research groups. As the
increased change in enzyme activity appears only several hours after the
application of stress, these methods are equally applicable in the field and
the laboratory. However, since the animals have to be sacrificed, the range
of application of this method is restricted.
HEART RATE AND BLOOD PRESSURE. As the sympathetico-adrenomedullary
system induces strong physiological reactions in the organism, indirect mea-
sures are sometimes used as indicators of sympathetico-adrenomedullary
activity. A particularly important measure in psychobiological research is
the heart rate and-mainly for medical purposes-blood pressure. Both
parameters usually increase rapidly with every activation of the sympathetic

nervous system. While the heart rate is activated by the sympathetic nervous
system (including the catecholamines of the adrenal medulla) and inhibited
by the parasympathetic nervous system, and correspondingly changes in
both parts of the autonomic nervous system can influence heart rate,
blood pressure is activated only in the short term by the sympathetico-
adrenomedullary system and the renin-angiotensin system. Chronic emo-
tional responses, however, can result in structural (arteriosclerotic) changes
of the cardiovascular system, which may lead to persistent hypertension
(Folkow et al., 1958, 1973).
The heart rate is particularly well suited for the detection of acute activa-
tions of the sympathetico-adrenomedullary system in socially or otherwise
stressful situations, and for the monitoring of chronic stressors and their
potential pathophysiological consequences. Implantable radio transmitters
have been used for many years to record heart rate telemetrically. Since
the working life and range of transmitter systems depend on battery size,
the weight of the transmitters usually determines their working life. An
extremely small radio transmitter (weight including battery <1 g) with a
range of above 30 m and a working life of 4-6 months has been developed
at our institute (Stohr, 1988). This transmitter enabled us to record heart
rate (and body temperature) continuously in animals differing greatly in
size (Mongolian gerbils, laboratory mice, tree shrews, and wild rabbits)
(Eisermann, 1992; Eisermann and Stohr, 1992).
The weights of most commercially available transmitter systems used in
stress research in nonhuman mammals are in the range of 5-10 g, with a
lifetime of 1-2 weeks and a range of a few meters. This usually restricts
their application to laboratory conditions and animals with body weights
of over 200 g. The working life is also usually too short to allow any more
than the acute effects of specific stressors to be recorded. This problem is
exaggerated by the fact that, depending on the species and the size of the
transmitters, a period of 1-2 weeks is necessary after implantation before
the animals have recovered and stable, low levels of their heart rate can
be recorded. In technically advanced systems, a magnetic on/off switch is
used to transmit the signals only at certain times, which conserves power
and may extend the working life up to several months.
Blood pressure has mainly been measured in animals in chronic emotion-
ally arousing situations, by direct arterial or indirect tail-cuff techniques at
weekly time intervals. For both techniques, the animals must be handled
and restrained before measurements can be taken. To reduce emotional
stress responses, the measurements are often performed after slight narco-
sis. Nevertheless, these procedures elicit stress responses in the animals,
which may influence the measurement. In laboratory animals, however,
the tail-cuff technique in particular has been proven to provide reliable

results, when the animals are used to this procedure (Bunag, 1984). Over
the past few years, the direct and continuous measurement of blood pressure
has also become possible in small mammals using implanted radio transmit-
ters; in one of the first studies using this method, Henry and co-workers
showed parallel blood pressure changes in rats during chronic stress, mea-
sured by radiotelemetry and by indirect tail-cuff techniques. The latter,
however, revealed levels 10% higher than the former (Henry et al., 1995).
c. Pituitary-Gonadal System. Apart from the generative testicular and
ovarian functions (the development of spermatozoa and ova, respectively),
these organs also have endocrine functions. They produce sex hormones,
the production of which is independent of spermatogenesis in males, but
closely linked to oogenesis in females.
Testes. The testes secrete hormones, which are collectively termed andro-
gens, of which testosterone is the main physiologically active one. Depend-
ing on the species and organ, testosterone sometimes has to be converted
enzymatically to dihydrotestosterone (17-DHT) in target cells in order to
become biologically effective.
Before birth, the testes of mammals release large amounts of androgens,
which are responsible for the development of male-specific sex organs
during fetal life, and for the organization of the central nervous structures
responsible for the expression of male-specific behavior in later life (e.g.,
Breedlove, 1994; Collaer and Hines, 1995; Goy et al., 1988; Gustafson and
Donahoe, 1994; Huhtaniemi, 1994; McCarthy, 1994; Phoenix et al., 1959;
Turner, 1984). In the absence of the male sex hormones, female-specific
sex organs and behavior patterns develop. These pre- or perinatal effects
of androgens are usually characterized as “organizing effects” (Phoenix et
al., 1959).
Following puberty, testosterone (and DHT) are necessary for the differ-
entiation and activation of spermatozoa during spermatogenesis, induce
growth and function of accessory sexual glands, and modulate sexual, ag-
gressive, and other testosterone-dependent behaviors. This occurs to a
greater or lesser extent depending on the species (e.g., Arnold and Breed-
love, 1985; Baum, 1992; Beach, 1975; Matochik and Barfield, 1991; Mo-
naghan and Glickman, 1992; Thiessen and Rice, 1976). These peri- and
postpubertal effects of androgens on morphology, physiology, and behavior
are usually called “activating effects.” Furthermore, marking behavior and
the secretory activity of glands, involved in marking or the characterization
of the hormonal state of a male, are modulated by testosterone. Thus,
decreased testosterone levels in a subordinate male can reduce the intensity
of its aggression-eliciting signals and thus reduce attacks against it.
Production and release of androgens by the testes is controlled by the

hypothalamus. Gonadotropin-releasing hormones (GnRH) induce the re-
lease of follicle-stimulating hormone (FSH), which stimulates spermatogen-
esis, and luteinizing hormone (LH), which induces synthesis and the release
of androgens from the Leydig cells of the testes.
Ovaries. In contrast to the male androgens, female sex hormones have
no organizational functions in mammals. Following puberty, dependent on
the development of the follicles and subject to the influence of the two
gonadotropins LH and FSH, the ovaries produce two main classes of sex
hormones-estrogens and progestins, which are responsible in female mam-
mals for their sex-specific morphology and physiology (e.g., Carter, 1992;
Takahashi, 1990).
Both classes of female sex hormones are also produced placentally in
gravid females, as well as by the adrenal cortex in both sexes. Estrogens
(especially the most effective estradiol) stimulate growth of the uterine
wall and have a variety of other functions contributing to the maintenance
of the condition of the female reproductive system. The biologically relevant
progestin is progesterone, which is secreted in most mammalian species
after ovulation by the corpus luteum of the ovary, and is necessary for
maintaining pregnancy. In small rodents no corpus luteum develops during
the estrous cycle and progesterone is synthesized by the interstitial tissue
of the ovary. Estrogens (in many rodents in combination with progesterone)
are responsible for female sexual receptivity, although their effect differs
very much between species. In addition, estrogens can also enhance the
attractivity of females by inducing the production of odors (sex phero-
mones) or other signals.
Most stress situations apparently inhibit the release of GnRH, thus modi-
fying fertility and the sex-hormone-dependent behavior of both sexes (e.g.,
Kime et al., 1980; Moberg, 1987; Orr and Mann, 1992; Rabin et al., 1988).
Because of their specific receptors, the Leydig cells of the testes can also directly
be influenced by glucocorticosteroid hormones (e.g., Evain et al., 1976; Stalker
et af., 1989). Thus, 2 hr of immobilization induces a fall in serum androgen
concentrations of rats without detectable changes in serum LH values.
The chronic stress of a daily 2-hr immobilization for 10 days, however,
results in decreased serum levels of both testosterone and LH (Maric et al.,
1996). Correspondingly, in vitro corticosterone directly inhibits testosterone
production by purified Leydig cells of rats, although only at high concentra-
tions (Gao et al., 1996). In contrast to long-lasting stressful situations, acute
stressors sometimes cause a transient elevation of plasma testosterone con-
centrations, despite the suppressed LH levels that precede the subsequent
decline in testosterone levels. The reason for this effect has not been clari-
fied yet, although increased testicular blood flow as a general consequence
of heightened sympathetico-adrenomedullary activity or of testicular

nerves, as well as testicular pro-opiomelanocortin-derived peptides, may
be involved in this function (e.g., Mann and Orr, 1990).
Assessment of gonadal activity
MORPHOLOGICAL AND HISTOLOGICAL DATA. As mentioned before, the testes
have endocrine as well as generative functions that are not necessarily
closely correlated. In order to obtain information on the generative func-
tions, data collection in animal studies on social stress focuses on the weights
of the testes, epididymides, and secondary sex glands, and in particular on
histological investigations on the testes and epididymides for the proof of
normal spermatogenesis, although the presence of intact spermatozoa does
not rule out functional damage.
In females, the weights of the ovaries provide a rather more superficial
indication of their function; histological investigations on the ovaries, how-
ever, provide relatively differentiated information on maturing follicles,
the number of ova released into the uterus, or impeded follicular maturation
(e.g., increased follicular atresia). In addition, investigations on the uterus
can provide information on implantation and also on abortions. All of these
methods, however, are unsatisfactory in connection with studies on social
stress and its consequences for the gonadal system, as they necessitate the
sacrifice of the animals.
HORMONE MEASUREMENTS.The determination of sex hormones in the blood,
urine, or feces is possible in both sexes. The problems involved in the
collection of these data (methods of blood sampling, time factors, methods
of sample collection) are equivalent to those involved in the collection of
glucocorticosteroids. Relatively reliable data on the influence of social and
other factors on endocrine activity in the testes can be collected by blood
sampling in males. In females, this is usually not possible except in extreme
conditions, as the changes in hormone values are dependent on the stage
of the cycle or pregnancy and daily hormone determination would be
required in order to pinpoint stress-related changes, especially in animals
with estrous cycles of a few days. This would be possible by determining
selected hormones and/or their metabolites in the urine or feces. However,
due to the many difficulties involved, very little information is as yet avail-
able on the quantitative influences of social and other stressors on the
female endocrine system. Extreme changes, such as those during estrus
and pregnancy can, however, be assessed by determination of hormones
in urine or feces (e.g., Schaftenaar et al., 1992).
d. Immune System. The immune system has two functional divisions:
the innate (unspecific) system and the adaptive (specific) system. The innate
immune system acts as a first line of defense against infectious agents
(viruses, bacteria, fungi, and parasites) and clears them before they establish
an overt infection. The adaptive immune system produces a specific reaction
to each pathogen (antigen), which normally eradicates that particular agent.
The adaptive immune system also remembers the infectious agent with the
aid of B and T memory cells, causing an immunity of variable duration
against the pathogen by inducing a much enhanced specific response at the
next contact with this antigen. Both divisions of the immune system consist
of a large number of humoral (molecules) and cellular (leucocytes) factors,
which circulate with the bloodstream and are distributed throughout the
entire body (Table 11).
During infections, both systems are usually activated and combat the
infectious agents in an integrated way: Following clonal activation after
contact with an antigen, T lymphocytes produce several soluble molecules
(cytokines), which stimulate the phagocytes to destroy the infectious agents
more effectively, and also stimulate antibody production by the B lympho-
cytes. These antibodies then also help the phagocytes to recognize their
targets. Depending on the antigen, the various parts of the immune system
are differently involved in the response pattern.
Although the immune system displays a certain degree of autonomy,
the research of the last two decades has revealed multiple channels of
communication between the central nervous system and the immune sys-
tem. Emotionally stressful situations are particularly associated with altered
immune function, and in some instances with altered health status, although
these two processes have not been linked causally (e.g., Adams, 1994; Ader
and Cohen, 1985; Ader et al., 1991; Dunn, 1989; Glaser and Kiecolt-Glaser,
1994; Kelley et al., 1994; Laudenslager and Fleshner, 1994; Monjan, 1981;
Solomon and Amkraut, 1981). In his first publication, Selye (1936) described
thymolymphatic involution as one of the most conspicuous signs of stress,
TABLE I1
MAJORCOMPONENTS I M M U N ESYSTEM
OF THE INNATEA N D THE ADAPTIVE
Innate (unspecific) system Adaptive (specific) system
Humoral factors Lysozymes

Complement system
Acute phase proteins (e.g., C-
reactive protein) Antibodies (produced by B
Interferons lymphocytes)
Cellular factors Phagocytes (polymorphs and
monocytes) T lymphocytes (e.g.. cytotoxic cells,
Natural killer (NK) cells helper cells)
and the connections between the hypothalamo-pituitary adrenocortical axis

and the immune system are among the best examined so far (e.g., Dhabhar
et al., 1995; Heinjen et al., 1991; Keller et al., 1991; Munck and Guyre, 1991;
Zwilling, 1994). A strong immunosuppressive effect is also exerted by the
catecholamines, but virtually every hormone investigated so far has some
effect on the immune system (such as growth hormone, prolactin, and sex
hormones), although not all of the effects are direct (e.g., Bernton et al.,
1991; Grossman, 1984; Kelley, 1991; McCruden and Stimson, 1991; Olsen
and Kovacs, 1996; Paavonen, 1994; Rabin et al., 1994).
Furthermore, lymphoidal tissues (e.g., bone marrow, thymus, spleen, and
lymph nodes) are innervated by sympathetic and sciatic fibers and hence
the central nervous system can directly influence immune function (e.g.,
Ackerman et al., 1991; Felten and Felten, 1991; Madden and Livnat, 1991).
On the other hand, the immune system is capable of modulating both
neuroendocrine responses and the behavior of mammals by nervous influ-
ences from lymphoidal structures and by lymphokines produced by leuco-
cytes, which act on hypothalamic and other central nervous structures (e.g.,
Anisman et al., 1993; Bateman et al., 1989; Besedovsky and del Rey, 1991;
Blalock, 1988; Carr and Blalock, 1991; Hall et al., 1991; Madden and Felten,
1995; O’Grady and Hall, 1991; Sternberg, 1988). These responses may
improve the defense reaction of the body against pathogens.
Thus, increased glucocorticosteroid levels may help to suppress an overly
strong immune response, which could in itself be dangerous (as can be seen
in allergic reactions) (e.g., Munck et al., 1984). Behavioral changes, such
as lethargy, anorexia, or reduced grooming, which are typically observed
in sick animals, can also be elicited by cytokines produced during the
immune response of an organism against an infection (e.g., Crnic, 1991;
Kelley et al., 1994; Myers and Murtaugh, 1995).
Assessment of immunological functions. There are two fundamentally
different-but not mutually exclusive-approaches to gaining information
on the activity and capacity of the immune system in stressful situations:
RELATIONSHIP BETWEEN STRESS A N D DISEASES
Diseases in natural populations. The first indications of the influence of
psychosocial factors on the functions of the immune system were provided
by investigations that demonstrated a relationship between certain social
situations and the outbreak of diseases (e.g., at high population densities
in rodents or after the death of a partner in human beings). This epidemio-
logical approach is of particular relevance for the human situation, as it
provides the only means of assessing the relevance of social influences on
disease. However, this approach usually requires large amounts of data
and provides no information on the specific underlying immunological
processes associated with the increased morbidity (e.g., Gentry, 1984;

Weiner, 1977; Wenar, 1983).
Induction of diseases. In this case, pathogens (such as parasites, bacteria,
or tumor cells) are injected into animals housed under different stress
conditions and the outbreak and progress of the disease is correlated with
the intensity of the stress (in some cases also the rejection of skin transplants
is used for the characterization of immune capacity). In general today, those
immunological parameters are also assessed that are capable of providing
information on the action of the stress-induced changes in resistance to
disease. This approach is of undeniable importance to the explanation of
the relationship between stress and resistance to disease, but is methodolog-
ically complicated ( e g , knowlege of suitable pathogens, complicated hous-
ing conditions) and hence has so far been used only in a small number of
investigations on standard laboratory animals (especially rats and mice)
(for details see Section III,B,Z).
Direct assessment of immunological parameters. No general statements
can as yet be made on the effect of social and psychological influences on
the immune system, as it is composed of many different subunits that can
exhibit synchronous or antagonistic changes, depending on the situation
or subsystem. We also have very little current understanding of the rele-
vance of changes in specific immunological subsystems to disease in an
individual. Therefore, the monitoring of many different parameters is neces-
sary, in order to obtain a picture as informative as possible of the reaction
of the immune system. This is, however, not the case in most experimental
research carried out on animals. Immune measures are usually selected on
the basis of financial or methodological constraints (e.g., availability of
antibodies and laboratory facilities). Studies based on a limited selection
of immune parameters may, however, result in a failure to detect immuno-
logical changes. Premature conclusions that a situation has no immunologi-
cal effects must therefore be avoided.
In the rest of this section, the most common immunological parameters
used in animal behavior studies are mentioned. A reduction of one or
several of these parameters is usually interpreted as a measurement of a
reduced immune function.
HUMORAL FACTORS AND LEUCOCYTE NUMBERS IN THE BLOOD. Very little blood
is required for the determination of humoral factors (e.g., serum concentra-
tions of immunoglobulins or C-reactive protein) as well as of the numbers
and types of leucocytes in the blood (e.g., total leucocyte number, leucocyte
subcategories such as neutrophils, eosinophils, basophils monocytes, lym-
phocytes), and subsets of lymphocytes (e.g., T and B lymphocytes, cytotoxic
T lymphocytes, T helper cells, natural killer cells). All these parameters
can therefore be determined at regular intervals from blood samples of
individuals. For the differentiation of the lymphocyte subsets and the deter-
mination of the different immunoglobulins specific antibodies are necessary,
which are currently available only for human beings and a few experimental
animal species.
The activity of the complement system in serum samples can easily be
measured by in v i m bioassays (e.g., CH5[)hemolysis test), by determining
the volume of serum necessary to induce hemolysis in 50% of red blood
cells (usually of sheep) used as antigens. The measurement can be carried
out repeatedly even in small mammals, as this test requires very little serum.
IN VITRO TESTS. The in vitro proliferative response of lymphocytes after
stimulation with mitogens is widely used as a functional test of the activity
of the cellular part of the immune system. Different mitogens can be used,
such as concanavalin A (Con A), that stimulates in most species predomi-
nantly T lymphocytes, or pokeweed, that is more specific to B lymphocytes.
These in vitro tests can be performed on cell cultures of lymphocytes from
the blood or from the spleen, as well as on unmanipulated blood samples. In
addition to the proliferation rate, the production of the different cytokines
(interferons, interleukins) by the lymphocytes after stimulation with mito-
gens can be determined in the supernatant of the cell suspension.
Phagocytic activity of blood cell suspensions (mainly of monocytes [mac-
rophages] and neutrophil leucocytes) can be determined after contact with
an antigen (e.g., cymosan A) and the hereby induced release of reactive
oxygen molecules. All tests mentioned so far can be performed on the cells
of all animal species, so long as the laboratory facilities are available. Most
studies so far have been done on animals living under constant laboratory
conditions: the application of some of these methods under field conditions
is, however, also possible.
Natural killer (NK) cell activity is usually determined in v i m by bioassays
using the destruction of certain tumor cell lines by the NK cells; the currently
available tumor cell lines are, however, specific to only a few laboratory
animal species.
Although originally most functional tests in animals (especially mice and
rats) used lymphocyte cultures received from the spleen of the animals,
studies on human beings are naturally based on cells from the blood. The
relationship between the activities of these cells of different origin is not
clarified. It is, however, accepted that changes in the blood reflect changes
in other immunological organs (such as spleen or lymph nodes). Of course,
tests on blood cells have the great advantage that changes in immunological
functions under different conditions of stress can be monitored on an
individual level. These tests are therefore being used increasingly in animal
research. For most of these tests, however, large numbers of blood cells
are needed, which means that sacrifice is necessary for animals below the
size of a rat.
I N VIVO ANTIBODY PRODUCTION AFTER A N ANTIGEN CHALLENGE. In these tests
the capacity of an intact organism to produce specific antibodies against
antigens (e.g., sheep erythrocytes or keyhole limpet hemocyanin [KLH])
is determined. The specific antibody concentrations can be measured indi-
rectly (by bioassays such as in sheep erythrocytes) or directly by immunolog-
ical methods, when specific antibodies against the antibodies produced by
the individual are available. For antibody determination very little blood
is needed, and therefore these functional tests can be performed even in
small animals. Since antibody concentrations do not change quickly due to
handling processes, these tests can be used in laboratory and field condi-
tions. To determine the time course of antibody production, several blood
samples must usually be taken over a period of several weeks, which may
raise some problems in field studies.
111. SOCIAL IN MAMMALS

STRESS
A. INTRODUCTION
While the research of Selye and his contemporaries was mostly concerned
with the effects of physical stressors, after 1950 many ecologically oriented
studies were published, indicating that social factors participate in pituitary-
adrenocortical regulation. This work originated from the hypothesis, ad-
vanced by John J. Christian in 1950, that regulation of population densities
of small mammals might be achieved by mechanisms intrinsic to the popula-
tion itself: Increasing population density, according to his hypothesis, results
in qualitative and quantitative changes in the behavior of the animals, which
in turn stimulate pituitary-adrenocortical activity and decrease pituitary-
gonadal activity. As a consequence of these endocrine stress responses,
the mortality of the animals increases and their natality decreases, thus
counteracting the increase of population density. In his first paper (1950),
Christian suggested adrenocortical exhaustion and, as a direct consequence,
mortality as the major cause of cyclical fluctuations in the population num-
bers of small mammals. In subsequent years, however, it became evident
that increased susceptibility to infectious and parasitic diseases, brought
about by increased adrenocortical activity and decreased reproduction,
is much more important (e.g., Christian, 1963, 1971, 1975; Christian et
al., 1965).
1. Self-Regulation of Mammalian Populations
The growth of mammalian populations usually stops at a more or less
stable level below the environmental capacity. Exceptions are the popula-
tions of many small rodents of the arctic regions, which undergo marked
fluctuations with a periodicity of 3-4 years. The causes of these cycles have
interested ecologists since the beginning of this century (Elton, 1942). Some
authors proposed extrinsic factors such as weather, food, predation, and
diseases as the predominant or sole factors influencing animal numbers
in natural populations. Without doubt all these extrinsic factors act on
populations, either alone or in combination, and can cause a population
decline or even an extinction in particular instances, but it is unclear whether
they can limit population increases under natural conditions (Krebs and
Myers, 1974). Therefore, alternative hypotheses were proposed, claiming
population regulation by intrinsic factors.
According to a hypothesis proposed by Chitty (1958,1960), the behavior
of animals changes with density as a consequence of selection on genetically
different behavioral types. Aggressive individuals might be at a selective
advantage as the population increases. Aggressive behavior among individ-
uals could then be the direct cause of the population decline, even though
the exact reasons for the mortality during the decline are so far unknown.
Several studies have demonstrated genetic changes associated with density
changes (using mostly gel electrophoretic variants as markers), but the
question still remains as to whether these genetic changes are the cause or
the result of the demographic changes (Krebs, 1996; Krebs and Myers,
1974). The hypothesis that they are the cause has never gained wide accep-
tance.
In contrast, from the outset, Christian’s hypothesis of a self-regulation
of mammalian populations by social stress was a subject of wide debate
(e.g., Krebs, 1996; Krebs and Myers, 1974), resulting in intensive stress
research in the laboratory as well as in the field (mostly on voles, mice,
rats, and rabbits). The strength of this concept is based on the results of
experimental crowding of small mammal populations in the laboratory,
which demonstrated all the changes in reproduction, growth, and mortality
typically found in natural populations of high densities. The relationships
between behavior, stress, and density under natural conditions, however,
remain far from clear. This is for many different and often methodological
reasons. In my opinion, the most important of these is the lack of detailed
behavioral studies. This is not surprising in the case of the generally cryptic
small mammals. Most studies have collected data only on density-dependent
changes in agonistic behavior, and even then these data are often inferred
indirectly by counting skin wounds on animals living in different population
densities. Such an approach to the determination of relevant behavioral
changes may be misleading, however, as some studies have demonstrated
that no relationship exists between the number of wounds and population
densities in rodents (e.g., Batzli and Pitelka, 1971; Christian, 1971; Krebs,
1964). As a consequence, Christian (1963) introduced the term “social

density” to explain discrepancies between animal numbers per unit space
and stress responses. This term, however, is far from clear and may include
any combination of aggressive behavior, social interactions, space, and
number of individuals. In addition, population densities in the field have
usually been estimated on the basis of trappings, which are subject to a
great number of potential shortcomings, such as differences in their success
rate, depending on the individuals’ trap experience, their social status, or
their food situation (e.g., Krebs and Myers, 1974). Statements on population
densities and changes therein are therefore affected by gross and possibly
density-dependent errors. Furthermore, the indices used in the measure-
ment of stress in individuals have usually been provided by data on adrenal
weights or by other indirect measurements of adrenocortical activity, both
of which give only rough indications of the endocrine state of the animals.
Finally, data on the sympathetico-adrenomedullary activities are generally
lacking. Thus, many contradictory conclusions on relationships between
population densities and stress may have been based on inadequate mea-
surements or interpretations of the data.
For these reasons I will not go any further into the question of self-
regulation of densities of individuals in mammals. In addition, several re-
views on the causes of population cycles in small mammals have been
written over the past 30 years and the understanding of these causes has
not subsequently improved (e.g., Christian, 1978; Krebs, 1996; Myers et al.,
1971; Nowell, 1980; Snyder, 1968; Watson and Moss, 1970).
2. Behavioral Stress Research in Biomedicine

While initially ecological questions were central to research on stress,
and factors relevant to populations such as fertility and mortality were
investigated, since the 1960s, interest has shifted into the biomedical area.
Three developments were responsible for this shift.
1. The development of modern techniques in hormone analysis, allowing
the recording of endocrine stress reactions in humans and revealing
the general correspondence between many different animal species
including humans.
2. The proof that in animals, as well as in humans, psychological processes
are decisive in the triggering of stress reactions. Psychosocial stressors
are capable of long-term modification of both the sympathetico-
adrenomedullary and the adrenocortical systems.
3. The epidemiological evidence that social challenges (life events) and
individual traits in personality (e.g., type A personality, characterized
STRESS AND ITS RELEVANCE FOR ANIMAL BEHAVlOR 45
after Rosenman, 1986, by enhanced competitiveness, aggressiveness,

impatience, and a chronic sense of time urgency) in humans are in-
volved in the development or outbreak of specific diseases.
Experimental investigations on animals are now primarily directed at
recording physiological processes potentially involved in the development
of disease in humans. The relevance of heart and circulatory diseases to
our society has resulted in most research focusing on this area. In recent
years, psychoneuroimmunological research has increasingly analyzed the
importance of social stress reactions in the development of infectious dis-
eases and tumors. Up to the present, research has concentrated almost
entirely on laboratory animals kept in highly standardized conditions. It is
only in recent years that a few ecologically oriented field research projects
have reappeared, such as on dasyurid marsupials in Australia, rabbits in
Europe and elsewhere, and mongooses, cheetahs, and several monkey spe-
cies in Africa (see Section 111,BJ).
Although the relevance of social stress for the “regulation” of populations
is still controversial, nowadays it is generally accepted that social interac-
tions or situations may result in strong stress responses in mammals (includ-
ing humans), which in certain circumstances can greatly reduce the vitality
and fertility of individuals and even lead to their death. Based on results
of our studies on tree shrews and wild rabbits, I shall present in the following
section my view of the relevance of recent stress research for the under-
standing of animal behavior.
B. SOCIAL THAT
CHALLENGES MAYRESULT
IN STRESS
RESPONSES
1. Social Conpict
a. Adrenocortical Activity and Fights. Fights over limited resources such
as food, shelter, territories, and/or rank are among the most conspicuous
behaviors in animal societies. All mammals react to acute stress, such as is
certainly induced by a fight, by immediate activation of their sympathetico-
adrenomedullary and hypothalamo-pituitary-adrenocortical systems. If
such fights occur only infrequently, they have no consequences for the
fertility and health of an individual: For instance, tree shrews can be sub-
jected to a 30-min confrontation daily over a period of several weeks without
suffering any detrimental physiological effects. If these daily fights occur
more frequently, however, they can have grave detrimental effects on the
fertility and health of the animal and even-as described at the beginning
of this chapter in Antechinus-rapidly result in death. One or two aggressive
conflicts per hour, not unusual in a natural environment, may be sufficient
to have a damaging effect, as the adrenocortical hormones require one to
several hours, depending on the species and the interaction, before they
(and hence many other parameters) revert to their original concentrations.
As is the case in Antechinus, and also in many other species with season-
ally restricted reproduction, this phase of increased aggression is character-
ized by correspondingly increased adrenocortical activity in individuals, as
was shown in the early 1960s by Bronson (1963, 1964) in his studies on
natural populations of woodchucks (Marmora monax). The same is also
true for other species (Saad and Bayle, 1985; Saboureau eta/., 1977; Schiml
el al., 1996), as shown later in more detail, based on our investigations on
European wild rabbits.
European wild rabbits (0ryctoIagu.y cnniculus) live in small territorial
groups of 1-3 males and approximately double the number of females. The
territories are intensively defended by the males against external rivals
during the reproductive period. Within the groups both sexes exhibit sepa-
rate intrasexual linear ranking systems. Aggression in both sexes is particu-
larly high at the beginning of the reproductive period, whereas for the rest
of the year wild rabbits live together largely in peace (e.g., Brambell, 1944;
Cowan, 1987; Lockley, 1961; Marsden and Holler, 1964; Myers and Poole,
1959; Mykytowycz, 1958; Southern, 1940).
In order to gain information on the influence of agonistic behavior on
physiological parameters, adult wild rabbits of a large natural population
on the North Sea island of Sylt were investigated at the beginning of the
reproductive period, when aggression was maximal (end of March), and
again after the end of the reproductive period (October/November). At
these two times of the year a total of 500 animals were shot between 18:OO
and 19:OO hours over a period of 3 years, and within less than 3 min of
their death blood samples were taken and different organs extracted for
endocrinological and other investigations. As food availability, tempera-
ture, and day length were more or less equivalent during the two hunting
seasons in spring and late autumn, differences in physiological stress param-
eters should be due largely to differences in aggression.
As expected, we found greatly increased adrenocortical and sympathet-
ico-adrenomedullary activities as well as many other changes in both sexes
in spring, indicating high levels of stress. That is, under natural conditions,
wild rabbits of both sexes show endocrine stress responses of the same
magnitude as those demonstrated mainly in rodents under laboratory condi-
tions (Fig. 13).
We have closely investigated these relationships between social behavior
and physiological stress responses over a period of 10 years in a population
of wild rabbits living in a seminatural environment in an enclosure covering
an area of approximately 22,000 m2 (e.g., Eisermann et a/., 1993; Kiinkele
and von Holst, 1996). In these conditions, both sexes also start fighting
STRESS AND ITS RELEVANCE €OR ANIMAL BEHAVIOR 47
Adrenal gland Corticosterone Cholesterol

400 I (ng/ml serum) 32 1 (mall00 ml . serum)
300 24
200 16
100 8
0 0
TH activity Adrenal medulla PNMT activity
7.5 (nmollh adrenals) 500 (Cell nucleus volume 36 1 (nmollh adrenals)
in LP)
6.0 400 27
4.5 300 18
3.0 200 9
1.5 1no n
I
Males Females Males Females Males Females

Spring: 89 males, 76 females Autumn. 25 males, 20 females
FIG. 13. Indices of adrenocortical and sympathetico-adrenomedullary activities of adult

male and female wild European rabbits in spring and autumn. The volume of the cell nuclei
of adrenal medullary cells was histologically determined from 10 males and females in each
season; their enlargement indicates a markedly increased adrenomedullary activity in spring.
All data means ( + SEM). Significant differences between the spring and autumn data:
* p < .05, * * p < .01.
heavily in spring and many injuries result (Fig. 14). In the males these fights
are over territories and ranking positions and decrease only slightly in the
course of the reproductive period. Females exhibit a bimodal pattern of
aggression with maxima at the beginning and the end of the reproductive
period. At the beginning of the reproductive period, fights usually occur
over ranking positions within their groups as well as with external females,
while in autumn most aggression is directed against young animals attempt-
ing to enter the groups (Fig. 15). In concert with these behavioral changes,
there are marked changes in adrenocortical activity in both sexes. Conflict
avoidance by the restriction of aggression mainly to the reproductive period
is therefore a useful means of avoiding such stress reactions.
4.0 1 Male sexual behavior

(interactionshour)
[7 sexual following
urination
3.0
2.0
1.o
0.0
2.4 Fresh wounds

(number)
1.8
1.2
0.6
0.0
250 Corticosterone
200
150
100
50
Dec Feb Apr Jun Aug Oct Dec
FIG. 14. Sexual behavior (medians), wounds, and serum corticosterone ( M _f SEM) of
wild European rabbits, kept in a 22,000-m2 enclosure under natural conditions. Data from
about 25 males and 50 females, observed over a 4-year period. Blood samples forcorticosterone
determination were taken monthly from the animals 1 hr after maximal stimulation of their
corticosterone release by an injection of ACTH. For behavior analysis each individual was
observed about 8 hr per month. Unpublished data; after Schonheiter (1992).
b. Physical versus Psychosocial Processes. At the onset of mammalian

stress research some authors suggested that intensive physical effort during
fighting and wounds resulting from these fights might be the principal
factor in adrenal enlargement; however, most studies found no correlation
between injuries from fights and adrenal weight, and suggested psychologi-
cal (psychosocial) factors as underlying causes (e.g., Christian, 1963). One
of the first studies supporting this hypothesis was carried out by Davis
and Christian (1957), who demonstrated a significant correlation between
Aggressive behavior (interactions per hour)

9.0 Males
6.0
3.0
0.0
Females
0adult group members
Jan Mar May Jul Sep Nov Jan
FIG. 15. Aggressive behavior of adult male and female wild European rabbits toward
juveniles and adults (medians; for animal numbers see Fig. 14).
adrenal weight and social rank in groups of male house mice. Another
study by Barnett (1958) found increased adrenal weights only in subordinate
wild Norway rats (Rattus norvegicus) in a resident intruder paradigm. Fur-
thermore, Bronson and ElefthCriou (1965b) showed that even the mere
exposure without physical contact of subordinate mice to fighters-if they
had previously been defeated in a confrontation-produces adrenocortical
responses of the same magnitude as those observed in mice actually attacked
and defeated. Similar findings have been obtained from research on Syrian
hamsters (Mesocricetus auratus) (Human et al., 1992). Furthermore, re-
search by Fokkema and Koolhaas, using chronically catheterized laboratory
rats, showed that defeated males exhibited over twice the increase in blood
pressure during brief dyadic encounters than their superior rivals (Fok-
kema, 1985; Fokkema and Koolhaas, 1985). If an animal had previously
been defeated and was then threatened by exposure to the victor, while
penned in a small wire mesh cage, the mere presence of the victor raised the
former victim’s blood pressure to the same level as during the direct defeat.
Extreme psychosocial stress can even rapidly cause death, as has been
demonstrated in wild Norway rats (Barnett, 1958, 1964, 1988; Barnett et
al., 1975), tree shrews (von Holst, 1972a,b, 1985a), rhesus monkeys (Hamil-
ton and Chaddock, 1977), and humans (Stumpfe, 1973). Death in these
cases is always associated with behavioral impairment, indicating a state
of helplessness or loss of control, and extremely heightened adrenocortical
activity (as shown in rats and tree shrews). An example of these mechanisms
is given in the following section.
c. Social Stress in Tree Shrews. Tree shrews (Tupaia belangeri, order
Scandentia) are small diurnal mammals distributed throughout Southeast
Asia. In the wild, tree shrews live in pairs in territories that they defend
vigorously against intruders of their own sex. In the laboratory, adult males
(and females) also immediately attack intruders of their own sex and nor-
mally defeat them within a few minutes. A short time after the fight, the
winners show no further signs of arousal and pay virtually no attention to
the defeated animals. The losers, in contrast, creep into any hiding place,
which they leave only to eat and drink. During the following days, fights
between the animals are extremely rare or nonexistent. Nevertheless, the
losers die within a few days. Death is not a result of physical exertion
during the fights, nor are wounds the cause of death, as the animals usually
inflict only superficial scratches and bites on one another. Death is rather
more the consequence of the continual presence of the winner, as was
shown by the following experiments: An adult male was placed in the cage
of a male conspecific (an experienced fighter), which usually immediately
attacked the intruder and subdued him in less than 2 min. Afterward, both
animals were separated either by a nontransparent partition or by a wire
mesh partition, so that the loser could no longer be attacked but could
continually see the threatening winner. Short fights were repeated every
1-3 days.
Losers separated by a nontransparent partition from the winner recov-
ered from the fights just as fast as the winners and did not die prematurely,
even when they were subjected to short daily fights over weeks. The situa-
tion of the losers within sight of the winners, however, was completely
different: As from the first subjugation, all the submissive animals sat in a
corner of their part of the cage or in the sleeping box attached to the cage
for practically the whole day, hardly responding to external stimuli. During
confrontations, they did not even attempt to escape the attacks of the
dominant animals, but usually suffered them without any attempt t o defend
themselves or to flee. After the first subjugation their body weight decreased
daily at an individually stable rate of 2-8% of their initial weight and all
animals died within 2-20 days, if not separated earlier.
As these results show, death of the submissive animals is not a direct
result of the fights and their physiolgoical consequences, but is rather a
result of central nervous (emotional) processes in the defeated animal,
induced by the constant presence of the threatening winner.
In all submissive animals dramatic stress responses were seen: Among

other responses, serum glucocorticosteroid levels rose to more than six
times their initial levels, serum concentrations of both thyroidal hormones
(thyroxine and triiodothyronine) decreased to less than 40%, and those of
testosterone to less than 10% within a few days. As shown by histological
examinations, within a few days spermatogenesis ceased completely and
all animals became sterile. In addition, a dramatic drop in the numbers of
lymphocytes as well as of basophil and eosinophil granulocytes to less
than 20% of their initial values indicates strong immunosuppression (von
Holst, 1985a).
The physiological cause of death differed among the animals. In animals
dying within 8 days of the confrontation, urea nitrogen (and creatinine)
levels in the serum rose to more than ten times their initial values, leading
to death by uremia, while in those animals surviving longer, these increased
only slightly. The cause of death in the latter is not known (von Holst,
1972a,b, 1985a). (Of course, the introduction of an individual into the cage,
or “territory,” of an experienced fighter without any possibility of avoiding
the rival is an unnatural and extremely stressful situation, especially in a
territorial animal such as the tree shrew.)
To obtain information on a less severe form of stress or even on adapta-
tion to it, two male tree shrews unknown to each other were put together
in a cage with two separate sleeping boxes, water bottles, and feeding
dishes. In this situation the animals did not start to fight immediately, but
first hesitantly explored and marked the cage. Slight fights usually began
within the first few hours, leading in most cases to clear dominance relation-
ships within 1-3 days. While the behavior of all males was more or less
comparable before the fights, it changed considerably after the dominance
relationship was established, depending on the social position of the ani-
mals. Although both animals continued to live together in the relatively
small cage, the winners more or less ignored the losers and attacks on the
latter were rare or even completely absent. The losers, on the other hand,
drastically moderated their behavior. On the basis of behavioral differences,
two types of losers could be distinguished: subdominant and submissive an-
imals.
Submissive animals corresponded to those of the first experiment: They
crouched in a corner of the cage or in a sleeping box and left their hiding
place only to drink and eat hastily. They even tolerated the infrequent
attacks of the winners without any attempts to defend themselves or to
flee. They ceased grooming completely and their fur became rough and
dirty. To the human observer they gave an apathetic or depressive impres-
sion. All submissive animals died within 2 weeks.
Sudominant animals, in contrast, showed greatly increased locomotor

activity, watching the movements of the dominants continually and attempt-
ing to avoid possible confrontations by giving way or fleeing. If a confronta-
tion could not be avoided, they even defended themselves. Under these
conditions, subdominants were capable of living in the presence of the
dominants for weeks, albeit with a reduced freedom to move.
In concert with these behavioral changes, body weights and many differ-
ent physiological parameters in the animals changed drastically. In ac-
cordance with the stress concept, the confrontation had the immediate
effect of activating the sympathetico-adrenomedullary and the pituitary-
adrenocortical systems in both rivals: Accordingly, the serum concentra-
tions of catecholamines and glucocorticosteroids as well as the heart rates
of all animals were greatly increased (Fig. 16).
As soon as the dominance relationship was clearly recognizable in the
behavior of the tree shrews, all stress reactions in the dominant animals
disappeared in spite of continued occasional fights. Moreover, glucocorti-
costeroid concentrations in the blood dropped marginally below the original
values, their body weights increased, and their gonadal functions improved:
After about 3 weeks, the dominant animals were significantly heavier than
before the confrontation and serum testosterone concentrations had in-
creased by approximately 100% (Figs. 17 and 18).
Cortisol Epinephrine Norepinephrine
6.0 1 (ng/ml serum)
I
10.0
(nglml serum)
20 5.0 8.0
15 4.0 6.0
10 3.0 4.0
5 2.0 7
D SD SM D SD SM D SD SM
FIG. 16. Serum concentrations ( M -+ SEM) of cortisol and catecholamines o f , minant

(D), subdominant (SD), and submissive (SM) tree shrews 8 days before (light bars) and two
days after (cross-hatched bars) start of a confrontation. Blood samples were always taken
2 hr before the activity period. Cortisol data (20-30 animals per group) are initial values,
catecholamine data (10 animals per group) are BSCT values. Significant differences to initial
values: * p < .05; * * p < .01; ***p < ,001.
Controls Dominants Subdominants Submissives
!
40 cortisol
(nglml serum)
20
301
B 2 10 20 B 2 10 20 B 2 10 20 6 2 10 20
FIG. 17. Serum concentrations (M t SEM) of cortisol and testosterone of male tree shrews
before (B) and at different days after start of the experiment. 20 controls remained in their
living rooms, but were treated as the other animals (handling for weighting and blood sampling)
and 25-40 males per group that confronted each other. See text for further details.
Subdominants
Submissives
0 6 12 18 24
Days after start of experiment
FIG. 18. Body weight changes of the animals after the start of the experiment ( M ? SEM):
for animal numbers see Fig. 17. Initial body weights of the males: 190-220 g.
Defeated tree shrews were characterized by reduced body weights and

a number of hormonal and other physiological changes; among other ef-
fects, the serum concentrations of testosterone, insulin, and thyroid hor-
mones decreased drastically. Overall, these effects were the same in sub-
dominant and submissive animals, differing only in their extent (Fig. 19).
Qualitative differences were, however, to be found in their sympathetico-
adrenomedullary and pituitary-adrenocortical systems.
Active and passive stress reponses in tree shrews. In dominant animals,
the activity of the sympathetico-adrenomedullary system reverted back to
Testosterone Protein Kidneys

9.0 j (g/100mlserum)
30 8.0 4 *
20 7.0
10 6.0
0 5.0
Triiodthyronine Insulin Triglycerides
1.0 1 (ng/ml serum) 36 1 (ng/ml serum) 150 1 (mg/l00 ml serum)
0.8 27
125 i I I
0.6 18 100
0.4 9 75
0.2 0 50
Hemoglobin Epididymides
1 (g/lDo
20
18
ml blood)
!joo
400
1 ;mg)
16 300
14 200
3I 7L ion
c D SD SM C D SD SM C D SD SM
FIG. 19. Several physiological measures and organ weights ( M ? SEM) of controls and
experimental male tree shrews 10 days after start of the experiment. Animal numbers as in
Fig. 17. Significant differences to controls: * p < .05: **p < .01; ***p < ,001. C: controls:
D: dominants: SD: subdominants: SM: submissives.
its original state after the dominance relationship had been established,
and the serum catecholamine levels even showed a tendency to levels lower
than those before the confrontation. In contrast, subdominant animals
exhibited a continually increased sympathetico-adrenomedullary activity.
While, correspondingly, the heart rate in the dominant animals returned
to normal once dominance was established, it remained high in subdominant
animals, not only throughout the day (when an attack by the dominant
animal was always possible), but also at night when they were sleeping in
their own sleeping box. Nocturnal heart rates in subdominant animals were
almost equivalent to diurnal values, thus abolishing the original day-night
rhythm (Fig. 20). The tyrosine hydroxylase activity in their adrenal
glands-an index of sympathetic activity-also increased on average by
loo%, in comparison to dominant or control animals (Fig. 21). Contrary
to the Selyean concept, the serum levels of the glucocorticosteroids de-
creased to initial levels; their adrenocortical secretory capacity, however,
500 Male dominant
400
-
d
300
E6= 200
100
n
u)
iii 500
al
a
E 400
u)
c
m 300
E
3 200
L
g
2
100
0)
5 500 1 Male submissive
72
400
I
300
200
100
Days before and I after start of confrontation
FIG. 20. Heart rates of a dominant, a subdominant, and a submissive male tree shrew
before and after the start of the confrontation.
Cortisol Corticosterone Adrenal weight

40 (nglmlserum) *** 60 **
I nglrnl serum) ***
I
I
30 45 36
20 30 32
10 15 28
0 0 24
Urea nitrogen irosine hydroxylase PNMT
100 (mg/l00
ml *** 10.0 :nmol/h *
serum) I adrenal) I adrenal) *
80
7.5 2.4
60
5.0 1.6
40
20 2.5 0.8
0 0.0 0.0
Lymphocytes Eosinophils Erythroblasts
2800 600 (n/ pl blood)
2100 450 540
1400 300 360
700 150 180
0 0 0
C D SD SM C D SD SM C D SD SM
Data 10 days after start of experiment
FIG.21. Several measures of the adrenocortical and adrenomedullary systems as well as

some blood cell numbers ( M i SEM) of controls and confronted male tree shrews 10 days
after start of the experiment. Animal numbers and abbreviations as in Figs. 17 and 19.
Significant differences to controls: * p < .05: **p < .01: ***p < ,001. See text for further details.
increased to the same level as in submissives, indicating a heightened corti-

sol release and elimination under active stress (Figs. 17, 21, and 22).
Submissive animals exhibited the opposite reactions: a tendency to a
decreased sympathetico-adrenomedullary activity (Fig. 21), as indicated by
their lowered adrenal tyrosine hydroxylase activities; and a substantial
increase in their serum levels of glucocorticosteroids, as well as of their
adrenal capacities (Figs. 17, 21, and 22), which was probably responsible
*** ***
.L
D SD SM D SD SM
FIG.22. Serum cortisol levels ( M t SEM) 15 min after start of blood sampling challenge
tests of dominants (n = 11). subdominants (n = 11) and submissives (n = 6 ) before (left)
and 10 days after (right) start of the confrontation. Significant differences: * p < .05: ***p <
.001. Note the significantly lower initial values of the males that became submissive during
the confrontation.
for the increased loss of muscle and adipose tissue, leading to a dramatic
loss of weight averaging 5% daily (von Holst, 1986a, 1994; Stohr, 1986).
Interestingly, the adrenal capacities of prospective submissive animals were
significantly lower before the confrontation than those of the other two
groups (Fig. 22).
Confrontation also triggered marked immunological changes: No signifi-
cant change was observed in leucocyte numbers and subsets in the blood
of dominant and subdominant animals. In contrast to dominant animals,
however, the efficiency of lymphocytes and phagocytes in the subdominant
animals was clearly reduced (Fig. 23). In submissive animals, substantial
changes were also found in those types of leucocytes indicative of strong
immune suppression (e.g., Fig. 21: lymphocytes and eosinophil granulo-
cytes). Accordingly, the proliferation capacity of their lymphocytes was
reduced to less than 20% of original values following a 10-day confronta-
tion period.
Subordination and presence of the dominant animal therefore affected
the immune system in subdominant and submissive animals. These effects
were qualitatively equivalent in both categories of animals as far as their
inhibiting effect on proliferation of lymphocytes was concerned, but differ-
ent regarding the distribution of the different types of leucocytes in the
blood: General statements on the possible qualitative differences in immu-
nological reactions in these ethologically and physiologically distinct subor-
dinates are not yet possible based on available data.
Before After Before After
FIG.23. In vitro lymphocyte proliferation (LP) after stimulation with the mitogen conca-
navalin A (Con A) and in vitro phagocytosis of monocytes and granulocytes ( M ? SEM) of
16 subdominant male tree shrews before and 10 days after start of the confrontation (for
details see Section 11,C: direct assessment of immunological parameters). While their lympho-
cyte numbers during the confrontation were not different from initial values, their proliferation
capacity decreased markedly ( p < ,001).
d. Assessment of Dominance between Rivals. As the results of research

on tree shrews show, prospective winners and losers exhibited differences
in body weights and cortisol values after only 2 days (Figs. 17 and 18),
even though it was not usually possible to predict the outcome of the
confrontation based on the animals’ behavior at this point.
Similiar results were found in domestic guinea pigs (Sachser and Lick,
1989): At the age of 30 days, the authors removed juvenile males from
their breeding colonies and housed each of them with a female in a 1-m2
enclosure. At the age of about 8 months, two males were confronted by
removing the partition between their enclosures. In all cases these confron-
tations escalated into fights between the two opponents, which declined to
low levels after the first day. However, dominance relationships between
the rivals could not be distinguished before day 4 of confrontation. Never-
theless, at day 2-3, prospective winners and losers already differed signifi-
cantly in nearly all physiological parameters measured: Prospective losers
exhibited a higher loss of body weight, their serum concentrations of cortisol
and catecholamines were two to three times higher than those of prospective
winners, while their testosterone levels decreased to about 30% of their
original value. Commencing with the fourth day of controntation, the physi-
ological changes became so dramatic that the losers died within a few days.
As these data show, the outcome of the agonistic encounters could be
confidently predicted from physiological data 1-2 days before it could be

derived from the behavior of the individuals.
These results of research on tree shrews and guinea pigs demonstrate
the animals’ capability of predicting the outcome of a confrontation while
it is taking place and of producing the corresponding physiological reactions.
How quickly endocrine changes occur during the changing assessment of
the situation by the animals is demonstrated by Schuurman’s results (1981)
with laboratory rats, which had been provided with chronic jugular vein
cannulae for frequent blood sampling. Schuurman introduced a male rat
into the home cage of an aggressive male conspecific. At first both animals
were separated by a wooden partition, which was removed after a habitua-
tion period of 2.5 hr for a l-hr confrontation. During the first quarter of
the confrontation fierce fighting took place, but with no apparent dominance
relationship between the two rivals: Both showed offensive as well as defen-
sive behavior, and their plasma corticosterone levels rose to about five
times their initial values. After about 15 min a dominance relationship
developed, which resulted in distinctly different adrenocortical responses
in both animals: Although the victors continued offensive fighting, their
plasma corticosterone levels declined. On the other hand, plasma corticoste-
rone levels of losers continued to rise and the animals exhibited much
defensive and submissive behavior, but no longer any offensive behavior.
At the end of the encounter, plasma corticosterone level in losers was more
than twice the level found in victors.
Once a dominance relationship has been established in mammals, fights
between rivals usually decrease. Nevertheless, as shown for tree shrews
(see earlier discussion), the permanent presence of a dominant rival can
cause dramatic stress responses in subordinate animals and ultimately even
lead to their death within a few days. In spite of these highly negative
consequences of subordination, fights for dominance are often astonishingly
slight or even completely absent in tree shrews. This is apparently due to
the fact that male tree shrews are able to recognize the potential “strength”
or “dominance” of a rival before the first physical interaction, as is shown
by the following experiments.
Two animals were transferred to an experimental room and housed in
one cage, but separated for 20 days by a wooden partition. Before the
transfer into the experimental room, and on days 10 and 20, blood samples
were taken from all animals for the determination of several endocrine
and immunological parameters. On days 21-23, the two animals confronted
each other daily for 10 min, while the rest of the time they were separated
by the wooden partition. Depending on the results of these confrontations,
the animals were designated prospective “dominants” and “subordinates.”
From the first day in the experimental cage both animals apparently
recognized the presence of the rival: They sniffed intensively at the wooden
partition and marked it. Animals that later turned out to be subordinate
in the confrontation seemed more alert and exhibited more locomotor
activity, as was also evident from their slightly decreased daily resting times
compared to their prospective dominant rivals. While few overt behavioral
differences were observed between the two groups, there were significant
immunomodulatory effects (Fig. 24). These effects were in opposite direc-
tions in the animals of the two groups: Tree shrews that later became
subordinate showed indications of an immunosuppression, while in prospec-
tive dominants the activity of the immune system improved. In contrast to
these immunomodulatory effects, we found only very slight changes in
adrenocortical or sympathetico-adrenomedullary activities. Amazingly, the
immunosuppressive effects in subordinate animals before any physical con-
tact between the rivals were of the same magnitude as those in direct
confrontations with constant physical presence of the rival (Fig. 25).
To exclude the possibility that these opposite immunomodulatory reac-
tions were consequences of differences in the constitution of the rivals, the
Lymphocyte proliferation Interleukin 1

200
150
100
I
3
7
-
50
m
._
-
._
c
._
0
c o
400
Interferon gamma
c
._
111 300 ***
0
5 200
100
0
H 10 20 H 10 20 H 10 20 H 10 20
Dominants Subdominants Dominants Subdominants
FIG.24. Several immunological measures taken from male tree shrews housed in the same
room with a potential rival ( M 2 SEM); data 8 days before (H) as well as 10 and 20
days after transfer into the experimental room: 20 prospective dominant and 20 prospective
subdominant males. Significant differences: **p < .01; ***p < ,001. See text for further details.
Lymphocyte proliferation Interferon gamma
1
120 - (lo3 cpm/lo5 cells)
+ r *
80 - 6.0
40 - 3.0
0- 00
10 2 12 2 12 10 2 12 2 12
Home Exp Conf Home Exp Cod
Experimental conditions and days of blood sampling
FIG.25. In v i m lymphocyte proliferation after stimulation with the mitogen Con A and
interferon production of 20 male subdominant tree shrews in their home room (Home: days
0 and lo), 2 and 12 days after transfer into an experimental room with a rival (Exp), and 2
and 12 days after start of the confrontation (Conf); significant differences to initial values
(= Home 10): * p < .05: * * p < .01. See text for further details. Unpublished data from
Vitek (1996).
tests were repeated 3 months later with all experimental animals in different
combinations: Previous dominants were confronted with previous domi-
nants, and previous subordinates with previous subordinates. In all cases,
the animals formed new dominance relationships, with identical immuno-
logical consequences for the final winner and loser, just as in the first
experiment. Thus, the differences in these immunomodulatory reactions
of prospective dominants and subordinates were only dependent on the
situation, that is, on the “quality” of the prospective rival.
As preliminary data show, male tree shrews are capable of obtaining
information on the potential strength of a rival through olfactory signals,
which originate from urine and glandular areas used for territorial marking.
It is entirely unknown, however, what determines this olfactorily communi-
cated “strength” of a rival. As our experiments with tree shrews have
shown, body size or body weight are of no predictive value for the outcome
of a fight between unknown rivals; the same is true for differences between
animals in testosterone serum concentrations or excretion rates.
2. Dominance Relations and Stress Responses in Other Mammals

a. Pituitary-Adrenocortical and Sympathetico-Adrenomedullary Sys-
tem. As shown in the previous section, subordinate tree shrews exhibit
behaviorally dependent differences in their stress responses, with only the
behaviorally inactive and apathetic submissive animals corresponding phys-
iologically to the Selyean concept. These findings strongly support the
concept of Henry and associates, which proposes that mammals exhibit

two types of stress with differing neuroendocrine responses and differing
ultimate disease states (e.g.? Ely and Henry, 1978; Henry and Meehan,
1981; Henry and Stephens, 1977): Cannon’s flight or fight response, ac-
companied by heightened sympathetico-adrenomedullary activity, which
may eventually lead to cardiovascular deterioration (“active stress”); and
Selye’s stress response, characterized in particular by heightened pituitary-
adrenocortical activity (“passive stress”).
In species such as the tree shrews, which live in territorial pairs in natural
conditions, adult individuals of the same sex cannot be kept together for
any length of time, since, depending on their passive or active coping
behavior, this may result in the death of subordinate animals within a few
days to weeks. For this reason, our confrontation experiments with males
were always terminated after 3 weeks. Following this time span, subdomi-
nant animals also succumbed to apparent signs of coronary insufficiency.
The situation is different in animals that normally live in groups with
idiosyncratic yet relatively stable social roles. In most species, hierarchical
systems with dominant and subordinate animals develop in such a way that
subordinates can usually live in the presence of dominants with only very
slight or no signs of stress at all. In situations of continuing conflict that
result from social instability within the group, life may be stressful for
subordinates as well as for dominants, even in the absence of overt aggres-
sion, as was shown by Henry and associates in their extensive studies on
laboratory mice (e.g., Ely, 1981; Ely and Henry, 1978; Henry and Stephens-
Larson, 1985).
After weeks of isolation, the authors introduced several male and female
laboratory mice into colony cages consisting of small boxes connected by
tubular runways and designed to induce frequent social interactions. This
always resulted in intensive fights among the males, which led within a few
weeks to a relatively stable social system with one dominant male and
several subordinates in each colony. During the time of colony formation
the physiological responses of all animals were typical of general nonspecific
arousal, with increased activities of their adrenocortical and adrenomedul-
lary systems. Once the social hierarchy was established, overt aggression
and strong stress responses decreased. Nevertheless, the behavior and physi-
ological response patterns of dominants and subordinates differed: Domi-
nant animals were significantly more active and vigilant than the subordi-
nate animals; they constantly visited the boxes of the cage system and tried
to exert control over all other animals. The subordinate males, on the other
hand, were restricted to very small areas of the population cage and showed
behavioral withdrawal, which according to the authors minimized aggressive
encounters with the dominants. As demonstrated previously by other au-
thors in crowding experiments on mice and other species (e.g., Bronson,

1973; Chapman et al., 1969; Louch and Higginbotham, 1967; Popova and
Naumenko, 1972), even in this more naturalistic experimental design subor-
dinate mice maintained slightly elevated adrenocortical activities. In con-
trast, dominant mice developed heightened adrenomedullary activities and
a moderate hypertension, which apparently permitted the animals to main-
tain the vigilant patrolling behavior necessary to the control and stabiliza-
tion of the colony. Testosterone levels in subordinate mice decreased to
about 30% of that of controls kept in standard laboratory cages, while the
levels in the dominants did not differ from those in controls, in spite of
their heightened sympathetico-adrenomedullary stress responses. All these
physiological response patterns typical of dominants and subordinates could
be reversed by experimentally induced changes of the social positions of
individuals (e.g., removal of dominant males). When the animals were
separated before 5 months of social conflict, all these stress values returned
more or less to those of control animals.
If, however, the situation persisted, within a few months pathophysiologi-
cal changes began to develop in all animals: Fixed hypertension and in-
creased PNMT values, increased heart weights, and histopathological dete-
rioration, such as interstitial nephritis, aortic arteriosclerosis, intramural
coronary arteriosclerosis, and myocardial fibrosis developed, which after 9
months of colony living remained irreversible even after months of isolation,
and led to the premature death of many individuals (Henry and Stephens-
Larson, 1985; Henry et al., 1971; Vander et al., 1978).
As shown by these data, in mice kept in complex population conditions,
dominant individuals show predominantly active stress responses, while
the subordinate animals show slight passive stress responses. These re-
sults are in contrast to our data on dominant tree shrews, which show no
sympathetico-adrenomedullary activation. However, this is probably due
to the striking differences in the social organization of these two species.
In tree shrews, which live in territorial pairs in the wild, a defeated rival
is apparently no longer threatening, and therefore, even in small cages,
does not elicit any behavioral or physiological arousal. In contrast to domi-
nant mice, dominant tree shrews even show slightly decreased adrenocorti-
cal activity. Lundberg and Frankenhaeuser (1980) particularly emphasize
this “bidirectional nature of the pituitary-adrenal response.” In their study
on humans, adrenocortical suppression was demonstrated in conditions
characterized by high levels of control and predictability. According to the
authors, this is consistent with the conclusion by Levine and associates
(1979), that reinforcement is an important cognitive factor mediating sup-
pression of the adrenocortical system. This also seems to be the case in
dominant tree shrews. Wild mice in natural conditions, however, live in
64 DIETRICH VON HOLS’I
groups containing one dominant and many subordinate males. The domi-
nant male must constantly control and keep in check the subordinate males
and this apparently necessitates heightened adrenomedullary activity
(Crowcroft, 1955; Lloyd, 1973).
Thus, it is not the social position that determines the physiological state
of an animal, but the effort of achieving and maintaining the status (i.e.,
whether the position is endangered or not). This has been clearly demon-
strated by Lundberg and Frankenhaeuser (1980) in the experiments on
humans, which demonstrate that pituitary-adrenocortical activation is asso-
ciated with negative feelings of distress, and sympathetico-adrenomedullary
activation with feelings of alertness and a readiness to act. A similar conclu-
sion was drawn by Ursin and associates (1978), who identified a “cortisol
factor” and a “catecholamine factor” in their analysis of data collected in
a study of trainee parachutists.
It has to be emphasized, however, that in these and many similar studies
with mice chronic high levels of stress were induced by using males that
were housed singly after weaning for many weeks before colony formation
(“unstable social situation”). In contrast to these socially deprived individu-
als, mice that had been raised in groups were able to live together in stable
social groups without overt aggression and stress responses (Fig. 26). As
pointed out already (Section II,B,3,c), these results again demonstrate the
crucial role of social experience after weaning for an animal to cope with
social conflict in a more or less stress-free way.
These results lead to the conclusion that, depending on species, group
composition, and group stability, dominant individuals may be character-
ized by lower or higher hypophyseo-adrenocortical and/or sympathetico-
adrenomedullary activities than their subordinate conspecifics. Our knowl-
edge of the effects of social positions and behavior on differing endocrine
stress responses in mammalian species, however, is scanty, as most research-
ers have chosen to work with one system only (usually the adrenocortical).
Accordingly, countless publications on many different species have dem-
onstrated that repeated subjugations or a subordinate social position result
in increased adrenocortical and decreased gonadal activities. However,
little is known about the species-specific and context-dependent positive
or negative consequences of dominant positions for the sympathetico-
adrenomedullary system (including heart rate and blood pressure). An
exception to this is provided by several studies on rats and monkeys (mainly
on the effects of psychosocial stress on blood pressure) and, as mentioned
earlier, to some extent also by studies on wild rabbits, tree shrews, and
guinea pigs.
An overall analysis of our wild rabbits housed in large enclosures revealed
a distinct relationship between the social rank of males and females and
Systolic blood pressure Tyrosine hydroxylase
I
240 1 (oh of controls) ,
140
120
100
Renin Corticosterone
1(nglml plasma x hr) 1 (nglml plasma)
10 150
5 100
0 50
FIG.26. Effects of chronic psychosocial stress on several physiological measurements of

male mice. Systolic blood pressure, adrenal tyrosine hydroxylase activity, and plasma values
of renin and corticosterone of mice housed singly (light bars) in small cages as well as of
mice housed in mixed-sex groups for 6 months. Significant differences between animals housed
in unstable (cross-hatched bars) social groups and the other housing conditions (stable groups-
striped bars) are indicated: **, < .01; ***, < .001. See text for further details. Adapted from
Henry (1992). with kind permission from Transaction Publishers, New Brunswick.
their adrenocortical activities during the reproductive season (Fig. 27). At

the group level, however, marked differences existed (Fig. 28): Generally,
males living without rivals within their groups had the lowest corticosterone
challenge test values. Depending on the number of subordinate rivals in a
territory (and hence the social instability within their groups), the adreno-
cortical activities of dominant males rose to the values of subordinate males;
thus no rank-related differences were evident. Furthermore, the number
of females living within a group had some influence on the adrenocortical
activities, especially of the dominant males. Dominant females, in contrast,
had generally the lowest adrenocortical activities of all animals within their
groups, although their activities differed to some extent among the groups.
Taken together, these data indicate that the adrenocortical activities of
dominant wild rabbits depend on the composition and stability of their
Corticosterone values Heart rates

(bpm)
I
250 240
200 200
monthly mean)
1 2 3 >3 1 =.I
Social rank of animals
FIG.27. Relationship between social ranks, adrenocortical activities, and heart rates of
wild European rabbits, which lived in large field enclosures. Corticosterone measures of
four reproductive seasons from animals from a 22,000-m2 field enclosure. Blood samples for
hormone analysis were taken from all animals once every month 1 hr after maximal stimulation
of their corticosterone release by an injection of ACTH; corticosterone values of the males
are absolute serum concentrations; since corticosterone measures of females show marked
variations during the reproductive season (see Fig. 14). their values are given as deviations
from the monthly mean of all females. All data are means ( 2 SEM) of the mean of 4-6 values
of each animal from one reproductive season. Heart rates of the animals were determined
telemetrically by transmitters implanted ( M 2 SEM of 30-60 days of measurement per
individual) in animals that lived in about 150-m2 enclosures. Social ranks were determined
by behavioral observations (> 40 hr per animal). Animal numbers are shown at the bottom
of the bars.
groups, which may explain that, especially under unstable social conditions
in wild rabbits as well as in other species, no rank-dependent differences
in adrenocortical activities are found.
The heart rate in dominant individuals of both sexes, living in smaller
enclosures (about 150 m2) in groups of 2-3 males and as many females,
was also lower compared to that in subordinates, and every change of
rank resulted in a corresponding change of heart rate (Fig. 27; see also
Eisermann, 1992).
Most stress research has been performed on laboratory rats. Rats are
highly social and intensive fighting is present only for as long as the animals
are unfamiliar with each other and no stable hierarchies have developed
Two or more females per group
1-male groups I 2-3 males per group
5 280
b
In
.
E
c
0)
240
._
(1 200
In
-
W
- 160
L
0
> 3 males per group
al I I
280
0
&
+
In 240
8
._
5
0 200
160
alone 1 2 3
Social rank of males
FIG. 28. Relationship between social ranks and corticosterone challenge values of wild
European rabbits in relation to their group composition. Data from animals living without
male rivals in their territories are shown in top and bottom parts. See Fig. 27 and text for
further details.
(Barnett, 1975; Calhoun, 1963). During this period all animals show typical
stress responses including elevated blood pressure. Once a stable hierarchy
with a dominant male has developed, fighting more or less ceases and blood
pressure decreases (Henry et al., 1993). Nevertheless, differences between
dominants and subordinates persist, as has been shown by Dijkstra and
colleagues (1992) in male Wistar rats housed in mixed-sex groups in complex
colony cages. Compared to pairwise housed controls, the dominant animals
exhibited significantly heightened testosterone plasma levels, while those
of the subordinates were in the range of the controls; corticosterone plasma
levels were increased in both ranks, but in subordinates the increase was
about 150%, three times higher than in dominants.
If the composition of a mixed-sex group of rats is changed regularly,
thus preventing the establishment of a hierarchical social system, this persis-
tent stress can result in a progressive rise of systolic blood pressure over
a period of months (Fig. 29). There are, however, marked differences
between different strains of rats in their cardiovascular response to chronic
160 -
h
0 Long-Evans rats
I .+ . pairwise housed (n = 15) H*
E 4 . stable colony ( n = 14)
5 150 - 6 unstable colony (n= 14)
2
2
ln
u)
g 140 - *
0 1 2 3 4 5 6
Months after start of experiment
FIG. 29. Effects of housing conditions on the systolic blood pressure of male Long-Evans
rats. Significant differences to initial levels: *p < .OS; **p < .01; ***p < ,001 (see text for
further details). Adapted from Henry er af. (1995). with kind permission from Lippincott-
Raven Publishers, Philadelphia.
stress, which is found to correlate with their aggressiveness: The very aggres-
sive Long-Evans rats show a great increase of blood pressure, the less
aggressive Sprague-Dawley rat, a modest increase, and no change is ob-
served in the peaceable Wistar-Kyoto (hyperactive) strain (Henry ef al.,
1993).
The same relationships found between aggressive behavior and blood
pressure responses also seem to apply to individual differences within a
strain (Bohus et al., 1987; Fokkema, 1985, Fokkema and Koolhaas, 1985;
Fokkema et al., 1988; 1995). These authors tested the aggressive behavior
of male laboratory rats (strain TMD-S3) in several resident-intruder tests.
Following these precolony tests, 10 males together with 5 sterilized females
were transferred into a large colony cage, which was fitted with small boxes
in which the animals could find shelter. Cannulas were attached to most
males for intermittent direct blood pressure measurements and blood sam-
pling. In this seminatural situation, the levels of aggressive behavior in
individuals correlated with those levels determined in the precolony resi-
dent-intruder tests: The more aggressive in the precolony tests the more
competitive were the rats during confrontations in the colony, whether they
became dominant animals exhibiting offensive behavior, or subdominant
animals exhibiting defensive behavior or flight. However, blood pressure
as well as plasma corticosterone levels in dominant animals tended to be
lower than those in equally competitive subdominant animals.
In contrast to the actively competing dominant and subdominant animals,

lower blood pressure was observed in nonaggressive rats (“subordinates”)
as well as in formerly dominant animals that had lost their position after
severe defeat (“outcasts”) and made no further attempts to defend them-
selves. Barnett (1975) termed such frequently defeated males “omegas”;
behaviorally they correspond to submissive tree shrews: They are inactive,
socially withdrawn, and die after a few weeks without having suffered any
wounds. This was demonstrated by the study by Fokkema (1985), men-
tioned earlier, as well as in a life span study of laboratory rats under
seminatural conditions (Blanchard et al., 1988).
An even more dramatic decline in blood pressure in the course of chronic
conflict has been reported by Adams and Blizard (1987) in S/JR rats (a
salt-sensitive strain), which were repeatedly exposed to the presence of a
trained fighter rat (Long-Evans) and subjugated by it. Thus, chronic high
blood pressure is the result of continuous attempts of socially active animals
to adapt to an environment that is both threatening and demanding; on
the other hand, loss of control (as also seen in submissive tree shrews)
results in decreased blood pressure.
The relevance of the stability of a social position for the physiological
status of an animal is also evident from studies by Russian scientists on
hamadryas baboons (Papio hamadryas) and rhesus monkeys (Macaca mu-
latfa).The heart rates of these monkeys were recorded telemetrically using
a transmitter placed on the monkeys’ backs in the pocket of a jacket.
Dominant males of both species, kept in groups of 2 males and 1-2 females,
always had lower heart rates than the subordinates and these differences
could be reversed after experimentally induced changes of the social posi-
tions of the individuals. The higher heart rates in subordinate monkeys
were not related to increased locomotor activity, but, according to the
authors, reflected the degree of emotional tension (Cherkovich and Ta-
toyan, 1973).
Any challenge to the stable position of a dominant male results in dra-
matic cardiovascular stress responses. An example of this was shown in
dominant hamadryas baboons, which had lived for months with a harem
of several females and their young. If the dominant male was not allowed
access to his former group, which now lived with a rival male in an adjacent
enclosure, the former harem owner at first tried fiercely to attack the new
harem owner through the bars again and again. This behavior, however,
ceased after a few weeks. Nevertheless, over a period of several months,
hypertension, coronary insufficiency, myocardial infarction, and other so-
matic diseases developed, leading to the death of many former harem
owners (Lapin and Cherkovich, 1971).
Social stress as a determining factor in coronary artery disease has also

been implicated by Hamm and associates (1983) in Java monkeys (Macaca
fascicularis), which were kept in groups of 5 males for nearly 2 years.
Subordinate individuals in stable groups had significantly heavier adrenal
glands and more extensive coronary artery stenosis than did their dominant
counterparts. In repeatedly reorganized “unstable groups,” dominant males
developed greater blood pressure and arteriosclerosis of coronary arteries,
but this occurred only in the more aggressive and highly competitive individ-
uals, which retained dominant status over the whole study (Kaplan et al.,
1982, Manuck et af., 1983; Shively and Kaplan, 1984). Subsequent studies
have demonstrated that individuals of both sexes, exhibiting a heightened
cardiac response to a standard stressor (threat of capture), probably sympa-
thetic in origin, also develop the most extensive coronary lesions (Manuck
et al., 1986; 1989, 1995). This is consistent with observations in humans on
relationships between behavioral reactivity (“type A behavior”), sympa-
thetic arousal, and cardiovascular disease (Dembroski et al., 1983; Hous-
ton, 1992).
Similar results have also been found in field studies on monkeys. Domi-
nant male olive baboons (Pupio anubis) living in stable groups in the East
African savannah exhibited lower “initial” levels of cortisol (10 min after
darting), but responded relatively faster and more strongly following stress
due to anesthesia. In this way, differences between high- and low-ranking
males were compensated (Sapolsky, 1982), as with findings in rhesus mon-
keys (Sassenrath, 1970). Additionally, subordinate olive baboons were less
responsive to dexamethasone-induced cortisol suppression than were domi-
nant males, which was due to a selective decrease of glucocorticosteroid
receptors in the hippocampus (Brooke et al., 1994; Sapolsky, 1983, 1990).
Since dexamethasone resistance is a typical indicator for reactive depression
in humans, these results may indicate a similar state in animals with a
long history of social instability and lack of control. Finally, there were
indications of cardiovascular pathologies following prolonged periods of
subordination. Compared to dominant individuals, subordinate animals
exhibited significant reductions in high-density cholesterol, which can pro-
mote arteriosclerosis and coronary heart disease (Sapolsky and Mott, 1987).
In studies on male rhesus monkeys, placed in groups of four in large
cages for several months, Hamilton and Chaddock (1977) even demon-
strated death of apathetic (“submissive”) individuals after a rank order
had developed among the males. In contrast to all other males, the two
animals concerned neither battled for dominance nor did they flee from
attack. In general they crouched in the corner of the cage and, although
attacked on occasion, they were not grossly maltreated. They appeared
helpless in dealing with the situation and did not display the “fight or
flight” syndrome.
In New World monkeys, rank-related differences have also been demon-
strated. Dominant squirrel monkeys (Saimiri sciureus) in stable heterosex-
ual colonies as well as in newly formed groups have lower plasma cortisol
levels than subordinate individuals (Candland and Leshner, 1974; Manogue
et al., 1975). Furthermore, during group formation, concentrations of uri-
nary catecholamines increased only in the midranking individuals who suc-
cessfully fought to maintain their status, but decreased in those animals
who were unsuccessful and became further subordinated.
Similar endocrine findings have also been described for the African
talapoin (Miopifhecustalapoin) (Eberhart et al., 1983; 1985). Sandra Vel-
lucci (1990) manipulated the behavior of dominant and subordinate male
talapoin monkeys with drugs that are used in the treatment of human
psychiatric disorders, such as anxiety and depression. In order to maximize
the number of interactions between dominant and subordinate animals,
groups of males were allowed to interact daily for a period of 50 min with
females. This led to intense fights for control between dominant individuals,
while subordinate individuals retreated, huddled in corners, moved very
little, and showed high levels of visual monitoring. As her results indicate,
the behavior of dominant individuals is more susceptible to drugs that
are known to decrease levels of anxiety in humans, whereas subordinate
individuals appear more susceptible to treatment with antidepressant drugs.
This clearly demonstrates different emotional states in the individuals,
depending on their social position within this stressful situation.
Overall, these results indicate lowered adrenocortical and sympathetico-
adrenomedullary activities in dominant monkeys living in stable groups,
while in unstable situations, heightened activities of both stress systems are
present in high-ranking individuals actively trying to attain control and/
or dominance.
There are, however, contradictory results even in closely related species
of primates. Thus, Shively and Kaplan (1984) found that in Java monkeys,
dominant males in well-established mixed-sex groups exhibited higher
blood pressure and more advanced arteriosclerosis than subordinates, while
the latter had heavier adrenal glands, indicative of heightened adrenocorti-
cal activity. Furthermore, McGuire and associates (1986) failed to detect
a clear relationship between cortisol levels and dominance status in estab-
lished colonies of vervet monkeys (Cercopifhecus aethiops sabaeus), while
during competition for dominance, plasma cortisol increased in all males.
The same has been demonstrated for baboons living in natural conditions
(Alberts et al., 1992; Sapolsky, 1990).
In contrast to males, data on relationships between social status and

adrenocortical and adrenomedullary stress responses in females are largely
missing. As shown by Christian (1980) in an extensive review, subordinate
females in most rodent species are characterized by larger adrenal glands,
indicating an increased adrenocortical activity. This was also demonstrated
by Schuhr (1987) through direct measurement of plasma corticosterone in
female laboratory mice housed in groups. As mentioned previously, in our
study on wild rabbits, we found lower adrenocortical activities and heart
rates in high-ranking males and females, but only when stable group compo-
sitions and a sex ratio of about 1 male to 1-2 females prevailed.
In one of the few studies on female monkeys, Gust and colleagues (1993b)
examined the relationship between specific social behavior and serum corti-
sol concentrations in rhesus monkeys. The subjects were 9 females living
in an established long-term (8 years) mixed-sex group with their young,
while a second group of 9 females was formed 5 months prior to the onset
of the study and made up of animals initially unfamiliar to each other.
During the 1-year study, the rank of the females correlated significantly with
cortisol levels in the established group, with higher serum levels exhibited by
subordinate individuals. This was not the case in the recently formed group.
In addition, the authors demonstrated that cortisol levels were not only
negatively influenced by aggressive interactions, such as receiving bites, but
also positively influenced by sociopositive interactions, such as being
groomed.
In contrast to most studies, Creel and associates (1996) described higher
fecal glucocorticosteroid levels in dominant African wild dogs (Lycaon
pictus) of both sexes, as well as in samples of urine of dominant female
dwarf mongooses (Helogale parvula), both living under natural conditions
in the wild. Although measurements made on urine and feces samples
must be interpreted cautiously, these data indicate higher adrenocortical
activities in dominant females in both species, and also in males in African
wild dogs. Details on group composition and stability are, however, not
provided by the authors.
To summarize, group formation in primates as well as in other species
requires the establishment of a social structure. This process is typically
characterized by high levels of aggression, particularly among males, with
a return to baseline levels within a few days or weeks (e.g., Bernstein and
Mason, 1963). As demonstrated by the data given above, the process of
establishing a dominance hierarchy represents a potent psychosocial stres-
sor in all mammalian species, and usually affects lower ranking animals
more greatly. Social subordination and defeat in aggressive encounters
usually leads to increased adrenocortical activity and this relationship has
been found in both recently formed and established social groups. Further-
more, active coping with social subordination in stable social systems, or

active efforts to maintain a dominant rank in unstable groups, may eventu-
ally have health-impairing consequences when effects extend to the cardio-
vascular system. Such stress-related pathological conditions are evident
even in those primate species in which overt fighting and injury are infre-
quent in the maintenance of dominant-subordination hierarchies. It must
be emphasized once more that social hierarchies do not always result in
rank-dependent stress states. In some species, life in well-established social
systems is possible without any negative physiological stress effects on any
group members, as has been shown by Sachser (1994b) in guinea pigs, and
the same may apply to other species as well.
b. Gonadal System. As already mentioned in the earlier sections, the
effects of social stress on reproduction are profound and in extreme in-
stances can result in sterility in both sexes within a few days. In their
classic studies on small mammals housed in stationary and in freely growing
populations, Christian and many others have demonstrated that the various
endocrine responses of animals in crowded situations decrease natality at
every possible physiological level: Crowding inhibits growth and develop-
ment of reproductive organs in males and females; in addition to inhibiting
spermatogenesis in males, it inhibits estrus and ovulation in females and
it may delay or inhibit implantation. induce fetal reabsorption, and cause
damage to or the loss of litters. Furthermore, changes in the endocrine
state of stressed females may influence the physiology and behavior of
their progeny (Sachser and Kaiser, 1996). Thus, prenatal stress has been
associated with feminized sexual behavior in males and altered behavior
in females, as well as with various changes in exploratory behavior, cognitive
performance, and aggression. In 1958, Christian and LeMunyan described
the effects of crowding of pregnant female laboratory mice on two genera-
tions of their offspring! These results have been repeatedly confirmed in
recent years. Furthermore, after birth, reduced lactation and the retarded
growth of progeny may delay their maturation and increase their morbidity.
As these data have been reviewed in many excellent papers (e.g., Christian,
1978, 1980; Christian et af., 1965; Collaer and Hines, 1995; Krebs, 1978;
Krebs and Myers, 1974; Lee and McDonald, 1985; Ward, 1984), I shall deal
here only with some more recent results.
Dominant male sugar gliders (Petaurus breviceps), living in stable colo-
nies consisting of four males and one female, are heavier than socially
subordinate males, have significantly higher plasma testosterone and lower
cortisol levels, are more active, and are the only males that exhibit scent-
marking behavior. When transferred into a foreign stable colony, former
dominant males became subordinate and exhibited a reduction or loss of
behavioral measures associated with dominance and a concomitant de-
crease in plasma testosterone and rise in cortisol over a period of 3 weeks

(Mallick et al., 1994).
The same relationships between high plasma levels of testosterone and
dominance have also been observed in males of many other species (e.g.,
humans: Booth et al., 1989; Elias, 1981; Mazur and Lamb, 1980; McGrady,
1984; nonhuman primates: Alberts et al., 1992; Coe et al., 1979; Keverne et
al., 1982; Leshner and Candland, 1972; Mendoza et al., 1979; Rose et al.,
1971, 1974, 1975; Sapolsky, 1982, 1983, 1985a,b; Schiml et al., 1996; rats,
voles, and mice: review, Christian, 1980; guinea pig: Sachser, 1994a; Sachser
and Prove, 1986; tree shrews: von Holst, 1969; see also Fig. 17).
As plasma testorterone levels as well as the weights of testosterone-
dependent organs are usually correlated with dominance rank and some-
times also with the frequency of aggressive behavior in stable social systems
(e.g., monkeys: Alberts et al., 1992; Rose et al., 1971; laboratory rats: Kool-
haas et al., 1980; Monder et al., 1994), it is sometimes assumed that individu-
als with higher initial testosterone levels and therefore heightened levels
of aggression will gain dominant rank positions. This conclusion is, however,
not justified. Mendoza and associates (1979) housed male squirrel monkeys
either alone or in groups of three males with or without a female. While
prospective dominant males housed alone had the lowest plasma testoster-
one levels compared to the subordinate individuals, their testosterone levels
were highest in all-male groups, and this effect became even more pro-
nounced in the presence of females (Fig. 30).
Gonadal endocrine activity changes very quickly during dominance inter-
actions, as was shown already in 1973 by Bronson and associates in their
;ii 280 0Males alone

5m
-
Male groups
Male-female groups
_n
.
E
m
c
210
140
c
2
Q)
5 70
0
I
u)
$ 0
Rank 1 Rank 2 Rank 3
FIG. 30. Relationships between social rank, housing conditions, and plasma levels of testos-
terone in male squirrel monkeys; 3 males per rank. Increase of testosterone levels in dominant
and decrease in subordinate males in the different test situations significant a t p < .01. Adapted
from Mendoza er al. (1979), with kind permission from Elsevier Science Ltd, The Boulvard,
Langford Lane, Kidlington OX5 IGB. United Kingdom.
studies on laboratory mice (Bronson, 1973; Bronson and Marsden, 1973;

Bronson et al., 1973). The authors grouped 4 adult males per cage for
periods of time ranging from 1 hr to 14 days. During the first hour all males
fought intensively to establish dominance orders; at the same time, their
plasma levels of corticosterone increased by about a factor of five and their
gonadotropin levels decreased by about 20% for FSH and by more than
90% for LH. Plasma corticosterone levels returned to baseline levels be-
tween days 1 and 3 in dominants and between days 3 and 6 in the case of
subordinates. In dominant mice, the most conspicuous effect was the in-
crease in weight of their preputial glands, which produce an aggression-
provoking pheromone, while preputial glandular weight decreased in subor-
dinates by about 30% within 14 days. The same relationship between prepu-
tial glandular weight and rank has also been found in laboratory rats (Dijks-
tra et af.,1992). In this study the authors also demonstrated a strong increase
in testosterone plasma levels in individuals as a consequence of a successful
fight for a dominant position.
In their elegant studies on laboratory rats, Koolhaas and associates (1980)
followed the endocrine changes in males during and after confrontations,
by repeated blood sampling using cannulas inserted into blood vessels.
During the 1-hr encounters, plasma testosterone concentrations rose in
victors as well as in losers, but the rise was significantly greater in victors
than in losers. About 30 min after the start of the confrontation, plasma
testosterone concentration in both victors and losers started to decrease.
Victors regained their original baseline levels about 90 min after the end
of the confrontation, whereas testosterone levels in losers continued to
decline, reaching about 20% of initial levels 4 hr after the end of the
confrontation, and most defeated rats maintained these lowered baseline
levels for several days (Schuurman, 1981).
In summary, increased testosterone levels, such as are usually found in
high-ranking males, are the consequences rather than the cause of high rates
of aggression, as exogenous manipulations of testosterone concentrations
within the physiological range do not cause parallel changes in rates of
aggression or other testosterone-modulated behaviors (e.g., Booth et af.,
1989; Dixon, 1979 Mendoza ef al., 1979; Monaghan and Glickman, 1992;
Rose, 1985). On the other hand, loss of control as evident in subordinate
individuals or loss of a dominant status is associated with suppressed testos-
terone levels, and can even lead to sterility within a few days (see also
Rose 1985; Rose et al., 1972; 1974).
As is the case in the adrenocortical and sympathetico-adrenomedullary
systems, the gonadal endocrine system is not activated by the physical
exertions of successful fighting but by the emotional processes induced by
it. Accordingly, Mazur and Lamb (1980) have shown in human males that
only those that win a contest (leading them to perceive that their status is
thereby improved) show an elevation of testosterone levels. This is appar-
ently also the case in nonhuman mammals, as can be deduced from the
findings of our research on tree shrews (Kaiser, 1996).
In order to differentiate between the physical and psychological effects
of confrontations on dominant and subordinate male tree shrews, 2 males
confronted each other for 10 min daily over a period of 14 days, in an
experimental cage that could be divided into two identical subdivisions by
a wall. The confrontations always led to low-key fights, which resulted
right from the beginning in definite dominance relationships. Outside the
confrontation periods each animal was separated from its rival by a wooden
partition (“without visual contact”) or a wire mesh partition (“visual
contact”).
Outside the confrontation periods, dominant individuals in visual contact
with their rivals were less active and rested more compared to the days
before the confrontation period. In addition, their daily excretion rates of
cortisol decreased after the start of confrontations, while the excretion of
testosterone and the in vitro proliferation rate of their lymphocytes in-
creased (Fig. 31). By contrast, dominants without visual contact with their
rivals showed no changes in behavior or physiological parameters in com-
parison with initial values. Thus, only constant visual contact with the
subordinate opponent and the emotional process of “elation” probably
thereby induced modulated the behavior and physiology of dominant indi-
viduals.
As expected, subordinates in visual contact with their dominant rivals
showed opposite reactions to those of their opponents: a slightly increased
locomotor activity and urinary cortisol excretion, as well as a decrease in
testosterone excretion and the in vitro proliferation rates of their lympho-
cytes (Fig. 31).
Surprisingly, subordinate animals without visual contact with their rivals
showed qualitatively similar reactions to dominant animals in visual contact
with their subordinate rivals (Fig. 31). Thus, low-key fights during the daily
confrontations had no negative effects on the behavior or physiological
parameters of the subordinates. Our data even point to an improved physio-
logical state of these individuals, which may be due to the high level of
control and predictability which these animals perceive in this situation
(Fig. 31).
In female mammals, social subordination is associated with a diminished
number of ovulatory cycles and hence also with impaired reproductive
success (Dittus, 1979; Drickhammer, 1974; Sade et al., 1976; Silk etal., 1981;
Walker et al., 1983; Wilson et al., 1978; Wise et al., 1985). Most studies have
been carried out on rodents and, as they have been reviewed extensively,
Locomotory activity
0
-5
-10
-15
-20
Testosterone excretion with urine
30
15
0
-15
-30 ’ I I
Lymphocyte proliferation
I *
100
50
0
-50
-100
Visual contact Without visual contact
FIG.31. Effect of visual contact on locomotory behavior (medians), testosterone excretion

( M 2 SEM), and in vitro lymphocyte proliferation after Con A stimulation ( M 2 SEM) of
14 dominant and 14 subdominant male tree shrews. After 10 days of habituation to the
experimental room all animals were daily confronted for 10 min over a 2-week period; at
other times they were separated by either a wooden partition (without visual contact) or a
wire mesh partition (visual contact). Blood samples were taken 1day before the first confronta-
tion and 1 day after the last confrontation. Urine was collected over the whole period and
the individual means of each animal’s excretion rates over 8 days before the confrontation
were used as initial values. Locomotory behavior was determined daily for 3 hr. Significant
differences: *p < .05; **p < .01.
they will not be covered here (e.g., Christian, 1978, 1980; Christian et a!.,
1965; Krebs, 1978; Krebs and Myers, 1974; Lee and McDonald, 1985). The
same relationships between rank and reproductive success were demon-
strated in primates as well as in other mammalian groups. Among macaques
living in natural or seminatural environments, subordinate females are less

likely to become pregnant, and their pregnancies are more likely to result in
abortion, stillbirths, or neonatal death than are those of dominant females.
Furthermore, in free-ranging populations of rhesus monkeys at Cay0
Santiago and La Parguera and in wild populations of toque macaques,
subordinate female genealogies were found to exhibit a lower intrinsic
rate of natural increase than those of dominant females (Dittus, 1979;
Drickhammer, 1974; Sade et al., 1976). This was also demonstrated in groups
of adult female Java monkeys housed in harem groups consisting of one
adult male and five to six females (Adams et al., 1985). To induce social
instability and social disruption in three groups, the females were redistrib-
uted every 12 weeks for a period of 24 months (unstable groups), while
the remaining groups served as stable controls for the duration of the study.
Compared to socially dominant females, subordinate individuals had fewer
ovulatory menstrual cycles, more cycles with deficient luteal plasma proges-
terone concentrations, increased adrenal weights, and increased heart
weights. Social instability, however, influenced none of these variables.
These results indicate that impaired reproductive success observed in subor-
dinate female macaques may be related, at least in part, to changes in
ovarian function. The same relationship between rank and reproductive
success has been found in many studies on European wild rabbits under
seminatural conditions (e.g., Garson, 1979; Myers and Poole, 1962; Mykyto-
wycz, 1959a,b).
In our studies on wild rabbits we also found higher reproductive success
and individual fitness in females, depending on their rank at the time
of their insemination and pregnancy: Compared to subordinate females,
dominant individuals gave birth to more litters per year, the weight of the
young was higher at birth and at weaning, and mortality during the nest
period was lower (Fig. 32). This last feature is due mainly to decreased milk
production in females of subordinate ranks, which leads to the starvation of
their young. The lower number of litters produced by subordinate females
is apparently not due to sterility or delayed implantation, but results from
a high rate of resorption and abortion of entire litters during pregnancy,
as was verified by hormone analysis. As a consequence of this higher
reproductive success in dominant females, there are also rank-dependent
differences in the fitness of the individuals (Fig. 33).
A particularly interesting effect of dominance on reproductive success
has been demonstrated in dwarf mongooses. In these group-living carni-
vores, the oldest male and female dominate reproduction, while the younger
and subordinate group members are reproductively suppressed and provide
care for the offspring of the oldest pair. Nevertheless, subordinate males
from several wild populations in Tanzania exhibited urinary testosterone
Litters per female @/year)

4.5 1 I
3.6
2.7
Birth weight of young (9)
50 1
45
40
Mortality of young before weaning (%)
30 1I .
15
0
1 2 3 >3
Social rank of females
FIG.32. Relationships between social rank and reproduction of female wild European
rabbits living in a 22.000-m2 field enclosure. Data from about 50 females and 4 years; the
numbers of young are indicated in the bars of the bottom figure. Data in figure “Mortality”
are means; other figures means ? SEM. See text for further details. Unpublished data from
H. Draxler. 1996.
levels corresponding to those of dominants. They were, however, apparently

prevented from mating by dominant male aggression. In contrast, subordi-
nate females exhibit a decreased ovarian function (Creel et al., 1992). As
Keane and associates have shown in a subsequent paper (1994), subordi-
nates of both sexes mate and about 20% of all young had subordinate
mothers or fathers. Those subordinates that reproduced were of higher
rank than those that did not.
Among the primates, marmoset monkeys and tamarins demonstrate an
extreme form of rank-dependent fertility. In laboratory colonies, as well
Social rank and reproductive success of males and females in %

100
139 animals
in 7 years
75
50
25
0
1 2 3 1 >I
Rank of females Rank of males
FIG. 33. Relationships between social rank and reproductive success of male and female
European rabbits. Data are percentages of all young that survived until the reproductive
season following the year of their birth. The mothers of the litters were determined by
observations, and the fathers were determined by multilocus DNA fingerprinting. See text
for further details. Unpublished data: after Zobelein (1996).
as under natural conditions, only the socially dominant female of each

group reproduces, while ovulation in subordinate females is always sup-
pressed. This infertility is immediately reversed when subordinate females
are removed from their group and housed singly. As shown by Abbott
and associates (1988), the social suppression of fertility in the subordinate
females is apparently mediated by impaired hypothalamic GnRH secretion.
The most impressive example of socially induced contraception is known
from naked mole rats (Heterocephalus glaber), which live in colonies of up
to 300 animals entirely underground in the semiarid regions of East Africa.
In the wild as well as in captivity, there is only 1 breeding female, the
“queen,” and 1-2 breeding males in each colony, while all other animals
are infertile workers or play defensive roles within the colonies. Suppression
of reproduction in nonreproductive females appears to be induced by ovula-
tory failure due to insufficient gonadotropin secretion from the anterior
pituitary gland and the same suppression of gonadotropin release is also
evident in the nonreproductive (subordinate) males (Abbott et al., 1989;
Sherman et al., 1991).
It is probable that in all of the cases mentioned so far, neural responses
associated with psychosocial stress operate through the hypothalamo-
pituitary-gonadal axis, to induce ovarian dysfunction and subsequent infer-
tility or pregnancy failure, such as has been demonstrated in small mammals
(Christian, 1980). Support for this is provided by the findings of elevated
plasma prolactin concentrations and failure of the estrogen-induced LH
surge in socially subordinate ovariectomized talapoins (Bowman et al., 1978;

Keverne et al., 1982), and lower fecal estrogen levels found in subordinate
wild yellow baboons during the luteal phase of ovarian cycles, when com-
pared to females of high rank (Wasser, 1996). Furthermore, Packer and
associates (1995) found some of the strongest evidence for the advantage
of high rank in primates in their 30-year study of olive baboons (Papio
cynocephalus anubis) at the Gombe National Park: Dominant females had
shorter interbirth intervals, improved infant survival, and accelerated matu-
ration of their daughters. It must be mentioned in this context, however,
that the authors also reported negative effects on several aspects of female
reproductive success, which they interpreted as the reproductive cost of
dominance. However, as pointed out by Altmann and colleagues (1995),
this conclusion is not supported by their data and might have resulted from
inadequate interpretation of external signs of early pregnancy.
c. Zmmune System. Most earlier research into the relationship between
the social behavior of animals and their immune system and resistance to
disease stemmed from crowding experiments and was conducted mainly
on mice. Davis and Read (1958) conducted a series of experiments on the
influence of daily fighting on wild-stock house mice that were infected
parenterally with about 125 Trichinella larvae. Each mouse was housed in
a separate cage and from day 3 through 11 after infection half of the mice
were placed in groups of 5-6 animals for 3 hr a day, while the other half
were left separated. All mice were killed on the 15th day after infection.
About 25% of the singly housed mice were infected with an average of 9
worms apiece, whereas all grouped mice were infected and had an average
of 32 worms. Furthermore, severe and prolonged fighting among crowded
male mice impeded the development of acquired immunity to the dwarf
tapeworm and also increased the reinfection rate in mice with well-
established acquired immunity (Weinmann and Rothman, 1967). The ef-
fects were clearly rank dependent: Four days after a second dose of
tapeworm eggs (3500 eggdmouse) the dominant mice had an intestinal
cysticercoid count of 27, comparable to that in nonstressed mice exposed
to the same infection; however, the counts in subordinate individuals ranged
from 108 (rank 2) to 685 cysticercoids (lowest rank of the 8 males). In a
similar study, Tobach and Bloch (1958) demonstrated a significantly re-
duced survival time to an acute tuberculosis infection in socially stressed
mice (20 individuals per cage) in comparison to singly housed controls.
Furthermore, Edwards and Dean (1977) found that laboratory mice of both
sexes kept at high animal numbers (30 and 60 animals per cage) exhibited
reduced antibody production (against typhoid paratyphoid vaccine) and
reduced resistance to disease. This was evident from the significantly higher
mortality rate following an injection of Salmonella typhimurium, in compar-
ison to groups with lower animal numbers (2 or 10 animals per cage). The
inflammatory response to subcutaneous implants of cotton pellets moist-
ened with turpine as well as the formation of granulation tissue is also
reduced in grouped mice (Christian and Williamson, 1958). Finally, Temo-
shok and Peeke (1988) found differences in induced tumor growth in two
experiments on adult female Syrian hamsters placed in groups of ten:
Females ranked as dominant by the authors exhibited reduced tumor
growth compared to the subordinate individuals.
These results indicate an influence of social disturbances on disease
susceptibility due to immunomodulatory processes. One of the earliest
experimental proofs of this stems from research carried out by Vessey
(1964), who examined the antibody production against bovine serum in
male laboratory mice. Previously isolated mice were placed together in
groups of 6 each for 4 hr daily. They were injected with bovine serum 5
days after grouping and were found to have significantly lower titers of
circulating antibodies than isolated control mice.
Vessey (1964) also provided the first indication of rank-dependent immu-
nological changes: The winners of confrontations exhibited substantially
higher titers of antibodies than did the losers; likewise, T lymphocytes of
subordinate mice showed a distinctly reduced in v i m response to mitogenic
stimulation and reduced interleukin 2 production compared to their domi-
nant counterparts (Hardy et al., 1990). Correspondingly, Ebbesen and asso-
ciates (1991) found a lower incidence of virus-induced leukemia in dominant
mice compared with subordinates.
Similar suppressive effects of defeat or subordinate social rank on immu-
nological parameters have also been found in many other species. In one
of the earliest studies on rats, Raab and associates (1986) found higher
tyrosine hydroxylase activities in both dominants and subordinates com-
pared to individually or pair-housed rats (controls) after 10 days of chronic
cohabitation. Only subordinates, however, lost body weight and they exhib-
ited plasma corticosterone levels more than twice as high as those in domi-
nants and controls. In addition, they had smaller thymus glands and a
reduced lymphocyte response to in vitro mitogenic stimulation, while the
values of dominants did not differ from those of controls. In laboratory
rats housed in colonies, rank-dependent alterations in various components
of the cellular and humoral systems have also been demonstrated (Bohus
et al., 1992). Taken together, the results of these authors indicate an im-
provement of the immune system in dominant rats and to a lesser degree
also in actively coping subdominants, while most immunological parameters
in subordinates and outcasts are clearly suppressed.
In piglets housed in mixed-sex groups in large pens for 80 days, Hessing
and associates (1994) demonstrated clear relations between rank and sus-
ceptibility to disease and immune reactivity. Based on dominance fights

and a food competition test, piglets were divided into high-, middle-, and
low-ranking groups. Dominant individuals showed a higher in vitro lympho-
cyte response to an Aujeszky disease virus, less severe clinical signs of
disease, and threefold lower mortality rates compared to the individuals
of lower rank. Similar findings have also been described in farmed red deer
hinds (Hanlon et al., 1995).
Corresponding results have also been demonstrated for primates. Gust
and associates (1991) studied the immunological consequences of social
stress associated with the formation of a new group of 8 unfamiliar adult
female rhesus monkeys, introduced into an outdoor enclosure along with
1 adult male. The establishment of a stable dominance hierarchy, apparent
within 48 hr, was accomplished without serious fighting and in a complete
absence of wounding. While humoral components of the immune system
(IgG, IgA, IgM) were not significantly influenced over the period of colony
formation, within 24 hr all females generally showed a significant increase
in cortisol plasma levels and a 30% decrease in absolute numbers of total
lymphocytes as well as CD4+ and CD8+ T cells. These changes were
significantly greater in the 4 lowest ranking females compared to those with
higher ranks. After 1 week the T cell subsets of the high ranking females
had returned to initial levels or exhibited even higher levels; however,
values of the low-ranking females returned more slowly to baseline levels
and were still low 9 weeks after group formation. This was in spite of the
fact that there were no significant differences in aggressive (offensive or
defensive) or affiliative behaviors between the two groups, with the excep-
tion of grooming: High-ranking subjects were groomed significantly longer
than the subordinates. Recent studies by Gust et al. (1996) on female pigtail
macaques (Macaca nernestrina) confirmed these results. Furthermore, high-
ranking males in small stable groups of male rhesus monkeys exhibited
significantly higher lymphocyte proliferation than middle- or low-ranking
individuals. Regrouping of the animals led to an increase in aggressive
behavior and plasma cortisol levels and a decrease in the lymphocyte prolif-
eration response to a mitogen (Clarke et aZ., 1996). The same effects have
been found in male Java monkeys, with particularly strong immunosuppres-
sive effects among those monkeys showing high levels of fear behavior
(Line et al., 1996).
In a field study, Alberts and colleagues (1992) found significantly lower
lymphocyte counts and a higher basal cortisol concentration in an adult
male that had entered a stable group of olive baboons, as well as in those
individuals that were victims of the intruder’s aggression, than in nonin-
volved individuals. The same suppressive effects of crowding stress, re-
peated regrouping or defeat, and social subordination on resistance against
parasites, bacterial and viral diseases, tumor growth, as well as on humoral

and cellular immunological parameters have also been found in many other
studies, conducted mainly on mice, rats, and rabbits (e.g., 1994; Brayton
and Brain, 1974a,b; Edwards and Dean, 1977; Edwards et al., 1980; Fleshner
et al., 1989; Hardy et al., 1990; Hoffman-Goetz et al., 1991; Mykytowycz,
1961; Plaut et al., 1969; Stefanski and Ben-Eliyahu, 1996; Stefanski and
Hendrichs, 1996; Stefanski et al., 1996; Tecoma and Huey, 1985; for reviews,
see also Ader and Cohen, 1985; Monjan, 1981; Plaut and Friedman, 1981;
Riley, 1981).
Amazingly, the odor from stressed laboratory mice alone can induce an
altered immune function in conspecifics after 24 hr of odor exposure, and
lead to a decrease in the number of cells forming antibodies to sheep red
blood cells (Zalcman et al., 1991), as well as to a decrease in production
of interleukin 2 by Con A-stimulated spleen cells, and decreased activity
of natural killer cells (Cocke et al., 1993). In contrast to the observed
suppression in cell-mediated responses, stress-odor exposed mice had an
enhanced humoral immune response to KLH. Thus, even in a given strain,
stressors do not necessarily affect all immune measures unidirectionally,
which cautions against premature conclusions based on a limited selection
of cellular or humoral immune parameters.
The general conclusion can be drawn from these and many other studies,
that “stress” as determined by adrenocortical activation can increase sus-
ceptibility to infectious diseases. However, there are exceptions to this
generalization (e.g., Moynihan et al., 1994).
In their study on male and female laboratory rats that were submitted
for 4 weeks to different forms of regrouping, Klein and associates (1992)
found clear indications of heightened adrenocortical activity (adrenal en-
largement and increased basal corticosterone levels) and thymus involution.
However, compared to undisturbed controls, neither natural killer cell
activity, splenocyte reactivity to mitogens, nor the rate of spontaneous
development of antibodies against a common pathogen of the respiratory
tract of mice (Mycoplasma pulmonis) were changed in the stressed animals.
This could be due to genetic differences in immunological responsiveness,
as was demonstrated in 1955 by Tobach and Bloch. These authors used
strains of rats and mice that varied in degree of “emotionality,” and found
that the most “emotional” strains had the shortest mean survival times
after a standard dose of intravenously administered tuberculosis bacilli.
Similarly, Friedman and Glasgow (1973) found, depending on mouse strain,
that grouped laboratory mice are more susceptible to Plasmodium berghei
than individually housed animals.
Likewise, Fauman (1987) demonstrated that, relative to subordinate ani-
mals and isolated controls, dominant laboratory mice have a reduced anti-
body response to an antigen (keyhole limpet hemocyanin). The extent of

the reduction of the antibody response in dominant mice was related to
the intensity of their aggressive behavior during confrontations. Similar
results have also been found in male laboratory rats (Bohus et al., 1993)
as well as in female chimpanzees (Pun trogfodytes) living in captive colonies
(Masataka et al., 1990).
Furthermore, according to Hausfater and Watson (1976), high-ranking
individuals in groups of free-living yellow baboons (Papio cynocephafus)
of both sexes exhibited a higher fecal parasite ova emission than more
subordinate individuals. Subadult individuals generally occupy lower ranks
and have lower egg counts than older ones. However, examination of the
mean egg counts in adult individuals only continued to show a correlation
between egg output and dominance rank in males, but not in adult females.
There are many possible reasons for these and other contradictions. In
relation to investigations carried out on laboratory animals, some of these
contradictions may be due to differences between the various strains. They
could also, however, be due to differences in housing conditions and in the
duration and type of stress. Very often the necessary information needed
to assess this question is missing, as are appropriate control groups. In
addition, the detailed observations required to supply reliable information
on social ranking and stress levels in individuals have often not been carried
out (see also Bohus and Koolhaas, 1991). Furthermore, data on the activity
of the sympathetico-adrenomedullary system, which could indicate the pres-
ence of social tension, are completely lacking. Finally, all of these studies
are based on a rather limited selection of immune parameters, which makes
general conclusions on the function of the immune system impossible. In
spite of these shortcomings, all investigations do indicate a strong response
of immune parameters to socially stressful situations.
d. Physiological Costs of Male Dominance. In general, engaging in social
conflict exposes individuals to the risk of injury and attacks by predators,
diverts precious energy from reproductive activities and feeding opportuni-
ties, and may enhance vulnerability to disease. In the long term these costs
are weighed against potential benefits for the dominant individual of ready
access to mates with high reproductive success (e.g., Huntingford and
Turner, 1987; Maynard Smith and Price, 1973; Riechert, 1988). In fact, one
of the most prominent views on subordinate animals is that they have less
access to mates and consequently leave fewer offspring than do dominant
animals, an idea that was advanced by Zuckerman (1932) and Maslow
(1936) in the 1930s for primates. This concept is widely accepted today and
its validity has been demonstrated for many species in the wild as well as in
captivity (e.g., Ellis, 1995; Miczek eta!., 1991). Several studies in laboratory
conditions have also shown that females, when given a choice, tend to
associate and mate with dominant males (e.g., lemmings: Huck and Banks,
1982; rats: Carr et al., 1982; bank voles: Hoffmeyer, 1982; Shapiro and
Dewsbury, 1986; hamsters: Brown et al., 1988; White, 1986; vervet monkeys:
Keddy, 1986).
The relationship between social status and susceptibility to disease is
currently of great interest among some evolutionary biologists, as parasite
burdens may influence several aspects of social and sexual behavior (e.g.,
Barnett and Sanford, 1982; Dobson and Hudson, 1986; Edwards, 1988;
Edwards and Barnard, 1987; Freeland, 1981; Kavaliers and Colwell, 1995;
Moore and Gotelli, 1990; Rau, 1983; Read, 1990; Toft and Karter, 1990;
Wedekind, 1994). In 1982, Hamilton and Zuk proposed that because of
the genetically based interactions between parasites and their hosts, females
are expected to choose mates based on their resistance to pathogens. Male
secondary sex characters or ornaments were supposed to have evolved at
least in part as indicators of this resistance. According to the authors,
females should prefer males with fewer parasites, an indication of which
is given by the degree of the development of secondary sex characters.
Many, though not all, tests designed to prove this hypothesis have been sup-
portive.
On the basis of the higher susceptibility of human males compared with
females to a variety of bacterial, viral, and parasitic diseases, Marlene Zuk
proposed in 1994 that high-ranking males are more vulnerable to diseases:
“The deleterious effects of testosterone may be an unavoidable price paid
by males for achieving reproductive success in a competitive environment.”
This hypothesis, however, seems rather unlikely, at least for mammals. As
shown in the previous sections, it is absolutely possible that under conditions
of social instability, dominant individuals fighting actively for control may
develop cardiovascular diseases that may shorten their life. Their immuno-
logical resistance and therefore their resistance to bacterial, viral, and para-
sitic diseases is, however, usually higher than that of subordinates, especially
in stable social situations.
Furthermore, the hypothesis is based on rather doubtful premises. The
author writes: “Social dominance has also been demonstrated to be testos-
terone dependent, with experimental castration generally reducing aggres-
sion and subsequent testosterone injections usually causing its return.”
Although these effects of castration and subsequent testosterone replace-
ment have been demonstrated in several mammalian species, the results
are of limited relevance to the hypothesis put forward by Zuk.
Behavioral endocrinology has shown that aggressive behavior in male
mammals is predominantly determined by genetic influences, which appar-
ently modify prenatally through testosterone those central nervous struc-
tures involved in the expression of aggressive behavior (e.g., de Ruiter et
al., 1993). Postnatally, these modifications are based on experience, so long

as certain minimal testosterone levels are present, and this is probably the
case in most mammals in natural conditions (e.g., Monaghan and Glickman,
1992; Sachser et al., 1994; Scott and Fredericson, 1951). Testosterone levels
are indeed usually higher in dominant males, but it must be emphasized
again that these increased testosterone levels are the consequence, rather
than the cause of the high-ranking positions. There are, at least to my
knowledge, no conclusive published data indicating that intact males of a
given species (or strain) with higher initial testosterone levels are more
aggressive and successful in confrontations than rivals with lower levels.
Rather, the opposite seems to be true (see also Figs. 17 and 30) and,
accordingly, it is not possible to increase aggression or other testosterone-
modulated behaviors in male mammals with normal serum androgen levels
by means of testosterone treatment (e.g., Clarke et al., 1996; Leshner, 1981;
Monaghan and Glickman, 1992; van Oortmerssen et al., 1987; Rose et al.,
1972; 1975).
Additionally, Zuk states in this context: “No one could be more stressed
than the males of many vertebrate species during the mating season, when
courtship displays are exhausting, the environment must be constantly scru-
tinized for competitors and those competitors fought off. . . .” Although
this statement applies to most species, a decrease in the response of immune
parameters during the mating season has not been demonstrated so far,
apart from in the highly stressed marsupial, Antechinus, mentioned in this
chapter’s introduction (for a recent review, see also Nelson and Demas,
1996). In our wild rabbits there is definite evidence of an improved immuno-
logical state during the mating season compared to the nonmating season
(Fig. 34).
The same objections have to be raised to the postulation by Zuk of a
relationship between testosterone levels of fertile males and their immuno-
logical resistance. In general, increased testosterone levels during fetal life
as well as in adult males after puberty are thought to reduce cell-mediated
immunological resistance, although it must be pointed out that these conclu-
sions are based on studies on very few laboratory animal species. Further-
more, there is considerable controversy concerning the effects of sex hor-
mones on antibody formation and unspecific biological resistance (e.g.,
Grossman, 1984; Madden and Felten, 1995; McCruden and Stimson, 1991;
Olsen and Kovacs, 1996; Schuurs and Verheul, 1990). Nevertheless, castra-
tion of adult males of those species examined so far does at least lead to
increases in their cellular immune resistance compared to that of fertile
males.
However, Zuk’s conclusion that this relationship also applies to fertile
individuals is not supported by the literature. In contrast, most data pre-
Testosterone Lymphocyte proliferation immunoglobulin G

0.5 1 I (ng/ml 36
32
20
24
20
Reproductive season Nonreproductive season
FIG. 34. Serum testosterone levels (40 males) and immunological parameters (117 males)
of wild European rabbits living in a 22,000-m2 field enclosure. Means ( 2 SEM) from the
reproductive period (April-September) and the nonreproductive period (October-March).
Differences between the seasons were always significant at p < ,001.
sented in the previous section indicate an improved immunological resis-

tance for dominant males, along with increased testosterone plasma levels.
Preliminary data from our laboratory collected on European rabbits, tree
shrews, and Long-Evans laboratory rats also contradict this hypothesis:
Injection of fertile males with physiological doses of testosterone over a
period of 2 weeks had no recognizable immunosuppressive effects in any
of these 3 species; on the contrary it even increased several cellular immune
parameters in rabbits that were kept under constant laboratory conditions.
e. Summary. In a stable dominance hierarchy, the dominant individuals
can predict and actively control the outcome of social interactions, they
have priority of access to food, mates, and other resources. On the whole,
this situation increases the fertility and health of dominant individuals,
while the opposite is usually true for subordinate individuals. This endocri-
nological advantage of a dominant social position may, however, be very
small or even nonexistent, depending on the social system and the species
(Table 111).
The contrasting situation of instability occurs in the wild, when new
animals migrate into a social group and destabilize the status quo, or when
individuals die. In captivity, such instability is evident when social groups
are first formed. In this case, the situation is very different for dominants
when compared to stable systems. Typically, the rates of aggressive interac-
tions are elevated, and are focused on animals in high-ranking positions.
Rank shifts may occur repeatedly and unexpectedly. This is a situation that
TABLE I11
RELATIONSHIP STATUS,
BBI'WEEN SOCIAL COPING
STYLE,
A N D PHYSIOLOGICAL
RESPONSE
PATTERN
Dominant Subordinate
Control of situation high low, rank low loss of control

or position threatened
Coping style active (fights usually active active (defensive/ passive and
not necessary) (offensive) offensive) apathetic
Pituitary- unchanged or slightly slightly elevated markedly
adrenocortical decreased elevated elevated
function
Sympathico- unchanged markedly greatly elevated unchanged or
adrenomedullary elevated even
function reduced
Pituitary-gonadal elevated elevated reduced markedly
function reduced
Immune function unchanged or often ? reduced greatly
improved reduced
exhibits anything but control and predictability. All individuals, especially

the top-ranking individuals, are therefore experiencing stress (Table 111).
These relationships between control and predictability of a social situa-
tion and the physiological response pattern of an individual may explain
most of the discrepancies between studies on social rank, physiology, and
stress-related diseases in different mammalian species. It must be empha-
sized again that, in order to gain reliable information on the physiological
state of an individual, it is not sufficient to focus on one stress system only,
as is the case in most studies so far conducted, but information has to be
collected at least on pituitary-adrenocortical as well as on sympathetico-
adrenomedullary systems. Furthermore, conclusions about the influence of
social situations on the immune status of an individual must be based on
a large number of different immune parameters, to avoid the premature
conclusion that a situation is without immunological effects.
3. Disruption of Social Bonds

a. Introduction. Dominance has been used in the preceding text in a very
general way, as a shorthand term that indicates the outcome of agonistic or
competitive interactions between two individuals. Dominance relationships
are usually more pronounced in males and are of overwhelming importance
to all aspects of the life of group-living mammals-they influence their
behavior, reproductive success, and health.
Dominance hierarchy has often tacitly been assumed to be an equivalent

term to social organization. However, dominance hierarchies based on
agonistic behavior are only one aspect of social systems. Although much
less conspicuous, social bonds, usually based on attachment between indi-
viduals, are at least as necessary for the establishment and stability of social
systems as are dominance relationships. Attachment between mothers and
their infants is usually a precondition for the survival of infants in mammals,
and later social bonds to peers and mates, as well as to adults of the same
sex, may develop, which profoundly influence the behavior of the animals.
Such bonds, in terms of social support, can also play a positive role in health
by presumably altering the way in which a potentially stressful situation is
perceived (House et al., 1988; Levine, 1993a; Unden et al., 1991). It is,
therefore, not surprising that the loss of a social bond and/or lack of social
support may result in strong stress responses and an increased risk of
mortality, as is indicated by numerous epidemiological studies in humans
(e.g., Berkman and Syme, 1979; Broadhead et al., 1983; Dyer et al., 1980;
Gilman et al., 1982; House et al., 1982; Kannel et al., 1987; Perrson et
al., 1994; Schoenbach et al., 1986). Although social relationships play an
important role in most mammalian societies, research into the relevance
of social bonds and the stress-buffering effects of sociopositive interactions
are usually neglected in stress research on nonhuman mammals.
b. Mother-Infant Bond. Mammalian young are born defenseless and
are highly dependent on their mothers for a relatively long period of time.
Therefore, they have to learn to bond to their mothers, which is essential
for their nurture and social development and has been dramatically demon-
strated in rhesus monkeys in Harlow’s classical studies (e.g., Harlow and
Suomi, 1974; Harlow et al., 1971; Rosenblum and Plimpton, 1981; Suomi,
1976). The mother-infant bond is crucial for the infant to learn to overcome
fear of novel stimuli and to control aggression in social settings in later life.
Maternal separation from infants is one of the most profound stressors
for monkeys, and usually results in the death of the young in the wild (e.g.,
Thierry et al., 1984). In the laboratory, it has been used as an animal
model for separation and depression. Extreme passive stress responses
characterized by increased excretion of urinary 17-hydroxycorticosteroids
and plasma cortisol levels, as well as strong immunomodulatory responses
occur in infant rhesus and squirrel monkeys in response to separation from
their mothers (e.g., Coe and Scheffler, 1989; Hrdina and Henry, 1981;
Levine, 1993b; Levine et al., 1985; Wiener et al., 1992). Infants (age around
30 weeks) of macaques also showed behavioral changes and depressed in
vitro lymphocyte proliferative responses to T-cell-specific mitogens over
the 14-day separation period, while response to a B-cell-selective mitogen
was not significantly affected. In addition, decreased natural killer cell
activity and significant alterations of different lymphocyte populations oc-

curred (Lubach et al., 1996). Following reunion, both behavior and immuno-
logical parameters returned to initial values. Maternal responses to separa-
tion were usually similar to those of their infants (Laudenslager et al., 1982;
Reite et al., 1981).
The behavioral responses of infants to separation from their mothers
differ among species, even when they are as closely related as the pigtail
and bonnet macaques. Both species exhibit an initial agitation phase, char-
acterized by distress vocalizations, high levels of locomotion, and other
behavioral attempts by the infant to relocate and reestablish contact with
its mother. Only pigtail macaques, however, exhibit a second depressive
phase in this response (Boccia et al., 1995). This difference is apparently
due to the different social environments of the young of these two species.
Because bonnet macaques exhibit lower levels of aggression and higher
levels of social contact than pigtail macaques, mothers are less restrictive
and permit their infants to freely interact with other group members. As
a consequence, when separated from their mothers, these infants are
adopted by one of the other females.
Boccia and collegues (1995) tested the effect of this different socialization
directly. They examined the behavior of two infant pigtail macaques, who
grew up in an environment of elevated aggression induced by a feeding
paradigm, which allowed the individuals to feed only one after another.
As a consequence, the high-ranking mother always had free access to the
food, while the subordinate mother became restricted. This situation had
striking effects on the social relationships of the infants of these two moth-
ers. The infant of the unrestricted dominant female exhibited close attach-
ments to four other group members, representing over 80% of the social
interactions, whereas the infant of the subordinate mother restricted his
social interactions to the mother. When the two infants were separated
from their mothers, the second infant without alternative attachments, but
not the first, became profoundly depressed and spent nearly 50% of its time
during the separation exhibiting a depressive slouched posture (Boccia et
al., 1991). Furthermore, the authors showed that social support from older
peers can be protective. In a social group containing six infants, three had
significant attachments with three older juveniles in the group, while the
others were attached only to their mothers. The authors removed all moth-
ers and juveniles from the group except for the three previously identified
juveniles. Thus, out of the six remaining infants, only three retained social
relationships with the juveniles with whom they had already had relation-
ships prior to the separation. As a measure of social support, the authors
took the number of affiliative behaviors directed by the juveniles to each
infant. This measure demonstrated the strong protective effects of social
support after mother-infant separation. Behaviorally, infants with social

support showed less evidence of depression, as was reflected in play and
eating behaviors. They also exhibited no change from baseline function in
natural killer cells, while infants without support showed a 40% decrease
from baseline function 2 hr after separation.
Similar studies have been performed on only a few nonprimate species,
such as the laboratory rat. Removal of mothers of laboratory rats at an
age of 2 weeks markedly decreased the heart rates of their infants to about
60% of the normal rate during the following 2 days, which was followed
by leveling off and recovery during the next few days (e.g., Hofer, 1981,
1994). In addition, increased plasma corticosterone baseline levels and
adrenal responsivity to acute stressors were evident even several days after
a single 24-hr period of maternal deprivation (Rosenfeld et al., 1992; Taka-
hashi, 1991).
The social bonds between mothers and their infants have also been
evaluated in guinea pigs. Guinea pigs are capable of coordinated locomotion
almost immediately following birth. Maternal care is minimal. When infants
at an age of about 2 weeks were separated from their mothers for 30 min
and transferred to an unfamiliar cage in an unfamiliar room, they exhibited
high rates of vocalization and almost doubled plasma cortisol levels. The
presence of their mothers reduced vocalization and plasma cortisol levels
significantly, while the presence of an unfamiliar lactating female produced
no effect over a period of hours (Hennessy and Ritchey, 1987).
c. Bonds between Juveniles and Adult Individuals. The separation from
peers can also result in strong physiological responses. The removal of
squirrel monkeys from their companions resulted in a strong decrease of
the lymphocyte proliferation to the mitogen Con A. The decrease reached
significance within the first day, was maximal on the second day, and re-
turned to initial levels within 7 days. As in common marmosets (Johnson
et al., 1996) and most other primate species, plasma cortisol levels in squirrel
monkeys peaked during the first day after separation, but took much longer
to return to baseline levels than did the immune parameter (Coe, 1993;
Friedman et al., 1991;Levine et al., 1989). The independence of adrenocorti-
cal activation and immune responses has also been demonstrated in juvenile
rhesus monkeys, which were removed from their natal social group to peer
housing at the age of 2 years. The highest plasma cortisol levels and greatest
decrease of total blood lymphocytes and several T cell subsets (CD4+ and
CD8+) were observed on the first day. While their adrenocortical activities
returned to baseline levels within about 2 weeks, immune measures re-
mained decreased for up to 2 months (Gust ef al., 1992). In contrast to
juveniles, adult male rhesus monkeys showed no stress response to separa-
tion from their group (Gust et al., 1993a), while separation of adult individu-
als from their group in rats and sheep causes a strong activation of the
pituitary-adrenocortical system (cited from Toates, 1987).
Ratcliffe and associates (1969) studied the psychological response of
swine to separation after the social bonds of grouped animals had been
established. Swine housed pairwise or in groups, responded to human visi-
tors with grunts and squeals for a handout. Competition among the males
was very low and was limited to pushing and shoving. By contrast, separated
swine, especially the normally sociable females, failed to respond to visitors,
lying unresponsive and refusing offers of added food. After a year of
isolation the separated females showed a significantly greater development
of arteriosclerosis than those that were grouped. These data suggest that
the lack of social bonds may result in sustained emotional disturbance and
pathophysiological changes.
Kaplan and associates (1991) examined the relationship between the
aggressive and affiliative behavior and cellular immune parameters in adult
male Java monkeys living in small groups, whose members were periodically
redistributed over months. While the authors did not find any influence of
social status on the immune parameters, the in vitro lymphocyte prolifera-
tion in reaction to the two T-cell-selective mitogens concanavalin A (Con
A) and phytohemagglutinin (PHA) was greatest in individuals that were
both highly affiliative and exhibited low levels of aggression. Furthermore,
natural killer cell activity was highest among highly affiliative males, regard-
less of their levels of aggression. These findings indicate that the cellular
immune competence may be enhanced among monkeys that, in response
to a disrupted social environment, spend large amounts of time in affiliation
with other males, or in males that seek and find social support.
4. Social Support and Its Stress-Reducing Effects
As these findings show, numerous factors influence the magnitude of the
physiological stress response, and one of the most important variables in
pairwise or group-living mammals appears to be the presence of a familiar
social partner. Social support generally reduces the magnitude of stress
responses and it has a stress-buffering effect, as was impressively demon-
strated by Levine and associates in their studies on squirrel monkeys
(Levine, 1993a,b). The authors exposed a well-established group of adult
squirrel monkeys to a live Boa constrictor that was confined in a plastic
box. Although direct physical contact between the monkeys and the snake
was prevented, all monkeys showed increased levels of vigilance, agitation,
and avoidance behavior. A strong adrenocortical activation, however, was
observed only when the monkeys were tested individually, but not when
tested together as a group. Surprisingly, this stress-buffering effect appears
only when adult squirrel monkeys are exposed to this situation together
with multiple partners. In pair-housed individuals no social buffering was

evident, although the behavioral signs of arousal were reduced (Coe et al.,
1982). In this context, the findings of Mendoza and Mason (1986) are of
special interest. They compared the effects of intruders on behavior and
adrenocortical activities of polygynous squirrel monkeys and monogamous
titi monkeys (Cullicebus rnoloch), housed as heterosexual pairs. In titi
monkeys, the presence of an intruder resulted in marked behavioral signs
of agitation, especially in the subjects of the same sex as the intruders.
Plasma cortisol levels of females showed no consistent changes to intruders
of either sex, while those of males were always increased in the presence
of a male rival. Squirrel monkeys of both sexes, on the other hand, re-
sponded to female intruders with a reduction in plasma cortisol to below
baseline levels, whereas a male intruder had no effect. Maintenance of a
monogamous social structure, such as in titi monkeys, is presumably based
on a bond between the male and female of the pair and the exclusion of
male rivals by the male. In squirrel monkeys, which usually live in large
groups of both sexes, life as a pair is lacking in the usual companionship
and could be improved by new individuals. Although this interpretation is
not without contradiction, these results nevertheless demonstrate impres-
sively that the social system of a species influences the behavioral and
physiological responses of the individuals to conspecifics in very different
ways.
The relevance of the quality of the relationships between individuals on
their stress-reducing effects are also evident from guinea pigs (Sachser et
ul., 1998). In mixed-sex colonies male guinea pigs develop long-lasting and
strong bonds to some females, while no such social ties exist to other
females. When male guinea pigs are taken from such colonies and placed
singly into unfamiliar cages their plasma cortisol levels increase for hours
by about 100%compared to initial levels. Presence of an unfamiliar female
from a different colony or a familiar female from their own colony to which
no social bonds exist has no stress-reducing effects. There is, however, a
sharp reduction in the endocrine stress response when each male is trans-
ferred into an unfamiliar cage together with a female with whom a social
bond exists (Fig. 35).
The relevance of social integration and social bonds to members of a
group is especially evident in our work on wild rabbits. Throughout the
reproductive phase, each female usually produces 5-6 litters at monthly
intervals, resulting in up to 30 progeny per year. Depending on the number
of adult females in our enclosure, up to 1000 animals are born each year.
However, about 70% of the young are taken by predators (e.g., cats, mar-
tens, weasels, hawks) before the onset of the winter season and only an
average of 5% of the original number actually survive the winter. The
6 225
m
c
c
200
s
- 175
0
u)
._
r
8 150
E
$ 125
h
100
familiar bonded unfamiliar
Females
FIG. 35. Cortisol values ( M -C SEM) of 10 male guinea pigs 2 hr after transfer into an
unfamiliar enclosure alone (dotted horizontal line), together with an unfamiliar female, a
familiar but unbonded female, or a bonded female. All data are percentages of initial levels.
Significant differences between the effects of the presence of bonded and unbonded conspe-
cifics: *p < .05; **p < .01. Adapted from Sachser et al. (1997).
number of surviving juveniles varies from one year to the next (from 0 to
more than 40 individuals) and is approximately equivalent to the number
of adults that have died. Consequently, the number of adult rabbits at the
beginning of each reproductive season has remained surprisingly constant
over the past 10 years (at 50-95 individuals).
Predation plays only a minor role in mortality during the winter months
(November to February). Death is usually due to an extreme loss of weight,
based on the breakdown of all fat reserves as well as large quantities of
muscle tissue, culminating in hypoglycemic shock. Although these findings
point to starvation, a general lack of food cannot be the reason for death,
as all the adults as well as those juveniles that survive the winter show no
loss in body weight. Rather, the moribund juveniles are incapacitated, in
spite of increased food intake, by extensive parasitic damage to their intesti-
nal epithelium, which prevents the digestion and/or resorption of food. In
addition, toxins produced by the changed intestinal flora probably also
contribute to the death of the animals. Within the last few weeks prior to
death, the number of oocysts and nematode ova in the feces of the moribund
juveniles increases dramatically and parallel to the loss in weight (Fig.
36). In comparison to the surviving individuals, this parasitic infestation is
probably due to a reduced immune resistance against the parasites, as
indicated by a reduced in v i m lymphocyte proliferation (Fig. 37), a de-
Body weight Nematodes Coccidia
1200
900
600 -
6 2 0 6 2 0 6 2 0
Weeks before death
FIG. 36. Changes of body weight and numbers of nematode eggs (predominantly Tricho-
strongylus retorfaefornzins and Gruphidium strigosum) and oocysts of several Elimera species
in the feces of 20 subadult European wild rabbits during the last weeks before their death
in the winter period. All data are means ( 2 SEM). See text for details.
creased number of T lymphocytes in the blood, and a 50% reduction in

the phagocytic capacity of the leucocytes.
Based on our current findings, mortality during the winter months ap-
pears to be a result of socially induced immune suppression: Young animals
usually leave their native groups in autumn and attempt to join other groups
(Kunkele and von Holst, 1996). In this process, all immigrants are initially
attacked and chased away by members of the group. However, some juve-
niles are tolerated after a while and integrated into the group, although
most do not achieve this social integration. The successful integration of a
juvenile is indicated by its spatial position within the group and by its
behavior toward the adults: Integrated animals restrict their whereabouts
more or less exclusively to the existing territory of a group of adults, while
nonintegrated animals tend to roam over a wide area and from one group
to the next. In addition, integrated animals are observed either in close
proximity to or in direct contact with individuals of a group during 30% of
observation time, while this is seldom the case in nonintegrated individuals
(Fig. 38: Spatial integration). Although aggressive reactions by adults are
directed against integrated and nonintegrated juveniles with almost equal
frequency, integrated animals are more often involved in friendly interac-
tions with adults. While in integrated animals two out of three interactions
with adults are of a friendly nature, nonintegrated animals are only in-
volved in one sociopositive interaction for every three aggressive ones (Fig.
38: Social integration). Finally, both groups also differ significantly from
each other in immunological measures: nonintegrated individuals clearly
exhibit lower values than integrated animals (Fig. 38: Immune measures).
Body weight Food intake LP after Con A
6.0
4.0
2.0
0.0
Nematodes Coccidia Intestinal villi
450
800 300
400 150
0 0
Surv Died Surv Died Surv Died
FIG. 37. Body weight, parasites in feces, and length of the intestinal villi of 20 subadult
wild rabbits at their death during the winter (Died) as well as their food intake and in vitro
lymphocyte proliferation (LP) after Con A stimulation 2-6 weeks before their death. All
measures were also determined at corresponding times from 20 animals of about the same
age that survived the winter period (Surv). All data are means (rf- SEM); significant differences:
**p < .01; * * * y < ,001.
The change in integration state was followed during the winter season
in several juveniles: Individuals that were first more or less integrated within
a group were expelled from it, and nonintegrated juveniles were accepted.
In all cases this also involved changes in immunological parameters: If
integration status deteriorated, then lymphocyte proliferation was reduced;
if integration status improved, that is, in the case of successful integration
into a group, proliferation increased (Fig. 39).
Based on these findings, an improved immune state and a reduced para-
sitic infestation in juveniles surviving the winter would appear to be the
result of successful integration into the existing social group. Accordingly,
out of more than 100 animals observed in detail over 5 years, only those
animals capable of successful integration into groups during the autumn
and winter months actually survived the winter.
The number of juvenile wild rabbits is therefore regulated during the
winter season, by giving only those individuals that have achieved integra-
Local attachment Being attacked Cellular (LP)

Interactions per hour
4.8
3.6
2.4
1.2
0.0
Distance to adults Sociopositive behavior Humoral (IgG)
4o 1 Sights < 2m in %
4.8
lnteractmns per hour
I 160
Deviation In %of the
mean of all subadults
30 3.6 120
20 2.4 80
10 1.2 40
0 0.0 n
IN NI IN NI IN NI
Spatial integration Social integration Immune measures
FIG. 38. Spatial integration. social integration, and immune measures of about 20 integrated
( I N ) and 30 nonintegrated (NI) subadult wild rabbits during the winter period ( M 2 SEM).
Immunological parameters: means of 1-3 measurements per animal; behavioral data: means
of 8-24 hr of observations per animal. Significant differences between IN and NI: **p < .01;
***p < ,001. See text for further details. Unpublished data from M. Kaschei (1996).
tion into an existing group a chance of surviving the winter. As the accep-
tance of juveniles into an existing group of adults is apparently dependent
on the size and composition of the group, this mechanism results in optimal
group composition before the onset of the reproductive season.
As mentioned previously, numerous epidemiological studies on humans
indicate that social bonds, in terms of social support, can play a positive
role in the health of an individual. The direct physiological mechanisms
are, however, far from being clear. As shown by the various studies on
nonhuman mammals described above, a breakdown of social bonds elicits
strong passive stress responses, while the presence of a bonded partner or
group has some stress-buffering effects. Furthermore, the development of
a bond can exert strong physiological consequences even in individuals that
apparently beforehand had lived an unstressed life.
Thus, tree shrews can be housed singly for more than 10 years in captivity
and be in excellent condition without any apparent signs of stress. Neverthe-
Change of social integration

-
E 400
J
m
U
2 300
-m
v)
’is 200
c
m
100
a,
5
$ 0
s 1 2 3 4
5 6 7 0
Number of animal
FIG. 39. Changes of the in virro lymphocyte proliferation after Con A stimulation of 8
subadult wild rabbits before and within 4 weeks after change of their social integration.
Unpublished data from M. Kaschei (1996).
less, the formation of a pair bond greatly improves their well-being, as

indicated by physiological data. Tree shrews usually live in pairs in the
wild. Putting a male and a female together, however, does not inevitably
lead to the formation of a pair bond. In some instances it can result in
intensive fights and-unless the animals are separated-in the death of
one of the opponents (male or female). In most cases, however, especially
in large enclosures, tree shrews of both sexes can coexist, although they
suffer from a certain amount of social tension, as evident from occasional
fights and avoidance behavior. At estrus, successful copulations may even
occur, but the offspring are always cannibalized by the parents shortly after
birth (von Holst, 1969). In all these unharmonious pairings, even if overt
aggression is not evident, the heart rates of the animals are constantly
increased, as is the case in subdominant males living together with a domi-
nant male under constant active stress (Fig. 40). In about 20% of all pairings,
however, contact between an unfamiliar male and female is characterized
from the outset by amicable behavior, which conveys the strong impression
of “love at first sight.” Both individuals “greet” each other frequently with
long bouts of mouth licking (up to 90 min per 12-hr observation day), they
move around in close contact, and mostly rest together. Copulations may
occur on the first day, but are not a necessary prerequisite for such a
harmonious pair bond. During the nights both animals always sleep together
in the same nest box, which is never the case in the previously mentioned
unharmonious pairs. In the laboratory harmonious pairs can live together
for more than 10 years and breed successfully and regularly in the absence
of any aggression.
In all harmonious pairs we found a drastic reduction in serum levels
of glucocorticosteroids and adrenocortical reactivity to standard stressors,
and-even more surprising-a reduction in heart rates (Fig. 40). Further-
more, all immunological parameters that were measured indicate an im-
provement of the immunological state of both individuals. The opposite is
true for unharmonious pairings. Amazingly, the quality of a pairing depends
on personal “sympathy” or “antipathy” between the individuals. Thus, a
male that has been fiercely rejected by one female can be accepted as a
“loved” partner by another female. Accordingly, the physiological status
of tree shrews kept as pairs changes depending on the quality of their pair
bond, as shown in studies in which females were paired with different
males. In standard tests females respond to males that they will accept as
partners with high marking responses, and to those that they will not accept
with low marking responses. Hence it was possible to pair females once
with males that they accepted as partners, and once with males that they
rejected (Fig. 41). As the results of these pairings demonstrate, both sexes
exhibited low levels of aggression and high levels of sociopositive behaviors
when females were combined with males to whose scent they had shown
the highest marking responses (harmonious pairs). The opposite was the
case when the females were unharmoniously paired with males whose
scent stimulated their marking behavior very little (Fig. 41). Furthermore,
harmonious pairings decreased serum levels of glucocorticosteroids and
epinephrine, while increasing those of gonadal hormones as well as improv-
ing cellular and humoral immune measures. The opposite was true in the
same individuals in unharmonious pairings (Fig. 42). Unfortunately, little
is known of the physiological effects of pair formation in other species,
with the exception of several studies on monogamous and polygamous
species of vole.
In the prairie vole (Microtus ochrogaster), long-term heterosexual pair
bonds are formed, which are characterized by affiliative behaviors, such as
side-by-side contact, and are independent of sexual behavior (Carter et al.,
1988, 1995; Winslow et al., 1993). In contrast to tree shrews, however,
prolonged mating of naive females with an unfamiliar male is necessary
for the induction of a pair bond in this species (Insel et al., 1995). The
-1 disharmonious pairing + single
500 I 1
0,
m
il
single harmonious pairing -
5 300
f
100
0 5 10 15
Days after beginning of pairing experiment
FIG.40. Effects of an unharmonious and a harmonious pairing on the heart rate of a female tree shrew. Night periods are striped.
Data higher and lower (unharmonious and harmonious pairing, respectively) than the mean of the last 3 days before the pairings are
accentuated by black. Adapted from von Holst (1987). with kind permission from Gustav Fischer Verlag, Stuttgart, Germany.
Marking behavior Sociopositive behavior Defensive behavior
90
u)
- (u
E
C 60 6.0
’c
0
b
.-
>
m
$ 30 3.0
m
0 - 0.0
Harm Unharm Harm Unharm Harm Unharm
FIG. 41. The marking activities of females in response to the scent of different males were
used to create harmonious (Harm) and unharmonious (Unharm) pairings (for details, see
von Holst, 1985b). The more that the scent of a male stimulates the marking activity of a
female the greater is the probability that the pairing with the female will result in a harmonious
pair bond. Each female was therefore paired for 14 days with that male whose scent elicited
the highest, and after 4 weeks of single housing with that whose scent elicited the lowest
marking response (“Marking behavior”). During the pairings the behavior of each male and
female was recorded for a total of 12 hr (“Sociopositive behavior” and “Defensive behavior”).
All data are means ( 2 SEM); significant differences are indicated: ***p < .001.
Tes Cor Nor Epi LTT IgG
FIG.42. The effects of harmonious (striped bars) and unharmonious (cross-hatched bars)
pairings on several physiological measures of 12 male tree shrews. All data ( M ? SEM) are
given as deviations from the initial levels of the males before the pairings. Abbreviations:
serum levels of testosterone (Tes), cortisol (Cor), norepinephrine (Nor), epinephrine (Epi),
and immunglobulin G (IgG); in vitro lymphocyte proliferation after Con A stimulation (LTT).
Significant differences are indicated: ***p < ,001. See text for further details.
results of many studies indicate that centrally released oxytocin during

mating may be critical to the formation of partner preferences in female
prairie voles, while vasopressin appears to be more important to pair bond-
ing in the males of this species (e.g., Carter er al., 1992; Insel and Hulihan,
1995; Williams et al., 1992,1994). Furthermore, plasma glucocorticosteroid
levels may respond to and influence the development of social attachments.
In naive female prairie voles, cohabitation with a male resulted in a dramatic
decline in serum corticosterone levels, which facilitated pair bonding. When
corticosterone levels were reduced via adrenalectomy, females developed
partner preferences after 1 hr of cohabitation, while sham-operated and
untreated females required 3 hr or more of cohabitation to establish partner
preferences (De Vries et al., 1995).
The role of oxytocin in the development of social bonds was first
proposed by Klopfer (1971), who suggested that the increased oxytocin
levels after birth facilitate the mother-infant attachment. In subsequent
studies, the relevance of centrally released oxytocin for the development
of maternal behavior and the development of mother-infant bonds was
shown in sheep, rats, and some other species (e.g., Da-Costa et al., 1996;
Kendrick et al., 1987; Pedersen and Prange, 1985; Uvnas-Moberg, 1994;
Yu et al., 1996).
Furthermore, many studies demonstrated the relevance of oxytocin for
sexual behavior (in addition to sex hormones). Thus, injections of oxytocin
in estrous rats stimulates sexual behavior in female rats, reduces aggression,
and increases physical contact with the males (Arletti and Bertolini, 1985;
Caldwell et al., 1986); similar results were also found in female Syrian
hamsters (Whitman and Albers, 1995). These data indicate that oxytocin
may be involved in the formation of social bonds between mothers and
their infants as well as between males and females in mammals (Carter et
al., 1990; Keverne, 1988). Shared sexual experience and the concomitant
oxytocin release usually found in both sexes may thereby facilitate social
bonds in mammals including human beings.
In addition, stress-buffering effects of intracerebroventricular injections
of oxytocin have been described. In laboratory rats, the development
of gastric lesions induced by cold and restraint stress or by the administra-
tion of cysteamine was reduced by oxytocin treatment (Grassi and
Drago, 1993). The physiological mechanisms involved are unsolved, but
nevertheless, these results indicate that social bonds may improve the
health of individuals by reducing their response to stressors (probably
by reducing the stress-inducing properties of stressors, due to the presence
of a security-providing partner), as well as by influencing the physiological
state of the individual.
C. CONCLUSIONS
Social relationships based on agonistic and sociopositive behaviors play
an important role in most mammalian societies. They determine not only
the stability of social systems but greatly influence almost all behavioral
elements of the individuals as well as their fertility and health. Disturbances
of social relationships may lead to stress responses that differ greatly in
quality and intensity, depending on the stressor and the coping behavior
of the individuals.
Disruption of social bonds usually initially elicits an alarm response
characterized by heightened physiological and behavioral arousal. Particu-
larly in infants separated from their mothers, this is followed by passive
stress responses characterized by apathetic behavior, withdrawal, and even-
tually death. Information on the long-term consequences of disruption of
social bonds for the health of adults is, however, lacking for nonhuman
mammals. In human beings, the loss of partners can have very strong health-
impairing effects.
Social conflict elicits an immediate acute alarm response in all animals,
characterized by increased sympathetico-adrenomedullary and pituitary-
adrenocortical activation. There is evidence that norepinephrine (the fight
hormone) predominates in this first response to challenges, which is proba-
bly characterized by the feeling of anger. If the stressful situation cannot
be resolved by behavioral responses (e.g., by fight or flight), differing chronic
stress responses may result, the degree of which is dependent on the percep-
tion of the amount of control over a social situation, and which are therefore
almost exclusively psychological phenomena. The perception that loss of
control is either possible or probable appears to lead to a change from
anger to fear, as is indicated by an increasing production of epinephrine (the
flight hormone) and mainly active subordinate behavior. As the threatening
situation continues, this active coping can shift to a more passive, apathetic
mode, accompanied by greatly increased adrenocortical activity and associ-
ated with the feelings of helplessness and depression. The adrenomedullary
epinephrine release can remain high or decrease by comparison to actively
coping individuals. Gonadal activity (at least in males) may actually increase
in early phases of successful responses to challenge (during anger), but
eventually declines as loss of control threatens. The immune system re-
sponds extremely sensitively to social challenges. Every challenge to control
(feelings of anger, fear, and depression) is usually accompanied by profound
indications of immunosuppression.
The relationships between social rank and stress response depend mostly
on the stability and predictibility of the social relationships. In stable so-
cial systems the dominants are usually not, or only slightly, stressed: Com-
pared to subordinate individuals, they exhibit lower adrenocortical- and

sympathetico-adrenomedullary activities as well as higher gonadal activities
and immune resistance. Subordinate individuals usually show active stress
responses, but in some cases also passive stress responses, which usually
lead to death within a short period of time. In socially unstable systems,
which are characterized by immigration processes or dominance conflicts,
all individuals usually show active stress responses of varying degrees of
strength as they fight to regain high control and predictability. The effects
on dominants and subdominants, however, may differ depending on the
species and the social situation. In many species, the highest active stress
responses are found in those individuals that are dominant and fight actively
to maintain their high social ranks. This may even lead to higher incidences
of cardiovascular diseases and premature death, in comparison to subordi-
nate individuals.
The neuroendocrine stress responses accompanying these subjective feel-
ings have a bipolar aspect: According to a concept proposed by Henry in
1986, the anger-fear (fight-flight) response is opposed by the serenity-
relaxation state, which is characterized by enhanced grooming and resting.
The opposite pole to the depression, loss-of-control, and loss-of-attachment
axis, is probably a subjective feeling of elation, such as in dominant tree
shrews in the presence of clear subordinate individuals-and probably also
in animals with strong bonds to partners (Fig. 43).
Because social relationships can influence the physiological state of indi-
viduals in so many positive or negative ways, it is not surprising that social
status alone cannot always predict stress-related measures in individuals,
particularly in natural conditions containing many uncontrollable social
influences. This means that, in order to understand the physiological conse-
quences of social interactions, an integrated approach is required to assess
what factors, including rank and social bonds, interact to affect an individu-
al’s fertility and health (see also Fig. 44). O r as Sapolsky (1988) puts it in
his discussion of individual differences in olive baboons and their stress re-
sponses:
Thus, among these primates, who you are. what your place is in your society, and what
sort of society it is appear to have everything to do with your physiology, both under
basal and stressed circumstances. Furthermore, one may argue at this stage that these
rank-related differences in physiology are of consequence, that the pattern observed in
dominant males seems to be the most adaptive.
And these comments on baboons apply to all mammalian species, from

rodents to humans, as pointed out as long ago as 1977 by Henry and Ste-
phens.
FIG.43. Schematic diagram of the stress-buffering emotional processes and their physiolog-
ical consequences. Adapted from Henry (1986). with kind permission from Academic Press,
Inc., New York.
IV. SUMMARY
Contact with conspecifics not only influences the behavior of individuals,

but is also associated with marked physiological changes, which can influ-
ence their vitality and fertility in positive or negative ways, depending on
the type of interaction. The term used to describe the negative effects is
+pq dominant
Acute stress
Chronic
passive stress
FIG. 44. Schematic diagram of the two stress axes (PAS, pituitary-adrenocortical system,
and SAS, sympathetico-adrenomedullarysystem) and their activation (+) or inactivation (-)
depending on the social control perceived by the individuals and the associated subjective
emotions. As an example, the physiological states of tree shrews in the differing social situation
described in this paper are given in circles. See text for further details.
social stress. In this chapter, the many negative physiological consequences

of social stress are addressed and the stress-reducing effects of sociopositive
contacts with conspecifics are described.
The changes that have taken place over the past 30 years in the concept
of stress are an important prerequisite to understanding the effects of social
interactions on the physiology of individuals. As the current concept of
stress has been developed within the boundaries of psychology and medi-
cine, largely excluding the field of zoology, a short synopsis of this develop-
ment is given. According to this concept, physiological stress reactions are
generally triggered by central nervous processes (emotions or feelings),
which always occur when a situation is characterized by uncertainty or
unpredictability, that is, when the individual’s control over the situation
is endangered or impossible. Differing physiological stress reactions are
induced, depending on the behavioral strategy used by an animal to either
obtain control or to cope with the situation. Active attempts at obtaining
control over a situation (e.g., fight or flight) are characterized initially by
activation of the sympathetico-adrenomedullary system and in the long

term by cardiovascular disease (active stress). Passive perception of defeat
or loss of control is characterized initially by pituitary-adrenocortical activa-
tion and in the long term by negative effects on almost all bodily functions
(passive stress).
In a second section, such peripheral physiological processes are described
as are essential to the understanding of stress reactions. Since various stress
reactions differ depending on the situation and the coping behavior of
the individuals, studies based on one or only a few measures can lead to
misleading or false conclusions. An introduction is also given into the most
frequently used methods in obtaining indications of the activities in the
pituitary-adrenocortical and sympathetico-adrenomedullary systems, as
well as in gonadal and immune functions. Particular attention is paid to
the limitations of these methods.
In a third section, based on our research on tree shrews, an overview of
the relationships between the social position of an animal and its physiologi-
cal state is given. General statements cannot be made due to the close
relationships between social system, social rank and rank stability, and the
impact of positive relationships with other conspecifics. However, in stable
social systems, an overall dominant position can improve the fertility and
vitality of the individual, while subordinate individuals exhibit no or only
slight active or passive stress reactions, dependent on the species. In unsta-
ble social systems, dominant individuals are characterized by particularly
high active stress reactions, due to their efforts of improving or retaining
their position through increased levels of aggression; subordinate animals
exhibit active or passive stress reactions of variable intensities. Finally, the
importance of social bonds for the health of individuals is assessed: The loss
of social bonds can provoke long-term stress reactions, while the presence of
bonded partners has stress-reducing effects.
Because social relationships can influence the physiological status of
individuals in so many positive or negative ways, social status alone cannot
always predict stress-related measures in individuals, especially in natural
conditions containing many uncontrollable social influences. Consequently,
in order to understand the physiological consequences of social interactions,
an integrated approach is required to assess which factors, including rank
and social bonds, interact to affect an individual’s fertility and health.
Acknowledgments
This chapter is dedicated to my late friend James Henry whose scientific results and concepts
were twenty years ahead of his time. I wish to thank Norbert Sachser for his critical comments,
which greatly improved this manuscript. Thanks are also due to Debby Curtis for her most
valuable help with the English version on this paper. 1 also appreciate the editors’ helpful
remarks on the manuscript.
References
Abbott, D. H.. Hodges, J. K., and George, L. M. (1988). Social status controls LH secretion and
ovulation in female marmoset monkeys (Callifhrixjacchus). J. Endocrinof. 117,329-339.
Abbott, D. H., Barrett, J., Faulkes, C. G . , and George, L. M. (1989). Social contraception in
naked mole-rats and marmoset monkeys. J. Zool. 219,703-710.
Ackerman, K. D., Bellinger, D. L., Felten, S. Y., and Felten, D. L. (1991). Ontogeny and
senescence of noradrenergic innervation of the rodent thymus and spleen. In “Psychoneu-
roimmunology” (R. Ader. D. L. Felten, and N. Cohen, eds.). pp. 71-125. Academic
Press, San Diego, CA.
Adams, D. 0. (1994). Molecular biology of macrophage activation: A pathway whereby
psychosocial factors can potentially affect health. Psychosom. Med. 56, 316-327.
Adams, M. R.. Kaplan, J. R.. and Koritnik, D. R. (1985). Psychosocial influences on ovarian
endocrine and ovulatory function in Macaca fascicularis. Physiol. Behav. 35, 935-940.
Adams. N., and Blizard, D. A. (1987). Defeat and cardiovascular response. Psychol. Rec.
37,349-368.
Ader, R., and Cohen, N. (1985). CNS-immune system interactions: Conditioning phenomena.
Behav. Brain Sci. 8, 379-394.
Ader, R.,Felten, D. L., and Cohen, N., eds. (1991). “Psychoneuroimmunology ” Academic
Press, New York.
Adler, R., Herrmann, J. M., Kohle, K., Schonecke, 0. W., von Uexkull. T., and Wesiak, W.,
eds. (1986). “Psychosomatische Medizin.” Urban & Schwarzenberg, Munich.
Aguilera, G., Kiss, A,, Hauger, R., and Tizabi, Y. (1992). Regulation of the hypothalamic-
pituitary-adrenal axis during stress: Role of neuropeptides and neurotransmitters. In
“Stress: Neuroendocrine and Molecular Approaches” (R. Kvetnansky, R. McCarty, and
J. Axelrod, eds.), pp. 365-381. Gordon & Breach, New York.
Alberts, S. C., Sapolsky, R. M.. and Altmann, J. (1992). Behavioral, endocrine, and immunolog-
ical correlates of immigration by an aggressive male into a natural primate group. Horm.
Behav. 26, 167-178.
Altmann, J., Sapolsky, R., and Licht, P. (1995). Baboon fertility and social status. Nature
(London) 377,688-689.
Anisman, H., Zalcman, S., and Zacharko, R. M. (1993). The impact of stressors on immune and
central neurotransmitter activity: Bidirectional communication. Rev. Neurosci. 4,147-180.
Arletti, R., and Bertolini, A. (1985). Oxytocin stimulates lordosis behavior in female rats.
Neuropeprides (Edinburgh) 6, 247-253.
Arnold, A. S., and Breedlove, S. M. (1985). Organizational and activational effects of sex
steroids on brain and behavior: A reanalysis. Horm. Behav. 19, 469-498.
Aus der Muhlen. K., and Ockenfels, H. (1969). Morphologische Veranderungen im Diencepha-
Ion und Telencephalon nach Storungen des Regelkreises Adenohypophyse-Nebennieren-
rinde. 111. Ergebnisse beim Meerschweinchen nach Verabreichung von Cortison und
Hydrocortison. Z. Zeflforsch. Mikrosk. Anat. 91, 126-138.
Barnett, S. A. (1958). Physiological effects of “social stress” in wild rats: I. The adrenal cortex.
J. Psychosom. Res. 3, 1-I 1.
Barnett, S. A. (1964). Social stress. Viewpoints Biol. 3, 170-218.
Barnett, S. A. (1975). “The Rat: A Study in Behavior.” University of Chicago Press, Chicago.
Barnett, S. A. (1988). Enigmatic death due to “social stress.” A problem in the strategy of
research. ISR, Inrerdiscip. Sci. Rev. 13, 40-51.
Barnett, S. A., and Sanford, M. H. R. (1982). Decrements in “social stress” among wild Rattus
rattus treated with antibiotic. Physiol. Behav. 28, 483-487.
Barnett. S. A.. Hocking, W. E.. Munro, K. M. H., and Walker, K. Z . (1975). Socially induced
renal pathology of captive wild rats (Ratrus villosissirnirs).Aggressive Behav. 1, 123-133.
Bateman, A., Singh. A., Krai, T., and Solomon, S. (1989).The immune-hypothalamic-pituitary-
adrenal axis. Endocr. Rev. 10, 92- 112.
Batzli. G. 0..and Pitelka, F. A. (1971). Influence of meadow mouse populations on California
grassland. Ecology 51, 1027-1039.
Baum. M. J. (1992). Neuroendocrinology of sexual behavior in the male. In “Behavioral
Endocrinology” (J. B. Becker, S. M. Breedlove, and D. Crews, eds.). pp. 97-130. MIT
Press, Cambridge, MA.
Beach. F. A. (1975). Hormonal modification of sexually dimorphic behavior. fsychoneuroendo-
crinology 1, 3-23.
Beato, M., and Doenecke. D. (1980). Metabolic effects and modes of action of glucocorticoids.
In “General, Comparative and Clinical Endocrinology of the Adrenal Cortex” (1. Chester
Jones and 1. W. Henderson, eds.), pp. 117-181. Academic Press, London.
Beerda, B.. Schilder, M. B. H., Janssen. N. S. C. R. M., and Mol, J. A. (1996). The use of
salivary cortisol, and catecholamine measurements for a noninvasive assessment of stress
responses in dogs. Horm. Behav. 30, 272-279.
Berkenbosch. F., de Rijk, R.. Schotanus, K., Wolvers, D., and van Dam, A. (1992). The
immune-hypothalamo-pituitary adrenal axis: Its role in immunoregulation and tolerance
to self-antigens. In “Interleukin-1 in the Brain. Pergamon Studies in Neuroscience” (N.
Rothwell and R. Dantzer, eds.), Vol. 5 , pp. 75-91. Pergamon, Oxford.
Berkman. L. F., and Syme, S. L. (1979). Social networks, host resistance and mortality:
A nine-year follow-up study of the Almameda County residents. A m . J . Epidemiol.
109, 186-204.
Bernard, C. (1859). “Leqon sur les ProprietCs Physiologiques et les AltCrations Pathologiques
des Liquides de I’Organisme,” Vol. 1. Baillibre. Paris.
Bernstein, I. S., and Mason, W. A. (1963). Group formation by rhesus monkeys. Anim. Behav.
11, 28-31.
Bernton, E. W., Bryant, H. U., and Holaday, J. W. (1991). Prolactin and immune function.
In “Psychoneuroimmunology” (R. Ader. D. L. Felten, and N. Cohen, eds.). pp. 403-428.
Academic Press, San Diego, CA.
Besedovsky, H. O., and del Rey, A. (1991). Physiological implications of the immune-
neuro-endocrine network. In “Psychoneuroimmunology” (R. Ader. D. L. Felten, and N.
Cohen, eds.). pp. 589-608. Academic Press, San Diego, CA.
Blalock, J. E. (1988). Immunologically-mediated pituitary-adrenal activation. A h . Exp. Med.
B i d . 245, 217-223.
Blanchard, D. C., Sakai. R. R., McEwen. B.. Weiss. S. M., and Blanchard. R. J. (1993).
Subordination stress: Behavioral. brain, and neuroendocrine correlates. Behav. Brain
Rex 58, 113-121.
Blanchard, R. J., Flannelly. K. J.. and Blanchard, D. C. (1988). Life span studies of dominance
and aggression in established colonies of laboratory rats. Physiol. Behav. 43, 1-7.
Boccia, M. L., Reite. M. L. and Laudenslager, M. L. (1991). Early social environment may
alter the development of attachment and social support: Two case reports. Infanf Behav.
Dev. 14,253-260.
Boccia. M. L., Laudenslager, M. L., and Reite, M. L. (1995). Individual differences in macac-
ques’ responses to stressors based on social and physiological factors: Implications for
primate welfare and research outcomes. Lab. Anim. 29, 250-257.
Bohus. B., and Koolhaas, J. M. (1991). Psychoimmunology of social factors in rodents and
other subprimate vertebrates. In “Psychoneuroimmunology” (R. Ader, D. L. Felten. N.
Cohen, eds.), pp. 807-830. Academic Press, San Diego, CA.
Bohus, B., Benus, R. F., Fokkema, D. S., Koolhaas, J . M., Nyakas, C., van Oortmerssen,
G. A., Prins, A. J . A.. de Ruiter, A. J. H., Scheuring, A. J . W., and Steffens. A. B. (1987).
Neuroendocrine states and behavioral and physiological stress responses. Prog. Brain
Res. 72, 57-70.
Bohus, B.. Koolhaas, M.. de Ruiter, A. J. H., and Heijnen, C. J. (1992). Psycho-social stress.
differential alterations in immune system functions and tumor growth. In “Stress: Neuro-
endocrine and Molecular Approaches” (R. Kvetnansky, R. McCarty, and J. Axelrod,
eds.), pp. 607-621. Gordon & Breach, New York.
Bohus, B., Koolhaas. J. M., Heijnen, C. J., and de Boer, 0. (1993). Immunological responses
to social stress: Dependence on social environment and coping ab es. Neuropsychobiol-
ogy 28, 95-99.
Booth, A., Shelley, G., Mazur, A,, Tharp, G.. and Kittok, R. (1989). Testosterone, and winning
and losing in human competition. Horm. Behuv. 23,556-571.
Bowman, L. A., Dilley. S. R., and Keverne. E. B. (1978). Suppression of oestrogen-induced
LH-surge by social subordination in talapoin monkeys. Nature (London) 275, 56-58.
Bradley, A. J., McDonald, I. R., and Lee, A. K. (1980). Stress and mortality in a small
marsupial (Anfechinusstuarfii, Mackuy). Gen. Comp. Endocrinol. 40, 188-200.
Brain, P. F., and Poole, A. E. (1974). The role of endocrines in isolation-induced intermale
fighting in albino laboratory mice. 1. Pituitary-adrenocortical influences. Aggressive Behav.
1,39-69.
Brambell, F. R. (1944). The reproduction of the wild rabbit Orycfolugus cuniculus (L.). Proc.
Zool. Soc. London 114, 1-45.
Brayton, A. R., and Brain, P. F. (1974a). Studies on the effects of differential housing on
some measures of disease resistance in male and female laboratory mice. J . Endocrinol.
61,48-49.
Brayton, A. R., and Brain, P. F. (1974b). Effects of differential housing and glucocorticoid
administration on immune responses to sheep red blood cells in albino “TO” strain mice.
J . Endocrinol. 64, 4-5.
Breedlove, M. (1994). Sexual differentiation of the human nervous system. Annu. Rev. Psychol.
45,389-418.
Broadhead. W. E., Kaplan. B. H., James, S. A,. Wagner, E. H.. Schoenbach, V. J., Crimson,
R., Heyden, S.. Tibblin. G., and Gehlbach. S. H. (1983). The epidemiological evidence
for a relationship between social support and health. Am. J. Epidemiol. 117,521-537.
Bronson, F. H. (1963). Some correlates of interaction rate in natural populations of wood-
chucks. Ecology 44,637-643.
Bronson, F. H. (1964). Agonistic behaviour in woodchucks. Anim. Behav. 12,470-478.
Bronson, F. H. (1973). Establishment of social rank among grouped mice: Relative effects
on circulating FSH, LH, and corticosterone. Physiol. Behav. 10, 947-951.
Bronson, F. H., and Eleftheriou, B. E. (1965a). Relative effects of fighting on bound and
unbound corticosterone in mice. Proc. Soc. Exp. Biol. Med. 118, 146-149.
Bronson, F. H.. and ElefthCriou, B. E. (1965b). Adrenal responses to fighting in mice: Separa-
tion of physical and psychological causes. Science 147, 627-628.
Bronson, F. H., and Marsden, H. M. (1973). The preputial gland as an indicator of social
dominance in male mice. Behuv. Biol. 9, 625-628.
Bronson, F. H., Stetson, M. H.. and Stiff, M. E. (1973). Serum FSH and LH in male mice
following aggressive and nonaggressive interactions. Physiol. Behav. 10, 369-372.
Brooke. S. M.. de Haas-Johnson, A. M., Kaplan, J. R.. Manuck, S. B.. and Sapolsky, R. M.
(1994). Dexamethasone resistance among nonhuman primates associated with a selective
decrease of glucocorticoid receptors in the hippocampus and a history of social instability.
Neuroendocrinology 60, 134-140.
Brown, M. R. (1991). Brain peptide regulation of autonomic nervous and neuroendocrine
functions. In “Stress, Neurobiology and Neuroendocrinology” (M. R. Brown, G . F. Koob.
and C. Rivier. eds.), pp. 193-215. Dekker, New York.
Brown, P. S., Humm, R. D., and Fischer, R. B. (1988). The influence of a males’s dominance
status on female choice in Syrian hamsters. Horm. Behav. 22, 143-149.
Buck, R. (1988a). Introduction. In “Human Motivation and Emotion” (R. Buck. ed.), pp.
1-89. Wiley, New York.
Buck, R. (1988b). Central nervous system mechanisms of motivation and emotion. In “Human
Motivation and Emotion” (R. Buck, ed.), pp. 90-151. Wiley, New York.
Bunag, R. D. (1984). Measurement of blood pressure in rats. In “Handbook of Hypertension”
(W. de Jong, ed.), Vol. 4, pp. 1-12. Elsevier, New York.
Bush, I. E. (1962). Chemical and biological factors in the activity of adrenocortical steroids.
Pharmacol. Rev. 14,317-445.
Caldwell, J. D., Prange, A. J., Jr., and Pedersen, C. A. (1986). Oxytocin facilitates the sexual
receptivity of estrogen-treated female rats. Neuropeptides (Edinburgh) 7 , 175-189.
Calhoun, J. B. (1963). “The Ecology and Sociology of the Norway Rat.” U . S. Public Health
Service Publ., Bethesda, MD.
Candland, D. K.. and Leshner. A. I. (1974). A model of agonistic behavior: Endocrine and
autonomic correlates. In “Limbic and Autonomic Nervous Systems Research” (L. V.
DiCara, ed.), pp. 137-163. Plenum, New York.
Cannon, W. B. (1929). “Bodily Changes in Pain, Hunger, Fear and Rage.” Branford, Boston.
Carr, D. J., and Blalock, J. E. (1991). Neuropeptide hormones and receptors common to the
immune and neuroendocrine systems: Bidirectional pathway of intersystem communica-
tion. In “Psychoneuroimmunology” (R. Ader, D. L. Felten, and N. Cohen, eds.), pp.
573-588. Academic Press, San Diego, CA.
Carr, W. J., Kimmel, K. R., and Anthony, S. L. (1982). Female rats prefer to mate with
dominant rather than subordinate males. Bull. Psychon. 20, 89-9 1.
Carter, C. S. (1992). Neuroendocrinology of sexual behavior in the female. In “Behavioral
Endocrinology” (J. B. Becker, S. M., Breedlove, and D. Crews, eds.), pp. 71-96. MIT
Press, Cambridge, MA.
Carter, C. S., Witt, D. M., Thompson, E. G . , and Carlstead, K. (1988). Effects of hormonal,
sexual, and social history on mating and pair bonding in prairie voles. Physiol. Behav.
44,691-697.
Carter, C. S., Williams, J. R., and Witt, D. M. (1990). The biology of social bonding in a
monogamous mammal. In “Hormones, Brain and Behaviour in Vertebrates” (J. Baltha-
zart, ed.). pp. 154-164. Karger, Basel.
Carter, C. S., Williams, J. R., Witt, D. M., and Insel, T. R. (1992). Oxytocin and social bonding.
Ann. N . Y. Acad. Sci. 652, 204-211.
Carter, C. S.. DeVries, A. C., and Getz, L. L. (1995). Physiological substrate of mammalian
monogamy: The prairie vole model. Nrurosci. Biobehav. Res. 19, 303-314.
Chapman, V., Desjardins, C., and Bronson, F. (1969). Social rank in male mice and adrenocorti-
cal response to open field exposure. Proc. Soc. Exp. Biol. Med. 130,624-627.
Cherkovich, G . M.. and Tatoyan, S. K. (1973). Heart rate (radiotelemetrical registration) in
macacques and baboons according to dominant-submissive rank in a group. Folia Primatol.
20,265-273.
Chitty, D. (1958). Self-regulation of numbers through changes in viability. Cold Spring Harb.
Symp. Quant. Biol. 22, 277-280.
Chitty, D. (1960). Population processes in the vole and their relevance to general theory.
Can. J. Zool. 38, 99-113.
Ciaranello, R. D. (1978). Regulation of phenylethanolamine N-methyltransferase synthesis
and degradation. I. Regulation by adrenal glucocorticoids. Mol. Pharmacol. 14,478-489.
Christian, J. J. (1950). The adreno-pituitary system and population cycles in mammals.
J. Mammal. 31, 247-259.
Christian, J. J. (1963). Endocrine adaptive mechanisms and the physiologic regulation of
population growth. I n “Physiological Mammalogy” (W. V. Mayer and R. G. van Gelder,
eds.), Vol. 1, pp. 189-353. Academic Press, New York.
Christian, J. J. (1971). Population density and reproductive efficiency. Biol. Reprod. 4,248-294.
Christian, J. J. (1975). Hormonal control of population growth. I n “Hormonal Correlates of
Behavior” (B. E. ElefthCriou and R. L. Sprott, eds.), pp. 205-274. Plenum, New York.
Christian, J. J. (1978). Neurobehavioral endocrine regulation in small mammal populations.
I n “Populations of Small Mammals under Natural Circumstances” (D. P. Snyder, ed.),
pp. 143-158. University of Pittsburg Press, Pittsburg.
Christian, J. J. (1980). Endocrine factors in population regulation. In “Biosocial Mechanisms
of Population Regulation” (N. N. Cohen, R. S. Malpass, and H. G. Klein, eds.), pp.
55-1 16. Yale University Press, New Haven, CT.
Christian, J. J., and LeMunyan, C. D. (1958). Adverse effects of crowding on lactation and
reproduction of mice and two generations of their offspring. Endocrinology (Ba/timore)
63,517-529.
Christian, J. J.. and Wiliamson, H. 0. (1958). Effect of crowding on experimental granuloma
formation in mice. Proc. Soc. Exp. Biol. Med. 99,385-387.
Christian, J. J.. Lloyd, J. A,, and Davis, D. E. (1965). The role of endocrines in the self-
regulation of mammalian populations. Recent Prog. Horm. Res. 21, 501-568.
Chrousos, G . P., Loriaux. L. D., and Gold, P. W. (1988). The concept of stress and its historical
development. Adv. Exp. Med. Biol. 245, 3-7.
Clarke, M. R.. Harrison, R. M., and Didier, E. S. (1996). Behavioral, immunological, and
hormonal responses associated with social change in rhesus monkeys (Macaca mularra).
Am. J. Primatol. 39, 223-233.
Cocke, R., Moynihan. J. A,, Cohen, N., Grota, L. J., and Ader, R. (1993). Exposure to
conspecific alarm chemosignals alters immune function in BALB/c mice. Brain, Behav.,
Immunol. 7 , 36-46.
Coe, C. L. (1993). Psychosocial factors and immunity in nonhuman primates: A review.
Psychosom. Med. 55,298-308.
Coe, C. L., and Scheffler. J. (1989). Utility of immune measures for evaluating psychological
well-being in nonhuman primates. Zoo B i d 1, 89-99.
Coe, C. L., Mendoza, S. P., and Levine, S. (1979). Social status constrains the stress response
in the squirrel monkey. Physiol. Behav. 23, 633-638.
Coe, C. L., Franklin, D., Smith, E. R., and Levine, S. (1982). Hormonal responses accompanying
fear and agitation in the squirrel monkey. Physiol. Behav. 29, 1051-1057.
Collaer. M. L.. and Hines, M. (1995). Human behavioral sex differences: A role for gonadal
differences during early development. Psychol. Bull. 118, 55-107.
Cowan, D. P. (1987). Aspects of the social organisation of the European wild rabbit (Oryctola-
gus cuniculus). Efhology 75, 197-210.
Creel, S., Creel, N., Wildt, D. E., and Montfort, S. L. (1992). Behavioural and endocrine
mechanisms of reproductive suppression in Serengeti dwarf mongooses. Anim. Behav.
43,231 -245.
Creel, S., Creel, N. M., and Monfort, S. L. (1996). Social stress and dominance. Nariire
(London) 379,212.
Crnic. L. S . (1991). Behavioral consequences of viral infection. In “Psychoneuroimmunology”
(R. Ader, D. L. Felten, and N. Cohen, eds.), pp. 749-787. Academic Press, San Diego, CA.
Crowcroft, P. (1955). Territoriality in wild house mice. J. Mammal. 36, 299-301.
Da Costa. A. P. C., Guevara-Guzman, R. G., Ohkura. S.. Goode, J. A,, and Kendrick, K. M.
(1996). The role of oxytocin release in the paraventricular nucleus in the control of
maternal behaviour in the sheep. J. Neiiroendocrinol. 8, 163-177.
Dallman. M. F. (1991). Regulation of adrenocortical function following stress. In “Stress,
Neurobiology and Neuroendocrinology” (M. R. Brown, G. F. Koob, and C. Rivier, eds.),
pp. 173-192. Dekker. New York.
Darwin, C. (1872). “The Expression of the Emotions in Man and Animals.” Murray, London.
Davis, D. E., and Christian, J. J. (1957). Relation of adrenal weight to social rank of mice.
Proc. Soc. Exp. Biol. Med. 94, 728-731.
Davis, D. E.. and Read, C. P. (1958). Effects of behavior on development of resistance in
Trichinosis. Proc. Soc. Exp. Biol. Med. 99, 269-272.
Davis, H., Porter, J. W., Livingstone, J., Herrmann, T., MacFadden, L., and Levine. S. (1977).
Pituitary-adrenal activity and lever press shock escape behavior. Physiol. Psycho/. 5,
280-284.
de Boer. S. F.. de Beun, R., Slagen, J. L., and van der Gugten. J. (1990). Dynamics of plasma
catecholamine and corticosterone concentrations during reinforced and extinguished op-
erant behavior in rats. Physiol. Behav. 47, 691-698.
De Jonge. F. H., Bokkers, E. A. M.. Schouten, W. G. P., and Helmond. F. A. (1996). Rearlin
piglets in a poor environment: Developmental aspects of social stress in pigs. Physiol.
Behav. 60,389-396.
Dembroski, T . M., Schmidt, T. H., and Blumchen, G., eds. (1983). “Biobehavioral Bases of
Coronary Heart Disease.” Karger, Basel.
Dempsher, D. P., and Gann, D. S. (1983). Increased cortisol secretion after small hemorrhage is
not attributable to changes in adrenocorticotropin. Endocrinology (Balrimore) 113,86-93.
de Ruiter. A. J. H., Feitsma, L. E., Keijser, J. N., Koolhaas, J. M., van Oortmerssen. G. A.,
and Bohus, B. (1993). Differential perinatal testosterone secretory capacity of wild house
mice testes is related to aggressiveness in adulthood. Horm. Behav. 27, 231-239.
Dess. N. K., Linwick, D., Patterson, J.. Overmier, J. B., and Levine, S. (1983). Immediate
and proactive effects of controllability and predictability on plasma cortisol responses to
shocks in dogs. Behav. Neiirosci. 97, 1005-1016.
DeVries. A. C.. DeVries, M. B., Taymans. S., and Carter, C. S. (1995). Modulation of pair
bonding in female prairie voles (Microtiis ochrogasrer) by corticosterone. Proc. Narl.
Acad. Sci. U. S. A. 92, 7744-7748.
Dhabhar, F. S., Miller, A. H., McEwen, B. S., and Spencer, R. L. (1995). Effects of stress on
immune cell distribution. Dynamics and hormonal mechanisms. J. Immunol. 154, 551 1-
5527.
Dijkstra, H., Tilders. F. H. J.. Hiehle, M. A,. and Smelik, P. G. (1YY2). Hormonal reactions
to fighting in rat colonies: Prolactin rises during defense, not during offense. Physiol.
Behav. 51,961 -968.
Dittus, W. P. J. (1979). The evolution of behaviors regulating density and age-specific sex
ratios in a primate population. Behavioiir 69, 265-302.
Dixon, A. (1979). Androgens and aggressive behavior in primates: A review. Aggressive
Behav. 6,37-52.
Dobson, A. P., and Hudson, P. J. (1986). Parasites, disease and the structure of ecological
communities. Trends Ecol. Evol. 1, 11-15.
Dohrenwend. B. S.. and Dohrenwend, B. P., eds. (1974). “Stressful Live Events: Their Nature
and Effects.” Wiley. New York.
Drickhammer. L. C. (1974). A ten-year summary of reproductive data for free-ranging Macaca
mulatfa. Folia Primafol. 21, 61 -80.
Dunn, A. J. (1989). Psychoneuroimmunology for the psychoneuroendocrinologist: A review
of animal studies of nervous system-immune system interactions. Psychoneicroendocrinol-
Ogy 14,25 1-274.
Dyer, A. R., Persky, V., Stamler, J., Oglesby, P., Shekelle. R. B.. and Berkson, D. M. (1980).
Heart rate as a prognostic factor for coronary heart disease and mortality: Findings in
three Chicago epidemiologic studies. Am. J . Epidemiol. 112, 736-749.
Ebbesen. P., Villadsen, J. A,, Villadsen, D., and Heller, K. E. (1991). Effect of subordinance.
lack of social hierarchy and restricted feeding on murine survival and virus leukemia.
Exp. Gerontol. 26, 479-486.
Eberhart, J. A,. Keverne, E. B., and Meller, R. E. (1983). Social influences on circulating
levels of cortisol and prolactin in male talapoin monkeys. Physiol. Behav. 30, 361-369.
Eberhart. J. A,, Yodyingyuad, U.. and Keverne. E. B. (1985). Subordination in male talapoin
monkeys lowers sexual behaviour in the absence of dominants. Physiol. Behav. 35,
673-677.
Edwards, E. A,, and Dean. L. M. (1977). Effects of crowding of mice on humoral antibody
formation and protection to lethal antigenic challenge. Psychosom. Med. 39, 19-24.
Edwards, E. A,, Rahe, R. H., Stephens, P. M., and Henry. J. H. (1980). Antibody response
to bovine serum albumin in mice: The effects of psychosocial environmental change.
Proc. Soc. Exp. Biol. Med. 164,478-481.
Edwards, J. C. (1988). The effects of Trichinella spiralis infection on social interactions in
mixed groups of infected and uninfected male mice. Anim. Behav. 36, 529-540.
Edwards. J. C.. and Barnard. C. J. (1987). The effects of Trichinella infection on intersexual
interactions between mice. Anim. Behav. 35, 533-540.
Ekkel, E. D., Dieleman, S. J.. Schouten, W. G. P., Portela, A,, Cornilissen, G., Tielen.
M. J. M.. and Halberg, F. (1996). The circadian rhythm of cortisol in the saliva of young
pigs. Physiol. Behav. 60, 985-989.
Eisermann. K. (1992). Long-term heart responses to social stress in wild European rabbits:
Predominant effect of rank position. Physiol. Behav. 52, 33-36.
Eisermann, K.,and Stohr, W. (1992). Diurnal heart rate rhythms in small mammals: Species-
specific patterns and their environmental modulation. In “Temporal Variations in Cardiac
Rhythms in Rodents” (T. H. Schmidt, B. T. Engel, and G. Bliimchen, eds.), pp. 87-91.
Springer, Berlin.
Eisermann, K., Meier. B., Khaschei, M., and von Holst, D. (1993). Ethophysiological responses
to overwinter food shortage in wild European rabbits. Physiol. Behav. 54, 983-980.
Elias, M. (1981). Cortisol, testosterone and testosterone-binding globulin responses to com-
petetive fighting in human males. Aggressive Behav. 7 , 215-222.
Ellis, L. (1995). Dominance and reproductive success among nonhuman animals: A cross-
species comparison. Ethol. Sociobiol. 16, 257-333.
Elton, C. S. (1Y42). “Voles, Mice and Lemmings.” Clarendon Press. Oxford.
Ely, D. L. (1981). Hypertension, social rank, and aortic arteriosclerosis in CBA/J mice. Physiol.
Behav. 26, 655-661.
Ely, D. L., and Henry, J. P. (1978). Neuroendocrine response patterns in dominant and
subordinate mice. Horrn. Behav. 10, 156-169.
Engel, B. T. (1985). Stress is a noun! No, a verb! No, an adjective. In “Stress and Coping”
(T. M. Field. P. M. McCabe, and N. Schneidermann, eds.), pp. 3-12. Erlbaum, Hills-
dale, NJ.
116 DIETRICH VON HOLS?
Evain, D., Morera. A. M., and Saez, J. M. (1976). Glucocorticoid receptors in interstitial cells
of rat testis. J. Steroid Biochem. 7, 1135-1 139.
Fagin, K. D.. Shinsako, J., and Dallman, M. F. (1983). Effects of housing and chronic cannula-
tion on plasma ACTH and corticosterone in the rat. Am. J . Physiol. 245, E515-E520.
Fauman, M. A. (1987). The relation of dominant and submissive behavior to the humoral
immune response in BALBlc mice. Biol. Psychiatry 22, 771-776.
Felten, S. Y., and Felten, D. L. (1991). Innervation of lymphoid tissue. In “Psychoneuroimmu-
nology” (R. Ader, D. L. Felten, and N. Cohen, eds.), pp. 27-61. Academic Press, San
Diego, CA.
Fenske, M. (1989). Application of a new, simple method for quantitative collection of 24-
hour urines in small laboratory animals: Determination of basal excretion of proteins,
creatinine, urea, electrolytes, and free steroids. Z . Versuchstierk. d. 32, 65-70.
Fenske, M. (1996). Saliva cortisol and testosterone in the guinea pig: Measures for the endo-
crine function of adrenals and testes? Steroids 61, 647-650.
Fleshner. M., Laudenslager, M. L., Simons, L., and Maier, S. F. (1989). Reduced antibodies
associated with social defeat in rats. Physiol. Behav. 45, 1183-1187.
Fokkema, D. S. (1985). Social behavior and blood pressure: A study of rats. Unpublished
Doctoral Dissertation, Rijksuniversiteit, Groningen, Netherland.
Fokkema, D. S., and Koolhaas, J. M. (1985). Acute and conditioned blood pressure changes
in relation to social and psychosocial stimuli in rats. Physiol. Behav. 34, 33-38.
Fokkema, D. S., Koolhaas, J. M., and van der Gugten, J. (1995). Individual characteristics of
behavior, blood pressure, and adrenal hormones in colony rats. Physiol. Behav. 57,
857-862.
Fokkema, D. S.. h i t , K., van der Gugten, J., and Koolhaas, J. M. (1988). A coherent pattern
among social behavior, blood pressure, corticosterone and catecholamine measures in
individual male rats. Physiol. Behav. 42, 485-489.
Folkow, B., Grimby, G., and Thulesius, 0.(1958). Adaptive structural changes of the vascular
walls in hypertension and their relation to the control of the peripheral resistance. Acta
Physiol. Scand. 44,255-272.
Folkow, B., Halback, M., and Lundgren, Y. (1973). Importance of adaptive changes in vascular
design for establishment of primary hypertension: Studies in man and in spontaneously
hypertensive rats. Circ. Res. 32/33, suppl. I, 2-13.
Freeland, W. J. (1981). Parasitism and behavioral dominance among male mice. Science
213,461-462.
Friedman, E., Coe. C. L., and Ershler, W. B. (1991). Time-dependent effects of peer separation
on lymphocyte proliferatin response in juvenile squirrel monkeys. Dev. Psychobiol. 24,
159-173.
Friedman, M., and Rosenman, R. H. (1974). “Type A Behavior and your Heart.” Alfred A.
Knopf, New York.
Friedman. S. B., and Glasgow, L. A. (1973). Interaction of mouse strain and differential
housing upon resistance to Plasmodium berghei. J. Parsitol. 59, 851-854.
Fukuhara, K., Kvetnansky, R., Weise, V. K., Ohara, H., Yoneda, R., and Kopin, I. J. (1992).
Correlation of plasma catecholamine levels with tissue tyrosine hydroxylase activity in
sart-stressed rats. In “Stress: Neuroendocrine and Molecular Approaches” (R. Kvetnan-
sky, R. McCarty, and J. Axelrod, eds.). pp. 881-889. Gordon & Breach, New York.
Gao, H. B., Shan, L. X., Monder, C., and Hardy, M. P. (1996). Suppression of endogenous
corticosterone level in vivo increases the steroigenic capacity of purified rat Leydig cells
in vitro. Endocrinology (Baltimore) 137, 1714-1718.
Carson. P. J. (1979). Social organization and reproduction in the rabbit: A review. In “Proceed-
ings of the World Lagomorph Conference 1979” (K. Myers and C. D. Maclnnes, eds.),
pp. 256-270. University of Guelph Press, Guelph.
Gentry, W. D., ed. (1984). “Handbook of Behavioral Medicine.” Guilford, New York.
Gilman, S. C., Schwarz, J. M., Mulner, R. J.. Bloom, F. E., and Feldmann, J. D. (1982). Beta-
endorphin enhances lymphocyte proliferative responses. Proc. Natl. Acad. Sci. U.S.A.
79,4426-4430.
Glaser, R., and Kiecolt-Glaser, J., eds. (1994). “Human Stress and Immunity.” Academic
Goy, R. W.. Bercowitch. F. B., and McBrair, M. C. (1988). Behavioral masculinization is
independent of genital masculinization in prenatally androgenized female rhesus ma-
caques. Horm. Behav. 22,552-571.
Grassi, M., and Drago, F. (1993). Effects of oxytocin on emotional stress and stress-induced
gastric lesion. J. Physiol. (London) 87, 261 -264.
Grossman, C. J. (1984). Regulation of the immune system by sex steroids. Endocr. Rev.
5,435-455.
Gust, D. A., Gordon, T. P., Wilson, M. E., Ahmed-Ansari, A.. Brodie, A. R., and McClure,
H. M. (1991). Formation of a new group of unfamiliar female rhesus monkeys affects
the immune and pituitary adrenocortical systems. Brain, Behav., Immunol. 5, 296-307.
Gust, D. A,, Gordon, T. P.. Wilson, M. E.. Brodie. A. R., Ahmed-Ansari, A,, and McClure,
H. M. (1992). Removal from natal social group to peer housing affects cortisol levels and
absolute numbers of T cell subsets in juvenile rhesus monkeys. Brain, Behav., Immunol. 6,
189- 199.
Gust, D. A,, Gordon, T. P., and Hambright, M. K. (1993a). Response to removal from and
return to a social group in adult male rhesus monkeys. Physiol. Behav. 53, 599-602.
Gust, D. A.. Gordon, T. P.. Hambright, M. K., and Wilson, M. E. (1993b). Relationship
between social factors and pituitary-adrenocortical activity in female rhesus monkeys
(Macaca mulatta). Horni. Behav. 27, 318-331.
Gust, D. A,, Gordon, T. P., Wilson, M. E., Brodie, A. R., Ahmed-Ansari, A,, and McClure.
H. M. (1996). Group formation of female pigtail macacques (Macaca nemestrina). Am.
J. Primatol. 39, 263-273.
Gustafson, M. L., and Donahoe. P. K. (1994). Male sex determination-current topics of
male sexual differentiation. Annu. Rev. Med. 45, 505-524.
Hall, N. S., O’Grady, M . P., and Farah, J. M. (1991). Thymic hormones and immune function:
Mediation via neuroendocrine circuits. In “Psychoneuroimmunology” (R. Ader, D. L.
Felten, and N. Cohen, eds.), pp. 515-528. Academic Press, San Diego, CA.
Hamilton. C. L., and Chaddock, T. (1977). Social interaction and serum insulin values in the
monkey (Macaca mulatto). Psychosom. Med. 39,444-449.
Hamilton, W. D., and Zuk, M. (1982). Heritable true fitness and bright birds: A role for
parasites? Science 218, 384-387.
Hamm, T. E., Kaplan. J . R., Clarkson, T. B., and Bullock. B. C. (1983). Effects of gender
and social behavior on the development of coronary artery atherosclerosis in cynomolgus
monkeys. Atherosclerosis (Shannon, Irel.) 48, 221-233.
Hanlon. A. J.. Rhind, S. M., Reid, H. W., Burrells, C., and Lawrence, A. B. (1995). Effects
of repeated changes in group composition on immune response, behaviour, adrenal
activity and lifeweight gain infarmed red deer yearlings. Appl. Anim. Behav. Sci. 44,57-64.
Hanson, J. D., Larson, M. E., and Snowdon, C. T. (1976). The effects of control over high
intensity noise on plasma cortisol levels in rhesus monkeys. Behav. Biol. 16,333-340.
Hardy, C. A., Quay, J., Livnat. S.. and Ader, R. (1990). Altered lymphocyte response following
aggressive encounters in mice. Physiol. Behav. 47, 1245-1251.
Harlow, H. F., and Suomi. S. J. (1974). Induced depression in monkeys. Behav. Biol. 12,
273-296.
Harlow, H. F., McGaugh, J. L., and Thompson, R. F. (1971). “Psychology.” Albion Publ.,
San Francisco.
Hausfater, G., and Watson, D. F. (1976). Social and reproductive correlates of parasite ova
emissions by baboons. Nature (London) 262, 688-689.
Heinjen, C. J., Kavelaars, A., and Ballieux, R. E. (1991). Corticotropin-releasing hormone
and the proopiomelanocortin-derived peptides in the modulation of immune function.
In “Psychoneuroimmunology” (R. Ader, D. L. Felten, and N. Cohen, eds.), pp. 429-513.
Hennessy, M. B., and Ritchey, R. L. (1987). Hormonal and behavioral attachment responses
in infant guinea pigs. Dev. Psychohiol. 20, 613-625.
Henry, J. P. (1986). Neuroendocrine patterns of emotional response. I n “Emotion: Theory.
Research and Experience” (R. Plutchik, ed.), Vol. 3. pp. 37-60. Academic Press, New
York.
Henry, J. P. (1992). Biological basis of the stress response. Integr. Physiol. Behav. Sci. 27,66-83.
Henry, J. P., and Meehan, J. P. (1981). Psychosocial stimuli, physiological specificity, and
cardiovascular disease. 111 “Brain, Behavior, and Bodily Disease” (H. Weiner, M. A.
Hofer, and A. J. Stunkard. eds.), pp. 305-333. Raven Press, New York.
Henry, J. P., and Stephens, P. M. (1977). “Stress, Health. and the Social Environment. A
Sociobiologic Approach to Medicine.” Springer, New York.
Henry, J. P., and Stephens-Larson, P. M. (1985). Specific effects of stress on disease processes.
In “Animal Stress” (G. P. Moberg, ed.), pp. 161-175. Am. Physiol. SOC.,Bethesda, MD.
Henry, J. P., Ely, D. L., Stephens, P. M., Ratcliffe, H. L., Santisteban, G. A,, and Shapiro,
A. P. (1971). The role of psychosocial factors in the development of arteriosclerosis in
CBA mice: Observations on the heart, kidney. and aorta. Atherosclerosis (Shannon Ire/.)
14, 203-218.
Henry, J. P., Ely, D. L.. and Stephens, P. M. (1972). Changes in catecholamine-controlling
enzymes in response to psychosocial activation of the defense and alarm reactions. In
“Physiology, Emotion and Psychosomatic Illness” (R. Porter and J. Knight, eds.), Ciba
Found. Symp. 8, pp. 225-25 I . Elsevier, Amsterdam.
Henry, J. P., Kross, M. E.. Stephens, P. M., and Watson, F. M. C. (1976). Evidence that
differing psychosocial stimuli lead to adrenal cortical stimulation by autonomic or endo-
crine pathways. In “Catecholamines and Stress” (E. Ursdin, R. Kvetnansky. and I. J.
Kopin, eds.), pp. 457-468. Pergamon, Oxford.
Henry, J. P.. Liu, Y. Y.. Nadra. W. E., Quian, C., Mormede, P., Lemaire, V.. Ely, D., and
Henley, E. D. (1993). Psychosocial stress can induce chronic hypertension in normotensive
strains of rats. Hypertension (Dallas) 21, 714-723.
Henry, J. P., Liu, J., and Meehan. W. P. (1995). Psychosocial stress and experimental hyperten-
sion. I n “Hypertension: Physiology, Diagnosis, and Management” (J. H. Laragh and
B. M. Brenner, eds.), pp. 905-921. Raven Press, New York.
Hessing. M. J. C., Scheepens. C. J. M.. Schouten. W. G. P., Tielen, M. J. M., and Wiepkema,
P. R. (1994). Social rank and disease susceptibility in pigs. Vet. Immunol. Immunopathol.
43, 373-387.
Hofer, M. A. (1981). Toward a developmental basis for disease predisposition: The effects
of early maternal separation on brain, behavior, and cardiovascular system. I n “Brain.
Behavior, and Bodily Disease” (H. Weiner. M. A. Hofer. and A. J . Stunkard, eds.). pp.
209-228. Raven Press, New York.
Hofer. M. A. (1994). Early relationships as regulators of infant physiology and behavior. Acta
Paediatr., Suppl. 397, 9-18.
Hoffman-Goetz, L., Simpson, J. R.. and Arumugam, Y. (1991). Impact of changes in housing
condition on mouse natural killer cell activity. Physiol. Behav. 49, 657-660.
Hoffmeyer, I. (1982). Responses of female bank voles (Clethrionomys glareolus) to dominant

vs subordinate conspecific males and to urine odors from dominant vs subordinate males.
Behav. Neural Biol. 36, 178-188.
House, J. S., Robbins, C.. and Metzner, H. L. (1982). The association of social relationships
and activities with mortality: Prospective evidence from the Tecumseh community health
study. A m . J . Epidemiol. 116, 123-140.
House, J. S., Landis. K. R., and Umberson, D. (1988). Social relationships and health. Science
241,540-544.
Hrdina, P. D., and Henry, M. E. (1981). Experimental models of mental illness: Separation-
induced depression in primates. In “Neuroendocrine Regulation and Altered Behavior”
(P. D. Hrdina and R. L. Singhal, eds.), pp. 263-275. Croom Helm, London.
Huck, W. W., and Banks, E. M. (1982). Male dominance status, female choice and mating
success in the brown lemming, Lemmus trimucronatus. Anim. Behav. 30,665-675.
Hucklebridge, F. H., Gamal-el-Din, L., and Brain, P. (1981). Social status and the adrenal
medulla in the house mouse ( M u s musculus, L.) Behav. Neural Biol. 33, 345-363.
Huhtaniemi, 1. (1994). Fetal testis: A very special endocrine organ. Eur. J. Endocrinol. 130,
25-31.
Human, K. L., Moore, T. O., Mougey. E. H., and Meyerhoff, J. L. (1992). Hormonal responses
to fighting in hamsters: Separation of physical and psychological causes. Physiol. Behav.
51, 1083-1086.
Huntingford, F. A., and Turner, A. K. (1987). “Animal Conflict.” Chapman & Hall, London.
Hutzelmeyer, H. (1987). Sympathicus-Nebennierenmarkaktivitatmannlicher Tupaia belang-
eri: Ein Mass zur Charakterisierung von Individuen mit unterschiedlichem Verhalten.
Unpublished Doctoral Dissertation, University of Bayreuth, Bayreuth, Germany.
Insel. T. R.. and Hulihan. T. J. (1995). A gender-specific mechanism for pair bonding: Oxytocin
and partner preference formation in monogamous voles. Behav. Neurosci. 109,782-789.
Insel, T. R., Preston, S.. and Winslow. J. T. (1995). Mating in the monogamous male: Behavioral
consequences. Physiol. Behav. 57, 615-627.
Johnson, E. O., Kamilaris, T. C., Carter, C. S., Calogero, A. E., Gold, P. W., and Chrousos.
G. P. (1996). The biobehavioral consequences of psychogenic stress in a small social
primate (Callithrix jacchus jacchus). Biol. Psychiatry 40,317-337.
Kaiser, C. (1996). Trennung physischer und psychischer Anteile sozialer Stressreaktionen bei
mannlichen Tupaia belangeri. Unpublished Doctoral Dissertation, University of Bayreuth.
Bayreuth, Germany.
Kannel, W. B., Kannel, C., Pfaffenbarger. R. S., and Cupples, L. A. (1987). Heart rate and
cardiovascular mortality. The Framingham Study. A m . Heart J. 116, 1369-1373.
Kaplan, J. R., Manuck, S. B.. Clarkson, T. B., Lusso, F. M., and Taub, D. M. (1982). Social
status, environment, and atherosclerosis in cynomolgus monkeys. Arteriosclerosis (Dallas)
2,359-368.
Kaplan. J. R., Heise, E. R., Manuck, S. B., Shively, C. A,, Cohen, S., Rabin, B. S.. and
Kasprowicz, A. L. (1991). The relationship of agonistic and affiliative behavior patterns
to cellular immune function among cynomolgus monkeys Macaca fascicularis living in
unstable social groups. Am. J. Primafol. 25, 157-174.
Kass, E. M., Hechter, 0..Macci, I. A,, and Mou, T. W. (1954). Changes in patterns of secretion
in rabbits after prolonged treatment with ACTH. Proc. Soc. Exp. Biol.Med. 85,583-587.
Kavaliers. M., and Colwell, D. D. (1995). Odours of parasized males induce aversive response
in female mice. Anim. Behav. 50, 1161-1169.
Keane, B., Waser, P. M., Creel, S. R.. Elliott, L. F., and Minchella, D. J. (1994). Subordinate
reproduction in dwarf mongooses. Anim. Behav. 4 7 , 6 5 7 5 .
Keddy, A. C. (1986). Female mate choice in vervet monkeys (Cercopithecus aefhivps sabaeus).
Am. J. Primatol. 10, 125-134.
Keller. S. E., Schleifer, S. J . , and Demetrikopoulos, M. K. (1991). Stress-induced changes in
immune function in animals: hypothalamo-pituitary-adrenal influences. In “Psychoneu-
roimmunology” (R. Ader, D. L. Felten. and N. Cohen, eds.). pp. 771-787. Academic
Kelley, K. W. (1991 ). Growth hormone in immunobiology. I n “Psychoneuroimmunology”
(R. Ader, D. L. Felten, and N. Cohen. eds.), pp. 377-402. Academic Press, San Diego. CA.
Kelley, K. W., Johnson, R. W., and Dantzer. R. (1994). Immunology discovers physiology.
Vet. Immunol. Immunopathol. 43, 157-165.
Kendrick, K. M., Keverne, E. B., and Baldwin. B. A. (1987). Intracerebroventricular oxytocin
stimulates maternal behavior in the sheep. Neuroendocrinology 46, 56-61.
Keverne, E. B. (1988). Central mechanisms underlying the neural and neuroendocrine determi-
nants in maternal behavior. Psychoneurvendocrinology 13, 127-141.
Keverne, E. G., Meller, R. E.. and Eberhart, J . A. (1982). Dominance and subordination:
Concepts and physiological states. I n “Advanced Views in Primate Biology” (A. B.
Chiarelli and R. S. Corrucini, eds.), pp. 81-94. Springer, New York.
Kime, D. E., Vinson. G. P., Major, P. W., and Kilpatrick, R. (1980). Adrenal-gonad relation-
ships. In “General. Comparative and Clinical Endocrinology of the Adrenal Cortex” (I.
Chester Jones and 1. W. Henderson, eds.), 183-264. Academic Press, London.
Kirschbaum, C., and Hellhammer, D. H. (1989). Salivary cortisol in psychobiological research:
An overview. Neurnpsychvbiology 22, 150-169.
Klein, F., Lernaire, V., Sandi. C., Vitiello, S.. Van der Logt. J., Laurent, P. E., Neveu, P., Le
Moal, M., and Mormede, P. (1992). Prolonged increase of corticosterone secretion by
chronic social stress does not necessarily impair immune function. Life Sci. 50,723-731.
Klopfer, P. H. (1971). Mother love: What turns it on? Am. Sci. 59,404-407.
Koolhaas, J., Schuurman. T., and Wiepkema, P. R. (1980). The organization of intraspecific
agonistic behaviour in the rat. Prvg. Neurobiol. 15, 247-268.
Kopin, I. J . (1980). Catecholamines, adrenal hormones and stress. I n “Neuroendocrinology”
(D. T. Krieger and J. C. Hughes, eds.), pp. 159-166. Sinauer, Sunderland, MA.
Kopin. 1. J., Eisenhofer, G., and Goldstein, D. (1988). Sympathoadrenal medullary system
and stress. In “Mechanisms of physical and emotional stress: Advances in experimental
medicine and biology, Volume 245” (G. P. Chrousos, L. D. Loriaux, and P. W. Gold,
eds.), pp. 11-23. Plenum Press, New York.
Krebs, C. J . (1964). The lemming cycle at Baker Lake, Northwest Territories, during 1959-1962.
Tech. Pap.-Arct. Inst. North Am. 15.
Krehs, C. J. (1978). “Ecology-The Experimental Analysis of Distribution and Abundance.”
Harper & Row, New York.
Krebs, C. J . (1996). Population cycles revisited. J. Mammal. 77, 8-24.
Krebs. C. J., and Myers, J . H. (1974). Population cycles in small mammals. Adv. Ecol. Res.
8,267-399.
Krum, A. A., and Glenn, R. E. (1965). Adrenal steroid secretion in rabbits following prolonged
ACTH administration. Proc. SOC. Exp. Biol. Med. 118, 225-258.
Kunkele, J.. and von Holst, D. (1996). Natal dispersal in the European wild rabbit. Anim.
Behav. 51, 1047-1059.
Kvetnansy. R., Weise, V. K., and Kopin, I. J . (1970). Elevation of adrenal tyrosine hydroxylase
and phenylethanolamine-N-methyltransferase by repeated immobilization of rats. Endo-
crinology (Baltimore) 87, 744-749.
Kvetnansy. R.. Weise, V. K.. Thoa. N. B., and Kopin, I. J. (1979). Effects of chronic guanethi-
dine treatment and adrenal medullectomy on plasma levels of catecholamines and cortico-
sterone in forcibly immobilized rats. J. Pharmacol. Exp. Ther. 209, 287-291.
Labhart, A,, ed. (1986). “Clinical Endocrinology. Theory and Practice.” Springer, Berlin.
Lapin, B. A.. and Cherkovich, G. M. (1971). Environmental changes causing the development
of neuroses and corticovisceral pathology in monkeys. I n “Society, Stress and Disease”
(L. Levi, ed.), Vol. 1, pp. 266-279. Oxford University Press, London.
Laudenslager, M. L., and Fleshner. M. (1994). Stress and immunity: Of mice, monkeys, models,
and mechanisms. I n “Handbook of Human Stress and Immunity” (R. Glaser and J. K.
Kiecolt-Glaser. eds.), pp. 161-181. Academic Press, San Diego, CA.
Laudenslager, M. L., Reite. M., and Harbeck, R. J. (1982). Suppressed immune response in
infant monkeys associated with maternal separation. Behav. Neural Biol. 36, 40-48.
Lazarus, R. S.. and Folkman, S. (1984). Coping and adaptation. I n “Handbook of Behavioral
Medicine” (W. D. Gentry, ed.), pp. 282-325. Guilford. New York.
Lee, A. K., and McDonald, I. R. (1985). Stress and population regulation in small mammals.
Oxford Rev. Reprod. Biol. 7 , 261-304.
Lemaire, V., Le Moal, M., and Mormkde. P. (1993). Regulation of catecholamine-synthesizing
enzymes in adrenals of Wistar rats under chronic stress. Am. J. Physiol. 264, R957-R962.
Leshner, A. I. (1981). The role of hormones in the control of submissiveness. I n “Multidiscipli-
nary Approaches to Aggression Research” (P. F. Brain and D. Benton, eds.), pp. 309-322.
Elsevier, Amsterdam.
Leshner. A. I., and Candland, D. K. (1972). Endocrine effects of grouping and dominance
rank in squirrel monkeys. Physiol. Behav. 8, 441-446.
Leshner. A. I., and Politch. J. A. (1979). Hormonal control of submissiveness in mice: Irrele-
vance of the androgens and relevance of the pituitary-adrenal hormones. Physiol. Behav.
22,53 1-534.
Leshner, A. I.. Korn, S. J., Mixon, J. F., Rosenthal, C., and Besser, A. K. (1980). Effects of
corticosterone on submissiveness in mice: some temporal and theoretical considerations.
Physiol. Behav. 24,283-288.
Levi. L.. ed. (1971). “Society. Stress and Disease,” Vol. 1. Oxford University Press, London.
Levine, S. (1993a). The psychoendocrinology of stress. Ann. N.Y. Acad. Sci. 697, 61-69.
Levine. S. (1993b). The influence of social factors on the response to stress. Psychofher.
Psychosom. 60, 33-38.
Levine, S., and Ursin, H. (1991).What is stress? I n “Stress,Neurobiology and Neuroendocrinol-
ogy” (M. R. Brown. G. F. Koob, and C. Rivier, eds.), pp. 3-21. Dekker, New York.
Levine, S., Weinberg, J., and Brett, L. P. (1979). Inhibition of pituitary-adrenal activity as a
consequence of consummatory behavior. Psychonellroendocrinology 4,275-286.
Levine, S., Johnson, D. F., and Gonzales, C. A. (1985). Behavioral and hormonal responses
to separation in infant rhesus monkeys and mothers. Behav. Neurosci. 99, 399-410.
Levine, S., Coe, C. L., and Wiener, S. G. (1989). Psychoneuroendocrinology of stress: A
psychobiological perspective. In “Psychoendocrinology” (F. R. Brush and S. Levine,
eds.), 341-378. Academic Press, San Diego, CA.
Lilly, M. P., Engeland, W. C., and Gann, D. S. (1983). Responses of cortisol secretion to
repeated hemorrhage in the anesthetized dog. Endocrinology (Balfimore)112, 681-688.
Line, S. W., Kaplan, J. R., Heise, E. R., Hilliard. J. K., Cohen, S.. Rabin, B. S., and Manuck,
S. B. (1996). Effects of social reorganization on cellular immunity in male cynomolgus
monkeys. Am. J. Primatol. 39,235-249.
Lloyd, J. A. (1973). Frequency of activity and endocrine response among male house mice
(Mus musculus) in freely growing populations. Proc. SOC. Exp. Biol. Med. 142,784-786.
Lockley, R. M. (1961). Social structure and stress in the rabbit warren. J. h i m . Ecol. 30,
385-423.
Louch, C. D.. and Higginbotham, M. (1967). The relation between social rank and plasma
corticosterone levels in mice. Gen. Comp. Endocrinol. 8, 441-444.
Lubach, G. R., Coe. C. L., Karaszewski, J. W., and Ershler, W. B. (1996). Effector and target
cells in the assessment of natural cytotoxic activity of rhesus monkeys. Am. J. Primatol.
39,275-287.
Lundberg, U., and Frankenhaeuser, M. (1980). Pituitary-adrenal and sympathetic-adrenal
correlates of distress and effort. J. Psychosom. Res. 24, 125-130.
Madden, K. S., and Felten, D. L. (1995). Experimental basis for neural-immune interactions.
Physiol. Rev. 75, 77-106.
Madden, K. S., and Livnat, S. (1991). Catecholamine action and immunological reactivity. In
“Psychoneuroimmunology” (R. Ader, D. L. Felten, and N. Cohen, eds.), pp. 283-310.
Magarinos, A. M., McEwen, B. S., Flugge, G., and Fuchs, E. (1996). Chronic psychosocial stress
causes apical dendrite atrophy of hippocampal CA3 pyramidal neurons in subordinate tree
shrews. J . Neitrosci. 16, 3534-3540.
Mallick, J., Stoddart, D. M., Jones, 1.. and Bradley, A. J. (1994). Behavioral and endocrinologi-
cal correlates of social status in the male sugar glider (Petaurus breviceps Marsupialia:
Petauridae). Physiol. Behav. 55, 1131-1134.
Mann, D. R.. and Orr. T. E. (1990). Effect of restraint stress on gonadal proopiomelanocortin
peptides and the pituitary-testicular axis in rats. Life Sci. 46, 1601-1609.
Manogue, K. R., Leshner, A. I., and Candland, K. (1975). Dominance status and adrenocortical
reactivity to stress in squirrel monkeys (Saimiri sciureus). Primates 16, 457-463.
Manuck, S. B., Kaplan. J. R., and Clarkson. T. B. (1983). Social instability and coronary artery
atherosclerosis in cynomolgus monkeys. Neurosci. Biobehav. Rev. 7,485-491.
Manuck, S . B., Kaplan, J. R., and Matthews, K. A. (1986). Behavioral antecedents of coronary
heart disease and atherosclerosis. Arteriosclerosis (Dallas) 6, 2-14.
Manuck, S. B., Kaplan, J. R., Adams, M. R.. and Clarkson, T. B. (1989). Behaviorally elicited
heart rate reactivity and atherosclerosis in female cynomolgus monkeys (Macaca fascicu-
laris). Psychosom. Med. 51, 306-318.
Manuck, S. B., Marsland, A. L., Kaplan. J. R., and Williams, J. K. (1995). The pathogenicity
of behavior and its neuroendocrine mediation: An example from coronary heart disease.
Psychosom. Med. 57, 275-283.
Maric, D., Kostic, T., and Kovacevic, R. (1996). Effects of acute and chronic immobilization
stress on rat Leydig cell steroidogenesis. J. Steroid Biochem. Mol. Biol. 58, 351-355.
Marsden, H. M., and Holler, N. R. (1964). Social behavior in confined populations of the
cottontail and the swamp rabbit. Wildl. Monogr. 13, 1-39.
Masataka, N., Ishida, T., Suzuki, J., Matsumura, S., Udono, S.. and Sasaoka, S. (1990). Domi-
nance and immunity in chimpanzees (Pan troglodytes). Ethology 85, 147-155.
Maslow, A. H. (1936). The role of dominance in the social and sexual behavior of infra-
human primates: 111. A theory of sexual behavior of infra-human primates. J . Genet.
Psychol. 48, 310-338.
Mason, J. W. (1968a). Organization of psychoendocrine mechanisms. Psychosom. Med. 30,
565-808.
Mason, J. W. (1968b). A review of psychoendocrine research on the pituitary-adrenal cortical
system. Psychosom. Med. 30, 576-607.
Mason, J. W. (1968~).“Over-all’’ hormonal balance as a key to endocrine organization.
Mason, J. W., and Brady, J. V. (1956). Plasma 17-hydroxycorticosteroidchanges related to
reserpine effects on emotional behavior. Science 124, 983-984.
Mason, J. W., Kenion, C. C., Collins, D. R., Mougey, E. H., Jones. J. A,, Driver, G. C., Brady.
J . V., and Beer, B. (1968a). Urinary testosterone response to 72-hr. avoidance sessions
in the monkey. Psychosom. Med. 30, 721 -732.
Mason, J. W., Jones, J. A,, Ricketts, P. T., Brady, J. V.. and Tolliver, G. A. (1968b). Urinary
aldosterone and urine volume responses to 72-hr. avoidance sessions in the monkey.
Matochik. J. A,, and Barfield. R. J. (1991). Hormonal control of precopulatory sebaceous
scent marking and ultrasonic mating vocaliztion in male rats. Horm. Behav. 25,445-460.
Maynard Smith, J., and Price, G. R. (1973). The logic of animal conflict. Nature (London)
246,lS-18.
Mazur, A,, and Lamb, T. A. (1980). Testosterone. status and mood in human males. Horm.
Behav. 14,236-246.
McCarthy. M. M. (1994). Molecular aspects of sexual differentiation of the rodent brain.
Physiol. Brhav. 19,415-427.
McCruden, A. B., and Stimson. W. H. (1991). Sex hormones and immune function. I n “Psycho-
neuroimmunology” (R. Ader, D. L. Felten, and N. Cohen. eds.). pp. 475-493. Academic
McDonald, I. R., Lee, A. K., Bradley, A. J.. and Than, K. A. (1981). Endocrine changes in
dasyurid marsupials with differing mortality patterns. Gen. Comp. Endocrinol. 44,
292-301.
McDonald, I. R., Lee, A. K., Than, K. A., and Martin, R. W. (1986). Failure of glucocorticoid
feedback in males of a population of small marsupials (Antechinus swainsonii) during
period of mating. J . Endocrinol. 108, 63-68.
McEwen, B. S., Albeck, D., Cameron, H., Chao, H. M., Could. E., Hastings, N., Kuroda, Y.,
Luine, V., Magarinos, A. M., McKittrick, C. R., Orchinik, M.. Pavlides, C., Vaher. P.,
Watanabe, Y., and Weiland, N. (1995). Stress and the brain: A paradoxical role for
adrenal steroids. Vitam. Horm. ( N .Y . ) 14, 371-402.
McGaugh, J. L.. and Gold, P. E. (1989). Hormonal modulation of memory. In “Psychoneuroen-
docrinology” (F. R. Brush and S. Levine, eds.). pp. 305-340. Academic Press, San
Diego. CA.
McGrady. A. V. (1984). Effects of psychological stress on male reproduction: A review. Arch.
Androl. 13, 1-7.
McGuire. M. T., Brammer, G. L.. and Raleigh, M. J . (1986). Resting cortisol levels and the
emergence of dominance status among male vervet monkeys. Horm. Behav. 10,285-292.
Mendoza, S. P., and Mason, W. A. (1986). Contrasting responses to intruders and to involuntary
separation by monoganous and polygynous New World monkeys. Physiol. Behav. 38,
795-801.
Mendoza, S. P.. Coe, C. L., Lowe, C. L., and Levine, S. (1979). The physiological response
to group formation in adult male squirrel monkeys. Psychoneuroendocrinology 3,221-229.
Miczek, K. A,, Thompson, M. L., and Tornatzky, W. (1991). Subordinate animals. Behavioral
and physiological adaptations and opiod telerance. In “Stress, Neurobiology and Neuroen-
docrinology” (M. R. Brown, G. F. Koob, and C. Rivier, eds.), pp. 323-357. Dekker,
New York.
Miller, M. W., Hobbs. N. T., and Sousa, M. C. (1991). Detecting stress responses in Rocky
Mountain bighorn sheep (Ovis canadensis canadensis): Reliability of cortisol concentra-
tions in urine and feces. Can. J. Zool. 69, 15-24.
Moberg, G. P. (1987). Influence of the adrenal axis upon the gonads. Oxford Rev. Reprod.
Biol. 9, 456-496.
Monaghan, E. P., and Glickman, S. E. (1992). Hormones and aggressive behavior. In “Behav-
ioral Endocrinology” (J. B. Becker, S. M.. Breedlove, and D. Crews, eds.), pp. 261-286.
MIT Press, Cambridge. MA.
Monder, C., Sakai, R. R., Miroff, Y..Blanchard, D. C., and Blanchard, R. J. (1994). Reciprocal
changes in plasma corticosterone and testosterone in stressed male rats maintained in a
visible burrow system: Evidence for a mediating role of testicular 11 beta-hydroxysteroid

dehydrogenase. Endocrinology (Baltimore) 134, 1193-1 198.
Monjan, A. A. (1981). Stress and immunological competence: Studies in animals. In “Psycho-
neuroimmunology” (R. Ader, ed.). pp. 185-228. Academic Press, New York.
Moore, J., and Gotelli, N. J. (1990). Phylogenetic perspective on the evolution of altered host
behaviours: A critical look at the manipulation hypothesis. In “Parasitism and Host
Behaviour” (C. J. Barnard and J. M. Behnke, eds.), pp. 193-233. Taylor & Francis,
London.
Mormtde, P., Lemaire, V., Castanon, N., Dulluc. J., Laval, M., and Le Moal, M. (1990).
Multiple neuroendocrine responses to chronic social stress: Interaction between individual
and characteristic situational factors. Physiol. Behuv. 47, 1099-1 105.
Moynihan, J. A., Brenner, G. J., Cocke, R., Karp, J. D., Breneman, S. E.. Dopp, J. M., Ader,
R., Cohen, N., Grota, L. J., and Felten, S. Y. (1994). Stress-induced modulation of immune
function in mice. In “Handbook of Human Stress and Immunity” (R. Glaser and J. K.
Kiecolt-Glaser, eds.), pp. 1-22. Academic Press, San Diego, CA.
Munck, A,. and Guyre, P. M. (1991). Glucocorticois and immune function. In “Psychoneuroim-
munology” (R. Ader. D. L. Felten, and N. Cohen, eds.). pp. 429-474. Academic Press,
San Diego, CA.
Munck, A., Guyre, P. M., and Holbrook, N. J. (1984). Physiological functions of glucocorticoids
in stress and their relation to pharmacological actions. Endocr. Rev. 5, 25-45.
Myers, K., and Poole, W. E. (1959). A study of the biology of the wild rabbit, Orycfolugus
cuniculus (L.), in confined populations. CSIRO Wildl. Res. 4, 14-27.
Myers, K., and Poole, W. E. (1962). A study of the biology of the wild rabbit, Orycfolagus
cuniculus (L.) in confined populations. 111. Reproduction. Aust. J. 2001.10, 225-267.
Myers, K., Hale, C. S., Mykytowycz, R.. and Hughes, R. L. (1971). The effects of varying
density and space on sociality and health in animals. In “Behavior and Environment.
The Use of Space by Animals and Men” (A. H. Esser. ed.). pp. 148-187. Plenum,
New York.
Myers, M. J., and Murtaugh, M. P. (1995). “Cytokines in Animal Health and Disease.” Dekker,
New York.
Mykytowycz, R. (1958). Social behaviour of an experimental colony of wild rabbits, Oryclolu-
gus cuniculus (L.). I. Establishment of the colony. CSIRO Wildl. Res. 3, 7-25.
Mykytowycz, R. (1959a). Social behaviour of an experimental colony of wild rabbits, Orycfolu-
gus cuniculus (L.). 11. First breeding season. CSIRO Wildl. Res. 4, 1-13.
Mykytowycz, R. (1959b). Social behaviour of an experimental colony of wild rabbits, Orycfolu-
gus cuniculus (L.). 111. Second breeding season. CSIRO Wildl. Res. 5, 1-20.
Mykytowycz, R. (1961). Social behaviour of an experimental colony of wild rabbits, Orycfolu-
gus cuniculus (L.). IV. Conclusions: Outbreak of myxomatosis, third breeding season,
and starvation. CSIRO Wildl. Res. 6 , 142-155.
Nelson, R. J., and Demas, G. E. (1996). Seasonal changes in immune function. Q. Rev. Biol.
71, 511-548.
Nowell, N. W. (1980). Adrenocortical function in relation to mammalian population densities
and hierarchies. In “General. Comparative and Clinical Endocrinology of the Adrenal
Cortex” (I. Chester Jones and I. W. Henderson, eds.), Vol. 3, pp. 349-393. Academic
Press, London.
O’Grady. M. P., and Hall, N. R. (1991). Long-term effects of neuroendocrine-immune interac-
tions during early development. In “Psychoneuroimmunology” (R. Ader, D. L. Felten.
and N. Cohen, eds.), pp. 561-572. Academic Press, San Diego, CA.
Olsen, N. J., and Kovacs, W. J. (1996). Gonadal steroids and immunity. Endocr. Rev. 17,
369-384.
Orr, T. E., and Mann, D. R. (1992). Role of glucocorticoids in the stress-induced suppression
of testicular steroidogenesis in adult male rats. Horm. Behav. 26, 350-363.
Ottenweller, J. E., Tapp, W. N., Burke, J. M.. and Natelson, B. H. (1985). Plasma cortisol
and corticosterone concentrations in the golden hamster (Mesocricefusaurafus).Life Sci.
37, 1551-1558.
Paavonen, T. (1994). Hormonal regulation of immune responses. Annals of Medicine 26,
255-258.
Packer, C., Collins, D. A,, Sindimwo, A,, and Goodall, J. (1995). Reproductive constraints
on aggressive competition in female baboons. Nature (London) 373, 60-63.
Pedersen, C. A,, and Prange, A. J., Jr. (1985). Oxytocin and mothering behavior in the rat.
Pharmacol. Ther. 28, 287-302.
Perrson, L., Gullberg, B., Hanson, B. S., Moestrup, T., and Ostergren, P. 0. (1994). HIV
infection: social network, social support, and CD4 lymphocyte values in infected homosex-
ual men in Malmo, Sweden. J. Epidemiol. Cornmunify Health 48, 580-585.
Phoenix, C. H., Goy, R. W., Gerall, A. A,, and Young, W. C. (1959). Organizing action of
prenatally administered testosterone propionate on the tissues mediating mating behavior
in the female guinea pig. Endocrinology (Baltimore) 65, 369-382.
Plaut, M., and Friedman. S. B. (1981). Psychosocial factors in infectious disease. In “Psychoneu-
roimmunology” (R. Ader, ed.), pp. 3-30. Academic Press, New York.
Plaut, S. M., Ader, R., Friedman, S. B., and Ritterson, A. L. (1969). Social factors and
resistance to malaria in the mouse: Effects of group vs. individual housing on resistance
to Plasmodium berghei infection. Psychosom. Med. 31, 536-552.
Politch, J. A., and Leshner, A. I. (1977). Relationship between plasma corticosterone levels
and levels of aggressiveness in mice. Physiol. Behav. 19, 775-780.
Popova, N. K., and Naumenko, E. V. (1972). Dominance relations and the pituitary-adrenal
system rats. Anim. Behav. 20, 108-111.
Price, V.A. (1982). “Type A Behavior Pattern. A Model for Research and Practice.” Academic
Press, New York.
Raab, A,, Dantzer, R., Mormede, M. P.. Taghzouti, K., Simon, H.,and Le Moal. M. (1986).
Behavioural, physiological and immunological consequences of social status and aggres-
sion in chronically coexistingresident-intruderdyads of male rats. Physiol. Behav. 36,
223-228.
Rabin. B. S.. Kusnecov, A,, Shurin, M., Zhou, D., and Rasnick, S. (1994). Mechanistic aspects
of stressor-induced immune alteration. I n “Handbook of Human Stress and Immunity”
(R. Glaser and J. K. Kiecolt-Glaser, eds.), pp. 23-51. Academic Press, San Diego, CA.
Rabin, D., Gold, P. W.. Margioris, A. N., and Chrousos, G. P. (1988).Stress and reproduction:
Physiologic and pathophysiologic interactions between the stress and the reproductive
axes. Adv. Exp. Med. Biol. 245,377-387.
Ratcliffe. H. L., Luginbuhl, H., Schnarr, W. R., and Chacko, K. (1969). Coronary arteriosclero-
sis in swine: Evidence of a relation to behavior. J. Comp. Physiol. Psychol. 68,385-392.
Rau, M. E. (1983). Establishment and maintenance of behavioural dominance in male mice
infected with Trichinella spiralis. Parasitology 86, 311-318.
Read, A. F. (1990). Parasites and the evolution of host sexual behaviour. I n “Parasitism
and Host Behaviour” (C. J. Barnard and J. M. Benke, eds.), pp. 117-157. Taylor &
Francis, London.
Reite, M. L., Harbeck, R. J., and Hoffman, A. (1981). Altered cellular immune response
following peer separation. Life Sci. 29, 1133-1136.
Riad-Fahmy, D.. Read, G. F., Walker, R. F., and Griffiths, K. (1982). Steroids in saliva for
assessing endocrine function. Endocr. Rev. 3, 367-395.
Riechert, S. E. (1988). The energetic costs of fighting. Am. 2001.28,877-884.
Riley, V. (1981). Psychoneuroendocrine influences on immunocompetence and neoplasia.

Science 212, 1100-1109.
Rivier, C. (1991). Neuroendocrine mechanisms of anterior pituitary regulation in the rat
exposed to stress. In “Stress, Neurobiology and Neuroendocrinology” (M. R. Brown.
G. F. Koob. and C. Rivier, eds.), pp. 119-136. Dekker, New York.
Rose, R. (1985). Psychoendorcinology. In “Williams Textbook of Endocrinology” (J. Wilson
and D. Foster, eds.), pp. 653-681. Saunders, Philadelphia.
Rose, R., Holaday, J., and Bernstein, I. (1971). Plasma testosterone, dominance rank and
aggressive behaviour in male rhesus monkeys. Nature (London) 231, 366-368.
Rose, R. M.. Gordon. T. P., and Bernstein. 1. S. (1972). Plasma testosterone levels in the
male rhesus: Influences of sexual and social stimuli. Science 178, 643-645.
Rose, R. M., Bernstein, 1. S., Gordon. T. P., and Catlin. S. F. (1974). Androgen and aggression:
A review and recent findings in primates. In “Primate Aggression, Territoriality and
Xenophobia” (R. L. Halloway. ed.). pp. 275-304. Academic Press, New York.
Rose. R. M., Bernstein. I. S., and Gordon, T. P. (1975). Consequences of social conflict on
plasma testosterone levels in rhesus monkeys. Psychosom. Med. 37, 50-61.
Rosenblum. L. A., and Plimpton, E. H. (1981). The infant’s effort to cope with separation.
In “The Uncommon Child” (M. Lewis and L. A. Rosenblum, eds.), pp. 225-257. Plenum.
New York.
Rosenfeld, P., Wetmore, J. B., and Levine, S. (1992). Effects of repeated maternal separations
on the adrenocortical response to stress of preweanling rats. Physiul. Behav. 52,787-791.
Rosenman, R. H. (1986). Current and past history of Type A behavior pattern. In “Biological
and Psychological Factors in Cardiovascular Disease” (T. H. Schmidt, T. M. Dembroski,
and G. Blumchen. eds.), pp. 15-40. Springer, Berlin.
Saad, M. B., and Bayle, J. D. (1985). Seasonal changes in plasma testosterone, thyroxine, and
cortisol levels in wild rabbits (Oryctolagus citniculus algirus) of Zembra island. Gen.
Comp. Endocrinol. 57,383-388.
Saboureau. M . , Laurent, G., and Boissin, J. (1977). Daily and seasonal rhythms of locomotory
activity and adrenal function in male hedghog (Erinaceus europaeus L.). J. Inrerdiscip.
Cycle Res. 10, 245-266.
Sachser, N. (1986). Different forms of social organization at high and low population densities
in guinea pigs. Behaviour 97, 253-272.
Sachser, N. (1987). Short-term responses of plasma norepinephrine, epinephrine. glucocorti-
coid and testosterone titers to social and non-social stressors in male guinea pigs of
different social status. Physiol. Behav. 39, 11-20.
Sachser, N. (1993). The ability to arrange with conspecifics depends on social experiences
around puberty. Physiol. Behav. 53,539-544.
Sachser, N. (1994a). “Sozialphysiologische Untersuchungen an Hausmeerschweinchen. Grup-
penstrukturen, soziale Situation und Endokrinium, Wohlergehen.” Parey, Berlin.
Sachser, N. (1994b). Social dominance and health in nonhuman mammals: A case study in
guinea pigs. In “Social Stratification and Socioeconomic Inequality” (L. Ellis. ed.). Vol.
2, pp. 113-121. Praeger. Westport, CT.
Sachser, N., and Kaiser, S. (1996). Prenatal social stress masculinizes the females’ behaviour
in guinea pigs. Physiol. Behav. 60, 589-594.
Sachser, N., and Lick, C. (1989). Social stress in guinea pigs. Physiol. Behav. 46, 137-144.
Sachser. N., and Lick, C. (1991). Social experience, behavior, and stress in guinea pigs. Physiol.
Behav. 50, 83-90.
Sachser, N., and Prove, E. (12186). Social status and plasma-testosterone titers in male guinea
pigs (Cavia aperea f: porcellits). Ethology 71, 103-1 14.
Sachser, N., and Renninger, S. V. (1993). Coping with social conflict: The role of social rearing
in guinea pigs. Ethol. EcoL Evol. 5, 65-74.
Sachser, N., Lick, C., and Stanzel, K. (1994). The environment, hormones, and aggressive
behaviour: A 5-year-study in guinea pigs. Psychoneuroendocrinology 19, 697-707.
Sachser, N., Diirschlag. M., and Hirzel, D. (1998). Social relationships and the management
of stress. Psychoneuroenr~ocrinology(in press).
Sade, D. S., Cushing, K., Cushing, P., Dunaif, A., Figueroa, J., Kaplan, R., Laver, C., Rhodes.
D., and Schneider. J. (1976). Population dynamics in relation to social structure on Cay0
Santiago. Yearb. Physiol. Anthropol. 20, 253-262.
Sapolsky, R. M. (1982). The endocrine stress-response and social status in the wild baboon.
Horm. Behav. 20,279-292.
Sapolsky. R.M. (1983). Individual differences in cortisol secretory patterns in the wild baboon:
Role of negative feedback sensitivity. Endocrinology (Baltimore) 113, 2263-2267.
Sapolsky, R. M. (1985a). Endocrine aspects of social instability in the olive baboon (Papio
anubis). A m . J. Primntol. 5, 365-379.
Sapolsky, R. M. (198Sb). Stress-induced suppression of testicular function in the wild baboon:
Role of glucocorticoids. Endocrinology (Baltimore) 116, 2273-2278.
Sapolsky, R. M. (1988). Individual differences and the stress response: Studies of a wild
primate. A d v . Exp. Med. B i d . 245, 399-411.
Sapolsky, R. M. (1990). Adrenocortical function, social rank, and personality among wild
baboons. Biol. Psychiatry 28, 862-878.
Sapolsky, R. M. (1991). Effects of stress and glucocorticoids on hippocampal neuronal survival.
I n “Stress, Neurobiology and Neuroendocrinology” (M. R. Brown, G. F. Koob, and C.
Rivier. eds.), pp. 293-322. Dekker, New York.
Sapolsky, R. M. (1992). “Stress, the Aging Brain, and the Mechanisms of Neuron Death.”
MIT Press, Cambridge, MA.
Sapolsky, R. M., and Mott, G. E. (1987). Social subordination in wild baboons is associated
with suppressed high density lipoprotein-cholesterol concentrations: The possible role of
chronic stress. Endocrinology (Baltimore) 121, 1605-1610.
Sassenrath, E. N. (1970). Increased adrenal responsiveness related to social stress in rhesus
monkeys. Horm. Behav. 1,283-298.
Schaftenaar, W., Buiter, R. M., and Dieleman, S. J., eds. 1992. Proceedings from the First
International Symposium on Faecal Steroid Monitoring in Zoo Animals. 1992 Feb 28-29,
Rotterdam, The Netherlands.
Schiml, P. A,, Mendoza, S. P., Saltzman, W., Lyons, D. M., and Mason, W. A. (1996). Seasonal-
ity in squirrel monkeys (Saimiri sciurew): Social facilitation by females. Physiol. Behav.
60, 1105-1113.
Schoenbach, V. C., Kaplan, B. H.. Fredman, L.. and Kleinbaum, D. G . (1986). Social ties and
mortality in Evans County, Georgia. Am. J. Epidemiol. 123, 577-591.
Schonheiter, R. (1992). Der Einfluss von Jahreszeit, Geschlecht und sozialen Faktoren auf
Verhalten und physiologische Parameter bei adulten Europaischen Wildkaninchen (Oryc-
tolagus cuniculus L.). Unpublished Doctoral Dissertation, University of Bayreuth, Bay-
reuth, Germany.
Schuhr, B. (1987). Social structure and plasma corticosterone level in female albino ice.
Physiol. Behav. 40,689-693.
Schuurman, T. (1981). Endocrine processes underlying victory and defeat in the male rat.
Unpublished Doctoral Dissertation, Rijksuniversiteit, Groningen, Netherland.
Schuurs, A. H. W. M.. and Verheul, H. A. M. (1990). Effects of gender and sex steroids on
the immune response. J. Steroid. Biochem. 35, 157-172.
128 DIETRICH VON HOLS?
Scott, J. P., and Fredericson. E. (1951). The causes of fighting in mice and rats. Physiol. Zool.
24,273-309.
Seligman, M. E. (1975). “Helplessness: O n Depression, Development, and Death.” Freeman,
San Francisco.
Selye, H. (1936). A syndrome produced by diverse nocuous agents. Nature (London) 138,
32-34.
Selye. H. (1950). “Stress.” Acta, Montreal.
Selye, H. (1952). “The Story of the Adaptation Syndrome.” Acta, Montreal.
Selye, H. (1976). “Stress in Health and Disease.” Butterworth, Boston.
Selye, H. (1981). The stress concept today. In “Handbook on Stress and Anxiety. Contempo-
rary Knowledge, Theory, and Treatment” ( I . L. Kutash and L. B. Schlesinger. eds.). pp.
127-143. Jossey-Bass, San Francisco.
Shapiro, L. E., and Dewsbury, D. A. (1986). Male dominance, female choice and male copula-
tory behavior in two species of voles (Microtus ochrogasrer and Microtus montanus).
Behav. Ecol. Sociohiol. 18, 267-274.
Sherman, P. W., Jarvis, J. U. M., and Alexander, R. A,, eds. (1991). “The Biology of the
Naked Mole-rat.” Princeton University Press, Princeton. NJ.
Shively, C.. and Kaplan. J. (1984). Effects of social factors on adrenal weight and related
physiology of Macaca fascicularis. Physiol. Behav. 33, 777-782.
Silk, J. B., Clark-Wheatley, C. B., Rodman, P. S.. and Samuels. A. (1981). Differential reproduc-
tive success and facultative adjustment of sex ratios among captive female bonnet ma-
caques (Macaca radiata). Anim. Behav. 29, 106-120.
Smelik, P. G., and Vermes, I. (1980). The regulation of the pituitary-adrenal system in mam-
mals. I n “General, Comparative and Clinical Endocrinology of the Adrenal Cortex” ( I .
Chester Jones and 1. W. Henderson, eds.), pp. 1-55. Academic Press, London.
Snyder, R. L. (1968). Reproduction and population pressures. Prog. Physiol. Psychol. 2,
119-160.
Solomon, G. F., and Amkraut, A. A. (1981). Psychoneuroendocrinological effects on the
immune system. Annu. Rev. Microhiol. 35, 155-184.
Southern. H. N. (1940). The ecology and population dynamics of the wild rabbit (Oryctolagus
cuniculus). Ann. Appl. B i d . 27, 509-526.
Stalker, A,, Hermo, L., and Antakly, T. (1989). Covalent affinity labeling, radioautography,
and immunocytochemistry localize the glucocorticoid receptors in rat testicular Leydig
cells. Am. J. Anar. 186, 36Y-377.
Stefanski, V., and Ben-Eliyahu, S. (1996). Social confrontation and tumor metastasis in rats:
Defeat and R-adrenergic mechanisms. Physiol. Behav. 60, 277-282.
Stafanski, V., and Hendrichs, H. (1996). Social confrontation in male guinea pigs: Behavior,
experience, and complement activity. Physiol. Behav. 60, 235-241.
Stefanski, V., Solomon, G. F.. Kling, A. S . , Thomas, J., and Plaeger, S. (1996). Impact of social
confrontation on rat CD4 T cells bearing different CD45R isoforms. Brain, Behavior, and
Immunity 10, 364-379.
Sternberg, E. M. (1988). Monokines, lymphokines, and the brain. In “The Year in Immunology
1988” (J. M. Cruse and R. E. Lewis, eds.). Vol. 5, pp. 205-217. Karger, Basel.
Stoddard, S. L. (1991). Hypothalamic control and peripheral concomitants of the autonomic
defense response. In “Stress, Neurobiology and Neuroendocrinology” (M. R. Brown,
G. F. Koob, and C. Rivier, eds.), pp. 231-253. Dekker, New York.
Stohr, W. (1986). Heart rate in tree shrews and its persistent modification by social contact.
In “Biological and Psychological Factors in Cardiovascular Disease’’ (T. H. Schmidt, T.
M. Dembroski, and G. Bliimchen, eds.), pp. 508-516. Springer, Berlin.
Stohr, W. (1988). Longterm heart rate telemetry in small mammals: A comprehensive approach
as a prerequisite for valid results. Physiol. Behav. 43, 567-576.
Stumpfe, K. (1973). “Der psychogene Tod.” Hippokrates, Stuttgart.
Suomi, S. J. (1976). Factors affecting responses to social separation in rhesus monkeys. In
“Animal Models in Human Psychobiology” ( G . Serban and A. Kling, eds.), pp. 9-26.
Plenum, New York.
Takahashi, L. K. (1990). Hormonal regulation of sociosexual behavior in female mammals.
Neurosci. Biohehav. Rev. 14, 403-413.
Takahashi, L. K. (1991). Ontogeny of stress-induced ultrasonic vocalization and pituitary-
adrenal hormone secretion in preweanling Norway rats. Psychol. Rec. 41, 159-174.
Tecoma, E. S., and Huey, L. Y. (1985). Psychic stress and the immune system. Life Sci.
36, 1799-1812.
Temoshok, L., and Peeke, H. V. S. (1988). Individual behavior differences related to induced
tumor growth in the female Syrian hamster: Two studies. Inr. J. Neurosci. 38, 199-209.
Thierry, B.. Steru, L., Chermat, R., and Simon, P. (1984). Searching-waiting strategy: A
candidate for an evolutionary model of depression? Behav. Neirral B i d . 41, 180-189.
Thiessen. D. D., and Rice, M. (1976). Mammalian scent gland marking and social behavior.
Psychol. Bull. 83, 505-539.
Thoenen, H., Mueller. R. A,, and Axelrod, J. (1969). Trans-synaptic induction of adrenal
tyrosine hydroxylase. J. Pharmacol. Exp. Ther. 169,249-254.
Toates, F. (1987). The relevance of models of motivation and learning to animal welfare. In
“Biology of Stress in Farm Animals: An Integrative Approach” (P. R. Wiepkema and
P. W. M. van Adrichem, eds.), pp. 151-186. Martinus Nijhoff, Dordrecht, The Netherlands.
Tobach, E., and Bloch, H. (1955). A study of the relationship between behavior and susceptibil-
ity to tuberculosis in rats and mice. Adv. Tuberc. Res. 6, 62-89.
Tobach, E., and Bloch, H. (1958). Effects of stress by crowding prior to and following tubercu-
lous infection. A m . J. Physiol. 187, 399-402.
Toft, C. A., and Karter, A. J. (1990). Parasite-host coevolution. Trends Ecol. Evol. 5,326-329.
Turner, J. W. (1984). Sex differences in the response to neonatal steroid treatment in the
mongolian gerbil. Physiol. Behav. 33, 173-176.
Unden, A. L., Orth-Gomer, K., and Elofsson, D. (1991). Cardiovascular effects of social
support in the working place: Twenty-four hour ECG monitoring of men and women.
Ungar, A,. and Phillips, J. H. (1983). Regulation of the adrenal medulla. Physiol. Rev. 63,
787-843.
Uno, H., Eisele, S., Sakai, A., Shelton, S., Baker, E., DeJesus, O., and Holden, F. (1994).
Neurotoxicity of glucocorticoids in the primate brain. Horm. Behav. 28, 336-348.
Ursin, H., and Olff. M. (1993). The stress response. In “Stress. From the Synapse to the
Syndrome” (S. C. Stanford and P. Salmon, eds.), pp. 3-22. Academic Press, London.
Ursin, H., Baade. E., and Levine, S. (1978). “Psychobiology of Stress.” New York: Academic
Press, New York.
Uvnas-Moberg, K. (1994). Role of efferent and afferent vagal nerve activity during reproduc-
tion: Integrating function of oxytocin on metabolism and behavior. Psychoneuroendocrin-
ology 19, 687-695.
Vander, A. J., Henry, J. P., Stephens, P. M., Kay, L. L., and Mouw, D. R. (1978). Plasma
renin activity in psychosocial hypertension of CBA mice. Circ. Res. 42, 496-502.
van Eck, M., Berkhof, H., Nicolson. N.. and Sulon, J. (1996). The effects of perceived stress,
traits, mood states, and stressful daily events on salivary cortisol. Psychosom. Med. 58,
447-458.
van Oortmerssen, G. A., Benus. I., and Dijk, D. J. (1985). Studies in the wild house mice:
Genotype-environment interactions for attack latency. Neth. J. Zool. 35, 155-169.
van Oortmerssen, G. A., Dijk, D. J., and Schuurman, T. (1987). Studies in wild house mice
11. Testosterone and aggression. Horm. Behav. 21, 139-152.
Vellucci, S. (1990). Primate social behavior-anxiety or depression. fharmacol. Ther. 47,
167-180.
Vessey, S. H. (1964). Effects of grouping on levels of circulating antibody levels in mice. Proc.
Soc. Exp. Biol. Med. 115, 252-255.
Vining, R. F., and McGinley, R. A. (1986). Hormones in the saliva. Crit. Rev. Clin. Lab. Sci.
23, 95-146.
von Holst, D. (1969). Sozialer Stress bei Tupajas (Tupai belangeri). Die Aktivierung des
sympathischen Nervensystems und ihre Bezihung zu hormone11ausgelosten ethologischen
und physiologischen Veranderungen. Z . Vergl. Physiol. 63, 1-58.
von Holst, D. (1972a). Renal failure as the cause of death in Tupaia belangeri exposed to
persistent social stress. J. Con7p. Physiol. 78, 236-273.
van Holst, D. (1972b). Die Funktion der Nebennieren mannlicher Tupaia belangeri. J. Cornp.
Physiol. 78, 289-306.
von Holst, D. (1985a). Coping behaviour and stress physiology in male tree shrews (Tupaia
belangeri). Fortschr. Zool. 31, 461-470.
von Holst, D. (1985b). The primitive eutherians 11: A case study of the tree shrew, Tupaia
belangeri. In “Social Odours in Mammals” (R. E. Brown and D. M. McDonald. eds.),
Vol. 1, pp. 155-216. Clarendon Press, Oxford.
von Holst, D. (1986a). Vegetative and somatic components of tree shrew’s behavior. J. Auton.
Nerv. Syst., Suppl., pp. 657-670.
von Holst, D. (1986b). Psychosocial stress and its pathophysiological effects in tree shrews
(Tccpaia belangeri). In “Biological and Psychological Factors in Cardiovascular Disease”
(T. H. Schmidt, T. M. Dembroski, and G. Bliimchen. eds.). pp. 508-516. Springer, Berlin.
von Holst, D. (1987). Physiologie sozialer Interaktionen-Sozialkontakte und ihre Auswir-
kungen auf Verhalten sowie Fertilitat von Tupajas. Physiol. Aktuel. 3, 189-208.
von Holst, D. (1994). Auswirkungen sozialer Kontakte bei Saugetieren. Biol. Unserer Zeit
24, 164-174.
Wade, S. E. (1991). An optimized method for measurement of salivary corticosteroids. In
“Assessment of Hormones and Drugs in Saliva in Biobehavioral Research” (C. Kirsch-
baum, C., F. R. Read, and D. H. Hellhammer, eds.), pp. 3-17. Hogrefe & Huber, Seat-
tle, WA.
Walker. M. L., Gordon, T. P., and Wilson, M. E. (1983). Menstrual cycle characteristics of
sexually breeding rhesus monkeys. Biol. Reprod. 29, 841-848.
Ward, I. L. (1984). The prenatal stress syndrome: Current status. fsychoneicroendocrinology
9, 3-11.
Wasser, S. K. (1996). Reproductive control in wild baboons measured by fecal steroids. Biol.
Reprod. 55, 393-399.
Watson, A., and Moss, R. (1970). Dominance, spacing behaviour and aggression in relation
to population limitation in vertebrates. In “Animal Populations in Relation to their Food
Resources” (A. Watson, ed.), pp. 167-218. Blackwell, Oxford.
Wedekind, C. (1994). Mate choice’ and maternal selection for specific parasite resistance
before, during and after fertilization. fhilos. Trans. R. Soc. London, Ser. B. 346,303-31 1.
Weiner. H. (1977). “Psychobiology and Human Diseases.” Elsevier, New York.
Weiner, H. (1991). Behavioral biology of stress and psychosomatic medicine. In “Stress,
Neurobiology and Neuroendocrinology” (M. R. Brown, G. F. Koob, and C. Rivier, eds.),
pp. 23-51. Dekker, New York.
Weinmann. C. J., and Rothman. A. H. (1067). Effects of stress upon acquired immunity to
the dwarf tapeworm, Hymenolepis nana. Exp. Parasitol. 21, 61-67.
Weiss, J. M. (1971). Effects of punishing the coping response (conflict) on stress pathology
in rats. 1. Conzp. Physiol. Psychol. 77, 1-13.
Weiss, J. M. (1972). Influence of pyschological variables on stress-induced pathology. In
“Physiology, Emotion and Psychosomatic Illness” (R. Porter and J. Knight, eds.), Ciba
Found. Symp. 8, pp. 253-279. Elsevier, Amsterdam.
Weiss. J. M. (1984). Behavioral and psychological influences on gastrointestinal pathology:
Experimental techniques and findings. In “Handbook of Behavioral Medicine” ( W. D.
Gentry, ed.), pp. 174-221. Guilford. New York.
Wenar, C. (1983). “Psychopathology from Infancy through Adolescence. A Developmental
Approach.” Random House, New York.
White, P. J. (1986). Female discrimination of male dominance by urine cues in hamsters.
Physiol. Behav. 37, 273-277.
Whitman, D. C.. and Albers. H. E. (1995). Role of oxytocin in the hypothalamic regulation
of sexual receptivity in hamsters. Brain Res. 680, 73-79.
Wiener, S. G., Lowe, E. L., and Levine. S. (1992). Pituitary-adrenal response to weaning in
infant squirrel monkeys. Psychobiology 20, 65-70.
Williams, J. R., Carter, C. S., and Insel, T. (1992). Partner preference in the development in
female prairie voles is facilitated by mating and central infusion of oxytocin. Ann. N. Y.
Acad. Sci. 652, 487-489.
Williams, J. R., Insel, T. R., Harbaugh, C. R.. and Carter, C. S. (1994). Oxytocin administered
centrally facilitates formation of a partner preference in female prairie voles (Mirrorus
ochrogaster). J . Neuroendocrinol. 6, 247-250.
Wilson. M. E., Gordon, T. P., and Bernstein, 1. S. (1978). Timing of births and reproductive
success in rhesus monkey social groups. J . Med. Primatol. 7, 202-212.
Winslow, J. T., Shapiro, L., Carter, C. S., and Insel, T. R. (1993). Oxytocin and complex social
behavior: Species comparison. Psychopharmacol. Bull. 29,409-414.
Wise, D. A,, Eldred. N. L., McAfee, J., and Lauber, A. (1985). Litter deficits of socially
stressed and low status hamster dams. Physiol. Behav. 35, 775-777.
Woolley, P. (1966). Reproduction in Antechinus spp. and other dasyurid marsupials. Symp.
Zool. Soc. London 15,281-294.
Wurtman. R. J., and Axelrod, J. (1966). Control of enzymatic synthesis of adrenaline in the
adrenal medulla by adrenal corticoid steroids. J. Clin. Chem. 241, 2301-2305.
Yu, G. Z., Okutani. F., Takahashi, S., and Higuchi, T. (1996). The olfactory bulb: A critical
site for oxytocin in the induction of maternal behaviour in the rat. Neurosciences 72,1083-
1988.
Zalcman. S., Kerr, L., and Anisman, H. (1991). Immunosuppression elicited by stressors and
stress-related odors. Brain, Behav., Immrcnol. 5, 262-274.
Zobelein. H. (1996). Sozialer Rang, Reproduktionserfolg und Fitness mannlicher Europaischer
Wildkaninchen Oryctolagtcs cuniculus L. Unpublished Doctoral Dissertation, University
of Bayreuth, Bayreuth, Germany.
Zuckerman, S. (1932). “The Social Life of Monkeys and Apes.” Harcourt, New York.
Zuk, M. (1994). Immunology and the evolution of behavior. I n “Behavioral mechanisms in
Evolutionary Ecology” (L. A. Real, ed.), pp. 354-368. University of Chicago Press,
Chicago.
Zwilling. B. S. (1994). Neuroimmunomodulation of macrophage function. I n “Handbook of
Human Stress and Immunity” (R. Glaser and J. K. Kiecolt-Glaser, eds.), pp. 53-76.
ADVANCES IN THE STUDY OF BEHAVIOR, VOL. ?I
Stress and Immune Defense
VICTORAPANIUS
DEPARTMENTOF BIOLOGICAL
SCIENCES
FLORIDAINTERNATIONALUNIVERSITY
PARK,MIAMI,FLORIDA
UNIVERSITY 33199
I. INTRODUCTION
It is a clichC to state that stress reduces immunocompetence. Although this

phrase is widely believed and supported by incontrovertible evidence, the
biological reality is much more complex and the topic remains an active
area of research by ethologists, endocrinologists, neurobiologists, immunol-
ogists, pathologists, and parasitologists. Interest in the relationship between
behavior and immune function has spawned the field of psychoneuroimmu-
nology, with several new journals reporting recent research. A resurgence
of interest in the role of parasites in the evolution of host life histories has
provided an impetus for integrating this biomedical information into a
Darwinian framework. The purpose of this review is to demonstrate the
simplistic nature of the phrase “stress suppresses immunity” and to discuss
how stress alters immunocompetence. A consideration of the words stress
and immunocompetence will show that these terms can be twisted to serve
any purpose. Progress in this field requires precise functional definitions
of stress and immunocompetence with an appreciation of the multifactorial
and nonlinear nature of these subjects. For example, the stress response
suppresses particular immunological mechanisms while enhancing others.
Stress, which can be immunosuppressive in the short term, can also enhance
immunological reactivity in the long term. It is hypothesized that these
immunological mechanisms have differential costs and benefits and that
these shifts enhance survival based on energetic considerations.
There is no shortage of reviews concerning the endocrinological aspects
of the stress response (Chrousos and Gold, 1992; Sapolsky, 1992), the
endocrine regulation of immunity (Blalock, 1989, 1994; Reichlin, 1993;
Leonard and Song, 1996), and the relationship between stress and immunity
(Khansari et al., 1990; Stein and Miller, 1993; Daynes et al., 1995; Besedov-
sky, 1996; Friedman et al., 1996; Ottaviani and Franceschi, 1996). The
133
0065-3454/’)8 $25 00
134 VICTOR APANIUS
purpose of this review is to highlight certain aspects of the relationship

between stress and the vertebrate immune system that cause generalizations
to be biologically simplistic. An additional goal is a synthesis based on
energetic considerations that may provide a framework for understanding
the adaptive significance of this phenomenon. It is hoped that the reader
will appreciate the daunting complexity of the neuroendocrine-immune
axis but will also be stimulated to think about the evolution of this vital
system from a functional rather than purely mechanistic viewpoint.
11. THENATURE
OF STRESS
Almost by definition, living organisms respond homeostatically to envi-

ronmental variation, whether it is the subtle variations in salinity experi-
enced by an estuarine crustacean or the transhemispheric migration of
shorebirds. The delineation between a normal homeostatic response and
a stress response is problematic. Typically, physiological parameters, for
example, plasma glucocorticoid concentration, are used to delineate the
range of environmental or social conditions that are stressful to an animal.
There are a number of problems with this approach, including the fact that
it is tautological. That is because we define stress as the state where stress
hormones are elevated above some arbitrarily defined threshold.
A more quantitative biochemical approach uses the ratio of energy-
charged adenine nucleotides to all such adenine nucleotides to give a mea-
sure called the adenine energy charge (AEC):
(ATP) + 1/2(ADP)
AEC =
(ATP) + (ADP) + (AMP)
(Hochachka and Somero, 1984). It has been argued that this energetically
based measurement reflects the physiological performance of the organism
in particular environments and provides a universal metric for comparisons
across taxa. Individuals living in stressful environments or experiencing
stressful events would have a lower AEC, refleding a lower biosynthetic
capacity for growth, reproduction, or storage. Although A E C can be conve-
niently measured, it is more problematic to relate it to survival rates or
lifetime reproductive success.
For the evolutionary ecologist or geneticist, a definition of stress should
logically incorporate the individual’s (inclusive) fitness. A stressful environ-
ment can be defined as one in which conditions do not allow population
persistence through local reproduction. Species ranges are limited because
of physical or biotic factors that would intrinsically be stressful to individu-
als. Selection would favor phenotypic plasticity as a response to the unpre-
STRESS A N D IMMUNE DEFENSE 135
dictable biophysical and biotic regimes that occur not only at the species
margin but throughout its range (Hoffmann and Parsons, 1991). Phenotypic
plasticity is one way that the stress can be detected but phenotypic plasticity
will also depend on the extent of gene flow.
It can be argued that all of these approaches simply provide operational
definitions, but the root of the problem is that stress is often considered
to be a dichotomous state, which is defined on the basis of continuously
varying physiological parameters. This allows proliferation of the situations
to which the word stress is applied to the point where it is meaningless.
Because glucocorticoid levels are generally higher during the active phase
than during sleep, one can talk about the stress of daily life. Glucocorticoid
levels are elevated during reproduction, allowing one to speak of the stress
of reproduction. Glucocorticoid levels are elevated in migrant songbirds,
suggesting that there is a stress of migration. Parasitic infections are associ-
ated with increased glucocorticoid levels, implying a stress of infection.
Glucocorticoid levels are elevated in subdominant individuals following
the initial agonistic encounters that establish dominance hierarchies, an
example of social stress. Yet all of these conditions are typical events for
these animals. Therefore, a useful definition of stress needs to capture the
intensity and transient nature of these conditions. Operational definitions of
stress are further complicated by the neuroendocrine adaptation to stressful
stimuli, which is manifested by the “coping” strategies of individual animals
to escapable versus nonescapable stress. Nonetheless, laboratory and field
experiments can informatively test hypotheses about the adaptive signifi-
cance of the stress response when stressful stimuli are clearly defined and
consistently applied to the subjects.
The animal’s capacity to respond to stressful situations has a genetic
component (Hoffmann and Parsons, 1991). The magnitude of the general-
ized stress response has been altered through artificial selection in poultry.
Turkeys have been selected for high and low responsiveness to cold stress
(Brown and Nestor, 1973). Chickens have been selectively bred for high
and low corticosterone production in response to social stress (Gross and
Siegel, 1985). As expected, the birds that responded to stress with higher
circulating levels of corticosterone showed a greater susceptibility to virus-
induced tumors and coccidiosis and increased ectoparasite populations.
This increased susceptibility to malignant and infectious disease in the birds
responding with high corticosterone levels was associated with lowered
immunocompetence, as measured by antibody and cell-mediated responses
(defined later). This research on domestic birds not only demonstrates a
genetic basis for the magnitude of the stress responses but also demonstrates
its lability in response to selection.
136 VICTOR APANIUS
Stress-induced immunosuppression is a paradoxical phenomenon. The

unfettered operation of the immune system is crucial to an individual’s
health, vigor, and, ultimately, survival. Yet, at the time when an animal’s
survival is challenged by extreme environmental change or by tremendous
physical exertion, the immune system appears to be downregulated. Pro-
longed stress can cause the premature regression of the primary lymphoid
organs and effectively truncate the ontogeny of immunity. Prolonged stress
can also induce atrophy of the secondary lymphoid organs, rendering the
animal more susceptible to infectious and malignant disease. Even acute
episodes of stress can reduce the effectiveness of the immune system to
control opportunistic parasites. It is likely that immunity is temporarily
downregulated to make nutrients available for other organismal processes
that have a higher priority, that is, the nervous system and the musculature.
The mechanistic basis of stress-induced immunosuppression is reasonably
well understood. The physiological basis for re-allocating nutrients from
lymphoid tissue to other somatic compartments is poorly understood. The
evolutionary basis for stress-induced immunosuppression is only dimly un-
derstood and will require a multidisciplinary approach to understanding
this paradox.
In conclusion, stress refers to a perturbation of the organism’s homeo-
static mechanisms that entails a definable suite of physiological responses.
For any particular organism, the stressful stimuli and the physiological
responses need to be operationally defined, especially in terms of the fre-
quency and duration of the perturbation. Stress leads to a reduction in the
energy and materials available for maximizing lifetime reproductive success.
Because the organism expends nutrients to surmount proximate physiologi-
cal and ecological threats to survival, less nutrients are available to fuel
reproductive functions. The measurement and interpretation of stressful
conditions remains a rigorous challenge to researchers investigating the
relationship between stress and immunocompetence.
111. THENATURE
OF IMMUNOCOMPETENCE
Resistance to disease and immunocompetence are often conflated and

used interchangeably in nonmedical writing. Properly, resistance is defined
as the ability of the host to prevent disease arising from endogenous (e.g.,
tumors) or exogenous (e.g., infectious agents) sources. Resistance is the
most general term and includes genetic, behavioral, and environmental
resistance modes. Allelic variants coding for alternative biochemical prod-
ucts can prevent infection or reduce the disease severity because the parasite
is not capable of efficient utilization of the variant protein of the host.
STRESS AND IMMUNE DEFENSE 137
The sickle cell allele of hemoglobin is the most widely cited example of
genetically based resistance. Preening or grooming to remove ectoparasites
exemplifies behavioral resistance. Commensal microbial populations in the
gut provide protection from more invasive enteric bacteria and provide an
example of environmental influences on disease resistance. Notice that all
of these examples of resistance lack an immunological component.
Immune-mediated resistance to infectious disease relies on two important
arms of the vertebrate immune system-natural (or innate) immunity and
acquired immunity. There is an unfortunate tendency for immunologists to
use the term “adaptive” in place of “acquired” immunity. Natural immunity
involves protective mechanisms that are constituitively expressed, such as
macrophages and natural killer cells, and does not require prior exposure
to the infectious agent. In contrast, acquired immunity is based on induced
responses that are very specific for particular parasite-derived antigens and
that retain memory of them. Acquired immune responses are generated
by lymphocytes derived from the thymus (T lymphocytes) and the bursa of
Fabricius (B lymphocytes), or the nonavian equivalent of the bursa-bone
marrow. Historically, acquired immunity has been divided into humoral
(antibody-mediated) and cellular (cell-mediated cytotoxic) responses, de-
pending on whether immunity can be transferred by soluble or cellular
elements, respectively. Currently, lymphocyte-mediated responses are clas-
sified as TH1or TH2,based on the pattern of cytokines (intercellular regula-
tory molecules) produced during the response. Generally speaking, TH1
responses involve cell-mediated cytotoxic responses against intracellular
pathogens, whereas TH2responses involve antibody responses that are most
effective against bacteria, extracellular protozoa, and most helminths. The
regulatory switch between TH1 and TH2immune responses is actively being
investigated and it appears that elevated steroid hormones are one factor
that shifts the balance from T H 1 to TH2 (Mason, 1991; Rook et af., 1994).
This simple overview not only provides the necessary background for the
ensuing discussion but also demonstrates that steroid hormones do not
have a simple, single effect on immune function.
Immune responses arise from the interaction of numerous cells types
and are regulated by endocrine and cytokine networks with complex feed-
back loops. As an example, bacteria entering through broken skin attract
granulocytes, such as neutrophils that emigrate through the endothelial
wall to the site of infection. These inflammatory cells degranulate and
release toxic compounds into the tissue, which results in killing of bacteria
and localized necrosis. Bacteria and antigens released from the lysed bacte-
ria encounter macrophages in the draining lymph node. These macrophages
ingest, proteolytically process, and then present antigen-derived peptides
on the cell surface to receptors on T helpers (TH2) and B lymphocytes.
138 VICTOR APANIUS
When activated, these lymphocytes proliferate rapidly, doubling every 14

hr. T helper lymphocytes secrete cytokines that drive the clonal expansion
of B lymphocytes bearing receptors with the highest affinity for antigen.
These cells secrete antibodies, which are used by granulocytes and mac-
rophages to clear the remaining bacteria. The essential features of a
lymphocyte-mediated immune response is the interaction with macro-
phages and the intense proliferation of lymphocytes at the peak of the
response. The combination of natural immune mechanisms and the induc-
tion and regulation of antigen-specific immune responses leads to the phe-
notypic trait called immunocompetence. This can be thought of as the
homeostatic regulation of host-parasite interactions, as animals are contin-
ually challenged by invasive microbes in the gut as well as opportunistic
parasite infections.
Research utilizing transgenic “knock-out’’ animal models has revealed
a surprising robustness of the vertebrate immune system. For example,
disrupting the antigen presentation pathway used for the detection of intra-
cellular parasites had a surprisingly modest effect when the deficient animals
were challenged with experimental infections using a diversity of intracellu-
lar infectious agents (reviewed in Apanius et al., 1997). From these studies,
one can infer that there is a remarkable redundancy in immune functions.
Deletion of antigen-specific responses in the mice mentioned here was
compensated by increased activity of natural killer cells, an arm of the
natural immune system. This underscores one of the paradoxes of the
relationship between stress and immunocompetence. For example, elevated
corticosteroids are related to increased traffic of granulocytes in the vascula-
ture and sequestration of lymphocytes in lymphoid tissue (Dhabhar et
al., 1995). This leads to an increased leukocyte count, which has been
misinterpreted by some to signal increased immunocompetence (Dufva and
Allander, 1995; Gustafsson et al., 1994). This redistribution of leukocytes
permits an immediate mobilization of granulocytes, which are preformed
and are able to enforce the first line of defense at lesions. Antigens released
from the site of infection drain to the lymph nodes where lymphocytes
have collected to cooperate with antigen-presenting cells in antigen-driven
proliferation. If the stressful condition continues and chronically elevates
glucocorticoid levels, then nonlymphocyte mediated immune mechanisms,
for example, natural killer cells, are upregulated to compensate (Fowles et
al., 1993). Thus, stress does not necessarily reduce immunocompetence
because, in some cases, it actually enhances some components of immuno-
competence as part of a compensatory response. This requires investigators
to be very specific about the form, intensity, and duration of stress and the
particular measure of immunocompetence that is taken.
Assessing immunocompetence is made difficult because of (1) the com-

plexity of a seemingly simple process, such as an antibody response against
a single antigen; (2) the overlapping and redundant nature of immune
mechanisms; and ( 3 ) the ability of one component of immunity to be
upregulated to compensate for ineffective responses by other components.
In practical terms, a battery of immunological tests are performed to assess
elements of the natural, humoral, and cellular immunity. These tests can be
performed only on captive animals, except under exceptional circumstances
such as in nestling birds, and usually entail euthanasia to collect spleen or
lymph node tissue. Under these circumstances, it is not difficult to under-
stand that immunocompetence of wild animals is often discussed only in
theoretical terms due to the paucity of empirical data. The situation is made
more difficult because of phylogenetic variation in immune mechanisms
and because immune function of wild animals may not correspond to that
measured in domestic animals living in the sanitary conditions of captivity.
A surprising gap in our understanding of immunocompetence is the
physiological cost of immune function. A number of studies have clearly
shown that nutrient limitation, especially protein deficiency, can reduce
immunocompetence. Yet, some immune mechanisms, principally elements
of natural immunity, are apparently enhanced by moderate nutrient limita-
tion, possibly as a compensation for reduced acquired immune responsive-
ness. In general, the nutritional studies show that immune function can be
altered by trace nutrient, calorie, and protein deficiency and that lympho-
cyte-mediated cytotoxic responses are most sensitive, followed by antibody
responses and then natural immunity (Chandra and Newberne, 1977;Gersh-
win et al., 1985; Klasing et al., 1991; Cunningham-Ruddles, 1993). Nonethe-
less, the metabolic cost of specific immune mechanisms is unknown and
the total fraction of basal metabolic rate or daily energy expenditure that
is allocated to immune function remains conjectural. This is unfortunate
because the immunoregulatory strategies under stressful conditions may
involve re-allocation of nutritional resources in the light of the cost of
particular immune mechanisms. Without a better understanding of the
physiological cost of immune function these hypotheses cannot be tested.
With these caveats in mind, this review will discuss the current under-
standing of how stress alters immunocompetence. Generalizations about
the relationship of stress and host susceptibility to parasitism will not be
directly addressed because so many additional factors influence host-
parasite interactions. This review will necessarily rely on detailed endocri-
nological and immunological information gathered using laboratory animal
models. To the extent possible, the relationship of the neuroendocrine-
immune axis to other physiological processes will be considered with the
140 VICTOR APANIUS
hope of reaching a better understanding of the strategy for stress-induced

immunomodulation.
LINKAGE
IV. NEUROLOGICAL OF STRESS
AND IMMUNOCOMPETENCE
The response to stress is a neurological cascade of events that begins at

the electrophysiological level in specific sites in the brain and continues
with sustained secretion of adrenal hormones that entrain long-term physio-
logical processes leading eventually to habituation or emelioration of the
stressful condition. This chain of events is mediated by two pathways: the
sympathoadrenal (SA) and the hypothalamo-pituitary-adrenocorticalaxis
(HPA). In this section, events within the central nervous system (CNS)
will be reviewed. Also considered will be immunological feedback to the
CNS, that is, the influence of immunological processes on neurological
activity and behavior of the host. This communication between the immune
and nervous systems is significant because most disease states can be consid-
ered as a source of stress and the response involves immune mechanisms
that are assisted by particular host behavior modifications, such as sleep.
Thus, the effect of stress on immune function is not a simple causal chain
but involves feedback loops that make the relationship more complex.
The immediate neurological response to stressful stimuli is activation of
cerebral catecholamine circuits, neurons which show an elevated secretion
and catabolism of norepinephrine, dopamine, and epinephrine. Cerebral
serotonin circuits also appear to play a role in mediating the stress response
in the central nervous system. Although these neurons project to many
regions of the brain, the turnover of norepinephrine in the hypothalamus
is notable (reviewed in Dunn, 1996).
Corticotropin-releasing factor (CRF) appears to be the central regulator
of the stress response at the neural level. Injection of CRF directly into
the brain mimics many of the endocrine and physiological responses to
stress, including (1) activation of the HPA and SA axes; (2) inhibition of
growth hormone and gonadotropins; and (3) behavioral responses such as
anorexia and reduced physical activity (Owens and Nemeroff, 1991). These
responses result from the pituitary release of adrenocorticotrophic hor-
mone, which leads to the production of glucocorticoids by the adrenal
cortex, which then induce these systemic effects. These will be covered in
more detail in the next section.
An important finding that has stimulated psychoneuroimmunological
research has been the discovery of autonomic nervous system innervation
of the peripheral lymphoid tissue. This provides an empirical basis for a
direct linkage of neural activity with immune function. To date, it has
been established that noradrenergic sympathetic nerve fibers innervate the

primary and secondary lymphoid organs. The primary lymphoid organs,
for example, thymus and bone marrow, are the sites of lymphopoiesis
during ontogeny. These organs are invested with noradrenergic fibers that
appear to influence cellular proliferation, differentiation, and emigration
rates. In the thymus, noradrenergic nerve fibers enter the tissue with the
vasculature and are found in highest density at the corticomedullary junc-
tion, which is a location that can potentially control the flux of differentiated
cells through the tissue. In secondary lymphoid tissues, for example, spleen,
lymph nodes, and gut, noradrenergic fibers radiate throughout the paren-
chmya and contact lymphocytes as well as macrophages and other antigen-
presenting cells. The precise role for these nerve tracts is not clear, but the
rich assortment of cognate receptors for neurotransmitters on immunocytes
suggests that the neurons are signaling to resident and circulating cells of
the immune system (reviewed in Ader et al., 1990).
It is these histological and neurochemical observations that lend credence
to a remarkable body of experiments on physiological and behavioral condi-
tioning of immune function. These experiments date back to the 1920s in
the Soviet Union and follow the classical conditioning paradigm of coupling
conditioned and unconditioned stimuli. In recent experiments conducted
by Ader and colleagues, the novel taste of saccharin (conditioned stimulus)
was paired with an immunosuppressive agent (unconditioned stimulus).
Upon reexposure to saccharin, depressed immune function was observed.
Because these experiments were greeted with great skepticism, they have
been repeated in a number of laboratories with additional immunological
assays and experimental psychological techniques. These experiments have
demonstrated that behavioral conditioning affects many components of
immune function, such as antibody production, lymphocyte proliferation,
foreign tissue rejection, as well as delayed-type allergic reactions. Most
importantly, enhancement of immune responses was also observed when
immunostimulatory drugs were paired with the conditioned stimulus. This
and other sources of evidence argue against conditioning of immune func-
tion being simply a stress response (reviewed in Husband, 1993).
The sympathetic innervation of the lymphoid compartments not only
complements the hypothalamo-pituitary-adrenal axis for modulating im-
mune function during stress but also forms another regulatory network
with feedback. Cytokines released in localized immune reactions can enter
the systemic circulation and bind to cognate receptors throughout the brain.
Interleukin 1 (IL-1) is secreted by macrophages at the initial stages of a
specific immune response and has profound effects at a number of physio-
logical levels. In the brain, elevated IL-1 levels are associated with an
increased time spent in slow-wave (non-REM) sleep and an increased
142 VICTOR APANIUS
hypothalamic set point for body temperature. IL-1 is also associated with
anorexia, lethargy, and inhibition of locomotory activity, which character-
izes the malaise of various disease states. Interestingly, a number of bacterial
cell-wall products, most notably muramyl peptides, have similar biological
properties. The neural and behavioral effects induced by IL-1 span a wide
spectrum. There is a circadian rhythmicity in physiological concentrations
of IL-1 that induces mild but measurable changes, in body temperature,
for example, and at the other extreme chronic malignant disease is often
associated with a chronic wasting syndrome that can be re-created in experi-
mental animals with exogenous IL-1 administration. The effects of IL-1 do
not occur in isolation. Not only is this cytokine part of a central immunoreg-
ulatory network, which includes glucocorticosteroids, but it also mediates
numerous other metabolic processes in the liver, skeletal muscle, vascular
system, and hematopoietic system (reviewed in Dinarello, 1992), all of
which ultimately affect the physiological condition and performance of
the individual.
The linkage of the nervous and immune systems through neurotransmit-
ters and cytokines appears to be deeply integrated with organismal pro-
cesses such as foraging behavior, digestion, energy expenditure, as well as
reproduction. The relationship between stress and immunocompetence can
be better rationalized if these additional behavioral and physiological pro-
cesses are also considered.
v. ENDOCRINE
LINKAGE
OF STRESS A N D IMMUNOCOMPETENCE
Glucocorticoids have had a long history in therapeutic medicine, yet

current research continues to enlarge our understanding of these central
mediators of stress and immunocompetence. The conventional wisdom is
that glucocorticoids suppress immune responses and reduce inflammation.
These hormones link stress, as perceived in the brain, with: (1) impaired
immunocompetence; (2) increased susceptibility to infectious and malignant
disease; and (3) decreased susceptibility to autoimmune diseases. Long-
term administration of high doses of glucocorticoids leads to a reduced
mass of primary and secondary lymphoid organs due to a depletion of
lymphocyte cellularity and regression to a condensed epithelial reticulum
(reviewed in Cupps and Fauci, 1982). These statements are supported by
a formidable body of experiments and underlie many modern medical
therapies. What is less widely appreciated is that glucocorticoid levels vary
on a daily basis and are increased during the course of an immune response.
Thus, glucocorticoids are an essential component of the endogenous immu-
noregulatory network.
Circulating glucocorticoid levels show a circadian oscillation. Levels in-

crease during an animal’s active phase and decline during the inactive
phase. IL-1 shows an inverse circadian pattern, with higher levels in the
animal’s inactive period (Zabel et af., 1990). These circadian oscillations
are mirrored by leukocyte traffic patterns (Abo et al., 1981; Kawate et af.,
1981) and antibody responsiveness (Fernandes et af., 1976). Moreover, sleep
deprivation impairs antibody responsiveness and the immunological deficit
can be restored by administration of IL-1 (Brown et al., 1989). It is now
known that IL-1 regulates the secretion of adrenocorticotrophic hor-
mone and other pituitary hormones, including growth hormone, thyroid-
stimulating hormone, and prolactin (Bernton et al.,1987). The combination
of pituitary hormones and IL-1 is associated with: (1) increased body tem-
perature; (2) increased catabolism of glycogen; (3) increased catabolism
of skeletal muscle; and (4) increased turnover of circulating amino acids
(Klasing, 1988). These metabolic changes can be viewed as a mobilization
of nutrient stores to provide substrates for activation and proliferation of
cells involved in inflammation and immunity. Thus, the flow of nutrients
into or away from the lymphoid compartment is principally controlled by
the levels of IL-1 and corticosterone, respectively.
When a specific immune response is induced by infection or immuniza-
tion, circulating glucocorticoid levels transiently rise in the first 24 hr of
the response. There is increasing evidence that a modest increase of cortico-
steroid levels does not always depress immunocompetence and in some
cases may actually be associated with enhanced immune responsiveness.
For example, prior handling or saline injections of mice to condition them
to the immunization procedure led to reduced antibody responses compared
to controls that were handled only at immunization. Pre- and postimmuniza-
tion glucocorticoid levels were comparable and stable in the mice condi-
tioned to injections or handling but hormone levels were elevated postim-
munization in the mice handled only once. The unconditioned animals had
higher antibody titers despite having higher corticosteroid levels following
immunization (Moynihan er af., 1989). This would not be expected from
the traditional view that glucocorticoids are immunosuppressive.
There are several plausible interpretations of this outcome. One line of
reasoning regards the elevated corticosterone levels in animals with ele-
vated antibody responses as evidence that glucocorticoids released during
an induced immune response are part of a negative feedback loop, whereby
the enhanced response is reduced by the concomitant increase in immuno-
suppressive hormone later in the response. This is supported by a great deal
of empirical evidence whereby hypophysectomized animals suffer increased
immunopathological disease from overreactive immune responses. Another
interpretation is that elevated glucocorticoid levels were responsible for
144 VICTOR APANIUS
increasing the flux of nutrients in the immunized animal and the dual signals
increased cell proliferation and antibody production in lymphoid tissues.
Because we currently lack the relevant information to address this hypothe-
sis it cannot be rejected outright. The important point is that physiological
and pharmacological levels of glucocorticoids are often associated with
reduced immunocompetence, but not unequivocally. The additional dimen-
sion of behavioral conditioning, through daily handling or repeated injec-
tions, shows that the effects of glucocorticoids on immunity are not simply
linear and additive but interact with other endocrine factors and the neu-
ral processes.
The relationship between energy metabolism and immunocompetence
is suggested by additional endocrine factors that regulate both of these
processes. Growth hormone, prolactin, and thyroid hormone generally have
stimulatory effects on lymphocytes. Congenital deficiency of growth hor-
mone is associated with fewer cells in the primary lymphoid organs as well
as reduced natural and acquired immunity, which can be reversed, in part,
by exogenous administration of the hormone in birds and mammals. Experi-
mental administration of growth hormone enhances immunocompetence
and even reverses some of the age-related decline in immune function
(Gelato, 1996). There is evidence that circulating growth hormone levels
increase (1) during infectious disease; (2) in response to elevated IL-1; and
(3) in response to bacterial cell-wall products called endotoxins (Gala,
1991). Furthermore, lymphoid cells of the thymus and spleen secrete growth
hormone as well as express growth-hormone-releasing hormone receptors
(Guarcello et al., 1991). Thus, growth hormone acts within the immune
system, probably in concert with insulinlike growth factor I, to positively
stimulate lymphocyte-mediated responses. Systemic effects of growth hor-
mone on nutrient mobilization would coincide with increased activity of
the lymphoid compartment.
Thyroid hormones, notably thyroxine, appear to play an immunostimula-
tory role within the endocrinekytokine network. Endogenous administra-
tion of thyroxine produces variable results in terms of altering the immuno-
competence of normal animals. However, animals with congenitally or
pharmacologically impaired thyroid function have reduced thymus and
spleen mass and decreased cell-mediated responses, which can be, at least
partially, restored with exogenous administration of thyroid hormones (re-
viewed in Marsh, 1992). More recently, thyroid-stimulating hormone has
been shown to be secreted by activated lymphocytes, which in turn promotes
the ability of helper T lymphocytes to increase antibody production by
B lymphocytes in vitro (Kruger et al., 1989). This provides another exam-
ple that hormones that regulate metabolism also control lymphocyte-
mediated immunity.
In the generalized stress response, prolactin secretion by the anterior

pituitary is increased. Growth hormone secretion may transiently increase
early in the stress response but is typically decreased in some species or
later in the response. Despite this difference in hormone secretion pattern,
prolactin has many of the same effects on immune function as growth
hormone. Markedly increased or decreased circulating levels of prolactin
are immunosuppressive, suggesting that normal physiological concentra-
tions of this hormone are necessary for immunological homeostasis. Immu-
nogenic stimulation leads to increased production of prolactin by the pitu-
itary as well as lymphocytes. Lymphocytes have prolactin receptors
indicating autocrine and paracrine functions for this hormone within the
immune system (Matera, 1996).
Prolactin levels are dramatically elevated in reproducing animals and
these same individuals are known to be immunosuppressed. It has been
suggested that the elevated prolactin levels in lactating sheep is the basis
for the well-documented phenomenon of periparturient increase in fecal
shedding of intestinal nematode eggs. However, pharmacological manipula-
tion of ewes has been successful in both increasing prolactin in nonlactating
females and decreasing prolactin in lactating females, while no effect has
been found on measures of immunity or on intestinal parasite egg shedding
(reviewed in Barger, 1993). Although additional endocrine factors have
been proposed to account for immunosuppression associated with parturi-
tion and lactation, it is equally likely that the high level of energy expendi-
ture associated with lactation (Thompson, 1992) may be the ultimate cause.
This hypothesis is supported by manipulation of parental effort in birds.
The proportion of individuals infected with blood parasites has been posi-
tively associated with artificially increased parental effort in a number of
studies (reviewed in Sheldon and Verhulst, 1996). It is also known that
daily energy expenditure (Deerenberg et al., 1995) and corticosterone levels
(Silverin, 1982; Hegner and Wingfield, 1987) are positively related to paren-
tal effort in birds. A recent study provides evidence that increased blood
parasitism in birds with artificially increased brood sizes is associated with
a hematological measure of stress, increased granulocyte : lymphocyte ratio
(Ots and Hbrak, 1996). In sum, these studies indicate that increased repro-
ductive effort (1) increases nutrient demand; (2) increases energy turnover;
(3) increases circulating glucocorticoid levels; and (4) increases the preva-
lence and intensity of parasitism. These studies link high levels of physical
activity and energy metabolism with increased susceptibility to parasitism.
VI. WHYSTRESS IMMUNOCOMPETENCE

ALTERS
It is tempting to finish with a diagram or table to summarize the bidirec-

tional communication between the neuroendocrine and immune systems,
146 VICTOR APANIUS
with pluses and minuses to show positive and negative effects. Readers
seeking this information should consult the reviews listed in Section I. These
summary figures fail to show the complex feedback networks, compensatory
cross-regulation, nonlinear dose dependence, and temporal dependence
that characterize these phenomena. As such, the depictions suggest that
the relationships are generalizable and consistent across species. The gener-
alizations that can be offered have already been made in this review and
any detailed listing of hormones and cytokines will soon be out of date. In
its place, I offer a synthesis that moves the question away from mechanisms
and pathways, which are only partially understood to date, to focus on the
adaptive significance of stress-induced alteration in immunocompetence.
What is the strategy underlying endocrine regulation of immunity and how
has this strategy been shaped by evolution through natural selection?
Stress requires that animals make an immediate response that can be
physiologically costly. Typically, nutritional reserves are mobilized and
nutrient acquisition is suppressed. This physiological regime entails a reduc-
tion of physiological processes that are not immediately vital. Immunity is
one of the processes that is reduced and this may be due to two factors.
First, immune function appears to be costly, although direct evidence is
lacking. It is possible that immune function requires the same quantity of
nutritional resources as the nervous system. Yet the cost of immune function
can be ameliorated by selectively downregulating the most costly compo-
nents without complete loss of immunity. The other factor that makes
immune function susceptible to downregulation during stress is that induc-
tion of immune mechanisms in the poststress period permits compensation
for the period of suppressed immunity. Thus, the increased level of micro-
parasites that have replicated in the host and the increased number of
infective stages of parasites acquired during the stress period may be cleared
afterward. At least, that would be the teleological expectation from the
host’s point of view.
This explanation is supported by empirical observations on the effect of
long-term stress on immunity in mice. Mice were subjected to daily auditory
stress. From the initial day and continuing for 20 days afterward, the mice
showed depressed lymphocyte function and elevated plasma cortisol levels.
After that point, and continuing for approximately 20 days, they showed
enhanced lymphocyte function and relatively low levels of plasma cortisol
(Monjan and Collector, 1977). This indicates that immune function does
not simply adapt to chronic stress by returning to the basal level of activity
but appears to be increased in a compensatory manner. This may have
occurred in these laboratory experiments because food was provided ad li-
bitum.
The relationship between imposed stress and the animal’s nutrient budget
is of critical importance in discussing pathogenesis of infections during
the stress period. If the stressful event is associated with depressed body
temperature or nutrient limitation, then replication of microparasites may
be inhibited because of their dependence on the host’s nutritional resources.
Viral replication and pathogenicity is often linked to the flux of nutrients
through host tissues, especially nitrogen-rich substrates for nucleic acid
synthesis. In these circumstances, stress-induced immunosuppression en-
tails less risk to the host than might be generally appreciated.
However, stress generally induces a short-term increase in the flux of
nutrients through host tissues, to support thermogenesis or muscular activ-
ity, and this can promote microparasite replication and macroparasite re-
production. The fact that the host is often immunosuppressed as well makes
it difficult to disentangle the two effects. Nevertheless, from first principles
one can infer that the stress response in a well-nourished animal is associated
with high nutrient turnover, which entails a greater risk of fulminating
infectious disease because of the additive and possibly synergistic operation
of these two factors.
Moderate levels of exercise in humans and captive animals also increases
nutrient turnover. This increased nutrient flux is associated with increased
circulating levels of epinephrine, adrenocorticotrophin, glucocorticoids, p-
endorphin, metenkephalin, prolactin, growth hormone, and thyroid hor-
mone (Smith and Weidemann, 1990). Since these hormones also regulate
immune function, then physical activity also affects immune function. Gen-
erally, moderate levels of activity enhance immunocompetence and parasite
resistance (Cannon and Kluger, 1984), possibly through elevated nutrient
flux. It is also possible that the increased levels of endotoxin associated
with exercise enhance immunocompetence (Cannon and Kluger, 1983).
Endotoxin is released in the digestive tract from the breakdown of gram-
negative bacteria and increased dietary throughput would thus result in
increased endotoxin levels. At higher levels of physical exercise, especially
among “elite athletes,” upper respiratory tract infections are more common
and a hormonal regime associated with immunosuppression is observed
(Hoffmann-Goetz and Pederson, 1994). These studies suggest that an in-
creased flow of nutrients in conjunction with normal immune function can
control opportunistic infections in all but the most demanding circum-
stances.
These considerations apply to parasites where there is a direct relation-
ship between host nutrient acquisition and pathogenesis. These parasites are
typically associated with economic disease observed in animal production
settings, for example, intestinal nematode infections of ruminants. Well-
nourished animals harbor low-grade, chronic infections and it appears that
148 VICTOR APANIUS
parasite populations are at equilibrium with immune-mediated expulsion.

This type of parasitism is probably widespread in wild animals and it has
been extremely difficult to measure the fitness consequences in nature. At
the same time, it is also widely recognized that these chronic infections
show an opportunistic increase in intensity in the stressed host. A good
example is the relapse or recrudescence of blood parasite infections ob-
served in wild bird populations. In avian malaria, there are seasonal in-
creases in prevalence and parasitemia associated with host reproduction.
Relapse of avian hematozoa infections can be induced by corticosterone
injections (Applegate, 1970). It is informative that although hematozoan
parasitemia or nematode egg-shedding increases during the reproductive
cycle, these infections are seldom associated with morbidity or mortality
in nature. It is tempting to suggest that the increased intensity of parasitism
that accompanies reproduction is an important component of the cost of
reproduction (Sheldon and Verhulst, 1996), but, to date, there is only one
study that shows that the increased parasitism reduces future reproductive
success (Mgller, 1993). However, these data are difficult to collect, so the
question remains an open one. The available data show that the intensity
of parasitism increases during the reproductive cycle and that the infections
are eventually controlled, presumably by host immune mechanisms. This
may be an example of adaptive modulation of immunity mediated by
endocrine factors associated with stress. The host strives to maximize its
reproductive output in a seasonal reproductive episode. It re-allocates nutri-
ents from immune function to reproductive effort during peak periods of
physical performance and incurs a greater parasite burden. This burden is
reduced by immunological mechanisms following the reproductive bout.
This hypothesis permits predictions that can be tested through comparative
and experimental approaches and provides a framework for viewing the
stress response as an adaptive strategy.
Numerous studies have documented the causes of death in wildlife popu-
lations and have attempted to identify the factors involved. A never-ending
controversy centers around the role of infectious agents in natural mortality.
Are the individuals that expire over the winter and that are found to harbor
larger parasite loads killed by environmental circumstances or because of
these parasitic infections? It is often stated that the higher parasite load
predisposes these individuals to stress-induced mortality. It is equally proba-
ble that individuals stressed by adverse weather may have been unable to
control their parasite load. Although it would appear to be simple to tease
these factors apart, in practice there are few studies that adequately address
the contributions of environmentally induced stress and parasitism to mor-
tality in wild animals.
It is this intertwined process, that is, the progression of infectious disease

in a stressed animal, that is crucial to understanding the adaptive nature
of stress-induced immunosuppression. Two polarized views can be ad-
vanced. One is that infectious disease in a stressed animal is biologically
self-limiting. As the host deteriorates in condition during a severely stressful
event, the nutrients to support parasite replication or reproduction diminish
and the infection becomes less pathogenic. Indeed, there are examples of
parasites that are highly pathogenic during the refeeding stage of hosts that
have been nutritionally deprived, for example, malaria in undernourished
human populations. In this view, stress-induced immunosuppression is
clearly advantageous due to the nutrient savings during the stress period
and because mortality solely due to parasitism is unlikely. The opposite
view is that stress-induced immunosuppression is advantageous only in the
short term where the nutrient savings may be critical during the stress
period. But as the period of immunosuppression lengthens, then the severity
of infectious disease increases in a runaway process in which disease pro-
gression contributes to mortality. Despite the intuitive appeal of this view,
the supportive evidence is anecdotal and seldom experimental. This stems,
in part, from the methodological difficulty of identifying multicausal factors
of mortality. Ultimately, selection must favor stress-induced immunosup-
pression because of the short-term advantages. Whether immunosuppres-
sion exacerbates mortality during prolonged stressful episodes remains to
be demonstrated.
VII. SUMMARY
Alterations in immune function by stress are a graded response that is

tightly regulated by neuroendocrine pathways. In mild stress responses,
the principal immunological mechanisms shift from lymphocyte-mediated
responses, which have high metabolic demands due to rapid cell prolifera-
tion, to natural immunity that utilizes the current store of cells and proteins.
This provides the individual with a saving in nutrients. Because the less
expensive immune mechanisms do not retain memory and are not as specific
as those mediated by lymphocytes, there is a modest decrease in immuno-
competence over the remaining lifetime of the individual. More intense
stress leads to the suppression of additional mechanisms with presumed
additional savings and costs.
As the stress period lengthens, either the immune mechanisms are re-
stored to their previous levels or they continue to decline, depending on
the intensity of the stressful stimuli. In the latter case, immunological control
of preexisting infections or new infections breaks down and an increased
150 VICTOR APANIUS
severity of infectious disease is observed. It is at this point that it is no

longer possible to interpret the outcome in adaptive terms. Up to this point
it can be consistently argued that stress-induced immunoregulation is part
of the generalized metabolic response to provide the resources for a short-
term solution to an immediate threat to survival. The stress response incurs
a fitness cost in decreased parasite resistance, which may be measured
proximally as increased susceptibility or ultimately through decreased im-
munological memory.
The assertion that long-term social stress induces immunosuppression,
which in turn allows malignant and infectious diseases t o progress to the
point of significant morbidity or mortality, is intriguing but remains conjec-
tural (Sapolsky, 1992). It is biologically plausible, but its relevance, except
in highly contrived circumstances, for example, in captive wildlife, is ques-
tionable. This assessment is based in part on the compensatory nature of
the immune system. Low-grade, chronic stress does alter immune function
but does not necessarily reduce disease resistance. More likely, chronic
social stress affects access to food and the likelihood of parasite transmis-
sion, and these may directly affect the animal’s health in nature.
References
Abo, T., Kawate, T., Itoh, K., and Kumagai, K. (1981). Studies on the bioperiodicity of the
immune response. I. Circadian rhythms of human T. B and K cell traffic in the peripheral
blood. J. Immunol. 126, 1360-1363.
Ader, R., Felten, D.. and Cohen, N. (1990). Interactions between the brain and immune
system. Annu. Rev. Pharmacol. Toxicol. 30, 561-602.
Apanius, V., Penn. D., Slev, P. R.. Ruff, L. R., and Potts, W. K. (1997). The nature of selection
on the major histocompatibility complex. Crii. Rev. Inmunol. 17, 179-224.
Applegate. J. E. (1970). Population changes in latent avian malaria infections associated with
season and corticosterone treatment. J. Parasitol. 56, 439-443.
Barger, 1. A. (1993). Influence o f sex and reproductive status on susceptibility of ruminants
to nematode parasitism. Inl. J. Parasiiol. 23, 463-469.
Bernton. E. W.. Beach, J. E., Holiday, J. W.. Smallridge. R. C., and Fein. H. G. (1987).
Release of multiple hormones by a direct action of interleukin-I on pituitary cells. Science
238, 519-521.
Besedovsky, H. 0. (1996). Immune-neuro-endocrine interactions: Facts and hypotheses. En-
doer. Rev. 17, 64-102.
Blalock, J. E. (1989). A molecular basis for bidirectional communication between the immune
and neuroendocrine systems. Physiol. Rev. 69, 1-32.
Blalock, J. E. (1 994). The syntax of immune-endocrine communication. Imrnunol. Today
15, 505-510.
Brown, K. I., and Nestor, K. E. (1973). Some physiological responses of turkeys selected for
high and low adrenal responses to cold stress. Poiilr. Sci. 52, 1948-1954.
Brown, R., Price. R. J., King, M. G.. and Husband, A. J. (1989). Interleukin-1 beta and
muramyl dipeptide can prevent decreased antibody response associated with sleep depri-
vation. Brain, Behav., Immunol. 3, 320-321.
Cannon, J. G.. and Kluger, M. J. (1983). Endogenous pyrogen activity in human plasma after
exercise. Science 220, 617-619.
Cannon, J. G., and Kluger, M. J. (1984). Exercise enhances survival rate in mice infected with
Salmonella typhimurium. Proc. Soc. Exp. Biol. Med. 175, 518-521.
Chandra, R. K.. and Newberne, P. M. (1977). “Nutrition, Immunity, and Infection.” Plenum,
New York.
Chrousos, G. P., and Gold, P. W. (1992). The concepts of stress and stress system disorders.
JAMA, J. Am. Med. Assoc. 267, 124441252,
Cunningham-Ruddles, S., ed. (1993). “Nutrient Modulation of the Immune Response.” Dek-
ker, New York.
Cupps, T. R., and Fauci, A. S. (1982). Corticosteroid-mediated immunoregulation in man.
Immunol. Rev. 65, 133-155.
Daynes, R. A,, Araneo, B. A,, Hennebold, J., Enioutina, E., and Mu, H. H. (1995). Steroids
as regulators of the mammalian immune response. J. Invest. Dermatol. 105, 14s-19s.
Deerenberg, C., Pen, I., Dijkstra, C., Arkies, B.-J., Visser, G. H., and Daan, S. (1995). Parental
energy expenditure in relation to manipulated brood size in the European kestrel. Zool-
ogy, Analysis of Complex Systems 99,38-47.
Dhabhar, F. S.. Miller, A. H., McEwen, B. S., and Spencer, R. L. (1995). Effects of stress on
immune cell distribution: Dynamics and hormonal mechanisms. J. Immunol. 154,551 1-
5527.
Dinarello. C. A. (1992). Role of interleukin-1 in infectious diseases. Immunol. Rev. 127,
119-146.
Dufva, R.. and Allander, K. (1995). Intraspecific variation in plumage coloration reflects
immune response in Great tit (Parus major) males. Funct. Ecol. 9, 785-789.
Dunn, A. J . (1996). Psychoneuroimmunology, stress and infection. In “Psychoneuroimmunol-
ogy, Stress, and Infection” (H. Friedman, T. W. Klien, and A. L. Friedman, eds.). pp.
25-46. CRC Press, Boca Raton, FL.
Fernandes, G., Halberg, F., Yunis, E. J., and Good, R. A. (1976). Circadian rhythmic plaque-
forming cell response of spleens from mice immunized with SRBC. J. Immunol. 116,
962-966.
Fowles, J. R., Fairbrother, A,. Fix, M., Schiller, S.. and Kerkvliet, N. I. (1993). Glucocorticoid
effects on natural and humoral immunity in mallards. Dev. Comp. Irnmunol. 17,165-177.
Friedman. H.. Klein, T. W., and Friedman, A. L., eds. (1996). “Psychoneuroimmunology,
Stress and Infection.” CRC Press, Boca Raton, FL.
Gala, R. R. (1991). Prolactin and growth hormone in the regulation of immunity. Proc. Soc.
Exp. Biol. Med. 198,513-519.
Gelato, M. C. (1996). Aging and immune function: A possible role for growth hormone.
Horm. Res. 45,46-49.
Gershwin, M. E., Beach, R. S., and Hurley. L. S. (1985). “Nutrition and Immunity.” Academic
Press, Orlando, FL.
Gross, W. B., and Siegel, P. B. (1985). Selective breeding of chickens for corticosterone
response to social stress. Poult. Sci. 64, 2230-2233.
Guarcello, V., Weigent, D. A,, and Blalock, J. E. (1991). Growth hormone releasing hormone
receptors on thymocytes and splenocytes from rats. Cell. Imrnunol. 136,291-302.
Gustafsson, L., Nordling, D., Anderson, M. S.. Sheldon, B. C., and Qvarnstrom, A. (1994).
Infectious diseases, reproductive effort, and the cost of reproduction in birds. Philos.
Trans. Soc. London, Ser. B 346,321-331.
Hegner, R. E., and Wingfield. J. C. (1987). Effects of brood size manipulations on parental
investment, breeding success, and reproductive endocrinology of house sparrows. Auk
104,470-480.
152 VICTOR APANIUS
Hochachka. P. W., and Somero, G. N. (1984). “Biochemical Adaptation.” Princeton University

Press, Princeton, NJ.
Hoffman-Goetz, and Pederson, B. K. (1994). Exercise and the immune system: A model of
the stress response. Irnmunul. Today 15, 382-387.
Hoffmann, A. A,. and Parsons, P. A. (1991). “Evolutionary Genetics and Environmental
Husband, A. J. (1993). Role of central nervous system and behavior in the immune response.
Vaccine 11, 805-816.
Kawate, T., Abo, T., Hinuma. S., and Kumagai, K. (1981). Studies on the bioperiodicity of
the immune response. 11. Co-variations of muringT and B cells and a role of corticosteroid.
J. Immunol. 126, 1364-1367.
Khansari. D. N., Murgo. A. J.. and Faith. R. E. (1990). Effects of stress on the immune system.
Immunol. Today 11, 170-175.
Klasing, K. C. (1988). Nutritional aspects of leukocyte cytokines. J . Nurr. 118, 1436-1446.
Klasing, K. C., Johnstone, B. J., and Benson, B. N. (199 1). Implications of an immune response
on growth and nutrient requirements of chicks. In “Recent Advances in Animal Nutrition”
(W. Haresign and D. J. A. Cole, eds.). pp. 135-146. Butterworth Heinemann, Stone-
ham, MA.
Kruger, T. E.. Smith, L. R., Harbour, D. V.. and Blalock, J. E. (1989). Thyrotropin: An
endogenous regulator of the in v i m immune response. J. Immunol. 142, 744-747.
Leonard. B. E., and Song, C. (1996). Stress and the immune system in the etiology of anxiety
and depression. Pharmacol., Biochern. Behav. 54,299-303.
Marsh, J. A. (1992). Neuroendocrine-immune interactions in the avian species- a review.
Poulr. Sci. Rev. 4, 129-167.
Mason, D. (1991). Genetic variation in the stress response: susceptibility to experimental
allergic encephalomyelitis and implications for human inflammatory disease. Immunol.
Today 12, 57-60.
Matera, L. (1996). Endocrine, paracrine and autocrine actions of prolactin on immune cells.
Life Sci. 59, 599-614.
Moiler, A. P. (1993). Ectoparasites increase the cost of reproduction in their hosts. J. Anim.
Ecol. 62, 309-322.
Monjan, A. A,, and Collector, M. I. (1977). Stress-induced modulation of the immune response.
Science 196, 307-308.
Moynihan. J., Koota, D., Brenner, G., Cohen. N., and Ader. R. (1989). Repeated intraperito-
neal injections of saline attentuate the antibody response to a subsequent intraperitoneal
injection of antigen. Brain, Behav., Immunol. 3, 90-96.
Ots. I., and Horak. P. (1996). Great tits Parus major trade health for reproduction. Proc. R.
Soc. London, Ser. 263, 1443-1447.
Ottaviani. E.. and Franceschi. C. (1996). The neuroimmunology of stress from invertebrates
to man. Prog. Neurohiol. 48, 421-440.
Owens, M. J., and Nemeroff, C. B. (1991). Physiology and pharmacology of corticotropin-
releasing factor. Pharmacol. Rev. 43, 425-449.
Reichlin, S. (1993). Neuroendocrine-immune interactions. N. Engl. J . Med. 329, 1246-1253.
Rook, G. A. W., Hernandez-Pando, R.. and Lightman, S. L. (1994). Hormones, peripherally
activated prohormones and regulation of the Thl/Th2 balance. Immunol. Today 15,
301-303.
Sapolsky, R. M. (1992). Neuroendocrinology of the stress response. In “Behavioral Endocrinol-
ogy” (J. B. Becker. S. M. Breedlove. and D. Crews, eds.). pp. 287-324. MIT Press,
Cambridge MA.
STRESS A N D IMMUNE DEFENSE 153
Sheldon, B. C., and Verhulst, S. (1996). Ecological immunology, costly parasite defenses and
trade-offs in evolutionary ecology. Trends Ecvl. Evol. 11, 3 17-321.
Silverin, B. (1982). Endocrine correlates of brood size in adult pied flycatchers Ficediiln
hypoleiicn. Geti. Cvmp. Etidvcrinvl. 47, 18-23.
Smith, J. A.. and Weidemann, M. J . (1990). The exercise and immunity paradox: A neuroendo-
crinelcytokine hypothesis. Med. Sci. Res. 18, 751 -755.
Stein, M., and Miller. A. H. (1993). Stress, the hypothalamic-pituitary-adrenalaxis, and immune
function. Adv. Exp. Med. B i d . 335, 1-5.
Thompson, S. D. (1992). Gestation and lactation in small mammals: Basal metabolic rate and
limits of energy use. 1n “Mammalian Energetics: Interdisciplinary Views on Metabolism
and Reproduction” (T. E. Tomasi and T. H. Horton. eds.). pp. 213-259. Comstock,
Ithaca. NY.
Zabel. P., Horst. H.-J.. Kreiker. C.. and Schlaak. M. (1990). Circadian rhythm of interleukin-
1 production by monocytes and the influence of endogenous and exogenous glucocorti-
coids in man. Klin. Wochmschr. 68, 1217-1221.
ADVANCES IN THE STUDY OF BEHAVIOK. VOL. 27
Behavioral Variability and Limits to Evolutionary

Adaptation under Stress
P. A. PARSONS
SCHOOL OF GENETICS A N D HUMAN VARIATION
LA TROBE UNIVERSITY
VICTORIA 3083
BUNDOORA,
AUSTRALIA
I. INTRODUCTION
A. STRESSA N D ENERGY
BALANCES
Interactions between organisms and environment are central for under-
standing evolution. For some evolutionary biologists, the occurrence of
environmental perturbations of an unpredictable nature emphasizes physi-
cal factors as the major determinants of the distribution and abundance of
organisms. Even so, the effects of abiotic factors can be modulated by
interactions with biotic factors (Dunson and Travis, 1991). However, the
abiotic environment should be tracked more predictably than the biotic as
the time scale lengthens.
From a diffuse literature Hoffmann and Parsons (1991) concluded that
abiotic stresses mainly of climatic origin are important in many evolutionary
and ecological processes. Furthermore, inadequate nutrition is usual in
free-living populations, so that animals normally struggle to exist in hostile
environments. White (1993) amassed much evidence, especially in herbi-
vores, indicating that the abundance of organisms is often determined by
a shortage of protein, especially for the young. For instance, pollen digestion
is important for early breeding of Darwin’s finches of the Galapagos Islands
(Grant, 1996). Consequently, many organisms are born but few are expected
to survive due to a combination of climatic stress interacting with and
causing nutritional stress.
Validity of this environmental model is suggested by the rarity of crea-
tures that commonly die of old age in free-living populations. In any case,
a reference point is provided as a boundary for comparisons with more
benign environments, especially certain human populations of recent times.
155
Copyright 0 1998 hy Academic Press
All rights of reproduction in any form reserved.
O(~S-34.541YX$25 00
156 P. A. PARSONS
The assumption of substantial stress contrasts with approaches to the envi-

ronment by many evolutionary biologists, whose ideas appear to be influ-
enced by the apparent existence of adequate nutrition in many human
populations today. But these populations may represent a benign environ-
mental outlier when considered in a historical context, both past and future.
One direct effect of stress is an increase in the expenditure of metabolic
energy, implying a cost (Odum et al., 1979). As exposure to stress is the
norm, there is a need for some energy to exist in any habitat. The habitats of
organisms can then be expressed by an interaction between stress intensity,
magnitude of environmental fluctuations, and energy from resources as a
first approximation. The interaction of stress of various types causing energy
costs and energy gains (provided from resources) is central in relating the
distribution and abundance of organisms to energy balances (Hall et al.,
1992). As energy costs increase as conditions deviate from optimal (Porter
and Gates, 1969), physical conditions can limit the occurrence of organisms
in particular habitats. Biotic variables, such as competition, can be incorpo-
rated into this model via an increase in energy costs, but these effects are
usually second-order compared with abiotic stresses (Parsons, 1996a,b).
Stress reduces the fitness of organisms by deflecting energy from pro-
cesses such as maintenance and survival, reproduction, growth, and genetic
adaptation (De Kruipf, 1991). Fitness therefore is inversely related to the
stress level as a first approximation. Furthermore, the impact of environ-
mental perturbations can be expressed as a stress gradient (Odum et af.,
1979), on which the potential for genetic adaptation falls as stress increases
to an extreme where survival is threatened.
The net energy balance of a species should be relatively high in central
regions of its distribution. However, the margins of distributions of at least
some animal species are limited by physiological constraints. Genes allowing
further adaptation may not arise or, if they do, the animals carrying them
may not survive (Parsons, 1991; Hoffmann and Blows, 1994). Physiological
constraints can therefore determine the location of species borders, for
example, in many North American bird species and in some small mammals
(Bozinovic and Rosenmann, 1989; Root, 1993). In any case, organisms
living at the very extremes of a species range are rarely the healthiest and
most vigorous members of that species.
B. HABITATS
PREFERRED
Habitats in which minimum energy is expended should be preferentially
occupied. Intermediate temperatures between limiting extremes should
therefore be preferred in an environment where temperature gradients
exist. In these regions, maximum energy should be available for behavior,
BEHAVIORAL VARIABILITY AND EVOLUTIONARY ADAPTATION 157
growth, and reproduction (Huey, 1991), and resistance to stress should be

higher than elsewhere (Klieber, 1961; Arking et al., 1988; Zotin, 1990). Of
course, these are the circumstances in which high population sizes may lead
to maximum competition, but in all but the most abiotically stable of
habitats such abiotic effects should be transient.
Insects living in habitats with steep abiotic gradients can be useful for
habitat preference studies. For example, in temperate zone rain forests,
adults of an Australian Drosophila species, D . inornata, tend to rest on the
fronds of tree ferns in the 15-20" range, with a mean of 17.7 ?I 2.0"C,
following behavioral responses to the microenvironment. Consequently,
flies attempt to select microhabitats as close as possible to optimal for
temperature/humidity conditions, where the energy cost from the physical
environment would be minimized (Parsons, 1993a). Similarly, Jones et al.
(1987) found behavioral flexibility for thermal niche preference in D . mela-
noguster, whereby the effects of temperature extremes were ameliorated
by habitat selection. Survival is thereby enhanced, since animals in early
developmental stages are intolerant of energetically costly extreme condi-
tions. In this context, the microenvironment (soil moisture and air tempera-
ture) experienced by a larva during wandering and pupation is important
for pupal survivorship (Rodriguez et al., 1992). Furthermore, larvae from
populations from dry habitats in Tunisia pupate closer to food than those
from wet habitats (Rodriguez et al., 1991).
Adult food-searching range is dependent on temperature. In D . melano-
gaster, the range searched is substantially smaller at the stressful tempera-
ture of 30°C than at 25°C (Good, 1993), presumably because of the high
energetic cost of surviving at 30°C. Genetic shifts in preferred temperatures
occurred in a gradient, when flies were reared at 25, 27, and 30°C for 15
generations. In addition, Yamamoto (1994) found temperature preferences
in natural populations of D. immigruns and D . virilis to be heritable.
Ye et al. (1994) found substantial genetic differences between strains of
D. melanogaster from six local populations for adult starvation resistance,
presumably as adaptations to cope with specific microhabitats. The six
strains ranked for search behavior in parallel with starvation resistance,
such that resistant strains searched over more extended ranges than did
sensitive strains. In the two-spotted spider mite, Tetranychus urticae, there
is genetic variation in aerial dispersal behavior, associated with resistances
to the environmental stresses of desiccation and starvation (Li and Margo-
lies, 1994).
These few examples suggest that behavioral flexibility can reduce the
energy costs of physical stresses. Furthermore, genetic changes can occur,
enabling adaptation to varying habitats, assessed abiotically. Hence, ecobe-
havioral genetic adaptation can evolve under varying stress levels in free-
158 P. A. PARSONS
living populations, which should ameliorate the direct effects of stress in

an evolutionary sense.
Turning now to the physiological state of insects, in D. melanogaster,
starved flies are less discriminatory in responding to alternative resources
in orchards than when unstressed (Hoffmann and Turelli, 1985). Resource
selection is therefore most efficient when organisms do not simultaneously
need to cope with extreme stress. However, nutritional stress appears to
be the norm in natural populations of Drosophifa. For example, within a
French population of D. melanogaster, high reproductive potential is not
normally expressed under natural conditions, because of substantial and
variable fluctuations in food availability (BoulCtreau-Merle et af., 1987).
Furthermore, in a widespread montane North American butterfly, Speyeria
mormonia, experiments with varying feeding regimes show that survival
claims resources as a priority over reproduction (Boggs, 1994).
In summary, more emphasis is needed on the study of habitat preferences
under a range of realistic abiotic environmental conditions. In the remainder
of this chapter, some background is provided for the consideration of limits
to adaptation under predominantly rigorous conditions.
11. ENERGY
LIMITSTO ADAPTATION
A. NONSEXUAL
BEHAVIOR
Energetically expensive behaviors are common, for example, web con-
struction in spiders, and insect and avian flight. However, the amount of
energy that can be assimilated from food is finite (Weiner, 1992). Oxygen
consumption can increase to meet a higher demand for ATP production,
but the maximum possible oxygen consumption (Bennett, 1991) sets a limit
to total behavioral activity.
Consequently, any superimposed environmental stress would be deleteri-
ous by increasing respiration and hence stress sensitivity. For example,
metabolic rate and whole-body thermal conductance increased in polar
bears exposed to oil pollution, which increased mortality during the stress
of a hard winter (Hurst et af., 1991). In D. melanogaster, high-metabolic-
rate “shaker” mutants show high levels of behavioral activity, and are very
sensitive to environmental stresses, including high temperature, desiccation,
and exposure to an unsaturated aldehyde acrolein (Barros et af., 1991;
Parsons, 1992). In fasting rats, minimum heat production occurs in the
thermoneutral zone and increases at higher and especially lower tempera-
tures in association with reduced stress resistance (Klieber, 1961; Blaxter,
1989). In the social Damara mole rat, Cryptomys damarensis, body tempera-
ture remains stable at ambient temperatures from 7 to 30°C, so that there

is a substantial metabolic cost at extremes; at 7°C the metabolic rate is more
than four times higher than in the thermoneutral zone (Lovegrove, 1986).
In C. damarensis, which is from arid regions in southern Africa, the rate
of metabolism is much lower than for comparable species from wetter
regions. This is an energy-saving device enabling adaptation to aridity stress.
In addition, cooperative searching and food sharing can reduce energy
demands further (Lovegrove, 1986). Such adaptations can occur seasonally,
or transiently during periods of bad weather. For instance, in the house
martin, Delichon urbica, energy is saved during transient problems in finding
food in the breeding season, by a complex of physiological and behavioral
adaptations, including low basal metabolic rate, low thermal conductance,
clustering behavior, high tolerance of the young to periods of low food
supply, and the ability to become torpid (Prinzinger and Siedle, 1988).
In subterranean blind mole rats of the Spalax ehrenbergi superspecies
complex, aggression tendency and basal metabolic rate decrease geographi-
cally across Israel as the climatic stresses of temperature, and especially
aridity, increase (Nevo, 1991). These responses would minimize water and
energy expenditure, and are adaptations to counter extreme stress. Further-
more, in isolates from the environmentally harsh Sahara Desert of northern
Egypt, mole rats were totally pacifist, presumably as an adaptation to an
environment that is even more extreme than that in Israel (Nevo et al.,
1992). Consequently, a behavioral-ecophysiological response has evolved
based on selection against aggression. This response has enabled the spread
of S. ehrenbergi into extremely arid environments (Ganem and Nevo, 1996).
In summary, energetically costly behaviors occur frequently, and can
determine limits of adaptation of organisms. Under these circumstances,
any additional stress would be rapidly restrictive. Adaptations to high stress
levels include reductions in resting metabolic rate, social behavior patterns
that conserve energy, and pacifist behavior.
B. SEXUAL
BEHAVIOR
AND SEXUAL
SELECTION
The energy used in calling can exceed resting levels by up to 20 times in
frogs (Ryan, 1988), suggesting that the mating process can be energetically
expensive. In some frog species, mating success of males increases with
increasing chorus tenure (Pough, 1989). As the time a male spends in the
breeding chorus is important in determining mating success, fitness assessed
in this way is likely to be related to available metabolic energy. In damsel
flies, Calopteryx maculata, territorial contests favor males with the greatest
energy reserves, measured by fat content (Marden and Waage, 1990). In
great tits, Parus major, and male pied flycatchers, Ficedula hypoleuca, in
160 P. A. PARSONS
breeding condition, resting metabolic rate is positively correlated with domi-

nance rank (Roskaft et al., 1986). In the fish Betta splendens, winners and
dominant individuals in a hierarchy consume more energy per unit time
than losers and submissives (Haller and Wittenberger, 1988). In pupfish,
Cyprinodon pecosensis, critical swimming speed is higher in territorial than
in nonterritorial males, indicating a positive correlation of vigor with social
status (Kodric-Brown and Nicoletto, 1993). In summary, these and other
examples imply that fitness in mating is normally correlated with energy con-
sumption.
Similarly, daily energy expenditure increased significantly with increasing
display rate and time spent in the lek in the male sage grouse, Centrocercus
urophasianus (Vehrencamp et al., 1989). Daily energy expenditure for the
most vigorously displaying males was two times higher than for a nondis-
playing male, and four times higher than the basal metabolic rate. The
increased levels of lek attendance and display levels appear fueled by
increased quantity or quality of food, since the more actively displaying
males can forage further from the lek. Furthermore, the abiotic environment
is relevant, since metabolic expenditure increases as temperature falls; this
would be ultimately restrictive.
The inadequate resources available for free-living organisms should be
used efficiently. Accordingly, resources are normally channeled to only
some of those seeking to use them, so that a dominant few survive; the
remainder in a population are vulnerable. This can be achieved by territorial
and social behaviors, largely restricting resources to the dominant few, as
demonstrated in passerine bird species by Moller (1991). Analogously,
polygyny can replace monogamy in traditional human societies, when there
are substantial fitness differences among men, following pathogen stress.
The minority of resistant men are dominant in mating, because they are
more skilled in promoting polygyny; these skills include hunting, winning
disputes, and resource acquisition (Low, 1990).
Considering stress from parasites, in the lizard, Sceloporus occidentalis,
Schall and Sarni (1987) found that the time males spend in social behaviors
was reduced when infected with the malarial parasite, Plasmodium mexi-
canum, and furthermore, infected males perch more often in shade. Hence,
the energy cost from parasites reduces social behaviors, and stressful micro-
habitats are avoided. In feral rock doves, Columbia liviu, lice reduced
feather and host body mass, and increased thermal conductance and meta-
bolic rate, indicating an energy cost. This is exacerbated in a deteriorating
abiotic environment during winter (Booth et ul., 1993). Finally, in the colo-
nially nesting cliff swallow, Hirundo pyrrhonutu, mark-recapture experi-
ments over an 8-year period showed that the annual survivorship of birds
BEHAVIORAL V A R I A B I L I T YAND EVOLUTIONARY ADAPTATION 161
parasitized with cimicid bugs, fleas, and chewing lice was .38, compared
with .57 for fumigated, nonparasitized birds (Brown et al., 1995).
Turning to reproduction, de Lope et al. (1993) found that the ectoparasitic
house martin bug, Oeciacius hirundinis, had larger negative effects on the
reproduction of its host, Delichon urbica, when nutritional conditions were
poor during the second compared with the first clutch in the season. In red
jungle fowl, Gallus gallus, chicks infected with parasites grew more slowly
than uninfected controls (Zuk et al., 1990). Since this effect was most
pronounced for secondary sexual traits, there is a channeling of resources
into the normal growth of nonornamental traits under parasite stress. Gen-
erally, birds and fish with high parasite loads engage in less courtship display
and obtain fewer mates than those with lower loads (Hamilton and Zuk,
1982; Kennedy et al., 1987; Clayton, 1990).
Therefore, the energy cost of parasites in combination with abiotic
stresses can preclude the full development of ornamental traits, and reduce
mating and fitness generally. Furthermore, if there are energy restrictions
from nutritional stress, an ornament can rapidly regress, as found for the
nuptial crest of male newts of the genus Triturus (see Halliday, 1978). In
addition, sexual ornamentation in some birds is restricted to the breeding
season, indicating an excessive cost in less favorable abiotic environments.
Sexual selection can be constrained by costs associated with mate choice,
when interacting with unfavorable abiotic circumstances. The same should
apply to biotic effects, although less obviously. Especially for predation,
theoretical models (e.g., Pomiankowski, 1987) predict that female prefer-
ence should decrease with increasing costs of mate choice. Accordingly,
individual females should modify their choice behavior to minimize this
risk. In terms of energy costs, this means that females should become less
discriminatory when given a choice among potential mates at times when
the predation risk is increased. For example, male pipefish, Syngnathus
typhle, exposed to the cod, Gadus morhua, as a predator, copulated infre-
quently and indiscriminately, whereas control males copulated more often
with large than with small females (Berglund, 1993); and predation from
the cichlid fish, Crenicichla alta, reduced female preference in the guppy,
Poecilia reticulata (Godin and Briggs, 1996). Even so, Magnhagen (1991)
has cautioned that only in a very few cases has it been shown that individuals
actively change their mating behavior according to predation risk. Addi-
tional studies would be useful.
Overall, the stressful scenario in nature should reduce the tendency for
traits involved in the sexual selection process to become progressively more
extreme, thereby limiting the runaway process in which the sexual ornament
is continuously exaggerated (Fisher, 1930; Lande, 1980). Therefore, a trade-
off occurs, since the energy cost of the development and maintenance of
162 P. A . PARSONS
ornaments of increasing size is countered by the cost of stress (Parsons,

1995a).
The most extreme ornaments should therefore occur when the stress
level is relatively low in the background environment, and the size of
ornaments should fall with increasing stress. Species with morphologically
complex sexual ornaments should be vulnerable during periods of environ-
mental stress (McLain, 1993), such as extinction events. For instance,
McLain et al. (1995) record that sexually dimorphic bird species are more
vulnerable to extinction than are monomorphic species, following their
introductions into the oceanic islands of Oahu and Tahiti. Certainly, in the
initial occupation of new adaptive zones, sexually dimorphic species would
appear to be at a disadvantage compared with more generalist species,
because of the energy costs in developing and maintaining secondary sexual
characters and in sexual display, which in total may approach the maximum
limit of available energy (Moller, 1994a). In many bird species the cost of
secondary sexual ornaments can be reduced by an investment in physiologi-
cal and anatomical adaptations. These adaptations coevolve with the sec-
ondary sexual characters, thereby permitting levels of sexual display consid-
erably higher than those observed in their absence (Moller, 1996). This
indicates strong selection at energetic limits, implying extreme vulnerability
of birds to any increase in stress. Finally, and in accord with the previous
considerations, a recent theoretical analysis concludes that sexual selection
in a changing environment enhances population extinction by increasing
selection intensities on a male trait (Tanaka, 1996).
C. SPECIES
BOUNDARIES
In general, the available data on species boundaries are restricted to
successful species, and represent end points of adaptive change during the
speciation process. Accordingly, it is appropriate to consider briefly the
boundaries between closely related species, especially those that are
mainly sympatric.
The predominantly sympatric sibling species, D. rnelanogaster and D.
sirnulans, are distinguishable physiologically, based on differing resistances
to environmental extremes, especially high and low temperatures, and toxic
levels of ethanol and acetic acid. Furthermore, within these extremes, ecobe-
havioral differences indicate very different microhabitats for the species in
nature, especially for larvae. Even so, these species of subgenus Sophophora
are ecobehaviorally and physiologically very similar, and contrast substan-
tially with another widespread species, D. irnrnigrans of subgenus Drosoph-
ila (Ehrman and Parsons, 1981).
Similarly, marine sympatric invertebrate species show distinct habitat

preferences defined by depth, salinity, exposure, or preference for host
substrates (Knowlton, 1993). In carnivorous stoneflies (Plecoptera), eggs
are laid at similar times, but newly hatched larvae rarely occur together in
the same habitat, defined by hatching temperature, incubation temperature,
and time of hatching, thereby largely eliminating competition btween the
species (Elliott, 1995). In two closely related hymenopteran parasitoid spe-
cies of Drosophila, there is odor-mediated avoidance of competition, as
Leptopilina heterotoma can recognize patches on stinkhorns where groups
of L. clavipes females occur (Janssen et al., 1995).
These examples indicate that in responding to abiotic stresses, physiologi-
cal and ecobehavioral traits are important in adapting to habitats, and are
likely to be important targets of selection in evolutionary shifts underlying
the speciation process. Consequently, the resources of the environment
become utilized efficiently, as the sibling species do not directly compete
with each other. Divergence at the ecobehavioral and physiological levels
is therefore primary to morphological divergence. Only to the extent that
it has a functional role can morphology be regarded as a direct target of
selection (Bonner, 1988).
Thermal constraints on the time and energy budgets of lizards have been
investigated extensively, and upper and lower critical thermal limits can
be determined (Adolph and Porter, 1993). Canyon lizards, Sceloporus mer-
riami, have a characteristic body temperature of 32.2”C, which is lower
than that of other North American desert iguanids. Under this thermal
environment, individual activities (movement rate, feeding strikes, and so-
cial displays) are restricted to a 2-hr period beginning around local sunrise
and to a brief period in the late afternoon. When the average temperature
was around 32.2”C, maximum activity and maximum use of microenviron-
ments occurred. However, as the temperature deviated from 32.2”C, the
use of microenvironments became more constrained (Grant and Dunham,
1988; Adolph and Porter, 1993). Presumably, energy costs would increase
in parallel with divergence from 32.2”C.
As thermal regimes deviate from optimal, lizard activity becomes more
restrictive, which is a behavior that influences home range size, population
density, fecundity, social history, and ultimately survival. From populations
across a range of elevations, complex relationships have been established
between biophysical constraints and fitness mediated through daily time
budgets and seasonal energy-mass budgets. These relationships underlie
life-history variation and the adaptation of organisms to specific environ-
ments (Dunham et al., 1989).
Adolph and Porter (1993) argue for the importance of activity as a
connecting link between thermal environment and lizard life histories. The
164 P. A. PARSONS
implication is that lizard activity is a target of selection, which is reasonable

because of its energy cost. When the energy available for activity becomes
so restrictive that there is no discretionary energy for reproduction, growth,
or storage, species boundaries are likely to be located. This is in accord
with the physiological constraints that appear important at the margins of
distributions of an increasing number of animal taxa (see Section 1,A).
A N D THE SURVIVAL
111. VARIABILITY OF VARIANTS
Under environments that are demonstrably extreme, heterozygotes tend

to be favored, especially for polymorphisms in natural populations. Even
under less extreme conditions characteristic of the laboratory, the level of
heterozygosity of individual organisms in populations tends in some cases
to correlate with measures of performance or fitness, in particular growth
rate and developmental stability. Enzyme loci influencing metabolism and
contributing to the amount of energy available for development and growth
show the most significant positive associations with heterozygosity (Mitton,
1993). Consequently, heterozygotes should be differentially favored in
growth and reproduction as stress increases, and when resources become
limited (Parsons, 1996b).
In the white-tailed deer, Odocoileus virginianus, antler size, body mass,
fat levels, and other dimensions were found to be correlated with heterozy-
gosity, dominance status, and reproductive success by Scribner et al. (1989),
who emphasized the importance of metabolic efficiency of the heterozy-
gotes. In bighorn sheep, Ovis canadensis, horn size largely determines
breeding superiority, as large horns give access to estrous ewes. In addition,
such rams have superior foraging ability, energy efficiency, and disease
resistance (Hogg, 1987). In the seventh year of life, which is around the
time of onset of breeding, 21% of variation in horn volume can be explained
by an association with heterozygosity. In contrast, the horns of young
rams show little variation in size that is attributable to genetic factors
(Fitzsimmons et al., 1995). Therefore, when energy demands from the devel-
opment and maintenance of horns and from the mating process itself are
high, heterozygote advantage is maximal.
In any case, mating is an expensive process energetically, and fitness
assessed by mating success can often be related to available metabolic
energy (see Section 11,B). Consequently, heterozygotes should be favored
during mating, as documented in a number of species, especially insects
and fish (Thornhill and Gangestad, 1993). Furthermore, Rolan-Alvarez et
al. (1995) found a positive association between heterozygosity and sexual
fitness for males in populations of the marine snail, Littorina mariae.
B E H A V I O R A L V A R I A B I L I T Y AND EVOLUTIONARY ADAPTATION 165
In the context of this discussion, competition can be used as an example

of stress under laboratory conditions. In offspring from a diallel cross
involving three inbred strains of mice, several traits were studied in a normal
cage, and a smaller cage with enhanced crowding. In the normal cages,
14% of inseminated females did not produce offspring; 29.4% did not in
the smaller cages, suggesting that crowding reduced reproductive fitness.
Additive genetic variability increased under crowding stress, especially for
preimplantation mortality, litter size, and relative adrenal weight. For pre-
implantation mortality and litter size, nonadditive effects and heterozygote
advantage increased under stress (Belyaev and Borodin, 1982). Genetic
variability for behavioral traits can therefore be high under stress (Par-
sons, 1988).
These and many other examples show that under highly stressed situa-
tions, especially in free-living populations, genetic variability is not normally
expected to be restrictive. In addition to heterozygote advantage under
stress, there is a substantial body of data indicating increased mutation,
recombination, developmental variability, and phenotypic variability as
stresses approach levels where extinctions become a real possibility (Par-
sons, 1987; Hoffmann and Parsons, 1991).
For novel variants, the issue then turns to the conditions under which their
survival and reproduction is likely. Following Fisher (1930), the chances of
survival of a novel variant should be inversely related to the magnitude of
its phenotypic change, or in the context of this discussion, the energy cost
of the change. In addition, survival should be inversely related to the
magnitude of energy cost of existing in a variable environment, so that
the more extreme the environment, the smaller the change that can be
accommodated by organisms for their survival (Parsons, 1996b).
In summary, the level of genetic variability is unlikely to be restrictive
for adaptation, but the ecological conditions determining the survival of
the variants may be. It is in this light that the issue of extending the limits
of adaptation in populations will now be considered, for a continuum of
environments, from extreme at species borders to benign, where learning
is possible. The survival of variants should increase as conditions become
less severe, even though variants can apparently appear under all conditions.
THE LIMITS
I v . EXTENDING OF ADAPTATION
A. ABIOTIC
STRESSES
A N D RESOURCES
Speciation necessarily involves a shift in the limits of adaptation of estab-

lished species. Therefore, the process of speciation, approached ecologi-
166 P. A. PARSONS
cally, implies energy costs (Van Valen, 1976a,b). Shifts in limits can in
principle involve changes in resistance to abiotic stresses, changes in re-
source availability and usage, or a combination of these variables.
Commencing with abiotic stress, thermophilia occurs in some desert ant
genera, enabling successful foraging for arthropods that have succumbed to
extreme heat. For instance, the Saharan silver ant, Cataglyphis bombycina,
scavenges for the corpses of insects and other arthropods that have SUC-
cumbed to the heat stress of their desert environment in a small thermal
window with a maximum width of 7°C (46.5-53.6"C). The boundaries of
the window are underlain by predatory pressure exerted by a desert lizard
at the lower limit and heat stress at the upper limit (Wehner et al., 1992).
Parallel situations occur in the Australian ant, Melophorus bagoti (Christian
and Morton, 1992), and in the burrowing spiders Seothyra in Namib desert
dunes (Lubin and Henschel, 1990). Finally, the diamond above, Geopelia
cuneara, an inhabitant of the arid savannahs and semideserts of Australia,
is extremely heat tolerant, and consequently activity occurs throughout
the day under dry and hot conditions when potential predators and food
competitors are reduced (Schleucher, 1993).
These are examples of extreme abiotic stress, where predators and com-
petitors are likely to be absent, enabling the occupation of extreme habitats.
Ultimately, as found in the lizard, Scleroporus merriami, at extreme temper-
atures, the energy for activity becomes so restricted that there is no discre-
tionary energy for reproduction, growth, or storage, and species boundaries
occur (Adolph and Porter, 1993). Heat shock protein synthesis may occur
in association with thermotolerance, as found in the ant Cataglyphis (Gehr-
ing and Wehner, 1995). However, the formation of heat shock proteins
is likely to have a metabolic cost, thereby reducing fitness (Krebs and
Loeschke, 1994).
Consequently, assuming that species boundaries are regions of energy
restriction, it seems difficult to envisage much widening of windows of
opportunity for direct abiotic extremes of climatic origin, especially as the
survival of novel variants would be unlikely.
Turning to resources, innovation can involve ecobehavioral traits in shift-
ing to alternatives at stressful times (Parsons, 1993b). Examples include:
(1) the evolution of specialization of D. sechellia onto a single resource,
Morinda citrifolia, from a stressful window of opportunity from D. simulans,
which finds the resource toxic (R'Kha et al., 1991); (2) the evolution of
host races of a stem-galling tephritid, Eurosta solidaginis, on two goldenrod
host species assisted by a 10- to 14-day difference in emergence times on
the two hosts (Craig et al., 1993); and (3) the evolution of races of the
tephritid fly, Rhagoletis pomonella, from its native host hawthorn to intro-
duced fruits maturing at different times (Feder et al., 1988).
Therefore, resource heterogeneity can underlie divergence, especially if

associated with the simultaneous need to adapt to some abiotic stress. The
summed environmental change must be intense enough to cause disruptive
selection for sufficiently long that any incipient divergence can become
established, and consequently to have the potential to lead to isolation. A
possible example comes from the intertidal snail, Littorina saxatilis, where
there is assortative mating leading to incipient reproductive isolation associ-
ated with habitat selection by two morphs; one of these occurs in the upper-
shore barnacle belt and the other in the lower-shore mussel belt, indicating
physiological and behavioral adaptations by the morphs to the two differing
environments (Johannesson etal., 1995). In contrast, in the Galapagos finch,
Geospiza conirostris, partial isolation of a population based on resource
heterogeneity occurred following a drought, but prolonged divergence was
prevented by extreme fluctuations in the abiotic environment (Grant and
Grant, 1989).
B. RESOURCE
POLYMORPHISMS
Genetically based resource and habitat polymorphisms permit the occu-
pation of more than one niche within a species, and can underlie divergence
( West-Eberhard, 1986; Stanhope et al., 1992). For instance, sympatric popu-
lations of the tropical sponge-dwelling coral-reef shrimp, Synalpheus
brooksi, occupy two alternative host species of sponge, and in laboratory
situations tend to choose native sponge species. This promotes assortative
mating and hence divergence, as shown by significant host-associated
genetic divergence of shrimp in two of three reefs based on protein-
electrophoretic variation (Duffy, 1996).
In the Arctic charr, Salvelinus alpinus, there are benthivorous, planktivor-
ous, and/or piscivorous forms in lakes in Iceland, which show substantial
morphological, developmental, and behavioral specialization for discrete
resource categories. The behavioral differences break down when food
is artificially superabundant, occurring only in the nutritionally restricted
environments of free-living populations (Skulason et al., 1993). Under these
latter conditions, energy returns to charr appear to be maximized by genetic
divergence among morphs, enabling the efficient exploitation of differing
resource categories. Novel variants promoting such divergence would be
favored on grounds of energetic efficiency.
Schluter and McPhail (1993) record multiple examples of fish in low-
diversity postglacial lakes, where there are sympatric species involving
limnetic and benthic forms. The limnetic forms, which exploit plankton in
open water, are typically smaller, with a narrower mouth and longer, more
numerous gill rakers than the benthic forms, which consume larger prey;
168 P. A. PARSONS
the morphological differences are adaptations to differing food require-

ments. The benthic-limrtetic split could therefore be a predictable first step
in the diversification of many fish taxa (Schluter and McPhail, 1993), and
this split would be favored by energy efficiency in resource utilization.
While Skulason and Smith (1995) argue that resource polymorphism has
been underestimated as an evolutionary force leading to divergence, this
could be precluded by abiotic instability, as noted in the finch G. conirosfris.
More generally, during a 2-year California drought in 1976-1977, pressures
on development time intensified in colonies of the specialist insect herbi-
vore, the butterfly Euphdryas editha, because host plants senesced rapidly
(Ehrlich et al., 1980). Conversely, continuous rainfall can retard postdia-
pause larval development so that adult flight is delayed beyond plant senes-
cence (Dobkin et al., 1987). Such climatic perturbations would appear suffi-
cient to swamp the selection for energy-use efficiency based on resource
polymorphisms that can occur under more stable and less stressful abiotic
conditions. The utilization of heterogeneous resources therefore is likely
to be the most efficient when organisms d o not simultaneously need to
cope with the energy costs of extreme stresses (see also Section 1,B).
C. LEARNING
The finch Pinaroloxias inornata, of Cocos Island, Costa Rica, has ex-
tremely generalist feeding habits, spanning those of several families of birds
on the mainland, encompassing insects, Crustacea, seeds, fruits, nectar from
many flower species, and perhaps lizards. In contrast, individuals feed as
specialists year-round, often using just one of the many feeding techniques
and resources observed at the population level. Apparently, these special-
izations are transmitted at least partly culturally, from the observation of
other individuals. Hence, these tropical birds have developed learning abil-
ity, permitting the exploitation of heterogeneous resources (Werner and
Sherry, 1987), which implies high energy-use efficiency. The ecological
situation on Cocos Island appears permissive of this situation, as it is an
aseasonal environment with very few competing species, and with high
availability, variety, and predictability of food resources. Under these cir-
cumstances of relatively low energy constraints from the environment,
specialization has apparently occurred. Ultimately, such behavioral special-
ization could be assimilated genetically, perhaps following varying biochem-
ical demands made on finches from differing food categories. For instance,
differences in feeding behavior in the crustacean Gammarus palustris are
associated with genetic variation in the properties of a digestive enzyme
(Guarna and Borowsky, 1993).
B E H A V I O R A L VARIABILITY AND EVOLUTIONARY ADAPTATION 169
Another example of learning comes from bluegill sunfish, Lepomis macro-

chirus, in North American freshwater lakes, where learning assisted the behav-
ioral modification needed to search efficiently in vegetated and open-water
habitats (Ehlinger, 1990). Habitat-specific foraging efficiency occurred,
thereby increasing the energetic effectiveness of resource exploitation.
Therefore, behaviors governing resource use may be influenced by the
previous experience of individuals. In insects, prior exposure to a particular
resource can enhance a female’s tendency to oviposit on that type of re-
source. For instance, in the true fruit fly, Rhagoletis pomonella, the propen-
sity to accept a particular fruit prior to the deposition of an egg can be
modified by previous ovipositional experience (Prokopy and Papaj, 1988).
Learning should assist in exploiting windows of opportunity presented by
introduced fruit species, which would be enhanced if mating occurred at
the resource. Learning may therefore assist in adaptation to new hosts from
the original host.
Hence, following Baldwin (1896) learning may be a factor in switches into
novel habitats, thereby assisting in the integration of genetic components of
behavior into the gene pool. As learning eases the process of genetic change
(Anderson, 1995), the energy costs in the occupation of novel habitats would
be reduced. For instance, fifteen-spined sticklebacks, Spinachia spinachia,
attack Gamrnarus and Artemia more efficiently as a result of experience.
By decreasing handling time, learning increased the profitability of specific
prey, expressed in terms of energy expended per given time period (Croy
and Hughes, 1991).
V. FROM
STRESS-RESISTANCE TO A CONNECTED
GENOTYPES METABOLISM
A. STRESS-RESISTANCE
GENOTYPES
Koehn and Bayne (1989) argue that high stress resistance is associated
with the efficient use of metabolic resources for growth and reproduction,
especially when resources are limited. Since stress-resistance phenotypes
tend to have a low metabolic rate (Hoffmann and Parsons, 1991), a low
maintenance requirement is implied. Consequently, growth should be sup-
portable over a wide range of conditions. In particular, the association
between metabolic efficiency and stress resistance suggests that genes for
stress resistance should be favored during the metabolically costly process
of the development and maintenance of sexual ornaments and mating itself
(Parsons, 1995a).
During mating, the preferred male trait may reflect the underlying genetic
quality of the male, so that females mating with these males gain additional
170 P. A. PARSONS
advantages for themselves or their offspring outside of mating (Thornhill

and Alcock, 1983). Such advantages are conventionally regarded as being
under the control of “good genes,” which enhance fitness both during
the mating process, conferring direct benefits to females, or by producing
offspring with superior fitness (Moore, 1994).
Wedekind (1994) argued that sexual selection for stress-resistance genes
is important in improving the survival chances of offspring. In other words,
mate preferences would be most efficient if coupled with resistance genes in
parents and offspring. Following Hamilton and Zuk (1982), this conclusion
comes from an assessment of parasite-driven sexual selection.
In the pheasant, Phasianus colchicus, male spur length correlates with
male viability, female mate choice, and offspring survival (von Schantz et
al., 1996). Genetic analyses show that the major histocompatibility complex
genotype is associated with variation in male spur length and male viability.
Von Schantz et al. (1996) conclude that these data directly support the
“good genes” hypothesis (Hamilton and Zuk, 1982) that females discrimi-
nate among males based on secondary sexual characters, and so pass on
genes for disease resistance that improve offspring fitness.
In any case, a premium on stress resistance and hence metabolic efficiency
conferring overall fitness is expected, assuming that populations are nor-
mally exposed to high levels of stress (Parsons, 1996b, 1997a). Furthermore,
because “good genes” reflect fitness under these environmental conditions,
it should be possible to incorporate other fitness traits into this scheme.
For instance, in an African cockroach, Nauphoeta cinerea, females have
offspring that develop relatively quickly following mating with the most
attractive males (Moore, 1994). This suggests that the choosing female
prefers individuals carrying “good genes,” which also underlie rapid devel-
opment. Additional examples cited by Moore (1994) include heritable varia-
tion in plumage as an indicator of viability in male great tits, Parus major
(Norris, 1993), and improved growth and survival of offspring of peacocks,
Pavo cristatus, with more elaborate trains (Petrie, 1994). In the damselfly,
Zschnura graellsii, Corder0 (1995) found that the best predictor of male
lifetime mating success was mature life-span. In barn swallows, Hirundo
rustica, Mdler (1994b) found that offspring longevity is positively related
to that of their fathers, and to the ornament size of the male parent.
Considering aging, assuming that a long life-span and rapid development
depend on metabolically efficient stress-resistance genes, individuals having
high inherited stress resistance should develop fastest and live longest
(Parsons, 1996b,c). Accordingly, ornament size, mating success, longevity,
and development time can perhaps be viewed as a coordinated suite of
characters assuming the stressful environments of free-living populations. If
a major target of selection of stress is at the level of energy carriers, “good
genes” therefore should be stress-resistance genes, and these should be

increasingly important for ensuring fitness as limits to adaptation are ap-
proached.
O n a cautionary note, the paucity of empirical observations for this
predicted relationship presumably relates to the point that studies carried
out under relatively benign laboratory conditions are unlikely to be efficient
in revealing such associations, because selection for stress resistance is
necessarily less intense than in free-living populations. On the other hand,
irrespective of the background environment, associations of development
time and life-span with mating success and the size of sexual ornaments
should be the most readily detectable, because mating and the development
and maintenance of sexual ornaments are normally energetically expen-
sive processes.
Lifetime reproductive success has not been considered to any extent in
this article, and in any case it is usually strongly correlated with longevity.
However, direct extrapolations from laboratory to natural populations can-
not be assumed. For instance, in D. melanogaster, substantial and variable
deficiencies in food availability under natural conditions preclude the ex-
pression of reproductive potential (BoulCtreau-Merle et af., 1987). In any
case, adults of English Drosophila populations had a mean life expectancy
of 1.3 to 6.2 days, which is at least an order of magnitude less than survival
under equable laboratory conditions (Rosewell and Shorrocks, 1987). In a
recent review of genetic variation and aging, Curtsinger et al. (1995) argued
for a model where old and young fitness components are correlated, which
is in accord with a prediction from the stress theory of aging (Parsons,
1993b). Accordingly, survival at any age should be a predictor of lifetime
reproductive success in free-living populations (see Parsons, 1997b, where
this conclusion is considered in the light of various evolutionary theories
of aging).
B. FITNESS
A N D METABOLIC
EFFICIENCY
In Section 111, it was noted that heterozygosity levels tend to be correlated
with fitness during the mating process, especially for enzyme loci controlling
metabolism and hence energy availability. For instance, in bighorn sheep,
Hogg (1987) argued that this association reflects a “good genes” strategy
favoring heterozygotes at an energetically demanding time. Extending to
other fitness traits, in particular development rate but also life-span, sub-
stantial evidence suggests that heterozygosity tends to be associated with
high fitness in a wide range of taxa, especially under stressful circumstances
(Mitton, 1993; Parsons, 1996b,c, 1997a).
172 P. A . PARSONS
Consequently, two approaches imply parallel associations for a range of

fitness traits. The first approach commences at the whole organism level
and leads to genes for stress resistance for promoting fitness, while the
second approach commences at the gene level using electrophoretic vari-
ants, and leads to generalized heterozygous advantage for promoting fitness
under stress. Although these approaches have developed largely indepen-
dently, they can be linked by a requirement for metabolic efficiency in the
face of the stress to which free-living populations are normally exposed.
Detailed gene location studies based on natural populations appear neces-
sary for additional elaborations. The generalized advantage of heterozy-
gotes under stress does, however, suggest that many interacting loci may
be involved in promoting metabolic efficiency, so that it appears more
appropriate to talk of “good genotypes” than good genes. This does not,
of course, preclude the involvement of some major genes, such as the more
anodal allozymehozyme at the phosphoglucose isomerase locus, which is
favored in a range of stressful situations, including high temperature, high
salinity, anoxia, and desiccation in natural populations of a wide range of
taxa (Riddoch, 1993).
In summary, a wide-ranging literature suggests that stress resistance and
metabolic efficiency are associated for a range of fitness measures (Table
I). The ranking 1 to 10 in Table I represents a continuum of organizational
levels ranging from the essentially molecular (1) to the organismic. The
items of main concern in this paper are categories 7 and 8, and correlations
with other life-history characteristics, especially 6 and 9, are noted under 10.
TABLE 1
ASSOCIATIONS
PREDICTED
I N STRESSED
FREE-LIVING
POPULATIONS“
~~~~~~~
1. Stress-resistance genes
2. High (electrophoretic) heterozygosity
3. High vitality, vigor. and resilience
4. High homeostasis in response to external
stresses
5. Low fluctuating asymmetry
6. Rapid development
7. High male mating success
8. Extremes of sexual ornaments
9. Long life span
10. Positive correlations among fitness traits
a See Parsons (199%. 1966b.c. 1997a) for detailed

discussions from which this table was derived.
BEHAVIORAL V A R I A B I L I T YAND EVOLUTIONARY ADAPTATION 173
Items 3-5 are various measures of homeostasis, from the morphological

to physiological levels. For instance, survival to an old age is associated
with high vitality, vigor, and resilience (3), and high homeostasis in response
to external stresses (4). Fluctuating asymmetry (FA) measures the degree
to which an individual can control development under given environmental
and genetic conditions ( 5 ) , and is a measure of individual phenotypic quality
or fitness (Zakharov, 1989; Parsons, 1990; Markow, 1994; Polak, 1997;
Moller, this volume). One manifestation of energy dissipation is increased
FA (Mitton, 1993), for instance the FA of antlers of reindeer is positively
correlated with parasite intensity (Folstad ef al., 1996). Accordingly, low
F A should occur in organisms in which metabolic efficiency or fitness is
highest (5). Similarly, low FA should be associated with genes for stress
resistance, which implies that FA should be heritable to some extent, as
noted from associations between development rate and life-span (Parsons,
1996d), but more generally from a meta-analysis of 29 studies of 13 species,
which revealed a mean heritability of FA of 0.27 (Moller and Thornhill,
1997).
Furthermore, because the level of heterozygosity of organisms tends to
correlate with performance or fitness, correlations with high FA should be
maximal in heterozygotes. There are now sufficient data sets to infer that
rapid development, a long life-span, success in mating, and extremes of
sexual ornaments tend to be associated with low FA, and this tends to be
clearest in heterozygotes. However, there is a need to devise laboratory-
based experiments to model the environments of free-living populations
to obtain additional empirical data to explore these apparent and rather
tentative generalizations more directly.
Finally, the extreme stress scenario, which is the basic assumption under-
lying this paper, gains support from Kauffman (1993) who argues that the
normal situation faced by organisms is an extremely perturbed world. Under
these circumstances, he argues that a connected metabolism is important for
the facilitation of adaptive change in response to environmental challenges.
There is a convergence with the model in this paper, as the associations in
Table I are underlain by selection by environmental stress, which targets
energy carriers in free-living populations. In any case, an energetic approach
to fitness has appeared previously. For instance, Van Valen (1976b) argued
that energy underlies fitness, which can then be viewed as the rate at
which resources, exceeding those needed for growth and maintenance, are
available for reproduction in the broadest sense (Brown et al., 1993). In
the context of this paper, mating, the development and maintenance of
sexual ornaments, and various nonsexual behaviors can extract substantial
energy from resources, and so may be critical in determining limits to
adaptation in extreme environments.
174 P. A. PARSONS
V1. SUMMARY
Energy expenditure is a prerequisite for organisms to exist in any habitat,

as exposure to biotic and especially abiotic stress is the norm in free-living
populations. Therefore, the distribution and abundance of organisms can
be related to energy balances, derived from the costs of various stresses
interacting with gains from resources. Consequently, the behavioral selec-
tion of preferred habitats imposing low energy costs is adaptive. On the
other hand, limits to adaptation occur when available energy becomes
totally restrictive. Therefore, energetically costly behaviors, especially those
involving sexual selection, are important in determining limits.
Assuming that species borders are regions of energy restriction, it is
difficult to envisage much widening of windows of opportunity for direct
stresses of climatic origin. However, when combined with resource hetero-
geneity, evolutionary divergence led by behavioral shifts appears more
likely, provided that abiotic perturbations are not extreme. In abiotically
benign environments, implying minimal energy constraints, resource use
specialization by learning appears possible.
Heterozygotes tend to be favored in extreme environments because of
their energy and metabolic efficiency. Therefore, genetic variability is un-
likely to be restrictive in stressed free-living outlier populations; however,
ecological circumstances can preclude the survival of novel variants. Conse-
quently, the primary key to understanding limits to adaptation for behav-
ioral traits is likely to be ecological.
Under stressed free-living conditions, favored “good genotypes” are
likely to be stress resistant and heterozygous. An association between suc-
cess in mating, the development of extreme sexual ornaments, rapid devel-
opment, and a long life can be postulated based on the metabolic efficiency
of stress-resistance genotypes. While these postulated associations among
fitness traits are supported by only limited empirical evidence, they may
be important in any habitat where organisms are close to their limits of
survival. If many organisms are born but few survive to reproduce because
of climatic stress interacting with and causing nutritional stress, this situation
may be quite normal. Although I contend that this model of the environ-
ment is generally valid, a reference point for comparisons with more benign
environments is certainly provided.
References
Adolph. S. C.. and Porter, W. P. (1993). Temperature, activity and lizard life histories. Am
Nnt. 142, 213-295.
Anderson, R. W. (1995). Learning and evolution: A quantitative genetics approach. J. Theor.

Biol. 175, 89-101.
Arking, R.. Buck, S., Wells, R. A,. and Pretzalff. R. (1988). Metabolic rates in genetically
based long lived strains of Drosophilu. Exp. Gerontol. 23, 59-76.
Baldwin, J. M. (1896). A new factor in evolution. Am. Nut. 30, 441-451.
Barros, A. R., Sierra, L. M.. and Comendador, M. A. (1991). Decreased metabolic rate as
an acrolein resistance mechanism in Drosophilu melunogaster. Behav. Genet. 21,321-328.
Belyaev, D. K., and Borodin. P. M. (1982). The influence of stress on variation and its role
in evolution. Biol. Zentralbl. 100, 705-714.
Bennett, A. F. (1991). The evolution of activity capacity. . IExp. B i d . 160, 1-23.
.
Berglund, A. (1993). Risky sex: Male pipefishes mate at random in the presence of a predator.
Anim. Behav. 46, 169-175.
Blaxter, K. (1989). “Energy Metabolism in Animals.” Cambridge University Press, New York.
Boggs, C. L. (1994). The role of resource allocation in understanding reproductive patterns.
In ”Individuals, Populations and Patterns in Ecology” (S. R. Leather, A. D. Walt, and
N. J. Mills, eds.), pp. 25-33. Intercept Ltd., Andover, Hampshire.
Bonner, J. T. (1988). “The Evolution of Complexity by Means of Natural Selection.” Princeton
University Press, Princeton, NJ.
Booth, D. T., Clayton, D. H., and Block, B. A. (1993). Experimental demonstration of
the energetic cost of parasitism in free-ranging hosts. Proc. R. Soc. London Ser. B
223, 123-129.
Boulttreau-Merle, J., Fouillet, P., and Terrier, 0. (1987). Seasonal variations and balanced
polymorphisms in the reproductive potential of temperate D. melanoguster populations.
Entomol. Exp. Appl. 43, 39-48.
Bozinovic, F., and Rosenmann, F. (1989). Maximum metabolic rate of rodents: Physiological
and ecological consequences in distributional limits. Acnct. Ecol. 3, 173-181.
Brown, C. R., Brown, M. B., and Rannala, B. (1995). Ectoparasites reduce long-term survival
of their avian host. Proc. R. Soc. London, Ser. B 262, 313-319.
Brown. J. H.. Marquet. P. A,, and Taper, M. L. (1993). Evolution of body size: Consequences
of an energetic definition of fitness. Am. Nut. 142, 373-384.
Christian, K. A., and Morton, S. R. (1992). Extreme thermophilia in a Central Australian
ant, Melophoriis bugoti. Physiol. Zool. 65, 885-905.
Clayton. D. H. (1990). Mate choice i n experimentally parasitized rock doves: Lousy males
lose. Am. Zool. 30, 251-261.
Cordero. A. (1995). Correlates of male mating success in two natural populations of the
damselfly Ischnuru gruellsii (Odonata: Coenagrionidae). Ecol. Entomol. 20, 213-222.
Craig, T. P.. Itami. J. K., Abrahamson. W. G., and Horner, J. D. (1993). Behavioral evidence
for host-race formation in Eitrostasolidaginis. Evolution (Lawrence, Kuns.)47,1696-1710.
Croy, M. I., and Hughes. R. N. (1991). The role of learning and memory in the feeding
behavior of the fifteen-spined stickleback. Spinachiu spinachia L. Anim. Behav. 41,
149-159.
Curtsinger, J. W., Fukui, H. H., Khazaeli, A. A,, Kirscher. A,, Pletcher, S. D., Promislow.
D. E. L., and Tartar, J. (1995). Genetic variation and aging. Annu. Rev. Genet. 29,553-575.
De Kruipf, H. A. M. (1991). Extrapolation through hierarchical levels. Comp. Biochem.
Physiol. IOOC, 291-299.
de Lope, F., Gonzalez. G., Perez, J. J.. and Moiler. A. P. (1993). Increased detrimental effects
of ectoparasites on their bird hosts during adverse environmental conditions. Oecologiu
95, 234-240.
Dobkin, D. S., Olivieri, I.. and Ehrlich. P. R. (1987).Rainfall and the interaction of microclimate
with larval resources in the population dynamics of checkerspot butterflies (Euphydryus
editha) inhabiting serpentine grassland. Oecologica 71, 161-166.
176 P. A. PARSONS
Duffy, J . E. (1996). Resource-associated population subdivision in a symbiotic coral-reef

shrimp. Evolution (Lawrence, Kans.) 50, 360-373.
Dunham, A. E., Grant, B. W., and Overall, K. L. (1989). Interfaces between biophysical and
physiological ecology and the population ecology of terrestrial vertebrate ectotherms.
Physiol. 2001.62, 335-355.
Dunson, W. A,, and Travis, J . (1991). The role of abiotic factors in community organization.
Am. Nat. 138, 1067-1081.
Ehlinger. T. J. (1990). Habitat choice and phenotype-limited feeding efficiency in bluegill:
Individual differences and tropic polymorphism. Ecology 71, 886-896.
Ehrlich. P. R., Murphy, D. D.. Singer, M. C., Sherwood. C. B., White, R. R., and Brown,
I. L. (1980). Extinction, reduction, stability and increase: The responses of checkerspot
butterfly (Euphydryas) populations to the California drought. Oecologia 46, 101-105.
Ehrman, L., and Parsons, P. A. (1981). “Behavior Genetics and Evolution.” McGraw-Hill,
New York.
Elliott, J . M. (1995). Egg hatching and ecological partitioning in carnivorous stoneflies (Plecopt-
era). C. R. Acad. Sci., Life Sci. 318, 237-243.
Feder, J . L., Chilcote, C. A,, and Bush, G. L. (1988). Genetic differentiation between sympatric
host races of the apple maggot fly Rhagoletis pomonella. Nature (London) 336,61-64.
Fisher, R. A. (1930). “The Genetical Theory of Natural Selection.” Clarendon Press. Oxford.
Fitzsimmons, N. N., Buskirk, S. W., and Smith, M. H. (1995). Population history, genetic
variability, and horn growth in bighorn sheep. Conserv. Biol. 9, 314-323.
Folstad, I., Arneberg, P., and Kanter. A. J. (1996). Antlers and parasites. Oecologia 105,
556-558.
Ganem, G., and Nevo, E. (1996). Ecophysiological constraints associated with aggression.
and evolution towards pacifism in Spalax ehrenbergi. Behav. Ecol. Sociobiol. 38,245-252.
Gehring, W. J., and Wehner. R. (1995). Heat shock protein synthesis and thermotolerance
in Cataglyphis, an ant from the Saharan desert. Proc. Natl. Acad. Sci. U.S.A.92,2994-2998.
Godin, J.-G. J., and Briggs, S. E. (1996). Female mate choice under predation risk in the
guppy. Anim. Behav. 51, 117-130.
Good, D. S. (1993). Evolution of behaviours in Drosophila melanogaster in high temperatures:
Genetic and environmental effects. J. Insect Physiol. 39, 537-544.
Grant, B. R. (1996). Pollen digestion by Darwin’s finches and its importance for early breeding.
Ecology 77, 489-499.
Grant, B. W., and Dunham, A. E. (1988). Thermally imposed time constraints on the activity
of a desert lizard Scleroporus merriami. Ecology 69, 167-176.
Grant, P. R., and Grant,B. R. (1989). Sympatricspeciation and Darwin’s finches. In “Speciation
and its Consequences” (D. Otte and J. A. Endler, eds.), pp. 433-457. Sinauer, Sunder-
land, MA.
Guarna, M. M., and Borowsky. R. L. (1993). Genetically controlled food preference: Biochemi-
cal mechanisms. Proc. Natl. Acad. Sci. U.S.A. 90, 5257-5261.
Hall, C. A. S., Stanford, J. A,, and Hauer. F. R. (1992). The distribution and abundance of
organisms as a consequence of energy balances along multiple environmental gradients.
Oikos 65,377-390.
Haller, J.. and Wittenberger, C. (1988). Biochemical energetics of hierarchy formation in
Betta splendens. Physiol. Behav. 43, 447-450.
Halliday, T. R. (1978). Sexual selection and mate choice. In “Behavioural Ecology: An
Evolutionary Approach” ( J . R. Krebs and N. B. Davies. eds.), pp. 180-213. Black-
well, Oxford.
Hamilton, W. D., and Zuk. M. (1982). Heritable true fitness and bright birds: A role for
parasites? Science 218,384-387.
Hoffmann, A. A.. and Blows. M. W. (1994). Species borders: Ecological and evolutionary
perspectives. Trends Ecol. Evol. 9, 223-227.
Hoffmann, A. A., and Parsons, P. A. (1991). “Evolutionary Genetics and Environmental
Hoffmann, A. A.. and Turelli, M. (1985). Distribution of Drosophila melanogasrer on alterna-
tive resources: Effects of experience and starvation. A m . Nai. 126, 662-679.
Hogg, J. T. (1987). Intrasexual competition and mate choice in Rocky Mountain bighorn
sheep. Ethology 75, 119-144.
Huey, R. B. (1991). Physiological consequences of habitat selection. Am. Nat. 137, S91-Sl15.
Hurst, R. J., Watts. P. D., and Oritsland, N. A. (1991). Metabolic compensation in oil-expressed
polar bears. J. Therm. Biol. 16, 53-57.
Janssen, A., Van Alphen, J. J. M.. Sabelis, M. W., and Bakker, K. (1995). Odour-mediated
avoidance of competition in Drosophila parasitoids: The ghost of competition. Oikos
73,356-366.
Johannesson. K., Rolan-Alvarez, E.. and Ekendahl. A. (1995). Incipient reproductive isolation
between two sympatric morphs of the intertidal snail. Lirrorina saxafilis. Evolution (Law-
rence, Kans.) 49, 1180-1 190.
Jones, J. S., Coyne, J. A., and Partridge, L. (1987). Estimation of thermal niche of Drosophila
melanogasrer using a temperature-sensitive mutation. Am. Nar. 130, 83-90.
Kauffman. S. A. (1993). “The Origins of Order: Self-organization and Selection in Evolution.”
Oxford University Press, New York.
Kennedy. C. E. J., Endler, J. A,, Poynton, S. L., and McMinn, H. (1987). Parasite load predicts
mate choice in guppies. Behav. Ecol. Sociobiol. 21, 291-295.
Klieber, M. (1961). “The Fire of Life: An Introduction to Animal Energetics.” Wiley, New
York.
Knowlton, N. (1993). Sibling species in the sea. Annic. Rev. Ecol. Syst. 24, 189-216.
Kodric-Brown, A., and Nicoletto, P. F. (1993). The relationship between physical condition
and social status in pupfish, Cyprinodon pecosensis. Anim. Behav. 46, 1234-1236.
Koehn. R. K., and Bayne, B. L. (1989). Towards a physiological and genetical understanding
of the energetics of the stress response. B i d . J. Linn. Soc. 37, 157-171.
Krebs, R. A., and Loeschcke, V. (1994). Costs and benefits of activation of the heat-shock
response in Drosophila melanogasier. Fiinci. Ecol. 8, 730-737.
Lande, R.(1980). Sexual dimorphism. sexual selection, and adaptation in polygenic characters.
Evolution (Lawrence, Kans.) 34, 292-305.
Li, J., and Margolies. D. C. (1994). Responses to direct and indirect selection on aerial dispersal
behaviour in Tetranychirs urticae. Heredity 72, 10-22.
Lovegrove. B. G . (1986). The metabolism of social subterranean rodents: Adaptation to
aridity. Oecologia 69, 551-55.5.
Low, B. S. (1990). Marriage systems and pathogen stress in human societies. Am. Zool.
30, 325-339.
Lubin, Y. D., and Henschel, J. R. (1990). Foraging at the thermal limit: Borrowing spiders
(Seothyru, Eresidae) in the Namib desert dunes. Oecologia 84,461-467.
Magnhagen, C. (1991). Predation risk as a cost of reproduction. Trends Ecol. Evol. 6,183-186.
Marden, J. H., and Waage, J. K. (1990). Escalated damselfly territorial contests are energetic
wars of attrition. Anim. Behav. 39, 954-959.
Markow, T. (1994). “Developmental Instability: Its Origins and Evolutionary Implications.”
Kluwer, Dordrecht, The Netherlands.
McLain, D. K. (1993). Cope’s rules, sexual selection, and the loss of ecological plasticity.
Oikos 68,490-500.
178 P. A. PARSONS
McLain, D. K., Moulton, M. P., and Redfearn, T. P. (1995). Sexual selection and the risk of
extinction of introduced birds on oceanic islands. Oikos 74, 27-34.
Mitton, J. B. (1993). Enzyme heterozygosity, metabolism, and developmental variability.
Genetica 89, 47-63.
Mdler, A. P. (1991). Clutch size, nest predation, and distribution of avian unequal competitors
in a patchy environment. Ecology 72, 1336-1349.
Mdler, A. P. (1994a). “Sexual Selection and the Barn Swallow.” Oxford University Press,
Oxford.
Meller, A. P. (1994b). Male ornament size as a reliable cue to enhanced offspring viability
in the barn swallow. Proc. NUB Acad. Sci. U.S.A. 91, 6929-6932.
Mdler, A. P. (1996). The cost of secondary sexual characters and the evolution of cost-
reducing traits. Ibis 138, 112-1 19.
M~iller,A. P., and Thornhill, R. (1997). A meta-analysis of the heritability of developmental
stability. J. Evol. B i d . 10, 1-16.
Moore, A. J. (1994). Genetic evidence for the “good genes” process of sexual selection.
Behav. Ecol. Sociobiol. 35, 235-241.
Nevo, E. (1991). Evolutionary theory and processes of active speciation and adaptive radiation
in subterranean mole rats, Spalax ehrenbergi superspecies, in Israel. Evol. B i d . 25,l-125.
Nevo, E., Simson, S., Heth, G., and Beiles, A. (1992). Adaptive pacifistic behaviour in subterra-
nean mole rats in the Sahara desert, contrasting to and originating from polymorphic
aggression in Israeli species. Behaviour 123, 70-76.
Norris, K. (1993). Heritable variation in a plumage indicator of viability in male great tits
Parus major. Nature (London) 326,537-539.
Odum, E. P., Finn, J. T., and Franz, E. H. (1979). Perturbation theory and the subsidy stress
gradient. BioScience 29, 349-352.
Parsons, P. A. (1987). Evolutionary rates under environmental stress. Evol. Biol. 21,311-347.
Parsons, P. A. (1988). Behavior, stress and variability. Behav. Genet. 18, 293-308.
Parsons, P. A. (1990). Fluctuating asymmetry: An epigenetic measure of stress. B i d . Rev.
Cambridge Philos. Soc. 65, 131-145.
Parsons, P. A. (1991). Evolutionary rates: Stress and species boundaries. Annu. Rev. Ecol.
Syst. 22, 1-16.
Parsons, P. A. (1992). Evolutionary adaptation and stress: The fitness gradient. E d . B i d .
26, 191-223.
Parsons, P. A. (1993a). Habitat preference: An interaction between genetic variability and
the costs of stress. Etologia 3, 1-9.
Parsons, P. A. (1993b). Evolutionary adaptation and stress: Energy budgets and habitats
preferred. Behav. Genet. 23, 231-238.
Parsons, P. A. (1995a). Stress and limits to adaptation: Sexual ornaments. J . Evol. B i d .
8,455-461.
Parsons, P. A. (1995b). Inherited stress resistance and longevity: A stress theory of ageing.
Heredity 75, 216-221.
Parsons, P. A. (1996a). Competition versus abiotic factors in variably stressful environments:
Evolutionary implications. Oikos 75, 129-132.
Parsons, P. A. (1996b). Stress, resources, energy balances, and evolutionary change. Evol.
Biol. 29, 39-72.
Parsons, P. A. (1996~).Rapid development and a long life: An association expected under a
stress theory of ageing. Experientia 52, 643-646.
Parsons, P. A. (1997a). Success in mating: A co-ordinated approach to fitness through genotypes
incorporating genes for stress resistance and heterozygous advantage under stress. Behav.
Genet. 27, 75-81.
BEHAVIORAL VARIABILITY A N D EVOLUTIONARY ADAPTATION 179
Parsons, P. A. (1997b). Stress-resistance genotypes, metabolic efficiency and interpreting

evolutionary change. In “Stress, Adaptation and Evolution” ( V . Loeschcke and R.
Bijlsma, eds.), pages 291-305. Birkhaueser, Basel.
Petrie, M. (1994). Improved growth and survival of offspring of peacocks with more elaborate
trains. Nature (London) 371, 598-599.
Polak, M. (1997). Parasites, fluctuating asymmetry and sexual selection. In “Parasites: Effects
on Host Hormones and Behavior” (N. E. Beckage, ed.). Chapman & Hall, New York.
In press.
Pomiankowski. A. (1987). The costs of choice in sexual selection. J. Theor. Biol. 128,195-218.
Porter, W. P., and Gates, D. M. (1969). Theormodynamic equilibia of animals with environ-
ment. Ecol. Monogr. 39,227-244.
Pough. F. H. (1989). Organismal performance and Darwinian fitness: Approaches and interpre-
tations. Physiol. Zool. 62, 199-236.
Prinzinger. R., and Siedle, K. (1988). Ontogeny of metabolism, thermoregulation and torpor
in the house martin, Delichon urbica (L.) and its ecological significance. Oecologia 76,
307-312.
Prokopy, R. J.. and Papaj, D. R. (1988). Learning of adult fruit biotypes by apple maggot
flies. J. Insect. Behav. 1, 67-74.
Riddoch, B. J. (1993). The adaptive significance of electrophoretic mobility in phosphoglucose
isomerase (PGI). Biol. J. Linn. Soc. 50, 1-17.
R’Kha, N., Capy. P., and David, J. R. (1991). Host-plant specialization in the Drosophila
melanogaster species complex: A physiological, behavioral, and genetical analysis. Proc.
Natl. Acad. Sci. U.S.A. 88, 1835-1839.
Rodriguez, L., Sokolowski, M. B., and Carton, Y. (1991). Intra- and inter-specific variation
in population behaviors of Drosophila from different habitats. Can. J. Zool. 69,2616-2619.
Rodriguez, L.. Sokolowski. M. B., and Shore, J. S. (1992). Habitat selection by Drosophila
mrlanogaster larvae. J. Evol. Biol. 5, 61-70.
Rolan-Alvarez, E., Zapata, C., and Alvarez, G. (1995). Multilocus heterozygosity and sexual
selection in a natural population of the marine snail Littorina mariae (Gastropoda: Proso-
branchia). Heredity 75, 17-25.
Root, T. ( I 993). Effects of global climate change on North American birds. In “Biotic Interac-
tions and Global Change” (P. M. Kareiva, J. G. Kingsolver, and R. B. Huey, eds.), pp.
280-292. Sinauer, Sunderland, MA.
Rosewell, J . , and Shorrocks, B. (1987). The implication of survival rates in natural populations
of Drosophila; Capture-recapture experiments in domestic species. Biol. J. Linn. SOC.
32,373-384.
Reskaft. E., Jarvi. T.. Bakken, M., Bech, C., and Reinertsen, R. E. (1986). The relationship
between social status and metabolic rate in great tits (Parus major) and pied flycatchers
(Ficedula hypolercca) Anim. Behav. 34, 838-842.
Ryan, M. J. (1988). Energy, calling, and selection. Am. Zool. 28, 885-898.
Schall, J. J., and Sarni, G. A. (1987). Malarial parasitism and the behavior of the lizard,
Sceloporus occidentalis. Copeia 1, 84-93.
Schleucher, E. (1993). Life in extreme dryness and heat: A telemetric study of the behaviour
of the diamond dove Geopelia cuneata in its natural habitat. Emu 93, 251-258.
Schluter, D., and McPhail, J. D. (1993). Character displacement and replicate adaptive radia-
tion. Trends Ecol. Evol. 8, 197-200.
Scribner, K. T . ,Smith, M. H.. and Johns, P. E. (1989). Environmental and genetic components
of antler growth in white-tailed deer. J . Mammal. 70, 284-291.
Skulason, S., and Smith, T. B. (1995). Resource polymorphism in vertebrates. Trends E d .
Evol. 10. 366-370.
180 P. A. PARSONS
Skulason. S.. Snorrason. S. S., Ota. D.. and Noakes. D. L. G. (1993). Genetically based
differences in foraging behaviour among sympatric morphs of arctic charr (Pisces: Salmo-
nidae). Anirn. Behav. 45, 1179-1 192.
Stanhope, M. J., Leighton, B. J., and Hartwick. B. (1992). Polygenic control of habitat prefer-
ence and its possible role in sympatric population subdivision in an estuarine crustacean.
Heredity 69, 279-288.
Tanaka, Y. (1996). Sexual selection enhances population extinction in a changing environment.
J. Theor. Biol. 180, 197-206.
Thornhill, R., and Alcock. J. (1983). “The Evolution of Insect Mating Systems.” Harvard
University Press, Cambridge, MA.
Thornhill. R., and Gangestad. S. W. (1993). Human facial beauty: Averageness, symmetry
and parasite resistance. Hum. Net. 4, 237-269.
Van Valen, L. (1976a). Ecological species, multispecies, and oaks. Taxon 25, 233-239.
Van Valen. L. (1976b). Energy and evolution. Evol. Theory 1, 179-229.
Vehrencamp, S. L., Bradbury. J. W.. and Gibson, R. M. (1989). The energetic cost of display
in male sage grouse. Anim. Behav. 38,885-896.
von Schantz,T.. Wittzell, H.. Goransson, G.. Grahn, M.. and Persson, K. (1996). MHC genotype
and male ornamentation: Genetic evidence for the Hamilton-Zuk model. Proc. R. Soc.
London, Ser. B 263, 265-271.
Wedekind, C. (1994). Handicaps not obligatory i n sexual selection for resistance genes.
J. Theor. Biol. 170, 57-62.
Wehner. R.. Marsh, A. C., and Wehner, S. (1992). Desert ants on a thermal tightrope. Nature
(London) 357,585-587.
Weiner, J. (1992). Physiological limits to sustainable energy budgets in birds and mammals:
Ecological implications. Trends Ecol. Evol. 7 , 384-389.
Werner, T. K., and Sherry, T. W. (1987). Behavioral specialization in Pinaroloxias inornara, the
“Darwin’s finch” of COCOS Island, Costa Rica. Proc. Natl. Acad. Sci. U.S.A.84,5506-5510.
West-Eberland, M. J. (1986). Alternative adaptations, speciation, and phylogeny (A review).
Proc. Natl. Acad. Sci. U.S.A. 83, 1388-1392.
White, T. C. R. (1993). “The Inadequate Environment: Nitrogen and the Abundance of
Animals.” Springer-Verlag. Berlin.
Yamamoto, A. H. (1994). Temperature preferences of Drosophila immigrans and D. virilis:
Intra- and interpopulation genetic variation. Jpn. J. Genet. 69, 67-76.
Ye, S., Aracena, J., Good. D. S., and Bell, W. J. (1994). Correlation between survival during
food deprivation and search behaviour in populations of Drosophilu melanogasfer.
J . Insect Physiol. 40, 137-142.
Zakharov. V. M. (1989). Future prospects for population phenogenetics. Sov. Sci. Rev.,Sect.
F: Physiol. Gen. Biol. 4, 1-79.
Zotin, A. I . (1990). “Thermodynamic Bases of Biological Processes: Physiological Reactions
and Adaptations.“ de Gruyter, Berlin.
Zuk, M., Thornhill, R., Ligon, J. D., and Johnson, K. (1990). Parasites and mate choice in
red jungle fowl. Am. Zool. 30, 235-244.
ADVANCES IN THE STUDY OF BEHAVIOR, VOL 27
Developmental Instability as a General Measure

of Stress
ANDERS
PAPE MBLLER
LABORATOIRE D'ECOLOGIE
U N I V E R S I T ~PIERRE ET MARIE CURIE
PARIS CEDEX 5, FRANCE
I. INTRODUCTION
Any organism, animal, plant, or fungus, is characterized by regularity of

its phenotype. That is why we often recognize species and sometimes even
sexes and age classes of particular species. A number of different mecha-
nisms ensure that development does not go wrong, and that developmental
processes are kept within certain limits. Disruptions of developmental tra-
jectories are caused by developmental noise from the environment, but
also inferior developmental processes caused by the genetic setup of the
individual. This ability to control development under given environmental
conditions is called developmental stability. Developmental stability cannot
be measured directly, but deviations from a regular phenotype provide
information on developmental instability. A number of such measures have
been proposed including fluctuating asymmetry, the frequency of phenode-
viants, and deviations from modal behavior, physiology, and immunology.
Most organisms display bilateral or radial symmetry, and random devia-
tions from such symmetry is termed fluctuating asymmetry (Ludwig, 1932;
Van Valen, 1962; Palmer and Strobeck, 1986; Parsons, 1990; Moller and
Swaddle, 1997). In other words, characters displaying fluctuating asymmetry
have signed right-minus-left trait values and normal frequency distributions
with a mean value of zero (Fig. 1). The level of asymmetry of an individual
belonging to a population demonstrating fluctuating asymmetry is simply
called its asymmetry or individual fluctuating asymmetry; the term fluctuat-
ing asymmetry, however, is a population parameter. It was only recently
that asymmetry at the level of the individual was used as a measure of
phenotypic quality (Moller, 1990). This individual approach is a most power-
ful tool in many different kinds of studies, as we shall see later. Two
other kinds of morphological asymmetry are common: Antisymmetry is
181
Cvpyriphf 0 I9YX hy Academic Press
All rights ol reproduction i n any iorm rcwrvcd.
IK)hS-34S4/98 $2S.M1
182 ANDERS PAPE MOLLER
A Directional asymmetry
-10 -8 -6 -4 -2 0 2 4 6 8 10
B Antisymmetry
2ol
16
-10 -8 -6 -4 -2 0 2 4 6 8 10
c Fluctuating asymmetry
-10 -8 -6 -4 -2 0 2 4 6 8 10
Right-minus-left character value
FIG. 1. Frequency distributions of signed right-minus-left character values for traits demon-
strating ( A ) directional asymmetry, (B) antisymmetry, and (C) fluctuating asymmetry. Adapted
from Moller and Swaddle (1997).
DEVELOPMENTAL INSTABILITY AND STRESS 183
characterized by individuals being asymmetric in a random direction, and

the frequency distribution of signed left-minus-right character values having
a deficiency of symmetric phenotypes (Fig. 1). Typical examples of antisym-
metry are the signaling claw of male fiddler crabs Uca spp. and the beak of
crossbills Loxia spp. The third kind of asymmetry is directional asymmetry,
which is usually displayed in a particular direction (Fig. 1). The mean value
of the frequency distribution of signed left-minus-right character values
therefore deviates significantly from zero. Examples of directional asymme-
try are the size of testes in mammals and the structure of ears in certain
species of owls. There is currently some controversy over whether only
fluctuating asymmetry, or also the two other kinds of asymmetry, reflects
developmental instability, although a number of cases clearly suggest that
antisymmetry and directional asymmetry may reflect poor developmental
conditions (review in Moller and Swaddle, 1997).
Developmental instability can also be estimated from other measures of
deviant phenotypes. Gross abnormalities such as a position of the heart in
the right side of the body cavity in some humans and four or six rather
than five fingers on each hand are termed phenodeviants. Their frequency
is actually positively correlated with fluctuating asymmetry and phenodevi-
ants therefore reflect developmental instability (e.g., Rasmuson, 1960). A
number of other valid and useful measures have been proposed for specific
kinds of organisms, and their common feature is the morphological invari-
ance. For example, snails grow their shells at a specific, constant angle,
which results in more and more narrow whorls, but when exposed to envi-
ronmental perturbations such as those caused by acid rain, the angle of
growth changes (Graham et al., 1993). The deviation from the normal angle
of growth is therefore a measure of developmental instability. Many kinds
of plants have composite leaves consisting of leaflets that are typically
exactly juxtaposed to one another. Deviations from commonly encountered
environmental conditions result in the stalks of the leaflets being displaced,
and the average deviation from perfectly juxtaposed leaflets is therefore a
measure of developmental instability (Freeman et al., 1993). A number of
other measures of developmental instability are listed by Graham ef al.
(1993) and Mgller and Swaddle (1997).
It is important to emphasize that phenomena other than developmental
problems may give rise to normal frequency distributions of signed left-
minus-right character values. Since individual fluctuating asymmetry is often
very small, with the majority of all individuals having asymmetries less
than 1% of the size of the character, measurement errors may contribute
significantly to the overall estimate of fluctuating asymmetry. In fact, mea-
surement errors also have normal frequency distributions with a mean value
of zero! Scientists working on fluctuating asymmetry traditionally test for
184 ANDERS PAPE M0LLER
the magnitude of measurement errors in a sample of individuals or in

the entire sample. The importance of measurement error can thereby be
evaluated or directly partialled out (see Palmer, 1994, and Moller and
Swaddle, 1997, for methods).
Organisms demonstrate regularity of their phenotypes because such regu-
larity promotes superior performance. Just think of the wealth of asymmetry
within the human body and contrast that with the exterior phenotype. Race
horses with symmetric skeletons win more races (Manning and Ockenden,
1994). Symmetry is the best solution to the engineering problem of con-
structing phenotypes that are well designed for locomotion. Many organisms
are sessile and do not need streamlined bodies for efficient locomotion.
Even though selection for symmetry may be most severe in mobile organ-
isms, there is still selection, albeit weaker, for symmetry in fungi, plants,
and sessile animals such as corals and sponges. The reason is that symmetry
also gives rise to more efficient resource use (such as light and nutrients)
and dispersal of propagules (such as spores, gametes, and seeds), but also
results in less severe effects of the abiotic environment such as wave action
and wind. Any deviation from symmetry is likely to impose performance
costs, and a number of such costs have been proposed or directly demon-
strated experimentally (see review in Moller and Swaddle, 1997). The opti-
mal solution therefore appears to be given a priori; it is a symmetric pheno-
type. Bilateral or radial symmetry therefore differs from any other
phenotypic measure because we know the optimum in advance. We might
be able to identify the optimum for other traits, but not without extensive
research and then only for an environment with particular characteristics.
If developmental control is a costly process, then perfect symmetry might
not be the optimal solution, since the incremental decrease in asymmetry
achieved by further investment is a function with diminishing returns. We
might be able to cope with average asymmetries of .1 mm in the length of
our fingers, while 1 mm or 10 mm might pose problems in certain situations.
Developmental instability is affected by a wide range of environmental
(external) and genetic (internal) factors that contribute to disruption of
the stable development of the phenotype. These factors are briefly reviewed
below. The control of developmental processes proceeds most efficiently
under the commonly encountered environmental conditions, and increasing
deviations from such conditions result in the use of energy for stress toler-
ance that could otherwise be used for developmental control, growth, repro-
duction, and survival (Hoffmann and Parsons, 1989; Parsons, 1990; Alek-
seeva et al., 1992; Ozernyuk et al., 1992). Development, like any other
process, is an energy-dissipating activity, and energy used for developmental
control has to be diverted away from other vital activities. As the future
performance of any organism depends on the developmental stability of
DEVELOPMENTAL INSTABILITY A N D STRESS 185
its phenotype, deviations from this goal can be used as a reliable measure
of the challenge experienced by an individual in its natural environment.
It is inherently impossible to generate a perfectly symmetrical phenotype;
the level of fluctuating asymmetry will provide extremely important infor-
mation about the developmental performance of an individual. This argu-
ment is of utmost importance because it allows us to obtain reliable informa-
tion about the state of individual organisms in their natural environment
as perceived by the organisms themselves. We cannot readily ask organisms
how they perceive their environment, but we can use their developmental
instabilities as an indirect answer to this question.
If fluctuating asymmetry provides reliable information on the well-being
of populations, and individual asymmetry does the same for individuals,
then it should suffice to measure a single character and extract the informa-
tion. However, it is a common finding that asymmetries in different charac-
ters often are not significantly positively correlated (review in Mdler and
Swaddle, 1997). Why should that be the case? As already stated, develop-
mental instability measured as individual fluctuating asymmetry integrates
the effects of a number of different environmental and genetic factors, and
if we were to rerun the development of an individual once more, we might
not end up with exactly the same level of asymmetry. The estimate of
developmental instability is basically an estimate of a variance based on a
single measurement of two different morphological characters, and such
an estimate is bound to have a high degree of uncertainty. The ability
of an individual to develop the same phenotype repeatedly is called the
repeatability of its individual fluctuating asymmetry. This quantity can be
estimated readily when we make some simplifying assumptions and know
the coefficient of variation of the asymmetry measure, the magnitude of
our measurement errors, and the phenotypic variance of the character in
question (Whitlock, 1996). The correlation in symmetry among a number
of different characters is likely to provide a serious underestimate of the
true correlation because of the lack of repeatability of developmental insta-
bilities. An unbiased estimate of the true correlation turns out simply to
be the correlation between the asymmetries divided by the square root
of the product of the repeatabilities (Whitlock, 1996). There are other
explanations for this general absence of a correlation between asymmetries
of different characters (review in Mgller and Swaddle, 1997), but the expla-
nation presented above is perhaps the most likely. A way of resolving this
problem when choosing characters for measurement is to choose a number
of different characters and use a composite measure of asymmetry as an
indicator of the overall level of developmental instability.
The ability to control developmental processes and generate a stable
phenotype differs among individuals, and both genetic and environmental
components affect this ability, as for any other character. A large number of
studies of different organisms have shown that there is indeed a statistically
significant heritability of measures of developmental instability (Mdler and
Thornhill, 1997a). Although the quality of these studies differs in a number
of ways that may affect the estimates of the genetic and environmental
components, independently of how the data are selected the conclusion
remains stable: developmental instability has a statistically significant herita-
bility (Houle, 1997; Leamy, 1997; Markow and Clarke, 1997; Palmer and
Strobeck, 1997; Pomiankowski, 1997; Swaddle, 1997; Whitlock and Fowler,
1997; Moller and Thornhill, 1997b). The signficance of this finding is that
relatives will resemble each other with respect to developmental instability,
and that developmental stability may evolve.
In the following three sections, I briefly review (1) the genetic and envi-
ronmental determinants of developmental instability, (2) the relationship
between developmental instability and mode of selection, and (3) the rela-
tionship between developmental instability and fitness.
A N D ENVIRONMENTAL
11. GENETIC DETERMINANTS
OF
DEVELOPMENTAL
INSTABILITY
Developmental instability as estimated from fluctuating asymmetry and

the frequency of phenodeviants has been investigated in many hundreds
of studies, and some general patterns have emerged concerning the factors
that contribute to increased developmental problems. These can basically
be divided into internal genetic and external environmental causes, which
are briefly reviewed here. An extensive review is provided by Mdler and
Swaddle (1997).
A. CAUSESOF DEVELOPMENTAL
GENETIC INSTABILITY
The genetic factors that increase developmental instability include in-
breeding, homozygosity, hybridization, and mutation. Inbreeding results in
a reduction in additive genetic variation, but also in the exposure of deleteri-
ous recessive alleles that become fully expressed in recessive homozygotes.
Either of these effects may disrupt the stable development of a phenotype.
A recent review of the literature has shown that a large majority of the
studies indeed found developmental instability to be negatively associated
with inbreeding (Moller and Swaddle, 1997). Exceptions to this finding may
be explained in a number of different ways of which the selective loss of
asymmetric homozygotes at early embryonic stages is a plausible one.
Heterozygosity at protein encoding loci is an indicator of genetic variabil-

ity, and individuals that are able to produce a more diverse array of biochem-
ical products should be capable of coping with a wide range of environmen-
tal conditions. There is considerable evidence in agreement with this
suggestion, and metabolism and therefore also growth have repeatedly
been shown to be most efficient among heterozygous individuals (review
in Mitton and Grant, 1984). If stable development is associated with efficient
metabolism and controlled growth processes, one might hypothesize a posi-
tive relationship between developmental homeostasis and heterozygosity
(as first suggested by Lerner, 1954). A large number of studies have ad-
dressed this question, but with very mixed results. A meta-analysis (a statis-
tically based review of overall effects and heterogeneity in effects among
studies) revealed that there was no consistent association between heterozy-
gosity and measures of developmental instability such as fluctuating asym-
metry (Vdlestad et al., 1997). However, there was statistically highly sig-
nificant heterogeneity in the effects among studies. Some of this could be
explained by whether the study organisms were heterothermic or homeo-
thermic (the latter providing a more stable and protected developmental
environment). Therefore, this review provided little evidence for a general
association. Vdlestad et al. (1997) discussed several explanations for this
lack of an association. A particularly likely explanation is that many studies
have been performed under relatively benign laboratory conditions that
do not result in severe environmental stress. Perhaps a high degree of
heterozygosity is beneficial only under stressful conditions. This hypothesis
has been tested explicitly by Mulvey et al. (1994) on the fish Gambusia
holbrooki. Fluctuating asymmetry and heterozygosity were negatively re-
lated at high temperatures, but completely unrelated at an optimal water
temperature. Another possibility is that the relationship holds only for
specific enzymes that are of particular functional importance in a specific
context (Mitton, 1995). This could certainly account for some of the results
for specific enzymes, but not for relationships between asymmetry and
general heterozygosity as found in a number of different studies.
Hybridization results in the genomes of two species being mixed, and this
may have severe fitness consequences because the functioning of biological
systems depends on the cooperation of the different components of the
genome. Of course, hybridization also results in the generation of novel
genetic variation and may therefore in some cases of closely related species
result in hybrid vigor. Divergence and reproductive isolation are outcomes
of the process of speciation, which may result in the acquisition of different
coadapted gene complexes. If different genomes are combined, this may
result in disruption of development because the gene combinations of the
hybrids have not been subject to natural selection. This will obviously
particularly be the case when different species have been isolated from
each other for a long time rather than recently (in evolutionary terms).
Experiments with eggs of the Florida large mouth bass, Micropterus salmon-
ides floridanus, fertilized by the sperm of ten different species resulted in
an increase in developmental deviants as the divergence measured by the
genetic distance increased (Parker et al., 1985). A review of the literature
demonstrated a clear general pattern of increased developmental instability
in hybrids as compared to the parental species, particularly if the species
in question had diverged considerably (Moller and Swaddle, 1997).
Mutations usually result in a deterioration of the phenotype, but every
now and then a slight improvement may arise. It has been known for a
long time that mutations usually result in deviant phenotypes with proper-
ties similar to those of asymmetric and phenodeviant individuals (Gold-
Schmidt, 1940, 1955). Particularly mutations with a low penetrance give
rise to deviant phenotypes, while highly penetrant mutations show less
phenotypic effects with respect to developmental regularity (Goldschmidt,
1940, 1955). Again, the explanation for this effect appears to be the lack
of genetic coadaptation. If genes are manipulated by use of recent molecular
techniques, the resultant phenotype becomes asymmetric for ETS2-alleles,
which have effects similar to those of Down’s syndrome (Sumarsono et al.,
1996). the vascular endothelial growth factor gene (Carmeliet et af., 1996),
and homeobox genes (Davis et al., 1995).
In conclusion, a number of different genetic factors contribute to the
development of a stable phenotype, although no consensus exists concern-
ing the relative roles of the different factors. The genetic mechanisms
involved in generating asymmetric phenotypes appear to be gene coadapta-
tion, but potentially also other mechanisms.
B. ENVIRONMENTAL
CAUSES
OF DEVELOPMENTAL
INSTABILITY
Environmental factors include temperature, food, pollutants, population
density, sound, light, and parasites. The diversity of environmental stresses
that have been shown to cause an increase in asymmetry is probably not
exclusive; many other kinds of stress might provide similar effects.
Temperatures that deviate from optimal conditions result in increased
energy expenditure for stress resistance. Increased temperature differences
from the normal range encountered have been shown to result in increased
asymmetry in Drosophila (Beardmore, 1960), rats Rattus norvegicus (Gest
et al., 1983, 1986), and a number of other organisms (review in Moller and
Swaddle, 1997).
Nutritional stress has been shown to increase asymmetry in a number of
different organisms under experimental conditions. For example, European
nuthatches, Sitta europaea, that had a couple of feathers removed during

winter, regrew these feathers more symmetrically when provided with extra
food than did control birds (Nilsson, 1994). Similar results have been ob-
tained for a wide variety of organisms (Mdler and Swaddle, 1997).
Free-living organisms encounter a diverse chemical environment during
development, particularly in species with external fertilization. Deviations
from commonly encountered conditions result in increased developmental
instability. This is the case for several different kinds of pollutants, but also
for chemicals found in the food consumed by animals. For example, gray
seals, Hulichoerus grypus, from the Baltic had increased asymmetry in their
skulls during the 1950s and 1960s (Zakharov and Yablokov, 1990), but
experienced a decrease in asymmetry as the concentrations of pollutants
decreased in recent decades (Zakharov ef al., 1989). In a similar vein,
alcohol consumption in pregnant women resulted in increased asymmetry
in their children (Kieser, 1992). Further examples are discussed in Sec-
tion V,A.
Increasing population density results in a reduction in the amount of
nutrients available per individual, but also in energy spent on stress resis-
tance that could otherwise be used for control of developmental processes.
Several studies have shown that asymmetry and phenodeviants increase as
a consequence of increased density. For example, elevated larval density
of Australian sheep blowflies, Lucilia cuprina, resulted in increased asym-
metry in the adult flies (Clarke and McKenzie, 1992). Similarly, although
under more natural conditions, skeletal asymmetry followed population
density in the small mammal cycles of the common shrew, Sorex araneus,
in Siberia (Zakharov et al., 1991). A final example derives from a study of
similarly aged clones of poplars, Popitlus arnericanus, planted at three
different densities (Rettig et aZ., 1997). The effect of density on asymmetry
increased linearly from a density of .167 to 2.0 plants per square meter.
Audiogenic stress may also increase asymmetry of the phenotype. Early
experiments by Siege1 and Smookler (1973) demonstrated that pregnant
rats that were exposed to noise had pups with increased dental asymmetry.
This effect has been repeated in a number of subsequent experiments on
rats and other rodents (review in Mdler and Swaddle, 1997).
A final example of an environmental component that can increase asym-
metry is exposure to predators (Witter and Lee, 1995). Molting starlings,
Sturnus vulgaris, were kept in aviaries with food provided near or away
from shelter. Hence, there was no differential exposure to predators per
se, but just a perceived difference in exposure. Starlings that developed
their feathers while feeding at an exposed food source developed signifi-
cantly greater feather asymmetry than did controls. As asymmetric individu-
als are more likely t o fall prey to predators (Section IV), increased perceived
190 ANDERS PAPE M 0 L L E R
risks of predation may actually result in increased predation, if increased

morphological asymmetries give rise to reduced performance.
In conclusion, the development of an individual integrates the effects of
a wide range of environmental and genetic factors that affect the stability
of developmental processes. This is an advantage for the scientists because
the overall effect of many different factors is added up into the phenotype,
but a disadvantage because we will not obtain information on the particular
factor that is causing an increase in asymmetry.
111. DIRECTIONAL
SELECTION
A N D DEVELOPMENTAL
INSTABILITY
The previous section dealt with a range of different genetic and environ-
mental factors that tend to increase the level of fluctuating asymmetry.
However, this is not the entire story, as characters of individuals even
when conditions are kept constant still may differ in their measures of
developmental instability. The recent (in evolutionary terms) history of
selection affecting a character may strongly influence the potential for
development of asymmetry. Characters that have been subject to a history
of directional or disruptive selection are generally less stable than characters
subject to a history of stabilizing selection (Moller and Pomiankowski,
1993a,b; Moller and Swaddle, 1997). The reason for this phenomenon ap-
pears to be that intense directional selection selects against any mechanisms
that control the full expression of a character. These control mechanisms
are also involved in the stable expression of the phenotype. Stabilizing
selection has the opposite effect of incorporating developmental mecha-
nisms that prevent the expression of extreme phenotypes, but also avoid
the expression of asymmetric phenotypes. The evidence for this scenario
comes from a range of different sources (sexual selection, life-history traits,
plant-animal interactions, the paleontological record, domestication, and
laboratory experiments) of which three are mentioned below.
Sexual selection results in the evolution of extravagant characters that
are costly to produce and maintain, and therefore do not ameliorate the
effects of natural selection (Darwin, 1871). The divergence in secondary
sexual characters is generally much larger than in ordinary morphological
traits, apparently leading to pre- and postcopulatory species isolation mech-
anisms. This extreme divergence is evidence of a recent history of evolution-
ary change. We should therefore expect secondary sexual characters in
general, but particularly those currently subject to intense directional selec-
tion, to have elevated levels of asymmetry. This appears to be the case in
some comparative studies (Moller 1992b; Moller and Hoglund, 1991), but
not in others (Balmford et al., 1993; Tomkins and Simmons, 1995). This
apparent discrepancy may be explained by differences in the patterns of

asymmetry between species in which there is currently a female mate prefer-
ence for the most ornamented traits and those in which there is no such
preference (Moller, 1993b). A number of different studies have investigated
in specific studies or experiments whether secondary sexual characters are
more susceptible to the negative effects of stress, and this appears generally
to be the case (review in Moller and Swaddle, 1997).
Animal and plant breeding results in intense directional selection during
a large number of generations to achieve preferred phenotypes. We know
from the appearance of plants and animals that there have been dramatic
responses to selection as determined from the extreme variance in pheno-
types (Darwin, 1868). Just attend an exhibition of cats, dogs, or poultry
and the diversity of morphology will become apparent. We should therefore
expect to find considerably more asymmetry in these domesticated forms
than in their wild ancestors. This appears to be the case. It is well known
to pet breeders that abnormal numbers of digits appear at a frequency
much higher than in free-living populations. Similarly, domestic strains of
chickens are on average considerably more asymmetric than their wild
ancestors in the jungles in Southeast Asia (Moller et al., 1995a). The null
expectation might reasonably be the opposite, as free-living jungle fowl,
Callus gallus, necessarily must be more severely restricted by limited access
to food and more frequent exposure to debilitating parasites. A confounding
factor is that many domesticated animals and plants may have suffered
from the negative effects of bottlenecks and inbreeding, factors that are
known to result in increased morphological asymmetry.
A more reliable source of information is the large number of laboratory
selection experiments performed over the years. This literature is reviewed
in Moller and Swaddle (1997). Directional and disruptive selection experi-
ments generally result in an increase in asymmetry, while stabilizing selec-
tion has the opposite effect of reducing asymmetry. Since most of these
experiments are performed in ways that avoid inbreeding, and as they
usually only last relatively few generations, before any depletion of additive
genetic variance has taken place, this provides the most firm evidence for
the relationship between mode of selection and developmental instability.
In conclusion, there is evidence from a number of different sources
suggesting that characters that have diverged because of a recent history
of directional selection are developmentally less stable than characters that
have been subject to a history of stabilizing selection. This observation
implies that the former kinds of traits may be more suitable for determining
the effects of adverse genetic and environmental conditions on develop-
mental instability.
192 ANDERS PAPE M 0 L L E R
IV. FITNESSCORRELATES
OF DEVELOPMENTAL
INSTABILITY
If performance generally depends on developmental stability, then one

should predict individual fluctuating asymmetry to be a reliable predictor
of fitness. Indeed, this appears to be the case. First, asymmetrical individuals
suffer more from intra- and interspecific competition and have an elevated
probability of becoming parasitized and falling prey to a predator. Second,
sexual selection and mating success in particular have been shown persis-
tently to depend on morphological asymmetry. Finally, other fitness compo-
nents such as growth performance, clutch size, offspring survival, and adult
survival have been shown often to be inversely related to asymmetry.
Ecological and behavioral studies of interactions usually depend on the
ability to identify individuals differing in their ability to perform well in an
interaction. Asymmetric individuals indeed appear to perform less well than
symmetric ones in a number of different kinds of interactions. Parasitism
differentially affects asymmetric hosts, and this is caused by greater suscepti-
bility in at least some cases (Moller, 1996~). Predation also affects asymmet-
ric indivduals differentially in organisms as diverse as domestic flies, Musca
domestica, preyed upon by dung flies, Scatophaga stercoraria, and by barn
swallows, Hirundo rustica, and barn swallows being preyed upon by Euro-
pean sparrowhawks, Accipiter nisus (Moller, 1996d; review in Moller and
Swaddle, 1997). Finally, intraspecific and interspecific competition has been
shown in a few cases to affect asymmetry. An experimental study of foliar
asymmetry in clones of poplars showed that both intraspecific competition
as determined from three levels of population density, but also interspecific
competition as determined from absence or presence of an herb layer
increased leaf asymmetry (Rettig et al., 1997). Given these negative relation-
ships between asymmetry and performance in interactions, it is perhaps
not surprising that the fitness of asymmetric individuals generally appears
to be reduced.
Sexual selection results from nonrandom variance in mating success being
associated with particular phenotypes. A general finding is that females of
a wide variety of species prefer males with more extreme phenotypes. Male
secondary sexual characters that are larger and brighter, vocalizations that
are louder, and pheromones that are more powerful result in higher mating
success (Andersson, 1994). Because secondary sexual characters have been
subject to a recent history of intense directional selection (as stated earlier),
they often have elevated levels of fluctuating asymmetry. These asymmet-
ries are often the direct target of sexual selection, as demonstrated by a
large number of observational and experimental studies. A recent meta-
analysis of developmental instability and sexual selection based on 61 stud-
ies of 36 species of animals revealed compelling evidence for an intermedi-
ately sized effect of asymmetry on mating success (Moller and Thornhill,

1997~).An extremely large number of unpublished, negative results would
have to exist in order to nullify this effect. There was considerable statis-
tical heterogeneity in the data, and some of this could be explained
by (1) lack of tests for fluctuating asymmetry and measurement errors,
(2) poor experimental design that resulted in unwanted side effects of
the manipulation, (3) experimental studies demonstrating stronger effects,
apparently because experiments control potentially confounding variables,
and (4) asymmetry in secondary sexual characters generally showing
stronger effects than asymmetry in ordinary morphological traits (Moller
and Thornhill, 1997~).There is also evidence suggesting that asymmetry
plays an important role in pollinator preferences and sexual selection in
plants (Moller and Swaddle, 1997).
A number of other fitness components have also been shown to be
associated with morphological asymmetry (Moller, 1997). A review showed
that 10 of 12 studies found increased growth performance, 16 of 17 studies
found increased fecundity, and 19 of 21 studies found increased survival
rates of the more symmetric individuals (Mdler, 1997). These findings
are particularly remarkable, given the different methods of study and the
observational nature of the approach in most studies. As developmental
instability appears to have a heritable component (Mdler and Thornhill,
1997a,b), asymmetric parents should also on average produce relatively
asymmetric offspring. A review has demonstrated that viability selection
often acts against asymmetric gametes, embryos, and juvenile individuals,
and that parents may use developmental selection against offspring with
deviant phenotypes as a way of allocating resources to viable offspring
(Mgller, 1997b). Such developmental selection against asymmetric pheno-
types is widespread in both animals and plants.
The final part of this chapter considers the potential uses of develop-
mental instability as a tool in a number of different contexts. Behavioral
biologists will find information of interest for their own field of research.
V. PRACTICAL
USESOF DEVELOPMENTAL
INSTABILITY
With the information on the causes and consequences of developmental

instability provided in the previous paragraphs in mind, we can start consid-
ering how this kind of information about the adaptation of individuals to
their natural environment can be used. A few examples are provided here
and more can be found in Mdler and Swaddle (1997). The topics covered
include (1) environmental monitoring, (2) conservation biology, (3) animal
welfare, (4) human and veterinary medicine, and ( 5 ) behavioral studies.
A. ENVIRONMENTAL
MONITORING
Large numbers of biologists are employed to monitor the environment
for regional, national, and international bodies. Vast sums of money are
spent on monitoring environmental quality, which is usually done by simply
determining measures of quality such as the presence or abundance of so-
called indicator species as well as the pollution levels of air and water.
When indicator species disappear, this is often a sign of severe, irreversible
damage. We would often like to interfere before this state has been reached.
Sublethal effects of environmental deterioration would therefore often be
preferred. How should we proceed if we want to know how animals and
plants perceive their environment? Measures of developmental instability
become elevated well before severe effects on fitness components appear
(e.g., Graham et al., 1993c), and asymmetry and similar measures may
therefore be useful early indicators of the level of stress in the environment
as experienced by free-living animals. Alternatively, the same strain of
laboratory animals such as fruitflies or mosquitoes could be used to assess
the environmental conditions across a range of sites. This would allow
testing of environmental conditions with homogeneous strains, but also
allow the use of particularly susceptible strains for monitoring. For example,
the shaker mutant of Drosophila melanogaster is particularly susceptible
to environmental stress because of the high activity level of flies with this
mutation. They may for this reason be particularly useful for assessment
of the effects of pollutants on developmental instability (Parsons, 1991).
The sublethal effects of pollutants on phenotypic expression is an aspect
of utmost importance. Because there is usually strong natural and sexual
selection against individuals with asymmetric phenotypes (see Section IV),
severe negative effects of pollution may appear to become hidden or even
absent if asymmetric individuals are selectively removed from the popu-
lation.
There is a relatively large number of studies of the effects of a range of
different kinds of pollutants on the asymmetry of plants and animals. The
pollutants range from heavy metals and organic compounds to electromag-
netic radiation and radioactivity (review in Maller and Swaddle, 1997).
Early work on Arabidopsis thaliana revealed that radiation caused an in-
crease in developmental instability (Bagchi and Iyama, 1983). Further ex-
amples are now available from Chernobyl in Ukraine. Maller (1993a) inves-
tigated the relationship between morphological asymmetry and exposure
to radiation in the barn swallow. Individuals were captured and measured in
a contaminated area near the nuclear power plant and in an uncontaminated
control area southeast of Chernobyl. Samples predating the contamination
accident were obtained from museum collections for both areas. There was
a statistically significant increase in asymmetry in tail length (a secondary

sexual character) in the Chernobyl area after the contamination, but not
in other characters in males or in any characters in female barn swallows.
This result shows that there was a differential effect on the secondary sexual
character. A subsequent study five years later confirmed these results and
demonstrated increased mutation rates in swallows from Chernobyl (Elle-
gren et al., 1997). Asymmetry in three species of plants (black locust tree,
Robinia pseudoacacia, rowan, Sorbus aucuparia, camomile, Matricaria per-
forata) sampled along a gradient from Chernobyl toward uncontaminated
areas in Southeast Ukraine revealed elevated asymmetries near Chernobyl
(Moller, 1997a). The asymmetries of the three species were concordant
across the gradient, indicating that the asymmetry was responding to the
same environmental conditions, and asymmetries were positively correlated
with radiation from a radioactive isotope of cesium ('37Cs) in soil samples.
A number of studies have determined the effects of less severe pollutants
on the stable expression of phenotypes. For example, heavy metal pollution
is common around large melter factories in various parts of the world. Leaf
asymmetry in two species of birch Betufa was severely elevated near such
sources of pollution in Finland and Russia, and the degree of foliar asymme-
try was directly related to the concentration of the pollutant (Kozlov et al.,
1996). Additional studies around a range of different kinds of chemical
factories in Russia have documented similar effects on different measures
of developmental instability in plants (Freeman et al., 1993). A number of
studies concerning animals are mentioned in Mraller and Swaddle (1997).
In conclusion, a diversity of organisms respond to sublethal exposure to
pollutants by developing increased asymmetries. The effects of pollutants
may thereby be assessed directly without using the traditional LD50 crite-
rion (the dose at which 50% of a population has died), and the sublethal
effects may also be more consistent with the generally accceptable levels
of exposure to pollutants.
B. BIOLOGY
CONSERVATION
Conservation biology is concerned with the factors that determine the
sustainability of viable populations of animals in their natural habitats (e.g.,
Meffe and Carroll, 1994). Species are threatened for a number of different
reasons including habitat destruction and other kinds of human activity as
well as reductions in genetic variability of populations. These factors can
be considered to result in a deterioration of the environment as perceived
by the organisms in question. Although some species are better able to cope
with environmental deterioration than others (Parsons, 1994), a number of
different environmental and genetic factors are directly associated with
increased stress. Again, the argument is the same as that already developed
at the start of this chapter: Different kinds of stress give rise to a deteriora-
tion of the energy balance of an individual, and may result in poor develop-
mental control independent of whether this is due to a reduction in the
availability of food, shelter, or other kinds of vitally important resources.
A reliable and objective insight into the ways that the environment is
perceived by organisms can be gained directly from their individual fluctu-
ating asymmetries or other measures of developmental instability.
One of the first studies to adopt this approach was based on dental
asymmetry in gorillas Gorilla gorilla (Manning and Chamberlain, 1993).
The habitats of gorillas have suffered from continuous destruction with
severe reductions in the living conditions of large parts of the populations.
If secondary sexual characters are particularly susceptible to stress, for the
reasons stated previously, we should especially expect to see an increase
in the degree of asymmetry of gorilla canine teeth, but less so for other
kinds of teeth. This was exactly the pattern that was found. Asymmetry
in the sexually size-dimorphic canines has increased considerably since
the beginning of the last century; this is not the cause for sexually size-
monomorphic teeth. This study suggests that the living conditions of gorillas
indeed have deteriorated during the last 150 years.
Large proportions of threatened animals are currently found in national
parks throughout the world, and they are superficially safe from threats
that otherwise may cause declines and extinction of less well protected
populations. Park populations are often sold or culled because of rapid
increases in numbers, and decisions have to be made concerning which
animals to remove. These decisions are not easily made on reasonable
scientific grounds. One possibility that has not been considered is that
individuals with symmetric phenotypes may have characteristics that allow
them to cope better with stressful conditions. Selective culling of asymmetric
individuals would actually be indirect selection for increased stress resis-
tance. A study addressing this question concerns the gemsbok, Oryx gazella,
a large antelope confined to very dry habitats such as dry savannahs and
deserts in southern Africa (Moller et al., 1996). Gemsboks have long, lance-
like horns that are used for interactions and antipredator defense. Adults
of both sexes were in better condition if they had symmetrical horns. Fights
between individuals of the same sex with similarly sized horns were also
most often won by individuals with symmetric horns, which therefore had
differential access to limiting resources such as drinking water. Adult fe-
males more often had a calf if their horns were symmetric, and males with
symmetric horns more often had access to females than did the flock-living
males with asymmetric horns. A couple of gemsbok eaten by predators
all had asymmetric horns. These observations suggest that symmetrical
individuals indeed were more fit than asymmetrical individuals, and that
symmetrical individuals would contribute more disproportionately to the
maintenance of a growing population than would asymmetric ones.
A final example concerns a number of different species of butterflies
that have become threatened and in many cases have disappeared from
large areas during recent decades. Poulsen (1997) investigated whether
butterfly species that were currently threatened in Denmark differed in
their wing asymmetry from closely related, common species. This pairwise
comparative approach helped control for a number of different factors
that potentially could affect developmental instability. Butterflies that were
currently threatened had considerably higher degrees of asymmetry than
their sister species (Poulsen, 1997). A second comparison determined
whether there had been a temporal increase in asymmetry in the threatened
species. Again, there was evidence for a significant increase in wing asymme-
try in the threatened species, but not in the common sister species. Although
the direct cause of the asymmetry cannot be pinpointed, this study provided
evidence for fluctuating asymmetry being a reliable predictor of future
conservation status.
In conclusion, conservation biological studies may benefit from the use
of measures of developmental instability for assessment of how organisms
perceive their environment. A number of other ways in which measures
of developmental instability can be of use in the context of conservation
biology are discussed by Clarke (1995), Sarre et al. (1994), and Moller and
Swaddle (1997).
C. ANIMAL
WELFARE
Scientific questions of animal welfare consider ways in which to decide
objectively about the state of animals and ways in which to ameliorate poor
conditions. A number of different solutions to these problems of welfare
have been suggested (e.g., Broom and Johnson, 1993; Toates, 1995), but
none of these approaches has satisfied the farmer community and the
decision makers. The reason is that there are no objective, a priori ways
of determining whether a specific criterion for rearing animals will improve
conditions in any appreciable way. Common measures of stress such as
behavioral or physiological variables are themselves subject to selection
during the domestication process, and they may not provide reliable infor-
mation about the welfare state of animals. Developmental instability can
be used as a direct measure of how animals and plants perceive their
environment, since we know a priori the optimal phenotypic solution to
the engineering problem of constructing a well-functioning organism; this
is a symmetric phenotype. Deviations from perfect symmetry can therefore
be used to objectively assess environmental conditions in whichever way

is of current interest. The optimal solution is a very low level of asymmetry,
and if a particular environmental gradient is suspected to cause increased
levels of stress, this can be assessed directly from the relationship between
asymmetry and increased values of the environmental variable.
A concrete example of this approach can be found in a study of chicken
asymmetry in relation to rearing density (MGller et al., 1995a). Chickens
of two different breeds were reared at three different densities of 20, 24,
and 28 chickens per square meter (the latter being the normal density in
commercial chicken farms), and their level of skeletal asymmetry was as-
sessed when slaughtered. There was a considerable increase of on average
30% in the degree of asymmetry across this range of densities (Fig. 2 ) . The
next step in this line of research would obviously be to extend the relation-
ship to even lower densities until a minimum has been found for the curve
relating asymmetry to density. Of course, this approach is not restricted to
density alone, and other conditions of rearing can be investigated as well.
In a second study, Mdler et al. (1998) investigated the relationship between
light regime and skeletal asymmetry in chickens. Commercial chicken farm-
ers expose their chickens to continuous light, but we managed to manipulate
the light to some extent: (1) continuous light, ( 2 ) a changing light regime,
(3) a 16 :8 hours light cycle with the exception of the first and last days of
life of the chickens; the lack of a strict difference in light regime was caused
by constraints imposed by the chicken farmers (MGller et al., 1998). Chickens
20 24 28
Density (inds. per square meter)
FIG. 2. Skeletal fluctuating asymmetry in relation to rearing density of chickens of the
breeds ScanBrid and Ross 208. Values are means (SE). Adapted from Moller et al. (1995).
reared in continuous light had an average 40% larger asymmetry than

chickens reared in the two other treatments, again suggesting that continu-
ous light imposed significant amounts of stress on the chickens.
A final example concerns cow asymmetry and milk production (J. T.
Manning, pers. comm.). Cows were measured on a number of skeletal traits
for asymmetry, and both the single asymmetry values and a composite
index of overall asymmetry were found to be negatively related to milk
production and the quality of milk as determined by the dairy companies.
This result differs to some extent from the studies of chickens because the
conditions that will improve animal welfare (a reduction in the stress factors
causing fluctuating asymmetry), will also improve productivity. Hence, in
this case the interests of people studying animal welfare and farmers are con-
gruent.
In conclusion, problems of animal welfare arise from the fact that we
cannot make the theoretical inferences about the optimal conditions under
which animals are reared. This problem can be resolved if measures of
fluctuating asymmetry are used as a way of determining the conditions
under which morphological asymmetry reaches a minimum level. These
conditions will reflect the environment in which a specific animal is reared
with a minimum amount of stress.
D. HUMAN
AND VETERINARY
MEDICINE
The practice of medicine has for a long time been isolated from evolution-
ary theory, and several evolutionary biologists believe that progress has
been prevented by this lack of scientific knowledge about the interactions
between pathogens and human hosts. Darwinian medicine based on evolu-
tionary theory may be a way of resolving these problems (Nesse and Wil-
liams, 1995). Similar arguments can be raised for veterinary medicine.
Since measures of developmental instability provide information about
the developmental state of individuals, this information may be useful for
understanding interactions between parasites and their human hosts, but
also for making inferences about the current health status of individuals
(Thornhill and Mgller, 1997). A number of examples of this approach are
listed in Thornhill and Mgller (1997) and Mgller (1996~).Here I will provide
only a couple of examples.
A number of different kinds of cancer associated with the use of contra-
ceptives have increased dramatically in frequency during recent decades,
and they are presumably more abundant as a consequence of dramatic
hormonal fluctuations experienced by regularly menstruating women (Ea-
ton et al., 1994). Breast cancer is one of these now common kinds of
reproductive cancers of women, and large sums of money have been used
to identify predictors of susceptibility to cancer. Preliminary studies showed

that breast asymmetry was a relatively important predictor of low lifetime
fecundity and hence potentially of exposure to repeated surges of estrogen
and other kinds of reproductive hormones (Moller et al., 199%; Manning
et al., 1996). A large data set of mammograms from a hospital was used to
determine whether breast asymmetry was a reliable predictor of breast
cancer (Scutt et al., 1997). Previous work had indicated that a number of
factors associated with increased exposure to estrogens such as high body
mass, large breasts, and repeated menstrual cycles were correlated with
elevated risks of acquiring cancer. However, breast asymmetry proved to
be an even better predictor of breast cancer than any of the previous
variables (Scutt et al., 1997). This has important implications for prevention
and treatment of breast cancer. It is possible that a number of cases of
cancer can be prevented simply by asking women with elevated breast
asymmetry to report for more regular checks than other women. Another
possibility is to attempt to reduce exposure to estrogens and thereby reduce
the risks of breast cancer. Cancers resemble to some extent developmental
instability in the sense that both phenomena are the result of uncontrolled
growth processes. If the lack of control of the two types of growth processes
has a similar cause, measures of developmental instability may be useful
for predicting other kinds of cancer.
A large number of studies have investigated the relationship between
developmental instability and parasitism in a wide variety of plants and
animals (review in Moller, 1996~).Asymmetric individuals throughout this
range of organisms usually have higher parasite loads than do symmetric
individuals, although the causal relationship between symmetry and parasit-
ism is not known in most of these cases. Parasitism was the cause of
increased asymmetry in studies of elms Ulmus glabra, fruitflies Drosophila
nigrospiracula, barn swallows, and reindeer Rangifer tarandus (Moller,
1992a, 1997b; Polak, 1993, 1997; Folstad et al., 1996). Studies of elms,
fruitflies D. nigrospiracula, domestic flies, barn swallows, and humans have
found increased susceptibility of asymmetric individuals (Fig. 3; Moller,
1992a, 1996d, 1997b; Polak, 1993, 1996; Shapiro, 1992). This might have
important implications for prevention of disease, but also for treatment, and
selection for disease resistance in domesticated organisms. If the reduced
resistance of certain individuals is caused by a low level of stress resistance
in general, and thereby loss of energy that could otherwise be used for
coping with parasite attacks, then selection for increased stress resistance
might be an important management tool.
In conclusion, a number of areas in human and veterinary medicine are
likely to benefit from the use of information on developmental instability
Uninfected Infected
Infection status
FIG. 3. Wing and tibia asymmetry in male domestic flies Musca domestica that acquired
or did not acquire a fungus infection after exposure to fungal spores. Values are means (SE).
Adapted from Moiler (1W6d).
at the level of individuals. Future research will decide the extent to which
this approach will be of general use.
E. BEHAVIORAL
MEASURES
OF DEVELOPMENTAL
INSTABILITY
In this section on behavioral measures of developmental instability I
adopt two approaches. First, I briefly review behavior as affected by deviant
morphology. Second, I suggest that phenodeviant behavior also can be
considered a measure of developmental instability.
1. Behavioral Invariance
The approach adopted for morphological developmental instability is to
consider deviations from the morphological invariant to reflect instability.
In a similar way, behavior also goes through an ontogenetic phase after
which a behavioral phenotype is developed. If this phase of ontogeny is
disturbed by mutations, an inability to learn, or by exposure to deviant
models from which a behavioral pattern can be learned, then this will
result in a deviant behavioral phenotype. The main problem for behavioral
studies, as for morphological ones, is to identify the invariance against
which phenotypes can be compared. For morphology this is bilateral or
radial symmetry against which deviations can be determined. Irregular
behavior can be considered phenodeviant with respect to modal behavioral
patterns. An explicit model for this approach was developed by Markow

and Gottesman (1993) for schizophrenia, a mental disease with a presumed
polygenic threshold inheritance. Many behavioral patterns have a quantita-
tive genetic inheritance with heterozygotes displaying modal phenotypes
and homozygotes exhibiting extreme behavior. High levels of homozygosity
were presumed in the model to result in poor developmental stability, which
would be reflected not only in elevated asymmetry of the morphological
phenotype, but also in deviant behavior and phenodeviance with respect
to the anatomy and biochemistry of the central nervous system. Some
support for this model was obtained for schizophrenia, but also for abnormal
behavior in Drosophila.
Many behavioral patterns are incredibly repetitive, particularly when it
comes to display and signals. Typical examples of such repetitiveness is the
same song being sung over and over again by a bird during an entire
breeding season and the same courtship display being performed repeatedly
by a male during a reproductive season. The function of this repetitiveness
could be that it reveals behavioral phenodeviance; only individuals that
are able to repeatedly perform the same behavior over and over again can
be considered to possess a developmentally stable phenotype.
A measure of the ability to perform such repetitiveness can be considered
a behavioral invariant (MGller and Swaddle, 1997). One measure of the
ability to perform a behavior in a similar way is the repeatability of the
activity. Repeatability is a measure from quantitative genetics of the vari-
ance among individuals in relation to a measure of the variance within
individuals (Falconer, 1989). Since the repeatability is a measure of the
genotypic variance and the general environmental variance relative to the
entire phenotypic variance, it also has the property of being an upper
estimate of the heritability of a trait (Falconer, 1989). Repeatabilities vary
from 0 to 1 with low values reflecting extreme variation within as compared
to among individuals, and values of 1 reflecting a complete ability to repeat
the same behavior again and again. Repeatabilities have been used in
studies of behavior for some time in various contexts (Boake, 1989), but
never as a measure of developmental instability. A high value of repeatabil-
ity can be considered the developmentally stable state of a behavioral
parameter, while lower values are unstable states. The use of repeatability
of behavioral traits as a measure of developmental instability also has the
advantage of different traits being directly comparable because of the range
of repeatabilities from 0 to 1.
I would predict that repeatability would be particularly high for behav-
ioral traits that are closely related to fitness, while less important behavior
will have lower repeatabilities. For example, alarm calls will undoubtedly
have high repeatabilities, while contact calls will have lower repeatabilities.
This argument parallels that for morphological developmental instability,

which appears to be inversely related to the functional importance of a
morphological character. An important exception to this prediction will be
behavioral traits that have been subject to intense directional selection.
These are predicted to be developmentally relatively unstable (see Section
111). Of course, it is important when estimating repeatabilities, for example,
in the behavior of an individual on different days, that comparisons among
individuals are made while controlling for the time scale of sampling, the
reproductive state of the individual, and other potentially confounding
factors. The kinds of behavioral traits that can be evaluated with respect
to repeatability are limited only by the imagination of the scientist; examples
include duration, interval, frequency, volume, and complexity.
A possible behavioral example of this approach, although not couched
in terms of repeatability per se, is a study of drift in the song of the great
tit, Parus major (Lambrecht and Dhondt, 1988a,b). Great tit males repeat
a song many times during a song bout, but if the frequency of the song
(measured in kHz) is estimated, males differ considerably in their ability
to maintain a consistently high frequency of their songs. The reduction
during a song bout in the frequency of the song, which is termed drift, can
be considered an indirect measure of repeatability. Interestingly, male great
tits that are able to sing without experiencing considerable drift have higher
reproductive success than other males. Obviously, the underlying mecha-
nism that generates differences in drift may well be morphological asymme-
tries that result in muscle fatigue being reached at an early stage.
Another suggestion for a behavioral invariance is the fractal dimension
of behavioral sequences (Esc6s et al., 1995). Fractals reflect a measure of
the scaling constant that describes the relationship between size of a charac-
ter and the scale of measurement (Hastings and Sugihara, 1993). For exam-
ple, the length of a coastline depends on the ruler that is used for measuring
the coast; as the ruler decreases in size, the length of the coastline increases,
although the pattern of the coastline is always the same. Anybody can
convince themselves of this fact by taking a map of an arbitrary archipelago
like the British Isles. Start out by placing a transparent grid with side length
of say 10 cm (this choice is not really important) on the map and count
the number of squares that include the coastline. Then redo this using
increasingly smaller squares with a side length of 5 cm, 2.5 cm, 1.25 cm,
and .625 cm. The regression coefficient relating the number of squares
necessary to cover the entire coastline to the size of the squares is an
unbiased estimate of the fractal dimension. Esc6s et al. (1995) used this
approach t o determine the fractal dimension of foraging and scanning
bouts in Spanish ibex, Capra pyrenaica. As is typical for such sequences
of behavior, periods of foraging are interspersed randomly with periods of
204 ANDERS PAPE MBLLER
scanning for predators or competitors. Any bout of foraging and scanning

measured at a particular time scale resembles a bout at another time. This
constancy is described by the fractal dimension, which was recorded to be
1.15 in the present case. Two groups of Spanish ibex were compared:
Individuals affected by scabies (an ectoparasitic disease) and healthy indi-
viduals. Disease status changed the fractal dimension from 1.15 to .94. The
difference in fractal dimension between the modal phenotype and the
alternative phenotype can be considered a measure of developmental insta-
bility. This conclusion obviously depends on the certainty with which the
baseline fractal dimension has been determined. While symmetry a priori
can be considered the optimal phenotype, it is not straightforward to make
the same claim for the fractal dimension of a particular behavioral pattern.
In fact, the baseline fractal dimension will depend enormously on the spatial
and temporal homogeneity of the sample obtained.
2. Behavioral and Morphological Developmental Instability
All behavior has at least two morphological bases; the morphology that
is used directly in the production of the behavior and the neural tissue
used for performing the behavior. Developmental instability in both these
morphological bases may have important consequences for the regularity
of the behavioral output. Even small morphological differences in any of
these morphological bases may result in large behavioral differences. Just
a small asymmetry in the size of a paired, bilaterally symmetrical muscle may
after considerable repetition of a behavior translate into large differences in
behavioral performance.
The study of the morphological basis of phenodeviant behavior is still
in its infancy. Early studies of morphological asymmetries demonstrated
experimentally that there were direct effects on locomotion. This was the
case for asymmetry of tail and wing feathers of birds (Mflller, 1991; Evans
et al., 1994; Swaddle et al., 1996). These studies demonstrated what one
might predict from aerodynamic theory. Perhaps the question of whether
morphological asymmetries affect other kinds of behavior differs only in
degree.
The most well studied example of behavioral asymmetry is vocalizations.
Vocalizations play an important role in animal communication and there
is ample evidence for considerable intraspecific variation in such calls. The
first study addressing the question of quality of calls and morphological
asymmetry dealt with mating calls of the field cricket, Cryflus campesfris
(Simmons and Ritchie, 1996).Males produce a mating call with an unpaired
structure called the harp, and the level of asymmetry in this morphological
structure translates directly into perceivable differences in calls. Males with
more asymmetrical harps produce calls that are more attractive to females
in phonotaxis experiments. Given the inherent asymmetry of the harp, and

the potential disadvantages in terms of natural selection of such asymmetry,
asymmetry can be maintained only by an oppositely directed selection
force such as sexual selection. A similar coupling between morphological
asymmetry and call characteristics was found in the oilbird, Steatornis
caripensis, although the functional importance of these differences in vocal-
izations was not studied (Suthers, 1994). Although these few studies all
concern calls, there is no a priori reason why similar principles may not
apply equally well to other modes of communication.
The second kind of morphological basis of behavior is neural anatomy.
Nerve cells and brains do not differ quantitatively from any other structure
with respect to their development, and deviations from regular neural
phenotypes will invariably affect the ways and the efficiency with which
various kinds of behavior can be performed. Stresses of various kinds are
predicted to severely affect the regulated development of the nerve system,
and similar kinds of environmental and genetic stresses that affect the
stable development of the ordinary phenotype will also affect the stable
development of the neural system. Deviant neural systems will arise from
insults to the stable development of the phenotype during ontogeny. There
is already evidence for this prediction from the relationship between genetic
deviants and the functioning of the brain in humans affected by chromo-
somal abnormalities and other genetically based diseases (Thornhill and
Mdler, 1997). I would also predict that deviant neural systems will give
rise to behavioral phenodeviance; a prediction that still needs to be rigor-
ously tested.
Recent progress in the study of developmental instability at the neural
level has important implications for these kinds of predictions. Morphologi-
cal asymmetries in the human brain were determined from magnetic emis-
sion scanning (MES) and magnetic resonance imaging (MRI) studies in a
sample of subjects from a university in the United States (Thoma, 1996).
Asymmetries in various components of the brain were subsequently corre-
lated with body asymmetries that were recorded blindly (i.e., without know-
ing the identity of the subjects with respect to the brain scans). There
were relatively strong correlations between brain asymmetry and body
asymmetry, indicating that the same factors responsible for the development
of fluctuating asymmetry in the skeleton also caused asymmetry in brain
structures (Thoma, 1996). This result can be interpreted in a number of
different ways. One possibility is that nerve cells play an important role in
stable morphogenesis of the phenotype, and that this causes covariation in
asymmetry in brains and the skeleton.
A second study investigated this relationship further by testing whether
the performance of the brain as estimated from a so-called “culture free”
version of a standard IQ test was directly related t o body asymmetry (Fur-

low et al., 1997). IQ tests have been severely criticized for not providing
reliable estimates of the intelligence of individual humans. However, a
recent study has demonstrated that there is a statistically significant positive
correlation between IQ and pH in the brain, suggesting that the IQ score
provides an estimate of a trait that has a physiological basis (Rae etaf.,1996).
Again, the data were collected blindly, so there was no prior knowledge of
the IQ of subjects that were assessed for fluctuating asymmetry, and vice
versa. There was a significant negative relationship between IQ and body
asymmetry, suggesting that asymmetry on the outside of people was directly
correlated with asymmetry on the inside. This result is based on a correlation
and the causation can go either way; or, perhaps most parsimoniously, the
stable development of the phenotype may be controlled by growth processes
controlled by neural signals. Any deviant neural anatomy may subsequently
give rise to a deviant body morphology and as well a deviant neural system.
A number of potentially confounding factors that are known from previous
studies to be correlated with IQ were controlled statistically, but did not
affect the relationship. These results are potentially of general biological
interest and should be investigated experimentally in other organisms. The
potential for a better understanding of the functioning of the brain and the
neural underpinnings of behavior is certainly present.
In conclusion, behavioral phenodeviance can be estimated when the
invariance of a behavioral pattern has been determined. Such measures of
behavioral invariance are the repeatability of a behavioral pattern and the
fractal dimension of a behavior. Behavior has morphological bases in the
structures that are involved in production of the behavior, but also in the
neural underpinnings of behavior. Developmental instability in behavior
may simply arise from developmental instability in either of these morpho-
logical bases. Several important advances have been made recently in at-
tempts to investigate the relationship between behavior and developmental
instability. There are many possibilities for further developments, as sug-
gested in the following section.
AND PROSPECTS
VI. CONCLUSIONS FOR FUTURE
STUDIES
Measures of developmental instability are fascinating in many different

ways. A particularly important point is that they are easy to obtain and do
not require any especially sophisticated and expensive equipment. The
generality of the concept also implies that there are no obvious limits to
the applications. Any aspect of biological phenomena can in one way or
another be linked either directly or indirectly to the degree of stability in
the development of the phenotype. Measures of developmental instability

such as fluctuating asymmetry are unique in the sense that the optimal
phenotype is known a priori; it is the symmetric phenotype. This makes
developmental instability different from any other phenotypic character.
Finally, measures of developmental instability, because of their strong cor-
relations with fitness components such as growth, mating success, fecundity,
and survival, provide the most readily accessible measure of fitness that
we usually can achieve under natural conditions for free-living organisms.
Behavioral biologists could contribute t o further development of the
developmental instability approach to biology in a number of different
ways. An obvious starting point would be to consider Tinbergen’s (1951)
four classical approaches to the study of ethology: ontogeny, mechanism,
function, and evolution. I will not provide extensive lists of studies that could
be done, but only suggest a few areas that might prove particularly fruitful.
There is a deficiency of studies, particularly experimental ones, investigat-
ing the relationship between morphology and behavior. I have reviewed
three studies considering the effects of morphological asymmetries on sig-
nals in the auditory domain. When measures of developmental invariance
regularly become adopted in studies of behavior, we may start investigating
the relationship between morphology and behavior. This approach will also
open up for studies of the link between behavioral developmental instability
and the consequences of behavior in terms of fitness. Developmental insta-
bility of behavior at different levels that subsequently result in mating
success and reproductive success (from display and handling of mates to
copulation, fertilization, and parental care) can potentially be investigated
as a series of behavioral events that lead to the fitness of the individual.
The previous section dealt with the association between brain asymmetry
and body asymmetry in humans. Future studies of neuroethology may
benefit considerably from considering measures of developmental instabil-
ity of the neural system, but also of the phenotype in general, as a means
of exploring the association between the environment and the way in which
the brain functions.
Learning is a central theme in ethology, and the mechanisms resulting
in the learning of particular tasks, and the variability among individuals in
this ability, are of great theoretical and practical importance. Learning
inability among humans has in several studies been associated with develop-
mental instability (review in Thornhill and Mflller, 1997). It is possible that
fluctuating asymmetry may provide a phenotypic marker in future studies
of learning. Recent studies of humans have also indicated that various
measures of psychological stability and mood are directly associated with
body fluctuating asymmetry (Shackelford and Larsen, 1997). It remains to
be determined whether similar relationships exist among other animals.
208 ANDERS PAPE MBLLER
VII. SUMMARY
Developmental stability reflects the ability of organisms to buffer their

developmental trajectories against disturbance. An inability to fulfill this
predetermined goal can be assessed at the phenotypic level in terms of
fluctuating asymmetry, the frequency of phenodeviants, o r other measures
of developmental instability. Such random deviations from perfect asymme-
try are related to a number of different kinds of deviant environmental
and genetic factors. Organisms are generally believed to be adapted to
the most commonly encountered environmental conditions, and deviations
from such optimal conditions result in energy being spent on maintenance.
The control of growth processes is energetically costly, and since the total
amount of energy available has to be allocated to either maintenance,
growth, storage, or reproduction, a larger fraction of the total energy budget
allocated to maintenance reduces the amount available for developmental
control. A range of environmental stresses such as food deficiency, pollut-
ants, parasites, and deviant temperatures gives rise to elevated develop-
mental instability. Similarly, a range of genetic factors such as mutations,
inbreeding, and hybridization increases developmental instability. In other
words, measures of developmental instability provide an integrated estimate
of the quality of the environment of an individual with a given genetic
background, as experienced by the individual itself. Developmental instabil-
ity may be particularly useful for studies of environmental monitoring,
conservation biology, animal welfare, human and veterinary medicine, and
behavioral studies, as shown by a large number of examples.
Acknowledgments
I am grateful for constructive criticism provided by J. T. Manning, M. Milinski. and

P. J. B. Slater. This paper was written while I was supported by the Danish Natural
Science Research Council.
References
Alekseeva, T. A,, Zinichev, V. V., and 2 0 t h . A. I. (1992). Energy criteria of reliability and
stability of development. Acta Zool. Fenn. 191, 159-165.
Andersson, M. (1994). “Sexual Selection.” Princeton University Press, Princeton, NJ.
Bagchi, S., and Iyama, S. (1983). Radiation induced developmental instability in Arabidopsis
tlialiana. Theor. Appl. Genet. 65, 85-92.
Balmford, A., Jones, I. L., and Thomas, A. L. R. (1993). On avian asymmetry: Evidence of
natural selection for symmetrical tails and wings in birds. Proc. R. Soc. London, Ser. B
252,245-251.
Beardmore, J. A. (1960). Developmental stability in constant and fluctuating temperatures.

Heredity 14, 41 1-422.
Boake. C. R. B. (1989). Repeatability: Its role in evolutionary studies of mating behavior.
Evol. Ecol. 3, 173-182.
Broom, D. M., and Johnson, K. G. (1993). “Stress and Animal Welfare.” Chapman &
Hall, London.
Carmeliet, P., Ferrteira, V., Breier, G., Pollefeyt, S., Kieckens, L.. Gertsenstein, M., Fahrig,
M.. Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L.,
Collen, D., Risau, W.. and Nagy, A. (1996). Abnormal blood vessel development and
lethality in embryos lacking single VEGT allele. Nature (London) 380,435-439.
Clarke, G. M. (1995). Relationships between developmental stability and fitness: Application
for conservation biology. Conserv. Biol. 9, 18-24.
Clarke, G. M., and McKenzie, J. A. (1992). Fluctuatingasymmetry as a quality control indicator
for insect mass rearing projects. J. Econ. Entomol. 85, 2045-2050.
Darwin, C. (1868). “The Variation of Animals and Plants under Domestication.” John Mur-
ray. London.
Darwin, C. (1871). “The Descent of Man, and Selection in Relation to Sex.” John Murray,
London.
Davis, A. P., Witte. D. P.. Hsieh-Li, H., Potter, S. S.. and Capecchi, M. R. (1995). Absence
of radius and ulna in mice lacking hoxa-11 and hoxd-21. Nature (London) 375,791-795.
Eaton, S. B., Pike, M. C . , Short, R. V.. Lee, N. C.. Trussell, J.. Hatcher, R. A,, Wood. J. S.,
Worthman. C. M., Blurton Jones. N. G.. Konner. M. J., Hill, K. R., Bailey, R., and
Hurtado. A. M. (1994). Women’s reproductive cancers in evolutionary perspective. Q.
Rev. Biol. 69, 353-367.
Ellegren, H., Lindgren. G., Primmer, C. R., and Mgiller, A. P. (1997). Fitness loss and germ
line mutations in barn swallows breeding in Chernobyl. Nafure (London) 389, 593-596.
Escos. J. M., Alados, C. L., and Emlen, J. M. (1995). Fractal structures and fractal functions
as disease indicators. Oikos 74, 310-314.
Evans, M. R., Martins, T. L. F., and Haley. M. (1994). The asymmetrical cost of tail elongation
in red-billed streamertails. Proc. R. Soc. London, Ser. B 256, 97-103.
Falconer, D. S. (1989). “Introduction to Quantitative Genetics,” 3rd ed. Longman, New York.
Folstad, I., Arneberg, P., and Karter, A. J. (1996). Antlers and parasites. Oecologia 105,
556-558.
Freeman, D. C.. Graham, J. H., and Emlen, J. M. (1993). Developmental stability in plants:
Symmetries. stress and epigenesis. Generica 89, 97-1 1Y.
Furlow, F. B.. Armijo-Prewitt, T.. Gangestad. S. W., and Thornhill, R. (1997). Fluctuating
asymmetry and psychometric intelligence. Proc. R. Soc. London, Ser. B 264, 823-829.
Gest, T. R.,Siegel, M. I., and Anistranski, J. (1983). Increased fluctuating asymmetry in the
long bones of neonatal rats stressed in cold, heat, and noise. Am. J. Phys. Anthropol.
60, 196-197.
Gest, T. R.. Siegel, M. I., and Anistranski, J. (1986). The long bones of neonatal rats stressed
by cold, heat, and noise exhibit increased fluctuating asymmetry. Growth 50, 385-389.
Goldschmidt, R. B. (1940). “The Material Basis of Evolution.” Yale University Press, New
Haven, CT.
Goldschmidt, R. B. (1 955). “Theoretical Genetics.” Cambridge University Press, Berkeley.
CA.
Graham, J. H., Freeman, D. C.. and Emlen, J. M. (1993). Developmental stability: A sensitive
indicator of populations under stress. ASTM Spec. Tech. Publ. STP 1179, 136-158.
Graham, J. H.. Roe, K. E., and West, T. B. (1993~).Effects of lead and benzene on the
developmental stability of Drosophilu melunogaster. Ecotoxicology 2, 185-195.
Hastings, H. M.. and Sugihara, G. (1993). “Fractals: A User’s Guide for the Natural Sciences.”
Oxford University Press. Oxford.
Hoffmann. A. A., and Parsons, P. A. (1989). “Evolutionary Genetics and Evironmental
Houle. D. (1997). Comment on ‘A meta-analysis of the heritability of developmental stability’
by Msller and Thornhill. J. Evol. Biof. 10, 17-20.
Kieser, J. A. (1992). Fluctuating odontometric asymmetry and maternal alcohol consumption.
Ann. Hiim. Biol. 19, 513-520.
Kozlov, M. V., Wilsey, B. J., Koricheva, J.. and Haukioja. E. (1996). Fluctuating asymmetry
of birch leaves increases under pollution impact. J. Appl. Ecol. 33, 1489-1495.
Lamhrecht. M., and Dhondt, A. A. (1988a). Male quality, reproduction, and survival in the
great tit (Punts major). Behav. Ecol. Sociobiol. 19, 57-64.
Lambrecht, M., and Dhondt, A. A. (1988b). The anti-exhaustion hypothesis: A new hypothesis
to explain song performance and song switching in the great tit. Anim. Behav. 36,327-334.
Leamy, L. (1997). Is developmental stability heritable? J. Evol. B i d . 10, 21-29.
Lerner, 1. M. “Genetic Homeostasis.” Oliver & Boyd, London.
Ludwig, W. (1932). “Das Rechts-Links Problem in Tierreich und beim Menschen.” Springer-
Verlag. Berlin.
Manning, J. T., and Chamberlain, A. T. (1993). Fluctuating asymmetry in gorilla canines: A
sensitive indicator of environmental stress. Proc. R. Soc. London, Ser. B 255, 189-193.
Manning, J. T., and Ockenden, L. (1994). Fluctuating asymmetry in racehorses. Nutiire (Lon-
don) 370, 185-186.
Manning, J. T., Scutt, D., Whitehouse, G. H.. and Leinsteb. S. J. (1996). Breast asymmetry
and phenotypic quality in women. Efhol. Sociobiol. (in press).
Markow, T. A,. and Clarke, G. M. (1997). Meta-analysis of the heritability of developmental
stability: A giant step backward. J . Evol. Biol. 10, 31-37.
Markow, T. A,, and Gottesman, 1. 1. (1993). Behavioral phenodeviance: A Lerneresque
approach. Genetica 89,297-305.
Meffe. G. K., and Carroll. C. R.. eds (1994). “Principles of Conservation Biology.” Sinauer,
Sunderland, MA.
Mitton. J. B. (199.5). Enzyme heterozygosity and developmental stability. Acfa Theriol., Siippl.
3, 33-54.
Mitton. J. B., and Grant. M. C. (1984). Associations among protein heterozygosity, growth
rate, and developmental homeostasis. Annu. Rev. Ecol. Syst. 15, 479-499.
Msller, A. P. (1990). Fluctuating asymmetry in male sexual ornaments may reliably reveal
male quality. Anim. Behav. 40, 1185-1187.
Msller, A. P. (1991). Sexual ornament size and the cost of fluctuating asymmetry. Proc. R.
Soc. London, Ser. B 243, 50-62.
Moller, A . P. (1992a). Parasites differentially increase fluctuating asymmetry in secondary
sexual characters. J. Eva/. B i d . 5, 691-699.
Msller, A. P. (1992b). Patterns of fluctuating asymmetry in weapons: Evidence for reliable
signalling of quality in beetle horns and bird spurs. Proc. R. Soc. London, Ser. B 248,
199-206.
Msller. A. P. (1993a). Morphology and sexual selection in the barn swallow Hirundo rusficu
in Chernobyl. Ukraine. Proc. R. Soc. London. Srr. B 252, 51-57.
Moiler, A. P. (l993b). Patterns of fluctuating asymmetry in sexual ornaments predict female
choice. J. Evol. Biol. 6, 481-491.
Moller. A. P. (1997). Developmental stability and fitness: A review. Am. Nut. 149, 916-932.
M@ller,A. P. (1997b). Developmental selection against developmentally unstable offspring
and sexual selection. 1. Theor. B i d . 185, 415-422.
DEVELOPMENTAL INSTABILITY A N D STRESS 21 1
Moller, A. P. (1996~).Parasitism and developmental instability of hosts: A review. Oikos

77, 189- 196.
Mprller, A. P. (lYy6d). Sexual selection. viability selection, and developmental stability in the
domestic fly Musca domesticn. Evolution (Lawrence, Kans.) 50, 746-752.
Moller, A. P. (1997a). Developmental instability of plants and radiation from Chernobyl.
Oikos (in press).
Moller, A. P. (1997b). Elm Ulmus gfabra leaf asymmetry and Dutch elm disease. Oikos
(submitted for publication).
Moller, A. P., and Hoglund, J. (1991). Patterns of fluctuating asymmetry in avian feather
ornaments: Implications for models of sexual selection. Proc. R. SOC. London, Ser. B
245, 1-5.
Moller. A. P.. and Pomiankowski, A. (1993a). Fluctuating asymmetry and sexual selection.
Genefica 89, 267-279.
M0ller. A. P., and Pomiankowski, A. (1993b). Punctuated equilibria or gradual evolution:
Fluctuating asymmetry and variation in the rate of evolution. J. Theor.Biof. 161,359-367.
Moller, A. P., and Swaddle, J. P. (1997). “Asymmetry, Developmental Stability and Evolution.”
Moller. A. P.. and Thornhill. R. (1997a). A meta-analysis of the heritability of developmental
stability. J . Evol. Biof. 10, 1-16.
Moller. A. P., and Thornhill, R. (1997b). Developmental instability is heritable. J . Evol. Biol.
10,69-76
Moller. A. P.. and Thornhill, R. (1997~).Bilateral symmetry and sexual selection: A meta-
analysis. Am. Nut. (in press).
Moiler. A. P., Sanotra, G. S., and Vestergaard, K. S. (1995a). Developmental stability in relation
to population density and breed of chickens Gulfus galfus. Poult. Sci. 74, 1761-1771.
Mdler, A. P., Soler, M.. and Thornhill, R. (199%). Breast asymmetry, sexual selection and
human reproductive success. Erhof. Sociobiol. 16, 207-219.
Moller, A. P., Cuervo, J. J., Soler, J. J., and Zamora-Muiioz, C . (1996). Horn asymmetry and
fitness in gemsbok Oryx g. gazeffa. Behav. Ecol. 7,247-253.
Moller. A. P., Sanotra, G. S., and Vestergaard, K. S. (1998). Developmental instability and
light regime in chickens. Appl. Anim. Behav. Sci. (in press).
Mulvey. M., Keller, G. P.. and Meffe, G. K. (1994). Single- and multiple-locus genotypes and
life-history responses of Gamhusia holbrooki at two temperatures. Evolution (Lawrence,
Knns.) 48, 1810-1819.
Nesse, R. M., and Williams, G. C. (1995). “Evolution and Healing.” Weidenfeld & Nicol-
son, London.
Nilsson, J.-A. (1994). Energetic stress and the degree of fluctuating asymmetry: Implications
for a long-lasting, honest signal. Evol. Ecol. 8, 248-255.
Ozernyuk. N. D., Dyomin, V. I., Prokofyev, E. A,, and Androsova, I. M. (1992). Energy
homeostasis and developmental stability. A c f a Zoof. Fenn. 191, 167-175.
Palmer, A. R. (1994). Fluctuating asymmetry analyses: A primer. In “Developmental Instabil-
ity: Its Origins and Evolutionary Implications” (T. A. Markow, ed.), pp. 335-364. Kluwer,
Dordrecht. The Netherlands.
Palmer, A. R., and Strobeck, C. (1986). Fluctuating asymmetry: Measurement, analysis, pat-
terns. Annri. Rev. Ecol. Sysr. 17, 391-421.
Palmer, A. R., and Strobeck, C. (1997). Fluctuating asymmetry and developmental stability:
Heritability of observable variation vs. heritability of inferred cause. J. Evol. Biol. 10,
39-49.
Parker, H. R., Philipp. D. P., and Whitt, G. S. (1985). Gene regulatory divergence among
species estimated by altered developmental patterns in interspecific hybrids. Mol. Biol.
Evol. 2, 217-250.
Parsons, P. A. (1990). Fluctuating asymmetry: An epigenetic measure of stress. Biol. Rev.

Parsons, P. A. (1991). Can atmospheric pollution be monitored from the longevity of stress
sensitive behavioural mutants in Drosophila? Funct. Ecol. 5, 713-715.
Parsons, P. A. (1994). The energetic cost of stress: Can biodiversity be preserved? Biodiv.
Lett. 2, 11-15.
Polak, M. (1993). Parasites increase fluctuating asymmetry of male Drosophila nigrospiracu/at
Implications for sexual selection. Genetica 89, 255-265.
Polak, M. (1997). Ectoparasitism in mothers causes higher fluctuating asymmetry in their
sons: Implications for sexual selection. Am. Naf. 149, 955-974.
Pomiankowski, A. (1997). Genetic variation in fluctuating asymmetry. J. Evol. Biol. 10,51-55.
Poulsen, M. G. (1997). Fluctuating asymmetry and conservation status of butterflies. MSc
thesis, Univ. of Copenhagen, Denmark.
Rae, C . , Scott, R. B., Thompson, C. H., Kemp, G. K., Dumughin, I., Styles. P., Tracey. I.,
and Radda. G. K. (1996). Is pH a biochemical marker of IQ? Proc. R. Soc. London, Ser.
B 263, 1061-1064.
Rasmuson, M. (1960). Frequency of morphological deviations as a criterion of developmental
stability. Heredifas 46, 51 1-536.
Rettig, J. E., Fuller, R. C., Corbett, A. L., and Getty, T. (1997). Indicates levels of competition
in an even-aged poplar clone. Oikos 80, 123-127.
Sarre, S., Dearn, J. M., and Georges, A. (1994). The application of fluctuating asymmetry in
the monitoring of animal populations. Pac. Conserv. B i d . 1, 118-122.
Scutt, D.. Manning, J. T.. Whitehouse, G. H., Leinster, S. J., and Massey, C. P. (1997). The
relationship between breast asymmetry, breast size and the occurrence of breast cancer.
Br. J. Radio/. (in press).
Shackelford, T. K., and Larsen, R. J. (1997). Facial asymmetry indicates psychological, emo-
tional, and physiological illness. J. Personal. Sue. Psycho/. (in press).
Shapiro, B. L. (1992). Development of human autosomal aneuploid phenotypes (with an
emphasis on Down syndrome). Acta Zool. Fenn. 191, 97-105.
Siegel, M. I., and Smookler, H. H. (1973). Fluctuating dental asymmetry and audiogenic stress.
Growth 37, 35-39.
Simmons, L. W., and Ritchie, M. G. (1996). Symmetry in the songs of crickets. Proc. R. Soc.
London, Ser. B 263,305-3 11.
Sumarsono, S. H., Wilson, T. J.. Tymms, M. J.. Venter, D. J.. Corrick. C. M., Kola, R.,
Lahoud, M. H., Papas, T. S., Seth, A,, and Kola, I . (1996). Down’s syndrome-like skeletal
abnormalities in ETS2 transgenic mice. Nature (London) 379, 534-537.
Suthers, R. A. (1994). Variable asymmetry and resonance in the avian vocal tract: A structural
basis for individually distinct vocalizations. J. Comp. Physiol. A 175, 457-466.
Swaddle, J. P. (1997). On the heritability of developmental stability. J. Evol. B i d . 10,57-61.
Swaddle, J. P., Witter, M. S., Cuthill. I. C., Budden. A,, and McCowen, P. (1996). Plumage
condition affects flight performance in starlings: Implications for developmental homeo-
stasis, abrasion and moult. J. Avian Biol. 27, 103-111.
Thoma, R. (1996). Developmental instability. handedness, and brain lateralization: MES and
MRI correlates. MSc Thesis, University of New Mexico, Department of Psychology, Albu-
querque.
Thornhill. R.. and Mgiller, A. P. (1997). Disease and developmental stability. B i d . Rev.
Cambridge Philos. Soc. (in press).
Tinbergen, N. (1951). “The Study of Instinct.” Oxford University Press, Oxford.
Toates, F. (1 995). “Stress: Conceptual and Biological Aspects.” Wiley, Chichester.
Tomkins. J. L., and Simmons, L. W. (199.5). Patterns of fluctuating asymmetry in earwig

forceps: No evidence for reliable signalling. Proc. R. Soc. London, Ser. B 259, 89-96.
Van Valen, L. (1962). A study of fluctuating asymmetry. Evolufion (Lawrence, Kans.) 16,
125-142.
V~llestad,L. A., Hindar, K., and M ~ l l e r .A. P. (1997). Patterns of correlation between
heterozygosity and fluctuating asymmetry. J. Evol. Biol. (submitted for publication).
Whitlock, M. (1996). The heritability of fluctuating asymmetry and the genetic control of
developmental stability. Proc. R. Soc. London, Ser. B 263, 849-854.
Whitlock, M. C., and Fowler, K. (1997). The instability of studies of instability. J. Evol. Biol.
10,6347.
Witter. M. S., and Lee, S. J. (199.5).Habitat structure, stress and plumage development. Proc.
R. Soc. London, Ser. B 261,303-308.
Zakharov, V. M.. and Yablokov, A. V. (1990). Skull asymmetry in the Baltic grey seal: Effects
of environmental pollution. Amhio 19, 266-269.
Zakharov, V. M., Olsson. M., Yablokov, A. V.. and Esipenko. A. G . (1989). Does environmen-
tal pollution affect the developmental stability of the Baltic grey seal (Halichoerus gry-
pus)? In “Influence of Human Activities on the Baltic. Proceedings of a Soviet-Swedish
Symposium, Moscow, 14-18 April 1986” (A. V. Yablokov and M. Olsson. eds.). pp.
96-108. Gidrometeorizdat, Leningrad.
Zakharov, V. M.. Pankakoski, E., Sheftel, B. I., Peltonen, A., and Hanski, I. (1991). Develop-
mental stability and population dynamics in the common shrew, Sorex araneus. Am. Nut.
138, 797-810.
ADVANCES IN THE STUDY OF BEHAVIOR. VOL. 27
Stress and Decision Making under the Risk of

Predation: Recent Developments from Behavioral,
Reproductive, and Ecological Perspectives
STEVEN
L. LIMA
DEPARTMENT OF LIFE SCIENCES
INDIANA STATE UNIVERSITY
TERRE HAUTE. INDIANA 47809
I. INTRODUCTION
My objective here is to provide a comprehensive review of recent

empirical and theoretical work on antipredator decision making. The
ways in which predators influence the behavioral decisions made by
their prey is now the subject of a large and growing literature. This
sustained interest in the behavioral aspects of predator-prey interactions
is readily traced to the fact that virtually all animals are subjected to
some form of predation, and many biological and ecological insights can
be gained from an understanding of the ways in which predators influence
their prey’s behavior.
Prey decision making under the risk of predation essentially allows an
animal to manage predator-induced stress. Stress is not a term commonly
associated with the study of antipredator &cision making, but this is largely
a matter of semantics, and one can relate stress to such decision making
in several contexts. If one defines stress as an environmental condition that
diminishes Darwinian fitness through either reproduction or survival (e.g.,
Sibly and Calow, 1989), then few aspects of an environment would lead to
more stress than predators. Note that death due to predation is not the
sort of stress that I consider here: observable predator-induced stress in
animals is (in part) a result of prey decision making itself, such as the
energetic stress caused by choosing to feed less in the presence of predators.
One might thus consider the adaptive management of this sort of predator-
induced stress as a main function of antipredator decision making. Forms
of non-predator-induced stress, such as energetic stress caused by food
shortages, will also influence such decision making. This view of stress is
215
Copyright C IYYK by Academic Press
0065-3454198 $25 00
216 STEVEN L. LIMA
ecological or evolutionary in perspective, and most of the existing literature

deals with stress in this context.
A more classical definition of stress concerns the rapid increase in certain
hormones (e.g., glucocorticoids) in response to some threatening situation
(Weiner, 1992). This hormonal response is considered to be a biological
marker of fear (Boissy, 1995), and there exists a substantial literature on
fear and the physiological (neuroendocrine) stress response (for a review,
see Boissy, 1995). However, relatively few studies on physiological stress
have worked with predators, and research relating such stress to antipreda-
tor behavior is still in its infancy (Bercovitch ef al., 1995; Boissy, 1995).
Further research into these physiological aspects of stress may ultimately
have several important implications for how we view antipredator decision
making, several of which I summarize in a closing section.
Regardless of the topic being addressed, all of the published works in-
cluded in this review share certain characteristics. First, the behaviors/
decisions in question respond in ecological time to changes in some compo-
nent of the risk of predation (sensu Lima and Dill, 1990). That is, this
review concerns plastic behavioral traits that respond to short-term per-
ceived changes in the risk of predation. Thus, I do not consider in detail
those aspects of behavior that respond to predation over evolutionary time
(see Edmunds, 1974; Endler, 1991). Second, figuring prominently in most
studies included herein is the inevitable trade-off between the benefits of
avoiding possible predation and the costs of doing so, in terms of feeding,
survival, or reproduction (i.e., stresses as defined earlier). Third, for reasons
of manageability, I include work published primarily during the last 7-8
years. This covers roughly the time period since the publication of several
relevant reviews that were written in the late 1980s (Dill, 1987; Sih, 1987;
Lima and Dill, 1990). The present review nevertheless encompasses about
twice the number of papers covered by Lima and Dill (1990), whose compre-
hensive coverage extended over almost a 15-year period!
I have strived to provide perspectives on antipredator decision making
that encompass several levels of biological organization. I thus cover the
spectrum from short-term decision making by individuals to the conse-
quences of such decision making for long-term fitness, population dynamics,
and species interactions. Work on short-term decision making under the
risk of predation has a relatively long history of study (Milinski, 1986; Sih,
1987; Lima and Dill, 1990), whereas most of the work on its consequences
has appeared in recent years. My choices for the topics organizing this
review reflect an attempt to provide a representative perspective on the
current state of the field-I hope they succeed. In the hope of synthesizing
available studies as much as possible, I have also classified studies across
several topics to the extent warranted.
PREDATOR-INDUCED STRESS AND BEHAVIOR 217
OF FEEDING
11. BEHAVIOR ANIMALS:
CLASSICAL
MOTIVATIONS
It is appropriate to begin with an examination of recent empirical and

theoretical work on foraging behavior. By “classical,” I refer to studies
motivated directly or historically by optimal foraging theory (Stephens and
Krebs, 1986). Work in this area still forms the main empirical, theoretical,
and philosophical basis for the study of decision making under the risk of
predation. Note that while this section focuses mainly on classical issues,
subsequent sections often deal with the behavior of feeding animals to one
degree or another.
A. ENERGETIC
STRESS RISKTAKING
AND STATE-DEPENDENT
1. Empirical Studies
One of the best ways to demonstrate that animals trade off safety against
feeding is to manipulate their internal (energetic) state (Milinski, 1993).
Such a manipulation is usually accomplished via a period of food depriva-
tion. Provided that riskier behavioral options are also those that allow for
a higher rate of energy intake, then an energetically stressed animal should
accept a relatively high risk of predation while feeding. This idea goes back
to the very earliest of studies on anti predator trade-offs (Milinski and
Heller, 1978; Dill and Fraser, 1984). Work on state-dependent risk taking
began in earnest during the late 1980s, and the pace of research has acceler-
ated in recent years (Table I).
Recent demonstrations of state-dependent risk taking make clear that
such behavior is widespread. Almost without exception, over a wide range
of decision making and taxa (Table I), energetically stressed animals will
accept relatively great risk to obtain food. Most studies manipulated an
animal’s energetic state (hunger), with a few exceptions addressing issues
such as the effects of reproductive or migratory state. Moore’s (1994) study
is particularly interesting in this regard; warblers in a migratory state (and
thus in need of large energetic reserves for long-distance flight) took greater
risks than control birds even though the former had higher energetic re-
serves than the latter. Another unusual result concerns the demonstration
that bumblebee workers accept greater risks for increased food intake when
their colony is experiencing energetic stress (Cartar, 1991). In related work,
Weary et al. (1996) found that slower growing piglets accept a higher risk
of maternal crushing to secure increased milk intake.
2. Theory: The Rise of Stochastic Dynamic Programming
State-dependent decision making under the risk of predation is at the
heart of stochastic dynamic programming (SDP). The introduction of SDP
TABLE I
RECENT
STUDIES
EXAMINING
STATE-DEPENDENT
RISK-TAKING
IN ANIMALS
Animal Statelstress Context and result Source
Invertebrates
Bumble bee (Bombus occidentalis) Energetic Foraging workers are reluctant to flee from predator when Cartar (1991)
colony’s reserves are low
Barnacle (Balanus glandula) Energetic Poorly fed barnacles resume feeding faster following Dill and Gillett (1991)
encounter with predator
t! Stonefly larvae (Paragnerina media) Energetic
DO No apparent effect of hunger on use of space Feltmate and Williams (1989a)
Mayfly larvae (Baetis tricaudatus) Energetic Hungry larvae increase feeding by spending less time in Kohler and McPeek (1989)
refuges
Stonefly larvae (Acroneuria and Energetic No effect of hunger on tendency to enter drift Rader and McArthur (1995)
Paragnetina, 2 spp.)
Whirligig beetle (Dineutes assimilis) Energetic Hungry beetles occupy profitable but risky outer portion Romey (1995)
of group
Whirligig beetle (D. assindis) Energetic Hungry beetles adopt solitary foraging to increase Romey and Rossman (1995)
energetic gain
Backswimmer (Notonecta hoffmnni) Energetic Hungry individuals resume feeding faster following Sih (1992a)
encounter with predator
Dogwhelk (Nucella lapillus) Energetic Hungry individuals move more and spend less time in Vadas et al. (1994)
aerial refuges
Vertebrates
Ground squirrel (Spermophihrs Body mass Individuals with low body mass show reduced vigilance Bachman (1993)
beldingi) following alarm calls
Stickleback (Spinachia spinachia) Energetic Hungry fish choose riskier but more profitable patches Croy and Hughes (1991)
Stickleback (Gasterosteus aculeatus) Energetic Hungry fish choose increasingly safer but less profitable prey Godin (1990)
as they satiate
Stickleback (C. aculeatus) Energetic Hungry fish increase predator inspection. reflecting a greater Godin and Crossman (1994)
need for information (?)
Atlantic salmon (Salmo salar) Energetic Hungry fish resume feeding faster following encounter Gotceitas and Godin (1991)
with predator
Pika (Ochotona collaris) Reproductive Lactating females feed in riskier but more profitable Holmes (1991)
microhabitats
Frog larvae (2 Rana spp.) Energetic Hungry tadpoles increase activity under all levels of risk Horat and Semlitsch (1994)
Stickleback (C. aculeatus) Energetic Hungry fish feed on dense but dangerous portions of prey Jakobsen era/. (1994)
swarms
Willow tit (Parus montanus) Energetic Hungry birds resume feeding faster following encounter Koivula ef 01. (1995)
with predator
Roach (Rutihts rutilus) Energetic Hungry fish occupy profitable but risky periphery of the Krause et al. (1992); Krause
t! group (1993a)
Dark-eyed junco (Junco h y e m a h ) Energetic Hungry birds increase rate of energy intake by reducing Lima (1995)
vigilance
Coho salmon (Oncorhynchus kisurch) Energetic Hungry fish are more willing to attack distant prey following Martel and Dill (1993)
recent exposure to predator
Yellow-rumped warbler (Dendroica Migratory Birds in migratory state resume feeding faster following Moore (1994)
coronata) encounter with predator
Crucian carp (Carassius carassius) Energetic Hungry fish feed in riskier but more profitable microhabitats Petterson and Bronmark (1993)
Porcupine (Erethizon dorsatum) Body mass Individuals with low body mass feed in risky but Sweitzer and Berger (1992)
profitable microhabitats
220 STEVEN L. LIMA
to behavioral ecology was spurred by the need to combine disparate quanti-

ties like predator avoidance and food intake into a common framework
for making predictions about state-dependent behavior (McNamara and
Houston, 1986; Mangel and Clark, 1988). SDP models use the numerical
technique of backward induction to develop an optimal behavioral “pro-
gram” in which optimal behavior is specified for all possible internal states
and environmental contingencies. The intuitive and conceptual appeal of
such behavioral programs, and the relative accessibility of SDP modeling
to biologists (via Mangel and Clark, 1988), have led to much recent interest
in SDP.
Published SDP models cover a wide range of behavioral issues in decision
making under the risk of predation. The more “classically” oriented models
examine issues of state dependence (typically energetic stress) in diet selec-
tion (Godin, 1990; Burrows and Hughes, 1991) or patch use (Newman,
1991; see also Houston et al., 1993; McNamara and Houston, 1994), while
others have explored state-dependent foraging activity/effort (Werner and
Anholt, 1993; Crowley and Hopper, 1994). Rosland and Giske (1994) and
Fiksen and Giske (1995) have developed models of optimal die1 vertical
migration in aquatic animals. Houston and McNamara (1989) have used
SDP to explore the issue of foraging effort in closed versus open experimen-
tal systems. The use of rules of thumb regarding uncertainty about predation
risk has also been addressed with SDP (Bouskila and Blumstein, 1992).
In addition, a series of SDP models incorporating body-mass-dependent
predation in birds addresses issues of optimal body mass (McNamara and
Houston, 1990; Houston and McNamara, 1993; Bednekoff and Houston,
1994; see also Bull et al., 1996, for related work with fish), the decision to
hoard food (Lucas and Walter, 1991), and the temporal patterning of daily
foraging behavior (Bednekoff and Houston, 1994; McNamara et al., 1994).
McNamara and Houston (1992), Houston et al. (1993), and Bednekoff
(1997) have used SDP to explore several issues surrounding the trade-
off between feeding and antipredatory vigilance. SDP models have also
addressed the influence of predation risk on optimal sociality (Szekely et
al., 1991; Paveri-Fontana and Focardi, 1994), parental behavior (Clark and
Ydenberg, 1990a,b), and various aspects of mating behavior (Sargent, 1990;
Crowley et al., 1991; KAlAs et al., 1995; Lucas and Howard, 1995; Lucas et
al., 1996).
This interest in SDP modeling has not yet produced a corresponding
increase in empirical tests of such models. Few SDP models are even
accompanied by much empirical information (but see Godin, 1990; Burrows
and Hughes, 1991; Lucas and Walter, 1991; Rosland and Giske, 1994; Bull
et al., 1996). Furthermore, most studies demonstrating state-dependent anti-
predatory decision making (Table I) do not directly address SDP theory.
The problem here may lie in (1) the sometimes extremely complex nature
of SDP models (e.g.. Burrows and Hughes, 1991; Crowley and Hopper,
1994; Rosland and Giske, 1994; Fiksen and Giske, 199.5; Lucas and Howard,
199.5), which may have outstripped the empiricist’s ability to provide even
qualitative tests of theory, and (2) the fact that qualitative predictions
regarding state-dependent behavior often do not require SDP modeling.
The value of SDP models is nonetheless clear and important, especially
with regard to the link between short-term decision making and life-history
phenomena (Clark, 1994; McNamara et al., 1995). I return to the issue of
testability and the importance of models later in this section.
B. THEp/g RULEFOR OPTIMAL

BEHAVIOR
The p / g rule specifies that an animal can maximize its fitness, or optimally
manage its predator-induced stress, by choosing the behavioral option that
minimizes the rate of mortality ( p ) per unit increase in growth rate (8).
Gilliam (1982; see also Werner and Gilliam, 1984) derived this rule for
animals that experience continuous growth up to some reproductive size,
but it has since been broadened to other animals in the form the p/f rule,
where f represents feeding rate (Gilliam and Fraser, 1988; Gilliam, 1990).
In all of its guises, the p/g rule has undeniable appeal. It has been applied
to the question of patch choice (Gilliam and Fraser, 1987, 1988; Moody et
al., 1996; Sih, 1998), diet choice (Gilliam, 1990), foraging effort (Werner and
Anholt, 1993), avian migration (Lindstrom, 1990), and life-history evolution
(Werner, 1986; Aksnes and Giske, 1990). Furthermore, the p/g rule has
been derived in several contexts (Clark and Levy, 1988; McNamara and
Houston, 1992,1994; Houston ef a/., 1993) and without the dynamical theory
used in its original formulation (Aksnes and Giske, 1990; Leonardsson,
1991; Brown, 1992; Clark, 1994; Clark and Dukas, 1994; Hugie and Dill,
1994; Dukas and Clark, 199.5).
Several recent papers caution that the p/g rule has its limitations (many
of which were noted in Gilliam, 1982). Ludwig and Rowe (1990) and Rowe
and Ludwig (1991) show that time constraints in reaching reproductive size
can negate the simple p/g rule. Clark (1994) adds that the p/g rule implies
an unlikely scenario in which reproductive value does not change over
time. McNamara and Houston (1994; see also Houston et al., 1993) show
further that the p/g rule requires no stochasticity or state dependency in
p or g (or f ) . Most importantly, McNamara and Houston show that the p/
g rule applies only when long-term foraging options are not subject to
change. Such an environment is unlike that in which the p/g rule might be
tested experimentally.
222 STEVEN L. LIMA
Despite these apparent limitations, the p/g rule can perform well even
when some of the above conditions are clearly violated (Werner and An-
holt, 1993; Crowley and Hopper, 1994). This suggests that animals might
actually use some p/g-like rule in their decision making. However, only
Gilliam and Fraser (1987) provide quantitative empirical support for such
a rule. Gotceitas (1990) claims empirical support for the p/g rule, but his
results suggest a more simple alternative explanation (see McNamara and
Houston, 1994) in which the fish studied simply acted t o minimize p . Other
tests (Bowers, 1990; Turner and Mittelbach, 1990) provide only qualitative
support that appears consistent with the general expectations of several
different models. In any case, the p/g rule remains a powerful heuristic
tool in the study of decision making under the risk of predation.
C. FORAGING ENVIRONMENT
I N A PATCHY
Here, I address primarily the relatively abstract theoretical and empirical

aspects of foraging in a patchy environment, typically a laboratory or “math-
ematical” environment. I address the more ecologically motivated studies
of habitat use in a later section. This distinction is not always easily made,
but it is a useful one.
1. Patch Choice
Recent theoretical developments in this area concern the Ideal Free
Distribution (IFD) model of patch choice. This model posits that animals
with perfect (ideal) information are free to choose patches such that they
maximize their fitness, subject to the choices made by other animals. A
common prediction is that the distribution of animals among patches will
eventually stabilize (at the IFD) and match the distribution of food re-
sources among patches (Milinski and Parker, 1991). Predators can certainly
disrupt the IFD, and Moody et al. (1996) provide a much-needed theoretical
perspective on this phenomenon. They show that (1) undermatching of
food resources is a universal expectation when resource-rich patches are
also the riskier patches (e.g., Abrahams and Dill, 1989), and that (2) multiple
stable distributions are possible under some circumstances. Other recent
IFD-based models examine situations in which predators respond to the
distribution of prey, and prey, in turn, respond to the distribution of both
their food resources and predators (Schwinning and Rosenzweig, 1990;
Hugie and Dill, 1994; Sih, 1998; see also van Baalen and Sabelis, 1992).
The overall results indicate that stable distributions across patches of both
predator and prey are possible outcomes in many situations (but see Schwin-
ning and Rosenzweig, 1990). These multi-trophic-level models also make
the counterintuitive predictions that (1) the distribution of predators should
tend to match their prey’s resource distribution (Hugie and Dill, 1994; Sih,
1998), and (2) the prey distribution may not closely match the distribution
of prey resources (Hugie and Dill, 1994; but see Sih, 1998). Empirical tests
of these predictions ought to be feasible, but none has been reported (but
for related empirical studies, see Sih, 1984; Formanowicz and Bobka, 1989).
Empirically, predator-induced deviations from the Ideal Free Distribu-
tion have been used to assess the “energetic equivalence” of predator
avoidance in predation-risk-dependent patch choice (Abrahams and Dill,
1989; Todd and Cowie, 1990; Utne et af., 1993); the ultimate goal here is
to express food intake and predator avoidance in the common currency of
energy (see also Kotler and Blaustein, 1995, for a different perspective on
this matter). Kennedy et al. (1994) criticized such IFD-based studies for
assuming an IFD rather than assessing the possibility of systematic devia-
tions from the IFD. Kennedy et al. also present a non-IFD-based alternative
to assessing the energetic equivalence of predator avoidance, but Moody
et al. (1996) warn that this alternative has no functional basis. Moody et
al. caution further that the entire enterprise of determining such energetic
equivalencies may rest on shaky conceptual ground.
There have been relatively few non-IFD-related developments regarding
patch choice under the risk of predation. Theoretically, Gilliam and Fraser
(1988) extend the pulg rule to patch choice with depleting resources. Houston
et al. (1993) provide a cogent discussion and review of the relationships
among models of optimal patch choice under the risk of predation. Empiri-
cally, there have been several recent demonstrations that patch choice
represents an energy-predation trade-off when dangerous patches are also
energetically profitable (e.g., Gotceitas, 1990; Gotceitas and Colgan,
1990a,b; Brown et al., 1992a,b; Pettersson and Bronmark, 1993; Scrimgeour
and Culp, 1994a; Scrimgeour et al., 1994). These studies complement many
similar studies reviewed in Lima and Dill (1990). Nonacs and Dill (1990)
provide the unique result that a worker ant’s decision to feed in a risky
patch reflects the contribution that its efforts make to colony growth.
2. Time in Patches
Recent theoretical treatments of patch use differ considerably in their
predictions. Newman (1991) indicates that optimal patch residence time
may be influenced little by the risk of predation. In contrast, Brown (1992)
develops several models in which optimal patch residence times are highly
predation-risk dependent. This discrepancy may reflect disparate assump-
tions about whether patches vary in predation risk or energetic quality.
Empirically, there is much evidence that the degree to which small mammals
exploit patches is predation-risk dependent (see Section VII1,A).
224 STEVEN L. LIMA
3. Choice of Foraging Location

Several recent (and somewhat difficult to categorize) papers come under
this general heading, which addresses within-patch decisions about where
to feed. For instance, Jakobsen et al. (1994) found that sticklebacks forage
on denser portions of zooplanktonic swarms only when energetically
stressed or safe from attack; this reflects a trade-off between predator
detection and feeding rate (see also Milinski and Heller, 1978; Godin and
Smith, 1988). Vasquez (1994) found that a small cricetid rodent becomes
a refuge-seeking, central-place consumer of food when feeding under a
threat of predation. Peterson and Skilleter (1994) found that clams shift
feeding location from substrate (risky but profitable option) to water col-
umn (safe but less profitable option) after suffering partial siphon loss to
foraging fish. This shift is consistent with an energy-predation trade-off,
but it is not clear whether clams can assess the risk of (partial) predation
independent of the act itself, or whether they could effectively employ both
foraging options after partial siphon loss (see also Lindsay and Woodin,
1995).
D. DIETSELECTION
Recent work provides much-needed theoretical perspectives on diet se-
lection under the risk of predation. Gilliam (1990) describes a particularly
insightful extension of the p / g rule to the question of diet selection. This
model exhibits quasi-classical behavior (see Stephens and Krebs, 1986) in
which prey-specific predation risks are a determinant of prey ranking. Godin
(1990) developed an SDP model that predicted that profitable but risky
prey (large prey whose consumption interferes with predator detection)
should be consumed preferentially only by energetically stressed animals.
Burrows and Hughes (1991) presented an ambitious SDP model in which
mortality and digestive constraints combine to cause a general contraction
of the diet with increasing risk of predation.
Empirical work on diet selection has been limited. Godin (1990)
provided empirical evidence that diet selection in guppies is predation-
risk and state dependent as predicted (qualitatively) by his SDP model;
further support for this model lies in the observation that fish may
prefer large, profitable items only under low predation risk (Ibrahim
and Huntingford, 1989). Phelan and Baker (1992) suggested that predation-
risk-related travel costs influence diet selection in mice, but their test suf-
fered from a lack of any manipulation of risk. Brown and Morgan (1995)
showed that a squirrel’s apparent preference for certain food types can
be predation-risk dependent, even though one food type may be inherently
preferred over others.
E. TESTA~ILITY
AND THE ROLEOF THEORETICAL
MODELS
It is appropriate at this point to address some important issues regarding
the role of modeling in the study of predator-induced stress and antipredator
decision making, as “classically motivated” work is the most theory-rich
area that I consider in this review. The following discussion, however,
applies generally to subsequent sections.
These issues regarding the role of modeling concern the virtual absence
of quantitative tests of theory. Besides the efforts of Gilliam and Fraser
(1987) and Gotceitas (1990), few attempts at quantitative tests have been
reported. There are probably several reasons for this phenomenon. First,
many simple models are obviously caricatures of reality that demand no
quantitative test. At the opposite extreme, some ambitious SDP models
may outstrip the ability of empiricists to provide even qualitative tests of
predictions. A more fundamental problem concerns our inability to measure
the risk of predation itself (or its various components). Only a few field
studies have much quantitative information on the risk of predation (e.g.,
Watts, 1990;Harfenist and Ydenberg, 1995), and none provides information
that relates an animal’s conceivable behavioral options to particular risks
of predation. This sort of information is critical to making quantitative
behavioral predictions about the adaptive management of predator-
induced stress.
To what extent are we limited by our inability to provide quantitative
tests of theory? Two lines of argument suggest that this limitation is not
too severe. First, qualitative tests of carefully reasoned predictions should
prove enlightening in most situations. Second, even without quantitative
tests, there has been an invaluable interplay between theory and empiricism
in the study of decision making under the risk of predation, and I see no
reason why this will not continue. On the other hand, Brown (1992) argues
that models with rather disparate fitness formulations can yield similar
qualitative predictions. Quantitative tests may ultimately be needed to
determine which fitness formulation is superior.
Given our ongoing inability to provide quantitative tests of theory, model-
ers have little choice but to strive for qualitative predictions that distinguish
among various hypotheses. I personally prefer relatively simple models
with broad heuristic value, but Abrams (1993a) argues that simple models
can also be misleading. In any case, a pluralism of modeling approaches
should continue to provide a strong conceptual basis for further empirical
and theoretical progress.
111. PAITERNS
OF ACTIVITY
“Activity” studies examine the influence of predators on both the level

and the temporal patterning of prey activity. I consider each of these areas
226 STEVEN L. LIMA
in turn. These studies on prey activity provide some of the best documented
behavioral responses by prey to the presence of predators, and form the
foundation for much behaviorally explicit ecological research on predator-
prey interactions (e.g., Werner, 1992; Wooster and Sih, 1995).
OF ACTIVITY
A. LEVELS
I distinguish between two types of activity, movement and refuging. An
animal can in principle vary these two types of activity independently in
response to the risk of predation (Sih and Kats, 1991; Werner and Anholt,
1993). By movement, I refer to things like speed of movement, length of
moves, frequency of movement, and so on. Refuging refers to a situation
in which an animal retreats to a refuge and emerges infrequently and for
only brief periods; a refuge is, for example, a burrow or rock crevice (as
opposed to a safe habitat) in which an animal cannot readily feed, locate
mates, and so on (e.g., Sih et al., 1988). Categorizing a given activity as
either movement or refuging is usually straightforward, although many
studies do not define in detail the behaviors under examination.
A decrease in prey activity following a heightened threat of predation
has been a reasonably well established result for some time (Sih, 1987;
Lima and Dill, 1990; Kolar and Rahel, 1993; Wooster and Sih, 1995), one
that also figures prominently in studies related to physiological (neuroendo-
crine) stress (Boissy, 1995). Work in recent years indicates that such a
response is indeed ubiquitous across diverse taxa (Table 11). Almost all
species studied exhibit decreased movement, increased refuging, or both
(if both types of behavior were examined) in response to an increase in
the risk of predation. Several studies indicate that many aquatic (and even
some terrestrial) animals respond to the chemical evidence of predators as
well as the actual presence of predators (for an extensive review, see Kats
and Dill, 1998).
There were exceptions to the general result of decreased activity with
increasing risk (Table 11). Some cases with no response to predator manipu-
lation may have involved prey large enough to be invulnerable to predators
(e.g., Willman et al., 1994), while in others a nonsignificant effect was in
the typical direction (e.g., Walls, 1995). Houtman and Dill (1994) found a
decrease in movement by marine sculpins only if the background provided
some degree of crypticity. Larval Ambystoma salamanders decreased move-
ment only in the absence of a refuge; otherwise, movement increased in
an effort to reach a refuge (Sih and Kats, 1991). The case of increased
movement in toad larvae in response to an alarm substance (Hews, 1988)
may also represent refuge-seeking behavior. Sih and Krupa (1992) argued
that female water striders take advantage of a predator-induced decrease
TABLE I1
RECENT
STUDIES EXAMINING
CHANCES I N PREY ACTIVITY
I N RESPONSE
TO PREDATOR PRESENCE OR PERCEPTION THEREOF
Change in activity' with
Species Predatof Activityh Risk t Food t Hunger Source
Invertebrates
Aquatic snail (Physella and Alarm substance R Incd NP Alexander and Covich
Planorbella, 2 spp.) (1991a.b)
Crayfish (Pacifastacus leniusciilus) Fish C M Dee - Blake and Hart (1993)
R Inc -
Damselfly larvae (2 Enallagrna Fish P. larval M Dec (1 sp.). - Blois-Heulin el al. (1990)
SPP.) odonate P NR (1 sp.)
Shrimp (Atya lanipes) Large shrimp P M Dee NP Crow1 and Covich (1994)
R Inc NP
Larval mayfly (Paraleptophlebia Fish P R Inc NP Culp et al. (1991)
heteronea)
Grass shrimp (Palaernonetes pugio) Fish P R Inc NP Everett and Ruiz (1993)
Stonefly larvae (Paragnetina Fish P M Dee NP Feltmate er al. (1992)
media)
Crayfish (Orconecfesvirilis) Alarm substance M Dec NP Hazlett (1994)
Crayfish (3 Orconectes spp.) Fish P R Inc - Hill and Lodge (1994)
Isopod (Lirceus fontinah) Fish P M Dec - Holomuzki and Short
(1990)
Amphipod ( Garnrnarus minus) Fish C M Dec NP Holomuzki and Hoyle
(1990)
lsopod ( L . fontinalis) Fish P M Dec - Huang and Sih (1990, 1991)
Damselfly larvae (2 Enallagnia Larval dragonfly P M Dee Inc Jeffries (1990)
SPP.1
Damselfly larvae (Coenagrion Larval dragonfly P M Dec NR Johansson (1993)
hastidaturn)
Mayfly larvae (Baetis fricaudatus) Fish P M Dec NR Kohler and McPeek (1989)
R Inc Dec
(continued)
TABLE I1 (Continued)
Change in activity’ with
Species Predatof Activity6 Risk t Food t Hunger f Source
Chironomid larvae (Chironomus Fish P R Incd NP Macchiusi and Baker (1991)

tentans)
Chironomid larvae (C. tentans) Fish P R Inc Dec Macchiusi and Baker (1992)
Damselfly larvae (4 Enallagma Fish P, larval M Dec - McPeek (1990)
SPP.1 dragonfly P
Worker ants (Lasius pallitarsis) Large ant P M Dec Inc - Nonacs (1990)
Caddis larvae (Rhyacophila nubila) None R - NP Dee Otto (1993)
Water flea (2 Daphnia spp.) Copepod P M Dee (1 sp.), inc - - Ramcharan and Sprules
(1 SP.1 (1991)
Marine snail (Stramonita Crab P R Inc Richardson and Brown
haemastoma) (1992)
Ostracod (Cypridopsis vidua) Fish C R Inc - Roca et al. (1993)
Mayily larvae (Baetis, Ephemerella, Fish P, stonefly P R Inc NP Scrimgeour et al. (1994)
Claassenia spp.)
Isopod (L.fontinalis) Fish C M Dec Short and Holomuzki
(1992)
Water strider (Aquarius remigis) Fish P M Dec (male), Inc Sih and Krupa (1992,1995)
inc (female)
Isopod (Saduria entomon) Larger isopod P R Inc NP Sparrevik and
Leonardsson (1995)
Mayfly larvae (2 Baetis spp.) Fish P M NR Tikkanen et al. (1994)
R Incd
Aquatic snail (Physella gyrina) Alarm substance R Inc - Turner (1996)
Dogwhelk (Nucella lapillus) Crab C, alarm M Dec Inc Vadas et al. (1994)
substance R Inc Dec
Lobster (Homarus americanus) Fish P R Inc Wahle (1992)
Crayfish (3 Orconectes spp.) Fish P R Dec, 1sp., NR, Willman et al. (1994)
2 spp!
Fish
Fathead minnow (Pimephales Alarm substance R Inc Brown et al. (1995)
promelas) M Dec
Fathead minnow (P.promelas) Alarm substance, R Inc Chivers and Smith (1994.
fish P 1995)
Brook stickleback (Culaea Alarm substance, M Dec Gelowitz et al. (1993)
inconstans) fish C
Marine sculpin (Oligocottus Alarm substance M Dec' Houtman and Dill (1994)
maculosus)
Coho salmon (Oncorhynchus Duck C M Dec Martel and Dill (1993)
kisutch)
Coho salmon (0.kisutch) Duck P M Dec Martel and Dill (1995)
Fathead minnow (P.promelas) Alarm substance, M Dec Mathis and Smith (1993a):
fish P R Inc Mathis er al. (1993)
Brook stickleback (C. inconstans) Alarm substance M Dec Mathis and Smith (1993b)
Darter (3 Etheostoma spp.) Fish P M Dec' Radabaugh (1989)
Lumpfish larvae (Cyclopterus Fish P M Dee Williams and Brown (1991)
h)
W
h) lumpus)
Amphibians
Frog larvae (Rana catesbeiana) Larval dragonfly P M - Dee - Anholt and Werner (1995)
Newt larvae (Taricha torosa) Adult newt C R Inc NP - Elliott et al. (1993)
Salamander larvae (Ambystoma Fish P R Inc Figiel and Semlitsch (1990)
maculatum)
Frog larvae (Ascaphus truei) Fish and R Inc Feminella and Hawkins
salamander C (1 994)
Toad (Bufo americanus) Snake P M Dec - - Hayes (1989)
Toad (B. americanus) Snake P M Dec NP Inc Heinen (1994a,b)
Toad larvae (B. americanus) Alarm substance M Inc NP - Hews (1988)
Frog larvae ( 2 Rana spp.) Fish C M Dec Dec or NR Inc Horat and Semlitsch (1994)
R NR NR NR
Salamander larvae (A. texanum) Fish P R Inc - - Huang and Sih (1990,1991)
Frog, toad larvae (Hyla and Bufo, Fish, newt, M Dec - - Lawler (1989)
4 SPP.) dragonfly P
-
(continued)
TABLE I1 (Continued)
Change in activity‘with
Species Predatof’ Activity’ Risk 7 Food 7 Hunger 7 Source
Frog larvae (R. temporarin) Fish, crayfish C M Dec Manteifel (1995)

Salamander larvae ( A . babouri) Fish P R Inc Sih et al. (1988)
Salamander larvae ( A . babotcri) Fish C M Dec Sih and Kats (1991)
Salamander larvae ( A . baboirri) Fish P R Inc Sih et al. (1992)
Toad larvae (B. americanus) Larval dragonfly M Dec Skelly and Werner (1990)
Frog larvae ( H y l a versicolor) Salamander P M Dec Skelly (1992)
Frog larvae (2 Pseudncris spp.) Dragonfly larva P M Decd Skelly (1995)
Frog larvae (2 R a m spp.) Fish P,C M Dec Stauffer and Semlitsch
(1993)
Salamander larvae (2 Ambysroma Large salamander R Inc (1 sp.), NR NP - Walls (1995)
SPP.) P ( 1 SP.)
Frog larvae (R. aurora) Alarm substance M Dec NP - Wilson and Lefcort (1993)
3
Other Vertebrates
Gerbils (2 Gerbilhus spp.) Owl P M Dec Abramsky et al. (1996)
Rat (Ratt~issp.) Cat P M Dec Blanchard and Blanchard
R Inc (1989)
Bank voles (Clethrionornys Mammals C M Dec Jedrzejewska and
glareolus) Jedrzejewski (1990);
Jedrzejewski et al. (1993)
Field vole (Microtus agrestis) Falcon P M Dec Korpimaki et al. (1996)
Desert rodents (4 heteromyids, 1 Owls P M Dec Longland and Price (1991)
cricetid)
Lizard (Lacerta viviparn) Snake C M Dec NP - Van Damme et al. (1990)
P, Predator present; C. chemical scent of predator: “alarm substance” usually refers to a chemical emanating from a killed or injured conspecific.
R, Refuging; M. movement.
NR, No response: Dec and Inc, decrease or increase, respectively, in the activity in question; -, no manipulation; NP, food not present.
‘Response varied according to body size; some size classes may have been invulnerable to predators.
‘May see only under cryptic conditions.
in male activity to pursue their own activities free from male harassment,
hence their atypical response to predator presence.
Surprisingly few studies have examined an animal’s level of activity in
the context of managing stress caused by a lower rate of feeding. In fact,
food (or an identifiable impetus for nonzero activity) was not present in
approximately 40% of the studies in Table 11. Food was present but unmani-
pulated in an additional 40% of studies; presumably, under these circum-
stances, a reduction in activity led to a decreased feeding rate. Studies
manipulating food levels show mainly a decrease in activity with increasing
food availability. Such a decrease is consistent with theoretical expectations
(Abrams, 1991; Werner and Anholt, 1993), provided that risk increases with
activity. The few studies manipulating an animal’s state show a consistent
increase in activity (increased movement, decreased refuging) in energeti-
cally stressed animals. Such a state-dependent response is indicative of a
trade-off between activity and the risk of predation (see Section 11,A).
Underlying any functional explanation for a predator-induced decrease in
activity is the assumption that increased activity raises the risk of predation.
Presumably, increased activity raises the probability of being detected or
encountered by a predator (but see also Houtman and Dill, 1994). This
assumption receives support from several recent studies involving diverse
predator-prey systems (e.g., Vaughn and Fisher, 1988; Daly et al., 1990;
FitzGibbon, 1990; Rahel and Kolar, 1990; Everett and Ruiz, 1993; Otto,
1993; Heinen, 1994a; Anholt and Werner, 1995; Martel and Dill, 1995).
Skelly (1994) provides a particularly nice demonstration of this effect by
comparing predation on active and partially anesthetized tadpoles. Further-
more, a predator-induced increase in activity in Daphnia oregonensis
(Ramcharan and Sprules, 1991) actually led to greater mortality. Interspe-
cific patterns in predation linked to differing levels of activity (Hershey,
1987; Lawler, 1989; Chovanec, 1992; Azevedo-Ramos et al., 1992; Juliano
et al., 1993; Grill and Juliano, 1996) provide further support for this impor-
tant assumption.
B. TEMPORAL
PATTERNS
I N ACTIVITY
1. Die1 Vertical Migration by Zooplankton

Zooplankton undertaking die1 vertical migration (DVM) descend to the
depths during the day, and ascend to the surface at night; cases of reverse
DVM (the opposite activity pattern) are also known (Ohman, 1990). Gliwicz
and Pijanowska (1988) and Lampert (1989) note that, by the mid-l980s,
many studies suggested that DVM is an adaptation against visually feeding
predators rather than one related to the reduction of energetic stress, as
232 STEVEN L. LIMA
once thought. Recent work on DVM collectively provides an unusually

comprehensive view of predator-induced stress and decision making.
The antipredator hypothesis posits that animals engaged in DVM trade
off the energetic benefits of remaining in the warm and food-rich surface
layers against the safety of the dark, but relatively cold and food-poor
deeper water (Lampert, 1989; Fiksen and Giske, 1995). Accordingly, the
addition of a predatory stimulus to experimental test chambers induces or
enhances DVM in many cladocerans (primarily Daphnia spp.; Dodson,
1988; Leibold, 1990; Dawidowicz and Loose, 1992; Dini and Carpenter,
1992; Young and Watt, 1993; Loose and Dawidowicz, 1994), copepods
(Bollens and Frost, 1989b;Neill, 1992),and Chaoborus midges (Dawidowicz
er al., 1990; Leibold, 1990;Tjossem, 1990). Similar effects occur upon whole-
lake additions or removals of planktivorous fish (Dini et al., 1993). These
experimental results have been corroborated by field work showing that
changes in DVM correspond closely to behavioral and distributional
changes in planktivorous fish (Dini and Carpenter, 1988; Bollens and Frost,
1989a, 1991; Dodson, 1990; Levy, 1990a; Ohman, 1990; Ringelberg er al.,
1991; Frost and Bollens, 1992).
The way in which predation risk interacts with nonpredatory factors
(e.g., food abundance, water temperature) to influence DVM is relatively
unexplored territory. However, recent work suggests that DVM can be
enhanced with the addition of food near the water’s surface (Leibold,
1990) or can be diminished with food addition to deeper water (Dini and
Carpenter, 1992); observational evidence also suggests a strong effect of
resource depth distribution on DVM (Gliwicz and Pijanowska, 1988). Fik-
sen and Giske (1995) suggest further that the effects of food abundance
on optimal DVM may be markedly nonlinear and circumstance dependent.
Theory also suggests that factors such as light transmission and water tem-
perature may be important determinants of the optimal depth of DVM
(Aksnes and Giske, 1990; Levy, 1990b; Fiksen and Giske, 1995), but there
appears to be relatively little experimental work in this area.
Gabriel and Thomas (1988) present a game-theoretical model of DVM
suggesting that at evolutionary stability some members of a population
may not engage in DVM. There is no strong evidence for such an effect
(but see Guisande et al., 1991), although clonal (genetic) differences in
DVM are known to occur (De Meester, 1993; De Meester et al., 1995). It
is also known that species or size classes most vulnerable to fish predation
tend to be those whose migratory behavior is most affected by changes in
the predatory regime (Dodson, 1988; Ohman, 1990; Leibold, 1991; Neill,
1992;Watt and Young, 1994;see also Fiksen and Giske, 1995). Conspicuous,
egg-carrying females may also be reluctant to ascend to the surface even
under relatively dark conditions (Bollens and Frost, 1991).
The proximate factors influencingDVM have also been examined. Chem-

icals emitted by predators are sufficient (and perhaps necessary) to induce
DVM in most species studied (see Larsson and Dodson, 1993,for a review).
Some progress has been made in characterizing the chemical(s) that signal
the presence of predators (Parejko and Dodson, 1990; Loose et al., 1993).
Rapidly changing light levels may also induce DVM (Ringelberg, 1991a,b;
see also Clark and Levy, 1988), but zooplankton may initiate migration
well in advance of changing light levels (Young and Watt, 1993).
Studies examining the long-term stress induced by DVM associate slower
growth (Dawidowicz and Loose, 1992; Gliwicz, 1994; Loose and Dawido-
wicz, 1994) and delayed reproduction (Vuorinen, 1987) with descending
into the depths during the day. Loose and Dawidowicz (1994) argue that
these costs of DVM are due mainly to the colder temperatures of deep
water (see also Aksnes and Giske, 1990). Despite these costs, demographic
analyses (Ohman, 1990;Bollens and Frost, 1991) suggest that DVM confers
a net advantage if it results in even a modest lowering of the risk of pre-
dation.
2. Die1 Migration in Fish
Fish may also engage in diel migrations, both vertical and horizontal
(Helfman, 1986; Clark and Levy, 1988; Levy, 1990a,b;Gliwicz and Jachner,
1992). Clark and Levy (1988) outline several hypotheses for such migratory
behavior, which parallel those proposed for zooplankton (see earlier discus-
sion). One of these hypotheses suggests that DVM in planktivorous fish
reflects little more than the DVM of their prey, but this alone cannot
explain DVM in such fish (Clark and Levy, 1988; Levy, 1990b; Rosland
and Giske, 1994). Furthermore, these fish may undergo DVM even in the
absence of DVM in their prey (Gliwicz and Jachner, 1992; Rosland and
Giske, 1994). Theoretical and empirical evidence suggests that DVM in
planktivores reflects in part the risk imposed upon them by piscivores
(Clark and Levy, 1988; Gliwicz and Jachner, 1992; Rosland and Giske,
1994). In any case, there appears to be little definitive experimental work
on diel migration in fish.
3. Nocturnal versus Diurnal Activity
Several recent studies show that animals will switch between nocturnal
and diurnal activity, depending on the activity patterns of predators. Fenn
and Macdonald (1995) showed that normally nocturnal rats may shift to
diurnal activity in response to nocturnal activity by foxes. Such flexibility
in rat activity was anticipated in recent psychological work on the patterning
of rat behavior in response to threatening stimuli (Lester and Fanselow,
1992; Helmstetter and Fanselow, 1993). A literature review by McNeil et
234 STEVEN L. LIMA
al. (1992) suggests further that some birds may shift to nocturnal activity
to avoid a strong diurnal risk of predation. Similarly, tiger moth (Spibsoma
congrua) larvae become more nocturnal after diurnal encounters with wasps
(Stamp and Bowers, 1993). On the more aquatic side of things, Culp and
Scrimgeour (1993) and Cowan and Peckarsky (1994) showed that mayflies
(Baetis spp.) switch from largely aperiodic to nocturnal feeding in the
presence of visually hunting fish. Juvenile crayfish become more nocturnal
in the presence of fish, but more diurnal in the presence of larger (and
nocturnal) adult crayfish (Blake et al., 1994).
4. Die1 Drift Periodicity in Stream Insects
The tendency for large benthic stream insects to enter the nocturnal drift
(to move via the current to a downstream site) has long been interpreted
as an antipredator response, as these insects would be at risk to size-
selective fish predators in the diurnal drift (Allan, 1978). Flecker (1992)
found support for this idea in a comparative study of streams with and
without fish, and suggested that such nocturnal drift periodicity was a
fixed (evolutionary) response to predation (see also Anderson et al., 1986;
Malmqvist, 1988). However, much recent work shows clearly that at least
some stream insects actively decide to enter the nocturnal drift in response
to an increased local risk of predation (Williams, 1990; Poff et al., 1991;
Andersen et al., 1993; Douglas et al., 1994; Forrester, 1994a,b; McIntosh
and Townsend, 1994; Tikkanen et al., 1994). Rader and McArthur (1995)
show further that the tendency of stoneflies to enter the nocturnal drift is
reduced in habitats with abundant refuges.
5. Daily Activity Patterns and Body Mass in Birds
Bednekoff and Houston (1994) and McNamara et al. (1994) argue theo-
retically that patterns in the daily feeding activity of birds should reflect a
trade-off between the costs (reduced speed or maneuverability) and benefits
(reduced energetic stress) of carrying high fat reserves. These models sug-
gest that such a trade-off can produce the bimodal daily feeding pattern
commonly seen in birds (McNamara et al., (1994) even in the absence
of die1 cycles in temperature, food availability, and so on. However, the
behavioral consequences of such trade-offs have received little experimen-
tal attention (but see Witter et al., 1994). Observational evidence neverthe-
less suggests an important role for fat-reserve-related predatory effects in
avian biology (Witter and Cuthill, 1993).
6. Nondiel Temporal Patterns in Activity
a. Activity and the Lunar Cycle. The brighter portion of a lunar cycle
represents a period of elevated risk for animals hunted by predators like
owls. Accordingly, recent studies have demonstrated repeatedly that small,

nocturnal mammals are relatively inactive under bright moonlight. This is
the case in gerbils (Kotler et al., 1991, 1993a,b; Hughes and Ward, 1993;
Hughes et af., 1994), for whom Kotler ei al. (1991) verify an elevated risk
of owl predation under bright conditions. Kotler et al. (1994a) also found
that gerbils reduce activity in anticipation of moonrise, indicating that the
simple avoidance of light is not necessarily the proximate factor controlling
lunar-based activity cycles.
Recent work on heteromyid rodents also shows strong moonlight avoid-
ance (Bowers, 1990; Daly et al., 1992; Bouskila, 1995; see also Lockard and
Owings, 1974; but see Longland and Price, 1991). Daly et al. (1992) found
that heteromyid kangaroo rats compensate for the lack of activity during
periods of full moon by increased crepuscular activity, which actually makes
them more vulnerable to diurnal predators. Work on murid rodents (in
addition to gerbils; Wolfe and Summerlin, 1989; Simonetti, 1989; Dickman,
1992; Vasquez, 1994) and Old World porcupines (Brown and Alkon, 1990)
indicates the same general trends in moonlight avoidance.
The generality of moonlight avoidance in small nocturnal mammals is
clear, but there appears to have been little recent work on nonmammalian
species. However, Gliwicz (1986) and Dodson (1990) suggest that the lunar
cycle can also affect the strength of die1 vertical migration in zooplankton.
b. Activity on Other Time Scales. Nondiel patterns in activity have re-
ceived relatively little attention outside of the context of the lunar cycle.
However, tidal cycles may influence risk taking by refuging barnacles (Dill
and Gillett, 1991) and migrating intertidal-feeding fish (Burrows and Gib-
son, 1995). On a shorter time scale, Speakman er al. (1995) suggest that
temporal clumping in the nightly emergence of bats from maternity colonies
represents an attempt by individuals to dilute the risk of owl predation.
Kalcounis and Brigham (1994) nevertheless found that the presence of a
vocal owl model had no impact on any aspect of bat emergence patterns.
Activity cycles expressed over an entire season have received almost no
attention. Lucas et al. (1996) provide an interesting exception in their
dynamic game analysis of chorusing behavior in male frogs. Their analysis
suggests that an interaction between predation risk, energetic stress, male
density, and female behavior may produce pulses (or waves) of chorusing
activity over the breeding season.
WITH A PREDATOR
AN ENCOUNTER
IV. AFTER
Recent work on postencounter decision making covers a variety of topics,

such as the resumption of activity, the choice of escape behavior, and flight
236 STEVEN L. LIMA
initiation distance. In covering these topics, I focus on behavior that is

flexible with respect to changes in the predatory environment; papers de-
scribing simple evasive behaviors in response to attack are outside the
scope of this review.
A. POSTENCOUNTER
RESUMPTION
OF ACTIVITY
Prey typically reduce activity via reduced movement, increased refuging,

or both, upon an encounter with a predator (Table 11). An animal must,
of course, resume its normal activity at some point. The period of reduced
activity may range from a few seconds in hermit crabs (Scarratt and Godin,
1992) to several days in small mammals (Jedrzejewski and Jedrzejewska,
1990; Kotler, 1992; Saarikko, 1992). However, despite the many activity-
related studies in Table 11, there is relatively little work on the factors
affecting an animal’s decision to resume activity.
One factor influencing the decision to resume activity is the nature of
the predatory threat, with animals remaining inactive for longer periods in
riskier situations (Scarratt and Godin, 1992; Sih, 1992a; Gotceitas and
Godin, 1993; Johansson and Englund, 1995). Several recent studies also
demonstrate that energetically stressed animals resume activity sooner than
those well fed (Dill and Gillett, 1991;Gotceitas and Godin, 1991;Sih, 1992a;
Koivula et al., 1995). Moore (1994) found that birds in a migratory state
(with large energy reserves and a need to acquire even more) were more
eager to resume feeding than nonmigratory birds after exposure to a hawk.
Theoretical studies on the resumption of activity are few. However, Sih
(1992a) provides a good theoretical discussion of the ways in which energetic
stress and information combine to influence the postencounter resumption
of activity in refuging prey. Stochastic dynamic programming could also
be usefully applied to this temporal phenomenon, but apparently only
one such model has been presented (KBlBs et al., 1995, dealing with the
resumption of lekking following a predatory encounter). Johansson and
Englund (1995) present a much-needed (but brief) game-theoretical per-
spective on the resumption of activity, which suggests that prey will gener-
ally outwait all but the most persistent predators.
B. PURSUIT-DETERRENCE
SIGNALS
Upon detecting a predator, an animal may signal that (1) the predator
has been detected, and (2) it is able to escape; such signals should deter
further pursuit. This mutually beneficial form of communication between
prey and predator (Hasson, 1991) should be subject to some form of cost-
benefit analysis on the part of prey (Caro, 1995),but few studies have taken
such a perspective. Car0 (1994) and Car0 et ul. (1995) provide exceptions
in their thoughtful consideration of antipredator signaling in ungulates.
Car0 (1994) argues convincingly that much antipredator behavior in African
ungulates is pursuit-deterrence signaling. Similarly, Car0 et ul. (1995) con-
clude that tail flagging in white-tailed deer (Odocoileus virginiunus) func-
tions as a pursuit-deterrence signal (see also Smith, 1991). The tail-flicking
response of rails (Aves) to various aspects of predation risk also suggests
that such behavior functions as a pursuit-deterrence signal (Alvarez, 1993).
Furthermore, predator inspection behavior has been implicated as a form
of pursuit-deterrence signaling in fish and mammals (see later discussion).
On a theoretical note, Vega-Redondo and Hasson (1993) suggest that
“honest” antipredator signaling can be evolutionarily stable depending on
the processes by which predators and prey encounter each other.
C. FLIGHT
INITIATION
Prey often allow a predator to approach up to a certain point (the flight
initiation distance, FID) before initiating escape behavior. Several recent
studies complement earlier work (see Ydenberg and Dill, 1986; Lima and
Dill, 1990), suggesting that FIDs increase in riskier situations, and are thus
the outcome of a cost-benefit analysis by prey. A good example of such
decision making occurs in woodchucks (Murmotu monax), which increase
FIDs with an increase in the distance to the nearest refuge burrow (Bonen-
fant and Kramer, 1996) and when the predator approaches from the side
opposite such a refuge (Kramer and Bonenfant, 1997);these studies comple-
ment similar work on tree squirrels (Dill and Houtman, 1989). Fish may
increase their FID when far from a refuge (Dill, 1990) or when in smaller
groups (Abrahams, 1995). Bulova (1994) also found a positive relationship
between distance to refuge and FID in two iguanid lizards, and (surpris-
ingly) a tendency toward shorter FIDs when approached directly by a
predator (as opposed to a more tangential approach).
There are still few studies examining nonpredatory influences on deci-
sions regarding flight initiation. However, Scrimgeour and Culp (1994a)
and Scrimgeour et al. (1994) found that FIDs in mayflies were shorter in
patches with a better food supply. Gravid female lizards may have lower
FIDs than nongravid females, perhaps reflecting the former’s relative inabil-
ity to flee from predators (Braiia, 1993).
OF ESCAPE
D. CHOICE BEHAVIOR
Animals generally have several escape options and may perform various
escape maneuvers at differing intensities. Legault and Himmelman (1993)
238 STEVEN L. LIMA
showed that the intensity of evasive behavior in several molluscs and echino-
derms varied positively with the danger posed by an encounter with a
starfish; these results imply a cost to escalated escape behavior, but the
nature of this cost was not clear. Dill etal. (1990) found that alarmed aphids
were less likely to drop off high-quality plants than poor-quality plants,
and suggested that aphid escape behavior is a function of both lost feeding
opportunities (post-escape) and mortality associated with the extreme es-
cape option of dropping off a plant. However, Stadler el al. (1994) found
that aphids drop off plants more readily under better feeding conditions;
this contradiction may be related to reproductive considerations. In related
work, Cartar (1991) found that threatened worker bumblebees were rela-
tively unlikely to initiate escape maneuvers (i.e., cease feeding) when their
colony was under energetic stress. Finally, badgers faced with a dangerous
predatory encounter choose the nearest available burrow for escape; they
may seek a more distant but safer burrow with a lesser threat (Butler and
Roper, 1994).
A N D INSPECTING
E. APPROACHING PREDATORS
There are many possible benefits and costs associated with the odd
behavior of approaching predators, many of which are discussed by Dugat-
kin and Godin (1992a) in a wide-ranging review. Here, I focus my attention
on the phenomenon of “predator inspection” by fish, which has received
much attention in recent years.
Predator inspection by fish usually involves one or more fish breaking
away from a larger group to approach a predator (Dugatkin and Godin,
1992a). Such inspections may serve to gain information about the type
of predator encountered (Magurran and Girling, 1986) or the predator’s
readiness to attack (Licht, 1989). Dugatkin (1992) demonstrates a mortality
cost to such behavior (but see Godin and Davis, 1995), and evidence sug-
gests that inspectors assess these costs when approaching a predator. For
instance, inspectors approach more closely when in larger groups, avoid a
moving predator, and approach preferentially the tail end of the predator
(Pitcher el al., 1986; Magurran and Seghers, 1990a; Dugatkin and Godin,
1992b). Larger fish, with presumably better escape abilities, may inspect
more closely (Kiilling and Milinski, 1992) than smaller individuals. Energeti-
cally stressed fish may also inspect more than others, presumably because
such fish must feed and thus have a greater need for information on preda-
tion risk (Godin and Crossman, 1994; McLeod and Huntingford, 1994).
Predator inspection may also serve as a form of pursuit-deterrence signaling
(Magurran, 1990; Godin and Davis, 1995; see also FitzGibbon, 1994, for a
possible mammalian example).
PREDATOR-INDUCED STRESS A N D BEHAVIOR 239
A great deal of controversy surrounds the related claims that (1) pairs
of inspecting fish are caught in the “prisoner’s dilemma,” and (2) such fish
engage in a strategy of conditional cooperation resembling the tit-for-tat
(TFT) strategy of Axelrod and Hamilton (1981). Evidence in favor of TFT
cooperation suggests that inspecting fish exhibit the sort of reciprocation,
retaliation, and forgivingness that one might expect in a TFT-like strategy
(Milinski, 1987; Dugatkin, 1988; Milinski ef al., 1990a,b; Dugatkin and Al-
fieri, 1991a,b; Huntingford et al., 1994; see also Pitcher, 1992; Chivers et
al., 1995b). Evidence against such a strategy suggests that inspectors may
not be caught in the prisoner’s dilemma in the first place (and thus the
TFT strategy would not apply; Magurran and Nowak, 1991; Murphy and
Pitcher, 1991; Magurran and Seghers, 1994; Godin and Davis, 1995; Ste-
phens et al., 1997). I cannot resolve this controversy, but much work clearly
remains to be done regarding the nature of predator inspection.
V. SOCIAL
SITUATIONS
A. ADAPTIVE
SOCIALITY
Decision making by individuals ought to influence the nature of sociality
under the risk of predation (e.g., Pulliam and Caraco, 1984). The last few
years have seen considerable progress in the study of such decision making,
but there are still surprisingly few studies in this area (see also Lima and
Dill, 1990; Krause, 199413). Recent years have also seen advances in the
comparative study of predation and sociality (notably in primates; e.g.,
Boesch, 1991; Cowlishaw, 1994; van Schaik and Horstermann, 1994; Stan-
ford, 1995), but such work is outside the scope of this review.
1. Spatial Position in Groups

Fish may seek out the innermost (safest) area in a group when threatened
by predators (Krause, 1993b). However, energetically stressed fish (Krause
et al., 1992; Krause, 1993a) and aquatic beetles (Romey, 1995) may seek
better feeding opportunities at their group’s (risky) periphery. A similar
“spatial conflict” between feeding and safety may influence the location
of web-building spiders within the larger colony (Rayor and Uetz, 1990,
1993). Krause (1994b) provides a cogent review of these and related studies
on spatial positioning in social animals.
2. Choice of Group
Larger groups should provide greater safety from predators than smaller
ones, all else being equal. Accordingly, fish given a choice prefer larger
240 STEVEN L. LIMA
groups, especially when under a heightened risk of predation (Hager and

Helfman, 1991; Ashley et al., 1993; Krause and Godin, 1994). Startled fish
may also join the largest available group, unless this group is much farther
away than a nearby, smaller group (Tegeder and Krause, 1995). A larger
group in a risky area may also be avoided (Ashley et al., 1993), and a
preference for larger groups may be overridden by a preference for individu-
als of a similar size (Krause and Godin, 1994; see also later discussion).
Krause and Godin (1995) found that large groups of fish may suffer more
attacks, but argue that prey are still better off in large groups (see also
Wrona and Dixon, 1991; Uetz and Hieber, 1994). Poysa (1991) suggests
that a duck’s choice of group may not be influenced by the risk of predation,
although these ducks may have realized that the predator in question was
not much of a threat.
An SDP model by Szekely et al. (1991) suggests that energetically stressed
birds should be less social (to avoid competitors) than those better fed. I
know of no studies testing this prediction in birds, but Romey and Rossman
(1995) describe such an effect in aquatic beetles. Paveri-Fontana and Fo-
cardi (1994) developed a model of optimal herd size selection in ungulates;
they related the results to various ecological processes, but the model’s
predictions for sociality per se were unclear.
3. Size-Assortative Grouping
A small individual in a group of large individuals (or vice versa) may be
conspicuous to predators and thus suffer a greater risk of attack (Wolf,
1985; Theodorakis, 1989). Such an effect may explain why fish in a group
associate preferentially with others of their size under a heightened risk of
predation (Theodorakis, 1989; Ranta et al., 1992a,b; Krause, 1994a; Krause
and Godin, 1994). However, under such conditions larger fish may aggres-
sively occupy the group’s central position, and thus preclude the intermin-
gling of size classes irrespective of any effect of conspicuousness per se
(Theodorakis, 1989; Krause, 1994a).
B. VIGILANCE
Many animals face a constant conflict between the need to be alert for
attack and the need to feed. A ubiquitous observation is that individuals
become progressively less vigilant (alert) as group size increases (see Elgar,
1989, for a brenchmark review). This “group size effect” is seen as an
outcome of the fact that individual group members can devote less time
to vigilance (i.e., more time to feeding) with increasing group size without
detracting from the group’s collective ability to detect attack (Elgar, 1989).
Interest in antipredatory vigilance has remained high in recent years,

and has entered a period of retrospection and reassessment of old ideas.
Insightful theory (Packer and Abrams, 1990; McNamara and Houston,
1992) and empirical studies (Krause and Godin, 1996) have elucidated
some key issues in the maintenance of social vigilance in selfish animals.
Refinements and challenges to the basic concept of collective detection
have appeared (Lima and Zollner, 1996; Roberts, 1996). The group size
effect itself has received better documentation (e.g., Roberts, 1995). Behav-
ioral sequences involving vigilance have received much needed attention
(Desportes et af., 1989; Roberts, 1994). Some exceptional observational
studies of predator-prey interactions shed further light on social vigilance
(e.g., Cresswell, 1994a). These studies and other developments have been
reviewed by Roberts (1996).
VI. REPRODUCTION
Sih (1994) summarizes the current state of affairs with regard to reproduc-
tive decision making under the risk of predation: “Although predation risk
is often viewed as an important component . . . of the evolution of mating
behavior, . . . little effort has gone into gaining a deep, ecologically-rooted
understanding of how predation risk influences reproductive behavior.” A
similar sentiment is expressed in Lima and Dill (1990), Magnhagen (1991,
1993), and Reynolds (1993). Recent years have nonetheless seen consider-
able progress in understanding such reproductive behavior in many con-
texts. I review this work below, and in keeping with my overall theme, I
focus on the management of predator-induced stress in ecological time.
Sih (1994) provides an excellent discussion of the more general evolutionary
and ecological aspects of reproductive behavior. I should note that “stress”
in this section refers ultimately to a loss of reproductive output, which may
or may not reflect a more standard form of stress (e.g., energetic) on the
animal in question.
A. MATECHOICE
Crowley et af.’s (1991) ground-breaking model of mate choice suggests
that females should become less choosy with an increase in the risk of
predation associated with locating mates. In other words, a given class of
males will enjoy a diminished mating advantage under a high risk of preda-
tion. This prediction is supported by observations of predator-induced ran-
dom mate choice in fish (Forsgren, 1992; Berglund, 1993). Godin and Briggs
(1996) also report a predator-induced lowering of female choosiness in
242 STEVEN L. LIMA
guppies, but only in females from high-risk streams (but mate choice copy-
ing by such guppies may not be influenced by predation risk, Briggs et
al., 1996). Similarly, the mating advantage enjoyed by longer-calling male
crickets may be overridden if females can approach short-calling males in
relative safety (Hedrick and Dill, 1993). On the other hand, large male
water striders enjoy an increased mating advantage under a high risk of
predation (Sih et al., 1990; Sih and Krupa, 1992, 1995,1996). This unusual
result may reflect the female-harassment-based mating system in water
striders (Krupa and Sih, 1993). Under a heightened risk of fish predation,
males harass females less (i.e., become less active), which may then allow
females to be more selective in their choice of mates or avoid mating alto-
gether.
B. ALTERNATIVE TACTICS
MALEMATING
Male guppies may court females via conspicuous visual displays, or at-
tempt “sneaky” forced copulations. Endler (1987) found that male guppies
attempted more sneaky copulations in the presence of predators. Similar
results have been reported in captive (Magurran and Seghers, 1990b) and
free-living guppies (Godin, 1995). It is perhaps intuitive that male guppies
would adopt the less conspicuous “sneaker” strategy in risky situations
(see also Lucas and Howard, 1995; Lucas et al., 1996), but sneaky males
may also be taking advantage of a female’s preoccupation with predator
inspection in the presence of predators (Magurran and Nowak, 1991; Godin,
1995). One might envision other scenarios of predator-induced flexibility
in alternative male mating tactics, but there appear to be no other reported
cases. However, Magnhagen (1995) found that the riskier tactics used by
sneaker and territorial common gobies (Pornatoschistus rnicrops) are used
less frequently in the presence of predatory fish.
C. MATING
DYNAMICS
The act of mating itself may be influenced by the risk of predation. For
instance, Travers and Sih (1991) found that male semiaquatic hemipteran
insects accept lowered mating success under a high risk of predation by
spending less time in tandem (copulating) with a female; tandem pairs
make tempting targets for predators (Sih, 1988). Sih and Krupa (1995,1996)
also found a decrease in mating duration and frequency in water striders
in the presence of fish, presumably at some reproductive cost to males;
tandem pairs once again are at greater risk than singletons (Fairbairn, 1993;
Rowe, 1994). Razorfish (Xyrichtys splendens) spawn closer to the (safe)
sea floor in high-risk situations, which may limit the dispersal success of
resulting zygotes (Nemtzov, 1994). Finally, copulation frequency and num-

ber in pipefish (Syngnathus typhle) may decrease in the presence of preda-
tors (Berglund, 1993; Fuller and Berglund, 1996), but copulation time may
increase to compensate (Berglund, 1993).
D. COURTSHIP
Conspicuous activities associated with courtship can lead to a higher risk
of predation for males (Lima and Dill, 1990; Magnhagen, 1991). Hence,
one might expect lowered courtship activity in the presence of predators.
This has been observed in several fish species (Endler, 1987; Berglund,
1993; Forsgren and Magnhagen, 1993; Nemtzov, 1994; Chivers et af., 1995~).
Area-specific differences in courtship activity by male fish may also be
determined by the local abundance of predators (Hastings, 1991); Lister
and Aguayo (1992) suggest that similar effects occur in lizards. Predators
may also inhibit courtship and spermatophore deposition by male salaman-
ders (Uzendoski et al., 1993). Following a predatory disturbance, the re-
sumption of courtship chorusing by male frogs is quicker in larger groups,
perhaps reflecting a greater dilution of risk in such groups (Jennions and
Blackwell, 1992).
E. OVIPOSITIONAL
BEHAVIOR
Mating dragonflies are sensitive to the presence of frogs in their choice
of oviposition sites (e.g., Michiels and Dhondt, 1990). However, dragonflies
appear unable to detect frogs lying in ambush (Rehfeldt, 1992). This inabil-
ity may explain why dragonflies are attracted to groups of ovipositing
pairs, as such groups form only in the absence of frog attacks (Rehfeldt,
1990, 1992).
Regarding theory, Mange1 (1989) and Weisser et af. (1994) developed
models of optimal ovipositional behavior by parasitoids searching in danger-
ous, patchy environments (see also Iwasa et af., 1984). The results suggest
that optimal patch residence times should be sensitive to the risk of mortality
experienced by ovipositing females. These models challenge the standard
view that parasitoids should act only to maximize their rate of oviposition,
but I know of no explicit tests of their predictions.
F. PREGNANCY
AND PARENTING
Observational evidence suggests that pregnant or lactating ground squir-

rels (MacWhirter, 1991) and bighorn sheep (Berger, 1991) take greater
risks in order to meet the energetic stresses of mammalian reproduction.
244 STEVEN L. LIMA
In contrast, egg-carrying marine copepods avoid food-rich surface waters

(Bollens and Frost, 1991); their opaque eggs make them vulnerable to
detection by surface-feeding fish, even at night.
Nest building and defense by male fish may also be predation-risk depen-
dent. Magnhagen (1990) showed that nest building by male black gobies
(Gobius niger) diminished in the presence of predators. The lack of such
an effect in sand gobies (Pomatoschistus minutus) was attributed to their
brief life-span (Magnhagen, 1990), which puts a premium on reproducing
as soon as possible. Magnhagen and Vestergaard (1991) found that male
common gobies took greater risks to defend their broods as their young
matured (and presumably became more vulnerable); Magnhagen (1992)
provides a general review of brood defense and parental risk taking in fish.
Surprisingly few studies on nestling provisioning in birds consider
risk to the parent to be an important determination of parental behavior
(Ydenberg, 1994). However, Harfenist and Ydenberg (1995) suggest that
rhinoceros auklet (Cerorhinca monocerata) chicks fledge younger and at
lower body mass in areas frequented by eagles because parents terminate
feeding earlier in high-risk areas. Such a decision is in accord with the
predictions of Clark and Ydenberg (1990a,b).
G. BREEDING
SUPPRESSION
A growing body of work, focused almost exclusively on small boreal
mammals (but see Fraser and Gilliam, 1992), addresses the issue of preda-
tion risk and the decision to engage in reproduction. Ylonen (1989) first
reported that bank voles (Clethrionomys glareolus) strongly suppress repro-
duction upon exposure to mustelid predators. Similar degrees of breeding
suppression have been observed in several other laboratory experiments
on bank voles (Ylonen et al., 1992; Ronkainen and Ylonen, 1994; Ylonen
and Ronkainen, 1994), other Clethrionomys voles (Ylonen et al., 1992;
Heikkila et al., 1993), and Microtus voles (Koskela and Ylonen, 1995).
Korpimaki et al. (1994) also demonstrate long-term breeding suppression
in bank voles under field conditions.
The mechanism behind this breeding suppression is not well understood.
However, female Clethrionomys voles aggressively avoid male advances
upon exposure to the scent of mustelid predators (Ylonen and Ronkainen,
1994; Ylonen, 1994). Male Microtus voles may themselves show less sexual
activity in high-risk situations (Koskela and Ylonen, 1995). Energetic stress
resulting from reduced feeding under high-risk conditions may also be
involved (Heikkila et al., 1993) in suppressing breeding. Research into
the estrous cycle of voles suggests that the mechanism behind breeding
suppression has a strong physiological component (Koskela et al., 1996).
This physiological link might conceivably relate to the negative effects of

physiologicalheuroendocrine stress (caused by exposure to predators) on
reproduction (Weiner 1992). Hansson (1995) suggests that reproduction in
some boreal voles may be sensitive to physiological stress of any sort, not
just that induced by predators.
VII. LONG-TERM
CONSEQUENCES
OF DECISION
MAKING
Most studies on antipredatory decision making accept the idea that any
decision has associated with it both a fitness cost (some form of predator-
induced stress) and benefit (avoiding an early death). How much do we
really know about these issues?
There are now several studies demonstrating that antipredator decision
making does indeed lower an animal’s risk of predation (as per examples
mentioned throughout this review). However, such benefits of antipredator
decision making remain a presumption in many research programs, espe-
cially those involving terrestrial vertebrates. It is thus perhaps disturbing
that a few studies have found antipredator responses to be inadequate in
some way. For instance, strong refuging behavior in larval salamanders can
be inadequate as a defense against fish predators (Sih et al., 1988; Sih,
1992b);a similar scenario is apparent in an amphipod predator-prey system
(Sparrevik and Leonardsson, 1995). McPeek (1990), Werner and McPeek
(1994), and Skelly (1995) report cases in which reduced activity in the
presence of predators failed to prevent predation; however, these cases
involved a lack of coevolutionary history between predator and prey.
Demonstrations of the long-term costs of antipredator behavior are rela-
tively uncommon. Recent years have nevertheless witnessed considerable
progress in identifying and quantifying these costs (Table 111). A common
theme in this work is that antipredator decisions that lower risk (usually
habitat shifts or decreased activity) also lead to some form of energetic
stress, typically manifest in lower growth rates. Slower growth may lead to
a smaller size at maturity (Skelly and Werner, 1990) or prolonged develop-
ment (Skelly, 1992). Exceptionally complete analyses of such predator-
induced stress, covering growth, development, and fecundity, have been
possible in mayflies (Peckarsky et al., 1993; Scrimgeour and Culp, 1994b)
and chironomids (Ball and Baker, 1995,1996). These insects have nonfeed-
ing adult life stages, and thus reduced larval growth translates directly into
reduced adult fitness (see also Feltmate and Williams, 1991; but see Duvall
and Williams, 1995, for a more complicated situation in stoneflies). It is
possible that a smaller size at maturity may reflect not only predator-
induced stress but also a predator-induced change in life history. There is,
TABLE 111
RECENT
EXPERIMENTAL
DEMONSTRATIONS
O F A LONG-TERM
COSTOF ANTIPREDATOR
DECISION
MAKING
Prey Predatof Prey response Conditions Cost Source
Invertebrates
Chironomid larvae Fish P Reduced activity Laboratory Slower growth and development, Ball and Baker (1995,
(Chironomus tentans) lower adult mass at emergence, 1996)
fewer eggs
Cladoceran spp. Copepod P,C Reduced activity (?) Laboratory Slower growth Gliwicz (1994)
2 Water flea (Daphnia Fish C Vertical migration Laboratory Deeper migrators experience Dawidowicz and Loose
magna) slower growth (1992); Loose and
Dawidowicz (1994)
Damselfly larvae None Reduced activity Laboratory Simulated predator-induced DixonandBaker (1988)
(Zschnura verticalis) reduction in feeding slows growth
and development
Ant (Lusius pallitarsis) Large ant P Reduced activity Laboratory Slower colony growth Nonacs and Dill (1990)
Dogwhelk (Nucellu Crab C Reduced activity Laboratory Slower (or zero) growth Palmer (1990)
lapillus)
Mayfly larvae (Baetis Stonefly P Escape-induced loss Semifield Adults emerge at lower mass, with Peckarsky et al. (1993)
bicaudatus) of feeding fewer eggs (no effect on
development time)
Marine snail (Strarnonita Crab P Reduced activity Laboratory Slower growth Richardson and Brown
haemastoma) ( 1992)
Mafly larvae ( B . Model fish P Reduced activity Laboratory Slower growth, lower adult mass, Scrimgeour and Culp
tricaudatus) longer development, fewer and (1994b)
smaller eggs
Buckmoth larvae Wasp P Microhabitat shift Semifield Slower growth Stamp and Bowers
(Hemileuca lucina) (1991)
Copepod (Eurytemora None Vertical migration Laboratory Simulated vertical migration leads Vourinen (1987)
hirundoides) to longer development
Vertebrates
Juvenile perch (Perca Fish P Habitat shift Semifield Slower growth (due to increased Diehl and Eklov (1995)
flu viatilis) competition)
Salamander larvae Fish P Reduced activity Laboratory Slower growth Figiel and Semlitsch
(Ambystoma (1990)
maculatum)
Guppy (Poecilia reticulata) Fish P Reduced activity, Field Reduced egg production and Fraser and Gilliam
habitat shift growth (1992)
Juvenile roach (Rutilzu Fish P Habitat shift SemifieId Slower growth Persson and Eklov
rutilus) (1995)
Toad larvae (Bufo Larval Reduced activity Laboratory Metamorphose at smaller size Skelly and Werner
N
P americanus) odonate P (1990)
4
Tree frog larvae (Hyla Salamander Reduced activity Semifield Slower growth and development Skelly (1992)
versicolor) P
Tree frog larvae (2 Larval Reduced activity Laboratory Slower growth Skelly (1995)
Pseudacris spp.) odonate P
Crucian carp (Carassius Fish P Habitat shift Field Slower growth (due to increased Tonn et al. (1992)
carassius) competition)
P, predator present; C, chemical scent of predator.

248 STEVEN L. LIMA
however, no clear evidence for such adaptive life-history changes (Skelly

and Werner, 1990; Ball and Baker, 1996).
The current emphasis on predator-induced reductions in growth rates is
entirely appropriate as most animals develop as free-living organisms for
whom successful reproduction means reaching adult size (cf. Werner and
Gilliam, 1984). However, for many birds and mammals, growth is often
largely complete before they strike out on their own. For such creatures,
the stress resulting from antipredator behavior is probably manifest in
decreased body condition (e.g., Hik, 1995; Sinclair and Arcese, 1995), lead-
ing ultimately to lower female fecundity or male competitive ability.
Such costs may also be manifest in energetic stress experienced by young
being provisioned by parents attempting to avoid predation (Harfenist
and Ydenberg, 1995).
The survival-growthireproduction trade-offs apparent in Table 111 seem
adaptive, given that an early death is the likely alternative to incurring some
form of predator-induced stress. However, the degree to which “adaptive”
approximates “optimal” is unknown. This should come as no surprise, given
our inability to quantify many aspects of predation risk (see Section 11,E).
Perhaps Nonacs and Dill (1990) come closest to making this distinction.
They estimated the benefits to an ant colony from extra foraging and the
cost of losing workers to predators, and found that the risks taken by
workers reflected the potential increase in colony growth as a result of
extra foraging.
VIII. ECOLOGICAL AND IMPLICATIONS

INFLUENCES
Decision making under the risk of predation can influence the nature of
ecological systems. Understanding these influences has long been a major
driving force in the study of antipredator decision making (Sih, 1980; Wer-
ner et al., 1983). Here, I discuss recent work in this area within three main
contexts: the use of space by individuals, population-level consequences,
and species interactions. This work involves mostly field or semifield experi-
mentation. Although often not achieving the controlled rigor of laboratory
experimentation, this work nevertheless illustrates the potential ecological
effects of predator-induced stress and antipredatory decision making.
A. USEOF SPACE
Table IV summarizes studies indicating that predators have a pervasive
effect on the use of space by a variety of animals. This work adds to the
many (but far fewer) studies on the use of space discussed in Lima and
TABLE IV
RECENT
STUDIES
EXAMINING UNDER THE RISKOF PREDATION
THE USEOF SPACE
Prey Predatof Scaleb Results Source
Invertebrates
Chironomid larvae Fish m No preference for predator-free areas (main Baker and Ball (1995)
(Chironomus tentans) response involved lower activity)
Juvenile crayfish Fish, adult m No consistent preference for safer microhabitats Blake et al. (1994)
(Pacifmtacus leniusculus) crayfish
Mayfly larvae (Baetis Fish C m Avoid profitable but risky feeding locations Cowan and Peckarsky (1994)
bicaudatus)
Stonefly larvae Fish m Strong preference for color-matching substrate, Feltmate and Williams (1989b)
(Paragnetina media) (undiminished in absence of predator)
Epibenthic invertebrates Fish M Choose areas rich in refuges (woody debris) Everett and Ruiz (1993)
(several spp.)
Mayfly larvae (B. Fish m Avoid profitable but risky feeding locations Kohler and McPeek (1989)
tricaudatus)
Damsellly larvae (Ischnura Fish m Strong preference for dark (safe) substrates, Moum and Baker (1990)
venicalis) which may be enhanced in the presence of
predators
Whelk (Baccinum Starfish m, M m: attracted to feeding starfish Rochette et al. (1995)
undarum) M: avoid areas with abundant starfish
Isopod (Saduria entomon) Large isopod m Avoid areas with abundant predators Sparrevik and Leonardsson (1995)
Juvenile lobster (Homarus Fish. crab m, M Predator-induced preference for safe, cobble Wahle and Steneck (1992)
americanus) substrate
Hermit crabs (Clibanarius, Crab C, alarm m Crabs with ill-fittingshells seek areas with recently Rittschoff et al. (1992)
Pagurus spp.) substance killedshell occupants;other crabsflee such areas
Flsb
Roach (Rutilus rutilus) Fish M Shift from pelagic to littoral zone after Brabrand and Faafeng (1993)
predator introduction
Stream fish (4 spp.) Fish m, M Juveniles and smaller species shifi to shallow Brown and Moyle (1991)
water at both micro- and macroscales
Bluegill and shad (Lepomis Fish m Only bluegill shift to shallow water in presence De Vries (1990)
and Dorosoma spp.) of predator
(continued)
TABLE IV (Continued)
Prey Predator" Scaleh Results Source
Juvenile perch (Perca Fish m, M m: remain close to refuge in presence of predator Diehl and Eklov (1995);
fluviatilis) M: avoid profitable but risky refuge-poor habitats Eklov and Diehl (1994);
Persson and Eklov (1995
Perch and rudd (Perca and Fish M Species segregate into pelagic vs littoral habitats Eklov and Hamrin (1989)
Scardinus spp.) based in part on vulnerability to predator
Perch and roach (Perca and Large perch m, M m: remain close to refuge in presence of predator Eklov and Persson (1995);
Rutilris spp.) M: prefer refuge-rich habitat in presence of Christensen and Persson 1993);
predator Persson (1991, 1993)
Small stream fish (Riuuhcs Fish M Avoid stream pools with predators; move to riffles Fraser and Gilliam (1992)
and Poecilia spp.)
Killifish (Rivulus hartii) Fish m, M Fish avoid streams populated by predators Fraser et al. (1995)
Juvenile salmon Bird, fish m Prefer deeper water under nonturbid conditions Gregory (1993)
(Oncorhynchus
tshawytscha)
Juvenile cod (Gadus Large cod m Predator-induced preference for safe, cobble Gotceitas and Brown (1993)
morhua) substrate
Sculpin (Cottus bairdi) Fish m Microhabitat use unaffected by predator presence Grossman et al. (1995)
Small stream fish (mainly Fish m, M m: shift to shallow water in presence of predator Harvey (1991)
juv. Lepomis) M: avoid pools with predators
Small, soft-rayed fish (4 Fish M Much emigration from lake (into outlet stream) He and Kitchell (1990)
SPP.) following predator introduction
Stickleback (Gasterosteus Fish m Stay close to bottom in presence of predator Ibrahim and Huntingford (1989)
aculeatus)
Bleak (Alburnus alburnus) Fish, alarm m Preference for vegetated habitats is enhanced by Jachner (1995a,b)
substance predators and diminished by food in open water
Arctic charr (Salvelinus Fish M Ontogenetic shift to pelagic habitat is delayed L'AbCe-Lund et al. (1993)
alpinus) under risky conditions
Fathead minnow Alarm substance m Avoid areas marked with alarm substance Mathis and Smith (1992); Chivers
(Pimephales promelas) et al. (1995a)
Perch (P.puviatilis) Fish M Choice of littoral (safe) or pelagic zone of lake Persson et al. (1996)
determined by presence of non-gape-limited
predator
Small fish (several spp.) Fish, crab M Preference for shallow water reflects risk in Ruiz e f al. (1993)
deep water
Juvenile pollock (Theragra Large pollock m Predator-induced preference for vegetated Sogard and Olla (1993)
chalcogramma) habitats
Mosquitofish (Gambusia Fish m Larvae may avoid adult cannibals by associating Winkleman and Aho (1993)
holbrooki) with predators that are avoided by adults
Brook stickleback (Culaea Alarm substance m Avoid areas marked with alarm substance Wisenden ef al. (1994)
inconstans)
Amphibians and Reptiles
Salamander larvae Fish m, M m: predator-induced preference for shallow water Sih et al. (1992)
(Ambystoma barbouri) M: avoidance of pools with predators
Salamander larvae (2 Large salamander m Shift to deeper water in presence of predator (one Walls (1995)
Ambystoma spp.) species only)
fj Birds
3
Himalayan snowcock Eagle m, M Escape tactic may constrain birds to steep terrain Bland and Temple (1990)
( Tetraogallus
himalay ensis)
Redshank (Tringa totanus) Raptors m, M m: juveniles feed in risky, profitable microhabitats Cresswell (1994b)
M: prefer less profitable but safe habitat
(mussel beds)
Titmice (2 Parus spp.) Raptor m Feed in open (away from vegetation) only when Hinsley et al. (1995)
forced to do so by aggression
Willow tit (P. montanus) Raptors m Feed in open only when forced to do so by Koivula et al. (1994)
aggression
Sparrows (2 emberizid spp.) Raptors m, M Willingness to feed in open related to escape tactic Lima (1990a)
Anna’s hummingbird Terrestrial birds m Avoid profitable feeding opportunities close to Lima (1991)
(Calypte anna) ground
Small granivores (7 spp.. Raptors M Large-scale habitat choice influenced by escape Lima and Valone (1991)
mostly emberizids) tactics
(continued)
Prey Predatof Scaleb Results Source

Downy woodpecker Choice of feeding site reflects vigilance-escape Lima (1992)
(Picoides pubescem) trade-off
Brambling (Fringilla Raptors M Prefer forest habitat over profitable but risky Lindstriim (1990)
montifringilla) open habitat
Duck (Anas penelope) Raptors, humans m Reluctant to feed far from water (refuge) Mayhew and Houston (1989)
Sparrows (3 emberizid spp.) Raptors m Avoid open areas, even those with high food Repasky and Schluter (1994)
density (except one sp.)
Small granivores (several Raptors m General avoidance of relatively profitable but Schluter (1988)
spp.. old-world open areas
granivores)
White-crowned sparrow Raptors m Feed in open only when forced to do so by Slotow and Rothstein (1995)
(Zonotrichia leucophrys) aggression
Titmice (2 Parus spp.) Raptors m Feed in open only when forced to do so by Suhonen (1993a,b;
aggression Suhonen et al. (1993)
Small birds (several spp.. Raptors M Small (vulnerable) species avoid nesting in Suhonen et al. (1994)
mostly passerines) vicinity of (up to 1 km or more from) falcon
nests
Blue tit (P. cueruleus) Raptors m Avoid profitable but open (risky) feeding sites Todd and Cowie (1990)
Sparrows ( 2 emberizid spp.) Raptors m, M Choice of feeding location influenced by escape Watts (1990)
tactics
Savannah sparrow Raptors m Reluctant to feed far from vegetated refuge Watts (1991)
(Passerculus
sandwichensis)
Mnmmals: Rodents
Gerbils (2 Gerbillus spp.) Owls m Avoid open (nonvegetated) areas when risk is Abramsky et al. (1996)
increased under field conditions
Kangaroo rat (Dipodomys Owls (?) m Avoid profitable but risky open microhabitats Bowers (1990)
merriami)
White-footed mouse Mammals, raptors m, M Avoid feeding opportunities in open habitats Bowers and Dooley (1993)
(Peromyscus leucopus)
Squirrels (Sciurus and Raptors m Avoid profitable but risky open microhabitats Bowers et al. (1993)
Tamias, 2 spp.)
Kangaroo rats ( 2 Snakes m Avoid feeding opportunities in vegetation that Bouskila (1995)
Dipodomys spp.) might be occupied by active snakes
Desert rodents (2 Raptors (mainly) m Avoid open areas, but kangaroo rats are more Brown (1989)
heteromyid, 1 sciurid) likely to be in open habitat than other species
Crested porcupine (Hystrix Large mammals m, M Avoid profitable feeding opportunities in open Brown and Alkon (1990)
indica) habitats
Fox squirrel (S. niger) Raptors/mammals m Avoid profitable but risky open microhabitats Brown et al. (1992a);
Brown and Morgan (1995)
Gerbil (G. allenbyi) Owls (?) m (M?) Perceive increased risk in rocky habitats, which Brown et al. (1992b)
are usually avoided
Guinea pig (Cavia aperea) Raptordmammals m Appear to perceive higher risk when away Cassini (1991);
from vegetation Cassini and Galante (1992)
Prairie dog (Cynomys Raptors m Avoid feeding far from refuge (burrow) unless Devenport (1989)
ludovicianus) feeding in groups
W
House mouse (Mus Mammals m, M Seek out vegetated habitats under increased risk Dickman (1992)
domesticus)
Gerbil ( G . tytonis) Raptors/mammals m Avoid profitable but risky open microhabitats Hughes and Ward (1993);
Hughes et al. (1994)
Field vole (Microtus Raptors/weasels m Avoid open areas in presence of kestrel, may Korpimaki et al. (1996)
agrestis) avoid cover when in presence of weasels
Gerbils (2 GerbiZZus spp.) Owls m Avoid profitable but risky open microhabitats Kotler (1992);
(can distinguish risk posed by different owl Kotler et al. (1991, 1994a);
species) Kotler and Blaustein (1995)
Gerbils (2 Gerbillus spp.) Snakes m Avoid feeding opportunities in vegetation that Kotler et al. (1992, 1993a,b)
might be occupied by active snakes
Degu (Octodon degus) Raptors m Appear to perceive higher risk when away Lagos et al. (1995a,b)
from vegetation
Desert rodents (4 Owls m Avoid open areas, but kangaroo rats are more Longland and Price (1991)
heteromyids, 1 cricetid) likely to be in open habitat than other species
(continued)
Prey Predator“ Scale” Results Source
Townsend’s vole ( M . Mammal C m Avoid feeding opportunities in open habitats Merkens et al. (1991)
townsendit)
Desert rodents (2 Snake m No consistent effect of snakes on use of space (on Pierce e f al. (1992)
heteromyids, 1 cricetid) very constrained spatial scale)
Small rodents (5 spp.. Raptors m Avoid open areas (which may not be very Simonetti (1989)
mostly cricetids) profitable)
Ground squirrels Raptors, m Avoid profitable but risky open microhabitats: Smith (1995)
(Spermophilus and mammals faster species feeds farther from cover
Tamias, 2 spp.)
Porcupine (Enthizon Mammals m, M Avoid feeding in open but more profitable Sweitzer and Berger (1992)
dorsafum) habitats
g Mammals: Nonrodent
* Bighorn sheep (Ovis Large mammals m, M Pregnant sheep leave relative safety of steep Berger (1991)
canadensis) terrain for better foraging
Hedgehog (Erinaceus Badger M May choose habitats in which predators are absent Doncaster (1993, 1994)
europaeus)
Pika (Ochotona cdlaris) Raptors, m Avoid profitable but risky microhabitats away Holmes (1991)
mammals from refuge
Ibex (Capra ibex) Large mammals m (M?) Preference for cliffs over flat terrain may be due Kotler e f al. (1994b)
to increased perceived predation risk in latter
habitat
Jackrabbit (Lepus Raptorsimammals m Perceive higher risk when away from vegetation Longland (1991)
californicus)
Buffalo (Syncerus cafer) Lions m. M No clear indication that lions influence use of Prins and Iason (1989)
space, despite spatial variation in predation risk
C, chemical scent of predator only; otherwise predators were present in environment.

” m. microscale: M. macroscale.
Dill (1990). In Table IV, the “microscale” category refers to an animal’s

use of its immediate surroundings, very often in the vicinity of a refuge
from attack. The “macroscale” category is more difficult to specify, but
refers to a scale at which changes in the use of space require a significant
investment in movement. The absolute spatial scale of macro- and micro-
habitat use is, of course, species-specific.
1. Invertebrates
Studies on invertebrates (Table IV) show a tendency for individuals to
avoid risky micro- or macrohabitats, even if such habitats offer good feeding
opportunities. Rochette et al. (1995) describe an unusual case in which
whelks avoid predatory starfish on a macroscale, but feed close to prey-
consuming starfish on a microscale; starfish occupied by prey consumption
are not dangerous, and produce “scraps” on which whelks can feed. Simi-
larly, hermit crabs with ill-fitting shells may be attracted to areas of recent
predation on gastropods in an attempt to obtain a better fitting shell;
individuals with proper-fitting shells often flee from such area (Rittschoff
et al., 1992).
Note that I have already reviewed the use of space by certain inverte-
brates in other contexts. For instance, die1 vertical migration in zooplankton
(see Section II1,B) involves a macroscale change in the use of the water
column. Nocturnal drift in stream-dwelling arthropods (Section II1,B) also
involves a macroscale change in location within a stream. Sih and Wooster
(1994) and Wooster and Sih (1995) provide excellent reviews of drift behav-
ior in stream animals and its consequences for local prey population regula-
tion; subsequent work by Crow1 and Covich (1994), Forrester (1994a,b),
Rader and McArthur (1995), and Kratz (1996) will also interest anyone
working in this general area.
Taking a different perspective, inadequate antipredator behavior may
be a major determinant of the large-scale distribution of certain inverte-
brates. For instance, Daphnia aregonensis is largely absent from lakes occu-
pied by a predator toward which its antipredator behavior is ineffective
(Ramcharan and Sprules, 1991). Larval damselflies typical of fish-free ponds
exhibit antipredator responses that are inadequate against the fish in perma-
nent ponds, and vice versa (Blois-Heulin et al., 1990; McPeek, 1990;McPeek
et al., 1996. Henrikson (1988) suggests similarly that inappropriate escape
responses toward fish limit a libellulid dragonfly larva to fish-free lakes.
Note, however, that these odonate larvae do not directly make decisions
regarding their distribution among ponds or lakes; such decisions are made
by ovipositing adults.
2. Fish
Many recent studies demonstrate that predators are a major determinant
of the use of space by fish. At the microscale, fish tend to remain in or
256 STEVEN L. LIMA
near safe habitats (e.g., shallow water, vegetation, safe substrates), while
at the macroscale they avoid predator-rich or refuge-poor habitats (Table
IV; see also Sih, 1987; Lima and Dill, 1990; Milinski, 1993). Eklov, Persson,
and colleagues provide an unusually complete look at the use of space by
small fish, which covers the spectrum from mechanistic studies of prey
behavior (Eklov and Persson, 1996) to field studies examining whole-lake
phenomena (Persson et al., 1996). Most fish-related studies in Table IV
deal with lake systems, but the distribution of fish within and among stream
pools is also influenced by predators (Brown and Moyle, 1991; Harvey,
1991; Fraser and Gilliam, 1992; see also Power et al., 1985; Schlosser, 1987;
but see Grossman et al., 1995). Along these lines, Fraser et al. (1995) link
small-scale decisions regarding the use of space to whole-drainage patterns
in the distribution of killifish.
3. Amphibians and Reptiles

Work on the use of space by these animals has been limited (Table IV),
and there is a clear need for work on reptiles. The few existing studies
suggest that predator-induced effects in larval amphibians are similar to
those seen in fish (see also Lima and Dill, 1990). Some studies also show that
larval amphibians stay as far from predators as possible in small laboratory
containers (e.g., Hews, 1988; Skelly and Werner, 1990), suggesting that
their microhabitat use might be predation-risk dependent. Morey (1990)
and Heinen (1993, 1994b) also found that frogs and toads, respectively,
choose substrates against which they are most cryptic; this has obvious
implications for the use of space under the risk of predation. On a large
scale, some studies link inadequate antipredator behavior to the distribution
of larval amphibians within streams (Sih, 1992b; Feminella and Hawkins,
1994) or among temporary versus permanent ponds (Kats et al., 1988;
Werner and McPeek, 1994). As in similar cases with invertebrates, however,
the choice of temporary versus permanent ponds is made not by these
larvae but by adults (Resetarits and Wilbur, 1989).
4. Birds
An emerging avian theme is that the use of space relative to vegetative
cover is determined to a large extent by escape tactics (Lima, 1993). Birds
with vegetation-dependent escape tactics are reluctant to feed far from
vegetative cover (Table IV). Observations of raptor predation on birds
confirm the adaptive nature of this reluctance to feed in the open (Watts,
1990; Suhonen, 1993a,b; Hinsley et al., 1995). Although less well studied,
birds with vegetation-independent tactics may avoid vegetative cover alto-
gether (Lima, 1993).
Most avian studies take a microscale perspective (Table IV), but patterns
at this scale may also translate to larger spatial scales (Lima and Valone,
1991; Watts, 1991). Bland and Temple (1990) describe a situation in which
a bird’s gravity-assisted, downhill escape tactic may explain its geographic
restriction to mountainous terrain. Birds may enter macrohabitats not well
suited to their escape tactics, or relatively risky macrohabitats, only (1) if
forced to do so by aggression (e.g., Cresswell, 1994b), or (2) if such habitats
offer exceptional foraging opportunities (e.g., Lindstrom, 1990). On a differ-
ent note, the location of falcon nests may also influence the large-scale
distribution of breeding passerines (Suhonen et al., 1994).
5. Mammals
Recent work shows convincingly that small mammals (mostly rodents)
avoid feeding far from protective cover (e.g., vegetation), even at the cost
of forgoing high-quality feeding opportunities (Table IV). Thermophysio-
logical stress in the open cannot account for the avoidance of open areas
(Bozinovic and Simonetti, 1992; Sweitzer and Berger, 1992; Kotler et al.,
1993d; Bowers et al., 1993; Lagos et al., 1995a), but such effects deserve
more attention. The strong attraction of woody vegetative cover for desert
rodents can be reduced or reversed when such vegetation harbors predatory
snakes (Table IV). In this regard, Kotler et af. (1992) and Korpimaki et al.
(1996) note the possibility of “predator facilitation” in which the avoidance
of vegetative cover makes prey more available to open-hunting predators
(or vice versa; see also Daly et al., 1992). Schooley et al. (1996) note further-
more that vegetation may present obstacles to escape and predator detec-
tion for some diurnal rodents, hence their preference for open areas.
Work on large mammals is sparse and mixed (Table IV). Predation risk
may be a factor in the use of space by bighorn sheep (Berger, 1991) and
ibex (Kotler et al., 1994b), but perhaps not by African buffalo (Prins and
Iason, 1989).
Work on the use of space by mammals usually focuses on small spatial
scales (Table IV). Doncaster’s (1993, 1994) work on hedgehogs provides
a notable exception. It nevertheless seems likely that the ubiquitous micro-
scale avoidance of open areas by small mammals (Table IV) will translate
to larger spatial scales. In other words, habitats with little vegetative cover
will probably be avoided by animals reluctant to forage away from such
cover (see also Price ef al., 1994).
B. POPULATION-LEVEL
CONSEQUENCES
Antipredatory decision making could in principle influence many aspects
of prey population dynamics and regulation (e.g., Desy et al., 1990; Chesson
258 STEVEN L. LIMA
and Rosenzweig, 1991; Schluter and Repasky, 1991; Sinclair and Arcese,
1995). This possibility is readily apparent given the long-term negative
effects of predator-induced stress (Table 111). However, translating behav-
ioral decisions to their population-level consequences has proven difficult.
Actually, the extent to which this is true depends on the scale of analysis.
The influence of predators on local population dynamics can often be
understood in terms of decisions affecting the large-scale distribution of
animals (see previous section). Nevertheless, studies covering whole popu-
lations are unusual.
The “whole population barrier” has been broken by some experimental
studies focusing on small lakes in which entire populations of predators
and prey can be manipulated and monitored (although often with limited
replication). He and Kitchell (1990) provide a particularly good case in
point. They showed that the “crash” in the prey population following the
introduction of pike into a lake was caused by a large-scale movement of
prey fish out of the lake and into the outflow stream (see also H e and
Wright, 1992). Tonn et al. (1992) also performed a whole-lake manipulation
of predators. In this case, predatory perch induced an almost exclusive use
of the shallow (safe) littoral zone by young crucian carp. This led to a
competitive bottleneck that ultimately limited recruitment to adult life
stages relative to a control population (see also Diehl and Eklov, 1995).
Individuals surviving this bottleneck grew much larger than control fish
after shifting to the competition-free pelagic zone. This scenario parallels
that in Werner el al.’s (1983) landmark study in a bass-sunfish system.
Recent work in similar systems suggests that such bottlenecks can alter the
competitive relationship among prey species (Brabrand and Faafeng, 1993;
see also next section). Furthermore, an understanding of these predator-
induced bottlenecks can provide insight into the nature of stock-recruitment
relationships of importance to fisheries management (Walters and
Juanes, 1993).
Models of predator-prey population dynamics abound (Crawley, 1992),
but very few incorporate adaptive antipredator behavior. Abrams (1993b)
argues that most predator-prey models actually suffer from assumptions
not easily supported by adaptive antipredator behavior. Ruxton (1995)
found that adaptive antipredator behavior acts to stabilize otherwise oscilla-
tory predator-prey population dynamics, complementing results from ear-
lier modeling (Ives and Dobson, 1987). Crowley and Hopper (1994) present
an extraordinary modeling attempt linking a stochastic-dynamic game be-
tween predator and prey to stock-recruitment curves and resulting popula-
tion dynamics.
Predator-prey population cycling might also be influenced by antipreda-
tor decision making by prey. Hik (1995) presents evidence that energetic
stress following a predator-induced microhabitat shift by snowshoe hare

(Lepus arnericanus) causes a lowering of hare reproductive output, which
then hastens the decline and lengthens the recovery phase in the cyclic
population dynamics of hare and their mammalian predators. Similarly,
Ylonen (1994) and Oksanen and Lundberg (1995) suggest that predator-
induced breeding suppression (see Section VI,G) hastens the crash phase in
the cyclic population dynamics of boreal voles and their mustelid predators.
Ylonen (1994) outlines the specific idea that breeding suppression repre-
sents an attempt by female voles to ride out (in a high-survival, nonrepro-
ductive state) the high-predation part of a population cycle, after which
they and their offspring would have a better probability of survival. Lambin
et al. (1995) leveled some harsh criticism against this idea regarding breeding
suppression and vole population dynamics, claiming that many of its key
assumptions are unsupported (especially the assumption of enhanced survi-
vorship in nonreproductive females). Ylonen’s idea still has considerable
merit, but there is clearly a need for critical experimentation and quantita-
tive modeling regarding the role of breeding suppression in predator-prey
population dynamics.
C. SPECIES
INTERACTIONS
Recent studies illustrate how antipredator decision making might influ-
ence species interactions. These studies emphasize the role of indirect
interactions between predators and other species mediated by the preda-
tors’ effect on the behavior of a third (transmitter) species (Abrams, 1995).
Such indirect interactions have been termed higher order interactions
(Werner, 1992) or trait-transmitted indirect effects (Abrams, 1995), but for
clarity I will use the term behaviorally transmitted indirect effects.
Behaviorally transmitted indirect effects may act in a variety of ways to
alter the outcome of interspecific competition (Werner, 1992). For example,
similar refuging behavior under a high risk of predation may lead to one
(transmitter) species excluding another from the refuges. This has the effect
of leaving the lesser competitor exposed to greater predation, which may
ultimately tip the competitive balance in favor of the transmitter species.
Such a scenario may apply in fish-crayfish systems (Hill and Lodge, 1994;
Soderback, 1994) and a fish-salamander-isopod system (in which fish con-
sume both salamanders and isopods; Huang and Sih, 1990). Werner (1991)
argues that greater larval bullfrog activity (movement) in the presence of
predators gives them a competitive advantage over larval green frogs; these
two species are evenly matched competitors in the absence of predators.
This effect of differential activity ultimately interacts with direct predatory
effects in determining the distribution of these two species among perma-
260 STEVEN L. LIMA
nent versus temporary ponds (Werner, 1994; Werner and McPeek, 1994).
Similar movement-related effects may influence competition between larval
mosquitos ( Juliano et al., 1993; Grill and Juliano, 1996). On the other hand,
Tayasu et af. (1996) argue on empirical and theoretical grounds that similar
levels of predator-induced inactivity in two shrimp species may allow for
coexistence that would not otherwise be possible. Here, lowered activity
in the superior competitor favors coexistence via a reduction in the overall
level of interference competition.
Behaviorally transmitted indirect effects may also be evident when preda-
tors influence a particular species’ use of space (Werner, 1992). Leibold
(1991) describes a case in which competitive exclusion between two zoo-
plankton species may be prevented by a predator-induced habitat shift in
the superior competitor (the transmitter species). Cases have also been
reported in which the similar use of space in the presence of predators
intensifies interspecific competition among fish (Person, 1991, 1993; Bra-
brand and Faafeng, 1993) and desert rodents (Hughes et al., 1994). Finally,
recent work on gerbils provides a cautionary tale regarding the use of space
and its ultimate effects on species interactions. Despite the fact that two
competing gerbil species may use space differently in the presence of preda-
tors (Kotler et al., 1991), the temporal partitioning of activity appears to
form the basis for their coexistence (Kotler et al., 1993c; Ziv et al., 1993;
Brown et al., 1994).
Behaviorally transmitted indirect effects have also been implicated in
cases of strong “top-down’’ ecosystem regulation; such regulation dictates
that a change in the abundance of top predators causes indirect ecological
effects, which are transmitted all the way down to the lowest trophic levels
of a food web (Power, 2992). For instance, Turner and Mittelbach (1990)
found that the strong indirect effect of piscivorous bass on zooplankton
communities is transmitted by predator-induced changes in the use of space
by planktivorous sunfish. Diehl and Eklov (1995) and Person ef al. (1996)
describe a very similar situation in a piscivore-+perch-+invertebrate trophic
system (arrows indicate predator-prey relationships). In a sunfish-sal-
amander+isopod system simulated by Huang and Sih (1991). a positive
effect of fish on isopods is transmitted primarily via a strong refuging
response by salamanders to the presence of fish. Turner (1997) provides
an extreme case of behaviorally transmitted top-down effects in a simulated
predator-+snail+algae system in which the mere chemical scent of preda-
tion drives the system. Finally, Hill and Lodge (1995) describe a case in
which the (nonlethal) presence of predators mediates top-down effects
in a fish+crayfish-+macroinvertebrate+plant system via both behavioral
changes and increased mortality in crayfish (the latter being caused by
increased fighting for refuges).
The importance of behaviorally transmitted indirect effects in ecolog-

ical systems has also been explored theoretically in recent years. Abrams
(1992,1995) and Abrams and Matsuda (1993) make a convincing case that
(1) community-level models ignoring such indirect effects may be mislead-
ing, and (2) a variety of indirect effects may be expected if both predator
and prey can change their behavior adaptively (see also Kotler and Holt,
1989). Abrams (1995) notes also that such adaptive behavioral traits may
make it difficult to even distinguish and classify direct versus indirect effects.
Indirect effects also figure prominently in models suggesting that ecological
communities will be more speciose if prey exhibit predator-specific rather
than generalist antipredator behavior (Matsuda el al., 1993,1994,1996; see
also Brown and Vincent, 1992, for a different perspective on this issue).
IX. ADDITIONAL
CONSIDERATIONS
In this section I group four disparate topics about which relatively little
is known. These topics nonetheless address several important issues in the
study of decision making under the risk of predation.
A. PHYSIOLOGICAL AND DECISION

STRESSRESPONSE MAKING
A threatening situation often induces the classic “fight or flight” physio-
logical (neuroendocrine) stress response, which involves (among other
things) the immediate production of hormones like cortisol, epinephrine,
and norepinephrine (Weiner, 1992); recent work suggests that this response
is even greater than previously thought (Le Maho et al., 1992). One of the
short-term physiological effects of the basic stress response is to make more
energy available for immediate action like escape (Weiner, 1992). Many
stimuli will produce this stress response, such as aggressive conspecifics,
unfamiliar terrain, novel objects, and so on (Boissy, 1995). Of course,
predators may also induce such a response, but relatively little work ad-
dresses the effects of predators per se (but see Levine et al., 1993; Boissy,
1995). However, work on stress caused by being approached or handled
by humans (Le Maho etaf.,1992; Boissy, 1995) has an obvious relationship
to physiological stress caused by predators.
The physiological stress response is well known, but its relationship to
antipredator decision making represents unexplored territory. Indeed, the
relationship between the basic stress response and subsequent behavior is
not always clear (Boissy, 1995). Experimental work in which the stress
response is chemically blocked does suggest, however, that elevated levels
of stress hormones affect (in part) various antipredator behaviors (Berco-
262 STEVEN L. LIMA
vitch et al., 1995). Boissy (1995) argues further that individual differences
in “fearfulness” among animals are related causally to such differences in
the stress response.
It thus seems likely that the physiological stress response is mechanisti-
cally linked in some way to antipredator decision making. It is, in fact,
conceivable that the stress response is to a significant extent a target of
selection in the evolution of antipredator behavior in general, especially
as it relates to short-term changes in responsiveness to predators. It is also
conceivable that an unusually extreme stress response may actually impair
decision making in some way; Mesa et al, (1994) suggest such a possibility
with regard to non-predator-induced physiological stress, but the same
might well hold for stress caused by chronic exposure to unusually high
predation risk (see also following discussion). It is also tempting to speculate
further that certain aspects of antipredator decision making are designed
to avoid the long-term effects of a chronic physiological stress, such as
stress-induced diseases and suppression of the immune system (Ader et al.,
1991); such a realization may have important implications for the design
of experiments on antipredator behavior (see following discussion). As
mentioned earlier, the reproductive effects of such physiological stress may
also impinge on our interpretation of predator-induced breeding suppres-
sion (see Section V1,G).
All of the forgoing discussion on physiological stress pertains to verte-
brates. In fact, most research has been conducted on only a small number
of mammals, birds, and fish of economic or medical importance (Schreck,
1990; Mesa etaf., 1994; Boissy, 1995). The results obtained thus far probably
apply to most vertebrates, but their relevance (if any) to physiological stress
and the antipredator behavior of invertebrates seems largely unexplored.
THE RISKOF PREDATION

B. ASSESSING
An assessment of the risk of predation must in some way form the basis
for antipredator decision making (Blumstein and Bouskila, 1996), but little
is known about the way in which such assessments are made. A great deal
is known about the sorts of predatory stimuli that animals interpret with
alarm (see Curio, 1993, for an excellent discussion), but the way in which
animals integrate information on predator abundance, the likelihood of
escape, and so on, into some sort of assessment of predation risk is unknown.
Following Lima and Dill (1990), it seems likely that animals use “rules of
thumb” in assessing the prevailing risk of predation. It also seems likely
that any such assessment will be fraught with uncertainty. In this regard,
Bouskila and Blumstein (1992) argue that animals might adaptively overes-
timate the risk of predation to avoid the relatively high costs of underesti-
mating risk. Abrams (1994) cautions, however, that underestimating the

risk of predation can be favored under certain circumstances.
Work on the chemical detection of predators might shed light on this
issue of assessing risk. Scores of studies show that a variety of animals can
detect a threat of predation via chemicals emitted by predators (for recent
reviews, see Weldon, 1990; Smith, 1992; Larsson and Dodson, 1993; Dodson
et al., 1994; Kats and Dill, 1998). As argued by Kats and Dill (1998), the
concentration of such chemicals might provide an accurate estimate of
predation risk. This might explain why the strength of antipredator behavior
in zooplankton (Ramcharan et al., 1992; Loose and Dawidowicz, 1994) and
tadpoles (Horat and Semlitsch, 1994) increases with the concentration of
fish-emitted chemicals. However, very few studies examine behavioral re-
sponses to varying chemical concentrations, nor have such concentrations
been related to mortality, predator abundance, and so on. Future work in
this area might well demonstrate that predator-emitted chemicals provide
many types of animals with an accurate estimate of the risk of predation
(Kats and Dill, 1998).
C. PREYACTION
A N D PREDATOR
REACTION
The study of antipredatory decision making is hindered by a lack of
information on the way in which predators respond (in ecological time) to
the antipredatory actions of their prey. In fact, a tacit assumption in the
vast majority of studies reviewed herein is that factors like attack rate are
fixed entities to which prey determine their optimal response. There are
nonetheless many scenarios in which prey behavior might influence preda-
tor behavior (and thus the components of risk controlled by predators, e.g.,
Lima, 1990b).
The smattering of studies addressing this issue of “action and reaction”
cover a wide range of phenomena. Johansson and Englund (1995) consider
explicitly the behavioral interaction between a refuging prey and a persis-
tent predator. Piscivorous perch change from an active to a sit-and-wait
foraging mode when their prey shift from an open to a refuge-rich habitat
(Eklov and Diehl, 1994). Of conceptual importance in the study of vigilance
are observations that predators avoid attacking relatively vigilant prey
(FitzGibbon, 1989; Krause and Godin, 1995). O n a different note, piscivo-
rous pike may defecate away from their feeding areas so as to avoid being
detected chemically by prey (Brown et al., 1995). Recent attempts to model
multi-trophic-level games of habitat selection (Schwinning and Rosenzweig,
1990; Hugie and Dill, 1994; Sih, 1998) provide notable instances in which
the crux of the matter is the real-time interaction between prey response
and predator reaction.
264 STEVEN L. LIMA
D. SCALING
TO THE REALWORLD
To what extent d o small-scale laboratory microcosms simulate the situa-

tion faced by animals avoiding predators in their natural environment?
Lima and Dill (1990) raised this question with regard to the common
experimental situation in which predator and prey are maintained in very
close proximity. Under such situations, the prey’s response to predators
may be so strong as to be potentially misleading. Richardson and Brown
(1992) report just such a situation in which a strong response by snails to
nearby crabs in the laboratory could not be replicated under field conditions.
Similarly, the relatively brief reduction in gerbil activity following an en-
counter with an owl under semifield conditions (Abramsky et al., 1996) did
not reflect the marked reduction in gerbil activity following an exposure
to captive owls at close quarters (Kotler, 1992). Perhaps the application of
most laboratory studies to the real world would not be so problematic, but
I nevertheless urge caution in the use of experimental protocols in which
prey and predator are in close proximity. Such caution may also be war-
ranted in light of the possibility that decision making may be impaired by an
abnormally intense physiological stress response under these circumstances
(see Mesa et al., 1994).
The general issue of “scaling to the real world” concerns not just the
spatial proximity of predator and prey, but also the temporal scale of the
interaction. Many studies demonstrate that animals respond markedly to
a brief but acute exposure to predators, perhaps with a complete cessation
of feeding. In effect, these animals are able to “ride out” a short period
of high risk. However, such strong responses may not be indicative of those
to a chronic exposure to high risk; animals must eventually eat.
X. CONCLUSIONS
A N D SUMMARY
Recent years have witnessed increasing interest in the study of antipreda-

tory decision making and its consequences. This recent work is much too
vast to summarize in detail, but some notable recent advances include clear
demonstrations that antipredatory decision making (1) may influence many
aspects of reproductive behavior, (2) has demonstrable long-term conse-
quences for individual fitness, and (3) may influence the nature of ecological
systems themselves. There have also been many advances in the theory of
antipredator behavior, which should provide a sound conceptual basis for
further progress. Overall, combined with earlier work (Sih, 1987; Lima and
Dill, 1990), these recent advances lead to the inescapable conclusion that
the risk of predation may influence any aspect of animal decision making.
Just about all of the areas covered in this review deserve more attention.
This is particularly true of areas that have emerged most recently. In this
regard, of great value would be further work on the effects that predator and
prey have on the other’s behavioral decisions. The range of reproductive
behaviors influenced by the risk of predation also requires much more
investigation. Work on the long-term costs of antipredator decision making
needs more empirical documentation and greater taxonomic diversity.
Work on the ecological implications of antipredatory decision making has
only “scratched the surface,” especially with regard to population-level
effects and species interactions. Theoretical investigations should also play
a prominent role in future work. While I am not sanguine about the possibili-
ties that such theoretical models can be tested quantitatively, theory is
nevertheless essential to the continued conceptual development of the field.
Finally, I suspect that research exploring the link between antipredator
decision making and the physiological stress response will prove rewarding.
What are the next “big steps” in the study of decision making under the
risk of predation? Two areas seem to have particularly good prospects.
The first concerns the aforementioned application of antipredator decision
making to the understanding of ecological systems. Such work will be
particularly interesting given that the early development of behavioral
ecology was spurred (in part) by the prospect that behavioral studies might
provide key insights into the workings of ecological systems; this prospect
may well be realized in the study of predator-prey interactions. The second
area concerns the development of a view of antipredator decision making
that encompasses phenomena expressed over both ecological and evolu-
tionary time. Work in this area promises to integrate the study of antipreda-
tor decision making with recent advances in the larger field of evolutionary
biology. I have not been able to cover this emerging area to any great extent,
but Sih (1992b) and McPeek etal. (1996) provide thoughtful discussions and
examples of how such an integration might proceed.
Acknowledgments
I thank Peter Slater. Manfred Milinski, and Anders M d l e r for their comments on the
manuscript, and their efforts regarding this volume on stress and behavior. Peter Bednekoff
and Patrick Zollner also commented on the manuscript. Chris Mathews provided competent
assistance with the literature search. Hilary Philpot helped in the preparation of the References
section. Finally, much of the background work in preparing this review was made possible
by a sabbatical leave granted by Indiana State University, for which I am most grateful.
References
Abrahams, M. V. (1995). The interaction between antipredator behaviour and antipredator

morphology: Experiments with fathead minnows and brook sticklebacks. Can. J. Zool.
73,2209-2215.
266 STEVEN L. LIMA
Abraham, M. V., and Dill, L. M. (1989). A determination of the energetic equivalence of

the risk of predation. Ecology 70, 999-1007.
Abrams, P. A. (1991). Life history and the relationship between food availability and foraging
effort. Ecology 72, 1242-1252.
Abrams, P. A. (1992). Predators that benefit prey and prey that harm predators: Unusual
effects of interacting foraging adaptations. Am. Nat 140,573-600.
Abrams, P. A. (1993a). Optimal traits when there are several costs: The interaction of mortality
and energy costs in determining foraging behavior. Behav. Ecol. 4, 246-253.
Abrams, P. A. (1993b). Why predation rate should not be proportional to predator density.
Ecology 73,726-733.
Abrams, P. A. (1994). Should prey overestimate the risk of predation? Am. Nat. 144,317-328.
Abrams, P. A. (1995). Implications of dynamically variable traits for identifying, classifying. and
measuring direct and indirect effects in ecological communities. A m . Nat. 146, 112-134.
Abrams, P., and Matsuda, H. (1993). Effects of adaptive predatory and anti-predator behaviour
in a two-prey-one-predator system. Evol. Ecol. 7, 312-326.
Abramsky. Z . , Strauss, E., Subach, A,, Kotler, B. P., and Riechman, A. (1996). The effect of
barn owls (Tyto alba) on the activity and microhabitat selection of Gerbillus allenbyi and
G. pyramidum. Oecologia 105, 313-319.
Ader, R., Felten, D., and Cohen, N., Eds. (1991). “Psychoneuroimmunology.” Academic
Aksnes, D. L., and Giske, J. (1990). Habitat profitability in pelagic environments. Mar. Ecol.:
Prog. Ser. 64,209-215.
Alexander, J . E., Jr., and Covich, A. P. (1991a). Predator avoidance by the freshwater snail
Physella virgata in response to the crayfish Procambarus simulans. Oecologia 87,435-442.
Alexander, J. E., Jr., and Covich, A. P. (1991b). Predation risk and avoidance behavior in
two freshwater snails. Biol. Bull. (Woods Hole, Mass.) 180, 387-393.
Allan, J . D. (1978). Trout predation and the size composition of stream drift. Limnol. Oceanogr.
23, 1231-1237.
Alvarez, F. (1993). Alertness signalling in two rail species. Anim. Behav. 46, 1229-1231.
Andersen, T. H., Friberg, N., Hansen, H. 0.. Iverson, T. M., Jacobsen, D., and Krojgaard,
L. (1993). The effects of introduction of brown trout (Salmo trutta L.) on Gammarus
pulex L. drift and density in two fishless Danish streams. Arch. Hydrobiol. 126,361-371.
Anderson, K. G., Bronmark. C.. Herrmann, J., Malmqvist, B.. Otto, C., and Sjorstrom, P.
(1986). Presence of sculpins (Conus gobio) reduces drift and activity of Gammarus pulex
(Amphipoda). Hydrobiologia 133, 209-215.
Anholt. B. R., and Werner, E. E. (1995). Interaction between food availability and predation
mortality mediated by adaptive behavior. Ecology 76, 2230-2234.
Ashley, E. J.. Kats. L. B., and Wolfe. J. W. (1993). Balancing trade-offs between risk and
changing shoal size in northern red-belly dace (Phoxinus eos). Copeia, pp. 540-542.
Axelrod, R., and Hamilton, W. D. (1981). The evolution of cooperation. Science 211, 1390-
1396.
Azevedo-Ramos, C., Van Sluys, M., Hero, J.-M., and Magnusson, W. E. (1992). Influence of
tadpole movement on predation by Odonate naiads. J. Herpetol. 26, 335-338.
Bachman, G. C. (1993). The effect of body condition of the trade-off between vigilance and
foraging in Belding’s ground squirrels. Anim. Behav. 46, 233-244.
Baker, R. L., and Ball, S. L. (1995). Microhabitat selection by larval Chironomus tentans
(Diptera: Chironomidae): Effects of predators, food, cover, and light. Freshwater Biol.
34, 101-106.
Ball, S. L., and Baker, R. L. (1995). The non-lethal effects of predators and the influence of
food availability on life history of adult Chironomus tentans (Diptera: Chironomidae).
Freshwater Biol. 34, 1-12.
Ball. S. L.. and Baker, R. L. (1996). Predator-induced life history changes: Antipredator
behavior costs or facultative life history shifts? Ecology 77, 1116-1124.
Bednekoff, P. A. (1997). Mutualism among safe, selfish sentinels: A dynamic game. Am. Nut.,
(in press).
Bednekoff, P. A., and Houston, A. 1. (1994). Avian daily foraging patterns: Effects of digestive
constraints and variability. Evol. Ecol. 8, 36-52.
Bercovitch, F. B., Hauser. M. D.. and Jones, J. H. (1995). The endocrine stress response and
alarm vocalizations in rhesus macaques. Anim. Behav. 49, 1703-1706.
Berger. J. (1991). Pregnancy incentives, predation constraints and habitat shifts: Experimental
and field evidence for wild bighorn sheep. Anim. Behuv. 41, 61-77.
Berglund, A. (1993). Risky sex: Male pipefishes mate at random in the presence of a predator.
Anim. Behuv. 46, 169-175.
Blake, M. A,, and Hart, P. J. B. (1993). The behavioural responses of juvenile signal crayfish
Pacifasfacrrs lenircsciilus to stimuli from perch and eels. Freshwater Biol. 29, 89-97.
Blake, M.. Nystrom, P., and Hart, P. J. B. (1994). The effect of weed cover on juvenile signal
crayfish (Pucifustucus leniusculus Dana) exposed to adult crayfish and non-predatory fish.
Ann. 2001.Fenn. 31,297-306.
Blanchard. R. J., and Blanchard. D. C. (1989). Antipredator defensive behaviors in a visible
burrow system. J. Comp. Psychol. 103, 70-82.
Bland, J. D., and Temple, S. A. (1990). Effects of predation-risk on habitat use by Himalayan
snowcocks. Oecologiu 82, 187-191.
Blois-Heulin, C., Crowley, P. H., Arrington, M.. and Johnson, D. M. (1990). Direct and indirect
effects of predators on the dominant invertebrates of two freshwater littoral communities.
Oecologiu 84, 295-306.
Blumstein, D., and Bouskila, A. (1996). Assessment and decision-making in animals: A mecha-
nistic model underlying behavioral flexibility can prevent ambiguity. Oikos 77,569-576.
Boesch, C. (1991). The effects of leopard predation on grouping patterns in forest chimpanzees.
Behuviour 117, 220-241.
Boissy. A. (1995). Fear and fearfulness in animals. Q. Rev. Biol. 70, 165-191.
Bollens, S. M., and Frost, B. W. (1989a). Zooplanktivorous fish and variable die1 vertical
migration in the marine planktonic copepod Culanus pacificus. Limnol. Oceunogr. 34,
1072-1083.
Bollens, S. M., and Frost, B. W. (1989b). Predator-induced die1 vertical migration in a plank-
tonic copepod. J . Plankton Res. 11, 1047-1065.
Bollens, S. M., and Frost, B. W. (1991). Ovigerity, selective predation, and variable die1 vertical
migration in Euchaefa elongura (Copepoda: Calanoida). Oecologiu 87, 155-161.
Bonenfant, M., and Kramer, D. L. (1996). The influence of distance to burrow on flight
initiation distance in the woodchuck, Marmotu monux. Behuv. Ecol. 7, 299-303.
Bouskila, A. (1995). Interactions between predation risk and competition: A field study of
kangaroo rats and snakes. Ecology 76, 165-178.
Bouskila. A,, and Blumstein, D. T. (1992). Rules of thumb for predation hazard assessment:
Predictions from a dynamic model. Am. Nut. 139, 161-176.
Bowers, M. A. (1990). Exploitation of seed aggregates by merriam’s kangaroo rat: Harvesting
rates and predatory risk. Ecology 71, 2334-2344.
Bowers. M. A., and Dooley, J. L., Jr. (1993). Predation hazard and seed removal by small
mammals: Microhabitat versus patch scale effects. Oecologiu 94, 247-254.
Bowers, M. A,. Jefferson, J . L.. and Kuebler, M. G. (1993). Variation in giving-up densities
of foraging chipmunks (Tumias sfriatus) and squirrels (Sciurus carolinensis). Oikos 66,
229-236.
268 STEVEN L. LIMA
Bozinovic, F., and Simonetti, J . A. (1992). Thermoregulatory constraints on the microhabitat

use by cricetid rodents in central Chile. Mammalia 56, 363-369.
Brabrand, A,, and Faafeng, B. (1993). Habitat shift in roach (Rutilus rutihts) induced by
pikeperch (Stizostedion htcioperca) introduction: Predation risk versus pelagic behaviour.
Oecologia 95, 38-46.
Braiia, F. (1993). Shifts in body temperature and escape behaviour of female Podarcis muralis
during pregnancy. Oikos 66, 216-222.
Briggs, S. E., Godin, J.-G., J., and Dugatkin, L. A. (1996). Mate-choice copying under predation
risk in the Trinidadian guppy (Poecilia reiiculaia). Behav. Ecol. 7, 151-157.
Brown. G. E., Chivers, D. P., and Smith, R. J. F. (1995). Localized defection by pike: A
response to labelling by cyprinid alarm pheromone? Behav. Ecol. Sociohiol. 36, 105-1 10.
Brown, J. S. (1989). Desert rodent community structure: A test of four mechanisms of coexis-
tence. Ecol. Monogr. 59, 1-20.
Brown, J. S. (1992). Patch use under predation risk I. Models and predictions. Ann. 2001.
Fenn. 29, 301-309.
Brown, J. S., and Alkon. P. U. (1990). Testing values of crested porcupine habitats by experi-
mental food patches. Oecologia 83, 512-518.
Brown, J. S., and Morgan, R. A. (1995). Effects of foraging behavior and spatial scale on diet
selectivity: A test with fox squirrels. Oikos 74, 122-136.
Brown, J. S., and Vincent. T. L. (1992). Organization of predator-prey communities as an
evolutionary game. Evoluiion (Lawrence, Kans.) 46, 1269-1283.
Brown, J. S., Morgan, R. A,, Dow, B. D. (1992a). Patch use under predation risk: 11. A test
with fox squirrels, Sciurus niger. Ann. Zoo/. Fenn. 29, 311-318.
Brown, J. S., Arel, Y.,Abramsky, 2..and Kotler, B. P. (1992b). Patch use by gerbils (Gerhillus
allenhyi) in sandy and rocky habitats. J. Mammal. 73, 821-829.
Brown. J. S., Kotler, B. P., and Mitchell, W. A. (1994). Foraging theory. patch use, and the
structure of a Negev Desert granivore community. Ecology 75, 2286-2300.
Brown, L. R.. and Moyle, P. B. (1991). Changes in habitat and microhabitat partitioning
within an assemblage of stream fishes in response to predation by Sacramento squawfish
(Ptychocheihcs grandis). Can. J. Fish. Aquat. Sci. 48, 849-856.
Bull, C. D., Metcalfe, N. B., and Mangel, M. (1996). Seasonal matching of foraging to antici-
pated energy requirements in anorexic juvenile salmon. Proc. R. Soc. London, Ser. B
236, 13-18.
Bulova, S. J. (1 994). Ecological correlates of population and individual variation in antipredator
behavior of two species of desert lizards. Copeia, pp. 980-992.
Burrows. M. T., and Gibson, R. N. (1995). The effects of food, predation risk and endogenous
rhythmicity on the behaviour of juvenile plaice, Pleuronectes plaiessa L. Anim. Behav.
50,41-52.
Burrows, M. T.. and Hughes. R. N. (1991). Optimal foraging decisions by dogwhelks, Nucella
lapillits (L.): Influences of mortality risk and rate-constrained digestion. Funci. Ecol.
5,461-475.
Butler, J. M.. and Roper, T. J. (1994). Escape tactics and alarm responses in badgers Meles
melest A field experiment. Ethology 99, 313-322.
Caro. T. M. (1994). Ungulate antipredator behaviour: Preliminary and comparative data from
African bovids. Behaviour 128, 189-228.
Caro, T. M. (1995). Pursuit-deterrence revisited. Trends Ecol. Evol. 10, 500-503.
Caro. T. M.. Lombardo, L., Goldizen, A. W., and Kelly, M. (1995). Tail-flagging and other
antipredator signals in white-tailed deer: New data and synthesis. Behav. Ecol. 6,442-450.
Cartar, R. V. (1991). Colony energy requirements affect response to predation risk in foraging
bumble bees. Eihology 87, 90-96.
Cassini. M. H. (1991). Foraging under predation risk in the wild guinea pig Cavia aperea.
Oikos 62, 20-24.
Cassini, M. H.. and Galante. M . L. (1992). Foraging under predation risk in the wild guinea
pig: The effect of vegetation height on habitat utilization. Ann. Zool. Fenn. 29,285-290.
Chesson, P., and Rosenweig. M. (1991). Behavior, heterogeneity, and the dynamics of interact-
ing species. Ecology 72, 1187-1195.
Chivers, D. P., and Smith, R. J. F. (1994). Fathead minnows, Pimephales promelas, acquire
predator recognition when alarm substance is associated with the sight of unfamiliar fish.
Anim. Behav. 48, 597-605.
Chivers, D. P., and Smith, R. J. F. (1995). Free-living fathead minnows rapidly learn to
recognize pike as predators. J . Fish Biol. 46, 949-954.
Chivers, D. P., Wisendon, B. D., and Smith. R. J. F. (1995a). The role of experience in the
response of fathead minnows (Pimephales promelas) to skin extract of Iowa darters
(E/heos/oma exile). Behavioirr 132, 665-674.
Chivers. D. P., Brown. G. E.. and Smith, R. J. (1995b). Familiarity and shoal cohesion in
fathead minnows (Pimephales promelas): Implications for antipredator behaviour. Can.
J. Zoo/. 73, 955-960.
Chivers, D. P.. Wisenden, B. D.. and Smith. R. J. F. (1995~).Predation risk influences reproduc-
tive behaviour of Iowa darters, E/heos/oma exile (Osteichthyes, Percidae). E/hology
99,278-285.
Chovanec, A. (1992). The influence of tadpole swimming behaviour on predation by dragonfly
nymphs. Amphibin-Rrplilia 13, 341-349.
Christensen. B., and Persson. L. (1993). Species-specific antipredatory behaviours: Effects on
prey choice in different habitats. Behav. Ecol. Sociobiol. 32, 1-9.
Clark, C. W. (1994). Antipredator behavior and the asset-protection principle. Behav. Ecol.
5, 159-170.
Clark, C. W.. and Dukas. R. (1994). Balancing foraging and antipredator demands: An advan-
tage of sociality. Am. Nat. 144, 542-548.
Clark, C. W., and Levy, D. A. (1988). Die1 vertical migrations by juvenile sockeye salmon
and the antipredation window. A m . Nut. 131, 271-290.
Clark, C. W., and Ydenberg, R. C. (1990a). The risks of parenthood. I. General theory and
applications. Evol. Ecol. 4, 21-34.
Clark, C. W., and Ydenberg. R. C. (1990b). The risks of parenthood 11. Parent-offspring
conflict. Evol. Ecol. 4, 312-325.
Cowan, C. A., and Peckarsky, B. L. (1994). Die1 feeding and positioning periodicity of a
grazing mayfly in a trout stream and a fishless stream. Can. J. Fish. Aqua/. Sci. 51,450-459.
Cowlishaw, G. (1994). Vulnerability to predation in baboon populations. Behaviour 131,
293-304.
Crawley. M. J. (1992). Population dynamics of natural enemies and their prey. In “Natural
Enemies: The Population Biology of Predators. Parasites. and Diseases” (M. J. Crawley.
ed.), pp. 40-89. Blackwell. Oxford.
Cresswell. W. (1994a). Flocking is an effective anti-predation strategy in redshanks, Tringa
/o/aiii~s.Anim. Behav. 47, 433-442.
Cresswell. W. ( 1994b). Age-dependent choice of redshank (Tringa /o/ani~s) feeding location:
Profitability or risk’?J . Anim. Ecol. 63, 589-600.
Crowl. T. A., and Covich. A. P. (1994). Response of a freshwater shrimp to chemical and
tactile stimuli from a large decapod predator. J. North Am. Benthol. Soc. 13,291-298.
Crowley, P. H., and Hopper, K. R. (1994). How to behave around cannibals: A density-
dependant dynamic game. A m . Na/. 143, 117-154.
270 STEVEN L. LIMA
Crowley. P. H.. Travers, S. E.. Linton, M. C., Cohn, S. L., Sih, A., and Sargent. R. C. (1991).
Mate density, predation risk, and the seasonal sequence of mate choices: A dynamic
game. Am. Nut. 137,567-596.
Croy. M. I., and Hughes, R. N. (1991). Effects of food supply, hunger, danger and competition
on choice of foraging location by the fifteen-spined stickleback. Spinachia spinachia L.
Anim. Behav. 42, 131-139.
Culp, J. M., and Scrimgeour, G. J. (1993). Size-dependent die1 foraging periodicity of a mayfly
grazer in streams with and without fish. Oikos 68,242-250.
Culp, J. M., Glozier, N. E., and Scrimgeour, G. J. (1991). Reduction of predation risk under
the cover of darkness: Avoidance responses of mayfly larvae to benthic fish. Oecologia
86, 163-169.
Curio, E. (1993). Proximate and developmental aspects of antipredator behavior. Adv. Study
Behav. 22, 135-238.
Daly. M., Wilson, M., Behrends. P. R., and Jacobs, L. F. (1990). Characteristics of kangaroo
rats, Dipodomys merriami, associated with differential predation risk. Anim. Behav.
40, 380-389.
Daly, M., Behrends, P. R., Wilson, M. I., and Jacobs, L. F. (1992). Behavioural modulations
of predation risk: Moonlight avoidance and crepuscular compensation in a nocturnal
desert rodent. Anim. Behav. 44, 1-9.
Dawidowicz, P., and Loose, C. J. (1992). Metabolic costs during predator-induced die1 vertical
migration of Daphnia. Limnol. Oceanogr. 37, 1589-1595.
Dawidowicz, P., Pijanowska, J., and Ciechomski, K. (1990). Vertical migration of Chaohorus
larvae is induced by the presence of fish. Limnol. Oceanogr. 35, 1631-1637.
De Meester, L. (1993). Genotype, fish-mediated chemicals, and phototactic behavior in Daph-
nia magna. Ecology 74, 1467-1474.
De Meester, L., Weider, L. J., and Tollrian, R. (1995). Alternative antipredator defences and
genetic polymorphism in a pelagic predator-prey system. Nature (London) 378,483-485.
Desportes, J.-P., Metcalfe, N. B., Cezilly, F.. Lauvergeon, G., and Kervella. C. (1989). Tests
of the sequential randomness of vigilance behavior using spectral analysis. Anim. Behav.
38,771-777.
Desy, E. A,. Batzli, G. 0..and Liu, J. (1990). Effects of food and predation on behaviour of
prairie voles: A field experiment. Oikos 58, 159-168.
Devenport, J. A. (1989). Social influences on foraging in black-tailed prairie dogs. J. Mammal.
70, 166-168.
DeVries, D. R. (1990). Habitat use by bluegill in laboratory pools: Where is the refuge when
macrophytes are sparse and alternative prey are present? Environ. B i d . Fishes 29,27-34.
Dickman. C. R. (1992). Predation and habitat shift in the house mouse, Miis domesticus.
Ecology 73, 313-322.
Diehl, S., and Eklov, P. (1995). Effects of piscivore-mediated habitat use on resources, diet,
and growth of perch. Ecology 76, 1712-1726.
Dill, L. M. (1987). Animal decision making and its ecological consequences: The future of
aquatic ecology and behaviour. Can. J. Zool. 65, 803-811.
Dill, L. M. (1990). Distance-to-cover and the escape decisions of an African cichlid fish,
Melanochromis chipokae. Environ. Biol. Fishes 27, 147-152.
Dill. L. M., and Fraser. A. H. G. (1984). Risk of predation and the feeding behavior of juvenile
coho salmon (Oncorhynchus kisutch). Behav. Ecol. Sociohiol. 16,65-71.
Dill. L. M.. and Gillett, J. F. (1991). The economic logic of barnacle Balanus glandula (Darwin)
hiding behavior. J. Exp. Mar. B i d . Ecol. 153, 115-127.
Dill, L. M.. and Houtman. R. (1989). The influence of distance to refuge on flight-initiation
distance in the gray squirrel (Sciurus carolinensis). Can. J. Zool. 67, 232-235.
PREDATOR-INDUCED STRESS AND BEHAVIOR 27 1
Dill, L. M.. Fraser, A. H. G., and Roitberg, B. D. (1990). The economics of escape behaviour
in the pea aphid, Acyrthosiphon pisum. Oecologia 83, 473-478.
Dini, M. L., and Carpenter, S. R. (1988). Variability in Daphnia behavior following fish
community manipulations. J. Plankton Res. 10, 621-635.
Dini, M. L., and Carpenter, S . R. (1992). Fish predators, food availability and diel vertical
migration in Daphnia. J. Plankton Res. 14, 359-377.
Dini, M. L., Soranno, P. A,, Scheuerell, M.. and Carpenter, S . R. (1993). Effects of predators
and food supply on diel vertical migration of Daphnia. In “The Trophic Cascade in
Lakes” ( S . R. Carpenter and J. F. Kitchell, eds.), pp. 153-171. Cambridge University
Press, Cambridge, UK.
Dixon, S . M., and Baker. R. L. (1988). Effect of size on predation risk, behavioral response
to fish, and cost of reduced feeding in larval Ischnura verticalis (Coenagrionidae: Odonata).
Oecologia (Berlin) 76, 200-205.
Dodson, S. (1988). The ecological role of chemical stimuli for the zooplankton: Predator
avoidance behavior in Daphnia. Limnol. Oceanogr. 33, 1431-1439.
Dodson, S . (1990). Predicting diel vertical migration of zooplankton. Limnol. Oceanogr.
35,1195-1200.
Dodson, S. I., Crowl, T. A,, Peckarsky, B. L., Kats, L. B., Covich, A. P., and Culp, J. M.
(1994). Non-visual communication in freshwater benthos: On overview. J. North Am.
Benthol. SOC. 13, 268-282.
Doncaster, C. P. (1993). Influence of predation threat on foraging pattern: The hedgehog’s
gambit. Rev. Ecol. 48, 207-213.
Doncaster. C. P. (1994). Factors regulating local variations in abundance: Field tests on
hedgehogs, Erinaceus europaeus. Oikos 69, 182-192.
Douglas, P. L., Forrester. G. E., and Cooper, S. D. (1994). Effects of trout on the diel
periodicity of drifting in baetid mayflies. Oecologia 98, 48-56.
Dugatkin, L. A. (1988). Do guppies play tit for tat during predator inspection visits? Behav.
Ecol. Sociobiol. 25, 395-399.
Dugatkin, L. A. (1992). Tendency to inspect predators predicts mortality risk in the guppy
(Poecilia reticulata). Behav. Ecol. 3, 124-127.
Dugatkin, L. A,, and Alfieri, M. (1991a). Guppies and the tit-for-tat strategy: Preference
based on past interaction. Behav. Ecol. Sociobiol. 28, 243-246.
Dugatkin, L. A,, and Alfieri, M. (1991b). Tit-for-tat in guppies (Poecilia reticulata): The
relative nature of cooperation and defection during predator inspection. Evol. Ecol.
5,300-309.
Dugatkin, L. A,, and Godin, J.-G. J. (1992a). Prey approaching predators: A cost-benefit
perspective. Ann. Zool. Fenn. 29, 233-252.
Dugatkin, L. A., and Godin, J . 4 . J. (1992b). Predator inspection, shoaling, and foraging
under predation hazard in the Trinidadian guppy, Poecilia reticulata. Environ. Biol. Fishes
34,265-276.
Dukas, R., and Clark, C. W. (1995). Sustained vigilance and animal performance. Anim.
Behav. 49, 1259-1267.
Duvall, C. J., and Williams, D. D. (1995). Individuality in the growth of stonefly nymphs in
response to stress from a predator. Arch. Hydrobiol. 133, 273-286.
Edmunds, M. (1974). “Defence in Animals.” Longman, New York.
Eklov, P., and Diehl, S. (1994). Piscivore efficiency and refuging prey: The importance of
predator search mode. Oecologia 98, 344-353.
Eklov, P.. and Hamrin, S. F. (1989). Predatory efficiency and prey selection: Interactions
between pike Esox lucius, perch Perca puviatilis and rudd Scardinus erythrophthalmus.
Oikos 56, 149-156.
272 STEVEN L. LIMA
Eklov, P.. and Persson, L. (1995). Species-specific antipredator capacities and prey refuges:
Interactions between piscivorous perch (Percu ,puviatilis) and juvenile perch and roach
(Rutilics rittilus). Behav. Ecol. Sociobiol. 37, 169- 178.
EklBv. P., and Persson. L. (1996). The response of prey to the risk of predation: Proximate
cues for refuging juvenile fish. Anim. Behav. 51, 105-115.
Elgar, M. A. (1989). Predator vigilance and group size in mammals and birds: A critical review
of the empirical evidence. B i d . Rev. Cambridge Philos. Soc. 64, 13-33.
Elliott, S. A,, Kats, L. B., and Breeding. J. A. (1993). The use of conspecific chemical cues
for cannibal avoidance in California newts (Taricha torosa). Ethology 95, 186-192.
Endler. J. A. (1987). Predation. light intensity, and courtship behaviour in Poecilio retictilafa
(Pisces: Poeciliidae). Anim. Behav. 35, 1376- 1385.
Endler. J. A. (1991). Interactions between predators and prey. In ”Behavioral Ecology: An
Evolutionary Approach” ( J . R. Krebs and N. B. Davies, eds.). 3rd ed., pp. 169-201.
Blackwell, Oxford.
Everett. R. A.. and Ruiz. G. M. (1993). Coarse woody debris as a refuge from predation in
aquatic communities: An experimental test. Oecologia 93, 475-486.
Fairbairn, D.J. (1993). Costs of loading associated with mate-carrying in the waterstrider,
Aquarius reniigis. Behav. Ecol. 4, 224-231.
Feltmate, B. W.. and Williams. D. D. (1989a). Influence of rainbow trout (Oncorhynchiis
mykiss) on density and feeding behaviour of a perlid stonefly. Can. J. Fish. Aquat. Sci.
46, 1575-1580.
Feltmate, B. W., and Williams, D. D. (1989b). A test of crypsis and predator avoidance in
the stonefly Paragnetitin media (Plecoptera: Perlidae). Anirn. Behav. 37, 992-999.
Feltmate, B. W.. and Williams. D. D. (1991). Evaluation of predator-induced stress on field
population of stoneflies (Plecoptera). Ecology 72, 1800-1806.
Feltmate, B. W., Williams. D. D.. and Montgomerie. A. (1992). Relationship between diurnal
activity patterns, cryptic coloration. and subsequent avoidance of predaceous fish by
perlid stoneflies. Can. J. Fish. Aquat. Sci. 49, 2630-2634.
Feminella, J. W., and Hawkins, C. P. (1994). Tailed frog tadpoles differentially alter their
feeding behavior in response to non-visual cues from four predators. J. North Am. Benthol.
Soc. 13, 310-320.
Fenn, M. G. P., and Macdonald, D. W. (1995). Use of middens by red foxes: Risk reverses
rhythms of rats. J. Mammal. 76, 130-136.
Figiel. C.R.. Jr.. and Semlitsch, R. D. (1990). Population variation in survial and metamorphosis
of larval salamanders (Ambysroma nzaculaturn) in the presence and absence of fish
predation. Copeia, pp. 818-826.
Fiksen, 0.. and Giske, J. (1995). Vertical distribution and population dynamics of copepods
by dynamic optimization. ICES J. Mur. Sci. 52, 483-503.
FitzCibbon, C. D. (1989). A cost to individuals with reduced vigilance in groups of Thomson’s
gazelles hunted by cheetahs. Anim. Behav. 37, 508-510.
FitzGibbon. C.D. (1990). Anti-predator strategies of immature Thomson’s gazelles: Hiding
and the prone response. Anim. Behuv. 40, 846-855.
FitzGibbon, C. D. (1994).The costs and benefits of predator inspection behaviour in Thomson’s
gazelles. Behrtv. Ecol. Sociobiol. 34, 139-148.
Flecker, A. S. (1992). Fish predation and the evolution of invertebrate drift periodicity:
Evidence from neotropical streams. Ecology 73, 438-448.
Forrnanowicz, D.R., Jr.. and Bobka, M. S. (1989). Predation risk and microhabitat preference:
An experimental study of the behavioral responses of prey and predator. Am. Mid/. Nut.
121,379-386.
Forrester, G. E. (1994a). Influences of predatory fish on the drift dispersal and local density
of stream insects. Ecology 75, 1208-1218.
Forrester, G. E. (1994b). Die1 patterns of drift by five species of mayfly at different levels of
fish predation. Can. J . Fish. Aqrmf. Sci. 51, 2549-2557.
Forsgren. E. (1992). Predation risk affects mate choice in a gobiid fish. A m . Nat. 140,1041-1049.
Forsgren. E.. and Magnhagen, C. (1993). Conflicting demands in sand gobies: Predators
influence reproductive behaviour. Behaviour 126, 125-135.
Fraser, D. F., and Gilliam, J. F. (1992). Nonlethal impacts of predator invasion: Facultative
suppression of growth and reproduction. Ecology 73, 959-970.
Fraser. D. F.. Gilliam. J. F., and Yip-Hoi. T. (1995). Predation as an agent of population
fragmentation in a tropical watershed. Ecology 76, 1461-1472.
Frost, B. W., and Bollens. S. M. (1992). Variability of die1 vertical migration in the marine
plankton copepod Pseitdocalonus newmani in relation to its predators. Can. J. Fish. Aquat.
Sci. 49, 1137-1 141.
Fuller, R., and Berglund. A. (1996). Behavioral responses of a sex-role reversed pipefish to
a gradient of perceived predation risk. Behav. Ecol. 7, 69-75.
Gabriel, W., and Thomas, B. (1988). Vertical migration of zooplankton as an evolutionarily
stable strategy. A m . Nut. 132, 199-216.
Gelowitz. C. M., Mathis, A,. and Smith, R. J . F. (1993). Chemosensory recognition of northern
pike (Esox lucius) by brook stickleback (Culaea inconstans): Population differences and
the influence of predator diet. Behaviour 127, 105-1 18.
Gilliam, J. F. (1982). Habitat use and competitive bottlenecks in size-structured fish popula-
tions. Ph.D. Dissertation, Michigan State University, East Lansing.
Gilliam. J. F. (1990). Hunting by the hunted: optimal prey selection by foragers under predation
hazard. N A T O AS1 Ser., Ser. G 20, 797-818.
Gilliam, J. F.. and Fraser, D. F. (1987). Habitat selection when foraging under predation
hazard: A model and a test with stream-dwelling minnows. Ecology 68, 1856-1862.
Gilliam, J. F.. and Fraser, D. F. (1988). Resource depletion and habitat segregation by competi-
tors under predation hazard. In “Size-Structured Populations” (B. Ebenman and L.
Persson. eds.), pp. 173-184. Springer-Verlag. Berlin.
Gliwicz, Z. M. (1986). A lunar cycle in zooplankton. Ecology 67, 883-897.
Gliwicz, 2. M. (1994). Retarded growth of cladoceran zooplankton in the presence of a
copepod predator. Oecologia 97,458-461.
Gliwicz. Z . M., and Jachner. A. (1992). Die1 migrations of juvenile fish: A ghost of predation
past or present? Arch. Hydrobiol. 124, 385-410.
Gliwicz, Z . M.. and Pijanowska, J. (1988). Effect of predation and resource depth distribution
on vertical migration of zooplankton. Bull. Mar. Sci. 43, 695-709.
Godin. J.-G. J. (1990). Diet selection under the risk of predation. N A T O AS1 Ser., Ser. C
20,739-769.
Godin. J.-G. J. (1995). Predation risk and alternative mating tactics in male Trinidadian
guppies (Poecilia reticulafa).Oecologia 103, 224-229.
Godin, J.-G. J., and Briggs, S. E. (1996). Female mate choice under predation risk in the
guppy. Anim. Behav. 51, 117-130.
Godin. J.-G. J., and Crossman. S. L. (1994). Hunger-dependent predator inspection and
foraging behaviours in the threespine stickleback ( Gasterosfeus aculeatus) under predation
risk. Behav. Ecol. Sociobiol. 34, 359-366.
Godin, J . 4 . J., and Davis, S. A. (1995). Who dares, benefits: Predator approach behaviour
in the guppy (Poecilia reticulata) deters predator pursuit. Proc. R. Soc. London, Ser. B
259, 193-200.
274 STEVEN L. LIMA
Godin, J.-G. J., and Smith, S. A. (1988). A fitness cost of foraging in the guppy. Nature
(London) 333,69-71.
Gotceitas, V. (1990). Foraging and predator avoidance: A test of a patch choice model with
juvenile bluegill sunfish. Oecologia 83, 346-351.
Gotceitas, V., and Brown, J. A. (1993). Substrate selection by juvenile Atlantic cod (Gadus
morhua): Effects of predation risk. Oecologia 93, 31-37.
Gotceitas. V., and Colgan, P. (1990a). Behavioural response of juvenile bluegill sunfish to
variation in predation risk and food level. Ethology 85, 247-255.
Gotceitas, V., and Colgan, P. (1990b). The effects of prey availability and predation risk on
habitat selection by juvenile bluegill sunfish. Copeia, pp. 409-417.
Gotceitas, V., and Godin. J.-G. J. (1991). Foraging under the risk of predation in juvenile
Atlantic salmon (Salmo salar L.): Effects of social status and hunger. Behav. Ecol. Socio-
b i d . 29, 255-261.
Gotceitas, V. and Godin, J . 4 . J. (1993). Effects of aerial and in-stream threat of predation
on foraging by juvenile Atlantic salmon (Salmo sahr). Can. Spec. Publ. Fish. Aquar. Sci.
118, 35-41.
Gregory, R. S. (1993). Effects of turbidity on the predator avoidance behaviour of juvenile
chinook salmon (Oncorhynchus rshawytschu). Can. J. Fish. Aquat. Sci. 50, 241-246.
Grill, C. P., and Juliano, S . A. (1996). Predicting species interactions based on behaviour:
Predation and competition in container-dwelling mosquitoes. J. Anim. Ecol. 65, 63-76.
Grossman, G. D., Ratajczak, R. R., Jr.. and Crawford. M. K. (1995). Do rock bass (Ambloplites
rupestris) induce microhabitat shifts in mottled sculpin (Cottus bairdi)? Copeia, pp.
343-353.
Guisande. C., Duncan, A,, and Lampert, W. (1991). Trade-offs in Daphnia vertical migration
strategies. Oecologia 87, 357-359.
Hager, M. C., and Helfman, G. S. (1991). Safety in numbers: Shoal size choice by minnows
under predatory threat. Behav. Ecol. Sociobiol. 29,271-276.
Hansson, L. (1995). Is the indirect predator effect a special case of generalized reactions to
density-related disturbances in cyclic rodent populations? Ann. Zool. Fenn. 32, 159-162.
Harfenist, A,, and Ydenberg, R. C. (1995). Parental provisioning and predation risk in rhinoc-
eros auklets (Cerorhinca monocerata); Effects on nestling growth and fledging. Behav.
E d 6, 82-86.
Harvey, B. C. (1991). Interactions among stream fishes: Predator-induced habitat shifts and
larval survival. Oecologia 87, 29-36.
Hasson. 0. (1991). Pursuit-deterrant signals: Communication between prey and predator.
Trends Ecol. Evol. 6, 325-329.
Hastings, P. A. (1991). Flexible responses to predators in a marine fish. Ethol. Ecol. Evol.
3, 177-184.
Hayes. F. E. (1989). Antipredator behavior of recently metamorphosed toads (Bufo a. ameri-
canus) during encounters with garter snakes (Thamnophis s. sirtulis). Copeia, pp. 1011-
1015.
Hazlett, B. A. (1994). Alarm responses in the crayfish Orconectes virilis and Orconectes
propinquus. J. Chem. Ecol. 20,1525-1535.
He, X., and Kitchell. J. F. (1990). Direct and indirect effects of predation on a fish community:
A whole-lake experiment. Trans. Am. Fish. Soc. 119, 825-835.
He, X., and Wright, R. A. (1992). An experimental study of piscivore-planktivore interactions:
Population and community responses to predation. Can. J . Fish. Aquat. Sci. 49,1176-1 183.
Hedrick, A. V., and Dill, L. M. (1993). Mate choice by female crickets is influenced by
predation risk. Anim. Behav. 46, 193-196.
Heikkila, J., Kaarsalo, K., Mustonen, O., and Pekkarinen, P. (1993). Influence of predation
risk on early development and maturation in three species of Clethrionomys voles. Ann.
Zool. Fenn. 30, 153-161.
Heinen, J. T. (1993). Substrate choice and predation risk in newly metamorphosed American
toads Bufo americanus: An experimental analysis. A m . Midl. Nat. 130, 184-192.
Heinen, J. T. (1994a). Antipredator behavior of newly metamorphosed American toads (Bufo
a. americanus). and mechanisms of hunting by eastern garter snakes (Thamnophis s.
sirtalis). Herpetologica 50, 137-14.5.
Heinen, J. T. (1994b). The significance of color change in newly metamorphosed American
toads (Bufo a. americanus). J. Herpetol. 28, 87-93.
Helfman. G. S. (1986). Behavioral responses of prey fishes during predator-prey interactions.
In “Predator-Prey Relationships: Perspectives and Approaches from the Study of Lower
Vertebrates” (M. E. Feder and G. V. Lauder, eds.), pp. 135-156. University of Chicago
Press, Chicago.
Helmstetter, F. J., and Fanselow. M. S. (1993). Aversively motivated changes in meal patterns
of rats in a closed economy: The effects of shock density. Anim. Learn. Behav. 21,168-175.
Henrikson, B.-I. (1988). The absence of antipredator behaviour in the larvae of Leucorrhinia
dubia (Odonata) and the consequences for their distribution. Oikos 51, 179-183.
Hershey, A. E. (1987). Tubes and foraging behavior in larval Chironomidae: Implications for
predator avoidance. Oecologia 73, 236-241.
Hews, D. K. (1988). Alarm response in larval western toads, Bufo boreas: Release of larval
chemicals by a natural predator and its effect on predator capture efficiency. Anim.
Behav. 36, 125-133.
Hik, D. S. (1995). Does risk of predation influence population dymamics? Evidence from the
cyclic decline of snowshoe hares. Wild. Res. 22, 115-129.
Hill, A. M., and Lodge, D. M. (1994). Die1 changes in resource demand: Competition and
predation in species replacement among crayfishes. Ecology 75, 21 18-2126.
Hill, A. M., and Lodge, D. M. (1995). Multi-trophic-level impact of sublethal interactions
between bass and omnivorous crayfish. J. North A m . Benthol. SOC. 14,306-314.
Hinsley, S. A,, Bellamy, P. E., and Moss, D. (1995). Sparrowhawk Acciprer nisus predation
and feeding site selection by tits. Ibis 137, 418-428.
Holmes, W. G. (1991). Predator risk affects foraging behaviour of pikas: Observational and
experimental evidence. Anim. Behav. 42, 111-119.
Holomuzki, J. R.. and Hoyle, J. D. (1990). Effect of predatory fish presence on habitat
use and die1 movement of the stream amphipod, Cammarus minus. Freshwater Biol.
24,509-517.
Holomuzki, J. R., and Short, T. M. (1990). Ontogenic shifts in habitat use and activity in a
stream-dwelling isopod. Holarctic Ecol. 13, 300-307.
Horat, P., and Semlitsch, R. D. (1994). Effects of predation risk and hunger on the behaviour
of two species of tadpoles. Behav. Ecol. Sociobiol. 34, 393-401.
Houston, A. I., and McNamara. J. M. (1989). The value of food: Effects of open and closed
economies. Anim. Behav. 37,546-562.
Houston, A. 1.. and McNamara, J. M. (1993). A theoretical investigation of the fat reserves
and mortality levels of small birds in winter. Ornis Scand. 24,205-219.
Houston, A. I., McNamara, J. M., and Hutchinson, J. M. C. (1993). General results concerning
the trade-off between gaining energy and avoiding predation. Philos. Trans. R. SOC.
London, Ser. B 341, 375-397.
Houtman, R., and Dill, L. M. (1994). The influence of substrate color on the alarm response
of tidepool sculpins (Oligocotrus maculosiis; Pisces, Cottidae). Erhology %, 147-154.
276 STEVEN L. LIMA
Huang. C.. and Sih, A. (1990). Experimental studies on behaviorally meditated, indirect
interactions through a shared predator. Ecology 71, 15 15-1522.
Huang. C.. and Sih, A. (1991). Experimental studies on direct and indirect interactions in a
three trophic-level stream system. Oecologia 85, 530-536.
Hughes, J. J., and Ward, D. (1993). Predation risk and distance to cover affect foraging
behaviour in Namib desert gerbils. Anim. Behav. 46, 1243-1245.
Hughes, J . J.. Ward. D., and Perrin, M. R. (1994). Predation risk and competition affect
habitat selection and activity of Namib Desert gerbils. Ecology 75, 1397-1405.
Hugie. D. M., and Dill, L. M. (1Y94). Fish and game: A game theoretic approach to habitat
selection by predators and prey. J. Fish B i d . 45, 151-169.
Huntingford. F. A.. Lazarus, J.. Barrie, B. D.. and Webb. S. (1994). A dynamic analysis of
cooperative predator inspection in sticklebacks. Anim. Behav. 47, 413-423.
Ibrahim. A. A., and Huntingford. F. A. (1989). Laboratory and field studies of the effect of
predation risk on foraging in three-spined sticklebacks (Gasterosfeusaculeatus).Behavioitr
109,46-57.
lves, A. R., and Dobson, A. P. (1987). Antipredator behavior and the population dynamics
of simple predator-prey systems. Am. Nat. 130, 431-447.
Iwasa, Y., Suzuki. Y.. and Matsuda, H. (1984) Theory of oviposition strategy of parasitoids.
I. Effect of mortality and limited egg number. Theor. Popicl. Biol. 26, 205-227.
Jachner, A. (1995a). Chemically-induced habitat shifts in bleak (Alburnus alburnus L.) Arch.
Hydrobiol. 133, 7 1-79.
Jachner, A. (199%). Changes in feeding behavior of bleak (Alburnus alburnus L.) in response
to visual and chemical stimuli from predators. Arch. Hydrobiol. 133, 305-314.
Jakobsen, P. J., Birkeland, K.. and Johnsen. G. H. (1994). Swarm location in zooplankton as
an anti-predator defence mechanism. Anim. Behav. 47, 175-178.
Jedrzejewska, B., and Jedrzejewski. W. (1990). Antipredatory behaviour of bank voles and
prey choice of weasels: Enclosure experiments. Ann. Zool. Fenn. 27, 321-328.
Jedrzejewski. W.. and Jedrzejewska. B. (1990). Effect of a predator’s visit on the spatial
distribution of bank voles: Experiments with weasels. Can. J . Zool. 68, 660-666.
Jedrzejewski, W., Rychlik. L., and Jedrzejewska, 5. (1993). Responses of bank voles to odours
of seven species of predators: Experimental data and their relevance to natural predator-
vole relationships. Oikos 68, 251-257.
Jeffries, M. (1990). Interspecific differences in movement and hunting success in damselfly
larvae (Zygoptera: Insecta): Responses to prey availability and predation threat. Freshwa-
ter Biol. 23, 191-196.
Jennions, M. D., and Blackwell, P. R. Y. (1992). Chorus size influences on the anti-predator
response of a neotropical frog. Anim. Behav. 44,990-992.
Johansson. A., and Englund. G. (1995). A predator-prey game between bullheads and case-
making caddis larvae. Anim. Behav. 50, 785-792.
Johansson, F. (1993). Effects of prey type, prey density and predator presence on behaviour
and predation risk in a larval damselfly. Oikos 68, 481-489.
Juliano, S. A,. Hechtel, L. J., and Waters, J. R. (1993). Behavior and risk of predation in
larval tree hole mosquitoes: Effects of hunger and population history of predation. Oikos
68,229-241.
Kilis, J . A., Fiske. P.. and Szther, S. A. (1995). The effect of mating probability on risk
taking: An experimental study in lekking great snipe. Am. Nut. 146, 59-71.
Kalcounis, M. C., and Brigham. R. M. (1994). Impact of predation risk on emergence by little
brown bats, Myotic lucifzrgiis (Chiroptera: Vespertilionidae), from a maternity colony.
Ethology 98, 201 -209.
Kats, L. B., and Dill, L. M. (1998). The scent of death: Chemosensory assessment of predation
risk by prey animals. Ecoscience (in press).
Kats. L. B.. Petranka, J. W.. and Sih. A. (1988). Antipredator defenses and the persistence
of amphibian larvae with fishes. Ecology 69, 1865-1870.
Kennedy, M.. Shave, C. R., Spencer, H. G.. and Gray. R. D. (1994). Quantifying the effect
of predation risk on foraging bullies: No need to assume an IFD. Ecology 75,2220-2226.
Kohler, S. L.. and McPeek, M. A. (1989). Predation risk and the foraging behavior of competing
stream insects. Ecology 70, 1181-1825.
Koivula. K., Lahti. K., Rytkonen. S., and Orell. M. (1994). Do subordinates expose themselves
to predation? Field experiments on feeding site selection by willow tits. J. Avian Biol.
25, 178-183.
Koivula, K., Rytkonen, S.. and Orell, M. (1995). Hunger-dependency of hiding behaviour
after a predator attack in dominant and subordinate willow tits. Ardea 83, 397-404.
Kolar, C. S., and Rahel, F. J. (1993). Interaction of biotic factor (predator presence) and an
abiotic factor (low oxygen) as an influence on benthic invertebrate communities. Oecologia
95,210-219.
Korpimaki, E.. Norrdahl, K., and Valkama, J. (1994). Reproductive investment under fluctuat-
ing predation risk: Microtine rodents and small mustelids. Evol. Ecol. 8, 357-368.
Korpimaki, E.. Koivunen, V., and Hakkarainen. H. (1996). Microhabitat use and behavior
of voles under weasel and raptor predation risk: Predator facilitation? Behav. Ecol.
7, 30-34.
Koskela. E.. and Ylonen, H. (1995). Suppressed breeding in the field vole (Microtus agrestis):
An adaptation to cyclically fluctuating predation risk. Behav. Ecol. 6, 31 1-315.
Koskela, E., Horne, T. J., Mappes. T., and Ylonen, H. (1996). Does risk of small mustelid
predation effect the oestrous cycle in the bank vole, Clethrionomys glareolus? Anim.
Behav. 51, 1159-1163.
Kotler. B. P. (1992). Behavioral resource depression and decaying perceived risk of predation
in two species of coexisting gerbils. Behav. Ecol. Sociobiol. 30,239-244.
Kotler, B. P., and Blaustein, L. (1995). Titrating food and safety in a heterogeneous environ-
ment: When are risky and safe patches of equal value? Oikos 74, 251-258.
Kotler, B. P., and Holt. R. D. (1989). Predation and competition: The interaction of two types
of species interactions. Oikos 54, 256-260.
Kotler, B. P., Brown, J. S., and Hasson, 0. (1991). Factors affecting gerbil foraging behavior
and rates of owl predation. Ecology 72,2249-2260.
Kotler, B. P., Blaustein, L.. and Brown, J. S. (1992). Predator facilitation: The combined effect
of snakes and owls on the foraging behavior of gerbils. Ann. Zool. Fenn. 29, 199-206.
Kotler, B. P.. Blaustein, L., and Dednam, H. (1993a). The specter of predation: The effects
of vipers on the foraging behavior of two gerbilline rodents. Isr. J. Zool. 39, 11-21.
Kotler, B. P., Brown, J. S., Slotow. R. H., Goodfriend, W. L.. and Strauss. M. (1993b). The
influence of snakes on the foraging behavior of gerbils. Oikos 67, 309-316.
Kotler, B. P., Brown, J. S.. and Subach, A. (1993~).Mechanisms of species coexistence of
optimal foragers: Temporal partitioning by two species of sand dune gerbils. Oikos
67,548-556.
Kotler, B. P., Brown, J. S., and Mitchell, W. A. (1993d). Environmental factors affecting patch
use in two species of gerbilline rodents. J. Mammal. 74, 614-620.
Kotler, B. P.. Ayal, Y., and Subach, A. (1994a). Effects of predatory risk and resource renewal
on the timing of foraging activity in a gerbil community. Oecologia 100, 391-396.
Kotler, B. P., Gross, J. E., and Mitchell, W. A. (1994b). Applying patch use to assess aspects
of foraging behavior in Nubian ibex. J . Wild. Manage. 58,299-307.
278 STEVEN L. LIMA
Kramer, D. L., and Bonenfant, M. (1997). Direction of predator approach and the decision
to flee to a refuge. Anim. Behav. (in press).
Kratz, K. W. (1996). Effects of stoneflies on local prey populations: Mechanisms of impact
across prey density. Ecology 77, 157351585,
Krause, J. (1993a). The relationship between foraging and shoal position in a mixed shoal of
roach (Rutilus rutilus) and chub (Leuciscus cephaius): A field study. Oecologia93,356-359.
Krause, J. (1993b). The effect of ‘Shreckstoff’ on the shoaling behaviour of the minnow: A
test of Hamilton’s selfish herd theory. Anim. Behav. 45, 1019-1024.
Krause, J. (1994a). The influence of food competition and predation risk on size-assortative
shoaling in juvenile chub (Leuciscus cephalus). Ethology 96, 105- 116.
Krause, J. (1994b). Differential fitness returns in relation to spatial position in groups. Biol.
Rev. Cumbridge Philos. SOC.69, 187-206.
Krause, J., and Godin, J.-G. J. (1994). Shoal choice in the banded killifish (Fundulusdiaphanus,
Teleostei, Cyprinodontidae): Effects of predation risk, fish size, species composition and
size of shoals. Ethology 98, 128-136.
Krause, J., and Godin. J.-G. J. (1995). Predator preferences for attacking particular prey
group sizes: Consequences for predator hunting success and prey predation risk. Anim.
Behuv. 50, 465-473.
Krause, J., and Godin, J.-G. J. (1996). Influence of prey foraging posture on flight behavior
and predation risk: Predators take advantage of unwary prey. Behav. Ecol. 7, 264-271.
Krause, J., Bumann, D., and Todt, D. (1992). Relationship between the position preference
and nutritional state of individuals in schools of juvenile roach (Rutilus rutihs). Behav.
Ecol. Sociobiol. 30, 177-180.
Krupa, J. J., and Sih, A. (1993). Experimental studies on water strider mating dynamics:
Spatial variation in density and sex ratio. Behav. Ecol. Sociobiol. 33, 107-120.
Kiilling, D., and Milinski, M. (1992). Size-dependent predation risk and partner quality in
predator inspection of sticklebacks. Anim. Behav. 44,949-955.
L’Abte-Lund, J. H., Langeland. A,, Jonsson. E., and Ugedal. 0. (1993). Spatial segregation
by age and size in Arctic charr: A trade-off between feeding possibility and risk of
predation. J. Anim. Ecol. 62, 160-168.
Lagos, V. 0.. Contreras, L. C., Meserve, P. L., GutiCrrez, J. R., and Jaksic, F. M. (1995a).
Effects of predation risk on space use by small mammals: A field experiment with a
neotropical rodent. Oikos 74, 259-264.
Lagos, V. O., Bozinovic, F., and Contreras, L. C. (1995b). Microhabitat use by a small diurnal
rodent (Octoden degus) in a semiarid environment: Thermoregulatory constraints or
predation risk? J. Mammal. 76,900-905.
Lambin, X., Ims. R. A,, Steen, H., and Yoccoz, N . G. (1995). Vole cycles. Trends Ecol. Evol.
5, 204.
Lampert, W. (1989). The adaptive significance of die1 vertical migration of zooplankton. Funct.
Ecol. 3, 21-27.
Larsson, P., and Dodson, S. (1993). Chemical communication in planktonic animals. Arch.
Hydrobiol. 129, 129-155.
Lawler, S. P. (1989). Behavioural responses to predators and predation risk in four species
of larval anurans. Anim. Behav. 38, 1039-1047.
Legault, C.. and Himmelman, J. H. (1993). Relation between escape behaviour of benthic
marine invertebrates and the risk of predation. J. Exp. Mar. Biol. Ecol. 170,55-74.
Leibold, M. A. (1990). Resources and predators can affect the vertical distributions of zoo-
plankton. Limnol. Oceanogr. 35, 938-944.
Leibold, M. A. (1991). Trophic interactions and habitat segregation between competing Daph-
nia species. Oecologia 86, 510-520.
Le Maho, Y . , Karmann. H., Briot, D., Handrich, Y . , Robin, J.-P., Mioskowski, E., Cherel,
Y . .and Farni, J. (1992). Stress in birds due to routine handling and a technique to avoid
it. Am. J. Physiol. 263, R7755R781.
Leonardsson, K. (1991). Predicting risk-taking behaviours from life-history theory using static
optimization technique. Oikos 60, 149-154.
Lester, L. S.. and Fanselow, M. S. (1992). Nocturnality as a defensive behavior in the rat:
An analysis in terms of selective association between light and aversive stimulation.
Psychological Record 42,221 -253.
Levine, S.. Atha, K., and Wiener, S. G. (1993). Early experience effects on the development
of fear in the squirrel monkey. Behav. Neural B i d . 60, 225-233.
Levy. D. A. (1990a). Reciprocal die1 vertical migration behaviour in planktivores and zooplank-
ton in British Columbia lakes. Can. J. Fish. Aqccaf. Sci. 47, 1755-1764.
Levy, D. A. (1990b). Sensory mechanism and selective advantage for die1 vertical migration
in juvenile sockeye salmon, Oncorhynchus nerka. Can. J. Fish. Aquat. Sci. 47,1796-1802.
Licht, T. (1989). Discriminating between hungry and satiated predators: The response of
guppies (Poecilia reticulafa) from high and low predation sites. Efhology 82, 238-243.
Lima, S. L. (1990a). Protective cover and the use of space: Different strategies in finches.
Oikos 58, 151-158.
Lima, S. L. (1 990b). Evolutionarily stable antipredator behavior among isolated foragers: On
the consequences of successful escape. J . Theor. Biol. 143, 77-89.
Lima, S. L. (1991). Energy, predators and the behaviour of feeding hummingbirds. Evol. Ecol.
5,220-230.
Lima, S. L. (1992). Vigilance and foraging substrate: Anti-predatory considerations in a non-
standard environment. Behav. Ecol. Sociobiol. 30, 283-289.
Lima. S. L. (1993). Ecological and evolutionary perspectives on escape from predatory attack:
A survey of North American birds. Wilson Bull. 105, 1-47.
Lima, S. L. (1995). Back to the basics of anti-predatory vigilance: The group size effect. Anim.
Behav. 49, 11-20.
Lima, S. L.. and Dill, L. M. (1990). Behavioral decisions made under the risk of predation:
A review and prospectus. Can. J . Zool. 68, 619-640.
Lima, S. L., and Valone. T. J. (1991). Predators and avian community organization: An
experiment in a semi-desert grassland. Oecologia 86, 105-112.
Lima, S. L., and Zollner, P. A. (1996). Anti-predatory vigilance and the limits to collective
detection: Visual and spatial separation between foragers. Brhav. Ecol. Sociobiol. 38,
355-363.
Lindsay, S. M., and Woodin, S. A. (1995). Tissue loss induces switching of feeding mode in
spionid polychaetes. Mar. Ecol: Prog. Ser. 125, 159-169.
Lindstrom, A. (1990). The role of predation risk in stopover habitat selection in migrating
bramblings, Fringilla montifringilln. Behav. Ecol. 1, 102-106.
Lister, B. C., and Aguayo, A. G . (1992). Seasonality, predation, and the behaviour of a tropical
mainland anole. J . Anim. Ecol. 61, 717-733.
Lockard, R. B., and Owings, D. H. (1974). Moon-related surface activity of bannertail (Dipodo-
mys spectabilis) and Fresno (D. nitratoides) kangaroo rats. Anim. Behav. 22, 262-273.
Longland, W. S. (1991). Risk of predation and food consumption by black-tailed jackrabbits.
J. Range Manage. 44,447-450.
Longland, W. S., and Price, M. V. (1991). Direct observation of owls and heteromyid rodents:
Can predation risk explain microhabitat use? Ecology 72, 2261-2273.
Loose, C. J., and Dawidowicz, P. (1994). Trade-offs in die1 vertical migration by zooplankton:
The cost of predator avoidance. Ecology 75,2255-2263.
280 STEVEN L. LIMA
Loose. C. J., von Elert. E.. and Dawidowicz. P. (1993). Chemically-induced diet vertical
migration in Daphnia: A new bioassay for kairomones exuded by fish. Arch. Hydrohid.
126, 329-337.
Lucas, J. R., and Howard, R. D. (1995). O n alternative reproductive tactics in anurans:
Dynamic games with density and frequency dependance. Am. Nor. 146,365-397.
Lucas, J. R., and Walter, L. R. (1991). When should chickadees hoard food‘? Theory and
experimental results. Anim. Behav. 41, 579-601.
Lucas, J. R.. Howard, R. D.. and Palmer. J . G. (1996). Callers and satellites: Chorus behaviour
in anurans as a stochastic dynamic game. Anim. Behav. 51, 501-518.
Ludwig, D., and Rowe, L. (1990). Life history strategies for energy gain and predator avoidance
under time constraints. Am. Nar. 135, 686-707.
Macchiusi, F., and Baker, R. L. (1991). Prey behaviour and size-selective predation by fish.
Freshwater Biol. 25, 533-538.
Macchiusi. F.. and Baker, R. L. (1992). Effects of predators and food availability on activity
and growth of Chironomus tenfans (Chironomidae: Diptera). Freshwater Biol. 28,207-216.
MacWhirter, R. B. (1991). Effects of reproduction on activity and foraging behaviour of adult
female Colurnbian ground squirrels. Can. J. Zool. 69, 2209-2216.
Magnhagen. C. (1990). Reproduction under predation risk in the sand goby, Pomaroschistus
minutus, and the black goby, Gohirrs niger: The effect of age and longevity. Behav. Ecol.
Sociobiol. 26, 331 -335.
Magnhagen. C . (1991). Predation risk as a cost of reproduction. Trends Ecof. Evol. 6,183-186.
Magnhagen. C. (1992). Parental care and predation risk in fish. Ann. Zool. Fenn. 29,227-232.
Magnhagen, C. (1993). Conflicting demands in gobies: When to eat, reproduce. and avoid
predators. Mar. Behav. Physiol. 23, 79-90.
Magnhagen. C. (1995). Sneaking behaviour and nest defence are affected by predation risk
in the common goby. Anim. Behnv. 50, 1123-1128.
Magnhagen, C.. and Vestergaard, K. (1991). Risk taking in relation to reproductive investments
and future reproductive opportunities: Field experiments on nest-guarding common go-
bies, Pomatoschistus microps. Behav. Ecol. 2, 351-359.
Magurran. A . E. (1990). The adaptive significance of schooling as an anti-predator defence
in fish. Ann. Zool. Fenn. 27, 51-66.
Magurran, A . E., and Girling, S. L. (1986). Predator recognition and response habituation in
shoaling minnows. Anim. Behav. 34, 510-518.
Magurran. A. E., and Nowak. M. A. (1991). Another battle of the sexes: The consequences
of sexual asymmetry in mating costs and predation risk in the guppy. Poecilia reticulata.
Proc. R. Sue. London. Ser. B 246, 31-38.
Magurran. A. E., and Seghers. B. H. (199th). Population differences in predator recognition
and attack cone avoidance in the guppy Poecilia reticrtlata. Anim. Behav. 40, 443-452.
Magurran. A. E., and Seghers, B. H. (1990b). Risk sensitive courtship in the guppy (Poecilia
reticrtlata). Behaviour 112, 194-201.
Magurran. A . E., and Seghers. B. H. (1994). Predator inspection behaviour covaries with
schooling tendency amongst wild guppy. Poecilia reticulata. populations in Trinidad.
Behaviour 128, 121-134.
Malmqvist. B. (1988). Downstream drift in Madeiran levadas: Tests of hypotheses relating
to the influence of predators o n the drift of insects. Aqiiot. Insects 10, 141-152.
Mangel, M. (1989). Evolution of host selection in parasitoids: Does the state of the parasitoid
matter? Am. Nut. 133, 688-705.
Mangel. M.. and Clark, C . W. (1988). “Dynamic Modeling in Behavioral Ecology.” Princeton
University Press, Princeton, NJ.
Manteifel. Y. (1995). Chemically-mediated avoidance of predators by Rana temporaria tad-

poles. J . Herpetol. 29, 461-463.
Martel, G., and Dill, L. M. (1993). Feeding and aggressive behaviours in juvenile coho salmon
(Oncorhynchus kisutch) under chemically-mediated risk of predation. Behav. Ecol. Soczo-
b i d . 32, 365-370.
Martel, G., and Dill, L. M. (1995). Influence of movement by coho salmon (Oncorhynchus
kisutch) parr on their detection by common mergansers (Mergus merganser). Ethology
99, 139-149.
Mathis, A., and Smith, R. J. F. (1992). Avoidance of areas marked with a chemical alarm
substance by fathead minnows (Pimephales promelas) in a natural habitat. Can. J . Zool.
70, 1473-1476.
Mathis, A., and Smith, R. J. F. (1993a). Fathead minnows, Pimephales promelas, learn to
recognize northern pike, Esox lucius, as predators on the basis of chemical stimuli from
minnows in the pike's diet. Anim. Behav. 46,645-656.
Mathis, A,, and Smith, R. J. F. (1993b). Intraspecific and cross-superorder responses to chemical
alarm signals by brook stickleback. Ecology 74, 2395-2404.
Mathis, A,, Chivers, D. P., and Smith, R. J. F. (1993). Population differences in responses of
fathead minnows (Pimephales promelas) to visual and chemical stimuli from predators.
Ethology 93, 31-40.
Matsuda, H., Abrams. P. A,, and Hori, M. (1993). The effect of adaptive anti-predator behavior
on exploitative competition and mutualism between predators. Oikos 68,549-559.
Matsuda. H.. Hori, M.. and Abrams, P. A. (1994). Effects of predator-specific defence on
community complexity. Evol. Ecol. 8, 628-638.
Matsuda, H., Hori, M., and Abrams, P. A. (1996). Effects of predator-specific defence on
biodiversity and community complexity in two-trophic-level communities. Evol. Ecol.
10, 13-28.
Mayhew, P., and Houston, D. (1989). Feeding site selection by widgeon Anus penelope in
relation to water. Ibis 131, 1-8.
McIntosh. A. R., and Townsend, C. R. (1994). Interpopulation variation in mayfly antipredator
tactics: Differential effects of contrasting predatory fish. Ecology 75, 2078-2090.
McLeod. P. G . ,and Huntingford, F. A. (1994). Social rank and predator inspection in stickle-
backs. Anim. Behav. 47, 1238-1240.
McNamara, J . M., and Houston, A. I. (1986). The common currency for behavioral decisions.
Am. Nat. 127, 358-378.
McNamara, J. M., and Houston, A. I. (1990). The value of fat reserves in terms of avoiding
starvation. Acta Biotheor. 38, 37-61.
McNamara. J . M.. and Houston, A. 1. (1992). Evolutionarily stable levels of vigilance as a
function of group size. Anim. Behav. 43, 641-658.
McNamara, J. M., and Houston. A. I. (1994). The effect of a change in foraging options on
intake rate and predation rate. Am. Nut. 144, 978-1000.
McNamara, J. M., Houston, A. I.. and Lima, S. L. (1994). Foraging routines of small birds
in winter: A theoretical investigation. J. Avian Biol. 25, 287-302.
McNamara, J. M., Webb, J. N., and Collins, E. J . (1995). Dynamic optimization in fluctuating
environments. Proc. R. Soc. London, Ser. B 261,279-284.
McNeil, R.. Drapeau, P., and Goss-Custard, J . D. (1992). The occurrence and adaptive signifi-
cance of nocturnal habits in waterfowl. Biol. Rev. Cambridge Philos. Soc. 67, 381-419.
McPeek. M. A. (1990). Behavioral differences between Enalfagmaspecies (Odonata) influenc-
ing differential vulnerability to predators. Ecology 71, 1714-1726.
McPeek, M. A,. Schrot, A. K.. and Brown, J. M. (1996). Adaptation to predators in a new
community: Swimming performance and predator avoidance in damselflies. Ecology
77, 617-629.
282 STEVEN L. LIMA
Merkens, M., Harestad, A. S., and Sullivan, T. P. (1991). Cover and efficacy of predator-
based repellents for Townsend’s vole, Microtus townsendii. J. Chem. Ecol. 17, 401-41 2.
Mesa, M. G., Poe, T. P., Gadomski, D. M., and Petersen, J. H. (1994). Are all prey created
equal? A review and synthesis of diffrential predation on prey in substandard condition.
J . Fish Biol. 45, Suppl. A, 81-96.
Michiels, N. K., and Dhondt, A. A. (1990). Costs and benefits associated with oviposition
site selection in the dragonfly Sympetrum danae (Odonata: Libellulidae). Anim. Behav.
40,668-678.
Milinski, M. (1986). Constraints placed by predators on feeding behaviour. In “The Behaviour
of Teleost Fishes” (T. Pitcher, ed.). pp. 236-252. Croom Helm, London.
Milinski, M. (1987). Tit for tat in sticklebacks and the evolution of cooperation. Nature
(London) 325,433-435.
Milinski, M. (1993). Predation risk and feeding behaviour. In “Behaviour of Teleost Fishes”
(T. Pitcher, ed.), 2nd ed., pp. 285-305. Chapman & Hall, London.
Milinski, M., and Heller, R. (1978). Influence of a predator on the optimal foraging behaviour
of stickelbacks (Gasterosteus aculeatus L.). Nature (London) 275, 642-644.
Milinski. M., and Parker, G. A. (1991). Competition for resources. In “Behavioral Ecology:
An Evolutionary Approach” ( J . R. Krebs and N. B. Davies. eds.). 3rd ed., pp. 137-168.
Blackwell. Oxford.
Milinski, M., Kiilling, D., and Kettler, R. (1990a). Tit for tat: Sticklebacks (Gasterosteus
aculeatus) “trusting” a cooperating partner. Behav. Ecol. 1, 7-1 1.
Milinski. M.. Pfluger, D.. Kulling, D., and Kettler, R. (1990b). Do sticklebacks cooperate
repeatedly in reciprocal pairs? Behuv. Ecol. Sociobiol. 27, 17-21.
Moody, A. L., Houston, A. I., and McNamara, J. M. (1996). Ideal free distributions under
predation risk. Behav. Ecol. Sociobiol. 38, 131-143.
Moore, F. R. (1994). Resumption of feeding under risk of predation: Effect of migratory
condition. Anim. Behav. 48, 975-977.
Morey, S. R. (1990). Microhabitat selection and predation in the pacific treefrog, Pseudacris
regilla. J. Herpetol. 24, 292-296.
Mourn. S. E., and Baker, R. L. (1990). Colour change and substrate selection in larval Ischnuru
verticalis (Coenagrionidae: Odonata). Can. J. Zool. 68, 221-224.
Murphy, K. E., Pitcher, T. J . (1991). Individual behavioural strategies associated with predator
inspection in minnow shoals. Ethology 88, 307-319.
Neill, W. E. (1992). Population variation in the ontogeny of predator-induced vertical migration
of copepods. Nuture (London) 356,54-57.
Nemtzov, S. C. (1994). Intraspecific variation in sand-diving and predator avoidance behavior
of green razorfish, Xyrichtys splendens (Pisces, Labridae): Effect on courtship and mating
success. Environ. Biol. Fishes 41, 403-414.
Newman, J. A. (1991). Patch use under predation hazard: Foraging behavior in a simple
stochastic environment. Oikos 61, 29-44.
Nonacs, P. (1990). Death in the distance: Mortality risk as information for foraging ants.
Behaviour 112,23-35.
Nonacs, P., and Dill, L. M. (1990). Mortality risk vs. food quality in a common currency: Ant
patch preferences. Ecology 71, 1886-1892.
Ohman, M. D. (1990). The demographic benefits of die1 vertical migration by zooplankton.
Ecol. Monogr. 60, 257-281.
Oksanen, L., and Lundberg. P. (1995). Optimization of reproductive effort and foraging time
in mammals: The influence of resource level and predation risk. Evol. Ecol. 9,45-56.
Otto, C. (1993). Long-term risk sensitive foraging in Rhyacophila nubila (Trichoptera) larvae
from two streams. Oikos 68, 67-74.
Packer, C., and Abrams, P. (1990). Should co-operative groups be more vigilant than selfish
groups? J. Theor. Biol. 142, 341-357.
Palmer, A. R. (1990). Effect of crab effluent and scent of damaged conspecifics on feeding,
growth, and shell morphology of the Atlantic dogwhelk Nucella lapillits (L.). Hydrobio-
logia 193, 155-182.
Parejko. K.. and Dodson, S. (1990). Progress towards characterization of a predatodprey
kairomone: Daphnia pirlex and Chaoborus americanus. Hydrobiologia 198, 51-59.
Paveri-Fontana, S. L., and Focardi, S. (1994). A theoretical study of the socioecology of
ungulates. 11. A dynamic programming study of the stochastic formulation. Theor. Popul.
Biol. 46,279-299.
Peckarsky, B. L., Cowan, C. A,. Penton, M. A.. and Anderson, C. (1993). Sublethal conse-
quences of stream-dwelling predatory stoneflies on mayfly growth and fecundity. Ecology
74, 1836-1846.
Persson, L. (1991). Behavioral response to predators reverses the outcome of competition
between prey species. Behav. Ecol. Sociobiol. 28, 101-105.
Persson, L. (1993). Predator-mediated competition in prey refuges: The importance of habitat
dependent prey resources. Oikos 68, 12-22.
Persson. L., and Eklov, P. (1995). Prey refuges affecting interactions between piscivorous
perch and juvenile perch and roach. Ecology 76,70-81.
Persson, L., Anderson, J., Wahlstrom, E., and Eklov, P. (1996). Size-specific interactions in
lake systems: Predator gape limitation and prey growth rate and mortality. Ecology
77,900-91 1.
Peterson, C. H., and Skilleter, G. A. (1994). Control of foraging behavior of individuals within
an ecosystem context: The clam Macoma balthica, flow environment, and siphon-cropping
fishes. Oecologia 100,256-267.
Pettersson, L. B., and Bronmark, C. (1993). Trading off safety against food: State dependent
habitat choice and foraging in crucian carp. Oecologia 95, 353-357.
Phelan, J. P., and Baker, R. H. (1992). Optimal foraging in Peromyscus polionotus: The
influence of item-size and predation risk. Behaviour 121, 95-109.
Pierce, B. M., Longland, W. S., and Jenkins, S. H. (1992). Rattlesnake predation on desert
rodents: Microhabitat and species-specific effects on risk. J. Mammal. 73, 859-865.
Pitcher, T. J . (1992). Who dares, wins: The function and evolution of predator inspection
behaviour in shoaling fish. Neth. J. Zool. 42, 371-391.
Pitcher, T. J., Green, D. A,, and Magurran, A. E. (1986). Dicing with death: Predator inspection
behaviour in minnow shoals. J. Fish Biol. 28,439-448.
Poff, N. L., DeCino, R. D., and Ward, J. V. (1991). Size-dependent drift responses of mayflies
to experimental hydrolic variation: Active predator avoidance or passive hydrodynamic
displacement? Oecologia 88, 577-586.
Power, M. E. (1992). Top-down and bottom-up forces in food webs: Do plants have primacy?
Ecology 73, 733-746.
Power, M. E., Matthews, W. J., and Stewart, A. J. (1985). Grazing minnows, piscivorous bass,
and stream algae: Dynamics of a strong interaction. Ecology 66, 1448-1456.
Poysa, H. (1991). Effects of predation risk and patch quality on the formation and attractiveness
of foraging groups of teal, Anas crecca. Anirn. Behav. 41,285-294.
Price, M. V.. Goldingay, R. L., Szychowski, L. S.. and Waser, N. M. (1994). Managing habitat
for the endangered stephen’s kangaroo rat (Dipodomys stephensi): Effects of shrub
removal. Am. Midl. Nat. 131, 9-16.
Prins, H. H. T., and Iason. G. R. (1989). Dangerous lions and nonchalant buffalo. Behaviour
108, 262-296.
284 STEVEN L. LIMA
Pulliam, H. R., and Caraco. T. (1984). Living in groups: Is there an optimal group size? In
“Behavioural Ecology: an Evolutionary Approach” ( J . R. Krebs and N. B. Davies. eds.).
2nd ed., pp. 127-147. Blackwell, Oxford.
Radabaugh, D. C. (1989). Seasonal colour changes and shifting antipredator tactics in darters.
J. Fish Biol. 34, 679-685.
Rader, R. B., and McArthur, J. V. (1995). The relative importance of refugia in determining
the drift and habitat selection of predaceous stoneflies in a sandy-bottomed stream.
Oecologia 103, 1-9.
Rahel. F. J., and Kolar, C. S. (1990). Trade-offs in the response of mayflies to low oxygen
and fish predation. Oecologia 84, 39-44.
Ramcharan, C . W., and Sprules, W. G. (1991). Predator-induced behavioral defense and its
ecological consequences for two calanoid copepods. Oecologia 86,276-286.
Ramcharan, C. W., Dodson, S. I., and Lee, J. (1992). Predation risk, prey behavior, and
feeding rate in Daphnia pulex. Can. J . Fish. Aquut. Sci. 49, 159-165.
Ranta, E., Juvonen. S.-K., and Peuhkuri, N. (1992a). Further evidence for size-assertive
schooling in sticklebacks. J. Fish Biol. 41, 627-630.
Ranta, E., Lindstrom, K.. and Peuhkuri, N. (1992b). Size matters when three-spined stickle-
backs go to school. Anim. Behav. 43, 160-162.
Rayor, L. S., and Uetz, G. W. (1990). Trade-offs in foraging success and predation risk with
spatial position in colonial spiders. Behav. Ecol. Sociobiol. 27, 77-85.
Rayor, L. S., and Uetz, G. W. (1993). Ontogenetic shifts within the selfish herd: Predation
risk and foraging trade-offs change with age in colonial web-building spiders. Oecologia
95, 1-8.
Rehfeldt, G. E. (1990). Anti-predator strategies in oviposition site selection of Pyrrhosoma
nyrnphula (Zygoptera: Odonata). Oecologia 85, 233-237.
Rehfeldt, G . E. (1992). Aggregation during oviposition and predation risk in Syrnpetrum
vulgatum L. (Odonata: Libellulidae). Behav. Ecol. Sociohiol. 30, 317-322.
Repasky, R. R., and Schluter, D. (1994). Habitat distribution of wintering sparrows along an
elevational gradient: Tests of the food. predation and microhabitat structure hypotheses.
J . Anim. Ecol. 63,569-582.
Resetarits, W. J., Jr.. and Wilbur, H. M. (1989). Oviposition site choice in Hyla chrysoscelis:
Role of predators and competitors. Ecology 70, 220-228.
Reynolds, J. D. (1993). Should attractive individuals court more? Theory and a test. Am. Nar.
141,914-927.
Richardson, T. D., and Brown. K. M. (1992). Predation risk and feeding in an intertidal
predatory snail. J . Exp. Mar. Biol. Ecoi. 163, 169-182.
Ringelberg, J. (1991a). Enhancement of the phototactic reaction in Dnphnia hyalina by a
chemical mediated by juvenile perch (Perca puviatilis). J. Plankton Res. 13, 17-25.
Ringelberg, J. (1991b). A mechanism of predator-mediated induction of die1 vertical migration
in Daphnia hyalina. J. Plankton Res. 13, 83-89.
Ringelberg, J., Flik, B. J. G., Lindenaar, D., and Royackers, K. (1991). Die1 vertical migration
of Daphnia hyalina (sensu latiori) in Lake Maarsseveen: Part 1. Aspects of seasonal and
daily timing. Arch. Hydrobiol. 121, 129-145.
Rittschoff, D., Tsai, D. W., Massey. P. G., Blanco, L., Kueber. G . L., Jr., and Haas, R. J., Jr.
(1992). Chemical mediation of behavior in hermit crabs: Alarm and aggregation cues.
J. Chem. Ecol. 18,959-984.
Roberts. G . (1994). When to scan: An analysis of predictability in vigilance sequences using
autoregression models. Anim. Behav. 48, 579-585.
Roberts, G. (1995). A real-time response of vigilance behaviour to changes in group size.
Anim. Behav. 50, 1371-1374.
Roberts. G. (1996). Why individual vigilance declines as group size increases. Anim. Behav.
51,1077-1086.
Roca, J . R., Baltanas, A., and Uiblein, F. (1993). Adaptive responses in Cypridopsis vidua
(Crustacea: Ostracoda) to food and shelter offered by a macrophyte (Chara frugilis).
Hydrobiologia 262, 127- 13 I.
Rochette, R., Morissette. S.. and Himmelman. J. H. (1995). A flexible response to a major
predator provides a whelk Buccinum undatuni L. with nutritional gains. J. Exp. Mar.
B i d . Ecol. 185, 167-180.
Romey, W. L. (1995). Position preferences within groups: Do whirligigs select positions
which balance feeding opportunities with predator avoidance? Behav. E d . Sociobiol.
37, 195-200.
Romey, W. L., and Rossman, D. S. (1995). Temperature and hunger alter grouping trade-
offs in whirligig beetles. A m . Midl. Nor. 134, 51-62.
Ronkainen, H., and Ylonen, H. (1994). Behaviour of cyclic bank voles under risk of mustelid
predation: Do females avoid copulations? Oecologia 97, 377-381.
Rosland, R.. and Giske, J. (1994).A dynamic optimization model of the die1 vertical distribution
of pelagic planktivorous fish. Prog. Oceanogr. 34, 1-43.
Rowe, L. (1994). The costs of mating and mate choice in water striders. Anim. Behav. 48,1049-
1056.
Rowe, L., and Ludwig, D. (1991). Size and timing of metamorphosis in complex life cycles:
Time constraints and variation. Ecology 72,413-427.
Ruiz, G . M.. Hines, A. H., and Posey, M. H. (1993). Shallow water as a refuge habitat for
fish and crustaceans in non-vegetated estuaries: An example from Chesapeake Bay. Mar.
Ecol.: Prog. Ser. 99, 1-16.
Ruxton, G. D. (1995). Short term refuge use and stability of predator-prey models. Theor.
Popul. B i d . 47, 1-17.
Saarikko, J. (1992). Risk of predation and foraging activity in shrews. Ann. Zool. Fenn.
29,291-299.
Sargent. R. C. (1990). Behavioural and evolutionary ecology of fishes: Conflicting demands
during the breeding season. Ann. Zool. Fenn. 27, 101-118.
Scarratt, A. M., and Godin, J.-G.J. (1992). Foraging and antipredator decisions in the hermit
crab Pagtcrus acadianus (Benedict). J. Exp. Mar. Bid. Ecol. 156, 225-238.
Schlosser, I . J. (1987). The role of predation in age- and size-related habitat use by stream
fishes. Ecology 68, 651 -659.
Schluter, D. (1988). The evolution of finch communities on islands and continents, Kenya vs.
Galapagos. Ecol. Monogr. 58,229-249.
Schluter. D., and Repasky. R. R. (1991). Worldwide limitation of finch densities by food and
other factors. Ecology 72, 1763-1 774.
Schooley. R. L., Sharpe, P. B., and Van Horne, B. (1996). Can shrub cover increase predation
risk for a desert rodent? Can. J. Zool. 74, 157-163.
Schreck, C. B. ( 1990). Physiological, behavioral, and performance indicators of stress. A m .
Fish. SOC.Symp. 8, 29-37.
Schwinning, S., and Rosenzweig, M. L. (1990). Periodic oscillations in an ideal-free predator-
prey distribution. Oikos 59, 85-91.
Scrimgeour, G. J.. and Culp, J. M. (1994a). Foraging and evading predators: The effect of
predator species on a behavioural trade-off by a lotic mayfly. Oikos 69, 71-79.
Scrimgeour, G. J.. and Culp, J. M. (1994b). Feeding while evading predators by a lotic mayfly:
Linking short-term foraging behaviours to long-term fitness consequences. Oecologia
100, 128-134.
286 STEVEN L. LIMA
Scrimgeour, G. J., Culp, J. M., and Wrona, F. J. (1994). Feeding while avoiding predators:
Evidence for a size-specific trade-off by a lotic mayfly. J. Norrh Am. Benfhol. Soc. 13,
368-378.
Short, T. M., and Holomuzki, J. R. (1992). Indirect effects of fish on foraging behaviour and
leaf processing by the isopod Lirceus fonfinaiis. Freshwater Biol. 27, 91 -97.
Sibly, R. M.. and Calow, P. (1989). A life-cycle theory of responses to stress. B i d . J. Linn.
SOC. 37, 101-116.
Sih, A. (1980). Optimal behavior: can foragers balance two conflicting demands? Science
210, 1041-1043.
Sih, A. (1984). The behavioral response race between predator and prey. Am. Nut. 123,
143-150.
Sih, A. (1987). Predators and prey lifestyles: An evolutionary and ecological overview. In
“Predation: Direct and Indirect Impacts on Aquatic Communities” (W. C. Kerfoot and
A. Sih, eds.), pp. 203-224. University Press of New England, Hanover, NH.
Sih, A. (1988). The effects of predators on habitat use, activity and mating behaviour of a
semi-aquatic bug. Anim. Behav. 36, 1846-1848.
Sih, A. (1992a). Prey uncertainty and the balancing of antipredator and feeding needs. Am.
Nut. 139, 10.52-1069.
Sih, A. (1992b). Integrative approaches to the study of predation: General thoughts and a
case study on sunfish and salamander larvae. Ann. Zool. Fenn. 29, 183-198.
Sih, A. (1994). Predation risk and the evolutionary ecology of reproductive behaviour. J. Fish
Biol. 45, 111-130.
Sih, A. (1998). Three trophic level ideal free distributions: A game-theory approach to under-
standing the predator-prey behavioral response race. In “Advances in Game Theory and
the Study of Animal Behavior” (L. A. Dugatkin and H. K. Reeve, eds.), pp. 221-238.
Sih, A,, and Kats, L. B. (1991). Effects of refuge availability on the responses of salamander
larvae to chemical cues from predatory green sunfish. Anim. Behav. 42, 330-332.
Sih. A,, and Krupa, J. J. (1992). Predation risk, food deprivation and non-random mating by
size in the stream water strider, Aquarius remigis. Behuv. Ecol. Sociobiol. 31, 51-56.
Sih, A,, and Krupa, J. J. (1995). Interacting effects of predation risk and male and female
density on malelfemale conflicts and mating dynamics of stream water striders. Behav.
E d . 6,316-325.
Sih, A., and Krupa, J. J. (1996). Direct and indirect effects of multiple enemies on water
strider mating dynamics. Oecologiu 105, 179-188.
Sih, A., and Wooster, D. E. (1994). Prey behavior, prey dispersal, and predator impacts on
stream prey. Ecology 75, 1199-1207.
Sih, A., Petranka, J. W., and Kats, L. B. (1988). The dynamics of prey refuge use: A model
and tests with sunfish and salamander larvae. Am. Nat. 132,463-483.
Sih, A,. Krupa, J. J., and Travers, S. (1990). An experimental study on the effects of predation
risk and feedingregime on the mating behavior of the water strider. Am. Nut. 135,284-290.
Sih, A., Kats, L. B., and Moore, R. D. (1992). Effects of predatory sunfish on the density,
drift. and refuge use of stream salamander larvae. Ecology 73,1418-1430.
Sirnonetti, J. A. (1989). Microhabitat use by small mammalsincentral Chile. Oikos56,309-318.
Sinclair, A. R. E.. and Arcese, P. (1995). Population consequences of predation-sensitive
foraging: The Serengeti wildebeest. Ecology 76,882-891.
Skelly, D. K. (1992). Field evidence for a cost of behavioral antipredator response in a larval
amphibian. Ecology 73, 704-708.
Skelly, D. K. (1994). Activity level and the susceptibility of anuran larvae to predation. Anim.
Behav. 47, 465-468.
Skelly, D. K. (1995). A behavioral trade-off and its consequences for the distribution of
Pseudacris treefrog larvae. Ecology 76, 150-164.
Skelly, D. K., and Werner, E. E. (1990). Behavioral and life-historical responses of larval
American toads to an odonate predator. Ecology 71, 2313-2322.
Slotow, R., and Rothstein, S. I. (1995). Importance of dominance status and distance from
cover to foraging white-crowned sparrows: An experimental analysis. Auk 112,107-117.
Smith, R. J. (1995). Harvest rates and escape speeds in two co-existing species of montane
ground squirrels. J. Mammal. 76, 189-195.
Smith, R. J. F. (1992). Alarm signals in fishes. Rev. Fish Biol. Fish. 2, 33-63.
Smith, W. P. (1991). Ontogeny and adaptiveness of tail flagging behavior in white-tailed deer.
Am. Nat. 138, 190-200.
Soderback, B. (1994). Interactions among juveniles of two freshwater crayfish species and a
predatory fish. Oecologia 100, 229-235.
Sogard, S. M., and Olla, B. L. (1993). The influence of predator presence on utilization of
artificial seagrass habitats by juvenile walleye pollock, Theragra chalcogramma. Environ.
Biol. Fishes 37, 57-65.
Sparrevik, E., and Leonardsson, K. (1995). Effects of large Saduria entomon (Isopoda) on
spatial distribution of their small S. enfornon and Monoporeia afinis (Amphipoda) prey.
Oecologia 101, 177-184.
Speakman, J. R., Stone, R. E., and Kerslake, J. E. (1995). Temporal patterns in the emergence
behaviour of pipistrelle bats, Pipistrellus pipisfrellus, from maternity colonies are consis-
tent with the anti-predator response. Anim. Behav. 50, 1147-1156.
Stadler, B., Weisser, W. W., and Houston, A. I. (1994). Defence reactions in aphids: The
influence of state and future reproductive success. J. Anim. Ecol. 63,419-430.
Stamp, N. E., and Bowers, M. D. (1991). Indirect effect on survivorship of caterpillars due
to presence of invertebrate predators. Oecologia 88,325-330.
Stamp, N. E., and Bowers, M. D. (1993). Presence of predatory wasps and stinkbugs alters
foraging behavior of cryptic and non-cryptic caterpillars on plantain (Plantago lanceolafa).
Oecologia 95, 376-384.
Stanford, C. B. (1995). The influence of chimpanzee predation on group size and anti-predator
behaviour in red colobus monkeys. Anim. Behav. 49,577-587.
Stauffer, H.-P., and Semlitsch, R. D. (1993). Effects of visual, chemical and tactile cues of
fish on the behavioural responses of tadpoles. Anim. Behav. 46, 355-364.
Stephens, D. W., and Krebs, J. R. (1986). “Foraging Theory.” Princeton University Press,
Princeton, NJ.
Stephens, D. W., Anderson, J. P., and Benson, K. E. (1997). On the spurious occurrence of
Tit for Tat in pairs of predator-approaching fish. Anim. Behav. 53, 113-131.
Suhonen, J. (1993a). Predation risk influences the use of foraging sites by tits. Ecology 74,1197-
1203.
Suhonen. J. (1993b). Risk of predation and foraging sites of individuals in mixed-species tit
flocks. Anim. Behav. 45, 1193-1198.
Suhonen, J., Halonen, M., and Mappes, T. (1993). Predation risk and the organization of the
Pnrus guild. Oikos 66, 94-100.
Suhonen, J., Norrdahl, K., and Korpimaki, E. (1994). Avian predation risk modifies breeding
bird community on a farmland area. Ecology 75, 1626-1634.
Sweitzer, R. A,, and Berger, J. (1992). Size-related effects of predation on habitat use and
behavior of porcupines (Erethizon dorsatum). Ecology 73, 867-875.
Szekely, T., Sozou, P. D., and Houston, A. I. (1991). Flocking behaviour of passerines: A
dynamic model for the non-reproductive season. Behav. Ecol. Sociobiol. 28, 203-213.
288 STEVEN L. LIMA
Tayasu, I., Shigesada. N.. Mukai, H., and Caswell, H. (1996). Predator-mediated coexistence
of epiphytic grass shrimps that compete for refuges. Ecol. Morlell. 84, 1-10.
Tegeder, R. W.. and Krause, J. (199.5). Density dependence and numerosity in fright stimulated
aggregation behaviour of shoaling fish. Philos. Trans. R. Soc. London. Ser. B 350,381-390.
Theodorakis, C. W. (1989). Size segregation and the effects of oddity on predation risk in
minnow schools. Anim. Behav. 38, 496-502.
Tikkanen, P., Muotka, T., and Huhta. A. (1994). Predator detection and avoidance by lotic
mayfly nymphs of different size. Oecologia 99, 252-259.
Tjossem, S. F. (1990). Effects of fish chemical cues on vertical migration behavior of Chaoborus.
Limnol. Oceanogr. 35, 1456-1468.
Todd, 1. A., and Cowie, R. J. (1990). Measuring the risk of predation in an energy currency:
Field experiments with foraging blue tits, Parirs caemleus. Anim. Behav. 40, 112-1 17.
Tonn, W. M., Paszkowski, C. A,, and Holopainen, I. J. (1992). Piscivory and recruitment:
Mechanisms structuring prey populations in small lakes. Ecology 73, 951-958.
Turner, A. M. (1996). Freshwater snails alter habitat use in response to predation. Anim.
Behav. 51,747-756.
Turner, A. M. (1997). Contrasting short-term and long-term effects of predation risk on
consumer habitat use and resource dynamics. Behav. Ecol. 8, 120-125.
Turner. A. M., and Mittelbach. C. G. (1990). Predator avoidance and community structure:
Interactions among piscivores, planktivores, and plankton. Ecology 71, 2241-22.54.
Travers, S. E.. and Sih, A. (1991). The influence of starvation and predators on the mating
behavior of a semiaquatic insect. Ecology 72, 2123-2136.
Uetz, G. W.. and Hieber, C. S. (1994). Group size and predation risk in colonial web-building
spiders: Analysis of attack abatement mechanisms. Behav. Ecol. 5, 326-333.
Utne, A. C. W., Aksnes, D. L.. and Giske. J. (1993). Food, predation risk and shelter: An
experimental study on the distribution of adult two-spotted goby Gobiuscultrspavescens
(Fabricius). J. Exp. Mar. Biol. Ecol. 166, 203-216.
Uzendoski, K., Maksymovitch, E.. and Verell. P. (1993). D o the risks of predation and
intermale competition affect courtship behavior in the salamander Desmognathus ochro-
phaeus? Behav. Ecol. Sociohiol. 32, 42 1-427.
Vadas, R. L., Sr.. Burrows, M. T., and Hughes, R. N. (1994). Foraging strategies of dogwhelks.
Nucella lapillus (L.): Interacting effects of age, diet and chemical cues to the threat of
predation. Oecologia 100, 439-450.
van Baalen, M., and Sabelis. M. W. (1992). Coevolution of patch selection strategies of predator
and prey and the consequences for ecological stability. Am. Not. 142, 646-670.
Van Damme, R., Bauwens. D., Vanderstighelen, D., and Verheyen, R. F. (1990). Responses
of the lizard Lacerto vivipara to predator chemical cues: The effects of temperature.
Anim. Behav. 40,298-305.
van Schaik, C. P., and Horstermann. M. (1994). Predation risk and the number of adult males
in a primate group: A comparative test. Behav. Ecol. Sociohiol. 35, 261-272.
Visquez. R. A. (1994). Assessment of predation risk via illumination level: Facultative central
place foraging in the cricetid rodent Phyllotis darwini. Behav. Ecol. Sociobiol. 34,375-381.
Vaughn, C. C.. and Fisher, F. M. (1988). Vertical migration as a refuge from predation in
intertidal marsh snails: A field test. J . Exp. Mar. B i d . Eco/. 123, 163-176.
Vega-Redondo. F., and Hasson, 0.(1993). A game-theoretic model of predator-prey signaling.
J . Theor. B i d . 162, 309-319.
Vuorinen. I . (1987). Vertical migration of Eiirytemora (Crustacea, Copepoda): A compromise
between the risks of predation and decreased fecundity. J . Hankton Res. 9, 1037-1046.
Wahle. R. A. (1992). Body-size dependent anti-predator mechanisms of the American lobster.
Oikos 65,.52-60.
Wahle, R. A., and Steneck, R. S. (1Y92). Habitat restrictions in early benthic life: Experiments
on habitat selection and in situ predation with the American lobster. J . Exp. Mar. Biol.
Ecol. 157, 91-114.
Walls, S. C. (199.5). Differential vulnerability to predation and refuge use in competing larval
salamanders. Oecologia 101, 86-93.
Walters. C. J., and Juanes, F. (1993). Recruitment limitation as a consequence of natural
selection for use of restricted feeding habitats and predation risk taking by juvenile fishes.
Ciin. J . Fish. Ayuuf. Sci. 50, 2058-2070.
Watt, P. J.. and Young, S. (1YY4). Effect of predator chemical cues on Duphniu behaviour in
both horizontal and vertical planes. Anim. Behrrv. 48, 861-889.
Watts, B. D. (1990). Cover use and predator-related mortality in song and savannah sparrows.
Auk 107, 775-778.
Watts, B. D. (1991). Effects of predation risk on distribution within and between habitats in
savannah sparrows. Ecology 72, 1515-1.519.
Weary, D. M., Pajor. E. A,. Thompson, B. K., and Fraser, D. (1996). Risky behaviour by
piglets: A trade off between feeding and risk of mortality by maternal crushing? Anim.
Behav. 51,619-624.
Weiner. H. (1992). “Perturbing the Organism: The Biology of Stressful Experience.” Univer-
sity of Chicago Press, Chicago.
Weisser, W. W., Houston. A. I., and Volkl, W. (1994). Foragingstrategies in solitary parasitoids:
The trade-off between female and offspring mortality rates. Evol. Ecol. 8, 5877597.
Weldon. P. J. (1990). Responses by vertebrates to chemicals from predators. In “Chemical
Signals in Vertebrates 5” (D. W. Macdonald, D. Miiller-Schwarze. and S. E. Natynczuk,
eds.), pp. 500-521. Oxford University Press, Oxford.
Werner. E. E. (1986). Amphibian metamorphosis: Growth rate, predation risk, and the optimal
size at transformation. Am. Nut. 128, 319-341.
Werner, E. E. (1991). Nonlethal effects of a predator on competitive interactions between
two anuran larvae. Ecology 72, 1709-1720.
Werner, E. E. (1992). Individual behavior and higher-order species interactions. Am. Nat.
140, S5-S32.
Werner, E. E. (1994). Ontogenetic scaling of competitive relations: Size-dependent effects
and responses in two anuran larvae. Ecology 75, 197-213.
Werner, E. E., and Anholt, B. R. (1993). Ecological consequences of the trade-off between
growth and mortality rates mediated by foraging activity. Am. Nat. 142, 242-272.
Werner, E. E., and Gilliam, J. F. (1984). The ontogenetic niche and species interactions in
size-structured populations. Annu. Rev. Ecol. Sysf. 15, 393-425.
Werner, E. E., and McPeek. M. A. (1994). Direct and indirect effects of predators on two
anuran species along an environmental gradient. Ecology 75, 1368-1382.
Werner, E. E., Gilliam, J. F., Hall, D. J., and Mittelbach, G. G. (1983). An experimental test
of the effects of predation risk on habitat use in fish. Ecology 64, 1540-1548.
Williams, D. D. (1990). A field study of the effects of water temperature. discharge and trout
odour on the drift of stream invertebrates. Arch. Hydrohiol. 119, 167-181.
Williams, P. J., and Brown, J. A. (1991). Developmental changes in foraging-predator avoid-
ance trade-offs in larval lumpfish Cyclopterus lumpus. Mar. Ecol.: Prog. Ser. 76,.53-60.
Willman, E. J., Hill, A. M., and Lodge, D. M. (1994). Response of three crayfish congeners
(Orconectes spp.) to odors of fish carrion and live predatory fish. Am. Midl. Naf. 132,44451.
Wilson. D. J., and Lefcort. H. (1993). The effect of predator diet on the alarm response of
red-legged frog, Rana aurora, tadpoles. Anim. Behav. 46, 1017-1019.
Winkleman, D. L., and Aho, J. M. (1993). Direct and indirect effects of predation on mosquito-
fish behavior and survival. Oecologia 96, 300-303.
290 STEVEN L. LIMA
Wisenden, B. D., Chivers, D. P., and Smith, R. J. F. (1994). Risk-sensitive habitat use by
brook stickleback (Culaea inconstans) in areas associated with minnow alarm pheromone.
J. Chem. Ecol. 20,2975-2983.
Witter, M. S., and Cuthill, I. C. (1993). The ecological costs of avian fat storage. Philos. Trans.
R. SOC. London, Ser. B 340,73-92.
Witter, M. S., Cuthill, I. C., and Bonser, R. H. C. (1994). Experimental investigations of
mass-dependent predation risk in the European starling, Sturnus vulgaris. Anim. Behav.
48,201-222.
Wolf, N. G. (1985). Odd fish abandon mixed-species groups when threatened. Behav. Ecol.
Sociohiol. 17, 47-52.
Wolfe, J. L., and Summerlin, C. T. (1989). The influence of lunar light on nocturnal activity
of the old-field mouse. Anim. Behav. 37,410-414.
Wooster, D., and Sih, A. (1995). A review of the drift and activity responses of stream prey
to predator presence. Oikos 73, 3-8.
Wrona, F. J., and Dixon, R. W. J. (1991). Group size and predation risk: A field analysis of
encounter and dilution effects. Am. Nut. 137, 186-201.
Ydenberg, R. C. (1994). The behavioral ecology of provisioning in birds. b s c i e n c e 1,1-14.
Ydenberg, R. C., and Dill, L. M. (1986). The economics of fleeing from predators. Adv. Study
Behav. 16,229-249.
Ylonen, H. (1989). Weasels Mustela nivalis suppress reproduction in the cyclic bank voles
Clethrionomys glareolus. Oikos 55, 138-140.
Ylonen, H. (1994). Vole cycles and antipredatory behaviour. Trends Ecol. Evol. 9,426-430.
Ylonen, H., and Ronkainen, H. (1994). Breeding suppression in the bank vole as antipredatory
adaptation in a predictable environment. Evol. Ecol. 8, 658-666.
Ylonen, H., Jedrzejewska, B., Jedrzejewski, W., and Heikkila, J. (1992). Antipredatory behav-
iour of Clethrionomys voles: ‘David and Goliath’ arms race. Ann. 2001.Fenn. 29,207-216.
Young, S., and Watt, P. (1993). Behavioral mechanisms controlling vertical migration in
Daphnia. Limnol. Oceanogr. 38, 70-79.
Ziv, Y., Abramsky, Z . , Kotler, B. P., and Subach, A. (1993). Interference competition and
temporal and habitat partitioning in two gerbil species. Oikos 66, 237-246.
ADVANCES IN THE STUDY OF BEHAVIOR, VOL. 27
Parasitic Stress and Self-Medication

in Wild Animals
GEORGE
A. LOZANO*
DEPARTMENT OF BIOLOGICAL SCIENCES
UNIVERSITY OF CALIFORNIA
RIVERSIDE, CALIFORNIA 92521
I. INTRODUCTION
In the physical sciences, “stress” is defined as the force per unit area, or
pressure, acting upon a solid body, resulting in the deformation (strain) of
the solid. At low stresses the strain is said to be elastic, directly proportional
to the stress and reversible; the solid returns to its original shape after the
stress is removed. As the stress increases, the elastic limit is reached, after
which the strain is said to be plastic, increasing exponentially with increasing
stress and nonreversible. Plastic deformation continues until the rupture
strength is reached, at which point the material breaks.
The term stress was adopted by biologists to refer to factors that interfere
with the maintenance of homeostasis, the effects of which range from the
minor, temporary, and easily reversible, to the complete breakdown of
homeostatic mechanisms (Cannon, 1935). As applied to vertebrates, the
term “stress” is generally used to denote stimuli that elicit a specific set of
physiological responses, particularly the release of corticosteroids (Vander,
1981; Kopin, 1995; Mims et al., 1995). However, these responses are not
characteristic of all taxa, so this definition is not inclusive. Stresses can also
be defined more broadly as aversive stimuli (McGrath, 1970; Selye, 1976),
regarded as selective forces, and studied along with the adaptations that
have evolved to reduce their negative effects (see Thornhill and Furlow,
this volume). Under this view, stresses can take a myriad of forms, as
indicated by the wide range of topics included in this volume.
Along with competition and predation, parasitism is one of the main sources
of biotic stress facing all organisms. For the purposes of this discussion, para-
sites will be functionally defined as organisms that live in or on a heterospecific
* Present address: Behavioral Ecology Research Group, Department of Biological Sciences,
Simon Frazer University, Burnaby, British Columbia, V5A 156 Canada
291
Copyright 0 1YYX by Academic Prew
0065 3454/YX $25 00
292 GEORGE A. LOZANO
animal (the host), draw their nutrients primarily from the host, and have
the potential to reduce its fitness. Therefore, this definition includes both
endoparasites and ectoparasites, but excludes micropredators or animals that
use their hosts solely for shelter. Second, both macroparasites ( h e h n t h s ,
arthropods) and microparasites (viruses, bacteria, protozoa, fungi) are in-
cluded. Finally, parasites need not be harmful all the time, or even most of
the time. Parasites can often coexist with their hosts without causing any
measurable deleterious effects, but parasites are also opportunistic, and can
quickly increase in numbers and overwhelm a host weakened by other forms
of stress, such as malnutrition or reproduction (Walzer and Genta, 1989).
To counteract actual or potential fitness losses due to parasitism, animals
have evolved a variety of anatomical, physiological, and behavioral adapta-
tions, and parasites have developed an equally impressive array of counter-
measures to bypass these defenses (Behnke and Barnard, 1990). In some
cases parasites have even evolved ways to manipulate their hosts’ behavior
for their own interests (e.g., Bethel and Holmes, 1973; Brassard el af.,1982;
Maitland, 1994). The effects of parasites on host behavior include the
manipulation of host behavior by parasites (reviewed by Moore and Gotelli,
1990), and host behavioral adaptations for protection against parasitism
(reviewed by Hart, 1990; Mdler et al., 1993).
Recently, it has become recognized that animal diets may also be shaped
by the need for protection from parasites. Foraging behavior evolves pri-
marily to meet the need of a nutritionally adequate diet. However, just as
foraging behavior can be affected by predators (e.g., Milinski and Heller,
1978; Krebs, 1980; Sih, 1980; Edwards, 1983; Abrahams and Dill, 1989;
Lima and Dill, 1990) and competitors (e.g., Baker et al., 1951; Milinski,
1982; Millikan ef al., 1985), some features of diet selection seem to have
evolved to stave off, or reduce parasitism. These adaptations can include
the avoidance of foods that are also potential sources of parasitic infection,
the use of prophylactic substances, and the consumption of therapeutic
substances (Phillips-Conroy, 1986; Lozano, 1991). Self-medication includes
the latter two types of responses.
Although in this chapter I deal largely with self-medication in the con-
text of feeding, it may also occur under other circumstances, including
the use of plants with potentially antibacterial chemicals for nest material
(Wimberger, 1984; Clark and Mason, 1985), and the topical application of
antifungal and antibacterial compounds (Ehrlich ef al., 1986; Baker, 1996;
Gompper and Hoylman, 1993). In this chapter I first incorporate self-
medication into the broader phenomenon, namely, the effects of plant
chemicals across several trophic levels, and categorize self-medicating be-
havior into two basic forms: prophylactic and therapeutic. In the body of
the chapter I review in detail current evidence in the published literature
SELF-MEDICATION IN WILD ANIMALS 293
for the occurrence of self-medication in nonhuman vertebrates. Finally, I

discuss behavioral mechanisms that may play a role in self-medication, and
highlight potential implications for other areas of research.
11. SELF-MEDICATION
The effects of secondary plant metabolites are not always limited to the
herbivores that consume them, but may also affect the herbivores’ preda-
tors, parasites, and parasitoids. For example, in several herbivorous insects
susceptibility to pathogens differs depending on the plant on which the
hosts feed (e.g., Hare and Andreadis, 1983; Krischik et al., 1988). Such
interactions have long been studied in the general framework of chemical
ecology, mostly in insects (reviewed by Duffey, 1980; Price et al., 1980).
Nevertheless, animals in other taxa are also able to ingest secondary plant
metabolites and accumulate them in their tissues. These compounds can
make prey unpalatable to predators (e.g., Brower, 1958; Rothschild, 1972;
Hay et al., 1990; Pennings, 1994; Daly e f al., 1994), or less susceptible to
parasitoids (e.g., Campbell and Duffey, 1979). Sequestered compounds,
specifically carotenoids can also play a role in sexual selection by altering
the showiness of secondary sexual ornaments in males (e.g., Kodric-Brown,
1989; Zuk, 1992; Milinski and Bakker, 1990; Hill, 1994), although it is
unknown whether these traits are important in sexual selection because
they indicate foraging ability or immunocompetence (Endler, 1980; Lozano,
1994). It is therefore clear that plant chemicals can have effects across
several trophic levels. The use of secondary plant metabolites by vertebrates
for the purpose of self-medication can be viewed as a special case of this
broader phenomenon.
Janzen (1978) was probably responsible for bringing to the forefront of
western scientific inquiry the idea of self-medication in nonhumans. He
compiled anecdotal accounts of unusual feeding habits by several species
of mammals. For example, just before starting long trips, Indian elephants
(Elephas muximus) reportedly feed on Entuda schefferi (Leguminosae).
Indian wild boars (Sus scrofu) consume the roots of Boerhavia diffusa
(Nyctaginaceae), a plant used in traditional medicine as an anthelminthic.
Sumatran rhinoceroses (Didermocerus sumatrensis) have been observed
eating copious quantities of the tannin-laden bark of mangroves (Ceriops
candoleana, Rhizophoraceae). Janzen pointed out that energy requirements
and chemical avoidance were probably not adequate to explain these obser-
vations, and raised the possibility that animals use plant secondary metabo-
lites as stimulants, antihelminthics, laxatives, antibiotics, or even as anti-
dotes for previously consumed toxins.
Despite this apparent taxonomic and behavioral diversity, self-

medication can take only two functionally distinct forms, preventive (pro-
phylactic) and therapeutic (Phillips-Conroy, 1986; Lozano, 1991). The two
processes yield different predictions and require distinct behavioral mecha-
nisms. By viewing self-medication under a more general framework, these
behaviors need not be studied as a series of isolated cases, but rather can
be considered in terms of common elements.
For example, the consumption of food items for preventative purposes
would be related to the risk of parasitism, but not necessarily to the presence
of parasites. The biological effects of these medicines may be aimed solely
at the infectious stage of the parasite, and could have no effect at all on
established parasites. Furthermore, the consumption of medicinal sub-
stances may not vary substantially among individuals within a population,
but could differ considerably between populations. Lastly, if the risk of
parasitism is predictable, seasonally, for example, dietary shifts may be
largely genetically determined, and not depend on individual or social
learning. This also means that the consumption of prophylactic food items
will probably be difficult to demonstrate conclusively, even for a single
parasite-host-medicine system, because the consumption of these food
items would likely be integrated with the regular diet.
On the other hand, in cases of therapeutic self-medication, only sick
individuals would be expected to consume medicinal substances. These
food items would not be expected to be in the animal’s regular diet, and
would be consumed only upon infection. Therapeutic medications would
probably be more potent than preventative ones, and consequently would
carry a greater risk of negative side effects. Medicinal substances could be
aimed at the infection, in which case their biological effect would be directed
at parasites already established within the host. Alternatively, the purpose
of medicinal substances could be to alleviate discomfort, similar to the use
of medicines for the common cold by humans, and have no effect at all on
the parasites. In either situation, the ability to self-diagnose, prescribe, seek,
and consume the appropriate medicine requires a fairly complex mechanism
of individual and/or social learning.
111. PROPHYLACTIC
SELF-MEDICATION
Studies have not always made clear the distinction between preventative
and curative self-medication. As previously indicated, the difference is that
therapeutic self-medication is a specific response to a particular situation;
that is, the deliberate consumption of medicinal substances by ill individuals.
In this section I discuss instances in which secondary plant metabolites seem
to affect parasites or disease, but, so far, there is no evidence suggesting

intentionality. The classification of the following behaviors as preventative
self-medication is therefore not definitive, as further work may show that
these behaviors are also examples of therapeutic self-medication.
A. GEOPHACY
IN PRIMATES
Geophagy, the deliberate consumption of soil, dirt, or rock, has been

observed in several herbivorous and omnivorous mammals (reviewed by
Kreulen, 1985). Geophagy may be used to control gut p H (Oates, 1978;
Davies and Baille, 1988), to meet nutritional requirements of trace minerals
(Davies and Baille, 1988; Johns and Duquette, 1991), to satisfy a specific
hunger for sodium (Mahaney et al., 1990), andlor to detoxify secondary
plant metabolites (Johns and Duquette, 1991). Recently, it has also been
suggested that some primates may use geophagy to combat intestinal prob-
lems, particularly diarrhea (Mahaney et al., 1995a,b).
Geophagy has been studied in the context of self-medication in Japanese
macaques (Macaca fuscata) (Mahaney et al., 1993), rhesus macaques
(Macaca rnularta) (Mahaney et al., 1995a), mountain gorillas (Gorilla go-
rilla) (Mahaney 1993; Mahaney et al., 1995b), and chimpanzees (Pan troglo-
dytes) (Mahaney et al., 1996,1997). Analyses of the soils consumed by these
four species have detected at least one of three mineralogically similar
clays: halloysite, metahalloysite, and kaolinite, the last of which is the
principal ingredient of the commercial antidiarrheal Kaopectate TM (Ver-
meer and Ferrell, 1985). So far, support for the idea of geophagy as self-
medication is limited to these mineralogical analyses. There have been no
studies relating geophagy to the incidence or risk of diarrhea, nor have
there been studies on the physiological effects of these clays in nonhumans.
B. STIMULANT
USE IN BABOONS
Hamilton et al. (1978) classified food items consumed by chacma baboons
(Papio ursinus), into four categories: (1) animals, (2) fruits and seeds,
( 3 ) leaves, and (4) “euphorics.” The fourth group consisted of plants that
were widely available and consumed consistently, but only in minute quanti-
ties. Furthermore, these plants were known to be hallucinogenic and highly
toxic to humans, and presumably also to other mammals (Hamilton et al.,
1978). These “euphorics” included Croton megalobotrys (Euphorbiaceae),
Euphorbia avasmontana (Euphorbiaceae), Datura innoxia (Solanaceae),
and D. stramonium. Subsequent authors (Huffman and Seifu, 1989;
Wrangham and Goodall, 1989) have cited this study as an example of
self-medication; however, aside from labeling these plants as “euphorics,”
Hamilton et al. (1978) did not speculate on their possible role(s). There
has been no further work with this system.
C. ANTISCHISTOSOMAL
DRUGUSEBY BABOONS
Phillips-Conroy (1986) examined the diet of baboons along the Awash
River Valley, Ethiopia, which is divided by waterfalls into two distinct
habitats, with water flow being faster upstream, but slower after the falls.
The valley was populated by anubis baboons (Papio anubis) above the
falls, and hamadryas baboons (Papio hamadryas) and anubis-hamadryas
hybrids below the falls. The risk of schistosomiasis infection varied for
these populations because snails (Biomphalaria sp.), the intermediate hosts
of Schistosoma spp., were absent upstream from the waterfalls, but were
abundant downstream. Finally, although the shrub Balanites aegyptica (Ba-
lanitaceae) was common throughout the valley, only downstream from the
falls did baboons consume its leaves and fruits. Balanites fruits contain
diosgenin, a hormone precursor. Phillips-Conroy (1986) suggested that Ba-
lanites is consumed because it hinders the development of schistosomes,
but experimental work in schistosome-infected mice showed that ingestion
of diosgenin actually increases the number of schistosome eggs in the liver;
it enhances the disease (Phillips-Conroy and Knopf, 1986).
FOLIAGE
D. ANTIBACTERIAL AS NESTMATERIAL
Several bird species place in their nests fresh vegetation that does not
constitute part of the nests’ structure. Wimberger (1984) noted that fresh
plants probably contain more volatile secondary compounds than does
dried vegetation, and he hypothesized that birds use these plants t o repel
or even kill ectoparasites. Using data from egg collections of North Ameri-
can and European Falconiformes, and based on the premise that nest reuse
leads to increased parasite transmission, Wimberger (1984) showed that
Falconiformes that reused their nests in successive years were more likely
to use green foliage in their nests, and those that did not were less likely
to do so. Clark and Mason (1985) conducted a similar comparison using
selected North American passerines and found that cavity nesters were
more likely to use green foliage than were open cup nesters (Table I).
Clark and Mason (1985) also demonstrated that plant use by starlings
(Sturnus vulgaris) was not random, as the plants selected did not simply
reflect the availability in the surrounding areas. Furthermore, preferred
plants were more effective at reducing the hatching success of lice (Mena-
canthus sp.) eggs and inhibiting bacterial growth than a random subset of
the available vegetation. Subsequently, they showed experimentally that
leaves of wild carrot (Daucus carom, Umbelliferae), one of the preferred
TABLE I
USEOF GREENPLANTS AS NESTMATERIAL I N RELATIONTO NESTREUSE A N D TYPEOF NEST
AMONG FALCONIFORMES A N D NORTH AMERICAN PASSERINES. RESPECTIVELY (EXPECTED
FREQUENCIES
I N PARENTHESES: FROM WIMBERCER. 1984; CLARK A N D MASON.1985.)
Use of green
vegetation
Present Absent x? P
a) Falconiformes ( n = 48)
Reuse nests 22(17.5) 8(12.5) 8,28 <o,oo5
Build new nests 6(10.5) 12(4.5)
b) North American Passerines (PZ = 137)
Enclosed nests lg(9.1) 9(17.9) 16.4 <0.001
Open nests 28(36.9) 82(73.1)
species, significantly reduced the number of fowl mites (Ornithonysus sylvi-

arum) in starling nests (Clark and Mason, 1988). The decrease in mite
abundance had no effect on nestling growth, but nestlings from nests with
carrot leaves had higher hemoglobin levels than chicks from control nests.
Therefore, it seems fairly clear that starlings select nest material with
insecticidal and antibacterial properties. However, contrary to what would
be expected according to the chemically mediated parasite-protection hy-
pothesis, starlings add green vegetation to their nests only during nest
building, and, unlike Clark and Mason (1988), not while eggs or young are
in the nest. Also, males that reuse a nest box during one breeding season,
whether because the first brood fledged or was lost, collect less foliage than
males nesting concurrently but for the first time (Gwinner, 1997). Finally,
the hypothesis does not explain why only males add green vegetation to
their nests, and first-year males use less fresh vegetation than older males
(Clark and Mason, 1985). Several other hypothesis, not necessarily alterna-
tive, have been proposed to explain the use of green vegetation in nests.
Green foliage may serve to attract females (Fauth et al., 1991; Gwinner,
1997) and actually be a rudimentary bower; it may be used to cover debris
and keep the nest clean; it may advertise nest occupancy, or it may prevent
egg desiccation. It would be interesting to know whether other species
behave similarly, and whether starlings use more green foliage in response
to higher levels of parasitism.
E. ANTING
A N D FURRUBBING
Anting refers to a behavior in which birds rub crushed ants throughout

their plumage. Birds also ant by lying on ant nests and letting ants crawl
over their plumage. This behavior occurs in a variety of birds (Potter, 1970)
and it has been suggested that anting is used to soothe irritated skin, help
with feather maintenance, and prevent or reduce the abundance of skin
parasites (Potter, 1970; Clunie, 1976; Ehrlich et al., 1986). Birds also “ant”
with other invertebrates, plants, and inanimate objects, such as millipedes
(Clunie, 1976), lime fruit (Clayton and Vernon, 1993), and mothballs (Clark
et al., 1990), all of which have some antiparasitic properties. Anting has
also been observed in mammals (Bagg, 1952; Hauser, 1964; Longino, 1984).
An analogous behavior, fur rubbing, occurs in some mammals. Baker
(1996) observed capuchin monkeys (Cebus capucinus) in Costa Rica rub-
bing their fur with the fruits of several species of Citrus (Rutadeae), and
with the leaves or stems of the vines Piper rnarginaturn (Piperaceae) and
Clematis dioica (Ranunculaceae). These plants have a wide range of bioac-
tive compounds and are used in traditional medicine to treat a variety of
ailments. White-nosed coatis (Nasua narica) have been observed coating
their bodies with Trattinnickia aspera (Burseraceae) resin. Although infor-
mation on the medicinal uses of T. aspera is limited, Gompper and Hoylman
(1993) suggested this behavior serves a medicinal function. In conclusion,
support for the idea that anting and fur rubbing are primarily antiparasitic
behaviors is still largely anecdotal; more detailed and experimental studies
are presumably forthcoming.
IV. THERAPEUTIC
SELF-MEDICATION
In contrast to prophylactic self-medication, evidence for therapeutic self-

medication is more compelling, and has attracted considerably more atten-
tion. This evidence comes from a single source: chimpanzees at Gombe
National Park and the Mahale Mountains, Tanzania, but it is very diverse
in nature. Conclusions are based on direct observations of chimpanzees in
the wild, fecal analyses, traditional medicine, and biochemistry. Several
other plants may be involved, but most work has concentrated on the
possible therapeutic use by chimpanzees of three specific plants.
A. ASPILIA
The first report of a possible case of therapeutic self-medication was based
on several peculiarities of the consumption of leaves of Aspilia pluriseta
(Compositae), A . rudis, and A. rnossarnbicencis (Wrangham and Nishida,
1983). Field observations of chimpanzees and subsequent fecal analyses
revealed that entire leaves were swallowed without being chewed. Instead,
these leaves were taken singly and rolled around the mouth before being
swallowed. It was later suggested that this feeding technique may facilitate
the intake of any existing medicinal substances via the buccal mucosa
(Newton and Nishida, 1990). At Gombe consumption occurred only during
the morning, but at Mahale it occurred all day. Finally, there was no
between-individual variation in the tendency to consume Aspilia leaves.
Based on these observations and the widespread use of Aspilia in traditional
medicine, Wrangham and Nishida (1983) suggested that these leaves are
consumed because of their pharmacological effects. However, because
of the lack of individual variation and because, at Gombe, consumption
occurred only during the morning, Wrangham and Nishida (1983) con-
cluded Aspilia was probably used as a stimulant, rather than as a medicine.
It is difficult to draw conclusions about the seasonal variation of Aspilia
consumption. At Mahale, the percentage of chimpanzee feces containing
Aspilia leaves was highest in January and February (Wrangham and
Nishida, 1983). In contrast, at Gombe, the presence of Aspilia leaves in
fecal samples was highest during June and July, but behavioral observa-
tions indicated that consumption peaked in January, November, and May
(Wrangham and Goodall, 1989) (Fig. 1). Further work has shown that
--*-
Gombe feces
I
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
FIG. 1 . Seasonal variation of Aspilia consumption by chimpanzees at Mahale and Gombe,

based on fecal samples and behavioral observations. (From Wrangham and Nishida, 1983
and Wrangham and Goodall. 1989).
the prevalence of infection by the intestinal nematode Oesophagostomurn

stephanostornurn is highest during the rainy season (November to March),
but there are no seasonal patterns in the prevalence of two other intestinal
nematodes: Trichuris trichiura and Stongyloides fuelleborni (Huffman et
al., 1997).
Several secondary metabolites have been obtained from other Aspilia
species (Mabry et al., 1977; Bohlmann et al., 1984; Ganzer et al., 1992).
Methanol extracts of A . rnossarnbicensisleaves have limited biological activ-
ity against a variety of insects, herbs, and fungi (Ohigashi et al., 1991a). In
contrast, chloroform extracts of dried leaves yielded thiarubrine-A,
a naturally occurring phototoxic compound also found in other species
of Compositae (Rodriguez et al., 1985). In the presence of UV-A light,
thiarubrine-A is toxic to several bacteria and viruses, and at least one free-
living nematode, but its toxicity decreases in the absence of light. Under
acidic or alkaline conditions thiarubrine-A readily changes into thiophene-
A, which is toxic only in the presence of UV-A light (Towers et al., 1985;
Hudson et al., 1986, Table 11).
Page et al. (1992) found thiarubrine-A in the roots of A . mossarnbicensis,
but were not able to isolate it from samples of either fresh or dried leaves.
They did, however, isolate two diterpenes, kaurenoic acid and grandiflorenic
acid from dried leaves, and showed that these compounds stimulate contrac-
T A B L E I1
IN VITRO TOXICITY
OF THIARUBRINE-A
AND THIOPHENE-A"
Thiarubrine-A Thiophene-A
Organism Light Dark Light Dark Ref.
Caenorhabditis riegans ++ + nt nt 1
Saccharomyces cerevisinc, ++ ++ nt nt 1
Candida albicans ++ ++ ++ ~
1
Staphylococcrrs nlbrrs ++ - ++ - 1
Bacillus subtilis ++ ++ nt nt 1
Streptococcus fecalis ++ - nt nt 1
Escherichia coli ++ + + ~
1
Pseudotriotias ,floirrescerzs - - nt nt 1
Mycobacrrririm phlei ++ + nt nt I
Murine cytomegalovirus ++ - nt nt 2
Sindbis virus ++ ~
nt nt 2
T4 bacteriophage + - nt nt 2
M 13 bacteriophage - ~
nt nt 2
1 = Towers et al. (1985)

2 = Hudson er a/. (1986)
" ++ = highly toxic. + = weakly toxic, - = no effect. nt = not tested.
tions of a guinea pig uterus in vitro. Observations of chimpanzees revealed

that at Gombe more females than males selected Aspilia leaves, and the
number of days in which Aspilia leaves were consumed was also significantly
higher for females (Wrangham and Goodall, 1989), which led Page et af.
(1992) to suggest that pregnant chimpanzees consume the leaves of A .
mossambicensis to induce labor. This hypothesis would predict increases
in Aspilia consumption by pregnant females as they approach their due
dates, and perhaps even by females ill suited to carry their fetuses to term.
However, there is no information on Aspilia use by pregnant females, nor
any evidence to indicate that Aspilia induces labor in vivo. This idea has
not received further consideration.
Conclusions based on the chemical analyses must be considered tentative
for at least two reasons. First, it is difficult to build a case for self-medication
based on the ingestion of thiarubrine-A, the biological activity of which is
markedly decreased, or completely absent, without light. Given its proper-
ties, thiarubrine-A seems an unlikely medicinal substance, except if used
as an external antibiotic (see Ehrlich et al., 1986; Gompper and Hoylman,
1993; Baker, 1996). Second, two subsequent studies (Page et af., 1992;
Huffman et al., 1996) have failed to detect thiarubrine-A in leaf samples
of A . mossambicensis, as first reported by Rodriguez et al. (1985). If indeed
only the roots of A . mossambicensis contain thiarubrine-A, then its biologi-
cal activity is irrelevant to leaf-eating chimpanzees. However, it has recently
been suggested that the leaves of Aspilia sp. and other suspected medicinal
plants may be consumed not because of their chemical properties, but
rather because of their characteristically rough surfaces, which may aid in
the mechanical removal of intestinal parasites (Messner and Wrangham,
1996; Huffman et al., 1996).
B. VERNONIA
The recognition of Vernonia amygdafina (Compositae) as a possible
chimpanzee medicinal plant was also the result of detailed field observations
(Huffman and Seifu, 1989). An adult female, dubbed CH, was observed
during 2 consecutive days, for a total of about 11 hr. For 35 min during
the afternoon of the first day CH foraged almost exclusively on the branches
of V. amygdafina, a plant that was not consumed by other members of her
group. When feeding on Vernonia, she chewed the young branches, sucked
and swallowed the pith juice, and discarded the remaining fibers. During
the afternoon of the first day and the morning of the second day, CH spent
an unusually long time lying down and very little time foraging; she seemed
to have trouble defecting, and her urine seemed unusually dark. Her behav-
ior and urine color returned to normal in the afternoon of the second
day. Interestingly, CH had previously been observed swallowing leaves of

another plant, Lippia plicata (Verbenaceae), presumably also for medicinal
purposes (Takasaki and Hunt, 1987).
Huffman et al. (1993) observed another adult female, dubbed FT, for a
total of about 5 hr, over a period of 2 days. During this time she fed on
clay from a termite mound and on at least four species of plants, among
them V.amygdahza. Like CH, FT did not consume the leaves of V. amygda-
lina, but instead chewed, sucked, and discarded young branches. A fecal
sample obtained during the afternoon of the first day was yellowish and
liquid, and contained 130 eggs of the intestinal nematode Ternidens sp. per
gram of feces. A second stool sample, obtained in the morning of the next
day, was solid, and contained only 15 eggs per gram of feces. Huffman et
al. (1993) also presented data on normal infestation levels, based on re-
peated fecal sampling of seven other individuals. However, these data were
not detailed enough to determine whether the decrease in Ternidens eggs
in FT was within the normal range of daily variation, in the absence of
Vernonia or clay consumption.
Several factors, including herbivory, can affect the production and distri-
bution of secondary metabolites within individual plants (Rhoades, 1979;
Karban and Myers, 1989), so care must be taken to ensure that the leaves
used for analysis are a suitable representation of those consumed (e.g.,
Huffman et al., 1996). Representative samples collected from the actual
Vernonia plants consumed by FT showed that, contrary to expectations,
the concentration of two biologically active compounds, vernonioside B1
and vernodalin, was higher in young leaves than in young stems (Huffman
et al., 1993; Ohigashi et al., 1994). Several other secondary metabolites have
been extracted from V. amygdalina (Kupchan et al., 1969; Ganjian et al.,
1983; Ohigashi et al., 1991a,b; Jisaka et al., 1992a, 1993a), and, as expected,
the biological activity of these compounds is diverse. Extracts from V.
amygdalina deter insect herbivory (Ganjian et al., 1983), are toxic to schisto-
somes (Jisaka et al., 1992b; Ohigashi et al., 1994), and have antitumoral
(Kupchan et al., 1969), antibacterial (Jisaka et al., 1993b), and antioxidant
(Igile et al., 1994) properties. Like Aspilia, V. amygdalina is used widely
in Africa by humans as a medicinal plant for a variety of ailments.
C. RUBIA
Wrangham (1995) examined the relationship between a parasitic tape-
worm infection and the peculiar habit of leaf swallowing by chimpanzees.
Fecal droppings containing whole leaves of Aneilema aequinoctiafe (Com-
melinaceae) and Rubia cordifolia (Rubiaceae) were found sporadically
throughout 6 years. Tapeworm fragments were detected in these droppings
on 14 occasions, spanning 7 months. During this time, the frequency of

droppings containing tapeworm fragments and whole leaves was signifi-
cantly greater than expected. Wrangham (1995) concluded that heavily
infected chimpanzees purposely swallow whole leaves, which cause the
shedding of tapeworm proglottides. However, Wrangham conceded that,
because proglottid shedding is part of tapeworm’s normal life cycle, leaf
swallowing is not necessarily an effective method of tapeworm control.
Another study (Messner and Wrangham, 1996), also involving R. cordi-
folia, is the only one so far in which the biological activity of a presumed
medicinal plant has been compared with that of other plants comprising
the regular diet of chimpanzees. Messner and Wrangham compared R.
cordifolia to six other plants, but found no differences in their toxicity to
free-living adults or larvae of Strongyloides spp.
V. SKEPTICISM
Although the idea of therapeutic self-medication in animals has been

discussed for over a decade, only in the semipopular literature have we
seen a healthy dose of skepticism (Sapolsky 1994, his pun). Sapolsky raised
three main concerns: the absence of controls with which to compare the
biological effects of these presumed medicinal plants, the lack of and need
for studies in vivo, and the absence of clearly demonstrated behavioral
mechanisms by which therapeutic self-medication can arise and be main-
tained in a population. I deal with the first two points here, and with
behavioral mechanisms in the subsequent section.
Currently, there is adequate evidence that some plants are consumed
under unusual circumstances, and that the leaves or roots of these plants
have secondary compounds with uterotonic, antiviral, antibacterial, andlor
anthelminthic properties. Little else is required if the goal is merely to
identify bioactive compounds present in chimpanzees’ diets. However, if
we consider that all plants have secondary metabolites, and that the main
role of these chemicals is protection from herbivores, fungi, and bacteria,
it is not particularly surprising to find that, for any given plant, even if
selected at random, some of these secondary metabolites will be biologically
active. The presence of bioactive secondary metabolites in suspected medic-
inal plants is therefore not conclusive evidence. A t best, it can be concluded
only that, when ill, some chimpanzees deviate from their normal diet.
To demonstrate that plants are consumed to deal with specific diseases,
we need to know the ailment affecting an individual, and show not only
that the plant parts consumed alleviate this condition, but that their effect
is greater than that of plants that make up a healthy individual’s regular diet.
Alternatively, it is probably difficult to determine the exact relationships

between specific diseases and their corresponding medicines, so, when ill,
individuals may simply choose plants containing a wide spectrum of biologi-
cal effects, akin to taking a general antibiotic. In these cases, these plants
would not be necessarily effective against the disease, but they would be
expected to contain more and/or stronger biologically active compounds
than plants normally consumed. Third, therapeutic medicines may be con-
sumed to alleviate the symptoms of the disease, and have no effects at all
on the pathogen itself. In such cases medicinal plants would be expected
to contain analgesics and other compounds that affect only the host, and
not the parasite. Whether an animal consumes medicinal plants as general
antibiotics, or to deal with specific diseases, or merely the symptoms of
disease, comparisons with other plants comprising the animal’s regular diet
are needed before firm conclusions can be drawn.
There has been only one study in which such controls have been used.
Messner and Wrangham (1996) found no differences in the biological activ-
ity against Strongyloides spp. between methanol extracts of R. cordifolia and
six other plants regularly eaten by chimpanzees. Messner and Wrangham
pointed out that these results do not necessarily mean that R. cordifolia
does not affect intestinal nematodes because (1) the extraction method
may have failed to obtain all bioactive compounds, (2) the nematodes used
in this test were not the parasitic stage, but rather free-living adults and
larvae, and (3) in vitro tests are a poor replacement for the complex interac-
tions that may occur in vivo. Unfortunately, these same caveats would
have also been valid had Messner and Wrangham (1996) found significant
differences between the biological activity of R. cordifolia and the six
other plants.
However, Messner and Wrangham (1996) did raise an important point:
experimental tests in vivo are needed. Understandably, in vivo trials may
not be practical or ethical in wild chimpanzees, so they must take a back
seat to observational and to phytochemical studies. Furthermore, chimpan-
zees at Gombe and Mahale have been the focus of ongoing research for
several decades, so it would be undesirable to carry out invasive experi-
ments with these populations. Nevertheless, such studies are necessary if
we wish to understand the effects of these plants, and could perhaps be
conducted with chimpanzee populations elsewhere, captive chimpanzees,
or other primates.
VI. BEHAVIORAL
MECHANISMS
Therapeutic self-medication requires fairly complex and interesting be-

havioral mechanisms of food selection. However, as Sapolsky (1994)
SELF-MEDICATION I N WILD ANIMALS 305
pointed out, this aspect has received little attention in the self-medication
literature. In this section I discuss food selection mechanisms that may be
involved in therapeutic self-medication, dealing first with individual learning
and then with social learning. The following discussion is not meant to be
a critical review of the literature on the mechanisms of food selection, and
is based largely on several comprehensive reviews (Rozin and Kalat, 1971:
Galef, 1976, 1996; Rozin, 1976; Bandura, 1977).
A. INDIVIDUAL LEARNING
If therapeutic self-medication is learned individually, a series of steps
must take place. First, upon infection by a parasite, or when the infection
reaches a particularly uncomfortable level, the host must begin sampling
unfamiliar food items, and in some cases overcome their natural aversion
to new foods and bitter-tasting plants (Garcia and Hankins, 1975). The
infected animal must then chance upon a medicinal plant and fortuitously
consume it in sufficient quantities for the plant to be effective against the
offending parasite. Upon recovery, which may occur many hours after the
medicinal plant was consumed, the animal must return to its regular diet.
Several relevant mechanisms have been demonstrated experimentally in
rats (Rafrus norvegicus), apparently the preferred species for experimental
work on the mechanisms of food selection. Richter (1943) showed that rats
faced with a limited number of single-nutrient food items were able to
select a nutritionally adequate diet. Furthermore, rats with deficiencies of
specific nutrients were able to obtain these nutrients by altering their diets.
In theory, the ability to obtain a balanced diet may be the result of
specific hungers, under strict genetic control, without the flexibility of
learned behavior. For every single nutrient it requires, an animal could have
the ability to sense physiological deficiencies, and recognize its presence in
food. The animal would need the physiological mechanisms to identify
each nutrient individually, presumably by taste or smell, and to monitor
constantly for deficiencies. This would mean a separate monitoring and
identification system for each essential amino acid, vitamin, and mineral.
Clearly, such a system would seldom be necessary or particularly useful.
Specific hungers do exist, but they are limited to extremely important
nutrients. For example, carnivores need not be concerned with individual
nutrients, as each prey item provides them with a full range of essential
nutrients in an adequate balance. Domestic chickens (Gullus gallus) have
a specific hunger for water that includes the ability to taste it, but not
identify it visually (Hunt and Smith, 1967). It is therefore possible to have
dehydrated newly hatched chicks walking through water and being com-
pletely unaware of the obvious solution to their problem. They soon learn
to identify it visually, but only after having pecked at it and tasted it. The
adult blowfly (Phormia regina) has specific hungers for sugar, water, and
salt, and its feeding response is under direct control of separate internal
and external chemoreceptors for each of these nutrients (Dethier, 1969).
Most mammals have a specific hunger for water (Rozin, 1976). Among
primates, sodium hunger has been shown in humans (Beauchamp et al.,
1990) and baboons (Denton et al., 1993). Given the large number and
unusual nature of chemicals involved, it is doubtful that specific hungers
play a role in therapeutic self-medication. Whereas the rule of thumb “when
suffering from dehydration, drink plenty of water” could be solely under
genetic control, the directive “when suffering from malaria, drink water
from under a cinchona tree, or better yet, chew on the tree’s bark” is far
more complex and more likely to be a learned response.
In the absence of specific hungers, diet selection must be the result of
learned preferences for suitable diets, or learned aversions for inadequate
diets. Rozin (1967) observed that the behavior of rats toward their regular,
palatable, but nutritionally deficient diet was similar to their behavior to-
ward highly unpalatable diets. In both cases rats approached the food tray
tentatively, spilled some food, and then moved away and chewed on some
inedible object. These rats were quick to consume any new diet, regardless
of whether it was nutritionally adequate. These observations showed that
diet changes in rats are not the result of learned preferences for new or
nutritionally adequate diets, but rather the result of a learned aversion for
the initial, nutritionally deficient diet.
Whether diet changes are the result of aversions or preferences, several
problems arise when attempting to apply diet selection mechanisms to
therapeutic self-medication. Self-medication requires that animals consume
unusual food items temporarily and maybe exclusively, and then revert to
their regular diets. This process does not entail a permanent preference
for the alternative diet, or a permanent aversion to the regular diet. It may
be possible to explain self-medication in terms of a dual aversion, first to
the regular diet, and then to the medicine. However, this would be possible
only if the initial aversion to the regular diet is strong enough to cause the
initial shift, yet mild enough to be subsequently forgotten.
Another problem for a self-medicating animal is learning to associate its
recovery with its diet over the past several hours, and not with other events
that may have occurred concurrently. Experiments in rats have shown that
aversions do not develop to the location or the type of food container, but
are limited to the nutrient-deficient food itself. Garcia and Koelling (1966)
exposed rats to taste, sound, and light stimuli, paired with either electrical
shocks or poisoning, via injection or radiation. Poisoned rats learned to
avoid the taste, but not the sound or light, whereas shocked rats developed
an aversion to the light and sound, but not the taste. This experiment
showed that certain associations between stimuli are learned more easily
than others. Specifically, visceral responses are more likely to be associated
with food consumption, which suggests that intestinal ailments would be
more likely associated with medicines consumed orally, and external ail-
ments with topical medicines.
A self-medicating animal must learn not only to associate its recovery
with food, but also to determine which food, out of the several items that
may have been consumed, is responsible for its recovery. While trying to
find an adequate diet, rats do not sample alternative foods randomly. In-
stead, their sampling pattern facilitates the possibility of associating recov-
ery with a specific item. Feeding bouts are temporally separated and include
only one new food source, and only a few new foods are sampled each day
(Rozin, 1969). So far, no studies have dealt specifically with the food sam-
pling behavior of sick chimpanzees.
B. SOCIAL
LEARNING
Social interactions play an important role in every aspect of chimpanzee
behavior; hence much of their knowledge concerning ways to interact with
their environment does not necessarily come from individual experience.
Food preferences may be influenced by the food choices of conspecifics,
so self-medication may not be learned de n o w by every individual.
Although the effects of social learning on self-medicating chimpanzees
have not been studied yet, several potentially relevant mechanisms have
been demonstrated experimentally in other species. For example, in rats,
protein deficiency increases the effect of social learning on diet preferences
(Galef et uf., 1991). These results suggest that sick animals in poor condition
may be more likely to alter their diets. It has also been demonstrated that
rats are more likely to learn the unfamiliar, rather than the familiar or
usual diet of their demonstrators (Galef, 1993). In spotted hyenas (Crocutu
crocutu), individually learned food aversions can be attenuated and even
eliminated by the observation of conspecifics feeding on the avoided food
(Yoerg, 1991). In red-winged blackbirds (Agefuiusphoeniceus), aversions
can develop from observing conspecifics becoming ill after consuming a
food item (Mason and Reidinger, 1982), which shows that blackbirds learn
to associate visual cues of illness in conspecifics with particular foods.
However, there have not been any studies demonstrating the reverse: the
ability to associate the recovery of a sick conspecific with its consumption
of a specific food item.
VII. IMPLICATIONS
ECOLOGY
A. CONSERVATION
The many ways in which animals interact with their environment are
seldom readily apparent. Self-medication in wild animals may be one such
relationship that we are only now beginning to recognize, much less under-
stand. This lack of knowledge further demonstrates that we do not have the
ability to reconstruct natural ecosystems; therefore, conservation ecology
requires the protection of entire communities, with all their species and
interrelationships intact (Clayton and Wolfe, 1993). Captive breeding pro-
grams can be successful at preserving individual species, but do not preserve
the relationships of an animal with its natural environment. Hence, the
preservation of species should be considered as an important fail-safe op-
tion, but only part of more holistic conservation ecology strategies.
The existence of self-medication may also affect the ease with which
animals can be reintroduced to the wild, especially in cases for which
knowledge about self-medication is culturally transmitted. Depending on
the extent to which self-medication and other parasite avoidance behaviors
are culturally transmitted, naive animals being returned to their natural
environment may be subjected to unusually high parasite loads. The nega-
tive effects of parasites may be further exacerbated in host populations
with heavily fragmented habitats, a factor that should be considered in
designing biological reserves (Loye and Carroll, 1995).
THEORY
B. FORAGING
Optimal foraging models were initially based on the assumption that the
primary goal of foragers was to maximize net energy or protein intake
(Stephens and Krebs, 1986). Other factors, such as the risk of predation
(e.g., Milinski and Heller, 1Y78; Sih, 1980; Edwards, 1983; Abrahams and
Dill, 1989; Lima and Dill, 1990), and the effects of intraspecific (e.g., Baker
et al., 1981; Milinski, 1982) and interspecific (e.g., Millikan el al., 1985)
competition, have been subsequently incorporated into diet choice models,
and increased their predictive powers. The effects of parasitic infections
on foraging behavior have also been examined (e.g., Crowden and Broom,
1980; Milinski, 1984; Giles, 1987). So far, however, diet choice has been
largely ignored as a way in which potential hosts could actively reduce para-
sitism.
Diet choice has also evolved under the selection pressures brought about
by parasites. It is therefore reasonable to expect that optimal diets are not
only nutritionally and energetically adequate, but also take into account
the potentially detrimental effects of parasites. Hosts could alter their diets
to counteract the risks of parasitism by (1) avoiding food items that are
common sources of parasites, (2) selectively consuming certain food items
to decrease their susceptibility to parasites, and (3) acitvely consuming foods
with antiparasitic properties upon infection (Lozano, 1991). As evidence of
self-medication continues to accumulate, future diet choice models must
consider the effects of parasitism.
c. BEHAVIORAL OF FOOD
MECHANISMS SELECTION
Diet preferences evolve under many constraints, including parasitism, so
it is easy to envision that plants with antiparasitic properties may become
part of an animal’s regular diet. In contrast, therapeutic self-mediation
requires a sick animal to deviate away from its regular diet, and seek
and consume medicinal substances. It requires intentionality, and is, by
necessity, a learned behavior. It is sometimes difficult for even well-
documented phenomena to be generally accepted without clear mechanisms
by which they can occur. So far, self-medication cannot .be explained in
terms of experimentally demonstrated food selection mechanisms, so it may
prove to be an interesting challenge, if it is to be demonstrated conclusively.
D. HUMAN
MEDICINE
Although estimates vary, it is generally agreed that a large proportion
of our current medicinal drugs are derived from plants (Fansworth ef al.,
1985;Balandrin et al., 1985;Caldecott, 1987; McKenna, 1996). The resources
are not available to sample every species, so the identification of plants
with potential medicinal uses is a major impediment in the discovery and
development of new medicines. It has been repeatedly stated that the study
of self-medication in nonhuman animals may lead to the discovery of new
medicinal compounds (Cowen, 1990; Newton, 1991; Clayton and Wolfe,
1993; Rodriguez and Wrangham, 1993; Sapolsky, 1994). However, this is
not supported by the cases studied so far.
As noted earlier, one important reason to suspect that chimpanzees
consume certain plants for medicinal purposes is that these plants are also
used in human traditional medicine. So, in essence, it was already known
that these plants may have medicinal properties. It is possible that further
research will indeed yield new medicines. However, the search for new
pharmaceuticals cannot be considered the primary goal of this line of
research, for we would probably fare much better by exploring plants
used in traditional medicine (e.g., Johns, 1990; Johns and Chapman, 1995;
Wagner, 1995) instead of plants used by self-medicating chimpanzees.
VIII. SUMMARY
A N D CONCLUSIONS
Parasites present a ubiquitous selective force that has led to the evolution
of a vast array of behavioral adaptations. The need to avoid and reduce
parasites can affect foraging patterns and diet choice, and conceivably lead
to self-medicating behavior. Self-medication can be viewed as a specific
case of the more widespread phenomenon of chemical interactions across
trophic levels. Despite the many apparently disparate examples suggestive
of self-medication, it can take only two functionally distinct forms: preventa-
tive and therapeutic. These two processes require separate mechanisms,
and yield different and explicit predictions. By viewing it in a more general
framework, self-medication can be studied in terms of common elements,
instead of isolated examples.
Rodriguez and Wrangham (1993) proposed the term “zoopharmacog-
nosy” to describe the scientific study of the use of plants by wild animals
for the prevention and treatment of disease. Current research on therapeutic
self-medication is still solely limited to chimpanzees at Mahale and Gombe,
but work could be carried out with other populations or other taxa. The
multidisciplinary nature of this relatively new field means that problems
can be tackled from many different angles, and many avenues of research
are still open. Contributions are possible from ethologists, biochemists,
parasitologists, pharmacologists, behavioral ecologists, immunologists, psy-
chologists, and statisticians, whether working directly in the field, or simply
being aware of the possibility of self-medication in wild animals when
conducting other lines of research.
Finding out that baboons indulge in the recreational use of pharmaceuti-
cals or that chimpanzees practice a primitive form of medicine may chal-
lenge some individuals’ convictions regarding the uniqueness of humans.
Understandably, public interest is high, and discussions on self-medication
have not been limited to academic media (Bower, 1986; Cowen, 1990; Sears,
1990; Gibbons, 1992; Strier, 1993; Sapolsky, 1994). O n the other hand, most
scientists would probably consider such findings extraordinary, but not
necessarily disturbing. Scientific interest, therefore, results from more than
a mere fascination with newly discovered behaviors; as noted earlier, the
study of self-medication in wild animals may have implications for a variety
of related fields. Although several recent synopses (Wrangham and
Goodall, 1989; Newton, 1991; Huffman, 1993; Rodriguez and Wrangham,
1993; Huffman and Wrangham, 1994) have presented the evidence for self-
medication as being fairly conclusive, I must conclude that the evidence
for therapeutic self-medication in nonhumans is still only suggestive. Never-
theless, the possibility of prophylactic or therapeutic self-medication in
nonhumans remains a fascinating prospect, and is certainly a fertile ground
for further innovative research.
Acknowledgments
I thank C-L. Adams. D. Kristan, R. E. Lemon, L. Lefebvre, M. Milinski, A. P. Moller, J.

Mountjoy, M. Sclafani. P. J. B. Slater, and M. Zuk for taking the time to discuss ideas with
me, guiding me toward relevant literature, and forcing me to find better ways to express
myself. Financial support was provided by FCAR.
References
Abrahams, M. V., and Dill, L. (1989). A determination of the energetic equivalence of the
risk of predation. Ecology 70, 999-1007.
Bagg. A. M. (1952). Anting not exclusively an avian trait. J. Mammal. 33, 243.
Baker, M. (1996). Fur rubbing: Use of medicinal plants by capuchin monkeys (Cebus capuci-
nus). Am. J. Primatol. 38, 362-370.
Baker, M. C., Belcher, C. S., Deutsh, L. C., Sherman, G. L., and Thompson, D. B. (1981).
Foraging success in junco flocks and the effect of social hierarchy. Anim. Behav. 29,
137- 142.
Balandrin, M. F., Klocke. J. A,, Wurtele, E. S., and Bollinger, W. H. (1985). Natural plant
chemicals: Sources of industrial and medicinal materials. Science 228, 1154-1 160.
Bandura. A. (1977). “Social Learning Theory.” Prentice-Hall, Englewood Cliffs, NJ.
Beauchamp. G. K., Bertino, M., Burke, D., and Engelman, K. (1990). Experimental sodium
depletion and salt taste in normal human volunteers. Am. J. Clin. Nutr. 51, 881-889.
Behnke, J. M., and Barnard, C. J. (1990). Coevolution of parasites and hosts: Host-parasite arms
races and their consequences. In “Parasites: Immunity and Pathology” (J. M. Behnke, ed.),
pp. 1-22. Taylor & Francis, London.
Bethel, W. M.. and Holmes. J. C. (1973). Altered evasive behavior and responses to light in
amphipods harboring acanthocephalan cystacanths. J. Parasitol. 59, 945-956.
Bohlmann. F., Gerke, T.. Jakupovic, J., Borthakur, N., King, R. M., and Robinson, H. (1984).
Diterpene lactones and other constituents from Wedelia and Aspilia species. Phytochemis-
try 23, 1673-1676.
Bower, B. (1986). Herbal medicine: R, for chimps? Sci. News 129,38-39.
Brassard, P., Rau, M. E.. and Curtis, M. A. (1982). Parasite-induced susceptibility to predation
in diplostomiasis. Parasitology 85, 495-501.
Brower. J. V. (1958). Experimental studies of mimicry in some North American butterflies.
I: The Monarch, Danausplexippus, and Viceroy, Limenitis archippus archippus. Evolution
(Lawrence, Kans.) 12, 32-47.
Caldecott, J. (1987). Medicine and the fate of tropical forests. Br. Med. J. 295, 229-230.
Campbell, B. C.. and Duffey, S. S. (1979). Tomatine and parasitic wasps: Potential incompatibil-
ity of plant antibiosis with biological control. Science 205, 700-702.
Cannon, W. B. (1935). Stresses and strains of homeostasis. Am. J. Med. Sci. 89, 1-14.
Clark, C. C., Clark. L., and Clark, L. (1990). “Anting” behavior by common grackles and
European starlings. Wilson Bull. 102, 167-169.
Clark, L., and Mason, J. R. (1985). Use of nest material as insecticidal and anti-pathogenic
agents by the European starling. Uecologia 67, 169-176.
Clark. L., and Mason, J. R. (1988). Effect of biologically active plants used as nest material
and the derived benefit to starling nestlings. Uecologia 77, 174-180.
Clayton, D. H.. and Vernon, J. G. (1993). Common grackle anting with lime fruit and its
effect on ectoparasites. Auk 110, 951-952.
Clayton, D. H., and Wolfe, N. D. (1993). The adaptive significance of self-medication. Trends
Ecol. Evol. 8, 60-63.
Clunie, F. (1976). Jungle mynah “anting” with a millipede. Notornis 23, 77.
Cowen. R. (1990). Medicine on the wild side: Animals may rely on a natural pharmacy. Sci.
News 138,280-282.
Crowden. A. E., and Broom, D. M. (1980). Effects of the eyefluke. Diplostomum spafhaceum,
on the behaviour of dace (Letrciscits leuciscus). A n i m Behav. 28, 287-294.
Daly. J. W.. Garraffo, H. M.. Spande, T. F., Jaramillo. C., and Rand, A. S. (1994). Dietary
source for skin alkaloids of poison frogs (Dendrobatidae)? J . Cheni. Ecol. 20, 943-955.
Davies, A. G.. and Baillie, I. C. (1988). Soil-eating by red leaf monkeys (PrcJsbytisruhicundu)
in Sabah, Northern Borneo. BioTropica 20,252-258.
Denton. D. A,. Eichberg, J. W.. Shade. R.. and Weisinger, R. S. (1993). Sodium appetite in
response to sodium deficiency in babbons. Am. J . Physiol. 264, R539-RS43.
Dethier, V. G. (1969). Feeding behavior of the bowfly. A d v . Study Behav. 2, 111-266.
Duffey, S. S. (1980). Sequestration of plant natural products by insects. Annu. Rev. Entoviol.
25,447-477.
Edwards, J. (1983). Diet shifts in moose due to predator avoidance. Oecologia 60, 185-189.
Ehrlich, P. R.. Dobkin. D. S.. and Wheye, D. (1986). The adaptive significance of anting. Aitk
103, 835.
Endler, J. A. (1980). Natural selection on color patterns in Poecilia reticulafa. Evoliifion
(Lawrence, Kans.) 34, 76-91.
Fansworth, N. R.. Akerele, O., Bingel, A. S.. Soejarto, D. D., and Guo, Z . G. (1985). Medicinal
plants in therapy. Bull. W.H.O.63, 83-97.
Fauth, P. T., Krementz. D. G., and Hines, J. E. (1991). Ectoparasitism and the role of green
nesting material in the European starling. Oecologia 88,22-29.
Galef, B. G., Jr. (1976). Social transmission of acquired behavior: A discussion of tradition
and social learning in vertebrates. A h . Sritdy Behav. 6, 77-100.
Galef. B. G., Jr. (1093). Functions of social learning about food: A causal analysis of effects
of diet novelty on diet preference. Anini. Behav. 46, 257-265.
Galef, B. G.. Jr. (1996). Social enhancement of food preferences in Norway rats: A Brief
review. In “Social Learning in Animals: The Roots of Culture” (C. M. Heyes and
B. G. Galef. Jr., eds.), pp. 49-64. Academic Press. Toronto.
Galef. B. G.. Jr.. Beck, M., and Whiskin. E. E. (IYYl). Protein deficiency magnifies social
influence on the food choices of Norway rats (Rartits norvegicus). J . Conip. Psycho/.
105,ss-59.
Ganjian, 1.. Kubo. I., and Fludzinski, P. (1983). Insect antifeedant elemanolide lactones from
Vemonia amygtlalinu. Phytocheinistry 22, 252552526,
Ganzer. U., Jakupovic. J., Bohlmann, F., and King. R. M. (1992). A highly oxygenated
germacrane from Aspilia eenii and constituents from other Aspilia species. Phytochemistry
31, 209-212.
Garcia, J., and Hankins. W. G. (1975). The evolution of bitter and the acquisition of toxiphobia.
In “Olfaction and Taste: V ” (D. A. Denton and J. P. Coghlan. eds.). pp. 39-45. Academic
Press. New York.
Garcia, J.. and Koelling. R. A. (1966). Relation of cue to consequence in avoidance learning.
Psychon. Sci. 4, 123-124.
Gibbons, A. (1992). Plants of the apes. Science 255,921.
Giles, N. (1987). Predation risk and reduced foraging activity in fish: Experiments with parasit-
ized and non-parasitized three-spined sticklebacks, Gasterosteus aculeafirs L. J. Fish B i d .
31,37-44.
Gompper, M. E., and Holyman, A. M. (1993). Grooming with Trattinnickia resin: Possible
pharmaceutical plant use by coatis in Panama. J. Trop. Ecol. 9, 533-540.
Gwinner. H. (1997). The function of green plants in nests of European starlings (Sturnus
vulgaris). Behaviour 134, 337-351.
Hamilton, W. J.. 111. Buskirk, R. E.. and Buskirk. W. H. (1978). Omnivory and utilization of
food resources by chacma baboons, Pupio ursinus. Am. Nut. 1l2, 911-924.
Hare, J. D.. and Andreadis, T. G. (1983). Variation in the susceptibility of Leptinotarsa
decemlineata (Coleoptera: Chrysomelidae) when reared on different host plants to the
fungal pathogen, Beauveria bassiana in the field and laboratory. Environ. Entomol. 12,
1891-1896.
Hart, B. L. (1990). Behavioral adaptations to pathogens and parasites: Five strategies. Neitrosci.
Biobehav. Rev. 14, 273-294.
Hauser, D. C. (1964). Anting by gray squirrels. J. Mammal. 45, 136-138.
Hay, M. E., Duffy, J. E.. Paul, V. J., Renaud, P. E., and Fenical, W. (1990). Specialist herbivores
reduce their susceptibility to predation by feeding on the chemically defended seaweed
A vrainvillea longicaulis. Limnol. Oceanogr. 35, 1734-1743.
Hill. G. E. (1994). Geographic variation in male ornamentation and female mate preference
in the house finch: A comparative test of models of sexual selection. Behav. Ecol. 5,64-73.
Hudson, J. B., Graham, E. A,, Fong, R.. Finlayson, A. J., and Towers, G. H. N. (1986).
Antiviral properties of thiarubrine-A, a naturally occurring polyine. Planta Med., 51-54.
Huffman, M. A. (1993). An investigation of the use of medicinal plants by chimpanzees:
Current status and future prospects. Primate Res. 9, 179-187.
Huffman. M. A,, and Seifu, M. (1989). Observations on the illness and consumption of a
possibly medicinal plant Vernonia amygdalina (Del.), by a wild chimpanzee in the Mahale
Mountains National Park. Tanzania. Primates 30, 51-63.
Huffman. M. A,. and Wrangham, R. W. (1994). Diversity of medicinal plant use by chimpanzees
in the wild. I n “Chimpanzee Cultures” (R. W. Wrangham. W. C. McGrew, F. B. De
Waal. and P. G. Heltne. eds.), pp. 129-148. Harvard University Press, Cambridge, MA.
Huffman, M. A,, Gotoh. S., Izutsu, D., Koshimizu, K., and Kalunde, M. S. (1993). Further
observations on the use of the medicinal plant Vernoniu umygdalina (Del.), by a wild
chimpanzee, its possible effect on parasite load, and its phytochemistry. Afr. Stud. Monogr.
14,227-240.
Huffman, M. A,, Gotoh, S., Turner, L. A,, Hamai, M., and Yoshida. K. (1997). Seasonal
trends in intestinal nematode infection and medicinal plant use among chimpanzees in
the Mahale Mountains, Tanzania. Primates 38, 111-125.
Huffman, M. A,, Page, J. E., Sukhdeo, M. V. K., Gotoh, S., Kalunde, M. S., Chandrasiri, T.,
and Towers, G. H. N. (1996). Leaf-swallowing by chimpanzees: A behavioral adaptation
for the control of strongyle nematode infections. In/. J . Primatol. 17, 475-503.
Hunt, G. L., and Smith, W. J. (1967). Pecking and initial drinking responses in young domestic
fowl. J. Comp. Physiol. Psychol. 64, 230-236.
Igile, G. 0..Oleszek, W., Jurzysta, M., Burda, S., Fafunso, M.. and Fasanmade, A. A. (1994).
Flavonoids from Vernnnia amygdalinu and their antioxidant activities. J. Agric. Food
Chem. 42,2445-2448.
Janzen. D. H. (1978). Complications in interpreting the chemical defenses of trees against
tropical arboreal plant-eating vertebrates. I n “The Ecology of Arboreal Foliovores”
(G. G. Montgomerie, ed.), pp. 73-84. Smithsonian Institution Press, Washington, DC.
Jisaka, M., Ohigashi, H., and Takagaki, T., Nozaki, H., Tada, T. Hirota, M., hie, R., Huffman,
M. A., Nishida, T.. Kaji. M., and Koshimizu, K. (1992a). Bitter steroid glycosides, vernonio-
sides A,, A?, A3. and related B, from a possible medicinal plant, Vernonia amygdalinu,
used by wild chimpanzees. Tetrahedron 48,625-632.
Jisaka, M., Kawanaka, M., Sugiyama, H., Takegawa, K.. Huffman, M., Ohigashi. H., and
Koshimizu, K. (1992b). Antischistosomal activities of sesquiterpene lactones and steroid
glucosides from Vernonia amygdalina, possibly used by wild chimpanzees against parasite-
related diseases. Biosci. Biotechnol. Biochem. 56, 845-846.
Jisaka, M., Ohigashi. H., and Takegawa. K.. Hirota, M., hie, R., Huffman, M. A,, and Koshim-
izu. K. (1993a). Steroid glucosides from Vernonia amygdalina, a possible chimpanzee
medicinal plant. Phytochemistry 34, 409-413.
Jisaka, M., Ohigashi, H.. and Takegawa, K., Huffman, M. A,, and Koshimizu, K. (1993b).
Antitumoral and antimicrobial activities of bitter sesquiterpene lactones of Vernonia
amygdalina, a possible medicinal plant used by wild chimpanzees. Biosci. Biotechnol.
Biochem. 57, 833-834.
Johns, T. (1990). “With Bitter Herbs they Shall Eat It: Chemical Ecology and the Origins of
Human Diet and Medicine.” University of Arizona Press, Tucson.
Johns, T., and Chapman, L. (1995). Phytochemicals ingested in traditional diets and medicines
as modulators of energy metabolism. Recent Adv. Phytochem. 29, 161-188.
Johns. T., and Duquette, M. (1991). Detoxification and mineral supplementation as functions
of geophagy. Am. J. Clin. Nutr. 53, 448-456.
Karban, R., and Myers, J. H. (1989). Induced plant responses to herbivory. Annu. Rev. Ecol.
Syst. 20, 331-348.
Kodric-Brown. A. (1989). Dietary carotenoids and male mating success in the guppy: An
environmental component to female choice. Behav. Ecol. Sociobiol. 25, 393-401.
Kopin, I. J. (1995). Definitions of stress and sympathetic neuronal responses. Ann. N. Y. Acad.
Sci. 771, 19-40.
Krebs, J. R. (1980). Optimal foraging, predation risk and territory defense. Ardea 68,83-90.
Kreulen, D. A. (1985). Lick use by large herbivores: A review of benefits and banes of soil
consumption. Mamm. Rev. 15, 107-123.
Krischik, V. A., Barhosa, P., and Reichelderfer. C. F. (1988). Three trophic level interactions:
Allelochemicals. Manduca sexta (L.), and Bacillus thrtringiensis var. kurstaki Berliner.
Environ. Entomol. 17, 476-482.
Kupchan, S. M., Hemingway, R. J., Karim, A,, and Werner, D. (1969). Tumor inhibitors.
XLVII. Vernodalin and vernomygdin, two new cytotoxic sesquiterpene lactones from
Vernonia amygdalina Del. J. Org. Chem. 34, 3908-391 1.
Lima, S. L., and Dill, L. M. (1990). Behavioral decisions made under the risk of predation:
A review and prospectus. Can. J. Zool. 68,619-640.
Longino, J. T. (1984). True anting by the capuchin (Cebus capucinus). Primates 25,243-245.
Loye, J., and Carroll. S. (1995). Birds, bugs, and blood: Avian parasitism and conservation.
Trends Ecol. Evol. 10,232-235.
Lozano, G. A. (1991). Optimal foraging theory: A possible role for parasites. Oikos63,391-395.
Lozano, G. A. (1994). Carotenoids, parasites, and sexual selection. Oikos 70, 309-31 1.
Mahry, T. J., Gill, J. E., Burnett, W. C., Jr., and Jones, S. B., Jr. (1977). Antifeedant sesquiter-
pene lactones in the Compositae. ACS Symp. Ser. 62, 165-178.
Mahaney, W. C. (1993). Scanning electron microscopy of earth mined and eaten by mountain
gorillas in the Virunga mountains, Rwanda. Primates 34, 31 1-319.
Mahaney, W. C., Aufreiter, S., and Hancock, R. G. V. (199%). Mountain gorilla geophagy:
A possible seasonal behavior for dealing with the effects of dietary changes. Int. J.
Primatol. 16, 475-487.
Mahaney, W. C., Hancock, R. G. V.. Aufreiter, S., and Huffman, M. A. (1996). Geochemistry
and clay mineralogy of termite mound soil and the role of geophagy in chimpanzees of
the Mahale mountains, Tanzania. Primates 37, 121-134.
Mahaney. W. C.. Hancock, R. G. V., and Inoue, M. (1993). Geochemistry and clay mineralogy
of soils eaten by Japanese macaques. Primates 34, 85-91.
Mahaney, W. C., Milner, W. M.,Sanmugadas, K.,Hancock, R. G. V.. Aufreiter.S., Wrangham,
R., and Pier. H. W. (1997). Analysis of geophagy soils in Kinale Forest, Uganda. Primates
38, 159-176.
Mahaney, W. C.. Stambolic, A,, Knezevich, M.. Hancock. R. G. V.. Aufreiter. S., Sanmugadas.
K., Kessler. M. J., and Grynpas, M. D. (1995a). Geophagia amongst rhesus macaques on
Cay0 Santiago, Puerto Rico. Primates 36, 323-333.
Mahaney. W. C., Watts, D. P., and Hancock, R. G. V. (1990). Geophagia by mountain gorillas
(Gorilla gorillu heringei) in the Virunga mountains. Rwanda. Primates 31, 113-120.
Maitland, D. P. (1994). A parasitic fungus infecting yellow dungflies manipulates host perching
behaviour. Proc. R. Soc. London Ser. B 258, 187-193.
Mason, J. R., and Reidinger, R. F., Jr. (1982). Observational learning of food aversions in
red-winged blackbirds (Agelaius phoeniceus). Auk 99, 548-554.
McGrath, J. E. (1970). A conceptual formulation for research on stress. In “Social and
Psychological Factors in Stress’’ (J. E. McGrath, ed.), pp. 10-21. Holt, Rinehart & Win-
ston, New York.
McKenna, D. J . (1996). Plant hallucinogens: Springboards for psychotherapeutic drug discov-
ery. Behav. Brain Res. 73, 109-116.
Messner, E. J., and Wrangham. R. W. (1996). In vitro testing of the biological activity of
Rubia cordifolia leaves on primate Strongyloides species. Primates 37, 105-108.
Milinski, M. (1982). Optimal foraging: The influence of intraspecific competition on diet
selection. Behav. Ecol. Sociohiol. 11, 109-1 15.
Milinski, M. (1984). Parasites determine a predator’s optimal feeding strategy. Behav. Ecol.
Sociohiol. 15, 35-37.
Milinski, M.. and Bakker. T. C. M. (1990). Female sticklebacks use male coloration in mate
choice and hence avoid parasitized males. Nature (London) 344,330-333.
Milinski, M.. and Heller, R. (1978). Influence of a predator on the optimal foraging behaviour
of sticklebacks (Gasterosteus acitleatus L.). Nature (London) 275, 642-644.
Millikan, G. C.. Caddis, P.. and Pulliam, H. R. (1985). Interspecific dominance and the foraging
behavior of juncos. Anim. Behav. 33,428-435.
Mims, C. A., Dimmock, N., Nash. A., and Stephen. J. (1995). “Mims’ Pathogenesis of Infectious
Disease.”- 4th ed. Academic Press, London.
Maller, A. P., Dufva, R., and Allander, K. (1993). Parasites and the evolution of host social
behavior. Adv. Study Behav. 22, 65-102.
Moore, J., and Gotelli, N. J. (1990). A phylogenetic perspective on the evolution of altered
host behaviours: A critical look at the manipulation hypothesis. In “Parasitism and Host
Behaviour” (C. J. Barnard and J. M. Behnke, eds.). pp. 193-233. Taylor & Francis,
London.
Newton, P. (1991). The use of medicinal plants by primates: A missing link: Trends Ecol.
Evol. 6, 297-299.
Newton, P. N.. and Nishida, T. (1990). Possible buccal administration of herbal drugs by wild
chimpanzees. Pan troglodytes. Anim. Behav. 39, 798-801.
Oates, J. F. (1978). Water-plant and soil consumption by Guereza monkeys (Colobusguereza):
A relationship with minerals and toxins in the diet? BioTropica 10, 241-253.
Ohigashi, H., Takagaki, T., Koshimizu, K., Watanabe, K.. Kaji, M., Hoshino, J.. Nishida, T.,
Huffman, M. A.. Takasaki. H., Jato. J.. and Muanza. D. N. (1991a). Biological activities
of plant extracts from tropical Africa. Afr. Stud. Monogr. U,201-210.
Ohigashi, H., Jisaka. M., Takagaki, T., Nozaki, H., Tada, T., Huffman, M. A,, Nishida. T.,
Kaji. M.. and Koshimizu. K. (1991b). Bitter principle and a related steroid glucoside from
Vernonia arnygdalina, a possible medicinal plant for wild chimpanzees. Agric. B i d . Chem.
55, 1201-1203.
Ohigashi, H., Huffman. M. A,, Izutsu, D., Koshimizu, K., Kawanaka, M.. Sugiyama. H., Kirby,
G . C., Warhurst, D. C., Allen, D., Wright, C. W., Phillipson, J. D., Timon-David, P.,
Delmas. F., Elias, R., and Balansard. G . (1994). Toward the chemical ecology of medicinal
plant use in chimpanzees: The case of Vernonia arnygdalina, a plant used by wild chimpan-
zees possibly for parasite-related diseases. J . Chern. Ecol. 20, 541-553.
Page. J. E., Balza. F.. Nishida, T., and Towers, G . H. N. (1992). Biologically active diterpenes
from Aspilia rnossarnbicensis, a chimpanzee medicinal plant. Phytochemistry 31, 3437-
3439.
Pennings, S. C. (1994). Interspecific variation in chemical defenses in the sea hares (Opistho-
branchia: Anaspidea). J. Exp. Mar. B i d . Ecol. 180, 203-219.
Phillips-Conroy, J. E. (1986). Baboons, diet. and disease: Food plant selection and schistosomia-
sis. In “Current Perspectives in Primate Social Dynamics” (D. M. Taub and F. A. King,
eds.), pp. 287-304. Van Nostrand-Reinhold, New York.
Phillips-Conroy, J. E., Knopf, P. M. (1986). The effects of ingesting plant hormones on
schistosomiasis in mice: an experimental study. Biochemical Systematics and Ecology
14, 637-645.
Potter, E. F. (1970). Anting in wild birds. its frequency and probable purpose. Auk 87,692-713.
Price, P. W., Bouton, C. E., Gross, P., McPheron. B. A., Thompson, J. N., and Weis, A. E.
(1980). Interactions among three trophic levels: Influence of plants on interactions be-
tween insect herbivores and natural enemies. Annu. Rev. Ecol. Syst. ll, 41-65.
Rhoades, D. F. (1979). Evolution of plant chemical defenses against herbivores. In “Herbi-
vores: Their Interaction with Secondary Plant Metabolites” (G. A. Rosenthal and
D. H. Janzen, eds.), Chapter 1, pp. 3-54. Academic Press, New York.
Richter, C. P. (1943). Total self-regulatory functions in animals and human beings. Harvey
Lect. 38, 63-103.
Rodriguez, E., and Wrangham, R. (1993). Zoopharmacognosy: The use of medicinal plants
by animals. Recent Adv. fhytochern. 27, 89-105.
Rodriguez, E., Aregullin, M., Nishida, T., Uehara, S., Wrangham, R.. Abramowski, Z., Finlay-
son, A., and Towers, G . H. N. (1985). Thiarubrine-A, a bioactive constituent of Aspilia
(Asteraceae) consumed by wild chimpanzees. Experientia 41, 419-420.
Rothschild, M. (1972). Some observations on the relationship between pants, toxic insects
and birds. In “Phytochemical Ecology” (J. B. Halborne, ed.), pp. 1-12. Academic
Press, London.
Rozin, P. (1967). Specific aversions as a component of specific hungers. J. Comp. fhysiol.
Psychol. 64, 237-242.
Rozin, P. (1969). Adaptive food sampling patterns in vitamin deficient rats. J. Comp. Physiol.
fsychol. 69, 129-132.
Rozin. P. (1976). The selection of food by humans, rats, and other animals. Adv. Study Behav.
6, 21-76.
Rozin, P., and Kalat, J. W. (1971). Specific hungers and poison avoidance as adaptive specializa-
tions of learning. Psychol. Rev. 78, 459-486.
Sapolsky, R. M. (1994). Fallible instinct: A dose of skepticism about the medicinal “knowledge”
of animals. Sciences 34, 13-15.
Sears, C. (1990). The Chimpanzee’s medicine chest. New Sci. 1728, 42-44.
Selye, H. (1976). “Stress in Health and Disease.” Butterworth, Boston.
Sih, A. (1980). Optimal behavior: Can foragers balance two conflicting demands? Science
210, 104-1043.
SELE-MEDICATION IN WILD ANIMALS 317
Stephens, D. W., and Krebs. J . R. (1986). “Foraging Theory.” Princeton University Press.
Princeton, NJ.
Strier. K. B. (1993). Menu for a monkey. Nat. Hist. 102, 34-43.
Takasaki, H.. and Hunt, K. (1987). Further medicinal plant consumption in wild chimpanzees?
Afr. Stiid. Monogr. 8, 125-128.
Towers, G. H. N.. Abramowski. Z., Finlayson, A., and Zucconi. A. (1985). Antibiotic properties
of thiarubrine A,. a naturally occurring dithiacyclohexadiene polyine. Planfa Med.
225-229.
Vander, A. J . (1981). “Nutrition, Stress, and Toxic Chemicals.” University of Michigan Press.
Ann Arbor.
Vermeer, D. E.. and Ferrell, R. E.,Jr. (1985). Nigerian geophagical clay: A traditional antidiar-
rheal pharmaceutical. Sciznce 227, 634-663.
Wagner. H. K. M. (1995). Immunostimulants and adaptogens from plants. Recent Adv. Phyto-
chem. 29, 1-18.
Walzer, P. D., and Genta, R. M. (1989). “Parasitic Infections in the Compromised Host.”
Dekker, New York.
Wimberger. P. H. (1984). The use of green plant material in bird nests to avoid ectoparasites.
Auk 101, 615-618.
Wrangham. R. W. (1995). Relationship of chimpanzee leaf-swallowing to a tapeworm infection.
Am. J . Primatol. 37, 297-303.
Wrangham. R. W., and Goodall. J . (1989). Chimpanzee use of medicinal leaves. In “Under-
standing Chimpanzees” (P. G. Heltne, and L. A. Marquardt. eds.), pp. 22-37. Harvard
University Press. Cambridge, MA.
Wrangham, R. W., and Nishida, T. (1983). Aspilia spp. leaves: A puzzle in the feeding behavior
of wild chimpanzees. Primates 24,276-282.
Yoerg. S . 1. (1991). Social feeding reverses learned flavor aversions in spotted hyenas (Crocufa
crocufa).J . Comp. Psychol. 105, 185-189.
Zuk, M. (1992). The role of parasites in sexual selection: Current evidence and future directions.
Adv. Study Behav. 21, 39-68.
ADVANCES IN THE STUDY OF BEHAVIOR. VOL. 21
Stress and Human Reproductive Behavior:

Attractiveness, Women’s Sexual Development,
Postpartum Depression, and Baby’s Cry
THORNHILL
RANDY AND BRYANT
FURLOW
DEPARTMENT OF BIOLOGY
THE UNIVERSITY OF NEW MEXICO
ALBUQUERQUE, NEW MEXICO 87131
I. INTRODUCTION
Stress resistance is highly variable among individuals within populations,

and this variation seems to be an important mediator of the association
between phenotypic variation and individual variation in reproductive suc-
cess (i.e., an important mediator of fitness variance) (Parsons, 1996). Stress
tolerance and resistance requires energy that must be diverted from other
functional channels of the organism, which negatively affects the organism’s
energy allocations to the other functions.
Parsons (1996) has emphasized that, although biotic stressors have been
central in analyses by evolutionary biologists, abiotic stressors are often
important sources of selection. He emphasizes also that an interaction
between biotic and abiotic stressors is often seen, and that this interaction
can be subtle and easily overlooked in favor of interpretations based on
centrality of biotic stressors as selection agents. Because of the often inti-
mate association between abiotic and biotic stressors, we choose here to use
a unitary concept and not distinguish between abiotic and biotic stressors.
Stress can be considered in ecological and evolutionary terms. An ecologi-
cal stress becomes significant in microevolutionary terms when it generates
selection for coping with the stress. An ecological stress is significant in
long-term evolution when an adaptation for coping with the stress arises.
Stress-related selection generates specieswide adaptation over long periods
of evolutionary time when the selection is consistently directional and is
not completely opposed by selection in other contexts. Thus, a species’
adaptation that functions against a stress is unequivocal evidence that the
stress was important to fitness over the evolutionary history of the species.
31 9
All rights of reproduction in any form reserved.
0065-3454/98 $25.00
320 RANDY THORNHILL A N D BRYANT FURLOW
Sweating in humans is an adaptation for prevention of overheating of the

other adaptations of the body beyond their optimal performance tempera-
tures. Human hunger is an adaptation to deal with stressfully low levels of
nutrients. Women’s preference for socially successful men as mates is an
adaptation for obtaining mates who provide resource stability and thus
ameliorate resource-related stresses. In an important sense, all adaptations
are evolved stress reducers, because adaptations are caused by selection,
and all selection-each of Darwin’s hostile forces-involves stress. Said
differently, an adaptation is an evolved phenotypic solution to an environ-
mental (abiotic or biotic) problem or stressor. Accordingly, all human
behavior is stress related because all human behavior is the output of
information-processing psychological adaptation.
From the vast domain of human behavior and stress, we have selected
a number of topics that we feel are advancing in terms of empirical or
theoretical progress. Human attraction and attractiveness is one such topic.
Attractiveness may be largely a certification of the ability to cope with
environmental and genetic stresses acting during development. Attraction
appears to arise from adaptations designed to secure benefits for self and
offspring that offset stress. We examine the role of stress-related attraction
and attractiveness in the three major areas of human social behavior: mat-
ing, nepotism, and reciprocity.
Another topic that we address centers around the stresses of the home
environment during girls’ upbringing and how this may functionally mold
the development of women’s sexuality, particularly the variation among
women in sexual arousal. There is considerable evidence that women raised
in low-resource settings exhibit different sexual tactics than women raised
in settings with more resources. The former women show earlier maturity
and age of copulation, more sex partners, less stable pair-bonds, and less
emotional warmth toward mates. It has been proposed that these tactics
serve women’s interest in low-resource settings by promoting mating with
multiple males, each of whom can provide limited resources at best. We
argue that because coital female orgasm has a bonding effect, limited and
highly selective coital orgasm is expected in low-resource settings, because
it will promote relatively weak bonds with multiple males. There is evidence
that low-orgasmic women are concentrated in low-resource settings and
that reduced parental investment is associated with the ontogeny of low-
orgasmic response in women.
We also examine the stresses on human parents surrounding parental
investment in babies. Our focus is on postpartum depression of mothers
and on infants’ crying. We propose and test the hypothesis that postpartum
depression is an evolved manifestation of discriminative parental solicitude
that motivates mothers to eliminate newborn babies under ecological situa-
STRESS AND HUMAN BEHAVIOR 321
tions that were not conducive to infant survival in human evolutionary

history. The baby’s cry is hypothesized to be an evolved, honest signal of
offspring reproductive value. Data bearing on this hypothesis are reviewed.
Both the postpartum depression hypothesis and the baby cry hypothesis
receive support.
11. HUMAN
ATTRACTION
A N D ATTRACTIVENESS
We discuss and test the hypothesis that human sexual competition re-
volves around assessment of and displays of stress resistance (Thornhill
and Gangestad, 1993; Grammer and Thornhill, 1994). The stress resistance
is seen in developmental stability and in development of secondary sexual
traits. We treat developmental stability briefly, as it is covered in detail by
Moller (this volume). We then discuss the connection between secondary
sexual traits and stress.
After treating the role of stress in sexual attraction and attractiveness,
we turn to the importance that attractiveness based on stress resistance
may play in human nepotism and reciprocal altruism.
A. TYPESOF DEVELOPMENTAL
STABILITY
Developmental stability occurs when the evolved or adaptive develop-
mental trajectory is achieved despite environmental and genetic perturba-
tions during development. There are a number of kinds of developmental
instability (Mdler, this volume). Two important kinds in our discussion
are phenodeviance and fluctuating asymmetry. Phenodeviants reflect con-
spicuous deviations from the adaptive developmental trajectory. Human
birth defects and so-called minor physical anomalies (Waldrop et al., 1968)
are examples of human phenodeviants. Birth defects may arise from envi-
ronmental perturbations (e.g., maternally ingested toxins during pregnancy;
Profet, 1995) or genetic perturbations. Minor physical anomalies arise in
the first three months of pregnancy. Phenodeviance refers to any relatively
gross deviation from the adaptive phenotypic target of development,
whether it be morphological, behavioral, physiological, immunological, or
of other origin (Zakharov, 1992; Mdler, this volume). Phenodeviants are
relatively rare in populations of organisms because they reflect only the
most radical disruptions of development.
Developmental instability is most often measured as fluctuating asym-
metry because fluctuating asymmetry is a more sensitive measure of de-
velopmental disturbance than is phenodeviance. Fluctuating asymmetry
is a deviation from bilateral symmetry in normally bilaterally symmetrical
322 R A N D Y THORNHILL A N D BRYANT FURLOW
morphological traits for which signed differences between the left and right
sides have a mean of zero and are normally distributed within a population
(Van Valen, 1962). A long-standing interest in fluctuating asymmetry (Lud-
wig, 1932) derives from the notion that perfect bilateral symmetry is the
optimum and fluctuating asymmetry results from imprecise expression of
developmental design. The corresponding two sides of a bilaterally symmet-
rical trait are encoded by the same genes; therefore, fluctuating asymmetry
arises from environmental stressors or stressors from a hostile genetic envi-
ronment within the genome that lower developmental homeostasis. Accord-
ingly, fluctuating asymmetry increases with exposure to a wide range of
environmental insults during ontogeny such as parasites (Bailit et al., 1970;
Moller, 1992; Moller and Saino, 1994; Polak, 1997), pollutants (Parsons,
1990, 1992), radiation (Moller, 1993a), extreme temperatures, and other
adverse physical conditions (Parsons, 1990,1992), including marginal habi-
tats (Moller, 1995). Fluctuating asymmetry also increases with exposure
to genetic perturbations such as inbreeding (Lerner, 1954), deleterious
recessives (Lerner, 1954; Parsons, 1990), homozygosity (Lerner, 1954; Mit-
ton and Grant, 1984; Mitton, 1993), directional selection (Moller and Pomi-
ankowski, 1993), hybridization of genetically distinct populations (Moller
and Swaddle, 1997), and chromosomal abnormalities (Shapiro, 1992; Fra-
ser, 1994).
Fluctuating asymmetry can be used to compare populations or individuals
within populations. Within populations, asymmetry can vary considerably
across individual organisms. Individual fluctuating asymmetry refers to an
individual’s deviation in either direction from perfect bilateral symmetry.
This variation, in part, is due to genetic differences. A meta-analytic review
of 34 studies of 17 species revealed that fluctuating asymmetry has significant
heritability in general, and multiple studies of human fluctuating asymmetry
show significant additive genetic variance (Moller and Thornhill, 1997a,b).
Fluctuating asymmetry is importantly related to the action of Darwinian
natural and sexual selection. In a wide range of animal species, individuals’
asymmetry is negatively correlated with fitness components pertinent to
natural selection, specifically fecundity, growth rate, and survival (Watson
and Thornhill, 1994; reviews in Mdler, 1997a; Moller and Swaddle, 1997).
In the last several years, evolutionary biologists have studied the role of
fluctuating asymmetry in sexual selection. Sexual selection is the individual
variation in offspring production that is the consequence of individual
differences in traits that affect access to mates. In a diversity of animal
species low-asymmetry males tend to obtain more mates than high-
asymmetry males, through either advantage in male-male competition for
females or female mate choice (Moller and Pomiankowski, 1993; Watson
and Thornhill, 1994; Polak, 1997;review in Moller & Swaddle, 1997). Recent
evidence reveals that fluctuating asymmetry plays a role in human sexual

selection as well, as we discuss later.
B. SECONDARY
SEXUAL AND STRESS
CHARACTERS
Secondary sexual traits evolve by sexual selection and are fundamentally
involved in courtship and other intrasexual contests for mates. Thus, second-
ary sexual traits are signals. An important discovery in modern studies of
secondary sexual traits is that the expression of such traits often signals
reliable information about the phenotypic condition of the signaler. Such
traits function as reliable signals for two related reasons. First, their expres-
sion is facultatively plastic and thus is importantly dependent on the envi-
ronment in which they develop, including the bearer’s condition during
development of the trait. Second, many secondary sexual traits are survival
handicaps (Zahavi, 1975, 1977a). They are costly to produce, but more so
to individuals of low rather than high phenotypic quality. A high-quality
individual is therefore able to develop the most extravagant secondary
sexual character, but at a relatively low cost. The differential cost of the
secondary sexual trait is the mechanism that ensures reliable signaling of
quality, because only high-quality individuals with superior viability genes
will be able to survive with the most extreme levels of sexual display
(Haywood, 1989; Iwasa et al., 1991).
A special version of the handicap mechanism is the immunocompetence
handicap hypothesis (Folstad and Karter, 1992; Wedekind and Folstad,
1994), which suggests that secondary sexual characters develop in response
to circulating androgens (or other self-regulating biochemicals) that in-
crease the expression of secondary sexual characters, but reduce the func-
tioning of the immune system, and thus suppress the ability of individuals
to raise an immune defense against parasites. In other words, there may
be an intricate, negative feedback mechanism among host secondary sexual
characters, host hormones, host immune defense, and parasites. High-
quality males will be able to develop large sexual traits, cope with high
levels of androgens, and to only a relatively small extent compromise their
immune defense. Sex differences in the course of parasite infections (Bundy,
1989; Zuk, 1990) and relationships between sex hormones and parasitism
(Alexander and Stimson, 1989) are consistent with the immunocompetence
handicap hypothesis.
Secondary sexual traits typically show much more fluctuating asymmetry
than other traits (Mdler and Pomiankowski, 1993). This is expected because
such traits are often condition dependent, elaborate, and energetically de-
manding. This combination presents a developmental challenge for an indi-
vidual to make secondary sexual traits perfectly symmetrical. This is not
324 RANDY THORNHILL AND BRYANT FURLOW
to say that secondary sexual traits are the only traits in which fluctuating
asymmetry will correlate with phenotypic quality. Indeed, fluctuating asym-
metry in ordinary traits may often be a good predictor of Darwinian fitness
and genetic quality. It is to say that secondary sexual traits will often be
the most revealing of phenotypic and genetic quality differences among
individuals. Furthermore, it is likely that many bilateral secondary sexual
traits are designed to signal symmetry per se and thereby phenotypic quality;
this is supported by experimental evidence from studies of male secondary
sexual traits in birds (Mgller, 1993b; Swaddle and Cuthill, 1994a,b).
Secondary sexual traits are exaggerated by a directional sexual selection
process. Directional selection appears to increase fluctuating asymmetry
and developmental instability in general regardless of whether the selected
trait is sexual or ordinary (Mgller and Pomiankowski, 1993; Pomiankowski
and Mgller, 1995;review in M d l e r and Swaddle, 1997). Directional selection
is thought to increase fluctuating asymmetry by incorporation of new alleles
with exaggerated trait value and by disruption of the mechanisms that
stabilize development. The new alleles may cause genetic stresses during
development. The disruption of developmental mechanisms associated with
the evolutionary response to directional selection makes the developing
organism susceptible to both environmental and genetic perturbations.
Thus, human secondary sexual traits may importantly display stress resis-
tance because they may be immunocompetence handicaps and because
they may manifest relatively high developmental instability.
C. STRESS A N D HUMAN
RESISTANCE SEXUAL
SELECTION
Current knowledge of human attractiveness indicates that it may be a
certification of stress resistance and thereby a health certification. Human
attractiveness research is thus potentially important to the health profes-
sions. Human attractiveness relates to health through phenotypic quality
in general and immunocompetence and developmental stability, in particu-
lar, and through immunocompetence handicapping sex hormones (Thorn-
hill and Mgller, 1997).
Cross-cultural research has shown that, although men place more value
on physical attractiveness of a mate than do women, both sexes highly
value it (Buss, 1989). The aesthetic judgments of faces made by individuals
of different cultures tend to agree, and children as young as 6 months show
preferences similar to those of adults (see review in Cunningham et al.,
1995; Jones, 1995).
Certain human facial features are secondary sexual traits, arising or
increasing in size at sexual maturity under the proximate influence of andro-
gens and estrogens. Both sexes have both hormones, but the ratio at puberty
is sex specific. Relatively high testosterone levels lead to growth of lower

face and jaw, cheekbones and brow ridges, and projection of the central
face between the brow and bottom of the nose. Relatively high estrogen
at puberty prevents this growth, but yields increased lip size. These features
distinguish young men’s and women’s faces (see reviews in Enlow, 1990;
Symons, 1995). As estrogen declines with female age, testosterone masculin-
izes the female face. Thus, a highly estrogenized female face marks youth
and thus high fertility (Symons, 1995; but see Jones, 1995). Abundant
evidence indicates that markers of high estrogen such as small lower jaw
and lower face, in general, and large lips are attractive in women’s faces
(Johnston and Franklin, 1993; Perrett et af., 1994; Cunningham et af., 1995;
Jones, 1995). By contrast, a large lower jaw is rated as dominant and
attractive in the male face (reviewed by Symons, 1995). The dominance
rating of adult men’s faces correlates positively with their amount of previ-
ous sexual experience and with testosterone level during puberty (Mazur
et af., 1994).
Facial secondary sex traits, in addition to conveying sex and sexual matu-
rity, appear to be designed to advertise phenotypic and genetic quality,
because of their connection to sex hormones. As mentioned earlier, testos-
terone appears to be an immunosuppressor. Estrogen may also negatively
impact immunocompetence when at high titers (Grossman, 1985; Ahmed
and Talal, 1990). Sex-hormone-facilitated markers honestly advertise im-
munocompetence because the high hormone titer needed to produce attrac-
tive features simultaneously handicaps disease resistance accordingly, and
thus can be afforded only by individuals of extraordinary immunocompe-
tence. This model applied to human facial secondary sexual traits suggests
that the facial hormone markers are conditionally expressed advertisements
of phenotypic quality or overall health, and simultaneously supports the
view that attractiveness judgments based on these features arise from adap-
tation designed to detect mate quality.
Crammer and Thornhill (1994) found that opposite-sex health ratings
of facial photos positively correlate with the degree of development of the
sex-specific facial secondary sexual traits. Attractive expressions of human
facial secondary traits actually covary with health (T. Shackleford and R.
Larsen, personal communication, January 1996). Subjects completed daily
records of psychological, emotional, and health symptom status over a 2-
month period. Shackleford and Larsen found evidence that women with
highly estrogenized facial structure-prominent cheekbones, short chin,
and narrow lower jaw-and men with highly testosteronized facial struc-
ture-prominent cheekbones, large chin, and wide lower jaw-are indeed
emotionally, psychologically, and physiologically healthier. For example,
compared to women with wide lower jaws, women with small lower jaws
are less depressed, less emotionally labile, more excited and enthusiastic,
and experience fewer gastrointestinal problems and less muscle soreness.
Men with wide jaws are less neurotic, less emotionally labile, feel less
sluggish, and experience fewer physical symptoms of ill health daily than
men with narrow jaws. It should be noted that prominent cheekbones are
attractive in both sexes, but reflect sex-specific hormone effects (see Sy-
mons, 1995). They reflect high midfacial flatness in women due to high
estrogen. In men, they reflect lateral bone growth due to androgens.
There is evidence that parasites play a critical role in human attractive-
ness. Disease organisms causing skin infections and rashes influence human
attractiveness (Symons, 1995). Also, the value that people of either sex
place on physical attractiveness in choice of a mate correlates positively
with the prevalence of parasites across 29 human societies (Gangestad and
Buss, 1993). Moreover, maxillary sinusitis of probable bacterial etiology
appears to cause irregular growth of most facial bones of the lower face.
Children with these facial irregularities are recognizable on the basis of
their facial appearance (Yates, 1928). It is unknown, however, how sinusitis
may influence the structure of facial secondary sexual traits at puberty. Both
size and symmetry of facial structures may be influenced by sinus infections.
Other evidence that physical attractiveness relates to phenotypic quality
has come from research on facial symmetry. In three studies, in both sexes,
faces with high bilateral symmetry are rated more attractive than less sym-
metrical faces (Grammer and Thornhill, 1994; D. Perrett and M. Burt,
personal communication, July 1996; L. Mealey, R. Bridgstock, and G.
Townsend, personal communication, November 1996), and facially symmet-
ric individuals exhibit more emotional, psychological, and physiological
health (Shackleford and Larsen, 1997). Faces show no evidence of system-
atic directionality in size of hard or soft tissue, according to Sackheim’s
(1985) review of 50 years of research on facial asymmetry. Thus, most facial
asymmetry appears to be fluctuating (see also Hershkovitz et al., 1992).
Facial asymmetry is the rule. Probably all faces show some asymmetry of
both hard and soft tissue and the degree varies considerably among healthy
subjects (Farkas and Cheung, 1981; Sutton, 1969; Sackheim, 1985; Peck et
al., 1991; Hershkovitz et al., 1992).
Two studies involving making perfectly symmetrical faces from hemifaces
using computer techniques have shown that such faces are rated lower in
attractiveness than natural faces, which exhibit some asymmetry (Langlois
et al., 1994; Swaddle and Cuthill, 1995). This effect apparently is due largely
to the unnatural facial feature sizes and textures created by the hemiface
method (D. Perrett and M. Burt, personal communication, July 1996; L.
Mealey, R. Bridgstock and G. Townsend, personal communication, Novem-
ber 1996). Also, as Swaddle and Cuthill point out, the low attractiveness
ratings of hemifaces may reflect, in part, the fact that perfectly symmetrical
faces look unnatural, perhaps even dead, given that emotional expression
in the face is asymmetric (e.g., see Sackheim, 1985).
There is evidence that facial asymmetries correlate within individuals.
Peck et al. (1991) examined facial skeletal asymmetry in 52 adult subjects.
Eye orbit asymmetry was positively correlated with zygion (cheekbone)
asymmetry and zygion with gonium (mandible) asymmetry. Farkas and
Cheung (1981), in a large study of soft-tissue asymmetries of children and
young adults, also found significant correlations within individuals in certain
facial asymmetries. It is conceivable that these within-individual correla-
tions may be generated by variation in sinus infections prior to puberty or
in early adulthood among individuals, given the strong effect of sinusitis
on facial bone growth (Yates, 1928).
Also indicating that facial attractiveness marks phenotypic quality is the
relationship between facial asymmetry and facial sex hormone markers.
Facial symmetry in each sex correlates with the sex-specific attractive ex-
pression of facial secondary sexual traits (e.g., symmetry positively corre-
lates with chin size in men and negatively with chin size in women) (Ganges-
tad and Thornhill, 1997). Thus, people with symmetric faces tend to have
attractive sex hormone markers in their faces.
Moreover, composite faces made from many individual photos are rated
more attractive than the majority of the individual photos used to make
the composites (Langlois and Roggman, 1990; Langlois et al., 1994). This
means that, at least for certain facial features (not the secondary sexual
traits), being near the mean is associated with greater attractiveness (also
see Jones and Hill, 1993). It has been suggested that averageness in certain
facial features may mark phenotypic and genetic quality because average-
ness in traits under stabilizing selection often positively covaries with heter-
ozygosity (Thornhill and Gangestad, 1993). Heterozygosity is sometimes
associated with increased developmental stability and may be associated
with reduced susceptibility to parasites (see Thornhill and Gangestad, 1993;
Wedekind, 1994; Maller and Swaddle, 1997).
The immunocompetence model may also hold for nonfacial secondary
sexual traits of men. For example, male body mass may signal immunocom-
petence. Adult body mass is sexually differentiated, with males being larger.
This is due, in part, to men’s greater muscle mass, which arises at puberty
as a result of testosterone. Male body mass in the Ache Indians (who
do not use contraception) positively correlates with number of offspring
produced, apparently because heavier men are more attractive to women
(Hill and Hurtado, 1996). In the West, women rate male athletic builds
more attractive than other male builds (Gangestad and Thornhill, 1997), and
the mates of larger men tend to orgasm during copulation more frequently
(Thornhill et al., 1995). Copulatory orgasm may be a mechanism of female

choice (see later discussion).
Men’s mass and athleticism appear to signal men’s developmental health,
that is, low-fluctuating asymmetry, which may reflect immunocompetence.
Male body mass is positively correlated with male body symmetry (Man-
ning, 1995; Thornhill et al., 1995), and, as already mentioned, men of high
body symmetry are more muscular and vigorous. Such men also show lower
resting metabolic rates than men of low body symmetry (Manning et al.,
1997). Women’s body mass negatively correlates with their body symmetry.
Thus, large men and small women exhibit greater developmental stability
(Manning, 1995).
The relatively low metabolic rate of symmetric men implies that such
men have more energy available for other body functions, which may
account for their greater vigor. If so, vigor itself in men is an honest signal
of developmental health.
Men exhibiting developmental stability are sexually attractive to women.
Compared to men with high asymmetry, men with low asymmetry have
greater facial attractiveness, greater numbers of lifetime sex partners, more
extrapair copulations, and quicker access to copulation in romantic relation-
ships. Low-asymmetry men also are chosen more often as extrapair copula-
tory partners and begin sexual intercourse earlier in their life history (Gan-
gestad et al., 1994; Thornhill and Gangestad, 1994; Baker, 1997; Gangestad
and Thornhill, 1997). Also, the mates of symmetrical men show the most
reported copulatory orgasms (Thornhill et al., 1995; but see Baker, 1997).
Female copulatory orgasm may function in selective bonding with males
and in selective ejaculate retention. Thus, female orgasm may be a subtle
form of female choice involving choice of sire (Smith, 1984; Birkhead and
Moller, 1993; Baker and Bellis, 1995; Baker, 1997).
Nonfacial human secondary sexual features in women appear to signal
phenotypic quality, and higher quality expressions are judged more attrac-
tive. The sexual dimorphism in waist-to-hip ratio in humans arises at puberty
and is facilitated by sex hormones. Waist-to-hip ratio in women correlates
negatively with estrogen, fertility, and health, and positively with age, and
low waist-to-hip ratios (.6-.7) are maximally attractive in women (Singh,
1993, 1995). Adult female breasts develop at puberty under the influence
of estrogen. Breast-size symmetry was positively correlated with reported
fertility (number of children during their lifetime) in two separate samples
of women (United States and Spain) (Mdler et al., 1995). Breast symmetry
positively affects attractiveness judgments of men, and men’s interest in
both short-term and long-term relationships (Singh, 1995).
In theory, if symmetry marks high phenotypic and genetic quality, and
bilateral secondary sexual traits honestly signal quality as reduced asymme-
try, it is predicted that bilateral secondary sexual traits will exhibit high
levels of fluctuating asymmetry, specifically higher levels than in ordinary
traits not influenced by sexual selection. This has been found in numerous
nonhuman species (Mdler and Swaddle, 1997). This same pattern is found
in Homo sapiens. Absolute breast size asymmetry in a sample of 172 women
in Spain showed a mean of 1.23 cm (Moller et af.,1995). Similar high breast
size asymmetry values were obtained from samples in New Mexico (Mdler
et al., 1995) and in the United Kingdom (Manning et al., 1996). Average
relative breast size asymmetry was about 5% in these samples (i.e., absolute
breast size asymrnetry/breast size). Other morphological characters in hu-
mans usually show 2% or less relative fluctuating asymmetry (Livshits and
Kobyliansky, 1991; Thornhill and Gangestad, 1994: Manning, 1995).
Human facial asymmetry also may show the pattern of more fluctuating
asymmetry with facial secondary sexual traits (Peck et af.,1991). Eye socket
asymmetry showed relatively low asymmetry, whereas more asymmetry
was found in the two sexual traits zygion (cheekbone at most outward
lateral point) and lower jaw. Farkas and Cheung (1981) measured lateral
facial soft-tissue asymmetry of children and 18-year-olds of both sexes.
Only in the 18-year-olds was there a significant sex difference (69% of
males and 47% of females) in asymmetry in the facial trait nasion to tragion
distance. This is the distance on the lateral aspect of the face between a
point on the upper end of the nose (nasion) to the temple (tragion). This
distance would include the outward pubertal growth of the central facial
region (eyebrows to bottom of the nose) seen in males (Enlow, 1990;
Symons, 1995), which is facilitated by testosterone.
Current knowledge of human physical attractiveness leads to the conclu-
sion that the sexual selection responsible for designing many of the second-
ary sexual traits was not Fisherian, that is, male and female winners of
sexual competition in human evolutionary history did not sexually signal
with arbitrarily attractive traits. Human secondary sexual traits reveal devel-
opmental stress resistance, and their attractive expressions signal the ability
to resist developmental stress. Attractive adult faces of both sexes reflect
secondary sexual traits requiring high titers of sex-specific hormones that
may connote immunocompetence, and attractive facial secondary sexual
traits may connote greater emotional and psychological health. Also, attrac-
tive faces are developmentally healthy, as seen in their high developmental
stability. The adult male body is attractive when reflecting testosterone
effects and athleticism, which covary with physical and developmental
health. The adult female body is attractive when it shows good health
by developmental stability in breasts and high estrogen in low waist-to-
hip ratio.
The use of simple neural network models has led to the hypothesis that
preferences for exaggerated or symmetrical stimuli that arise from these
models are inevitable by-products of visual information processing and not
the result of selection for detection of signaler quality (e.g., Johnstone,
1994). However, Dawkins, and Guilford (1995) have pointed out that these
simple models do not behave like the visual systems of animals, and the
preferences they generate may be incidental to the inadequacies in the
models themselves. Furthermore, the by-product hypothesis of preferences
for exaggeration or symmetry is unlikely to apply to humans, because of
the association between preference and tangible benefits, specifically stress
resistance in a mate and the genetic and material benefits a stress-resistant
mate can provide.
D. STRESS
RESISTANCE SOCIAL
AND NONSEXUAL BEHAVIOR
The above discussion of human attractiveness in relation to stress focuses
on sexual selection, that is, on issues of competition within each sex by
display to the opposite sex and between sex mate preference. Human
attractiveness judgments occur in social behavioral domains other than
sexual competition. Humans everywhere are immersed in a complex set of
interactions with relatives and nonrelatives other than mates. Under the
hypothesis that attractiveness reflects phenotypic and genetic quality, spe-
cifically the ability to withstand developmental stresses, and the complimen-
tary hypothesis that attraction reflects a preference for stress resistance
during development, the general prediction is that human reciprocity and
nepotism will be importantly based on physical attractiveness (Thornhill
and Gangestad, 1993). If, as proposed here, physical attractiveness is a
stress-resistance certification and thereby a marker of phenotypic quality,
people are expected to have adaptations that use attractiveness information
(1) in choosing social allies for reciprocal alliances: (2) for dispensing bene-
fits to these social allies: and (3) for dispensing nepotism to relatives. The
reproductive benefits of aid in the form of reciprocity or nepotism depend
on survival of the beneficiary of that aid. According to our view, physical
attractiveness was a reliable and consistent predictor of individual survival
in human evolutionary history because attractiveness reflects phenotypic
quality. Thus, we anticipate that people will prefer to interact socially with
individuals exhibiting bodily health certificates (e.g., attractive secondary
sexual traits and low asymmetry). Freeland (1976) suggests that, in nonhu-
man primates, health of a potential social ally should be an important factor
influencing the willingness of individuals to engage in social interactions,
and that healthy individuals should be preferred social companions. We
agree with Freeland but add that physical beauty may be an important
means of assessing the health-related reproductive value of social allies

to self.
Abundant evidence indicates that people show greater interest in being
associated socially with physically attractive people than with unattractive
people and that people treat attractive people with favoritism (reviewed
in Alley and Hildebrandt, 1988; Eagly et al., 1991; Jackson, 1992). Physically
attractive people are viewed as having greater social skills and competence,
as having higher status, and as being more successful, interesting, educated,
and intelligent than unattractive people. That is, they are viewed as having
more social assets and as better social allies. Not only are attractive people
perceived as having these social assets, they are treated as if they have
them. Moreover, numerous studies reveal that the physical attractiveness
of a person has a major positive influence on the number and stability of
friendships with same-sex individuals. No study has tested for positive
effects of facial symmetry or beauty of secondary sexual traits on people’s
interest in and selection of potential allies for reciprocity.
The hypothesis that people base social interactions on physical attractive-
ness because it is related to phenotypic and genotypic stress resistance
ancestrally may apply to discriminative nepotism patterns in people. It is
likely that stress tolerance and resistance was a major determinant of sur-
vival and well-being of one’s genetic relatives during human evolutionary
history. Thus, everything else being equal, we would expect people to direct
more nepotistic benefits to attractive than to unattractive relatives. Studies
have provided positive results bearing on the predicted relation between
attractiveness and nepotism (see review in Alley and Hildebrandt, 1988;
also Langlois et al., 1995). Mothers of children with facial anomalies rate
their children more negatively than do mothers with facially normal chil-
dren. Parents have higher expectations for attractive than for unattractive
children. Also, mothers of more attractive newborns are more affectionate
toward their babies than are the mothers of unattractive newborns (Langlois
et af.,1995). All of these patterns suggest differential investment by parents
in offspring, with attractive offspring receiving more. We know of no study
that has directly examined the effect of body symmetry or other markers
of attractiveness on wealth inheritance or any other form of nepotism.
E. CONCLUSIONS
No finding in social psychology is more robust and replicable than the
finding that individuals respond more positively to physically attractive
than to unattractive persons. Indeed, attractiveness may be the single best
predictor of human social preference. It clearly affects how one is treated
by others. Attractiveness also appears to affect how one treats others.
Attractive individuals seem to be very discriminating socially-this is seen

in the positive assortment on the basis of looks both in mateships and
friendships (Jackson, 1992); it is also seen in the reduced investment of
attractive men in romantic relationships (Gangestad and Thornhill, 1997).
Only relatively recently has research focused on the most fundamental and
intriguing question about the power of looks in human everyday life. What
is the evolutionary basis of the priority placed on looks in the human
mind? There is increasing evidence that human physical attractiveness and
attraction owe their ultimate existence to Darwinian selection for displaying
health and for assessing health in others. Health or phenotypic quality can
be appropriately viewed as synonymous with stress tolerance and resistance.
Accordingly, physical attractiveness reflects (1) developmental stability, or
the ability to cope with environmental andfor genetic perturbations during
development; and (2) expressions of secondary sexual traits that appear to
mark immunocompetence and the ability to cope with sex hormones.
There is need for more experimental research on human attractiveness in
relation to symmetry and secondary sexual traits. Singh (1993,1995), D. Perrett
and M. Burt (personal communication, July 1996), and L. Mealey, R. Bridg-
stock and G. Townsend (personal communication, November 1996) have
made pioneering studies in this regard, but most studies are observational
and correlational with only statistical control of potential confounds.
111. PARENT-DAUGHTER
RELATIONS
A N D WOMEN’S
SEXUAL
BEHAVIOR
There is a voluminous literature dealing with the effects of parental

relationship behavior on women’s romantic relationship attachment. There
is also a large literature dealing with parental behavior in relation to the
development of women’s sexuality (Belsky et af., 1991). Evolutionary psy-
chologists have reviewed this literature in the light of their evolutionary
developmental theory of women’s behavior pertaining to relationships and
sexuality (Draper and Harpending, 1982; MacDonald, 1985, 1988, 1992;
Surbey, 1990; Belsky, 1990; Belsky et af., 1991; Cashdan, 1993). According
to this theory of women’s development, which is based on life-history theory
(Roff, 1992), the parents’ ability and willingness to invest in each other
and their children affects the sexual developmental pathway followed by
a daughter.
Parents and daughters are viewed as strategists. Parents’ patterns of child
rearing prepare daughters for the adult social environment they will be
likely to face. Daughters internalize the rearing setting and learn the emo-
tional and relationship skills that are suitable for the probable environment
they will face in adulthood. Factors such as marital discord, high stress in
STRESS AND HlJMAN BEHAVIOR 333
the home, inadequate resources, and father absence or emotional distance

are developmental stresses that cue in daughters early puberty and sexual
activity, short-term and unstable mateships characterized by exploitation,
emotional coolness and insensitivity, and limited parental care. On the
other hand, interparental harmony, adequate resources, father presence
and involvement in child rearing cue later puberty and restricted sexual
activity, stable relationships, greater parental investment, and emotional
warmth and positive affection. Daughters with the former set of traits are
viewed as best suited in terms of reproductive success for an environment
in which resources will be relatively limited. The latter set of traits equips
daughters reproductively for a social environment in which resources will
be abundant and consistently available. Women in environments of reliable
resources are expected to perform best in terms of reproductive self-interest
by positive and enduring interpersonal, social relations (including romantic
ones) based on trust. Sexual restraint may allow women in resource-rich
environments to secure a mate with resources because of the positive influ-
ence of women’s restricted sexual history on men’s perception of high
paternity reliability (Low, 1989). Women in resource-poor environments
are served by a mistrustful and opportunistic relationship style. Women
use their sexuality to access resources from men (e.g., Symons, 1979; Buss,
1994). In stressful environments, early maturity would lead to quicker access
to resources that are missing as a result of limited parental investment.
Given that individual men will be less able to provide sufficient investment
in resource-poor situations, women’s weak attachment with a mate would
adaptively promote multiple sex partners and thus a greater total amount
of male-provided resources than if they were faithful to a single mate.
The massive literature on female child development shows a remarkable
fit to this evolutionary view of sexual and adult relationship behaviors of
daughters being dependent on upbringing. This literature comprises many
detailed studies in the West spanning several decades, as well as cross-
cultural studies, including hunter-gatherer societies (see previous refer-
ences: also Low, 1989; Hurtado and Hill, 1992; other papers in Hewlett,
1992). This view sees female development as a conditional strategy with
tactics being selected by an adaptation that reads current rearing conditions
pertaining to resources and reliability of parental investment and forecasts
the future social environment based on rearing environment. The different
tactics involve different schedules of reproductive and somatic effort and
different allocations of reproductive effort to mating and parenting.
Because an unrestricted sexual history is a detriment for women in secur-
ing a mate with abundant resources-such males are choosy about females
in whom they invest (e.g., Symons, 1979; Buss, 1994)-it is expected that
early sexual activity and a sexual history of many partners will often track
a young woman into a promiscuous future. Thus, there may be considerable

continuity in the benefits and costs of unrestricted sexuality throughout the
life history of a woman in a low-resource setting.
Although there is reason to believe that a girl’s experience in a rearing
environment with a low- versus high-resource base may have long-term
consequences on her sexuality, it does not follow that women’s sexuality
will show no flexibility in terms of resource variation in the adult environ-
ment. The adaptive hallmark of condition dependence is the ability to shift
tactics to increase reproductive success if circumstances encountered are
promising in this regard. Thus, it is anticipated that women will have the
ability to assess current circumstances of male investment potential and
make adjustments in sexuality that will best capitalize on current conditions.
Research by Cashdan (1993) suggests women may have this lability. She
has shown that the pursuit of short-term versus long-term romantic relation-
ships by women (Buss and Schmidt, 1993) depends on whether a woman
perceives that men in the environment will invest and commit themselves
in romantic relationships. Compared with women who perceive that they
are in an environment in which men will invest, women who perceive that
they are in an environment where men are not likely to invest act and
dress in a more sexually provocative manner, and use copulation to attract
desirable men. This variation can be understood as a facultative response
of an evolved female sexual psychology designed for an output of sexual
restraint (thus giving cues of paternity reliability) when investing males
may be accessible, and less sexual restraint (to access material benefits)
when each male can or will invest little (Cashdan, 1993). However, it is
unclear from Cashdan’s research how much development versus current
environment affected the sexual differences among the women.
Father-daughter relations have figured importantly in the evolutionary
theory of women’s sexual development. Women developing in settings
of reduced paternal investment associated with divorce, marital strife, or
father’s death show earlier onset of menarche and sexual behavior and
greater numbers of sexual partners. Reduced paternal investment may have
impacted on daughters severely throughout human evolution. Ache Indian
girls without fathers are more likely to have infectious disease, suffer mortal-
ity, and be captured by men in raids (Hurtado and Hill, 1992). In the
Ache context, the advantage to girls without fathers of speeding up sexual
maturity and beginning sexual activity early is crystal clear.
This literature’s focus on the fathers’ effects on daughters’ sexual develop-
ment does not imply that mothers don’t influence daughters (see Patterson,
1986). Nor does it mean that maternal investment is insignificant in influ-
encing daughters’ life-history decisions about life pursuits. However, there
is good reason to believe that adult males provided certain essential parental
STRESS A N D HUMAN BEHAVIOR 335
resources more consistently than did females in human evolutionary history.

Presumably, this is the reason that male parental investment evolved in
humans in the first place. Based on human sex differences in parenting
behavior, the resources provided by ancestral males were meat from hunt-
ing, teaching hunting skills to sons, and physical protection (e.g., Hewlett,
1992). Even in cases in which mothers reject, neglect, or desert offspring,
an underlying cause may involve absent or unreliable paternal investment.
It is clear that social support from fathers is important in decisions by
mothers to kill offspring rather than invest in them (Daly and Wilson, 1988).
We hypothesize here that there is an additional component of women’s
sexuality that may be influenced significantly by rearing environment: varia-
tion among women in sexual arousal. Among women, sexual response
is highly variable (Fisher, 1973). For example, some women are totally
anorgasmic, others are anorgasmic during copulation, and others show
frequent copulatory orgasm. The evolutionary theory of women’s sexuality
discussed earlier includes the following tactical components: time of sexual
maturation/puberty, onset of sexual activity, stability of the pair-bond and
sex partner number, and emotional coldness/warmth toward the mate. Sex-
ual arousal is likely to be a tactical component of female strategy because
of its connection to pair-bonding and sperm use. Specifically, we suggest
that limited and highly selective orgasmic response is an evolved facultative
expression of female development in the stressful environments of limited
male spousal and parental investment that would complement the other
sexual tactics shown by women in such environments. Below we examine
first the hypothesis that female sexual arousal is a female choice adaptation.
We then more fully discuss variation in female sexual strategy and the
relationship between female sexual response and rearing environment, and
then we specifically treat the daughter-father developmental environment.
A. FEMALE
SEXUAL AS FEMALE
AROUSAL CHOICE
Female sexual arousal can be viewed as a female choice adaptation with
multiple outputs that range from no arousal to orgasm. There is considerable
evidence that the range of outputs reflects female choice. Many have com-
mented on the connection between women’s sexual arousal and female
choice. When female choice is circumvented, as in rape, women do not
show sexual arousal. This is a reason that women often label rape as a
violent act rather than a sexual act, whereas men view sexually coercive
sex as a sexual experience (Thornhill, 1994). Also, women’s sexual interest
and arousal are importantly tied to a man’s investment in the relationship:
female mate preference is also importantly predicted by the ability and
willingness of a man to invest (Symons, 1979; Grammer, 1993; BUSS,1994).
Moreover, women’s copulatory orgasmic frequency is significantly pre-

dicted by marital happiness and husband’s income and status, and by the
woman’s own socioeconomic status with low frequencies in women of lower
socioeconomic status (Fisher, 1973). Finally, as mentioned earlier, female
copulatory orgasm is positively correlated with the developmental stability
of the mate, a pattern predicted by the hypothesis that women are selectively
bonding with and retaining preferentially the sperm of men of high pheno-
typic and genetic quality (Thornhill et al., 1995; but see Baker, 1997).
Women are more sexually interested in and aroused by men capable and
willing to invest and by men of high phenotypic and genetic quality. Thus, all
components of female sexual arousal from absence of arousal to copulatory
orgasm may be strategically related to female choice of mate and sire.
Female orgasm could interfere with women’s sexual strategy when
women are in settings in which limited sexual and emotional commitment
to each man is in their evolved interests, that is, when men have reduced
resources. Said differently, like all traits, female orgasm has costs in addition
to benefits. The costs of orgasm may include strong pair bonding with a
mate in environments in which short-term mateships are optimal. Oxytocin,
a hormone known to have effects on social bonding, is released in large
quantities during female orgasm (see Thornhill et al., 1995, for evidence
and discussion). If oxytocin plays an important role in differential bonding
of women to men, it may lead to reducing female interest in male nonmates
and thereby limit resources females can obtain by exchanging sex for re-
sources from multiple mates.
Females with multiple mating partners could still benefit from orgasm’s
sperm retention function by limited and highly selective orgasm in relation
to male genetic quality. Also, women appear to control the fertilizing
potential of ejaculates in ways other than by copulatory orgasmic upward
suction. First, masturbatory, that is, cryptic, orgasm may be important
because it can result in acidic cervical conditions that are hostile to subse-
quently placed ejaculates (Baker and Bellis, 1995). Second, the timing of
copulation in relation to ovulation may allow women some control over
which male sires offspring.
Strategic faked orgasm may be especially common in women pursuing
resources from multiple sex partners. Romantically involved women who
are flirtatious with men other than their usual partner seem to fake more
orgasms (Thornhill et al., 1995).
The main idea here is that limited and highly selective female coital
orgasm is predicted in women who live in environments in which resources
are limited or unreliable. This form of female sexual response would compli-
ment the female’s other traits that have been viewed by others as strategic
in such settings. In a resource-rich, high-male investment setting, female
orgasm will bond the woman with her providing man, limit infidelity, and
result in his sperm being used. This is not to say that female infidelity is
uncommon by women paired with providing males. During infidelity, these
women may use orgasm for selective sperm retention. There is evidence
that women’s extrapair copulations result in higher copulatory orgasm
frequency than their inpair copulations (Baker and Bellis, 1995).
RELATIONS A N D FEMALE
B. FATHER-DAUGHTER SEXUAL AROUSAL
There has been some in-depth research on the relationship between
fathers’ and mothers’ behaviors and presence and daughters’ sexual arousal.
Most of this was inspired by Freudian psychoanalytic theory, which sees
daughters desiring sexual experience with father and thus competing sexu-
ally with mother. Particularly salient in Freudian theory is the need for
women to resolve incestuous desires for father in order to be orgasmic
during copulation. Copulatory orgasm in Freudian terms is the marker of
an emotionally mature and stable woman, and the more orgasmic a women
is during mating, the more mental health she possesses. We will not provide
a full critique of Freud’s views of women’s sexuality. We do point out that
evolutionary knowledge of mind design implies that Freud’s view is false:
women will not have evolved to desire father as a mate (Thornhill and
Thornhill, 1984). At any rate, the literature conains data bearing on the
connection between women’s sexual arousal and the parent-daughter rela-
tionship even though the data were not collected to examine Darwinian
hypotheses.
Fisher (1973) studied a sample of about 300 married, middle-class women
ranging in age from 21 to 45 years (x = 25). The women completed
questionnaires about the parent-child relation and parental behavior dur-
ing childhood. Separate questionnaires were used for mothers and fathers.
The questionnaire by Roe and Siegelman (1963) examined parental behav-
iors such as protective, demanding, rejecting, neglecting, casual, loving, and
punishment. The women rated each parent on a five-point scale for each
variable. The questionnaire also allowed women to judge their relationship
with each parent. The women separately provided tape-recorded oral narra-
tives of their impressions of the parent-child relationship, and included
any separation between parent and child (for each parent). The women
self-rated their consistency in attaining orgasm during sexual intercourse
on a six-category scale (1 = always, 6 = never). Retesting of the orgasm
rating over intervals of 3 to 7 days was highly significant.
There are several findings from Fisher’s research that are relevant to the
prediction that resource-poor childhood environment will cue less sexual
arousal in women. Parental loss or absence was examined in relation to
low- and high-orgasmic women. Childhood separations from mother were

essentially absent in the sample and too infrequent for analysis. References
to separation from father were more numerous. The low-orgasm consistency
women exceeded those with high consistency in the frequency with which
they described their fathers as having been dead, separated, or absent
during their childhood. Although not an ideal measure of resource limita-
tion during childhood, it is likely that, all else being equal, women reporting
father separation grew up in homes with fewer resources than women
reporting no father separation.
In the questionnaire data that systematically evaluated the women’s
attitudes toward their parents, there were differences dpending on sex of
parent. There was no statistically significant relationship between a woman’s
recall of how her mother had treated her and that woman’s ability to attain
orgasm or her orgasmic consistency. Some interesting patterns emerged,
however, in the women’s recall of how father had behaved. A statistically
significant negative correlation was observed between orgasm consistency
and casualness of paternal behavior. Orgasm consistency was lower in those
women who perceived the father as conforming to the following paradigm
of a casual parent from Roe and Siegelman’s questionnaire (1963, p. 357):
They [high-casual parents] will be responsive to him [the child] if they are not busy
about something else. They d o not think about him or plan for him very much, but take
h i d h e r as a part of the general situation. They don’t worry much about him and make
little definite effort to train him. They are easygoing, have few rules, & do not make
much effort to enforce those they have.
Demanding fathers are at the opposite end of the continuum of casual-

demanding fathering behavior. Demanding refers to the father’s imposition
of strict rules and demands of obedience, whereas casualness refers to the
father’s easygoing manner and the few rules that were rarely enforced.
There was a significant positive correlation between demanding paternal
behavior and women’s orgasm consistency.
Overall, then, the greater a woman’s coital orgasm consistency, the less
permissive and the more controlling she perceives her father to have been.
The casualness-demanding dimension may tap importantly into paternal
investment. The value of the daughter to the casual father is probably less
than the value of the daughter to demanding fathers. It would appear that
demanding fathers are investing more time and other parental effort into
daughters. That casual fathers are less investing in daughters is supported
by Fisher’s own conclusion (p. 263): “. . . these behaviors on a [casual]
father’s part could be viewed as being thinly separated from indifference
and lack of genuine concern about his daughter’s welfare.”
The oral narratives of high- and low-orgasmic women depicted their

fathers significantly differently. Low-orgasm consistency women, more fre-
quently than high-orgasmic women, described their fathers as unavailable
for a substantial or consistent relationship with them. The closeness-
distance factor did not show this pattern in the recollections of the women
about their mothers’ behavior.
The narrative aspect of Fisher’s study and the results on casualness-
demanding behaviors of fathers converge in support of the conclusion that
fathers of low-orgasmic women invest less in their daughters than fathers
of high-orgasmic women. Mothers’ behavior, however, was unrelated to
daughters’ sexual responsiveness during copulation. Also consistent is the
finding in many studies in diverse cultures (see Fisher, 1973, for review)
that women of low socioeconomic level consistently report significantly
lower copulatory orgasm frequency than women of the middle class. Pater-
nal investment is reduced and often absent in lower socioeconomic settings.
Another component of Fisher’s research is also consistent with our view
here. We have suggested that low-orgasm women are those who strategically
have less emotional commitment and sensitivity in mateships and less stable
pair-bonds. Fisher’s results on women’s fantasies and concerns in relation
to orgasm consistency show that low-orgasm women, in comparison to
high-orgasm women, are especially concerned about how transitory rela-
tionships in general are and how easily loved ones can be lost. Such concerns
in low-orgasm women may reflect a perception of greater likelihood of and
preparation for mate rejection or desertion.
Although Fisher’s various results discussed are supportive, they are only
moderately so. More research is needed to examine the relationship be-
tween women’s sexual arousal and resource level and parental involvement
in the home of origin. Measurement of resource availability and how paren-
tal behaviors actually relate to parental investment are critical. Women’s
recall of conditions in the home could be cross-checked by other observers.
Also, longitudinal studies of girls who subsequently become sexually active
women and who grow up in homes differing in parental investment could
address causal factors. The hypothesis proposed is one based on a condi-
tional expression of a sex-specific female sexuality adaptation. However,
genetic differences between families producing low- versus high-orgasmic
women could explain Fisher’s results as well as the socioeconomic pattern
in orgasm. Studies of female twins who are separated at birth and adopted
by families differing in the variables of interest (resource level, stability of
marriage, father presence/absence) could provide useful data to separate
the heritability of sexual arousal from condition-dependent effects on sex-
ual arousal.
C. CONCLUSIONS
There is little doubt that certain important aspects of women’s sexuality
and related heterosexual relationship behavior-age of puberty and first
sex, number of sex partners, stability of consortships, interpersonal attitude
toward mate (trust, sensitivity, interest, support)-are correlated with fac-
tors that would have reflected resource availability and reliability of parental
investment, especially father’s investment, in the rearing environments of
daughters in human evolutionary history. Women’s sexuality seems to show
two distinct modes, each composed of tactics that combine to form a sexual-
ity that is consistent with functioning sexually in either low- or high-resource
environments. Women raised under conditions of resource stress, that is,
limited resource base and parental investment, show early age of maturity
and sexual intercourse, more sexual partners, more infidelity, less enduring
pair-bonds, and a more opportunistic, insensitive, and exploitative orienta-
tion toward a mate. These tactics serve women’s interests in an environment
in which it was historically adaptive to pursue resources from multiple
mates. To the suite of tactics that are functional for women raised in limited
resource situations we have added the tactic of reduced and highly selective
sexual arousal, which we have suggested would promote pursuit of multiple
partners by reducing the emotional commitment shown by a woman toward
each partner.
The suite of sexual tactics shown by women reared in conditions of
resource availability and consistent parental investment also would fit them
to that adult environment. Accessing investment from single males would
be promoted by restricted sexual history, reduced infidelity, and greater
sensitivity and trust toward the mate. Orgasmic consistency may be an
important component of sexuality of women in environments in which single
males can provide sufficient resources for successful female reproduction.
The data bearing on the connection between orgasm consistency in
women and limited resources during development are far from convincing,
but are all consistent, from the father absence or disinterest data to the
concerns about loss of relationships shown by these women. Also, the effect
of socioeconomic level on women’s arousal corroborates the same trend.
As Fisher emphasized, more research needs to be done here, but there is
consistency in the data collected by different methods and investigators.
Fisher and others before him looked at many variables in attempts to
understand female copulatory orgasm variation. In all this literature, the
associations between orgasm frequency and the father-child relationship
and other measures of resource stress during upbringing emerge as the
most robust and consistent pattern.
Note that the argument we have provided makes no value judgments
about orgasmic capacity of women. Even a superficial survey of the female
sexual arousal literature reveals that many workers assume that female
sexuality is incomplete without copulatory orgasm, and women’s sexual
competence and general well-being are correlated with frequency of coital
orgasm (e.g., Fisher, 1973). This value system is based on a male sexuality
model: copulation equals orgasm. As Symons (1979) pointed out, women’s
more conditional sexual response often is seen as repressed from the ideal
of male sexual response. There is no way that our approach could inform
a value system based on orgasm as better among those who are naive about
the naturalist fallacy, because we are saying that sexual arousal patterns
of women from no arousal to orgasm are serving women’s evolved interests.
Of course, what has evolved or is otherwise natural can never be used
logically as a source of moral guidance. Those who define right and wrong
in terms of what is natural commit the naturalist fallacy.
Iv. POSTPARTUM DEPRESSION
Upon reaching reproductive maturity, human females become signifi-

cantly more likely than males to suffer serious psychological depressions
(Nolen-Hoeksema, 1990). This is not the case before puberty, and some
evidence even suggests that male children experience a greater incidence
of depressive symptomology than do girls (Nolen-Hoeksema, 1990). Tradi-
tional feminist interpretations of this increased vulnerability to depression
among reproductively mature women generally posit that psychological
stress due to social gender inequities are to blame, while biomedical re-
searchers have typically conducted studies based on the assumption that
sex hormones associated with reproductive status are at the root of women’s
greater incidence of psychological suffering. Such proximate-causation ap-
proaches do not address the evolutionary significance of psychological pain
itself, however, and thus lack an important component of any coherent
conceptual framework from which predictions about the nature of psycho-
logical pain can be generated and tested. In this section, we propose and
preliminarily test the hypothesis that one type of female depression-
postpartum depression (PPD)-may be an evolutionary psychological ad-
aptation for discriminative maternal solicitude, encouraging mothers to
cease investment in newborn offspring under circumstances in which the
offspring would have had low reproductive value in ancestral environments.
A. PSYCHOLOGICAL PAIN AS A N EVOLUTIONARY

ADAPTATION
Throughout much of this century, researchers treated emotions as though
they were little more than disruptions of normal psychological functioning
(see Nesse, 1990). Many behavioral biologists and psychologists, especially

those informed by a neo-Darwinian perspective, have since rejected this
assumed meaninglessness of emotions and have begun in recent years to
search for the functional significance of emotional states (e.g., Daly et al.,
1985; Leeper, 1948; Nesse, 1990; Thornhill and Thornhill, 1989; also see
Hamida, 1996). Yet, many biomedical researchers have continued to treat
depression as if it were always a “disorder,” a pathological malfunctioning
of neurological hardware rather than a fitness-buffering reaction to stress
(such researchers tend to emphasize the proximate causation of depressions
only, typically focusing on depression’s hormonal or other physiological
correlates; e.g., Goodwin et al., 1976; Livingston et al., 1978; Handley et al.,
1977; Ballinger et af., 1979; Okano and Nomura, 1992). This assumption
about the nature of depression has shaped doctors’ attitudes toward proper
treatment (usually, prescribing antidepressive drugs; Nesse, 1990 or manipu-
lating hormone levels; Gregoire et al., 1996), and has led to an inaccurate
presentation in the medical literature of “biological” and “environmental”
(or “psychosocial”) theories of psychological pain as alternative, mutually
exclusive explanations.
Evolutionary psychologists, on the other hand, have proposed that psy-
chological pain reflects evolved responses to environmental cues related
to impacts to a person’s inclusive fitness-that is, to evolutionarily relevant
sources of stress (Thornhill and Thornhill, 1989; Nesse, 1990). Thornhill
and Thornhill (1989) argue that psychological pain was selectively favored
through evolutionary time because it increased victims’ lifetime reproduc-
tive success and inclusive fitness by forcing “an assessment of the circum-
stances surrounding social problems in the lives of individual humans”
(Thornhill and Thornhill, 1989), and that depression is proximately caused
by dramatically fitness-reducing social events. Hence, psychological pain is
evolutionarily analogous to physical pain; pain receptors are found only
where they are useful in terms of correcting evolutionarily recurrent fitness
impacts (Alexander, 1986; Thornhill and Thornhill, 1989), and are designed
to demand immediate attention.
Evolutionary predictions about the proximal causes of different types
of psychological pain vary, reflecting the particular challenges to fitness
presented by different social circumstances during human evolution. If a
given form of depression is an evolved strategy for dealing with ancestral
fitness impacts, and occurs when current environments contain cues that
trigger those evolved states of psychological pain, then it should be possible
to make predictions about the social environment of the depressed person.
For instance, in a review of the largest investigation of victims’ anguish
following rape (McCahill et al., 1979), Thornhill and Thornhill (1989) pre-
dicted and confirmed that women of reproductive age suffer greater psycho-
logical distress following rape than do pre- or postreproductive-age victims,

reflecting the increased risk of pregnancy by an unwanted and uninvesting
mate. Another evolutionarily relevant cost of rape is the disruption of
support from an investing mate. Victims’ mates often view claims of rape
as denials of what was actually infidelity, and men frequently abandon
women after a sexual assault (McCahill et al., 1979). Hence, whereas
McCahill et al. expected that increased physical trauma during rape would
correlate positively with subsequent psychological pain, the evolutionary
psychological hypothesis of postrape anguish generated the opposite predic-
tion: that physical injury or signs of brutalization would constitute strong
evidence of actual rape victimization rather than consensual sexual infidelity
to mates, and would therefore correlate with lower levels of psychological
pain suffered by rape victims. A single episode of rape constitutes a less
serious paternity threat than multiple copulations during sexual affairs, and
is less likely to result in questions about fidelity in the minds of mates. The
counterintuitive prediction from the evolutionary hypothesis was supported
by the data presented by McCahill et al. (1979; Thornhill and Thornhill,
1989): physical trauma during rape negatively correlates with the amount
of psychological pain suffered by rape victims. (See Thornhill, 1997, for a
recent discussion of the evolutionary psychology of rape victim trauma.)
Postrape anguish is but one of the many forms of psychological pain
suffered by women. From an evolutionary perspective, it should be expected
that women would suffer a greater incidence of psychological pain, espe-
cially in response to events affecting their reproductive status. Women’s
lifetime offspring number is physiologically constrained such that each
reproductive event constitutes a larger proportion of a female’s lifetime
reproductive potential than a male’s.
B. POSTPARTUM
PSYCHOLOGICAL
PAIN
Maternal postpartum psychological pain was first described in the West
by Hippocrates (Jones, 1923). Many researchers believe that western
women are more likely to suffer serious depression soon after childbirth
than at any other time in life (but see O’Hara er al., 1990). In contrast
with pregnancy, which is associated with very few psychiatric conditions
(Paullekhoff, 1992), depression during the postpartum period is the most
common serious psychiatric syndrome among western women (Hamilton,
1988; Inwood, 1985), and appears to be more common than depression at
other times in life (Inwood, 1985; Paffenbarger, 1961; Whiffen, 1992). The
prevalence of PPD in Arab and Ugandan populations match that of western
nations (Cox, 1983; Ghubash and Abou-Saleh, 1997). Women with previous
histories of psychiatric disorders are at significantly more risk of postpartum
psychiatric hospitalization (Inwood, 1985). Because postpartum psychologi-

cal pain in these women cannot be disentangled from their preexisting
psychiatric illnesses, we will not consider this population in our review of
PPD. Further, because male postpartum depression does not share the
physiological etiology of maternal PPD (see Perry, 1992), we consider it
to be a distinct psychological phenomenon (although perhaps a convergent
one, shaped by related selective pressures during human evolution). Since
very little is known about paternal PPD, we deal only with maternal PPD
in this chapter.
Feminist perspectives on PPD have emphasized the distress inflicted on
women during child birth in male-dominated medical settings in which
“birth is manipulated to suit institutional requirements and women are
pressured to accept passive and dependent roles” (Oakley, 1980). However,
empirical evidence fails to support this view’s prediction that a greater
proportion of hospital births than home births in the West leads to PPD
(Pop et al., 1995).
Maternal postpartum dysphoria varies along a continuum of symptomol-
ogy from ubiquitous but mild and transient “baby blues” to “postpartum
psychosis.” Baby blues occur from a few hours to a few days after giving
birth, and involve bouts of tearfulness and relatively mild levels of anxiety
(Hamilton et al., 1992). More serious is PPD, a clinical depression involving
longer lasting feelings of grief, guilt, irritability, tearfulness, feelings of
hopelessness and failure, and apathy or hostility toward the newborn (Ham-
ilton et al., 1992). Typically, PPD is mild compared to nonpostpartum
clinical depressions and involves a markedly lower incidence of suicidal
ideation (Whiffen, 1992). Serious, clinical PPD symptomology often begins
after a few weeks postpartum and, unlike the baby blues, lasts up to several
months (Hamilton et al., 1992).
Unusually serious cases of PPD are sometimes called “postpartum psy-
chotic depression” if onset is acute immediately after childbirth, but in
diagnostic practice there is considerable overlap and confusion about such
differentiations (Hamilton et al., 1992; Paullekhoff, 1992). Postpartum psy-
chosis (PPP) is typified by delusional and obsessive thoughts about failure,
an inability to love or care for the newborn, and guilt resulting from these
feelings (Herzog and Detre, 1976; Paullekhoff, 1992). As Daly and Wilson
(1995) point out, however, the content of these “delusions” is not always
particularly illogical (see also Herzog and Detre, 1976). Beliefs in many
cultures around the world that deformed newborns are demons or demonic
progeny are irrational and may be considered psychotic by western psychia-
trists, but the resulting infanticidal behavior-while morally reprehensible
to us-is nevertheless adaptive, evolutionarily speaking (Daly and Wilson,
1995). We will treat PPD and PPP as different points along the same
continuum of psychological distress (collectively referred to by us simply

as “PPD”), because even if real psychiatric differences exist between the
two diagnoses, they give rise to similar fitness-affecting behaviors (most
notably, neglect or abuse of newborns).
Immediately after birth, a mother is apparently at special risk of de-
pressive symptomology because of reduced levels of the stress-mediating
corticotropin-releasing hormone (CRH; Magiakou et al., 1996), rendering
her vulnerable to sources of stress. Deficits in corticotropin-releasing hor-
mone are not a sufficient explanation of PPD’s origins until we answer
the question of why this postpartum vulnerability to stress (with its often
dramatically fitness-affecting results) was not expunged from human psy-
chological design by natural selection during human evolution. One intu-
itive and potentially popular answer to this question is that the female
human body plan is somehow physiologically constrained, such that no
escape from vulnerability to PPD has ever emerged throughout the course
of human evolution. For instance, CRH levels may regulate prenatal and
delivery events (Karalis et al., 1996), and until postpartum maternal hor-
mone levels are differentiated from a prenancy-typical profile, CRH may
confound postpartum behavioral adaptations. A more readily tested alter-
native hypothesis is that natural selection acting in ancestral populations
favored a maternal vulnerability to stresses after giving birth.
C. PARENTAL
INVESTMENT
BEHAVIORS
I N EVOLUTIONARY
CONTEXT
As stated earlier (Section II,C), parents do not invest maximally in all of
their offspring. To understand the evolved psychology of variable parental
investment, we must clarify the selective pressures responsible for shaping
human parental psychology. Parent-Offspring Conflict Theory was the first
systematic attempt to delineate the selective forces that shape interactions
between parents and their offspring in modern, neo-Darwinian terms (Triv-
ers, 1974). In this section, we introduce Trivers’s theory, and then describe
the more recent, complimentary model of differential parental solicitude.
Finally, we place the study of PPD in this evolutionary framework and
present predictions derived from our hypothesis about PPD’s evolved func-
tion. These predictions are tested against the medical literature in the
following section. While extensive and representative, our review of the
literature is almost certainly not exhaustive.
Trivers argued that the nonidentical genetic self-interests of diploid par-
ents and offspring leads to natural selection favoring behavioral strategies
of mutual exploitation (Trivers, 1974). To some degree, for instance, off-
spring who exaggerate their nutritional needs will be favored over siblings
who do not, even if their exaggeration reduces their parents’ long-term
fitness. Parental psychology and reproductive physiology, on the other hand,

are expected to be designed to counter exploitation by offspring (Haig,
1993; Peacock, 1991) and to favor investment in those offspring that have
higher probable eventual fitness (an offspring’s “reproductive value”).
Hence, contrary to popular notions of automatic and unconditional mother-
infant bonding, evolutionary theory leads to the prediction that parental
investment in offspring will be preceded by an evaluation of the newborn’s
reproductive value (Daly and Wilson, 1995).
This line of thinking-termed “differential parental solicitude” by Daly
and Wilson (e.g., 1988)-has been expanded in recent years, and now
includes an explicit model of maternal bonding psychology to replace the
simpler, naive version of unconditional, “automatic” bonding, which still
enjoys wide popularity (Daly and Wilson, 1995; Eyer, 1994). The evolution-
ary bonding model posits three stages of maternal attachment to a newborn
(Daly and Wilson, 1995):initial assessment of the baby’s reproductive value,
followed by discriminative emotional attachment to the baby, followed in
turn by a gradual deepening of appreciation and attachment to that individ-
ual as his or her reproductive value increases with age.
Infanticidal behaviors frequently correlate with low infant reproductive
value (Daly and Wilson, 1988), and involve a failure of the second step of
maternal attachment, as described above. In many cultures, including those
of the industrialized West, mothers react to cues of low infant reproductive
value with lethal neglect or active infanticide; in other words, mothers make
a negative assessment of offspring reproductive value and act on evolved
motivational states that encourage cessation of investment in the offspring
(Daly and Wilson, 1995). Infants’ deformities are grounds for socially ac-
cepted infanticide in many cultures, for instance (Daly and Wilson, 1988).
This infanticide is an example of what Moiler (1997b) calls developmental
selection against developmentally unstable offspring.
This postnatal assessment of offspring fitness may allow a more subtle
evaluation than prenatal maternal assessments, and hence, postnatal infanti-
cide may be a more fitness-enhancing strategy of investment withdrawal
than adaptive (spontaneous) abortion. Human mothers represent their off-
spring’s main or only source of nutrition for the first 6 months to 2 years
(Wood, 1994). Since lactation is significantly more metabolically costly to
the mother than gestation (Miller and Huss-Ashmore, 1989; Worthington-
Roberts et al., 1985), infanticide may constitute a strategic maternal invest-
ment withdrawal at the beginning of the truly costly stage of reproduction.
It is important to note that the first stage of maternal attachment (assess-
ment of reproductive value) includes factors in a mother’s life that are
extrinsic to the child’s inherent reproductive value, because there are cir-
cumstances under which even the healthiest newborn is unlikely to survive
to reproductive maturity. Hence, if PPD has been designed by natural

selection to motivate strategic investment withdrawal from infants of low
reproductive value, predictions other than higher incidence of PPD in
mothers of ill or deformed infants can be generated and tested.
We propose the hypothesis that PPD is a form of evolved psychological
pain, designed to encouarge maternal withdrawal of investment from off-
spring when doing so would have increased long-term maternal fitness in
ancestral environments. We therefore expect that stress factors that would
have related to the intrinsic and extrinsic reproductive value of a newborn
during human evolution will be highly correlated with the incidence of
maternal PPD. We derive six predictions from our hypothesis that we
believe must be true if our hypothesis is correct:
1. Cues of compromised health and/or developmental integrity of the
newborn will correlate positively with the incidence of maternal PPD.
2. Low levels of paternal investment will correlate positively with the
incidence of PPD.
3. Low levels of social support from kin and social allies other than an
investing mate will correlate postively with the incidence of PPD.
4. Cues of a compromised resource base (famine or poverty) will corre-
late with the incidence of PPD.
5. Increases in the incidence of infanticidal ideation by mothers will
accompany PPD.
6. Cultures in which ritual displays of social support and paternal invest-
ment do not accompany childbirth will have an increased incidence
of PPD.
In the following sections, these predictions are tested against the existing
English-language medical literature on PPD.
1. Infant Health and the Incidence of PPD
If postpartum depression is a psychological adaptation encouraging the
cessation of investment in offspring of low reproductive value, then the
most obvious correlate of PPD should be unhealthy infants. Indeed, mothers
of infants at higher medical risk report higher levels of emotional distress
and PPD symptomology, difficulty expressing affection toward their baby,
and greater dissatisfaction with the levels of social support they receive
(Bennett and Slade, 1991).
Kumar and Robson (1984) found a significantly increased incidence of
PPD in mothers of premature babies compared to mothers of full-term
newborns. Premature infants would have had slim survival prospects in
ancestral environments, so evolved discriminative parental investment
mechanisms may assess these infants as possessing low intrinsic reproduc-
348 R A N D Y THORNHILL A N D B R Y A N T FURLOW
tive value despite recent medical advances that enhance the odds of survival
for such newborns. Condition of the newborn negatively correlates with
PPD (Hopkins etal., 1987). However, Davidson (1972) found no correlation
between infant state and PPD.
Medical complications during delivery (e.g., excessive bleeding or fever
during delivery, emergency cesarean section) correlate significantly with
PPD in some studies (Campbell and Cohn, 1991; Morgenshy, 1982; O’Hara
et al., 1984; Paykel et al., 1980) but not others (O’Hara et af., 1982). Burger
et al. (1993) found that women with severe complications during pregnancy
are significantly more prone to suffer PPD, even after variables such as
premature birth and neonatal hospitalization were controlled for. It appears
possible that such pregnancy or delivery-related “emergency” cues can
negatively affect maternal assessment of infant viability.
2. Social Support and P P D
Low levels of social support are markedly related to a higher risk of
PPD (Cutrona and Troutman, 1986; Nilsson and Almgren, 1970; O’Hara
et al., 1983; Pitt, 1968; Richman et al., 1991; Spangenberg and Pieters,
1991). Even the additional social support afforded by volunteer “labor
companions” from the local community reduces the risk of PPD in middle-
class South African women (Wolman et al., 1993). Among Arab women,
presence of a housemaid may reduce the risk of PPD (Ghubash and Abou-
Saleh, 1997). A review of mild PPD (cases not characterized as psychotic)
reveals that a lack of social support is significantly correlated with higher
rates of PPD (Cutrona, 1982). Workplace social support may be relevant
to postpartum adjustment as well (Leathers et al., 1997). However, Hopkins
et af. (1987) report no significant correlation between PPD and low levels
of social support in the sample they studied. A recent review by Wilson et
al. (1996) reports “fair” evidence for a PPD/social support relationship.
Poor relationships of women with their mothers is related to their risk
of experiencing PPD (Douglas, 1963; Kumar and Robson, 1984; Richman
et af., 1991). These relationships may be viewed as indications of low levels
of kin support, an important component of social support networks for
new mothers. In adolescents, the correlation between poor relationship
with own mother and PPD may be due to the fact that the emotionally or
materially uninvesting mother is the new mother’s primary or only source
of support (see Quijano and Cobliner, 1983).
3. Paternal Investment and Maternal PPD
A perceived lack of support from a husband is associated with maternal
PPD (Buchwald and Unterman, 1982; Unterman el af., 1990). Help with
parenting duties, displays of interest in the newborn, and support for the
STRESS A N D HUMAN BEHAVIOR 349
mother correlate with a significantly reduced incidence of maternal PPD,

whereas marital discord or lack of investment behaviors by the father
correlate with a higher incidence of PPD (Campbell and Cohn, 1991;Collins
et al., 1993; Gordon and Gordon, 1967; Kumar and Robson, 1984; Logsdon
et al., 1994; Paykel et al., 1980; Richman et al., 1991; Sosa et al., 1980;
Spangenberg and Pieters, 1991). Arab women with PPD are significantly
more likely to report marital probkms than controh (Ghubash and Abou-
Saleh, 1997). Indeed, Close (1980) argues that the best treatment for PPD
is extra affection, understanding behavior, and help with parenting from
the husband (see also Hickman 1992). Barnett et al. (1996) report that
among adolescent mothers, social support from their mothers or from the
infants’ fathers was significantly associated with lower rates of PPD. Wilson
et al. (1996) report “good evidence for association” between poor marital
relations and PPD, in their recent review.
In this section, we have emphasized the hypothesis that PPD constitutes
a mechanism whereby maternal investment in a newborn is suspended
because of low newborn reproductive value. Another purpose for PPD
may be the demonstration of need to people likely to respond by changing
the environmental or social conditions that are extrinsic to the offspring
but nevertheless lower its reproductive value. In the only other evolutionary
psychological analysis of PPD that we are aware of, it has been proposed
that in addition to investment withdrawal from infants, defecting from
univesting or coercive mateships is a primary function of PPD (Hagen,
1996). This display-component hypothesis posits that PPD forces interested
parties to invest more in the mother and her offspring lest she abandon
the infant. Hence, PPD may serve as an unconscious strategy for displaying
need to a social support network via a threat of child neglect or infanticide,
and the vulnerability to PPD may have been evolutionarily retained not
just as an effective mechanism for strategic resource withdrawal from off-
spring, but also may have enhanced maternal fitness by forcing others to
correct circumstances that lowered offspring reproductive value in ancestral
environments. Symptoms such as weeping and irritability, if this is the case,
are best viewed as social signaling (Hagen, 1996), while apathy or hostility
toward the newborn reflect a resource-withdrawal function of PPD.
It is interesting to note that while adaptationist explanations have been
proposed for human grief displays in other contexts (such as the loss of a
loved one), the honesty of grief displays is suspect because of the potential
fitness benefits of exaggeration (Thornhill and Thornhill, 1989). Whether
a display is a reflection of real need or whether it is a counterfeit perfor-
mance is not easily assessed by onlookers. For displays to constitute an
evolutionarily stable signaling strategy, they either must make assertions
that are readily verified by signal receivers, who in turn reliably punish
exaggerations, or they must carry a physiological or fitness cost to the signal

sender (a “handicap”; Grafen, 1990; Zahavi, 1977b). Weeping is a species-
typical correlate of human grief, and is an obvious component of PPD.
Tearfulness in adult women is immunosuppressive (Labott et al., 1990;
Martin et al., 1993); sad mood alone does not affect immune system function-
ing, but shedding tears results in depressed salivary immunoglobulin-A
counts. Since immunological handicapping serves as a physiological en-
forcer of signal reliability in other signaling contexts (Folstad and Karter,
1992), we speculate that weeping may be a Zahavian social signal of need.
Assuming that maternal tearfulness during PPD is physiologically similar
to experimental subjects’ evoked tearfulness, immunosuppression may have
limited the exaggeration of need during PPD weeping displays to social
allies among our ancestors, leaving tears, on average, an honest indicator
of a woman’s psychological pain and real need. Even small degrees of
immunosuppression during the postpartum period may have had important
consequences in human evolutionary history because of the greater impact
of parasites than that existing in contemporary western societies.
4. Resource Base and PPD
Women living in lower socioeconomic strata (Davidson, 1972), recently
immigrated to a new nation (Zelkowitz and Milet, 1995), or expressing
perceptions of poor housing quality (Paykel et al., 1980) experience a higher
incidence of PPD than other women, but not a higher incidence of less
serious forms of postpartum dysphoria (“blues”). Socioeconomic level
involvement in PPD incidence may be confounded by some unmeasured
variable such as a higher incidence of poor marital relations in lower socio-
economic stratum marriages. Women who express concern over finances
or their husband’s employment reliability exhibit an increased incidence
of PPD (Heitler and McCrensky, 1976), while women who are themselves
employed experience lower rates of PPD than do women without jobs
(Richman et al., 1991; Zelkowitz and Milet, 1995). Western housewives
may perceive poor social support because they experience relative social
isolation, caring for infants in the home (E. H. Hagen, personal communi-
cation).
Unterman et al. (1990) state that “economic problems” constitute a risk
factor for PPD. During the Great Depression, rates of PPD in the United
States increased significantly, perhaps partly because of an increased inci-
dence of unemployment or general employment insecurity (Wick, 1941).
5. Infanticidal Ideation and PPD
As stated earlier, women suffering PPD exhibit decreased levels of sui-
cidal ideation compared to those suffering nonpostpartum clinical depres-
sions. However, a higher incidence of homicidal (infanticidal) ideation

associated with PPD is reported in the literature (Beck, 1992; Wisner et
al., 1994), as is an increased incidence of actual infanticides (Herz, 1992;
Kumar and Marks, 1992). Infanticidal thoughts appear to be a major source
of the guilt feelings suffered by women during PPD (Beck, 1992; Wisner
et al., 1994), and may be more common than is reported. A number of
authors report that women suffering from PPD are unable to feel love or
maternal concern for their newborns (e.g., Affonso and Arizmendi, 1986;
Kumar and Robson, 1984).
6. Cultural Traditions and PPD
PPD appears to be very rare in traditional, nonwestern cultures (Hark-
ness, 1987; Kelly, 1967; Kruckman, 1992; Stephenson et al., 1978). The
traditionally recognized “amakiro” postpartum illness in Uganda, typified
by mental confusion and a desire to kill the newborn, may represent an
exception (Cox, 1978,1979). Domination by colonial cultures and resulting
disruption of traditional native culture appears to result in the incidence
of PPD similar to that of Westerners (e.g., East Africans and New Zealand’s
aboriginal Maori; Harris, 1981; Webster et al., 1994). Apparently, people
from traditional cultures are no less susceptible to PPD, given certain
stresses, than Westerners are.
Those cultures that observe traditional, ritual displays of social support
to the new mother seem largely “immune” from PPD (Kelly, 1967; Kruck-
man et al., 1983; Upreti, 1979). Among the Nigerian Ibibio, for instance,
women are placed in a “fattening room” after giving birth, while others
take over their daily duties (Kelly, 1967). In China and Taiwan, a traditional
postpartum custom of “doing the month” (to yueh; 30 days of rest) similarly
demonstrates social support to the new mother, who is viewed as unusually
vulnerable and in need of great physical and emotional support (Fried and
Fried, 1980; Pillsbury, 1978). The Chinese and the Ibibio both apparently
enjoy an absence of PPD, according to ethnologists (although it should be
noted that ethnologists often have low sample sizes). It appears that displays
of social support from a woman’s kin and allies can mitigate the stress
leading to PPD, and that people around the world have noted a mother’s
special postpartum vulnerability and responded with cultural practices that
mitigate relevant sources of stress.
D. CONCLUSIONS
Our review of the medical literature on postpartum depression provides
preliminary, correlational support for our hypothesis that PPD constitutes
an evolved mechanism of differential maternal solicitude, discouraging in-
vestment in offspring under circumstances in which the offspring are un-

likely to survive to reproductive maturity. Two of the eighteen studies
reviewed with regard to our first two predictions failed to support our
hypothesis; Davidson (1972) found no clear correlation between infant
condition and incidence of PPD, and Hopkins et al. (1987) found social
support to play an unimportant role in subjects’ depressions. The variance
of infant conditions and the range of social support represented in a given
study is likely to affect detection of correlations between social stress and
PPD. Thus, these two studies may not be exceptions to our hypothesis;
however, even if they are treated as true exceptions, the overall pattern is
highly supportive of our hypothesis (sign test for predictions 1-6,35 positive
excluding the Wilson etaf., 1996 review [also positive], 2 negative,p < .OOl).
Available cross-cultural evidence suggests that many cultures observe
traditional displays of social support that reduce the incidence of PPD. If
similar preventative measures are adopted in western nations (for instance,
doctors could discuss with a woman’s kin and mate the importance of
displays of a willingness to help raise the coming child), we predict that
the incidence of PPD would drop dramatically. Little research has been
done to evaluate whether identifying and correcting the relevant social
variables after PPD has been diagnosed can improve subsequent mother-
offspring interactions, but this is a reasonable prediction of both our and
Hagen’s hypotheses. Both our and Hagen’s hypotheses identify selective
pressures, which, through evolutionary time, could have maintained mater-
nal vulnerability to PPD. Predictions from the two models overlap, and both
receive support in the medical literature. Future research should emphasize
distinctions between evolved social need signal components (e.g., weeping)
and resource withdrawal components (e.g., infanticidal urges) of PPD, so
that the relative importance of each of these apparent evolved functions
of PPD may be approximated.
V. INFANT
CRYING
AS A OF PHENOTYPIC
SIGNAL QUALITY
Care-soliciting vocalizations are not limited to human neonates. They

are common in bird and in other mammal species, and occur throughout the
primate order (Newman, 1985). Human neonatal crying generally enhances
caregiver proximity and investment (Ainsworth, 1969), but also carries
serious risks as an apparently causal focal point of child abuse and even
infanticide (Frodi, 1981; Steele and Pollack, 1968; Weston. 1968). This has
been called the “crying paradox” (Barr, 1990), because crying is apparently
necessary for the neonate to elicit care but may result in fitness-impacting
reactions by caregivers. A resolution of this apparent paradox must explain
both the beneficial and the fitness-reducing (for the baby) aspects of caregiv-
ers’ reactions. It has been proposed that the resolution of the crying paradox
rests with an understanding of evolved patterns of differential parental
solicitude (Furlow, 1997), discussed earlier in this chapter (Section IV,C).
Zahavi (1975, 1977a) and others (e.g., Grafen, 1990) argue that because
deceived signal receivers regularly suffer significant fitness impacts (unnec-
essarily abandoning resources to a bluffing competitor, for instance, or
mating with a genetically substandard suitor), signal receiving mechanisms
should evolve that favor signals that are costly to send. Signal costs (fitness
impacts) limit exaggeration during communication between organisms dur-
ing which “claims” are being made about a signal sender’s phenotypic
quality, because only high-quality signal senders can afford to suffer the
associated fitness reductions. Just as it is important for courted female birds
or competing red deer males to accurately evaluate the phenotypic quality
of the individuals before them, parents who discriminatively evaluate the
phenotypic quality of their offspring before responding to offspring solicita-
tion can avoid the fitness costs of investing in nonviable babies.
While attention has been paid to the potential for deceptive solicitation
by primate infants (e.g., Hauser, 1986), few attempts have been made to
analyze human crying in the context of the evolutionary theory of honest,
signaling. Godfray (1991) addresses the evolutionarily stable signaling of
offspring need in vertebrates, and presents a model that presupposes that
some sort of criteria exists for parental assessment of offspring reproductive
value, as the intrinsic reproductive value of a newborn is as important from
the parental perspective as its asserted level of need. Some species appear
to have such fitness-assessment components in their neonatal solicitation
displays (Lyon et al., 1994; Bustamante et al., 1992). In this section, we
evaluate predictions derived from the hypothesis that the human neonatal
cry itself contains criteria that allow parental assessment of offspring repro-
ductive value. Hence, whereas previous authors have studied the motiva-
tional (offspring need) component of crying, we focus instead on the ques-
tion of whether the acoustic structure of cry vocalizations accurately
indicates the condition of a crying neonate, and hence, correlates with its
intrinsic reproductive value. Parental bonding and investment patterns
hinge on an early critical assessment of offspring reproductive value (Daly
and Wilson, 1995), and we propose that the earliest example of human
vocal communication-the neonatal cry-constitutes an important source
of offspring fitness information for parents during this assessment period.
The previous section of this chapter, on postpartum depression, included
“infant condition” as one variable of importance to maternal assessments
of an offspring’s reproductive value. Many physical and behavioral cues of
offspring health are available to parents (e.g., rash, irregular breathing),
354 R A N D Y THORNHILL A N D B R Y A N T FURLOW
but because the cry involves neurological, cardiorespiratory, and vocal

phenotype, which are not as readily assessed as a rash on the skin, crying
may reveal information about the internal structural and functional integrity
of the offspring. Selection on parents to evaluate offspring reproductive
value should have favored discriminative responses to these cues. (Discrimi-
native investment is not limited to intrinsic offspring fitness, of course.)
An honest signal model of infant crying predicts that crying will be more
expensive to infants with relatively low phenotypic quality than to those
with relatively high intrinsic fitness. Indeed, the energetic expense of crying
is 11% higher than basal infant metabolism (Brignol et al., 1993). This
expense is not likely to impact on healthy, well-nourished infants, but
probably constitutes a serious threat to the fitness of ill and/or malnourished
infants who must partition energy to mounting immunological defenses or
who simply lack sufficient energy to afford the expense of crying. This is
consistent with the expectations of the Zahavian paradigm of signal evo-
lution.
If infant crying contains acoustic cues of reproductive value, and is a
costly honest signal in the Zahavian sense, then two testable predictions
become immediately apparent: acoustic cry parameters will be correlated
with infant phenotypic condition, and the relevant cry parameters will
correlate with parental investment behaviors (or with emotions, like hostil-
ity or concern, which probably inspire given investment behaviors, like
infanticidal abuse, neglect, or immediate investment).
A. OFFSPRING
CONDITION
A N D CRYACOUSTICS
Healthy human infants’ cries have an average fundamental frequency

(pitch) of approximately 300-600 Hz (Furlow, 1997; Zeskind, 1983). From
brain-damaged infants to malnourished, premature, asphyxiated, and para-
sitized infants, low infant fitness is correlated with significantly increased
cry pitch (Frodi, 1985; Morley et al., 1991; Wasz-Hockert et al., 1985). Cry
pitch is more exclusively indicative of serious illness than symptoms typically
noted by pediatricians, including a change in respiration or pulse rate,
rash, nasal discharge, stridor, pale skin, wheezing, temperature, sweating,
dehydration, or behavioral symptoms (Morley et af., 1991).
The magnitude of pitch abnormality correlates with the magnitude of
the infant’s fitness impacts. Diabetic mothers’ infants have a mean maximum
pitch of 1480 Hz, whereas infants of diabetic mothers who are also hypogly-
cemic produce cries with a mean maximum pitch of 1520 Hz; hypoglycemic
infants of diabetic mothers who have hyperbilirubinemia (jaundice) have
a mean maximum pitch of 1980 Hz (Wasz-Hockert et al., 1985). Likewise,
the severity of asphyxiation can be predicted by the magnitude of cry pitch
STRESS AND H U M A N BEHAVIOR 355
abnormality (Michelsson, 1971). Infants with only peripheral asphyxia have

a mean maximum pitch of 1000 Hz, but those with central asphyxia produce
cries with a mean maximum pitch of 1460 Hz. Central asphyxia causes
neurological damage, and constitutes a more serious impact to offspring
reproductive value. The severity of pitch abnormality in infants with bacte-
rial meningitis is more pronounced (higher pitched) in babies later diag-
nosed with neurological sequelae (Michelsson et af., 1977).
Cry pitch is also associated with subsequent cognitive development (Don-
zelli et af., 1995; Lester, 1987). At 18 months and 5 years of age, children’s
scores on cognitive tests were predicted by analysis of cry pitch during
early infancy-babies with higher and more variably pitched cries score
significantly lower on cognition tests than other children (Lester, 1987).
Neurodevelopmental integrity is evidenced by normal cry pitch (Donzelli
et al., 1995).
A N D PARENTAL
B. CRYING REACTIONS
Crying is recognized in many cultures as an important sign of a healthy
infant. Among the African Igbo people, for instance, babies who do not
cry vigorously are abandoned in the forest (Basden, 1966). Laboratory
studies of adult reactions to recorded cries report a negative emotional
reaction to high-pitched cries in western cultures (e.g., Crowe and Zeskind,
1992; Zeskind and Lester, 1978). Adult reactions to cry pitch are similar,
despite differences in caregiving experience, subject gender, and age (Fur-
low, 1997). Cardiac response and skin conductance levels support self-
reported emotional states in study participants (Frodi et af., 1978; Weisen-
feld et af., 1981). Heart rate increases are greater when participants are
played high-pitched cries than when they are played normal cries, and skin
conductance diminishes with time to normal cries, but not to high-pitched
cries (Frodi et af., 1978; Weisenfeld et af., 1981).
Pitch appears to be unrelated to aversiveness ratings under 610 Hz,
roughly the upper limit of the range of normal (healthy) cries (Bisping et af.,
1990). Hence, adults’ ratings of cry aversiveness correlate with infant health.
It should be noted that cry pitch may be linked with another, unmeasured,
acoustic parameter, and may not itself be the exact parameter adults re-
spond to in reaction-to-cry studies. This seems unlikely, however, as pitch
is obviously a salient acoustic parameter of cry vocalizations. The role of
cry pitch in the incidence of PPD has yet to be studied.
C. CONCLUSIONS
Crying conveys information about infants’ health and may serve as an
important cue of offspring reproductive value for parents. Parents, in turn,
react negatively to at least one acoustic correlate of low infant reproductive

value (pitch). If aversiveness ratings (and their physiological correlates) in
the lab are indicative of parental emotions to crying in more natural settings,
it would appear that unhealthy children are likely to suffer reduced invest-
ment or even parental hostility. Zahavian selection for a signal that accu-
rately communicates signal senders’ phenotypic quality seems to have
played a potentially important role in the evolution of the human neonatal
cry. A limitation to this review is that it is based on correlational data, and
with the exception of adult emotional reactions to variation in cry acoustics,
causality is therefore assumed rather than empirically established.
VI. SUMMARY
We addressed four major topics under the heading of stress and human
behavior that are currently advancing areas of research in human behavior.
The first topic is human attraction and attractiveness. It is well established
that looks or physical attractiveness matter a great deal in everyday human
life and that attractive juveniles and adults of both sexes have social advan-
tages. Only relatively recently have evolutionary psychologists and behav-
iorists explored in any detail the evolutionary basis of attraction and attrac-
tiveness in humans using hypotheses based on modern sexual selection
theory. The hypothesis that attractiveness is a phenotypic marker of stress
resistance is discussed, as is the complimentary hypothesis that attraction
involves assessment of stress resistance. These hypotheses entail selection
favoring individuals who viewed physical beauty as phenotypic stress resis-
tance, and selection favoring individuals who displayed their stress resis-
tance in physical features.
Stress resistance is equated with low developmental instability, particu-
larly low fluctuating asymmetry, and well-developed body and facial second-
ary sexual traits. Fluctuating asymmetry is known to be caused by various
environmental and genetic perturbations acting during development. The
development of secondary sexual traits in humans is mediated by sex hor-
mones, which handicapkompromise the immune system. Thus, low asym-
metry and highly developed secondary sexual traits signal an ability to cope
with stresses during development.
Evidence is reviewed showing that body and facial symmetry, as well
as secondary sexual traits, have positive effects on sexual attractiveness.
Attractive faces of both sexes exhibit relatively low asymmetry and rela-
tively high development of sex-specific facial sex hormone markers. Facially
symmetrical individuals and individuals with attractive expressions of facial
secondary sexual traits exhibit greater emotional and physiological health,
a further indicator of their ability to resist stress. Also, nonfacial body

symmetry correlates with attractiveness in men when attractiveness is mea-
sured as mating success (sex partner number, age of first sex, and other
related factors). Moreover, nonfacial body attractiveness in both sexes is
explicable by sex-specific sex hormone effects. It is concluded that human
attraction and attractiveness are importantly related to stress resistance or
general health and phenotypic and genetic quality.
Most of the research on sexual attractiveness has involved correlational
analyses with statistical control of confounding variables. However, some
of the recent work on facial attractiveness has been experimental. There
is need for further, careful experimentation with symmetry and secondary
sexual traits.
We argue that humans have been selected to make physical attractiveness
judgments in the context of nepotism and reciprocal altruism based on the
two markers of stress resistance: low asymmetry and secondary sexual
traits. There is vast evidence that attractiveness affects nepotism as well as
reciprocity such as friendships. However, the role of the two markers of
stress resistance per se in this effect has not been studied. This may be a
fruitful research domain in human behavior.
The second area of behavioral research explored in this paper is the
relationship between women’s sexual arousal and the stressful condition
during women’s upbringing of limited or absent paternal investment. It
has been hypothesized by developmental evolutionary psychologists that
paternal divestment during development has important consequences for
women’s sexuality. Empirically, the divestment appears to relate to earlier
sexual maturity and sexual intercourse, reduced stability of romantic pair-
bonds, more sex partners, and emotional coldness toward mates. This suite
of sexual traits has been interpreted by evolution-minded psychologists as
evolved condition-dependent tactics for securing material benefits from
many males in a social environment in which paternal investment from
individual males is not reliable. In an ecological setting of paternal divest-
ment, the suite of female sexual traits mentioned is viewed as maximizing
the amount of male-provided benefits by consortships with many men, each
of whom is capable or willing to invest to a small degree. We hypothesize
that reduced frequency and highly selective female orgasm would be func-
tional for women to couple with the other traits mentioned for exploiting
resources held by men in a social environment of male paternal divestment.
Female copulatory orgasm apparently has a selective bonding function.
Thus, limited orgasm would promote women’s strategic pursuit of multiple
partners by reducing the emotional commitment shown by a woman toward
individual partners. The literature of female orgasm frequency is examined
in the light of this hypothesis. As predicted, there is a positive relationship
between orgasm frequency in women and paternal investment during a

girl’s upbringing. We emphasize that more research is needed to clarify
the role of a father’s involvement on a developing female’s orgasm behavior.
Next, we turn to the fascinating behaviors and associated mental states
of women’s postpartum depression (PPD). Data derived from the medical
literature on PPD is consistent with the predictions of our evolutionary
psychological “resource withdrawal function” hypothesis for PPD. Through
human evolutionary history, females who withdrew resources from infants
exhibiting cues of low phenotypic quality (and therefore, low reproductive
value to parents), or who withdrew resources from offspring during times of
resource limitation would have improved their own long-term reproductive
success. Mothers who invest indiscriminately in unhealthy offspring or who
invested maximally during times of resource shortages would have been
selected against, and should not represent our ancestors. As we predicted,
cues of infant health, social support, mate support, and compromised re-
source base all correlate positively, in general, with the incidence of PPD.
Cultures that observe ritualized, exaggerated displays of social and material
support to new mothers seem to have a remarkably low incidence of PPD,
but in those cultures in which European imperialism has disrupted tradi-
tional cultural customs, the incidence of PPD appears to be markedly
higher. In addition, women suffering from PPD more often report thinking
about killing their babies, and the incidence of actual infanticides is higher
among PPD sufferers than among other mothers, as should be expected if
PPD is an evolved baby killing (resource withdrawal) adaptation. In this
way, vulnerability t o PPD could have been shaped and retained for its
fitness-enhancing benefits for sufferers living in ancestral environments.
We d o no argue or assume that PPD is currently adaptive.
Our final topic is the baby’s cry. To our knowledge, no researcher has
yet studied the role of abnormal acoustic quality in babies’ cries in the
etiology of PPD. Based on our review of the relevant literature, we predict
that abnormal cry quality contributes to maternal PPD, because the cry
represents an assay of internal phenotypic quality (and hence, reproductive
value) of offspring, and abnormal cry quality is a potent indicator of compro-
mised infant phenotypic quality. Human neonatal crying appears to have
been shaped by natural (parental) selection to reveal the soliciting off-
spring’s reproductive value, in addition to any need assertion or need-type
communication (pain, hunger, fear, etc.) conveyed by cries. As is the case
with our review of PPD, all published support for the Zahavian or “honest”
signal hypothesis of cry quality is correlational rather than experimental.
A further limitation is that studies of adult responses to infant cry quality
are almost entirely limited to the western, industrialized nations. One of
us (Furlow) is initiating cross-cultural studies of adult reactions to cry

quality in the neotropics.
Acknowledgments
R T and BF worked together on all topics in the manuscript. RT was invited to write the
manuscript and he is. therefore, first author. M. Milinski, A. Moiler, and P. J. B. Slater provided
provided helpful suggestions on the manuscript. J. Belsky, E. Cashdan, K. MacDonald, and
M. Surbey provided useful discussion about fathers and women’s sexuality. E. Hagen’s com-
ments on postpartum depression were helpful, as were references on lactation that he brought
to our attention. A. Rice’s help in preparation of the manuscript is greatly appreciated.
References
Affonso, D., and Arizmendi, T. (1986). Disturbances in post-partum adaptation and depressive
symptomatology. J . Psychosom. Obstet. Gynaecol. 5, 15-32,
Ahmed, S. A.. and Tala], N. (1990). Sex hormones and the immune system. Part 2. Animal
data. Bailliere’s Clin. Rheumatol. 4, 13-31.
Ainsworth, M. (1969). Object relations, dependency, and attachment: A theoretical review
of the mother-infant relationship. Child Dev. 40, 969-1025.
Alexander, J., and Stimson, W. H. (1989). Sex hormones and the course of parasitic infection.
Parasitol. Today 4, 1891-1 893.
Alexander. R. D. (1986). Ostracism and indirect reciprocity: The reproductive significance of
humor. Ethol. Sociobiol. 7 , 105-122.
Alley, T. R., and Hildebrandt, K. A. (1988). Determinants and consequences of facial aesthet-
ics. In “Social and Applied Aspects of Perceiving Faces” (T. R. Alley, ed.), pp. 101-140.
Erlbaum, Hillsdale, NJ.
Bailit. H. L., Workman, P. L., Niswander. J. D., and MacLean, C. J. (1970). Dental asymmetry
as an indicator of genetic and environmental conditions in human populations. Hum.
Biol. 42, 626-638.
Baker, R. R. (1997). Copulation, masturbation and infidelity: State-of-the-art. Proc. Int. SOC.
Hum. Ethol. (in press).
Baker, R. R., and Bellis. M. A. (1995). “Human Sperm Competition: Copulation, Masturbation
and Infidelity.” Chapman & Hall, London.
Ballinger. B. C., Buckley, D.. Naylor, G. J., and Stransfield, D. (1979). Emotional disturbance
following childbirth: Clinical findings and urinary excretion of cyclic AMP. Psychal. Med.
9(2), 293-300.
Barnett. B.. Joffe, A., Duggan, A,, Wilson. M.. and Repke, J. (1996). Depressive symptoms,
stress, and social support in pregnant and postpartum adolescents. Arch. Pediafr.Adolesc.
Med. 150, 64-69.
Barr. R. G. (1990). The early crying paradox. Hum. Nat. 1(4), 355-389.
Basden, G. T. (1966). “Among the Ibos of Nigeria.” Frank Cass & Co., London.
Beck, C. T. (1992). The lived experience of postpartum depression: A phenomenological
study. Niirs. Res. 41(3), 166-170.
Belsky. J. (1990). Parental and non-parental care and children’s socioemotional development:
A decade in review. .I. Marriage Family 52, 885-903.
Belsky, J., Steinberg. L., and Draper, P. (1991).Childhood experience. interpersonal develop-
ment and reproductive strategy: An evolutionary theory of socialization. Child Dev.
62,647-670.
Bennett, D. E., and Slade, P. (1991). Infants born at risk: Consequences for maternal postpar-
tum adjustment. Br. J. Med. Psychol. 64, 159-172.
Birkhead. T.,and Moller. A. P. (1993). Female control of paternity. Trends Ecol. Evol.
8, 100-104.
Bisping. R., Steingrueber, H.. Oltmann, M.. and Wenk. C. (1990).Adults’ tolerance of cries:
An experimental investigation of acoustic features. Child Dev. 61, 1218-1229.
Brignol, M., Rao, M., Blass, E., Marino, L.. and Glass. L. (1993).Effect of crying on energy-
metabolism in human neonates. Clin. Res. 41, A597.
Buchwald, J.. and Unterman, R. (1982).Precursors and predictors of postpartum depression:
A retrospective study. J. Prev. Psychiatry l(3). 293-308.
Bundy. D. A. P. (1989).Gender-dependent patterns of infection and disease. Parusitol. Todiry
4, 186-189.
Burger. J.. Horwitz, S., Forsyth, B., Leventhal. J.. and Leaf, P. (1993).Psychological sequelae
of medical complications during pregnancy. Pediatrics 91(3).566-571.
Buss, D. M. (1989).Sex differences in human mate preferences: Evolutionary hypotheses
tested in 37 cultures. Behav. Brain Sci. 12, 1-49.
Buss, D. (1994). “The Evolution of Desire: Strategies of Human Mating.” Basic Books,
New York.
Buss, D., and Schmidt, D. P. (1993).Sexual strategies theory: An evolutionary perspective
on human mating. Psycho/. Rev. 100,204-232.
Bustamante, J., Cuervo, J, and Moreno, J. (1992).The function of feeding chases in the chin
strap penguin, Pygoscelis antartica. Anim. Behav. 44,753-759.
Campbell, S.. and Cohn. J. (1991).Prevalence and correlates of postpartum depression in
first-time mothers. J . Abnornz. Psychol. 100, 594-599.
Cashdan, E. (1993).Attracting mates: Effects of parental investment on mate attraction.
Ethol. Sociobiol. 14, 1-24.
Close, S.(1980).“Birth Report: Extracts from Over 4000 personal Experiences. Nfer Publishing
Company, Windsor, Berks.
Collins, N.. Dunkel-Schetter. C.. Lobel, M.. and Scrimshaw. S. C. (1993).Social support in
pregnancy: Psychosocial correlates of birth outcomes and postpartum depression. J . Pres.
Soc. Psychol. 65(6), 1243-1258.
Cox, J. L. (1978).Some socio-cultural determinants of psychiatric morbidity associated with
childbearing. In “Mental Illness in Pregnancy and the Puerperium” (M. Sandler, ed.),
pp. 91-98. Oxford University Press, London.
Cox, J. L. (1979).An Ugandan puerperal psychosis. SOC. Psychiatry 14,49-52.
Cox, J. L. (1983).Postnatal depression: A comparison of African and Scottish women. Soc.
Psychint. IS,25-28.
Crowe, H., and Zeskind, P. (1992).Psychophysiological and perceptual responses to infant
cries varying in pitch: Comparison of adults with low and high scores on the Child Abuse
Potential Inventory. Child Ahiise Neglect 16, 19-29.
Cunningham, M. R.,Roberts, A. R., Barbee, A. P., Druen, P. B., and Wu, C.-H. (1995).Their
ideas of beauty are, on the whole, the same as ours: Consistency and variability in
the cross-cultural perception of female physical attractiveness. J . Pers. Soc. Psychol.
68,261-279.
Cutrona. C. ( 1982).Nonpsychotic postpartum depression: A review of recent research. Clin.
Psychol. Rev. 2(4),487-503.
Cutrona. C.. and Troutman, B. (1986). Social support, infant temperament. and parenting
self-efficacy: A mediational model of postpartum depression. Child Dev. 57, 1507-1518.
Daly, M., and Wilson, M. (1988). “Homicide.” de Gruyter. New York.
Daly, M., and Wilson, M. (1995). Discriminative parental solicitude and the relevance of
evolutionary models to the analysis of motivational systems. In “The Cognitive Neurosci-
ences” (M. S . Gazzaniga, ed.), pp. 1269-1286. MIT Press, Cambridge, MA.
Daly, M., Wilson, M., and Weghorst, S . J. (1985). Male sexual jealousy. Ethol. Sociobiol.
3, 11-27.
Davidson, J. R. (1972). Post-partum mood change in Jamaican women: A description and
discussion on its significance. Br. J . Psychiatry 121, 659-663.
Dawkins, M. S.. and Guilford. T. (1995). An exaggerated preference for simple neural network
models of signal evolution. Proc. R. Soc. London, Ser. B 261, 357-360.
Donzelli. G., Rapisardi, G., Moroni, M., Scarano, E.. Ismaelli. A,. and Bruscaglioni, P. (1995).
Neurodevelopmental prognostic significance of early cry analysis in preterm infants.
Pediutr. Res. 38, 432.
Douglas. G. (1963). Puerperal depression and excessive compliance with the mother. Br. J.
Med. Psychol. 36, 271-278.
Draper, P., and Harpending, H. (1982). Father absence and reproductive strategy: An evolu-
tionary perspective. J . Anthropol. Res. 38, 255-273.
Eagly, A. H., Ashmore, R. D., Makhijani, M. G., and Longo, L. C. (1991). What is beautiful
is good, but . . . : A meta-analysis review of research on the physical attractiveness
stereotype. Psychol. Bull. 110, 109-128.
Enlow, D. H. (1990). “Facial Growth,” 3rd ed. Harcourt Brace Jovanovich. Philadelphia.
Eyer, D. E. (1994). Mother-infant bonding: A scientific fiction. Hum. Not. 5(1), 69-94.
Farkas, L. G., and Cheung, G. (1981). Facial asymmetry in healthy North American Caucasians:
An anthropometrical study. Angle Orthodontist 51, 70-77.
Fisher, S. (1973). “The Female Orgasm: Psychology, Physiology, Fantasy.” Basic Books,
New York.
Folstad. I . , and Karter, A. J. (1992). Parasites, bright males and the immunocompetence
handicap. A m . Nut. 139, 603-622.
Fraser, F. C. (1994). Developmental instability and fluctuating asymmetry in man. In “Develop-
mental Instability: Its Origins and Evolutionary Implications” (T. A. Markow. ed.), pp.
319-334. Kluwer, Dordrecht. The Netherlands.
Freeland, W. J. (1976). Pathogens in the evolution of primate sociality. BioTropicu 8, 12-24.
Fried, M. N., and Fried. M. H. (1980). “Transitions: Four Rituals in Eight Cultures.” Norton.
New York.
Frodi, A. (1981). Contribution of infant characteristics to child abuse. A m . J . Ment. Defici.
85,341-349.
Frodi, A. (1985). When empathy fails: Aversive infant crying and child abuse. In “Infant
Crying” (B. Lester and C. Boukydis, eds.). pp. 107-123. Plenum, New York.
Frodi. A.. Lamb, M., Leavitt. L., Donovan, W.. Neff, C.. and Sherry, D. (1978). Fathers’ and
mothers’ responses to the appearance and cries of premature and normal infants. Dev.
Psychol. 14, 490-498.
Furlow. F. B. (1997). Human neonatal cry quality as an honest signal of fitness. Evol. Hum.
Behuv. 18, 175-193.
Gangestad, S. W., and Buss, D. M. (1993). Pathogen prevalence and human mate prefence.
Ethol. Sociobiol. 14, 89-96.
Gangestad. S. W.. and Thornhill, R. (1997). Human sexual selection and developmental
stability. In “Evolutionary Social Psychology” ( J . A. Simpson and D. T. Kenrick, eds.),
pp. 169-195. Erlbaum, Hillsdale, NJ.
Gangestad, S. W., Thornhill. R.. and Yeo, R. (1994). Facial attractiveness, developmental
stability and fluctuating asymmetry. Ethol. Sociobiol. 15, 73-85.
Ghubash, R., and Abou-Saleh, M. T. (1997). Postpartum psychiatric illness in Arab culture:
Prevalence and psychosocial correlates. Brit. J . Psychiat 171, 65-68.
Godfray, H. (1991). Signaling of need by offspring to their parents. Nature (London) 352,
328-330.
Goodwin. J . W., Godden. O., and Chance, G. W. (1976). “Perinatal Medicine: The Basic
Science Underlying Clinical Practice.” Williams & Wilkins, Baltimore, MD.
Gordon, R. E., and Gordon, K. K. (1967). Factors in postpartum emotional adjustment. Am.
J. Orthopsychiatry 37, 359-360.
Grafen, A. (1990). Biological signals as handicaps. J. Theor. Biol. 144, 517-546.
Crammer, K. (1993). “Signale de liebe: Die biologischen gesetze der partnerschaft. Hoffman
and Campe, Hamburg.
Grammer, K., and Thornhill, R. (1994). Human (Homo sapiens) facial attractiveness and
sexual selection: The role of symmetry and averageness. J . Comp. Psychol. 108,233-242.
Gregoire, A. J., Kumas, R., Everitt. B.. Henderson, A. F., and Studd, J . W. (1996). Transdermal
estrogen for treatment of severe postnatal depression. Lancet 347, 930-933.
Grossman, C. J. (1985). Interactions between the gonadal steroids and the immune system.
Science 227, 257-261.
Hagen, E. H. (1996). Postpartum depression: A model for depression as a strategy to defect
from coercive and exploitative relationships. Paper presented at the Human Behavior
and Evolution Society Meeting, 1996.
Haig, D. (1993). Genetic conflicts in human pregnancy. Q. Rev. Biol. 68, 495-532.
Hamida, S. B. (1996). Mate preferences: Implications for the gender difference in unipolar
depression. ASCAP Newsl. 9(5), 4-29.
Hamilton, J . A. (1988). “Rick Factors for Depression in Women and Diagnostic Subtypes.”
National Institute on Drug Abuse, Washington, DC.
Hamilton, J. A,, Harberger. P. N., and Parry, B. L. (1992). The problem of terminology. In
“Postpartum Psychiatric Illness” ( J . A. Hamilton and P. Harberger. eds.), pp. 33-40.
University of Pennsylvania, Philadelphia.
Handley, S. L., Dunn, T. L., Baker, J., Cockshott, C.. and Gould, S. (1977). Mood changes
in puerperium, and plasma tryptophan and cortisol concentrations. Br. Med. J . 78,18-20.
Harkness, S. (1987). The cultural mediation of postpartum depression. Med. Anrhropol. Q.
6(2), 194-209.
Harris, B. (1981). Maternity blues in East African clinic attenders. Arch. Gen. Psychiatry
38( 11). 1293-1295.
Hauser, M. (1986). Parent-offspring conflict: Care elicitation behavior and the “cry wolf”
syndrome. In “Primate Ontogeny, Cognition, and Social Behavior” ( J . Else and P. Less,
eds.), pp. 293-303. Cambridge University Press, Cambridge, UK.
Haywood, J. S. (1989). Sexual selection by the handicap principle. Evolurion (Lawrence,
Kans.) 43, 1387-1397.
Heitler, S., and McCrensky, K. (1976). Postpartum depression: A multidimensional study.
Diss. Absrr. Inf. 36(11B), 5792-5793.
Hershkovitz, I., Ring, B., and Kobyliansky. E. (1992). Craniofacial asymmetry in Bedouin
adults. Am. J. Hum. Biol. 4, 83-92.
Herz, E. K. (1992). Prediction, recognition, and prevention. In “Postpartum Psychiatric Illness”
( J . A. Hamilton and P. Harberger, eds.), pp. 65-77. University of Pennsylvania Press, Phil-
adelphia.
Herzog, A,, and Detre, T. (1976). Psychotic reactions associated with childbirth. Dis.Nerv.
Sys. 31, 229-235.
Hewlett, 9. S., ed. (1992). “Father-child Relations: Cultural and Biological Contexts.” de
Gruyter, New York.
Hickman, R. (1992). Husband support: A neglected aspect of postpartum psychiatric illness.
In “Postpartum Psychiatric Illness” ( J . A. Hamilton and P. Harberger, eds.), pp. 305-312.
University of Pennsylvania Press, Philadelphia.
Hill, K.. and Hurtado, A. M. (1996). “Ache Life History: The Ecology and Demography of
a Forest People.” de Gruyter, New York.
Hopkins, J., Campbell, S. B., and Marcus, M. (1987). Role of infant-related stressors in
postpartum depression. J . Abnorm. Psychol. 96, 237-241.
Hurtado. A. M., and Hill. K. R. (1992). Parental effect on offspring survivorship among Ache
and Hiwi hunter-gathers: Implications for modeling pair-bound stability. In “Father-child
Relations: Cultural and Biosocial Contexts” (B. S. Hewlett, ed.), pp. 31-56. de Gruyter,
New York.
Inwood, D. G., ed. (1985). “Recent Advances Postpartum Psychiatric Disorders.” American
Psychiatric Association Press, Washington, D.C.
Iwasa, Y., Pomiankowski, A,. and Nee, S. (1991). The evolution of costly mate preferences.
11. The “handicap” principle. Evolution (Lawrence, Kans.) 45, 1431-1442.
Jackson, L. A. (1992). “Physical Appearance and Gender: Sociobiological and Sociocultural
perspectives.” State University of New York Press, Albany.
Johnston, V. S.,and Franklin, M. (1993). Is beauty in the eye of the beholder? Ethol. Sociobiol.
14, 183-199.
Johnstone, R. A. (1994). Female preference for symmetrical males as a by-product of selection
for mate recognition. Nature (London) 372, 172-175.
Jones, D. (1995). Sexual selection, physical attractiveness and facial neotony. Curr. Anthropol.
36,723-748.
Jones, D., and Hill, K. (1993). Criteria of facial attractiveness in five populations. Hum. Nat.
4, 271 -296.
Jones. W. H. S. (1923). “Hippocrates with an English Translation.” Heinemann, London.
Kelly, J. V. (1967). The influence of native customs on obstetrics in Nigeria. Obster. Gynecol.
30, 608-612.
Karalis, K., Goodwin, G., and Majzoub, J. A. (1996). Cortisol blockade of progesterone: A
possible mechanism involved in the initiation of human labor. Nat. Med. 2, 556-560.
Kruckman, L. D. (1992). Rituals and support: An anthropological view of postpartum depres-
sion. In “Postpartum Psychiatric Illness” (J. A. Hamilton and P. Harberger, eds.). pp.
137-148. University of Pennsylvania Press, Philadelphia.
Kruckman, S., Kruckman, G.. and Kruckman, L. D. (1983). Multi-disciplinary perspectives
on postpartum depression: An anthropological critique. Soc. Sci. Med. 17, 1027-1041.
Kumar. R., and Marks, M. (1992). Infanticide and the law in England and Wales. In “Postpar-
tum Psychiatric Illness” ( J . A. Hamilton and P. Harberger, eds.), pp. 257-274. University
of Pennsylvania Press, Philadelphia.
Kumar, R., and Robson, K. (1984). A prospective study of emotional disorders in childbearing
women. Br. J. Psychiatry 144, 35-47.
Labott, S., Ahleman, S., Wolever, M., and Martin, R. (1990). The physiological and psychologi-
cal effects of the expression and inhibition of emotion. Behav. Med. 16, 182-189.
Langlois. J. H.. and Roggman. L. A. (1990). Attractive faces are only average. Psychol. Sci.
1, 115-121.
Langlois, J. H., Roggman, L. A., and Musselman, L. (1994). What is average and what is not
average about attractive faces? Psychol. Sci. 5, 214-220.
Langlois, J. H., Ritter. J. M. Casey, R. J.. and Sawin, D. B. (1995). Infant attractiveness
predicts maternal behaviors and attitudes. Dev. Psychol. 31,464-472.
Leathers, S. J., Kelley, M. A,, and Richman, J. A. (1997). Postpartum depressive symptomatol-
ogy in new mothers and fathers. J. Nerv. Ment. Dis. 185, 129-139.
Leeper. R. W. (1948). A motivational theory of emotions to replace “emotions as a disorga-
nized response.” Psychol. Rev. 55, 5-21.
Lerner, I. M. (1954) “Genetic Homeostasis.” Wiley, New York.
Lester, B. (1987). Developmental outcome prediction from acoustic cry analysis in term and
preterm infants. Pediatrics 80, 529-534.
Livingston, J. E., MacLeod, P. M., and Applegarth. D. A. (1978). Vitamin B6 status in women
with postpartum depression. Am. J . Clin. Nutr. 31(5), 886-891.
Livshits, G., and Kobyliansky, E. (1991). Fluctuating asymmetry as a possible measure of
developmental homeostasis in humans: A review. Hum. B i d . 63,441-466.
Logsdon, M.. McBride, A,, and Birkimer. J. C. (1994). Social support and postpartum depres-
sion. Res. Nurs. Health 11,449-457.
Low, B. S. (1989). Cross-cultural patterns in the training of children: An evolutionary perspec-
tive. J . Cornp. Psychol. 103, 311-319.
Ludwig. W. (1932). “Das rechts-links Problem in Tierreich und beim Menschen.” Springer.
Berlin.
Lyon, B., Eadle. J., and Hamilton, L. (1994). Parental choice selects for ornamental plumage
in American coot chicks. Noture (London) 371, 240-243.
MacDonald, K. (1985). Early experience, relative placisticity and social development. Dev.
Rev. 5, 99-121.
MacDonald, K.. ed. (1988). “Sociobiological Perspectives on Human Development.” Springer-
Verlag, New York.
MacDonald. K. (1992). Warmth as a developmental construct: An evolutionary analysis. Child
Dev. 63,753-774.
Magiakou. M. A.. Mastorakos, G., Rabin. D., Dubbert, D., Gold, P. W., and Chrousos.
G. P. (1 996). Hypothalamic corticotropin-releasing hormone suppression during the post-
partum period: Implications for the increase in psychiatric manifestations at this time.
J. Clin.Endocrinol. Matahol. 8, 1912-1917.
Manning, J. T. (1995). Fluctuatingasymmetry and body weight in men and women. Implications
for sexual selection. Ethol. Sociobiol. 16, 145-153.
Manning, J. T.. Koukourakis, K.. and Brodie, D. A. (1997). Fluctuating asymmetry. metabolic
rate and sexual selection in human males. Ethol. Sociobiol. 18, 15-21.
Manning, J. T., Scutt, D., Whitehouse, G. H., Leinster, S. J., and Walton, J. M. (1996).
Asymmetry and the menstrual cycle in women. Ethol. Sociobiol. 17, 129-143.
Martin, R., Guthrie, C., and Pitts, C. (1993). Emotional crying. depressed mood, and secretory
immunoglobulin-A. Behav. Med. 19, 111-1 14.
Mazur, A,, Halper, C., and Udry, J. R. (1994). Dominant-looking male teenagers copulate
earlier. Ethol. Sociohiol. 15, 87-94.
McCahill. T. W., Meyer, L. C.. and Fischmann. A. M. (1979). “The Aftermath of Rape.”
Heath and Co., Lexington. MA.
Michelsson, K. (1971). Cry analysis of symptomless low birth weight neonates and of asphyxi-
ated newborn infants. Acra Paediatr. Suppl. 216.
Michelsson. K.. Sirvio. P., and Wasz-Hiickert, 0. (1977). Sound spectrographic cry analysis
of infants with bacterial meningitis. Dev. Med. Child Neurol. 19,309-315.
Miller, J. E., and Huss-Ashmore. R. (1989). Do reproductive patterns affect maternal nutri-
tional status‘?An analysis of maternal depletion in lesotho. Am. J. Hum.Biol. 1,409-419.
Mitton. J. B. (1993). Enzyme heterozygosity, metabolism and developmental stability. Cenetica
89,47-65.
Mitton, J. B.. and Grant, M. C. (1984). Associations among protein heterozygosity, growth
rate and developmental homeostasis. Annu. Rev. Ecol. Syst. 15, 479-499.
Moller, A. P. (1992). Parasites differentially increase the degree of fluctuating asymmetry in
secondary sexual characters. J. Evol. Biol. 5, 691-699.
Moller. A. P. (1993a). Morphology and sexual selection in the barn swallow. Hirttndo rusticu,
in Chernobyl, Ukraine. Proc. R. Soc. London, Ser. B 252, 51-57.
Moller, A. P. (1993b). Female preference for apparently symmetrical male sexual ornaments
in the barn swallow. Hiricndo rustico. Behav. Ecol. Sociobiol. 32, 371-376.
Moller, A. P. (1995). Patterns of fluctuating asymmetry in sexual ornaments of birds from
marginal and central populations. A m . Nut. 145, 316-327.
Moller. A. P. (1997a). Developmental stability and fitness: A review. A m . Nut. 149, 916-932.
Moller. A. P. (1997b). Developmental selection against developmentally unstable offspring
and sexual selection. J . Theor. B i d . 185, 415-422.
Moller. A. P.. and Pomiankowski, A. (1993). Fluctuating asymmetry and sexual selection.
Genetico 89, 267-279.
Moller, A. P.. and Saino. N. (1994). Parasites, immunology of hosts. and host sexual selection.
J. Pnrasitol. 80, 850-858.
Moller. A. P., and Swaddle, J. P. (1997). “Asymmetry, Developmental Stability, and Evolu-
tion.” Oxford University Press, Oxford.
Moller, A. P., and Thornhill, R. (1997a). A meta-analysis of the heritability of developmental
stability. J . Evol. Biol. 10, 1-16.
Moller, A. P.. and Thornhill. R. (1997b). Developmental stability is heritable. J . Evol. B i d .
10,69-76.
Moller, A. P., Soler, M.. and Thornhill, R. (1995). Breast asymmetry, sexual selection and
human reproductive success. Ethol. Sociobiol. 16, 207-219.
Morgenshy, C. (1982). Psychological and attitudinal reaction to childbirth of recently parturient
women. Dissertation, Fordham University. Dissertation Abstracts, Ann Arbor, MI.
Morley, C., Thornton, A,, Cole, T., Fowler, M., and Hewson, P. (1991). Symptoms and signs
in infants younger than 6 months of age correlated with severity of their illnesses. Pediatrics
88,1119-1124.
Nesse. R. M. (1990). Evolutionary explanations of emotions. Hum. Not. l(3). 261-289.
Newman, J. (1985). The infant cry in primates: An evolutionary perspective. In “Infant Crying”
(B. Lester and C. Boukydis, eds.). pp. 307-323. Plenum, New York.
Nilsson. A,. and Almgren, P. (1970). Para-natal emotional adjustment: A prospective investiga-
tion of 165 women. Acto Pscyhintri. Scund., Siippl. 220, 9-61.
Nolen-Hoeksema. S. (1990). “Sex Differences in Depression.” Stanford University Press,
Stanford, CA.
Oakley. A. (1980). “Women Confined.” Schocken Books, New York.
O’Hara, M. W., Rehm, L., and Campbell, S. (1982). Predicting depressive syrnptomatology:
Cognitive-behavioral models and postpartum depression. J. Abnorm. Psychol. 91,
457-461.
O’Hara. M. W.. Rehm, L., and Campbell, S. (1983). Postpartum depression: A role for social
network and life stress variables. J. Nerv. Ment. Dis. 171(6). 336-341.
O’Hara. M. W.. Neunaber, D.. and Zekosko, E. (1984). Prospective study of postpartum
depression: Prevalence, course, and predictive factors. J . Abnorm. Psychol. 93, 158-171.
O’Hara, M. W.. Zekoski, E.. Philipps. L., and Wright, E. (1990). Controlled prospective study
of postpartum mood disorders: Comparison of childbearing and nonchildbearing women.
J . Ahtiortn. Pscyhol. 99, 3-15.
Okano, T., and Nomura, J. (1992). Endocrine study of the maternity blues. Prog. Neuropsycho-
phurmacol. B i d . Psychiatry 16, 921-932.
Paffenbarger, R. S. (1961). The picture puzzle of the post-partum psychosis. J. Chronic Dis.
13, 161-173.
Parry, B. L. (1992). Reproductive-related depressions in women: Phenomena of hormonal
kindling? In “Postpartum Psychiatric Illness’’ ( J . A. Hamilton and P. Harberger, eds.),
pp. 200-218. University of Pennsylvania Press, Philadelphia.
Parsons, P. A. (1990). Fluctuating asymmetry: An epigenetic measure of stress. Biol. Rev.
Parsons, P. A. (1992). Fluctuating asymmetry: A biological monitor of environmental and
genomic stress. Heredity 68, 361-364.
Parsons, P. A. (1996). Stress. resources, energy balances and evolutionary change. Evol. Bid.
29, 39-72.
Patterson. G. R. (1986). Maternal rejection: Determinants or product for deviant child behav-
ior? In “Relationships and Development” (W. W. Hartup and Z . Rublin, eds.), pp. 73-94.
Erlbaum, Hillsdale, NJ.
Paullekhoff, B. (1 992). Toward the diagnosis of postpartum psychotic depression. In “Postpar-
tum Psychiatric Illness” ( J . A. Hamilton and P. Harberger, eds.), pp. 239-249. University
of Pennsylvania Press. Philadelphia.
Paykel, E. S., Emms, E.. Fletcher, J., and Rassaby, E. (1980). Life events and social support
in puerperal depression. Br. J. Psychiatry 136, 339-346.
Peacock, N. (1991). An evolutionary perspective on the patterning of maternal investment
in pregnancy. Hum. Nut. 2(4), 351-385.
Peck, S.. Peck, L., and Kataja. M. (1991). Skeletal asymmetry in esthetically pleasing faces.
Angle Orthodontist 61, 43-48.
Perrett, D. I., May, K. A,, and Yoshikawa, S. (1994). Facial shape and judgements of female
attractiveness. Nature (London) 368, 239-242.
Pillsbury, B. (1978). Doing the month, confinement and convalescence of Chinese women
after childbirth. Soc. Sci. Med. 12, 11-22.
Pitt, B. (1968). Atypical depression following childbirth. Br. J . Psychiatry 114, 13251335.
Polak, M. (1997). Parasites, fluctuating asymmetry and sexual selection. In “Parasites: Effects
on Host Hormones and Behavior” (N. E. Beckage, ed.), Chapman & Hall, New York
(in press).
Pomiankowski, A,, and Mbller, A. P. (1995). A resolution of the lek paradox. Proc. R. SOC.
London, Ser. B 260,21-29.
Pop, V. J., Wijnen, H. A,, Vanmontfort, M., Essed, G., Degeus, C. A., Watson, M., and
Komproe, I. H. (1995). Blues and depression during early puerperium: Home versus
hospital deliveries. Br. J. Obstet. Gynaecol. 102(9), 701-706.
Profet, M. (1995). “Protecting your Baby to be: Preventing Birth Defects in the First Trimes-
ter.” Addison-Wiley. Reading, MA.
Quijano. E. C., and Cobliner. G. (1983). Postpartum depression in adolescents. J . Adolesc.
Health Care 4(3), 213.
Richman, J.. Raskin, V., and Gaines, C. (1991). Gender roles, social support, and postpartum
depressive symptomology. J. Nerv. Ment. Dis. 179, 139-147.
Roe, A.. and Siegelman. L. (1963). A parent-child relations questionnaire. Child Dev. 34,
355-369.
Roff, D. A. (1992). “The Evolution of Life Histories.” Chapman & Hall, London.
Sackheim, H. A. (1985). Morphologic asymmetries of the face: A review. Brain Cognition
4, 296-312.
Shackelford, T. K., and Larsen, R. J. (1997). Facial asymmetry as an indicator of psychological,
emotional and physiological distress. J. Pers. Soc. Psychol. 72, 456-466.
Shapiro, B. L. (1992). Development of human autosomal aneuploid phenotypes (with an

emphasis on Down syndrome). Acfa Zool. Fenn. 191,97-105.
Singh, D. (1993). Adaptive significance of female physical attractiveness: Role of waist-to-
hip ratio. J . Pers. SOC.Psychol. 59, 1192-1201.
Singh, D. (1995). Female health, attractiveness and desirability for relationships: Role of
breast asymmetry and waist-to-hip ratio. Erhol. Sociobiol. 16, 465-481.
Smith, R. L. (1984). Human sperm competition. I n “Sperm Competition and the Evolution
of Animal Mating Systems” (R. L. Smith, ed.). pp. 601-660. Academic Press, London.
Sosa, R., Kennell, J., Klaus, M., Robertson, S., and Urrutia, J. (1980). The effect of a supportive
companion on perinatal problems, length of labor, and mother-infant interaction. N.
Engl. J. Med. 303(1l), 597-599.
Spangenberg, J., and Pieters, H. (1991). Factors related to postpartum depression. S.Afr. J.
Psychof. 21, 159-165.
Steele. B., and Pollack, C. (1968). A psychiatric study of parents who abuse infants and small
children. I n “The Battered Child” (R. Helfer and C. Kempe, eds.), pp. 103-145. University
of Chicago Press, Chicago, IL.
Stephenson, R., Huxel, L., and Harui-Walsh, E. (1978). From wife to mother: An exploratory
study of Micronesian postpartum practices. Preliminary Report, unpublished manuscript,
University of Guam.
Surbey, M. K. (1990). Family composition, stress and human menarche. I n “The Socioendocri-
nology of Primate Reproduction” (T. E. Ziegler and F. B. Bercovitch, eds.), pp. 11-32.
Wiley-Liss, New York.
Sutton, P. R. N. (1969). Bizygomatic diameter: The thickness of the soft tissues over the
zygions. Am. J. Phys. Anthropol. 30,303-310.
Swaddle, J. P., and Cuthill. I. C. (1994a). Female zebra finches prefer symmetric males. Narure
(London) 367, 165-166.
Swaddle, J. P., and Cuthill, I. C. (1994b). Female zebra finches prefer males with symmetric
chest plumage. Proc. R. Soc. London, Ser. B 258, 267-271.
Swaddle, J . P., and Cuthill, 1. C. (1995). Asymmetry and human facial attractiveness: Symmetry
may not always be beautiful. Proc. R. Soc. London, Ser. B 261, 111-116.
Symons, D. (1979). “The Evolution of Human Sexuality.” Oxford University Press, Oxford.
Symons, D. (1995). Beauty is in the adaptations of the beholder: The evolutionary psychology
of human female sexual attractiveness. In “Sexual NaturelSexual Culture” (P. R. Abram-
son and S. D. Pinkerton, eds.), pp. 80-118. University of Chicago Press, Chicago.
Thornhill, N. W., and Thornhill, R. (1984). Book review: Incest: A biosocial view. Ethol.
Sociobiol. 5, 211-214.
Thornhill, R. (1994). Is there psychological adaptation to rape? Anal. Kritic 16, 68-85.
Thornhill, R. (1997). Rape-victim psychological pain revisited. In “Human Nature: A Critical
Reader” (L. Betzig, ed.), pp. 239-240. Oxford University Press. Oxford.
Thornhill, R., and Gangestad, S. W. (1993). Human facial beauty: Averageness, symmetry
and parasite resistance. Hum. Nat. 4, 237-269.
Thornhill, R., and Gangestad, S. W. (1994). Human fluctuating asymmetry and sexual behavior.
Psychol. Sci. 5,297-302.
Thornhill, R., and Moller, A. P. (1997). Developmental stability, disease and medicine.
Biol. Reviews 72, 497-528.
Thornhill, R., and Thornhill, N. W. (1989). The evolution of psychological pain. I n “Sociobiol-
ogy and the Social Sciences” (R. Bell, ed.), pp. 73-103. Texas Tech University Press,
Lubbock.
Thornhill, R., Gangestad. S. W., and Comer, R. (1995). Human female orgasm and mate
fluctuating asymmetry. Anim. Behav. 50, 1601-1615.
Trivers, R. (1974). Parent-offspring conflict. Am. Zool. 14, 249-264.

Unterman, R.. Posner, N., and Williams. K . (1990). Postpartum depressive disorders: Changing
trends. Birh 17, 131-136.
Upreti, N. S. (1997). A study of the family support system: Child bearing and child rearing
rituals in Kathmandu, Nepal. Unpublished doctoral dissertation, University of Wiscon-
sin. Madison.
Van Valen, L. (1962). A study of fluctuating asymmetry. Evolution (Lawrence, Kans.) 16,
125-142.
Waldrop. M. F., Pedersen, F. A.. and Bell, R. Q. (1968). Minor physical anomalies and
behavior in preschool children. Child Dev. 39,391-400.
Wasz-Hockert, O., Koivisto. M., Vuorenkoski. V., Partanen. T.. and Lind. J . (1985). Twenty-
five years of Scandinavian cry research. In “Infant Crying” (B. Lester and C. Boukydis,
eds.), pp. 187-215. Plenum. New York.
Watson. P. J., and Thornhill, R. (1994). Fluctuating asymmetry and sexual selection. Trends
Ecol. Evol. 9, 21-25.
Webster, M.. Thompson, J., Mitchell, E., and Werry. J. (1994). Postnatal depression in a
community cohort. Aitsf. N . Z. J. Psychiatry 28, 42-49.
Wedekind, C. (1994). Mate choice and maternal selection for specific parasite resistance
before, during and after fertilization. Philos. Trans. R. Soc. London, Ser. B 346,303-31 1.
Wedekind. C.. and Folstad, 1. (1994). Adaptive or non-adaptive immunosuppression by sex
hormones? Am. Nat. 143, 93-98.
Weisenfeld, A,. Malatesta, C. Z., and DeLoach, L. (1981). Differential parental responses to
familiar and unfamiliar infant distress signals. Infant Behav. Dev. 4(3), 281-290.
Weston. J. (1968). The pathology of child abuse. I n “The Battered Child” (R. Helfer and C.
Kempe, eds.). pp. 77-100. University of Chicago Press, Chicago.
Whiffen, V. E. (1992). Is postpartum depression a distinct diagnosis‘? Clin. P.Tycho1. Rev.
12, 485-508.
Wick. S. (1941). Puerperal psychoses. Wis.Med. J. 40, 299-302.
Wisner, K. L., Peindl. K., and Hanusa, B. H. (1994). Symptomatology of affective and psychotic
illnesses related t o childbearing. J. Affeciive Disord. 30(2), 77-87.
Wolman. W. L.. Chalmers, B., Hofmeyr, G . .and Nikodem. V . C. (1993). Postpartum depression
and companionship in the clinical birth environment: A randomized. controlled study.
Am. J. Ohster. Gynecol. 168(5). 1388-1393.
Wood. J. (1994). “Dynamics of Human Reproduction: Biology, Biometry, Demography.” de
Gruyter, New York.
Worthington-Roberts. B., Vermeersch, J., and Williams, S. (1985). “Nutrition in Pregnancy
and Lactation.” MirrorlMosby College Publishing, St. Louis. MO.
Yates. A . L. (1928). The role of nose and throat disease in the production of deformities of
the jaw. Dent. Rec. 48, 609-623.
Zahavi. A. (1975). Mate selection-A selection for a handicap. J. Theor. Biol. 53, 205-214.
Zahavi. A . (1977a). The cost of honesty (further remarks on the handicap principle). J . Theor.
Biol. 67, 603-605.
Zahavi, A. (l977b). Reliability in communication and the evolution of altruism. In “Evolution-
ary Ecology” (B. Stonehouse and C. Perrins, eds.), pp. 253-260. Macmillan. London.
Zakharov, V. M. (1992). Population phenogenetics: Analysis of developmental stability in
natural populations. Acia Zool. Fenn. 191, 7-30.
Zelkowitz, P., and Milet. T. H. (1995). Screening for postpartum depression in a community
sample. Con. J. Psychiatry 40(2), 80-86.
Zeskind, P. (1983). Production and spectral analysis of neonatal crying and its relation to
other biobehavioral systems in the infant at risk. In “Infants Born at Risk: Physiological,
Perceptual. and Cognitive Processes” (T. Field and A. Sostek. eds.), pp. 23-43. Grune &
Stratton, New York.
Zeskind, P.. and Lester, M. (1978). Acoustic features and auditory perceptions of cries of
newborns with prenatal and natal complications. Child Dev. 49, 580-589.
Zuk. M. (1990). Reproductive strategies and sex differences in disease susceptibility: An
evolutionary viewpoint. Pornsirol. Today 6, 231-233.
ADVANCES I N THE STUDY OF BEHAVIOR. VOL. 27
Welfare, Stress, and the Evolution of Feelings
M. BROOM
DONALD
DEPARTMENT OF CLINICAL VETERINARY MEDICINE
UNIVERSITY OF CAMBRIDGE
MADINGLEY ROAD
CB3 OES UNITED
CAMBRIDGE, KINGDOM
I. FEELINGS, ROLEA N D THEIR

THEIR EVOLUTION
A. FEELINGS,
EMOTIONS,
AND CONSCIOUSNESS
Three of the ways in which feelings can arise in an individual are as

follows. First, inputs to sensory systems may result in changes in the brain,
which we refer to as sensations or perceptions. Some of these have wide-
ranging effects within the brain in addition to information processing, stor-
age as memory, or initiation of activity modification. They lead to feelings
in the individual, for example, pain or sexual gratification (Ottoson, 1983;
Swenson and Reece, 1993). Second, various neural and hormonal changes
result in physiologically describable conditions in individuals, which we
refer to as emotions. The emotional state may involve electrical and neuro-
chemical activity in well-defined parts of the brain, hormone release, and
peripheral consequences. These various changes may also result in feelings,
such as lust or anxiety, although emotional states may exist without any
accompanying feeling, for example, as active regions of the amygdala with
no cortical activity or during sleep (Guyton, 1982). Third, even in the
absence of sensory input, or hormonal change, or activity in emotional
centers in the brain, complex or simple brain processing can lead to the
existence of feelings, for example, pleasure of achievement, guilt, or bore-
dom. Each feeling is an internal brain construct, which the individual con-
cerned may be able to describe but which will not necessarily have any
external manifestations. As Dawkins (1993, p. 142) has pointed out, feelings
may have observable signs, for example, happiness can lead to laughter,
but happiness is not laughter and cannot be defined in terms of laughter.
It is part of the state of the individual at that time. Some of the characteristics
of feelings and examples of feelings are detailed in Table I.
37 1
Copyright 0 1998 hy Academic Press
0065-3454/98 $25 00
372 DONALD M. BROOM
TABLE I
A N D EXAMPLES
CHARACTERISTICS OF FEELINGS
1. A feeling is a brain construct within an individual, sometimes with peripheral links.

2. The brain construct includes something additional to that required for other functioning.
3. A feeling is recognizable by the individual when it recurs.
4. Feelings may change behavior immediately or eventually hut need not do so.
5. Feelings often act as reinforcers when learning.
6 . Feelings can he positive or negative in that they promote approach or avoidance.
7. Examples of feelings are: pain, malaise. tiredness, hunger, thirst, thermal discomfort, fear,
anxiety. grief, frustration, guilt, depression, boredom. loneliness, general suffering, lust,
jealousy, anger, sexual pleasure. eating pleasure, exhilaration. other sensory pleasure.
achievement pleasure, general happiness.
Each kind of feeling can vary greatly in strength according to the magni-
tude and duration of the eliciting input. The mechanism for initiating the
feeling in the individual in response to an input can also vary. Indeed,
individuals will vary in sensory functioning, other physiological processing,
and analytical ability according to their genotype and environment during
life. Hence feelings will vary from one individual to another. As indicated
above, some feelings are largely elicited by low-level neural processing,
while others depend on very complex processing. Pain depends on inputs
from nociceptive pathways, usually originating in nociceptive receptor cells,
and does not require high-level processing in the brain. Similarly, thirst
is principally dependent on inputs from body fluid monitors and mouth
receptors, thermal discomfort results from local or general peripheral input,
and pleasure associated with food or sexual intercourse is principally due
to sensory input. Fear, in contrast to pain, requires high-level processing,
usually involving the comparison of sensory inputs with established models
in the brain of what constitutes a familiar or a dangerous stimulus. Likewise,
frustration is complex because it necessitates precise expectations to com-
pare with actual inputs. The most complex processes may be involved in
deriving pleasure from the solving of a difficult problem, or in some situa-
tions that lead to anxiety. Simpler, perhaps more primitive, processes lead-
ing to feelings are likely to be more widespread in the animal kingdom
than the most complex process. However, the existence of a range of levels
of complexity in the origins of different feelings does not mean that the
feelings themselves, or any behavioral consequences that they may have,
also differ in complexity. All may be equally simple. It is also likely that
most feelings, whatever the complexity of their initiation, can be amplified
by complex brain processes when much attention is devoted to the source
of the feeling, or diminished by active reduction in the brain resources
employed in that area.
WELFARE. STRESS. AND THE EVOLUTION OF FEELINGS 373
The variation in the extent of high-level processing, which is likely to be

involved in the initiation of feelings, is clearly relevant to the relationship
between having feelings and being conscious or aware. Emotion, in the
sense of neural activity in the emotion centers of the brain or specific
hormonal changes, can occur without any feelings being reported, but how
aware does an individual have to be in order to feel something such as
pain or the pleasure associated with eating a favorite food? Pain is univer-
sally referred to as a feeling, indeed the definition of pain given in a
veterinary dictionary by Blood and Studdert (1988) starts with the words
“a feeling.” The feeling is what distinguishes pain from nociception. Most
medical and veterinary usage of the words “conscious” or “aware” would
imply that an individual cannot have a feeling without being conscious
and aware. Blood and Studdert (1988) define conscious as “capable of
responding to sensory stimuli; awake; aware,” a definition that refers to
the lower threshold of what people might call conscious. The Oxford English
Dictionary defines a feeling as “pleasurable or painful consciousness, emo-
tional appreciation or sense.” Others wish to elevate the terms conscious
and aware to mean something more complex, in some cases defining them
so that they are exclusively human qualities. Griffin (1981) says that “aware-
ness involves the experiencing of interrelated mental images” and (1984)
defines conscious as “aware of what one is doing or intending to do, having
a purpose and intention in one’s actions.” It would seem likely that many
feelings involve mental images, and hence are associated with being aware
and conscious as defined by Griffin but we cannot know this for certain.
Statements by others, on the other hand, as Rodd (1990, pp. 51-54) has
pointed out, seem designed to specifically exclude feelings from conscious-
ness. Gallup (1983) refers to consciousness as the demonstrable capacity
to reflect about the self, specifically the ability to recognize oneself in
mirrors, and Humphrey (1986, pp. 93-94) states that there is no conscious-
ness in the development of a human baby until it recognizes itself in a
mirror or makes guesses about other people’s feelings. Dennett (1991) uses
the term consciousness in a much wider and vaguer way but tends (pp.
171-226) ot equate the complexity of the individual’s world with the exis-
tence of and necessity for consciousness. The definition of awareness that
will be used here is: “awareness is a state in which complex brain analysis
is used to process sensory stimuli or constructs based on memory.” It is
clear that there is a range of degrees of awareness (Sommerville and Broom,
1998), the highest of which might require a greater extent of intellectual
processing than is needed for most feelings; but it is also clear that, according
to the commoner meanings of the concept, feelings involve awareness. The
term consciousness, however, is so widely used in the Blood and Studdert
374 DONALD M. BROOM
sense (discussed earlier) that it would be simplest to limit its meaning to

this and use awareness when referring to different degrees.
When the word “feelings” is used, some authors qualify it with “subjec-
tive.” This usage seems redundant, as every feeling is necessarily limited
to one individual or subject. The word subjective can also be confusing
because it is sometimes used to refer to a conclusion that is based not on
the observations, but on self-examination by the observer.
The various feelings are discussed later with reference to their possible
evolutionary origins, but it is useful at this stage to consider the widely
used categorization of feelings into pleasant and unpleasant. Unpleasant
feelings are those that the individual concerned would avoid if possible,
whereas efforts are generally made to experience pleasant feelings. Single
unpleasant feelings or combinations of unpleasant feelings may sometimes
be referred to as suffering. Dawkins (1990) stated that “suffering occurs
when unpleasant subjective feelings are acute or continue for a long time
because the animal is unable to carry out the actions that would normally
reduce risks to life and reproduction in those circumstances.” However, as
Broom and Johnson (1993, p. 81) have pointed out, most people would
include all but the milder, briefer kinds of pain and malaise within the term
suffering and such problems do not necessarily involve inability to reduce
risks to life and reproduction. Hence, a better definition might be “suffering
is an unpleasant feeling, which is prolonged or severe.” An individual
in distress may well have unpleasant feelings, but could also be showing
physiological consequences of adverse conditions without having such feel-
ings. Hence “distress” is not listed as a feeling.
B. EVIDENCE
FOR FEELINGS
The existence of a feeling in an individual may result in a change in its

physiology or behavior that can be recognized by other individuals, includ-
ing a human scientist. If the change were unique to individuals with such
a feeling, the possibility would arise for the existence of the feeling to be
recognized whenever the change occurred. For example, persons who feel
grief exhibit particular facial expressions and may produce tears. Individual
actions are often not unique to particular feelings, but combinations of
actions can be good indicators. However, we know that such expressions
and tears can be faked, and although the observer might be quite efficient
in ability to identify the signs of grief and to discriminate between fake
and real signs, the achievable precision in the identification of grief will be
no more than a high probability. A person can also use words to express
aspects of his or her grief, but words are more easily faked than the behavior
changes, and so are generally less reliable as indicators of feelings. Despite
WELFARE. STRESS. A N D THE EVOLUTION OF FEELINGS 375
the possibilities of deliberate deceit or other simulation, words and behavior

observation can often provide evidence for reasonably confident recogni-
tion of feelings in other people. For most observers, the recognition of grief
is initiated after personal experience of such a feeling and facilitated by
previous experience of what was assumed, from observing context and
consequences, to be grief in others.
Examples of other feelings that can substantially affect behavior are pain
and guilt and those feelings easily related to particular needs. Some needs
are associated with feelings and these feelings are likely to change when
the need is satisfied (Broom, 1996). Examples of such feelings include
hunger, eating pleasure, lust, and sexual pleasure. As needs are part of
motivational state and motivational state alters the likelihood of occurrence
of behaviors, the existence of needs can be deduced from the frequency
and pattern of behavior (Fraser and Broom, 1990, pp. 31-38, 263-264).
Hence, reasonable predictions about various feelings can be obtained from
observations of behavior.
The major problem with the recognition of feelings from observations
of behavior is that feelings may often exist without any behavioral or
physiological change to indicate them. Abdominal pain can lead to particu-
lar postures and movements and leg pain can lead to limping (for other
examples, see Fraser and Broom, pp. 296-304), but severe pain can exist
without any detectable sign. In some cases an observer or experimenter
can contrive situations so as to maximize the likelihood that a measurable
behavioral or physiological change will indicate the existence and extent
of a feeling. For example, suspected localized pain may be identified by
palpation of the area and measurement of behavioral, heart rate, and adre-
nal cortex responses. Experimental studies of animal preferences may also
be carried out in order to obtain some information about a feeling such as
hunger, frustration, or an aspect of sensory pleasure. The existence of a
strong preference for some resource or possibility to carry out a behavior
gives some indirect information about what an animal is likely to be feeling,
but may not discriminate between working hard because of the existence
of a negative feeling and working hard in order to obtain a positive feeling.
Some feelings are not easy to investigate experimentally in this way, for
example, malaise or boredom, because, for different reasons, the feeling
may affect the likelihood of carrying out the preference test.
Some information that can help in the recognition of feelings can be
obtained from measurement of physiological changes including brain state.
Heart rate and adrenal cortex activity changes, as already mentioned, and
measurements of hormonal or neural activity changes may coincide with
or precede changes in feelings. Brain scanning techniques can indicate sites
and pathways in the brain where activity is occurring. Such activity in the
376 DONALD M. BROOM
brain may be related to reports or behavior changes that give other evidence
of the existence of feelings and hence may themselves come to be used as
evidence for particular feelings.
The general conclusion about evidence for the existence and extent of
feelings is that, even with sophisticated techniques, it is not possible to
know exactly the feelings of any other individual, whatever the species, but
reasonable predictions may be made using evidence, most reliably that
from carefully studied behavior. The argument that no feelings can be
recognized in others unless the individual can describe the feeling in words
is wrong.
C. AREFEELINGS OR JUST EPIPHENOMENA?

FUNCTIONAL
I. Feelings in General
It is thought by some scientists that all feelings are merely trivial by-
products of processes within the body. Skinner (1974, p. 17) said “what is felt
or introspectively observed is not some non-physical world of consciousness,
mind or mental life but the observer’s own body. This does not mean . . .
that what are felt or introspectively observed are the causes of behavior.”
Further, he said (1978, p. 124), “One feels various states and processes
within one’s body, but these are collateral products of one’s genetic and
personal history. No creative or initiating function is to be assigned to
them.” This view of feelings as solely an accident of individual development
with no function or relevance to any other individuals is certainly not held
by many people. As Dawkins (1993, p. 5 ) points out, the actions of people
are much affected by a belief that these might cause pain, happiness, or
sorrow in others.
Once it is accepted, as it is by most people, especially pet owners, that
feelings exist in other individuals of our own and other species, the idea
that they must have some function often follows. Y.-K. Ng (personal com-
munication) says that consciousness is a major mechanism in individuals and
hence it “must contribute to fitness to survive natural selection.” Although
adaptive characteristics will survive in populations because selection will
favor their survival, if feelings were just an accidental, nonadaptive conse-
quence of other adaptive mechanisms, would they persist in a population?
Provided that a nonadaptive characteristic has some genetic component
(as almost everything has) and some cost, it is likely that it would disappear
from the population. Because various feelings have persisted, they are
probably adaptive. The nature of any advantage to animals of having feel-
ings is explored here in some detail, but first it is necessary to consider
possible origins of feelings and the possibility that they are still accidental
and nonfunctional.
WELFARE, STRESS. A N D THE EVOLUTION OF FEELINGS 377
An assumption outlined in Section I,A is that feelings involve brain

activity additional to the minimum required for information processing,
storage as memory, and motor output. During the early stages of develop-
ment of systems such as those for recognizing and responding to predators
or for recognizing and ingesting food, it seems very likely that there would
have been some accidental activation of parts of the brain that were not
essential for the neural function. Also, as the system became more efficient,
it is likely that pathways that were at first necessary, perhaps to ensure
effective communications between receptor and effector, became redun-
dant but were not immediately eliminated from the functioning system.
Both of these are examples of nonfunctional epiphenomena of the system,
which might have the kinds of effects in the brain that became feelings.
Some of these epiphenomena might continue to be nonfunctional but inevi-
table side effects of an essential system. Others might have had effects that
eventually became functional, as discussed later in relation to the various
feelings. It is also possible that there are feelings that were once functional
in the ancestors of present-day animals but that now have little or no
function. Finally, it is important to consider the extent to which feelings
might sometimes be harmful (Broom and Johnson, 1993, p. 80), perhaps
by making it more difficult for individuals to show the most appropriate re-
sponses.
The idea that pain has the function of preventing body damage has been
espoused on many occasions, but the more general concept of all feelings
being functional is relatively recent. Dawkins (1977) stated that “It is rea-
sonable to assume that subjective feelings (like other characteristics)
evolved because animals which possessed them were fitter than those which
did not” and “feelings must be a product of natural selection. They are
part of biology.” In a much more extensive exposition, Cabanac (1979) said,
We experience feelings of hunger because that is part of our mechanism for rectifying
a food deficit and getting something to eat. We experience fear and pain because they
are part of our body’s way of removing us from situations that are life-threatening.
Conscious experiences are there as survival aids.
A similar argument was presented by Wiepkema (1985) who asserted that

feelings are involved in monitoring the effectiveness of regulatory actions,
being positive when the regulation is successful and negative when it is
not. A further statement that the evolutionary advantage of having feelings
is considerable was made by Dawkins (1990), and Broom and Johnson
(1993, p. 334) said,
A final point about the evolution of adaptation to the vicissitudes of the physical and
social environment is that a very important part of that evolution has been the develop-
378 DONALD M. BROOM
ment of the complex appreciation of the interactions of an individual with the world in
which it lives, which we call feelings. Complex brains, like those of vertebrates, have
complex systems for regulating those interactions which are not just the product of
automatic responses to stimuli. If an individual has a system of feelings which involves
changes in its mental, and perhaps in its hormonal, functioning because a certain kind
of body regulation or because an anticipated event has not occurred, such an individual
will have increased fitness in comparison with a genetically different individual which
has no such system.
Similarly, Y.-K. Ng (personal communication) argues that awareness con-

tributes to fitness but is limited to species in which there is plasticity in
brain and behavior, and that “affective feelings must have evolved fairly
quickly after the evolution of awareness, if not concomitantly.”
In order for feelings to confer an adaptive advantage it is essential that
they should have an effect (Dawkins, 1993, p. 169); the exploration of such
effects and how they might be recognized is a major part of this paper.
The effect of the feelings might be “that the individual is more likely to
carry out some adaptive action and hence more likely to survive” (Broom,
1996). The most likely way in which this would occur is that the feeling
acts as a reinforcer, which makes it more likely that the individual will
learn to carry out the adaptive action. Indeed “if the state of the individual
in certain conditions is desirable from an evolutionary viewpoint, there
should be a propensity for that individual to have good feelings. On the
other hand, if a state is one which should be quickly altered, it should be
associated with unpleasant feelings which prompt avoidance or some other
action” (Broom, 1996). If a feeling does have an effect such that it can act
as a reinforcer to promote adaptive behavior, the effect might be coincident
with the occurrence of the feeling, or it could be that the feeling is remem-
bered so that its beneficial effect occurs long after the feeling itself has
finished. Such effects would be difficult to distinguish from the effects of
other events in the life of the individual and hence to attribute to the feeling.
The general argument that consciousness in general and feelings in partic-
ular make a difference in the way in which organisms possessing it function
is presented with some force by Dawkins (1993, pp. 9, 143,167-181). One
straightforward but strong argument is that most people report that many
of the feelings listed in Table I can have a considerable effect on the way
in which they organize their lives. Observation of human behavior supports
this. Proposals about the effects and functions of consciousness have sug-
gested that it allows more effective analysis of the environmental and
prediction of the future (Crook, 1980, p. 31) or that it improves the efficiency
of information processing in general (Weiskrantz, 1987). Dawkins (1993)
suggests that “Consciousness might . . . make actions more decisive or
give better anticipation of the future” and “have an evolutionary effect
WELFARE. STRESS, AND THE EVOLUTION OF FEELINGS 379
and an effect which we could detect.” Dawkins does not claim that every
action will be more efficient if the individual carries it out consciously, and
refers to the observation by Baars (1988) that certain actions progress more
efficiently if the individual is not actively thinking about what is being done.
However, this argument about the function of consciousness in the sense
used by those authors is more relevant to the control of actions than to
areas where feeling plays an important role.
2. Pain
Although the word pain is used colloquially to refer to a wide range of
unpleasant experiences, its scientific and medical meaning is limited to refer
to a sensation, that is, to the immediate consequences of a particular sensory
input. The sensation elicits immediate avoidance or subsequent modifica-
tion of behavior whose effect is to reduce the likelihood of recurrence of
the sensation. Hence, a definition of pain is “A sensation which, without
involving higher level brain processing, such as that associated with fear,
is very aversive” (Broom and Johnson, 1993, p. 27). Pain usually involves
specialized nociceptive receptor cells and some degree of injury. Even in
the case of phantom limb pain, the specialized nociceptive neural pathways
are involved. Pain normally elicits protective reactions, causes emotional
responses, and results in learned avoidance behavior.
The pain system includes: specialized receptor cells, nerves in which
evoked electrical responses to mechanical or thermal damage can be de-
tected, neural pathways involving characteristic transmitters such as sub-
stance P, brain mechanisms, which include endogenous analgesic opioids,
and the propensity to initiate avoidance behavior. This system is present
in all vertebrates that have been studied, including fish, and most aspects
of it are also present in some invertebrates, for example, cephalopod mol-
luscs. As mentioned already, all pain is regarded as being a feeling and if
there is activity in the nociceptive system with no feeling, perhaps because
of naturally occurring opioid-induced inhibition or the use of an externally
applied analgesic, then it is not pain. We cannot know whether pain in
another individual is the same as that which we feel ourselves, but observa-
tion of behavior in all species with complex nociceptive systems suggests
that there is likely to be much similarity among the different kinds of
animals in the feeling of pain.
The importance of feeling pain in promoting individual survival is consid-
erable (Melzack, 1973). When pain is felt, the individual can take action
to minimize tissue damage being caused, and the greater the feeling of
pain, the faster the initial action is likely to be. Once pain has been felt in
a recognizable situation, the possibility of learning to avoid any future pain,
and hence damage, of the same kind is increased. Again, more intense
380 DONALD M. BROOM
pain is likely to have a greater effect. After physical trauma, or during a

pathological condition, recovery will often be facilitated by modification
of behavior so as to avoid further damage to the affected parts of the body
(Wall, 1979). Chronic pain can therefore be functional in that it increases
the chances that activity level and type will be modified in an adaptive way.
Hence, as explained by Broom and Johnson (1993, p. 29), to suppress pain
would in many situations be disastrous.
The existence of a feeling of pain in an individual may not be obvious
and it seems very likely that in various species extreme pain can occur
without any recognizable modification of behavior (Morton and Griffiths,
1985). Indeed, as pointed out by Fraser and Broom (1990 pp. 269-273),
pain is not necessary unless some action has to be taken, and animals of
different species will vary in the kind of behavioral response to pain that
is most adaptive. For example, vocalization when in pain may be advanta-
geous for a young animal or a social animal that might be helped by its
mother or by members of its own social group, but disadvantageous when
a dangerous predator is close and no effective help from any other individual
is likely. Hence, it is not surprising that young pigs, dogs, and humans,
which could be helped, make a lot of noise when in pain, but young sheep,
which are less likely to be helped against most predators, do not. Indeed,
the sheep response to tissue damage is considerable in terms of physiological
change (Shutt et al., 1987) and subsequent avoidance of the situation where
the damage occurred (Rushen, 1990), so it is extremely likely that they
are feeling pain, but they show little behavioral response at the time of
the damage.
Although it is often possible to hypothesize that in many situations pain
has a function that would result in natural selection favoring genotypes in
which the individuals were able to feel pain, in other situations it is not
easy to ascribe any function to the pain. Bateson (1991) refers to the
extreme pain associated with a kidney stone stuck in a ureter and other
forms of extreme pain that result in the individual writhing in agony. The
writhing could conceivably have a function, but this seems unlikely when
recovery from surgery seems to be facilitated by analgesia. Perhaps the
various gradations of the lower levels of pain are adaptive but some extreme
pain is an inevitable but nonfunctional consequence of the system. The
possibility that extreme pain interferes with various aspects of normal
functioning, and hence potentially impairs the efficacy of adaptive re-
sponses, is often quoted as a reason why endogenous opioid analgesic
mechanisms evolved. An individual might be in such pain that it could not
show an escape response unless the analgesic opioid inhibited or amelio-
rated the pain.
WELFARE. STRESS. A N D THE EVOLUTION OF FEELINGS 381
Another critique of the idea that pain is adaptive is that it should not
be necessary for there to be such an elaborate within-individual communica-
tion system when all of the cells of the body have the same genotype and
hence do not need any more than a very simple message. However, there
are other elaborate communications systems within the body and a fast,
effective system of messages about actual or potential body damage is im-
portant.
The conclusion of these arguments about the function of pain is that
many forms of pain are important for survival but some are likely to
be accidental and nonfunctional. Functional pain seems to occur in all
vertebrates that have been studied and in some invertebrates. Nonfunc-
tional pain probably also occurs in all animals. There is no reason why
there should be any differences between humans and other vertebrates in
the proportions of functional and nonfunctional pain.
3. Malaise
Malaise is a feeling of illness or discomfort associated with some
pathology or inadequacy of bodily function. It is more general in its
effects than pain, which is localized in a particular part of the body.
There could be wide-ranging effects on the body when there is a reduction
in the availability of energy providing resources because these are being
used to fight pathogens. Similarly, an accumulation of toxins could have
consequences in various parts of the body. In both of these examples,
the effects on the brain, perhaps mediated via the peripheral nervous
system, could lead to the feeling of malaise. The exact details of the
feeling are likely to vary according to the kinds of effects of the toxin
or pathogen, but a general feeling of lethargy is common to malaise
with various causes.
Most people think of malaise, or of feeling ill, as an unfortunate by-
product of infection, but it often affects behavior as well as physiology and
its major effect may well be adaptive. Although some effects of parasites
or pathogens on host behavior are induced by the parasite or pathogen for
its own benefit, many of such effects help the host. As Hart (1988, 1990)
points out, animals that are sick are often depressed, lethargic, show no
interest in eating, and fail to exhibit body care, but such behaviors “appear
to comprise an adaptive behavioral mode that facilitates recovery from
illness.” When the immunological and other defenses of the body are having
to work hard and consume a lot of energy, it is advantageous for the
individual to rest and to be able to concentrate available energy on fighting
the cause of the malaise. Even high fever would be adaptive if the net
benefit from killing pathogens and suppressing activity outweighed the net
cost of tissue damage (Kluger, 1979; Ewald, 1980, 1983; Hart, 1988). The
382 DONALD M. BROOM
general feeling of malaise can be thought of as part of the means by

which active defense against pathogens, with its multitude of consequences
including increased levels of interleukin 1, can act so that behavior is
adaptively modified.
There may also be specific aspects of malaise that are adaptive. For
example, disorders of the gut may be remedied fastest if no further food
is taken in; therefore, a feeling of nausea is advantageous in this circum-
stance. Too much consumption of a poison, such as alcohol, may overload
liver resources and hence also lead to nausea and reduction in further
intake of food or poison. Infection or poison-induced nausea is likely to
be adaptive, but some nausea, like that induced by certain forms of motion,
might be accidental. The nausea, vomiting, and food aversion that occurs
in early human pregnancy could be a mechanism for minimizing the risk
that toxins from various normally edible foods might adversely affect the
developing fetus (Profet, 1992). These feelings are also likely to suppress
activity in the mother and could be a consequence of a risk of mechanical
damage to the fetus.
4. Tiredness
Muscular fatigue is a sometimes unwanted consequence of muscular
activity, but it may have neural effects that are felt as tiredness. Other
forms of exertion and prolonged periods of being awake rather than asleep
can also lead to a feeling of tiredness. As sleep tends to occur with a fairly
regular rhythm, tiredness could sometimes indicate that the normal time
for sleeping has arrived rather than that the individual has been involved
in much exertion. Tiredness as an indicator of high levels of exertion can
be an adaptive feeling in that it tends to prevent levels of exertion that
might be damaging. Tiredness as a prompter of sleeping can be adaptive,
just as sleep can, in that it ensures that the individual is inactive and
inconspicuous at a time of day when accidents are more likely and predators
abound. There may also be some recuperative function.
5. Hunger
The feeling of hunger is generally associated with emptiness at some
level in the gut, or low levels of nutrients within some organ of the body,
or with input from an internal clock that indicates that the time of day
when some food intake normally occurs has arrived (Grossman, 1973). As
with other feelings, hunger varies from slight to very severe. The stronger
the feeling, the more likely it is that the motivational state will be greatly
influenced by causal factors relating to hunger. Hunger is generally rather
unspecific, but there can be specific hungers that are satisfied only by the
consumption of a particular nutrient. Although individuals can overeat and
WELFARE, STRESS, AND THE EVOLUTION OF FEELINGS 383
can select the wrong nutrients, in most circumstances the feeling of hunger
would seem to be adaptive.
6. Thirst
The sensory inputs that lead to feeling thirst come from body fluid
concentration monitors, including those in the mouth (Toates, 1986). Thirst
varies from a minor feeling, which has little effect on behavior, to an
all-pervading feeling, which dominates all behavior in that the individual
devotes all possible resources to finding water while conserving it. In most
circumstances, the feeling of thirst would seem to be adaptive, although it
is possible that some very thirsty individuals behave in a rash way in order
to obtain water.
7. Thermal Discomfort
The physiological responses to, and consequences of, exposure to very
high or very low temperatures have been described in detail (Milner, 1970).
The feeling of discomfort when too hot or too cold results in changes
in behavior of various kinds (Broom, 1981, pp. 108-113). Most of the
consequences of the feelings that have been described have the effect of
increasing survival chances.
8. Fear
Fear is a feeling that occurs when there is perceived to be actual danger
or a high risk of danger. Although a fast looming visual stimulus, a sudden
sound, or an acrid smell might elicit a startle response, the feeling of fear
depends on some more complex analysis in which current sensory input is
compared with memories of previously experienced events. Blood and
Studdert (1988) define fear as “a normal emotional response to consciously
recognized external sources of danger.” Hence, the recognition precedes
the feeling rather than being a part of it.
There are two very different kinds of response to fear (Broom, 1981, pp.
162-175). One is to actively escape or defend, while the other is to rapidly
and substantially reduce behavioral and physiological activity so as to render
the individual inconspicuous to an attacker. The propensity of predators
to notice fleeing prey and ignore immobile objects is utilized in this response.
The physiological changes associated with feeling fear are also of two kinds
(Broom and Johnson, 1993, pp. 92-107). If active responses are a possibility,
first the adrenal medulla response and then the adrenal cortex response
prepare the individual for precipitant activity. If suppression of movement
is important, then bradycardia, which tends to suppress movement, occurs.
In many cases, however, either the active or the passive kind of response
could be shown and animals may show first one and then the other. For
384 DONALD M. BROOM
example, when a young domestic chick is startled, it usually freezes for a

while and then shows escape activity, the proportions of each response
depending on the conditions and the previous experience of the bird
(Broom, 1969a,b).
The freezing response can be very effective as a means to survive predator
attack, but it would be extremely inappropriate as a response to the immedi-
ate risk of a large quantity of rocks falling onto the individual. Hence, the
efficiency of analysis of the situation and appropriate selection of response
is of great importance. If a predator was detected that would not be deceived
by a freezing response but that could be outrun, then it would again be
important to choose the response correctly.
The response to feeling fear could also be maladaptive if carried to an
extreme. For example, an extreme form of the freezing response is catalepsy.
This can be an adaptive response to the presence of a predator but if it
leads to hypovolemic shock, the individual can be damaged by the response
itself. Shock reactions are on the borderline between the life saving and
the clearly disadvantageous, being the former when predator attack is a
major factor in the life of the individual, but being the latter if predation
risk is insignificant.
9. Anxiety
Anxiety is “a feeling of uneasiness, apprehension or dread” (Blood and
Studdert, 1988). It depends on an ability t o predict future risk, usually
based on recent stimuli and always on previous experience. The horse that
is reluctant to pass a gateway where it once had a frightening experience
is showing the characteristic signs of an individual feeling anxiety. Whenever
information about disturbing events is stored, there must be a potential
that there will be recall of this information and consequences of the recall
that activate emotional systems in the individual. The recall can occur at
an entirely cerebral level, that is, with no current stimulus involved, but
with consequences of the feeling that may be physiological as well as
conceptual. The feelings of anxiety probably potentiate the response to the
risk. In some cases, showing an appropriately high level of response to a
risk can be life saving, while in other cases anxiety can be damaging to the
individual (Nesse and Williams, 1995). A genotype that promoted the ability
to explore previous experiences in the brain and to feel anxiety when high
levels of risk were deduced would be strongly favored in natural selection,
given the very considerable advantages of anxiety in some circumstances.
The disadvantages of unnecessary anxiety would not be sufficient to coun-
terbalance the advantages of having the feeling. It could be that anxiety
was much more important to our ancestors than to present-day humans so
that many individuals, and some in particular, feel much anxiety that confers
WELFARE. STRESS. AND THE EVOLUTION OF FEELINGS 385
no benefit to them. Indeed, the propensity to feel much anxiety might be

expected to be greater in females because of the difficulties of maternal
care when offspring are young and very vulnerable. The anxiety is of such
importance at that time that its selective advantage then outweighs any
disadvantages associated with its persistence throughout life during times
when it is unnecessary or harmful to the individual. Anxiety may indeed
be greater at times of parental responsiblity but may be difficult to switch
off at other phases in life.
10. Grief
Grief and sorrow are unpleasant feelings associated with unwanted life
events, particularly those connected with social relationships. Grief is a
stronger version of sorrow. For the most part they are private feelings with
no effect on others, although their outward manifestations may have an
effect of maintaining social position, preserving good relationships, or in-
forming others, honestly or dishonestly, of the properness of the individual’s
feelings. When a significant unpleasant event occurs during life, for example,
an important failure to succeed in some endeavor, or the loss of a social
relationship, increased brain processing might occur. However, that increase
in brain processing and activation of emotional centers that result in grief
could then increase even more by positive feedback in a way that may help
the individual to respond adequately to the event, both immediately and
in terms of later learning. Hence, the grief could be functional in that, by
amplifying the effects of the event, it ensures that something of importance
is not processed too rapidly and then stored without sufficient likelihood
of affecting future decision making.
Grief is only one example of a feeling as an amplifier of the significance
of important life events. Simple brain processing of perceived events may
not allow sufficient allocation of attentional and analytical resources to
those events that demand more consideration if the individual is to survive
and reproduce effectively. However, mechanisms whose general effect is
amplification can operate too strongly in a particular instance or may fre-
quently exaggerate the significance of real situations in certain individuals.
Grief, like other feelings, seems to have adaptive advantages on some
occasions but not on others. The most obvious situation in which grief is
felt by humans is when a relative or close friend departs or dies, a circum-
stance in which the amplifying effect of grief helps in the evaluation of the
importance of relationships. There are many reports of pets, especially
dogs, and monkeys (de Waal, 1996, pp. 54-56) showing the same sort of
behavior that humans show in such circumstances. This behavior is some-
times described as mourning and has also been described in horses, pigs,
and elephants. As the arguments for the selective advantages of grief would
386 DONALD M. BROOM
apply to other species, especially those that have a long-lasting and elaborate
social structure, it would seem likely that, as de Waal proposes, some degree
of grief occurs in other species.
11. Frustration
The existence of elaborate motivational mechanisms in a wide variety of
animals has already been mentioned. Whenever a well-organized decision-
making process exists, the individuals will sometimes be unable to do what
they most want to do. Broom (1985) said that an animal is frustrated “if
the levels of most of the causal factors which promote a behavior are high
enough for the occurrence of the behavior to be very likely but, because
of the absence of a key stimulus or the presence of some physical or
social barrier, the behavior cannot occur.” The feeling of frustration could
originally have arisen as an accident because of some alternative channeling
of output from the decision-making system. This feeling could be damaging
to the individual in that it might tend to make it carry out behavior such
as self-mutilation or to activate physiological systems that either use up
energy or promote pathological effects. Alternatively, the feeling of frustra-
tion could lead to behavioral and physiological changes that help the individ-
ual to cope with the frustrating situation. Some behavior changes resulting
from frustration, such as stereotypies, might be adaptive in some situations
but harmful in others (Mason, 1991; Broom and Johnson, 1993, pp. 139-
141).
Frustration is likely to be an important feeling in most complex animals.
Because there will have been strong selection for systems that allocate
resources effectively (Broom, 1981, pp. 79-96), as Houston (1997) points
out, animals will be strongly motivated to perform certain important activi-
ties, and suffering is likely if they are prevented from performing them.
The feeling of frustration, which at a high level might be referred to as
suffering, is therefore likely to occur in a wide range of animals.
12. Guilt
The feeling of individuals who have behaved in a way that they or their
social group members would normally condemn or punish is referred to
as guilt. People report that they feel guilt in a variety of situations. It is
clear that the brain processing underlying the feeling must often involve
sophisticated analysis of actions in relation to experience of the conse-
quences of such actions for the individual carrying them out and for others.
This brain activity could easily have been the origin of certain wide-ranging
effects that are now grouped together as the feeling of guilt.
The consequences of feeling guilt are often changes in behavior described
in some detail for dogs and primates by de Waal (1996). One of the human
responses is blushing, dogs may hang their heads, and monkeys may behave
in an unusually submissive way and show “grin” expressions on their faces.
The feeling of guilt could be advantageous to the individual in that it forces
attention channels and processing capacity to be allocated to consideration
of a situation of importance in relation to future decision making. However,
behavior generally associated with guilt is shown much more often in social
situations than by individuals that are by themselves. The information
conveyed to others in the social group could be of value to them and might
therefore increase the spread of a gene promoting the feeling and expression
of guilt because of close relatedness or the potential for reciprocal altruism.
However, the guilt expressor might also derive a direct social advantage
in that he or she would be perceived as continuing to be an honest and
constructive member of the group. Perhaps without the expressions of guilt
he or she would be expelled from the group. Another way of putting this
is that the blushing, o r other behavioral manifestation of guilt, is an honest
signal. Blushing can also be a contrived signal with the intent of making
the blushing individual more sexually attractive to the observer.
13. Depression
Depression is a clinical condition associated with certain neurological
disorders and with extreme malaise, fear, anxiety, grief, or frustration. It
can be described as a feeling separate from all of these, as it is reported
by people as being different. Depressed people describe how nothing mat-
ters to them and most of their normal activities are reduced or absent.
These behavioral signs can exist in individuals of other species, at least in
mammals and birds. An experimentally induced parallel is the “learned
helplessness” described in laboratory animals by Seligman (1975).
Some of the possible advantages of feeling depressed have already been
described as extreme forms of fear, grief, and so on. In social situations in
which the individual is performing very badly, behavior and physiology
indicative of depression may occur. Extreme examples are the tree shrews
and rodents that have been defeated in fights with rival conspecifics but
that remain caged with them (Koolhaas et al., 1983; von Holst, 1986). These
individuals show very depressed behavior and often die quickly. However,
passive responses in such a situation, and perhaps the feeling of depression
itself, may be an effective strategy for avoiding future attack (Mend1 et al.,
1992). Even the tree shrews might benefit from such behavior if they had
the opportunity to escape completely from the rival. Similarly, some human
depression could be specific to perceived failure in a specific situation, such
as inability to compete adequately in a particular segment of society, and
could reduce the chances of attack by dominant individuals (Nesse and
Williams, 1995). Depression, linked with the possibility to move to a differ-
388 DONALD M. BROOM
ent segment of society in which coping is easier, could be an effective

response. The depressed period could also allow time to work out a better
life strategy. However, it would appear that in some cases, feeling depressed
is maladaptive, perhaps because there is no viable alternative lifestyle.
14. Boredom
Boredom is a feeling associated with a lack of novel input, perhaps with
a lack of input in total. Its occurrence in various animals has been discussed
by Wemelsfelder (1993). She refers to animals in impoverished conditions
increasingly directing their behavior toward inadequate stimuli, exaggerat-
ing normal behavior, and establishing stereotypies. A threshold in the
increasing series, which she defines as boredom, is “that state of behavioral
fixation in which the animal’s orientation towards a novel stimulus loses
its inquisitive and manipulative character.”
The feeling of boredom may arise because of severe sensory deprivation
or because those potentially interesting stimuli that are detectable are
repeated exactly. In the first case, the individual might be in a plain, empty
cage; in the second case, there might be a machine present that continually
undergoes the same movements with the same periodicity. In either case
a sequence of behavior whose function is to carry out useful exploration
is continued until lack of useful consequences tend to result in its inhibition.
At the same time, or even in the absence of the attempts at exploration,
a feeling of boredom arises and increases, sometimes associated with the
abnormalities of behavior described here.
It may be that brain function can be impaired when there is too little
input, or too little new input, for a prolonged period. Studies of sensory
deprivation suggest that this is so. It may be that the existence of the
feeling of boredom, with its concomitant behavior abnormalities, provides
sufficient neural activity to maintain the brain in a low input situation.
Boredom that continues to the point where inadequate input results in
greater and greater risk of neural system damage may have stronger and
stronger effects on the individual, which, although more and more unpleas-
ant, are more likely to prevent the damage.
15. Loneliness
In social animals, social contact is often important for survival of the
individual as well as for reproduction. Hence, it is not surprising that a
specific feeling can occur when inadequate social contact occurs. In our
human ancestors, predator avoidance, food finding, and control of the
physical environment were all very much facilitated by, or perhaps possible
only in, a group-living situation. Hence, an individual separated from all
WELFARE, STRESS, A N D THE EVOLUTION OF FEELINGS 389
conspecifics might feel lonely and make all efforts to restore adequate
social contact.
16. General Suffering
All of the negative feelings mentioned so far could constitute suffering
if extreme enough. Combinations of feelings might also be referred to as
suffering. Individuals with a range of negative feelings might be able to
respond by moving to a new living place, so suffering could lead to adaptive
responses. In most cases, however, the most extreme forms of suffering
will not have beneficial effects. The various forms of suffering may well be
adaptive when moderate but neutral or harmful in relation to fitness when
severe. The form of suffering that occurs when there is real or perceived
failure to cope with the environment is sometimes called despair, which in
its extreme form will probably not help the individual to survive. However, a
genotype that promoted the existence of such feelings, and their consequent
behavioral manifestations, might spread because close relatives observing
the behavior might respond in a way that preserved their lives.
17. Jealousy
Jealousy is a feeling that occurs in social situations when another individ-
ual is perceived to have achieved something that the subject would like to
have achieved. The term envy may be used when possession of an object
or a position, rather than a social relationship, has been obtained. In both
of these cases there may be an element of frustration. Some of the most
extreme examples of jealousy are those feelings of male humans when they
perceive that the female whom they regard as their mate is courted by or
attracted to another male. A large proportion of murders of women arise
from male jealously and they refer to the relative uncertainty about pater-
nity as compared with the certainty about maternity (Nesse and Williams,
1995). Hence, it would seen that the feeling of jealousy can be adaptive in
that it reduces the chances that males will lose their mates or invest resources
in offspring that are not their own. The biological situation can exist in
various species in addition to humans; therefore, the feeling may also exist
in them. On some occasions, jealousy may involve a complex analysis of
what might have been achieved as compared with what has been achieved.
Such comparison can be useful to the individual and the feeling of jealousy
may draw attention to the situation in such a way that future performance
is improved. Excessive feelings of jealously or envy can be damaging.
18. Lust
There are times when eagerness to behave in a certain way or to obtain
a certain objective is so great that a generally pervading feeling of lust
390 DONALD M. BROOM
exists in the individual. A clear-cut example of this is the lust of a male

mammal to mate with a female. The sequence of events following exposure
to an actually or apparently receptive female is an increase in plasma
testosterone, which results in approach and display and, by means of a
positive feedback loop, more testosterone production, more approach and
display, and so on. The feeling of the individual is an increasing lust for
mating with that female. Other resources can elicit lust to various degrees.
The biological value of this feeling of lust is to increase the chances that
the objective, for example, successful mating, will be achieved. A hazard
of lust is that antisocial or directly damaging results may occur.
19. Anger
Anger is a feeling of intense displeasure in situations in which the individ-
ual is not able to control events adequately. The lack of control may be
because of the actions of other individuals, or changes in the physical world,
or inadequacies in the abilities of the individual who becomes angry. The
feeling of anger includes emotional components and may involve aggressive
and violent behavior. Anger may be preceded by other feelings, especially
frustration or pain but also fear, anxiety, grief, or lust.
An individual who is angry can sometimes achieve objectives that would
not otherwise be achieved. Some of these objectives are of considerable
importance in terms of individual survival or of reproduction. On other
occasions, however, anger can have damaging effects on the individual.
However, as these damaging effects may be relatively slight, even if anger
results in fitness increment only occasionally, those genotypes that facilitate
expression of the feeling could spread in the population. The effects of
anger are often more extreme where substantial increases in plasma testos-
terone can occur, that is, in males. The ideal balance, in terms of fitness,
between ready anger and retention of individual control over social relation-
ships and other environmental variables will vary according to the effective-
ness of the anger in achieving objectives and the kind of environment in
which the individuals live. Highly organized societies with much reciprocal
altruism leave less room for anger.
20. Sexual Pleasure
The feeling at orgasm is much more wide-ranging in its effects than might
be expected from the localized nature of the receptors involved. There can
be little doubt that genotypes that led to expression of substantial sexual
pleasure survived better in the population than those that led to little sexual
pleasure because their bearers would have made more efforts to achieve
mating. Indeed, it would be predicted that sexual pleasure would be one
of the most substantial forms of pleasure because otherwise individuals
would not make effective mating a sufficiently high priority objective. This
argument seems to be widely accepted in relation to males. However, Baker

and Bellis (1995) seem reluctant to use the same argument for females, as
they present (p. 49) two “hypotheses concerning the functioning of the
female copulatory orgasm,” which are (1) that it increases the chances that
the female will lie down after copulation in order to reduce sperm loss, and
(2) that the orgasmic movements tend to suck up sperm during copulation.
Hypothesis (1) receives little support from Baker and Bellis, but the more
important adaptive advantage of sexual pleasure in general and the orgasm
component in particular must surely be to increase the likelihood that
copulation will be repeated.
All species would be expected to have some feeling of sexual pleasure,
but in those species in which mating would reduce fitness in certain seasons,
this feeling might be restricted to the appropriate periods during the year.
Some difference in the mechanism that leads to sexual pleasure would also
be expected according to whether the ideal was one single mating or multi-
ple matings. A further possible advantage of sexual pleasure that would
be more likely to occur in species in which multiple matings and shared
parental care occur is that the feeling might serve to encourage bonding
between the partners.
21. Eating Pleasure
When animals eat, a variety of sensory systems are in operation, and it
is not surprising that a generalized feeling of pleasure resulting from the
procedure of eating and the taste sensations has arisen. Anticipation of the
pleasure of eating would result in one of the causal factors promoting
eating. Hence, this feeling would tend to make individuals work harder for
food than they would if there was no such pleasure. As some of the pleasure
is likely to emanate from sensations obtained during feeding on appropriate
foods, the feeling would also have the function of encouraging the individual
to eat appropriate rather than inappropriate foods. Some eating pleasure
occurs when specific foods that are damaging to the individual are consumed
or when overeating occurs, so the effects of this feeling are not always
adaptive.
It would seem likely that most species with complex nervous systems
have some eating pleasure. The particular actions or substances that elicit
this pleasure will vary from species to species. However, Cabanac (1979)
described how the relative proportions of different strengths of sugar solu-
tions drunk by rats and humans were very similar. Hence, it would appear
that this aspect of eating pleasure is similar in rats and humans.
22. Exhilaration
Pleasure is often reported by people as being felt when they walk in the
countryside, take various forms of exercise, reach a mountaintop, or stand
392 DONALD M. BROOM
by the sea or on a boat. Some of this feeling may be associated with inhaling
a high oxygen concentration or a certain small quantity of ozone. Other
aspects may reflect a high level of control over muscular activity. Yet
another aspect of the feeling may indicate an action of endogenous opioids
in the brain. Each of these is likely to be promoting the fitness of the
individual; therefore, a feeling that increases the chances of such effects is
of value to the animal.
23. Other Sensory Pleasure
A wide variety of events in life can have effects that are detectable by
sense organs and that result in pleasant feelings. Particular odors, sounds,
sights, or mechanoreceptor inputs can evoke pleasure. In some cases these
events are brief, or are subtle and difficult to appreciate. The feeling of
pleasure can be an indicator of good sensory and brain function and thus
help with the monitoring of body functioning. Other sensations are associ-
ated with important life events, so their occurrence should be promoted.
Just as with eating pleasure, some of the events may not be beneficial for
that individual at that time; not all of such feelings are adaptive.
24. Achievement Pleasure
A variety of intellectual tasks, as well as some that combine physical
effort and much brain activity, can result in feelings of pleasure. The feeling
could have been a by-product of the processing of information originally
but its continuation would have been promoted if the feeling itself conferred
an advantage on the individuals. Perhaps the pleasure resulting from effec-
tive high-level brain processing helps to make energy available for brain
functioning or changes the biochemical or electrical characteristics of a
region of the brain so that further efficient brain functioning can occur. In
the competitive and difficult social world in which many animals live, those
animals that process information in the brain efficiently are at a considerable
advantage over those that do not; feelings that promote this should be-
come widespread.
25. General Happiness
Intense pleasure of any kind can lead to people declaring themselves to
be happy. If several different kinds of pleasure are combined, this is more
likely. However, most people who are asked to say what happiness is will
refer to the absence of problems as an important part of happiness. When
Cabanac (1979) refers to a widespread control of behavior through pleasure
seeking, a major part of this must be the diminution of any bad feelings.
It is difficult to identify a general feeling of happiness, but the seeking of
this feeling is clearly a major factor in the life of complex animals.
D. FEELINGS
AS PART
OF COPING
SYSTEMS
Animals have a wide range of systems for trying to cope with their
environments. These coping systems include positive and negative aspects:
useful actions should be carried out and resources obtained but damaging
events should be avoided. Coping is having control of mental and bodily
stability (Fraser and Broom, 1990). Coping and control systems are de-
scribed at length by Broom and Johnson (1993, Chapters 2 and 3). A
substantial part of physiology and behavior plays a role in attempts to
cope with the environment. This varies from straightforward, low-energy
homeostatic mechanisms to high-urgency emergency responses. The gen-
eral idea that feelings often have a function and help individuals to cope
with their environment has been advanced (see Section CJ), especially by
Dawkins (1990,1993), Broom and Johnson (1993), and Broom (1996). The
central thrust of the arguments is that all feelings can be functional to a
greater or lesser extent. Feelings are a part of the biology of the individual
that has evolved. They are used in order to maximize its fitness, often by
helping it to cope with its environment.
The coping systems used by animals operate on different time scales.
Some must operate during a few seconds in order to be effectual, others
take hours or months. Optimal decision making depends not only on an
evaluation of energetic costs and benefits but on the urgency of action: in
other words, the costs associated with injury, death, or failure to find a
mate (Broom, 1981, p. 80). In the fastest acting, urgent coping responses,
such as avoidance of predator attack or risk of immediate injury, fear and
pain play an important role. In longer time scale coping procedures, where
various risks to the fitness of the individual are involved, feelings rather
than just intellectual calculations are among the causal factors affecting
what decisions are taken. In attempts to deal with long-term problems that
may harm the individual, aspects of suffering contribute significantly to
how the individual tries to cope. In the organisation of behavior so as to
achieve important objectives, pleasurable feelings and the expectation that
these will occur have a substantial influence. The general hypothesis ad-
vanced is that whenever a situation exists in which decisions are made that
have a big effect on the survival or potential reproductive output of the
individual, it is likely that feelings will be involved. This argument applies
to all animals with complex nervous systems, such as vertebrates and cepha-
lopods, and not just to humans. Feelings are not a minor influence on
coping systems, they are an important part of them.
In circumstances in which individuals are starting to lose control and fail
to cope, feelings may exist. These feelings might have a role in damage
limitation, which is functional. However, they might also occur when the
394 DONALD M. BROOM
individual is not coping at all, and hence the feelings have no survival
function. Extreme suffering or despair are probably not adaptive feelings,
but an observer of the same species might benefit, and a scientist might
use indications of such feelings to deduce that the animal is not coping.
11. WELFARE,
STRESS,A N D FEELINGS
A. DEFININGWELFARE, THERELEVANCE
STRESS,AND HEALTH:
OF FEELINGS
I. Welfare
Welfare is a term restricted to animals, including humans, and hence not
used for other organisms or inanimate objects. It is used in science and in
legislation and therefore must have a meaning precise enough for such use.
Welfare refers to a characteristic of an individual rather than to something
given to it, it must be measurable in a scientific way, and it must vary over
a scale from very good to very poor (Curtis, 1986; Duncan, 1987; Broom,
1988, 1991, 1996; Broom and Johnson, 1993, pp. 74-76). The original use
of the word welfare, meaning how well an individual fares or goes through
life, is followed in the definition proposed by Broom (1986): the welfare
of an individual is its state as regards its attempts to cope with its environ-
ment. Its state includes how well or how badly it is coping and how much
difficulty it is having in coping. As emphasized by Broom (1991, 1996) and
Broom and Johnson (1993, pp. 80-82) the feelings of the individual are an
important part of that state. The assessment of how good or how poor the
welfare is depends on a wide range of measures of behavior, physiology,
brain functioning, immune system functioning, pathology, injury, and life
expectancy .
This definition of welfare was not found satisfactory by Duncan (1993,
1996), who considered that it understated the importance of feelings. In-
deed, this was a valid criticism of the original paper (Broom, 1986), which
did not refer to feelings. An aim of the present chapter is to explain, first,
that feelings are important biologically, second, that they are very significant
contributors to coping systems, and hence third, that they are encompassed
properly in this definition of welfare. The key point, which has been made
in Section I, is that feelings have extremely important biological functions,
especially in coping systems. A further point is that although feelings are
part of coping systems, they do not make up all of them, so if the concept
of welfare is to be usable it must refer to all aspects of coping systems and
not just to feelings (Broom, 1993). The separation of feelings from all other
WELFARE. STRESS. A N D THE EVOLUTION O F FEELINGS 395
aspects of coping systems in biologically unsound and impractical when

welfare assessment is attempted.
2. Stress
Two other important terms in relation to welfare are stress and health.
As argued in detail by Broom and Johnson (1993, pp. 57-73), the term
stress should be limited to refer to something that is bad for the individual.
It should not be used to mean any kind of disturbance of homeostasis, and
should not be merely equated with adrenal responses. The use of stress to
refer to environmental impacts on the individual, whatever their conse-
quences, and hence the idea that stress can be good for an individual is
confusing and renders the term virtually useless. Similarly, the fact that
adrenal cortex activity can occur during beneficial activities such as mating
and prey catching and may not occur during hemorrhage, dehydration, or
harmful increases in body temperature means that such activity cannot be
used in defining stress. A measure of what constitutes adverse or harmful
effects is needed in the definition of stress, and hence, following earlier
versions of the definition by Broom (1983) and Fraser and Broom (1990),
Broom and Johnson (1993, p. 72) conclude that: “stress is an environmental
effect on an individual which overtaxes its control systems and reduces its
fitness or appears likely to do so.” The “environment” here is that which
is outside the brain. Hence, when an individual attempts to cope with its
environment but fails to do so, welfare will be poor and stress will occur.
The failure to cope would usually be associated with one or more of the
unpleasant feelings previously described and, although many aspects of
stress are psychological, stress is not itself a feeling. It is possible for stress
to occur without any associated feelings. An individual knocked uncon-
scious and otherwise injured so that, despite the operation of various coping
systems, death occurs without consciousness being regained is stressed but
does not suffer. The relationship between stress and welfare is straightfor-
ward. Welfare refers to a wide range of states from very good to very poor,
but the welfare of any stressed individual is poor. However, there can be
poor welfare without stress when the individual is coping but only with
considerable difficulty. For example, on some occasions when a person is
succeeding in coping with a minor injury, he or she may be in some pain
and may show some physiological and behavioral responses; the welfare is
poor, but there is no likelihood of any reduction of fitness and therefore
no stress.
3. Health
The word “health,” like “welfare,” can be qualified by “good” or “poor”
and varies over a range. However, health refers to the state of body systems,
396 DONALD M. BROOM
including those in the brain, which combat pathogens, tissue damage, or

physiological disorder. Welfare is a broader term covering all aspects of
coping with the environment and taking account of a wider range of feelings
and other coping mechanisms than those that affect health, especially at
the positive end of the scale. Although people regularly refer to poor health,
they sometimes use the word to mean absence of illness or injury in the
same way that people refer to welfare when they mean good welfare.
However, the precise and scientific use of health and welfare must refer
to states varying from very good to very poor. “Health” is encompassed
within the term “welfare,” and indeed is a very important part of welfare.
B. ASSESSING
WELFARE, A N D FEELINGS
STRESS,
f. What To Measure?
The assessment of welfare is discussed at length by Dantzer and Mormkde
(1979), Smidt (1983), and Broom and Johnson (1993), and is not described
in detail here. As Mason and Mend1 (1993) and Fraser (1995) have pointed
out, no single, all-embracing measure of welfare is available, and indeed,
all of the authors mentioned above refer to the necessity to use a range of
measurements in studies where welfare following different short-term or
long-term treatments is assessed. It is occasionally possible to recognize
very poor welfare in a set of individuals using a single measurement, for
example, when all of the animals show early death or severe abnormalities
of behavior, physiology, or immune system function. However, the range
of methods used to try and cope, the range of effects of adversity, and the
range of feelings that should be considered means that good studies should
use an array of measures. Fraser (1995) describes welfare as a “type 3
concept,” like “safety,” which has multiple attributes so that the results of
different kinds of assessment must be weighed one against the other. How-
ever, in comparing the welfare of animals in two conditions, each scientific
measurement and its interpretation must be objective and unaffected by
prejudice about which condition is better. Occasionally the data available
may not allow a conclusion to be reached, but eventually the scientific
approach will provide data that give information about how good or how
poor the welfare is in each case. Once this has been done, a moral judgment
about what is acceptable and what is not can be made (Broom, 1996).
Part of the assessment of welfare will involve the various indicators of
stress, that is, reduced fitness or potential reduction of fitness as deduced
from impaired growth, disease risk, serious injury, and so on. Another part
of the assessment will concern effects on health ascertained especially from
evaluating any pathological effect. A further part will concern observations
that indicate the extent to which the individual has good or bad feelings,
WELFARE, STRESS. AND THE EVOLUTION OF FEELINGS 397
for example, the extents to which strongly preferred behaviors can be

shown and the magnitude of physiological and behavioral changes usually
associated with pain. Many measures of behavior and physiology reveal
the extent of difficulty in coping without necessarily giving direct informa-
tion about unpleasant feelings, and some measures, such as those of failure
in immune system function, may not be related to feelings at all.
2. The Role of Preference Tests
Animals normally prefer what is good for them, and studies that provide
information about the strengths of individual preferences are of consider-
able value in designing conditions for animals and deciding what should
or should not be done in procedures involving animals (Dawkins, 1980,1990;
Duncan, 1993, 1996; Broom and Johnson, 1993, pp. 145-157). However,
preferences are affected by previous experience (Mendl, 1990), and there
are various examples of animals choosing something that is not good for
them (Duncan, 1978). Dawkins (1990) concludes that several studies with
different individuals will often be necessary as a consequence when trying
to find out which treatments or conditions are likely to result in good
welfare. The preference studies do not provide comprehensive information
about the welfare of the individuals, but should be followed up by studies
using other welfare indicators to make sure that what is preferred does not
lead to poor health or other problems.
Carefully designed preference studies can give good indications about
some feelings such as pain, thirst, or thermal discomfort. However, they
may not indicate whether the behavior is affected most by a positive or by
a negative feeling, they d o not always allow discrimination of what kind
of feeling is being avoided or sought, and they may not make it possible
to recognize some feelings such as guilt and jeolousy. Like measures of
abnormalities of behavior or physiology, preference tests cannot tell us
exactly what the individual studied is feeling.
Preference tests are valuable in some situations but not appropriate in
others. Studies of foods and physical and social conditions are readily
carried out by measuring how hard animals will work to obtain each re-
source, and measures of strength of aversion help in determining how
painful or frightening some events or stimuli are. However, it would be
difficult to assess stunning procedures or disease effects using preference
tests, and there are problems in interpreting strong preferences for some
harmful foods or drugs (Broom, 1996).
3. Welfare Assessment and Time
Sometimes poor welfare is very brief but very severe; at other times,
poor welfare is prolonged but slight. Similarly, pleasant experiences may
398 DONALD M. BROOM
be intense or mild, and momentary or extended in effect. Time should be

considered as a variable in decisions about welfare. In general (Broom and
Johnson, 1993,pp. 109-110), it would seem biologically meaningful to make
some assessment of intensity. Then, if PW is intensity of poor welfare, and
t is time, the multiplicative effect (PW X t ) can be calculated (Fig. 1). We
cannot be sure that a simple multiplicative relationship is correct, and this
requires experimental investigation, but the problem should not be ignored.
very
good
Poor welfare x time = X

Very
poor
Time
very
good
(b) Welfare
5 Neutral
Poor welfare x time = 3X

very
Time
very
good
........................
........................
........................
........................
........................
Poor welfare x time = 3 X

Very
poor
Time
FIG. I . Relationships between poor welfare and time. The effects of two levels of poor
welfare (PW) and two durations of environmental conditions on an individual are shown.
From Broom and Johnson (1993).
C. FEELINGS
AS PART
OF WELFARE
We are so aware of our own feelings that it is not surprising that they
are at the forefront of our minds when we think of our welfare; indeed,
some people refer to feelings as being the sum total of welfare. As discussed
at length earlier, the propensity to have feelings has evolved because some
of the feelings in each of the general categories considered have some
function, and feelings are an important part of mechanisms used in coping
with the environment. Hence, it is impossible to consider a concept of
welfare without feelings being included in that concept (Dawkins, 1990;
Broom and Johnson, 1993, pp. 33-34,8042; Broom, 1996). However, some
feelings may be epiphenomena of neural activity (Broom and Johnson,
1993, p. 80) and hence need not affect coping with the environment.
If the definition of welfare were limited to the feelings of the individual
it would not be possible to refer to the welfare of any individual that was
asleep, or anesthetized, or drugged, or suffering from a disease that affects
awareness. Neither would any disease be considered to affect welfare unless
it altered the feelings of the individual. The welfare of an individual who
was dying but was briefly euphoric because of drug administration would
be described as very good. Most people would say that welfare can be poor
in circumstances in which there are no bad feelings.
A further problem results if only feelings are considered in assessing
welfare: a great deal of important evidence in assessing welfare in practical
studies would not be used. Animals may be studied and found to have
neuromas, or extreme physiological responses, or abnormalities of behavior,
or immunosuppression, or disease, or inability to grow and reproduce, or
reduced life expectancy, but this evidence would not be used to indicate
poor welfare unless bad feelings could be demonstrated to be associated
with them in such individuals. This argument is already being used by some
people to say that systems for housing farm animals are not proved to
result in poor welfare, because the abnormalities of behavior and physiology
that occur in these conditions have not been linked with certainty to unpleas-
ant feelings. It is often difficult to convince scientists about evidence con-
cerning feelings, and some people are never convinced about feelings in
any other individual, even of their own species. Hence, it is dangerous to
let decisions about welfare depend entirely on evidence about feelings.
We should carry out research in which we try to find out as much as
possible about the feelings of individuals whose welfare we are trying to
assess. However, we should incorporate such studies in investigations of
the whole range of coping methods by studying coping systems in their
entirety. Individuals cope by using their immune systems, adrenal responses,
400 DONALD M. BROOM
and behavioral regulatory methods as well as by responding to their feelings.

Welfare assessment must take account of all of these. As Broom (1993)
and Broom and Johnson (1993, pp. 82-84) point out, welfare is worse when
there is an injury than when there is no injury, even if the individual is
narcotized or asleep, but the welfare is worse still if the individual is awake
and suffering. Similarly, an individual whose immune system functioning
is suppressed is coping with its environment less well than an individual
with normal immune system functioning, so its welfare is worse even if it
is unaware of the immunosuppression. However, the welfare of the immu-
nosuppressed individual would be worse still if it were diseased, especially
if the effects of the disease caused suffering. Welfare assessment should
involve a combination of studies of feelings and of other factors providing
information about coping.
111. SUMMARY
Animals have systems for recognizing harmful or favorable stimuli and

for monitoring the effects of environmental conditions on themselves. They
also have a range of methods for attempting to cope with perceived or actual
adversity. Mechanisms involving learning tend to maximize the chances that
things that enhance fitness will happen to them, and tend to minimize the
chances that things that reduce fitness will happen. As part of these various
methods and mechanisms, positive and negative feelings have evolved.
These feelings have a biological role that complements various other ana-
tomical, physiological, and behavioral mechanisms. Each of 24 different
kinds of feelings is discussed and the origin and possible function of each
is considered. All have some potential for improving fitness and most are
likely to have been the subject of considerable selection pressure, but some
aspects of feelings are likely to be just epiphenomena of neural mechanisms.
With this view that most aspects of feelings have evolved like other biologi-
cal mechanisms and that they help significantly in coping and responding,
a single view of welfare as the state of an individual as regards its attempts
to cope with its environment becomes clearer. Feelings are an important
part of the welfare of an individual and should be assessed as well as
possible. Other coping procedures and effects of the environment on the
individual should also be assessed. An effect on an individual that is adverse
in the long term is categorized as stress. Programs for trying to evaluate
and improve welfare should combine the use of experiments to assess what
is important to the individual by measuring the strengths of preferences,
with monitoring studies in which feelings and other aspects of welfare are
assessed more directly.
WELFARE, STRESS. AND THE EVOLUTION OF FEELINGS 401
References
Baars. J. (1988). “A Cognitive Theory of Consciousness.“ Cambridge University Press. Cam-

bridge. UK.
Baker, R. and Bellis. M. A . (1995). “Human Sperm Competition.” Chapman and Hall. London.
Bateson. P. (1991). Assessment of pain in animals. Anim. Behov. 42, 827-839.
Blood, D. C., and Studdert, V. P., eds. (1988). “Baillitre‘s Comprehensive Veterinary Diction-
ary.” Baillitre Tindall. London.
Broom, D. M. (1969a). Reactions of chicks to visual changes during the first ten days after
hatching. Anim. Behav. 17, 307-315.
Broom, D. M. (1969b). Effects of visual complexity during rearing on chicks’ reactions to
environmental change. Anim. Behav. 17,773-780.
Broom. D. M. (1981). “Biology of Behaviour.” Cambridge University Press, Cambridge, UK.
Broom, D. M. (1983). The stress concept and ways of assessing the effects of stress in farm
animals. Appl. Anim. Ethol. 11, 79.
Broom, D. M. (1985). Stress, welfare and the state of equilibrium. In ”Proceedings of the
Second European Symposium on Poultry Welfare” (R. M. Wegner, ed.), pp. 72-81.
World Poultry Science Association. Celle, Germany.
Broom, D. M. (1986). Indicators of poor welfare. Br. Vet. J. 142, 524-526.
Broom. D. M. (1988). The scientific assessment of animal welfare. Appl. Anim. Behav. Sci.
20, 5-19.
Broom. D. M. (1991). Assessing welfare and suffering. Behav. Process. 25, 117-123.
Broom, D. M. (1993). A usable definition of animal welfare. J . Agric. Environ. Ethics 6, Suppl.
2. 15-25.
Broom. D. M. (1996). Animal welfare defined in terms of attempts to cope with the environ-
ment. Acto Agric. Scarid.. Anini. Sci. Siippl. 27, 22-28.
Broom, D. M.. and Johnson, K. G . (1993). “Stress and Animal Welfare.” Chapman &
Hall, London.
Cabanac. M. (1979). Sensory pleasure. Q. Rev. B i d 54, 1-29.
Crook, J. H. (1980). “The Evolution of Human Consciousness.” Oxford University Press,
Oxford.
Curtis, S. E. ( 1 983). Perception of thermal comfort by farm animals. In “Farm Animal Housing
and Welfare” (S. H. Baxter, M. R. Baxter, and J. A. C. MacCormack, eds.), Curr. Top.
Vete. Med. Anim. Sci, pp. 59-66. Martinus Nijhoff, The Hague.
Curtis, S. E. (1986). The case for intensive forming of food animals. I n “Advances in Animal
Welfare Science” (M. W. Fox and L. D. Mickey, eds.), pp. 245-255. The Humane Society
of the United States, Washington, D.C.
Dantzer, R.. and Mormkde, P. (1979). “Le stress en 6lCvage intensii. Masson. Paris.
Dawkins, M. (1977). D o hens suffer in battery cages? environmental preferences and welfare.
Anim. Behav. 25, 1034-1046.
Dawkins, M. (1980). “Animal Suffering: The Science of Animal Welfare.” Chapman &
Hall, London.
Dawkins. M. (1990). From an animal’s point of view: Motivation, fitness and animal welfare.
Behav. Brain Sci. 13, 1-61.
Dawkins. M. (1993). “Through our Eyes Only.” Freeman, Oxford.
Dennett, D. C. (lY91). “Consciousness Explained.” Penguin, London.
de Waal. F. (1996). “Good Natured: The Origins of Right and Wrong in Humans and Other
Animals.” Harvard University Press, Cambridge, MA.
Duncan, 1. J. H. (1978). The interpretation of preference tests in animal behaviour. Appl.
Anim. Ethol. 4, 197-200.
402 DONALD M. BROOM
Duncan, I. J. H. (1987). The welfare of farm animals: An ethological approach. Sci. Prog.
71, 317.
Duncan, I. J. H. (1993). Welfare is to do with what animals feel. J. Agric. Environ. Ethics. 6,
Suppl. 2, 8-14.
Duncan, I. J. H. (1996). Animal welfare definition in terms of feelings. Actu Agric. Scand.,
Sect. A, Anim. Sci. Suppl. 27, 29-35.
Ewald, P. W. (1980). Evolutionary biology and the treatment of signs and symptoms of
infectious disease. J . Theor. Bioi. 86, 169-176.
Ewald, P. W. (1983). Host-parasite relations vectors and the evolution of disease severity.
Annic. Rev. Ecol. Syst. 14, 465-485.
Fraser. A. F., and Broom, D. A. (1990). “Farm Animal Behaviour and Welfare” 3rd ed.
C.A.B. International, Wallingford.
Fraser, D. (1995). Science, values and animal welfare: Explaining the inextricable convention.
Anim. Welfare 4, 103-117.
Gallup, G. G. (1983). Towards a comparative psychology of mind. In “Animal Cognition and
Behaviour (R. L. Mellgren, ed.). pp. 502-503. North-Holland Publ., New York.
Griffin, D. R. (1981). “The Question of Animal Awareness. 2nd ed. Rockefeller University
Press. New York.
Griffin, D. R. (1984). “Animal Thinking.” Harvard University Press, Cambridge, MA.
Grossman, S. P. (1973). “Essentials of Physiological Psychology. Wiley, New York.
Guyton, A. C. (1982). “Human Physiology and Mechanisms of Disease.” Saunders, Phila-
delphia.
Hart, B. L. (1988). Biological bases of the behavior of sick animals. Neurosci. Biobehav. Rev.
12, 123-137.
Hart. B. L. (1990). Behavioural adaptations to pathogens and parasites: Five strategies. Neu-
rosci. Biobehav. Rev. 14, 273-294.
Houston, A. I. (1997). Demand curves and welfare. Anim. Behav. 53,983-990.
Humphrey, N. K. (1986). “The Inner Eye.” Faber and Faber. London.
Kluger, M. J. (1979). Fever in eitotherms: Evolutionary implications. A m . Zool. 19,295-304.
Koolhaas. J . M., Schuurmann, T., and Fokkema, D. S. (1983). Social behaviour of rats as a
model for the psychophysiology of hypertension. In “Biobehavioural Bases of Coronary
Heart Disease” (T. M. Dembrowski. T. H. Schmidt, and G. Blumchen, eds.), pp. 391-400.
Karger, Basel.
Mason, G. J. (1991). Stereotypies and suffering. Behav. Process. 25, 103-115.
Mason, G. J., and Mend!, M. ( 1 9 9 3 Why is there no simple way of measuring animal welfare?
Anim. Werfare 2,301-319.
Melzack. R. (1973). “The Puzzle of Pain.” Basic Books, New York.
Mendl, M. (1990). Developmental experience and the potential for suffering: Does “out of
experience” mean “out of mind?” Behav. Brain Sci. 13, 28-29.
Mendl, M.. Zanella, A. J., and Broom, D. M. (1992). Physiological and reproductive correlates
of behavioural strategies in female domestic pigs. Anim. Behav. 44, 1107-1121.
Milner, P. M. (1970). “Physiological Psychology.” Holt, Rinehart and Winston, New York.
Morton, D. B.. and Griffiths, P. H. B. (1985). Guidelines on the recognition of pain, distress
and discomfort in experimental animals and an hypothesis for assessment. Vet. Rec.
116,431-436.
Nesse, R. M., and Williams, G. C. (1995). “Evolution and Healing.” Weidenfeld & Nicol-
son, London.
Ottoson, D. (1983). “Physiology of the Nervous System.” Macmillan. London.
Profet. M. (1992). The antiteratogen theory of morning sickness. In .‘The Adapted Mind”
( J . H. Barkow, ed.), pp. 327-365. Oxford University Press, New York.
WELFARE. STRESS, A N D THE EVOLUTION O F FEELINGS 403
Rodd, R. (1990). “Biology. Ethics and Animals.” Oxford University Press, Oxford.
Rushen. J. (1990). Use of aversion learning techniques to measure distress in sheep. Appl.
Anim. Behav. Sci. 28, 3-14.
Seligman, M. E. P. (1975). “Helplessness: On Depression, Development and Death.” Freeman,
San Francisco.
Shutt, D. A,, Fell, L. R., Cornell, R., Bell, A. K., Wallace, C. A,, and Smith, A. I. (1987).
Stress-induced changes in plasma concentrations of immunoreactive p endorphin and
cortisol in response to routine surgical procedures in lambs. Ausr. J. B i d Sci. 40,97-103.
Skinner, B. F. (1974). “About Behaviorism.” Jonathan Cape, London.
Skinner, B. F. (1978). “Reflections on Behaviorism and Society.” Prentice Hall, Englewood
Cliffs, NJ.
Smidt. D., ed. (1983). “Indicators Relevant to Farm Animal Welfare,“ Curr. Top. Vet. Med.
Anim. Sci.. Vol. 23. Martinus Nijhoff. Dordrecht. The Netherlands.
Sommerville, B. A.. and Broom, D. M. (1998). Olfactory awareness in domesticated animals.
Appl. Anim. Behav. Sci. (in press).
Swenson, M. J., and Reece, W. 0. (1993). “Duke’s Physiology of Domestic Animals,” 11th
ed. Comstock, Ithaca, NY.
Toates, F. (1986). “Motivational Systems.” Cambridge University Press. Cambridge, UK.
von Holst, D. (1986). Vegetative and somatic components of tree shrews’ behaviour. J. Aurvn.
Nerv. Syst., Sicppl., pp. 657-670.
Wall, P. (1979). On the relation of injury to pain. Pain 6, 253-264.
Weiskrantz, L. (1 987). Neuropsychology and the nature of consciousness. I n “Mindwaves”
(C. Blakemore and S. Greenfield, eds.). pp. 307-320. Blackwell, Oxford.
Wemelsfelder, F. (1993). The concept of animal boredom and its relationship to stereotyped
behaviour. I n “Stereotypic Animal Behaviour” (A. B. Lawrence and J. Rushen, eds.),
pp. 65-95. C.A.B. International, Wallingford.
Wiepkema. P. R. (1985). Abnormal behaviour in farm animals: Ethological implications. Neth.
J 2001.35,279-289.
ADVANCES IN THE STUDY OF BEHAVIOR, VOL. 21
Biological Conservation and Stress
HERIBERT
HOFER
AND MARION
L. EAST
MAX-PLANCK-INSTITUT FUR VERHALTENSPHYSIOLOGIE
D-82319 SEEWIESEN POST STARNBERG
GERMANY
I. INTRODUCTION
In this contribution we review why stress has important implications for

biological conservation and consider practical ways in which conservation-
ists can identify and tackle problems caused by stress. We take an evolution-
ary approach to stress, its consequences, and the adaptations that permit
animals to cope with stress. Using information from several scientific disci-
plines we outline the factors known to cause stress and the Darwinian
fitness consequences of stress. Increasingly anthropogenic factors such as
environmental pollution, tourism and leisure activities, hunting, noise, and
global warming are thought to cause stress. We investigate when anthropo-
genic factors are likely to generate stress, what conditions favor detrimental
fitness consequences, and which stress responses provide a reliable indicator
of a potentially harmful situation.
There is extensive literature on both stress and biological conservation,
but there has been little integration of the two disciplines (for an admirable
exception, see Hoffmann and Parsons, 1 9 9 1 ) , and there has been no compre-
hensive review of the importance of stress to conservation. In part it may
be because biological conservation studies and stress studies tend to focus
on different parameters. Biological conservation generally concentrates
on populations and species communities and rarely considers mechanisms
operating at the level of the individual that may be ultimately responsible
for population and community changes (e.g., Frankel and SoulC, 1981;
SoulC, 1986; Primack, 1993). Research on stress is normally concerned
with individual organisms and their developmental, behavioral, metabolic,
endocrinological, and immunological responses to stress, but it rarely (for
exceptions, see Fraser and Broom, 1990; Hoffmann and Parsons, 1991)
considers the consequences of stress for the reproductive success of individ-
uals and the persistence of populations. It is the purpose of this review
to show that (1) stress is a valuable concept in biological conservation;
405
Copyright 0 I Y Y X by Academic Press
All rights of reproduction In any form reserved
IW65-1454/YX $25 00
406 HERlBERT HOFER A N D MARION L. EAST
(2) conservation efforts often generate stress and unless this is appreciated
no effort will be made to minimize stress: and (3) the success of conservation
efforts could be improved if detrimental Darwinian fitness consequences
of stress were minimized.
Ideally, a theory of stress and biological conservation should guide con-
servation actions by accurately predicting the likely response to both natural
and anthropogenic forms of stress for individuals, populations. species,
and communities. Currently no such theory is available. However, useful
elements for such a theory can be drawn from many disciplines such as
population biology, evolutionary genetics, evolutionary ecology, physiologi-
cal ecology, ethology, animal welfare, immunology, and endocrinology. We
suggest that a Darwinian approach is essential for any theory of stress in
biological conservation. This approach considers the stress response of
organisms as an evolved trait with an adaptive value and provides a yardstick
(Darwinian fitness) to measure the consequences of anthropogenic and
natural sources of stress that is of direct relevance to conservation.
In this review we first consider the problem of defining stress and its
relevance to conservation, and how an evolutionary framework can be
incorporated into studies of stress that are relevant for conservation. We
then review how stress has been studied and develop a set of criteria to
evaluate published work. We review the natural history of stress, in particu-
lar factors influencing inter- and intrapopulation variation in the stress
response, and outline elements of a theory of stress in biological conserva-
tion. This is followed by a survey of indicators of stressed states. We then
discuss the potential €or anthropogenic environmental factors such as pollu-
tion, disturbance, hunting, noise, and climatic (global) warming to cause
stress and evaluate whether stress responses to these anthropogenic stimuli
can be equated to natural stress stimuli. We also consider whether research
and conservation activities should be considered potential stressors. We
finally discuss actions that minimize the occurrence and impact of stress in
conservation research and management. We show that such actions readily
fall into two categories: minimizing the occurrence of stress and maximizing
the ability of individuals to cope successfully with stress.
We have attempted to use examples from a diverse range of taxa, includ-
ing invertebrates, although as most studies have been done on birds and
mammals, this review is necessarily biased toward them. We use four case
studies to illustrate the difficulties and progress that stress studies in biologi-
cal conservation have made: (1) the mountain pygmy possum and develop-
ment of tourist facilities in Australia; (2) the mass die-off of seals in the
North Sea in 1988; (3) whale watching; and (4) disturbance of Antarctic
penguin breeding colonies by tourism.
BIOLOGICAL CONSERVATION AND STRESS 407
11. STRESS
IN A CONSERVATION
BIOLOGY
CONTEXT
In this section we develop a concept of stress for biological conservation

and a framework for studies of stress that is relevant to conservation.
A. STRESS, A N D STRESS
STRESSORS, RESPONSE
Currently there is no generally accepted definition of stress, or agreement
on how to use the word. Ambiguity arises because the word has been used
to describe changes in or states of the environment that may cause a change
of the organism’s internal state (e.g., “heat stress”), the internal state of
the organism, and the actions of the organism in response to environmental
stimuli (Toates, 1995). Changes in or states of the environment are often
called stressors, the internal state of the organism is called a stressed state,
and the actions of the organism, the stress response. The inability to agree
on one rigorously defined and universally accepted concept of stress has
led some authors to believe that stress as a scientific concept should be
abandoned (Rushen, 1986). If a similar rigor were applied to other biological
disciplines, then ethologists would abandon concepts like territoriality and
dominance (see Kaufmann, 1983), and population ecologists would dispense
with density dependence or population regulation (e.g., den Boer, 1990,
1991; but see Dennis and Taper, 1994; Sinclair and Pech, 1996). All these
terms continue to be used because they can be rigorously defined in some
way. They are also useful because they are parsimonious in the sense that
they summarize in an economic manner a host of phenomena that would
otherwise be cumbersome to describe.
For the purpose of biological conservation, a definition of stress should
be theoretically well founded, useful in practice, and help predict the conse-
quences of conservation actions. Before we explain which concept of stress
we consider useful for this purpose, it is necessary to look at the criteria
that have been used to define stress. A systems view of stress (Moberg,
1985; Toates, 1995) is helpful in this respect. It acknowledges that organisms
need to evaluate potential or actual challenges to internal homeostasis
(mental and bodily stability) and then act in such a way that their state is
restored to the one prior to that of a challenge (Fig. 1). The criteria that
distinguish concepts of stress include the following:
1. An organism’s experience of an environmental stimulus. A general
concept of a stress or stressor is anything that threatens (Chrousos et al.,
1988) or changes internal homeostasis (Sapolsky, 1994). This concept does
not distinguish between pleasant or aversive stimuli, and thus includes food
or sexual encounters as stressors. More restricted definitions of stressful
408 HERIBERT HOFER AND MARION L. EAST
INPUT OUTPUT
I
-
LateOi’ +
environ
ment \
00 internal states
lenviron /I
mental
st mrlus external
- actions
l ”
FITNESS CONSEQUENCESl
FIG. 1. A systems view of the links between environmental stimulus, organismal state, and
organismal response in the context of stress. Changes in or states of the environment may
be perceived by the organism as a potential challenge to internal homeostasis. The organism
then either finds itself in an actual state of deviation from homeostasis or anticipates such a
deviation. An evaluation of the likely success of potential internal andlor external actions to
restore homeostasis may preceed such actions. If the actions fail to restore homeostasis, then
the cycle starts again and eventually may cause pathological changes to the organism. Partially
derived from diagrams and discussions in Moberg (1985). Broom and Johnson (1991), and
Toates (1995). The numbers refer to the criteria that distinguish different concepts of stress
(see text).
situations stipulate that the organism’s experience of the stimulus is aversive

rather than pleasant, a view widely adopted in animal welfare and psychol-
ogy (see Broom and Johnson, 1991).
2. The anticipation of environmental stimuli. The absence of a stimulus
may create a physiological and behavioral response (“frustration”) reminis-
cent of a response that would in other circumstances be considered a stress
response (Toates, 1995). Anticipation of aversive stimuli under certain
conditions may trigger a more substantial internal response than the actual
stimulus itself (Arthur, 1987). Thus, the state of the environment in relation
to the animal’s control, familiarity, and expectation may be important
(Toates, 1995).
3. The temporal pattern of stimuli. In evolutionary ecology, stimuli are
distinguished by their predictability and their frequency of occurrence rela-
tive to the generation time of an organism. Single, unpredictable events
(e.g., an attempt to capture an animal) are contrasted with continuous
stimuli (e.g., pollution of the environment with heavy metals). Hoffman
and Parsons (1991) consider the first type to be stress, whereas Rollo (1995)
labels these as “(strong) disturbances” and reserves the term stress for the
second class of stimuli.
4. Kinds of internal changes. Selye’s (1946) original concept viewed the
stress response as composed of a specific response and a nonspecific re-
sponse common to a wide variety of environmental stimuli and determined
only by the intensity of the stimulus. He termed this the “general adaptation
syndrome.” The responses involved include cardiovascular effects and a
hormonal response associated with (1) the sympathetic-adrenal medullary
system (catecholamines such as epinephrine and norepinephrine, also
known as adrenaline and noradrenaline); and (2) the hypothalamic-
pituitary-adrenocortical (HPA) axis (also known as the adrenocortical re-
sponse: adrenocorticotropic hormone (ACTH), cortisol, corticosterone,
other corticosteroids). Thus, he tied the definition of stress to an explicit
response mechanism. Such a criterion may be too narrow and too unspecific
(Toates, 1995). Too narrow, because some life-threatening situations fail to
evoke a corticosteroid response (Freeman, 1985) and invoke physiological
responses other than these hormonal responses (Fraser et al., 1975;Moberg,
1987), and because it would exclude taxa such as invertebrates that do not
have these systems. Too unspecific, because similar behaviors or hormonal
changes may be observed under both pleasant and aversive stimuli. For
instance, toxins, infections, pain, sleep, and exercise all stimulate the hypo-
thalamus and lead to an increase in corticosteroid secretion (Rivier, 1989),
and field studies have sometimes found evidence for a positive, rather than
a negative, association between corticosteroid secretion and reproductive
activity (Wilson and Wingfield, 1992, 1994; Saltzman et al., 1994). Hence,
a corticosteroid response could be considered a condition that is neither
necessary nor sufficient, although it might be a helpful indicator in many sit-
uations.
5 . Presence or absence of external actions (behavior). Some concepts
require behavioral changes as a necessary condition (McCarty, 1983). Such
a restriction would exclude most cases of “prenatal stress” in mammals
and many interesting cases in evolutionary ecology and environmental
toxicology where the response involves changes in energy and resource
allocation within the body.
6. Kinds of behavior. Some concepts expect animals to show certain
kinds of behavior when in a stressful situation, for example, changes in
alertness and attention span, decreases in reflex time, and suppression of
feeding and sexual behavior (Chrousos et al., 1988). Toates (1995) argued
that this is not a very useful criterion, as any behavior needs to be evaluated
410 HERIBERT HOFER A N D MARION L. EAST
in terms of what the animal achieves by it and what would happen if it was
somehow prevented from executing it. For instance, both extreme agitation
and apathetic withdrawal might be considered responses to stressful situa-
tions (Mason, 1991).
7. The results of the organism's response. The most general concept in
this respect is that the results of the organism's action, its success in restoring
the state of homeostasis or stability prior to a stimulus, does not matter.
Thus, a temporary heat wave that the animal successfully escapes from by
seeking a cool refuge could be considered a stressor (cell groups A and B
in Fig. 2). More restrictive definitions consider situations stressful only if
the action of the organism fails to restore homeostasis (cell groups E and
F in Fig. 2), or if response mechanisms are being chronically stretched or
are failing (Toates, 1995), that is, the cell groups labeled 3 in Fig. 2. If this
4L
stimulus ~timulus stimulus
I
P
E
time time time
RESTORATION OF HOMEOSTASIS
within a
short period 1 ;;01
over a
yr;d
never restored;
organism remains
i
of time in a modified state
~-
I-1-I
~ ~~ ~ ~ ~
I EFFECT ON
unchanged
modified
1
I stretched/
failing
/B1I&-m ~ F1 ,
I
I
I
1 'EFFECTON
, NO 1 RESPONSE modified I
FIG.2. Separating the time course of restoration of homeostasis from fitness consequences
and the impact of the environmental stimulus on the organism's response system helps to
clarify and classify different concepts of stress. Training of the response system by successive
stimuli is indicated by the links between cells C1 and C2 and A1 and A2, and cells D1 and
D2 and B1 and B2. Definitions of stress commonly used in psychology and ethology (Toates.
1995) restrict the concept of stress to cells E3/F3, those used in animal welfare studies to the
shaded block of cells A3/C3/E3 and B3/D3/F3 (Broom and Johnson, 1991). The evolutionary
concept of stress advocated in this chapter includes all cases of cell groups A. C. and E.
BIOLOGICAL CONSERVATION AND STRESS 41 1
happens, then the organism’s state may change in a more permanent fash-
ion, and go through a prepathological state (Moberg, 1985) before processes
such as the development of an enlarged adrenal cortex, gastric ulcers, or
shrinkage of lymphatic tissues (Selye’s (1973) “stress syndrome”) create
pathological states. Some definitions demand evidence of prepathological
or pathological states before a situation can be termed stressful (Moberg,
1985). None of the restrictive definitions copes well with training effects (an
example are changes in the organismal response in successive challenges, as
indicated by arrows in Fig. 2).
8. Fitness consequences. Fraser and Broom (1990) suggested that the
Darwinian fitness of an organism may be used to evaluate objectively
whether an animal is coping, that is, regaining control of mental and bodily
stability. Following their lead, a restrictive definition of stress could be
constructed by arguing that stress encompasses only situations that, all else
being equal, are likely to or do lead to a reduction in fitness. In this form
the concept avoids arguments such as: capturing a wild animal and placing
it with conspecifics in captivity might result in an increase in fitness because
the threat of predation is removed (e.g., Toates, 1995). In fact, confinement
in captivity leads to a decrease in reproductive success. In evolutionary
ecology stressors are usually subsumed under the term environmental stress
if they exert selection pressure (Hoffmann and Parsons, 1991; Rollo, 1995).
9. Immediately lethal versus sublethal effects. Immediately lethal “stim-
uli” (e.g., successful predation) may be contrasted with effects that are at
least initially sublethal. Conventionally, only initially sublethal effects are
considered stressors, but this does not exclude sublethal effects that may
eventually cause the death of an organism. However, the distinction be-
tween immediately lethal and nonlethal stimuli is not as clear-cut. Suppose
there is variation between individuals in the tolerance to the concentration
of heavy metals in the soil. A high concentration causes outright death in
some but not all individuals in a population. Did only the survivors experi-
ence stress but not the nonsurvivors? Environmental stress as defined in
evolutionary ecology ignores this restriction. Here stressful effects are de-
fined as exerting selection pressure and thus effects that may be immediately
lethal are included (see Hoffmann and Parsons, 1991; Rollo, 1995).
CONCER
B. A N EVOLUTIONARY OF STRESS
Biological conservation strives to enhance the persistence of populations

and ecosystems in their natural setting or through ex situ measures. Popula-
tion persistence depends on individual attributes such as fecundity, survival,
and lifetime reproductive success as well as genetic and spatial population
structure and interactions between individuals (e.g., Huffaker, 1958; Ginz-

burg et a!., 1990; Karr, 1990; Walter, 1990; Renshaw, 1991; Boyce, 1992).
In this context, a concept of stress that considers the fitness consequences
of the organism’s response to stimuli appears eminently useful. Broom and
Johnson (1991) advance a similar argument in the context of animal welfare.
Thus, we define stress as an environmental effect that is likely to or does
reduce the Darwinian fitness of the organism. The Darwinian fitness of a
genotype, a particular assemblage of genes, is the per capita rate of increase
of its corresponding phenotype, the traits of the individual that carries the
genes (Sibly and Calow, 1983). In practice, fitness can be approximated by
the long-term reproductive capabilty and expected reproductive success of
individuals. These are linked to their fecundity, survival, and age at first
breeding (Sibly and Calow, 1986). Reductions in fitness can therefore be
measured as changes in the age at first breeding, survival to adulthood,
adult survival between breeding attempts, or fecundity (female offspring)
per female breeding attempt (Charlesworth, 1994).
For practical conservation activities it makes sense to distinguish suble-
thal from outright lethal effects (criterion 9) in many cases, so our preference
is to usually limit the application of the concept of stress to cases where
sublethal effects are observed. Thus, we would exclude deadly road traffic
accidents as incidences of stress but include the effect of noise due to road
traffic on the breeding performance of woodland birds. By requiring a
reduction in fitness we apply a restrictive version of criterion 1, as pleasant
stimuli would not be considered stressful (unless they have or are likely to
have detrimental fitness consequences). Rollo (1995) used the temporal
course and predictability of environmental stimuli to distinguish “stress”
from “disturbance” (criterion 3). For our purposes, we do not consider this
distinction useful. Thus, states of environment (criterion 2) are included
provided they have or are likely to have detrimental fitness consequences.
Also, the time course over which restoration of homeostasis may or may
not be completed (criterion 4) becomes irrelevant for our concept. Consider
Fig. 2: It is possible that a response may achieve restoration of homeostasis
quickly, slowly, or may fail, in which case the organism continues in a
modified state. In each case fitness may be reduced or may remain un-
changed. We include all cases where detrimental fitness consequences are
observed or are likely to happen. We are therefore not restricted to those
cases where the response mechanism is stretched or actually failing, as
others demand (Moberg, 1985; Toates, 1995).
Using a reduction in Darwinian fitness to define stress has a number of
advantages. The first is universality. It does not require specific mechanisms
(criteria 4, 5, and 6) that restrict the concept to certain taxa, or classes of
stimuli. The evolutionary concept also recognizes that organisms may em-
ploy a variety of mechanisms to cope with different stressful stimuli. Using

fitness consequences as the decisive factor provides a comparable measure
of the impact of stress across a wide variety of organisms, environmental
stimuli, and organismal response systems.
The second advantage is that it allows us to incorporate cases where the
restoration of homeostasis happens quickly but where a reduction of fitness
nevertheless occurs (criterion 7, cell group A in Fig. 2). Consider distur-
bance by boats to great crested grebes, Podiceps cristatus, incubating eggs
in vegetation on the edges of lakes. Grebes respond to approaching boats
by covering up the eggs, flying away, and returning to the nest after boats
have moved on. It is unlikely that coping mechanisms are being stretched
or are failing to restore internal homeostasis to the adult. Boating, however,
increases exposure of eggs because adults stay away longer after boating
disturbances than after other kinds of disturbance and egg losses due to
predation are likely if the eggs remain uncovered (Keller, 1989a,b). We
think it is useful to call this kind of disturbance stressful.
The third advantage is that it emphasizes that for the purpose of biological
conservation, studies of the effects of anthropogenic disturbance and envi-
ronmental change ought to measure fitness consequences. This is important
from both a theoretical and a practical point of view. The practical point
of view is that without measuring fitness consequences the interpretation
of changes in stress response may be erroneous. In her study of grebes,
Keller (1989a) showed that the distance between nests and boats at which
incubating grebes fled depended on the amount of boating. Grebes exposed
to a high amount of boating fled at shorter distances. This may be interpre-
ted as a positive sign that the grebes had successfully habituated to the
disturbance by boaters, and indeed grebes with shorter flight distances
had a higher reproductive success than those with longer flight distances.
However, grebes that had shorter flight distances and fled from approaching
boats were also more likely to cover up their nests incompletely than those
with longer flight distances (Keller, 1989a), and this may have increased
the likelihood of egg losses (Ingold et al., 1992), as even the most successful
grebes on sites with intensive recreational activity still had a lower reproduc-
tive success than on those without (Keller, 1989a).
Note that defining stress in conservation as a condition that reduces or
is likely to reduce fitness does not affect the study of potential stressors,
particularly anthropogenic ones, nor does it invalidate research that investi-
gates the organism’s stress response without looking at fitness consequences.
It merely emphasizes that for conservation purposes such studies remain
incomplete unless it can be convincingly argued that the kind and intensity
of the organism’s stress response predict detrimental fitness consequences.
Nor does such a concept prescribe whether the effects of a potential stressor
should be considered acceptable or desirable if no detrimental fitness conse-

quences are found. For instance, if plumage fouling of a species of penguin
by low amounts of crude oil has no detrimental fitness consequences, then
crude oil pollution of the investigated amount, consistency, and application
pattern would not be considered a stressor for this species. However, this
does not make oil pollution acceptable. The same application pattern of
crude oil pollution may have detrimental fitness consequences for other
species. For example, Fowler et al. (1995) demonstrated that even low levels
of oil fouling in Magellanic penguins, Spheniscus magellanicus, significantly
elevated plasma corticosterone concentrations in females and reduced their
fitness. Also, other organisms that were not investigated may suffer detri-
mental fitness consequences from pollution, habitat quality may decline,
or the composition of ecological communities may change.
Note that all concepts assume that there is an anticipated o r actual
deviation from homeostasis; the implication is that this can be recognized
easily. In practice, however, this may be a point of contention. For instance,
body mass losses of birds during the breeding season have been considered
an evolutionary adaptation designed to reduce increased energetic costs
of flight of foraging parents during the breeding season, and a stressful
consequence of increased energetic demands by the young that leads to
parents obtaining insufficient food so that they use up their own body
reserves (Moreno, 1989). The controversy hinges on the notion whether
the loss of body mass is an unexpected deviation from homeostasis or
whether it is an adaptive change in anticipation of the demands of the
breeding season. Theoretical analyses suggest that adaptive body mass
changes are expected as a consequence of trade-offs between costs and
benefits of fat storage (Lima, 1986; Houston and McNamara, 1993). The
evidence mostly favors the adaptationist view (Korpimaki, 1990; Merkle
and Barclay, 1996), although there have been cases of mass losses consistent
only with a stress explanation (Korpimaki, 1990). We return to the relation-
ship between stress and evolutionary adaptations in Section IV.
C. A FRAMEWORK
FOR STUDYING
STRESS
I N BIOLOGICAL
CONSERVATION
We now introduce a framework for studying stress in biological conserva-
tion. It concentrates on the natural history of stress and emphasizes the
evolutionary importance of stress (Fig. 3). This permits us to focus on
aspects that have been neglected in conceptual debates about stress and
that direct us to some interesting questions for biological conservation.
A natural history of stress explores the link between the variation in
environmental stimuli, the evaluation of these stimuli by organisms, and
the range of responses to and consequences of stressful stimuli. Current
knowledge of potential stressors (Fig. 3) is probably incomplete, but each

listed factor has been hypothesized or shown to be involved in stressful
situations. Natural stressors include physical and chemical characteristics
of the environment (radiation, temperature, pH, water/drought, salinity,
minerals, environmental catastrophes such as storms), and factors emanat-
ing from within a species (population density, social instability, breeding)
and other species (resource availability, pathogens, predators). Anthropo-
genic factors include global warming, pollution, noise (boats, powerful
acoustic communication and listening devices in the marine environment,
road and air traffic), disturbance by visitors and leisure activities (tourist
game viewing, hiking, paragliding, mountain biking, skiing, canoeing), seis-
mic exploration, development (often for tourist game viewing or leisure
activities), and sports hunting. Conservation-activity-related stressors in-
clude intervention and handling, confinement and captivity, transportation,
and translocation.
If stress is an evolutionary force, then an organism’s evaluation of an
environmental stimulus and its response may be an evolved trait. Variation
in evaluation and response may thus not be random but adaptive, and
hence predictable, in at least those cases where a particular type of stimulus
occurred sufficiently frequently during the evolutionary history of a popu-
lation. Section IV reviews evidence that variation in stress response is pre-
dictable and an evolved trait, and summarizes current knowledge about
the sources of variation in stress response relevant to biological conser-
vation and their consequences. The natural history of anthropogenic
stress is considered in Sections V (general anthropogenic stressors) and VI
(conservation-activity-related stressors).
For many species we know little about the kinds of stressors they face,
what factors influence the organism’s evaluation of the stimulus, what kinds
of response are available, and what consequences result from the response.
However, even if our knowledge of the natural history of the stress response
of a species to natural stimuli were well known, could we predict how they
would respond to anthropogenic stimuli? This is an important question
that has not been reviewed before (Sections V and VII).
D. QUESTIONS
ABOUT STRESS
IN THE CONTEXT
OF
CONSERVATION
BIOLOGICAL
We introduce our discussion of questions about stress in the context of
biological conservation by a case study on the development of tourist
facilities and its impact on the mountain pygmy-possum, Burramys parvus.
This case study is interesting because it illustrates many questions that
can be asked about stress-related issues and because a scientific study
(POTENTIAL) STRESSOR MODULATING STRESS CONSEQUENCI

FACTORS RESPONSE
transportation
behavioral
confinement / response
captivity (habitat
choice,
handling / movements,
intervention social inter-
sports
hunting
ieveloprnent
noise coids, cate-

cholarnines,
visitor androgens) immuno-
disturbance competence
pollution energy
requirements
global
foraging
efficiency
radiation survival
temperature breeding
success
PH
breeding
water I suppression
physiological
drought
salinity
I
I
allocation)
catastrophes
resource
availability I
I
population
density develop-
mental
breeding response
social
instability
pathogens
predators
BIOLOGICAL CONSERVATION A N D STRESS 417
accompanied the development of a ski resort from the design stage onward
so that the setup (one developed and disturbed versus one undeveloped
and undisturbed habitat) approximated a scientific experiment.
1. Case Study I : The Mountain Pygmy-Possum and
Tourist Development
The mountain pygmy-possum is one of the most threatened mammalian
species. Its distribution is restricted to less than 10 km2 total world habitat.
Its world population comprises roughly 2600 individuals in several dis-
jointed populations in the Australian Alps. Mountain pygmy-possums are
the longest lived small terrestrial mammal, they are female dominated, feed
on plants and insects, and hibernate during the winter (Mansergh and
Broome, 1994).
The development of a ski resort, accompanied by the building of roads
and the construction of ski slopes, had far-reaching consequences for the
pygmy-possum populations. Development involved clearing vegetation,
habitat destruction, and habitat fragmentation. Habitat destruction and
fragmentation in a key high-quality site reduced the proportion of adult
females and increased the proportion of adult males in the period prior to
hibernation. Female survival during hibernation was only half of the survival
in the undisturbed area and female fecundity was reduced (Mansergh and
Broome, 1994). Why? Hibernating animals rely on fat rather than food
caches to survive the winter (Geiser and Broome, 1991). Males and juveniles
are usually ejected by females from high-quality sites and this ensures that
the females accumulate sufficient fat in autumn. Prehibernation fattening
occurs at a time of reduced activity and energy expenditure (Kortner and
Geiser, 1995). It is thus possible that overcrowding the habitat with juveniles
and males disturbed the females and raised their activity levels so that
FIG. 3. A framework for the study of stress in biological conservation that emphasizes the
role of stress as an evoultionary force. Stressors are environmental stimuli that include both
natural (bottom box, left side of figure) and anthropogenic stimuli (top two boxes, left side
of figure), separated into stimuli arising from conservation activities (top box) and other
anthropogenic stimuli. If stress is an evolutionary force. then we would expect that the
evaluation of a stressor by the organism as reflected in the organism’s subsequent response
is shaped by natural selection. Thus, we expect the evaluation to vary, and, in the case of
natural environmental stimuli at least, to vary in an adaptive way. The suite of factors known
to modify response systems (modulating factors) are listed in the shaded block. The subsequent
response of the organism may activate a behavioral. hormonal, immune, developmental, and/
or physiological response. Stimulus, evaluation. and response may modify the organism’s
energy requirement and foraging efficiency, and its survival and reproductive activity and
success.
accumulation of fat reserves was insufficient (Mansergh and Broome, 1994).

Snow-grooming of ski slopes by graders, ski-mobiles, and skiers may also
have contributed because snow-grooming compacts snow and creates noise
and vibrations. This may affect microclimate and incite arousal from torpor
more often than usual. Excessive arousal would deplete energy reserves
and hence diminish survival chances.
Several conservation activities were considered or implemented. The
creation of a habitat corridor (a tunnel of rock boulders under the road)
reconnected key habitats on both sides of the main alpine highway, and
within a year the population structure in the key areas for females reverted
to the pattern known from undisturbed sites (Mansergh and Scotts, 1989).
Habitat restoration, however, is considered to be of limited use, as growth
of a key food plant, the mountain plum-pine, Podocarpus lawrencei, is slow
(.25 mm increase in diameter per year). The chance of successful restoration
is further reduced by siltation from areas where plum-pines were removed
by ski slope development, inadvertently lost due to grazing earlier this
century, or lost by fire (Mansergh and Broome, 1994). Ex situ conservation
efforts at first faced considerable difficulties. Captive-bred animals exposed
to autumn and wintering conditions in the laboratory approximating condi-
tions in the wild did not put on weight and did not hibernate at first (Geiser
e f a[., 1990), but did so under modified conditions (Broome and Geiser,
1995; Kortner and Geiser, 1995). Captive breeding and release efforts may
still require fine-tuning before they can be considered an “insurance”
against extinction (Mansergh and Broome, 1994). Captive breeding might
become valuable if the predicted winter temperature rise caused by global
warming becomes reality. Such a rise may increase winter mortality because
the energetic costs of hibernation critically depend on ambient winter tem-
perature (Geiser and Broome, 1993). Increased summer temperatures may
also limit dispersal. Animals quickly become hyperthermic at ambient tem-
peratures above 30°C (Fleming, 1985). In order to disperse during summer,
animals would have to cross valleys and hence descend to lower altitudes
where temperatures are higher. Models show that an increase of average
ambient temperature by 1°C might lead to species extinction because no
area would retain the climate pattern to which the mountain pygmy-possum
is assumed to be currently adapted (Mansergh and Broome, 1994; Brereton
et af., 1995). In addition, climatic change might also affect the migratory
behavior of the Bogong moth, Agrotis infusa, a major food source of the
pygmy-possum during the breeding season. This possibility has not yet
been investigated.
2. Questions about Stress
This case study illustrates that answers to basic questions about the
natural history of stress responses to natural and anthropogenic stressors
are crucial if conservation activities are to be effective. What are the fitness
consequences of a stressor and by what mechanism does the stressor reduce
fitness? Do animals react to stressors with developmental, behavioral, ener-
getic, hormonal, or immunological adjustments, and are these aspects of
the stress response linked? At what frequency of occurrence and/or magni-
tude does a stressor cause a reduction in fitness? Do characteristics of a
potential stressor (frequency of occurrence, magnitude) reliably predict the
magnitude of fitness consequences and the kind of stress response an animal
will exhibit? Is the pattern of the stress response a reliable predictor of
fitness consequences? Under what conditions do potential stressors cause
pathological states? The case study also illustrates that answers to questions
concerning stress may determine the success or failure of conservation
activities. To answer these questions a comparison of managed (experimen-
tal) and unmanaged (control) segments of the population may be required.
The case study shows that application of experimental principles are essen-
tial if the impact of management is to be understood.
Theories of conservation actions emphasize the preservation of genetic
diversity (SoulC, 1987). Genetic studies have not been undertaken on the
mountain pygmy-possum, but the pattern of several small populations with
potentially restricted gene flow between them (Mansergh and Broome,
1994) suggests that it would be valuable to look at genetic variation between
populations. Conservation actions often aim for maximizing “general” ge-
netic diversity, as measured by mean population heterozygosity across many
alleles, or the proportion of alleles that are heterozygous (see Nevo et al.,
1984). How compatible is such a goal with the goal that population persis-
tence may be improved by maximizing stress tolerance or stress resistance?
If past selection favored individuals that are stress resistant or stress toler-
ant, and if such selection contributed to genetic variation between popula-
tions, then the two goals may not be compatible. Futuyma (1983) argued
that if biotic stressors (pathogens, parasites, competitors, predators) are an
important threat to a population, then it may be important to preserve rare
alleles, rather than overall genetic diversity. This may be important for
conservation programs that involve the translocation of populations. Trans-
located populations and refuge populations may show reduced genetic
diversity because alleles rare in the parental population got lost (Stockwell
et al., 1996). Hoffmann and Parsons (1991) and Parsons (1989a, 1995) argued
that small marginal populations with restricted gene flow to large central
populations are more likely to experience strong selection for stress resis-
tance or stress tolerance. They expect marginal populations to contain
increased genetic diversity as well as a preponderance of stress-tolerant or
stress-resistant individuals. Their line of argument predicts that, if manage-
ment actions increased the genetic flow between central and marginal popu-
420 HERIBERT HOFER A N D M A R I O N L. EAST
lations, for example, by creating habitat corridors, then population persis-

tence may be reduced. For successful conservation efforts, detailed
knowledge about population structure, gene flow between populations,
and genetic variation for stress tolerance between marginal and central
populations might prove to be essential. Are individuals with higher stress
tolerance at a fitness disadvantage compared with normal individuals if
stressful conditions are rare (e.g., Krebs and Loeschke, 1994)? Because
inbreeding depression may be more severe under conditions of environmen-
tal stress (Miller, 1994), reduced genetic variability through inbreeding may
be a much greater problem for recently reintroduced free-ranging popula-
tions than for populations in a relatively benign captive environment. Does
captive housing of animals in temperature-regulated facilities over several
generations reduce the capacity to adapt to extreme temperatures (Kohane
and Parsons, 1988)? This type of knowledge may help guide the design of
captive facilities, the selection of individuals for breeding and reintroduction
programs, or the matching of reintroduction sites with individuals best able
to cope with the prevailing forms of stress.
These sets of questions summarize an outline of stress research in biologi-
cal conservation. This research focuses on the natural history of the stress
response and links developmental, behavioral, energetic, endocrinological,
and immunological aspects of the stress response with fitness consequences.
It identifies aspects of the natural history of a species that predict the
consequences of particular stressors. It assesses genetic variation in stress
tolerance or stress resistance between populations. It emphasizes minimal-
invasive or noninvasive monitoring and research techniques that minimize
stress caused by conservation activities. Our survey indicates that there are
many promising beginnings but that answers to many questions are not
yet available.
A CONSERVATION
111. DESIGNING STUDYT o MEASURE
STRESS
AND ITS IMPACT
Careful design of research studies and conservation activities can increase

the value of the results and foster more effective conservation. The value of
conservation management increases substantially if it conforms to standard
scientific protocol, as this improves the assessment of its impact and efficacy.
Although the recommendations in this section are not new, the violation
of scientific protocols in conservation-oriented studies suggests that a brief
summary of some important issues would be useful. Elaborations can be
found in Cochran (1977), Green (1979), Mead (1991), or Manly (1992).
A. MEASURINGSEVERAL
COMPONENTS
OF THE STRESS
RESPONSE
Is DESIRABLE
Stressed states and stress responses may be measured in several ways.
A comprehensive assessment of the impact of stress should include more
than one component of the response, as it is increasingly accepted that
there is activation of more than one physiological system (Fig. 3). A virus
infection, for instance, may trigger immunosuppression as well as neuro-
chemical and endocrine responses (Dunn et al., 1987, 1989), although the
physiological relevance and precise mechanisms that link different response
systems after an infection are only beginning to be explored (Olsen et al.,
1992; Stefan0 and Smith, 1996). Hormonal responses have also been shown
to interact with foraging behavior, energy storage, and release systems
(Astheimer et al., 1992; Rogers et al., 1993). If individuals vary in the extent
to which hormonal, resource allocation, immunological, and behavioral
systems are triggered (e.g., Hurst et al., 1996), and if these systems follow
different time courses (e.g., Sachser and Lick, 1989), then measuring only
part of the response may provide incomplete information about the magni-
tude and consequences of the full response. Thus, it is valuable to monitor
hormonal, behavioral, and immunological consequences of stress and con-
sider their links (Coe e? al., 1994; Toates, 1995; Hurst et al., 1996; Sapol-
sky, 1997).
B. THETIMESCALEOF MEASUREMENT WHICHEFFECTS

DETERMINES CAN
BE STUDIED
What time scales are required to detect fitness consequences and the full
range of the stress response? What time scales are required to assess the
efficacy of conservation efforts? Answers to such questions are vital for
the experimental design of a study. Short time scales in terms of minutes,
hours, or days usually cover most components of the immediate stress
response. The immediate hormonal consequences of ACTH stimulation
can be monitored over hours (Sapolsky, 1982) and even severe stresses
may cause changes in behavior that persist for only a few days (Astheimer
et al., 1995). However, some consequences of the immediate response, such
as a disturbance of ultradian and circadian rhythms, may take more than
one week before they are rectified, long after a challenge has terminated
(Harper et al., 1996). We know little about the time course of energy
reallocation processes.
Short-term monitoring is unlikely to identify pathological consequences.
The development of pathological states may require days (von Holst, 1986;
Sachser and Lick, 1989) or weeks (dasyurid marsupials: Lee and Cockburn,
1985; salmonids: Hare and Robertson, 1959; Robertson and Wexler, 1957,
1960). An exception to this rule is the phenomenon of capture myopathy,
a debilitating and often fatal syndrome associated with extreme muscular
exertion during pursuit, restraint, or handling (see Section V1,C). Death
due to capture myopathy may occur within hours (dugongs, Diigong dugon:
Anderson, 1981; African lion, Panthera Leo: Joubert and Stander, 1990) or
days (white-tailed deer, Odocoileus virginianus: Beringer et al., 1996).
The relationship between short-term and long-term measures of fitness
consequences of stress is complex. Short-term monitoring may sometimes
suffice to detect significant effects of handling or intervention in field studies
(see later discussion). However, long-term monitoring of the fate of individ-
uals subjected to intervention in field studies may often be useful because
interventions can result in delayed mortality, as in the case of capture
myopathy (Beringer et al., 1996), or other forms of detrimental fitness
consequences (Putman, 1995). In such cases, short-term measures would
conclude that there are no detrimental fitness consequences when in fact
they do occur. Short-term assessment of fitness consequences may also
misjudge long-term effects on population persistence. For instance, pollu-
tion may produce pronounced short-term effects on behavior of adults that
may be associated with a reduction of reproductive success during the
current breeding season. However, some studies found no significant long-
term effects on population persistence by the specific stress and application
method (insecticides and several species of passerines: Busby et al., 1990;
Millikin and Smith, 1990; crude oil and Leach’s storm petrel, Oceanodroma
leucorhoa: Butler et al., 1988).
We emphasize that long-term monitoring is particularly crucial for the
assessment of the efficacy of conservation activities. Such long-term moni-
toring is not standard practice. For instance, one set of guidelines that
eloquently summarizes protocols for vaccinating wildlife (Hall and Har-
wood, 1990) does not mention long-term monitoring of vaccinated individu-
als to judge vaccination success. Long-term monitoring is always advisable,
as it is unwise to assume that conservation actions cannot make things
worse (Thorne and Williams, 1988; Hall and Hanvood, 1990). Short-term
measures of significant seroconversion of a high proportion of vaccinated
individuals may not be a reliable guide to the long-term persistence of
protective levels of antibodies (rabies and domestic dogs: Sage et al., 1993).
Vaccination may also have detrimental effects on individual life expectancy
(canine distemper and black-footed ferret, Mustela nigripes: Carpenter et al.,
1976).These effects of vaccines may become apparent only after monitoring
individuals over several months, and the effects of other forms of interven-
tions may be apparent only after years of monitoring (rabies and African
wild dogs, Lycaon pictus: Burrows et al., 1994, 1995).
Temporal characteristics of the occurrence of stressors have not been

well studied but are of major importance to evolutionary models that relate
spatial and temporal characteristics of habitats to patterns of selection
pressure (Southwood, 1988; Wingfield et al., 1992a; Parsons, 1994; Rollo,
1995). Such temporal characteristics include the degree to which environ-
mental factors remain constant or change from season to season, or how
frequently these factors change in relation to the generation time or lifetime
of individuals. Unpredictable, severe events are considered to be rare but
may have pronounced effects on the pattern of hormonal and behavioral
response (Wingfield, 1988; Smith et al., 1994; Astheimer et al., 1995).
C. MEASURING
FITNESS
CONSEQUENCES
Is IMPORTANT
Why is it important to measure fitness consequences? Consider a study
that demonstrates that disturbance by people approaching an incubating
bird increases the animal’s heart rate by a certain percentage compared to
some baseline value. Does this result indicate a conservation problem?
Internal changes following the occurrence of a potential stressor cannot,
for the purpose of biological conservation, be properly interpreted without
measuring fitness consequences, or at least establishing a calibration curve
that relates the stress response to fitness consequences. As argued above,
population persistence depends on components of Darwinian fitness. Fit-
ness consequences of stressors that have been measured within the context
of intervention include failure to breed (Sapolsky, 1985), abandonment
of offspring (Colwell et al., 1988), a decline in the number of surviving
independent young (Rodway et al., 1996), an increase in mortality (Cotter
and Gratto, 1995), and a decline in longevity (Burrows et al., 1994).
D. GOODSTUDIESAND MANAGEMENT
ACTIVITIES
PAYATTENTION
TO
PRINCIPLES
OF EXPERIMENTAL
DESIGN
Good studies select subjects carefully, chose appropriate controls, avoid
confounding factors, optimize the number of subjects required, avoid pseu-
doreplication, and design the study in such a way that the power of statistical
tests is sufficiently high to recognize effects if there are in fact any (Green,
1979; Hurlbert, 1984; Machlis et al., 1985; Manly, 1992).
1. Selecting Individuals
Most statistical tests, including those commonly used in ecology, animal
behavior, or physiology, require a random sample of individuals. This re-
quirement is sometimes not easy to fulfill and often ignored. An example
will illustrate this. A study aims to assess the influence of vessel-based
nearshore whale watching (e.g., Stone et al., 1992) on the behavior and
fitness of migrating whales. Principles of experimental design specify that
a random sample of individuals should be constructed by randomly selecting
individuals prior to migration at, say, their overwintering site and assigning
them to either the control group (no whale watching) or the treatment group
(whale watching). The effect of whale watching could then be assessed by
applying a statistical test to selected measurements. It is unlikely that all
or even most whales assigned to one group will actually end up in that
group because the researcher is unlikely to influence the route of migration
of individual whales, so the actual composition of control and treatment
groups may be very different from the one decided by the observer prior
to migration. The measurements could still be evaluated as planned, but
now there is a caveat. Imagine that animals in poor condition stay closer
to the shore than animals in good condition and are therefore more likely
to experience whale watching. Any differences between treatment and
control groups now may be either due to body condition, or due to whale
watching, or due to the combined effect of both. In order to avoid such
problems, periods of whale watching might be alternated with periods
without, or the experimental manipulation could consist of “moving” whale
watching vessels rather than whales.
A common problem with observational studies or “natural” experiments
is that individuals within experimental and control groups have often en-
tered these groups by their own choice because of some unmeasured trait.
Such studies provide correlational evidence only if individuals are not
randomly assigned to experimental and control group. This does not detract
from the value of observational studies that often summarize empirical
evidence difficult to obtain otherwise. It just means that we ought to be
careful about how far the results of such studies can be interpreted. For
instance, in an observational study of the great skua, Catharacta s k u ,
Thompson et al. (1991) found no relationship between mercury concentra-
tions of individuals and their breeding performance or survival. They sug-
gested that mercury concentrations were not linked to fitness because there
was substantial individual variation in choosing feeding areas used before
the breeding season. Can we conclude from this study that mercury pollu-
tion does not have detrimental fitness consequences in great skuas? No,
because animals in the group with high concentrations of mercury may
have operated a different foraging tactic or used some other means to
reduce reproductive effort compared with animals in the group with low
concentrations of mercury, and the fitness costs of mercury pollution may
become apparent only at high levels of reproductive effort.
Another example is that if, understandably, researchers tend to radio-
collar those animals they consider particularly healthy in order to minimize
potentially detrimental effects of collaring, then a comparison of collared

and uncollared individuals becomes biased and cannot be done using stan-
dard statistical tests because of the requirement of random samples. For
instance, Creel (1996) and Creel et al. (1997) explicitly selected those indi-
vidual African wild dogs for radio-collaring that they considered were least
likely to suffer from any potential effects of collaring, and then went on
to compare this nonrandom sample with unhandled African wild dogs using
standard statistical procedures to argue that collaring did not cause “stress.”
Pilot studies may help to decide whether nonrandom sampling is a problem.
If the observer cannot be sure that individuals in control and treatment
groups are random samples from a larger population, then extrapolations
of the results of a study to other segments of the population may not
be warranted.
2. Controls Are Important
Having control groups, and selecting them wisely, is important. In the
absence of a control, it is often impossible to assess the effect of a stressor
on the success or otherwise of a conservation action. It is important to
observe control and experimental animals prior to the conservation action
as well as after the action (Green, 1979). Without knowing whether differ-
ences between control and experimental groups were already present prior
to the conservation action it is impossible to know whether differences
between experimental and control groups after the action are due to it or
some other factor. With a wisely chosen set of control animals or conditions,
the only difference between experimental and control groups is the stress
factor of interest, and thus any observed difference can be attributed to
the experimental factor. This is important for conservation activities that
are perceived to be beneficial, and thus applied to entire populations. An
example is the use of wildlife vaccinations where there is a tendency to
think that saving the entire population from the perceived threat of a
viral or bacterial infection is more important than considering such an
intervention as an experiment whose efficacy ought to be properly assessed.
When the Serengeti population of African wild dogs was vaccinated against
rabies, there were no unvaccinated and monitored control groups (Hein-
sohn, 1992). This is part of the reason why it is unlikely that the cause of
the extinction of the study population, which occurred shortly after the
vaccination, will be conclusively identified (Burrows et al., 1994).
3. Careful Experiments Attempt to Avoid Confounding Factors
To obtain unambiguous results, clearly only one factor should be changed
between experimental and control groups. Changing two or more factors
simultaneously in a conventional experimental design will ensure that any
observed differences between experimental and control groups cannot con-

clusively be assigned to one particular factor. With special designs it is
possible to study the effects of several factors simultaneously (Cochran,
1977; Green, 1979; Mead, 1991; Manly, 1992).
4. Optimizing the Number of Individuals Required
Optimizing the number of individuals required for a study is of both
practical and ethical importance, particularly if the experiments are conten-
tious, painful, or stressful. It is often possible to reduce the number of
individuals needed if the same individual is used both in control and in
experimental groups, where experimental conditions permit this. Such ex-
perimental data can be analyzed with paired tests (e.g., Wilcoxon signed-
rank test), which are particularly powerful. More specialist test statistics
designed for sequential samples can also be used (Gottman and Roy, 1990).
Still (1982) and McConway (1992) discuss these and other recommenda-
tions. A second problem is that sample sizes may be too small if the power
of a test is small, or if pseudoreplication occurs (see the next two sections).
The power of a test can be calculated using procedures described by Cohen
(1988). It is usually better to have less detailed data on more individuals
than to have very detailed data on only a few individuals.
5. The Power of a Test Decides Whether a Study Is Likely to Find
Significant Effects
The potential impact of anthropogenic factors is usually tested by formu-
lating a null and an alternative hypothesis, conducting observations or
experiments, and then applying a statistical test (e.g., de Swart et al., 1995a;
Ross et al., 1995). A significant effect emerges if the null hypothesis is
rejected. Such a test is associated with two types of error: rejecting the null
hypothesis when in fact it is true (Type I error with probability a ) , and
accepting the null hypothesis when it is in fact false (Type I1 error with
probability p). The power of a test is 1 - p, that is, the probability of
rejecting the null hypothesis when it is false, and can be calculated by
following the procedures outlined by Cohen (1988). If the power is high,
then the test is highly likely to find a significant difference between treat-
ment and control if a difference really exists. If the power is small, however,
then this is unlikely. In randomized trials, the power should be approxi-
mately .8-.9 (e.g., Parmar and Machin, 1995).
Ignoring the power of a test may lead to false conclusions: A lack of a
significant difference may be due to a lack of power to distinguish between
treatment and control, not because there is no real difference (Thomas
and Juanes, 1996). This is of practical importance. For instance, if the
purpose is to show that handling does not stress the animal, then a statistical
test with low power is unlikely to find an impact of handling even if there
was one. Thus, reporting the power of tests used to examine intervention
effects should become routine. However, such reports are an exception
and when values of power are reported they are usually low (.l-S), so
there is room for improvement (e.g., Houston and Greenwood, 1993; Gam-
monley and Kelley, 1995; Thirgood et al., 1995). Without information on
power, previously published conclusions that handling has no effect should
be considered tentative. An example is the debate about the potential
impact of researcher presence on the likelihood of predation on cheetah,
Acinonyx jubatus, cubs (Laurenson and Caro, 1994; O’Brien, 1994; May,
1995).
What does the power of a test depend on? Apart from the choice of test
statistic it depends on the sample size, the magnitude of the expected effect,
and the sample variance. Small sample sizes imply low power, large sample
sizes increase power. White and Garrott (1990) reanalyzed data from a
previous publication (Garrott e f al., 1985) on the effect of radio-collaring
of mule deer, Odocoileus hemionus, fawns on the chance of being preyed
upon by coyotes, Canis latrans. Garrott et al. (1985) concluded that 45
radio-collared fawns did not suffer ( p = .67) increased predation mortality
compared to 46 ear-tagged controls. White and Garrott (1990) then showed
that if collars were supposed to have doubled predation mortality, the
minimum sample size to detect a significant (at p = .05) difference between
collared animals and controls would have been 48 animals in each group,
so their sample size was insufficient to show this. To demonstrate that an
increase of merely 20% in predation mortality due to collaring was signifi-
cant would require a sample of 408 animals in each group! Because the
calculation of adequate minimum sample sizes requires an idea of the
magnitude of “treatment” effects, pilot studies may be useful to assess the
magnitude of effects likely to be encountered (Underwood and Ken-
nelly, 1990).
6. Pseudoreplication Inflates the Number of Independent
Sampling Events
Pseudoreplication occurs when individuals contribute more than one data
point to a data set and each data point is treated as if it were independent
(Hurlbert, 1984; Machlis et al., 1985). For instance, an experiment with six
fish provides six independent data points, one for each individual, even if
each individual was subjected to ten trials (for a discussion of a recent
case, see Lamprecht and Hofer, 1994; Lombardi and Hurlbert, 1996). The
frequent mistake is to assume that the sample size n required in standard
statistical tests is the number of experiments (60) rather than the number of
independent sampling events, the individuals (6). Following such a mistake
significant results are more likely because p-values depend on a. If the

number of data points varies between individuals, a second problem arises:
The number of data points per individual operates as a weighting factor.
In extreme cases, where one individual contributes a substantial proportion
of data points, the results of the statistical test depend on the behavior of
that individual.
There are several solutions to the issue of pseudoreplication. If there is
more than one data point per individual, data may be averaged, yielding
one composite value per individual, or they may be analyzed with special
tests for repeated sampling (Hurlbert, 1984; Machlis et al., 1985). Because
of the problem of pseudoreplication it is generally better to collect less
detailed data on more individuals than detailed data on very few. This also
increases the power of a test (see previous discussion).
IV. THENATURAL
HISTORY
OF STRESS
In this section we focus our discussion of the natural history of stress on

issues that are relevant to biological conservation. Our discussion will center
on factors that modulate an individual’s stress response. Figure 3 emphasizes
that there is a wide range of such factors, an aspect neglected in many
conceptual discussions of stress but of immediate and wide-ranging practical
importance for biological conservation. Such a range of modulating factors
may be expected because different populations and species have different
evolutionary histories, and optimal levels of environmental conditions,
where individuals experience no stress or a minimum of stress (Hoffmann
and Parsons, 1991), may vary between populations. Different species may
therefore respond in different ways to the same stimulus, there may be
variation in stress response between populations within a species, and there
are a host of attributes that cause variation in stress response between
individuals within a population (Fig. 3). We review this variation in three
contexts before we outline elements of a theory for such variation and
consider whether the stress response can have detrimental fitness conse-
quences and yet be considered an adaptive trait. We then review issues
surrounding the impact of stress on organisms relevant to biological conser-
vation and evaluate measures that may help to identify stressed states of
organisms and populations and thereby direct conservation work.
For a comprehensive discussion of the organismal stress response we ask
the reader to consult the chapters by Apanius, von Holst, Lima, Mlbller,
and Parsons in this volume. Before we look at the factors that modulate
the organismal stress response we briefly introduce the “cellular stress
response” as a reminder of the evolutionary continuum that links stress
responses of all organisms. We should point out that for many of the studies
we discuss in Section IV it is not clear whether the environmental stimuli
examined had detrimental fitness consequences and thus would be consid-
ered stressors according to our definition.
A. THECELLULAR
STRESSRESPONSE
The “cellular stress response” is the activation of a class of proteins
known as “heat-shock’’ proteins (HSP) when a cell is submitted to a tran-
sient rise in temperature, low pH, low oxygen level, or other physical or
chemical treatments (Morimoto et al., 1990; Nover, 1991; van Eden and
Young, 1996). These treatments lead to the accumulation of denatured
and/or aggregated proteins inside the cell. HSPs repair the damage resulting
from these conditions in a variety of ways (Watson, 1990; Burel et al., 1992;
Parsell et al., 1994). Hence, HSPs are often called stress proteins and an
increase in their production is described as a “cellular stress response.”
HSPs are present in all cells of all prokaryotic and eukaryotic organisms
and the amino acid sequences of these proteins are highly conserved in all
groups of organisms (Morimoto et al., 1990; Nover, 1991; van Eden and
Young, 1996). These characteristics indicate an evolutionary continuity of
selection pressures exerted by stress since the earliest life forms and suggest
that the cellular response may provide the molecular basis of universal
components of the organismal stress response. Links between molecular
processes on a cellular level and the organismal response are increasingly
recognized (van Eden and Young, 1996). For instance, the capacity to cope
with infections can be linked to the state of repair of cells of the immune
system (Macario, 1995). Thus, for an evolutionary understanding of stress,
the cellular stress response is of considerable interest.
B. THEORGANISMAL
STRESSRESPONSE
CANBE HIGHLY
VARIABLE
It is important to appreciate that individuals vary in their response to
the same stressor. This variation may be random, but more often it can be
predicted from modulating factors (Fig. 3). The following sections review
empirical evidence on factors that modulate the stress response and discuss
hypotheses that predict variation in response in three contexts: the ecology
of reproduction, metabolic rate and energetics, and social factors. It con-
cludes with an outline of elements that a theory would require to predict
individual variation in stress response relevant for biological conservation.
Predictions about intraspecific and interspecific variation in the stress re-
sponse would be valuable for anticipating the consequences of conservation-
related activities.
430 H E R B E R T HOFER AND MARION L. EAST
Experiments that involve situations in which the stressor has been stan-
dardized have been conducted in evolutionary studies with heat or desicca-
tion as a stressor (Hoffmann and Parsons, 1991), and in field endocrinologi-
cal studies with capture of immobilization as a stressor (Wingfield et al.,
1996). This involves first immobilizing or capturing and handling an individ-
ual and then measuring a hormonal response such as corticosteroid produc-
tion by blood sampling at regular intervals over a standard time period of
one or a few hours (Wingfield et al., 1996). An extension of this protocol
involves the initial injection of a standardized amount of ACTH (Sapolsky,
1982). In the following sections we concentrate on studies with this ap-
proach.
1. Factors Influencing the Stress Response
Differences between species in their corticosteroid stress response have
been documented in birds (e.g., Wingfield et al., 1992b) and mammals (e.g.,
Widmaier et al., 1994). In some cases, species differences were experimen-
tally shown to be a consequence of the differences in social organization
(new-world monkeys: Mendoza and Mason, 1986), or were presumed to
be related to differences in life-history parameters (Wingfield et al., 1995b;
Cockburn, 1997).
Within species, there may be significant differences in stress responses
(of any kind) between populations (Krebs and Loeschcke, 1994; Wingfield
et al., 1994a, 1995a; Dunlap and Wingfield, 1995). Evolutionary studies have
demonstrated significant genetic between-population variation in stress re-
sponse and stress tolerance to physical factors such as heat or desiccation,
for example, in Drosophila (see Hoffmann and Parsons, 1991; Krebs and
Loeschcke, 1994) and intertidal organisms (Etter, 1988). The successful
breeding of strains of laboratory animals for a specific stress response
demonstrates that the stress response may in part be genetically determined
(Satterlee et al., 1993; Backstrom and Kauffman, 1995; Lemaire and
Mormkde, 1995).
Field studies of between-population differences in the adrenocortical
response have only recently begun (Dunlap and Wingfield, 1995; Wingfield
et al., 1994a, 1995a). None of these studies provides direct evidence that
between-population variation may be genetic. Their results are nevertheless
instructive. For instance, in the western fence lizard, Sceloporus occidentalis,
the adrenocortical response to acute stress was higher in populations at
the margin of the species range than in central populations. This study
controlled for possible seasonal changes, differences in individual physio-
logical condition, seasonal changes in physiological condition, and popula-
tion differences in physiological condition. Population differences therefore
may have been generated by genetic variation, or possibly were a conse-
quence of differential ontogenetic sensitization, but this has not been inves-
tigated (Dunlap and Wingfield, 1995).
Significant differences between individuals of a population may be due
to sex (Matt et al., 1983; Wingfield, 1985; Astheimer et al., 1994; Logan and
Wingfield, 1995) and season (Bradley et al., 1980; Gustafson and Belt, 1981;
Wingfield et al., 1992b; Astheimer et al., 1994; O’Reilly and Wingfield,
1995). Individual attributes that may change within the lifetime of an indi-
vidual and significantly affect the stress response include age (Kock et al.,
1995), body condition (Smith et al., 1994; Wingfield et al., 1994a,b), molt
(Astheimer et al., 1995), lactation status in females (Higuchi et al., 1989),
stage of territoriality in males (Hannon and Wingfield, 1990), whether
females have arrived on a male territory or not (Beletsky et al., 1990),
social status (Sapolsky, 1982,1997; Sachser, 1987; Sapolsky and Ray, 1989),
social experience (Sapolsky and Ray, 1989; Sachser and Lick, 1989,1991),
social support (Sachser and Beer, 1995), personality (Sapolsky, 1997), and
individual differences in agonistic behavior (Line et al., 1996).
2. The Adrenocortical Response and the Ecology of Reproduction
The adrenocortical response (activation and tuning of the HPA axis) is
an important component of the hormonal stress response because it has at
least two potential fitness effects. First, glucocorticoids mobilize energy and
tissue nitrogen as a source of protein through gluconeogenesis, which may
improve fitness because it can improve reproductive success (Lee and Cock-
burn, 1985). Second, it may also be responsible for immunosuppression
and the creation of pathological states (Lee and McDonald, 1985;Wingfield,
1988). If the adrenocortical response is an evolved adaptation, then we
would expect that different species vary in the way in which this response
is fine-tuned. Figure 4 illustrates four options for tuning the adrenocortical
response known to vary between species or populations. The anterior pitu-
itary may or may not change its sensitivity to environmental stimuli (option
1). Adrenal activity and gonadal activity may or may not affect each other
in different species (option 2). Gonadal activity may increase or decrease
the rate of production of corticosteroid binding proteins (CBG), which
determines the concentration of biologically active glucocorticoids (option
3). Plasma concentration of glucocorticoids may or may not control adrenal
activity via negative feedback through the anterior pituitary (option 4).
a. Changing Sensitivity To Environmental Stimuli (Option 1). During a
severe snowstorm, Lapland longspurs, Calcarius lapponicus, had a higher
adrenocortical response to capture compared to that in other weather
conditions, suggesting increased sensitivity of the HPA axis in response to
severe conditions (Asheimer et al., 1995). In the common diving petrel,
Pelecanoides urinatrix, corticosterone levels usually decline with improved
I environment 1
.L@
1
I
1
I
roid binding gluco-
J
I gluconeogenesis I I pathological s t a t e s I
I mating s u c c e s s I jimmunosupressionl
FIG. 4. The basic components of the adrenocortical response system, the hypothalamic-
pituitary-adrenocortical (HPA) axis. The anterior pituitary secretes adrenocorticotropic hor-
mone (ACTH), which in turn stimulates the production of glucocorticoids in the adrenals.
The biologically active fraction of glucocorticoids is the fraction of total plasma glucocorticoids
that remains after most glucocorticoids are bound to corticosteroid binding proteins (CBG).
Negative feedback of glucocorticoids on the anterior pituitary regulates the rate of production
of ACTH. The actions of glucocorticoids may contribute both fitness benefits (gluconeogenesis.
sometimes associated with increased mating success) and fitness costs (pathological changes
in organs and immunosuppression). Comparisons between species suggest that natural selec-
tion has tuned the system in different ways in at least four compartments. (1) The anterior
pituitary may change sensitivity t o environmental stimuli. (2) Adrenal activity and gonadal
activity may be negatively coupled o r uncoupled. (3) Gonadal activity may either increase
or decrease rate of production of CBGs. (4) Negative feedback is either present or switched
off. Continuous lines: positive effects: dashed lines: negative effects: dotted lines: effects can
be positive o r negative o r the two compartments can be uncoupled.
body condition. This effect vanished during stormy weather even though
body condition differences persisted, suggesting a change in the sensitivity
of the HPA axis in some animals (Smith et al., 1994).
6. Negative Coupling of Adrenal and Gonadal Activity (Option 2). In-
creased plasma levels of glucocorticoids typically impair fertility and repro-
ductive function, and reproductive function typically reduces the capacity
to increase glucocorticoid production (Greenberg and Wingfield, 1987).
For instance, in wild baboons, Papio anubis, stress-induced suppression of
testicular function and testosterone production is due to cortisol (Sapolsky,

1985). Prolonged captivity changed adrenal gland morphology and led to
reproductive failure in female nine-banded armadillos, Dasypus novemcinc-
tus, in a manner similar to free-ranging individuals experiencing a harsh
winter (Rideout et al., 1985). The link between gonadal and adrenal activity
may pose a problem during periods of reproduction if predictable (seasonal
changes) or unpredictable environmental events (inclement weather) acti-
vated the HPA axis, perhaps maximizing survival but interfering with go-
nadal function. We would therefore expect different species exposed to
different kinds of environmental stimuli and under different breeding re-
gimes to vary in the way they weigh the importance of responding to
inclement conditions versus the continuation of reproduction.
Wingfield et al. (1995b) considered the following hypotheses: (1) In the
Arctic, a very short breeding season and the potential of summer snow-
storms can restrict food supplies to passerine birds. It should be adaptive
for arctic passerines to show less sensitivity to environmental stressors than
midlatitudinal species to avoid delays in breeding that might prevent a
breeding attempt in that season. This reasoning also applies to other habitats
where the severity of environmental conditions is similar. (2) If adverse
environmental conditions persist, then the HPA axis should become sensi-
tized to prevent debilitating consequences. (3) Larger species, or individu-
als, should have relatively larger energy reserves and thus should show a
muted response to acute stressors compared to smaller ones, and individuals
with a high fat score should respond less than individuals with a low fat
score. (4) Short-lived species should be more resistant to acute stressors
than long-lived species that have several opportunities to breed. Hence,
short-lived species should show a muted response compared to long-lived
species. ( 5 ) Species, or individuals, with a high level of parental care should
show a muted response compared to species, or individuals, with little
parental care.
Evidence is equivocal, both supporting and failing to support some of
these predictions. The adrenocortical response to acute stress was muted
in many bird species in the Arctic (see Wingfield et al., 1995b) and the
Sonoran Desert (Wingfield et al., 1992b), sensitized during Arctic summer
snowstorms (Astheimer et al., 1995), reduced in large individuals or those
with a high fat score (Wingfield et al., 1994a,b), and muted in species and
sexes that display high levels of parental care (Wingfield, 1986; Wingfield
et al., 1995b). In contrast to the predictions, the adrenocortical response
to acute stress was the same, or even higher and more variable, in some
populations in the more severe habitat and with the shorter breeding season
(Astheimer et al., 1994; Wingfield et al., 1994b, 1995a), and there appeared
to be no effect of body mass in some species, or of longevity in general

(Wingfield et al., 1995b).
c. Uncoupling Adrenal and Gonadal Activity (Option 2). In all observed
cases, an uncoupling of adrenal and gonadal activity was associated with
changes to CBG production (option 3) and the negative feedback loop
(option 4). The adaptedness of these changes has not yet been fully explored
and there are currently no hypotheses that convincingly explain why such
changes occurred in some species but not in others. The principal benefit
appears to be the mobilization of tissue nitrogen as a source of protein and
of energy reserves through gluconeogenesis at a time of peak mating activity.
This requires an increase in the concentration of biologically active gluco-
corticoids either by manipulating the amount of CBG (option 3) or by
increasing the total amount of glucocorticoids by suppressing negative feed-
back (option 4), which would otherwise reduce adrenal activity. We later
describe studies of a passerine and dasyurid marsupials where these effects
have been clarified. Similar systems operate in salmonids (Hare and Robert-
son, 1959; Robertson and Wexler, 1957, 1960) and the plaice, Pleuronectes
pfatessa (Wingfield and Grimm, 1977).
In Gambel’s white-crowned sparrow, Zonotrichia leucophrys gambelii,
adrenal activity during the breeding season is increased by uncoupling
(option 2) and lack of negative feedback (option 4; Astheimer et af., 1994).
Because testosterone simultaneously increased corticosterone-binding ca-
pacity (option 3) in males (Wingfield and Farner, 1980), corticosterone-
binding capacity did not decrease below maximum corticosterone plasma
levels. Concentration of biologically active corticosterone was still effec-
tively controlled, preventing the potentially pathological effects of high
concentrations of active glucocorticoids (Fig. 4).
In several species of dasyurid marsupials, males increase the rate of
gluconeogenesis during the breeding season by uncoupling, eliminating
negative feedback (McDonald et al., 1986; Bradley, 1990) and decreasing
corticosteroid-binding capacity through androgen stimulation, thereby los-
ing control over the concentration of active glucocorticoids (Bradley et al.,
1980;Bradley, 1987). Within a few days or weeks, the soaring concentrations
of active glucocorticoids cause immune suppression (Barker et af., 1981),
a reduction in imrnunocompetence, gastric ulcers, diseases, and lethal hem-
orrhage (Bradley et al., 1980; Bradley, 1987), leading to total male mortality
within two weeks of mating (Lee et al., 1977; Lee and Cockburn, 1985;
Bradley, 1987).
The evolutionary origins of this kind of mass male mortality at the end
of the mating season are unclear. In life-history terms its ultimate cause is
a question of the evolution of semelparity, which has been extensively
studied in other taxa (Roff, 1992; Stearns, 1992). Its proximate cause was
cited by Lee et al. (1977) as an example for the stress hypothesis advanced
by Christian (1971, 1978, 1980). This hypothesis states that an increase in
population density raises the level of agonistic behavior between individuals
provoking a stress response that ultimately impairs survival and reproduc-
tion. It viewed such stress responses as nonadaptive. Lee and McDonald
(1985) have reviewed the evidence and concluded that there is scant empiri-
cal evidence in support of this hypothesis. An alternative view (Lee and
Cockburn, 1985) suggested that such a stress response is adaptive because
increased gluconeogenesis during times of intense competition for mates
when food is in short supply reduces the need to spend time on foraging.
However, this and other adaptive hypotheses cannot accommodate all cases
of dasyurid species with total male mortality (Cockburn, 1997).
This debate suggests that even spectacular cases of mass mortality involv-
ing pathological states may not necessarily be a consequence of nonadap-
tive, aberrant pathological events. They may instead be part of a suite of
characters that maximize Darwinian fitness within the constraints set by
past phylogeny and current ecology.
3. Metabolic Rate and Energetics
Breeding experiments demonstrate that stress resistance is often related
to a genetically reduced metabolic rate (Parsons, 1993b). Cattle bred for
high growth rate under tropical conditions and increased resistance to high
parasite loads had lower mass-specific metabolic rates under fasting than
did controls (Frisch, 1981). When metabolic rate is genetically reduced,
genetic correlations may increase resistance to several environmental
stresses simultaneously (Hoffmann and Parsons, 1989). Stress-tolerant and
stress-intolerant individuals within the same population vary in their meta-
bolic efficiency and capacity for growth and reproduction, suggesting that
stress-resisting processes are energetically expensive (Sibly and Calow,
1989; Calow and Sibly, 1990; Parsons, 1990c; Tranvik et al., 1993; Wynn
et al., 1995). Experimental investigations of the link between metabolic
efficiency and the adrenocortical response have shown that activation of
the HPA axis reduces metabolic efficiency (e.g., Wingfield, 1988). Domestic
lambs that were immunized against ACTH and thus could not activate the
HPA axis increased metabolic efficiency by up to 20% during exercise, as
assessed by a reduction in oxygen consumption (Wynn et al., 1995). In
birds, corticosterone reduces responsiveness to external stimuli and thus
may promote nocturnal restfulness (Buttemer et al., 1991).
Thus, an energetic approach that considers the metabolic cost accompa-
nying major genetic changes in response to stressors of various types may be
useful (Parsons, 1990~). This may also be helpful in predicting geographical
distribution and physiological attributes of species. For instance, the range
436 HERIEERT HOFER AND MARION L. EAST
of wintering bird species in North America may be limited by the amount

of energy needed to keep warm, regardless of body size or habitat; geno-
types to function beyond these stressful limits probably do not exist because
of physiological limitations (Root, 1988). Lower metabolic rates of Galapa-
gos fur seal, Arctocephalus galapagoensis, females in comparison with Ant-
arctic fur seal, Arctocephalus gazella, females are probably an adaptation
to reduce thermal stress on land in the warm equatorial habitat (Costa and
Trillmich, 1988).
Careful experimental studies have shown that parent birds work well
below the maximum effort they can sustain when rearing young (Masman
et al., 1989), although increased daily work may have detrimental fitness
consequences (Daan et al., 1996) or may increase the incidence of malaria,
suggesting that the extra effort of tending a brood can reduce immunocom-
petence (Oppliger et al., 1996a,b). Measurements of corticosterone concen-
trations in parent birds that increased their effort during the breeding
season also provide an equivocal picture. In some studies adults that tended
larger broods were not unduly stressed by extra efforts of feeding more
nestlings (Hegner and Wingfield, 1987), whereas in others successful terri-
tory owners had the highest circulating levels of corticosterone during most
of the breeding season, implying greater energetic demands and stress
(Beletsky et al., 1989).
Trade-offs in energy allocation have been repeatedly studied in ecotoxi-
cology (Walker er al., 1996). Introducing animals from toxin-free environ-
ments to a toxic environment is likely to change the optimal allocation of
resources compared with a toxin-free environment (Holloway er al., 1990).
4. Social Factors
The importance of social factors has been effectively summarized by the
hypothesis that social skills, the control and predictability over the outcome
of social interactions, and the available social support, determine the level
of stress. Thus, it is the extent to which an organism feels challenged by a
social context, rather than the context as such, that determines its physiolog-
ical attributes (e.g.. cortisol or catecholamine concentrations) and its stress
response (change in these concentrations; Henry and Stephens, 1977). This
hypothesis predicts that individuals (1) with appropriate social skills and
an experience of a high degree of control of social interactions, and
(2) that have high social support and may reliably predict the outcome of
interactions, should show low levels of glucocorticoids and a small hormonal
and cardiovascular stress response. Individuals with a low degree of control
over social interactions, that cannot predict the course and outcome of
interactions, and that lack social support or appropriate skills should show
elevated levels of glucocorticoids and a strong response to socially stressful
situations. Supporting evidence has come from many studies of mammals

in both captive and wild settings (Sapolsky and Ray, 1989; Sachser and
Lick, 1991; Sachser et al., 1994; Sachser and Beer, 1995; Sapolsky, 1997).
Individuals that are predicted to show a strong response often, but not
invariably, include individuals of low social status (Sapolsky, 1997). How-
ever, because predictability is an important component, stability or instabil-
ity of social situations may be more important than social status in some
cases. Several studies have demonstrated that in stable contexts low-ranking
individuals may respond similarly to high-ranking individuals, whereas dif-
ferences in baseline levels and in stress response are pronounced in both
high- and low-ranking individuals if situations change from social stability
to social instability (e.g., Sachser, 1994).
Thus, the same context may present itself as a different challenge to
different individuals of the same population, or to individuals from different
species. Holding captive primates in large groups may be very stressful for
socially monogamous species but may cause little stress for species living
naturally in large groups (Mendoza and Mason, 1986).
5. Elements of a Theory of the Stress Response
How can the existence and effects of modulating factors be explained?
The appropriate evolutionary theory for such phenomena would consider
the stress response as a trait within the context of an individual’s life history
and use an optimality approach to identify traits favored by natural selection
(Sibly and Calow, 1986, 1989; Roff, 1992; Stearns, 1992; McNamara and
Houston, 1996). Throughout its life, an organism allocates resources to
production (growth and future reproduction) and to defense mechanisms
that reduce the chances of mortality. It is often the case that improving
defense mechanisms (e.g., by building detoxification procedures against
environmental pollutants) results in a reduced allocation of resources to
production. Trade-offs between life-history traits, particularly mortality and
production, are central to life-history theory, which aims to explain how
selection acts on traits when such trade-offs are involved (Sibly and Calow,
1986,1989; Roff, 1992; Stearns, 1992). How animals respond to such trade-
offs may depend on their state, for example, the amount of fat stored in
body tissue. Many modulating factors in Fig. 3 characterize variation in the
state of an individual. State-dependent models of life histories (McNamara
and Houston, 1996) will therefore be an important tool for future theoretical
conservation-oriented studies. An example is the verbal model of Wingfield
et al. (1997, in press) that describes a general state-dependent “emergency
life history stage”.
The development of life-history models for stress response traits is un-
even. Within the context of environmental pollution (ecotoxicology), there
438 HERIBERT HOFER A N D MARION L. EAS?
are testable models that examine the impact of stress on resource allocation
within an individual and its consequences for population dynamics and
population persistence (Kooijman et al., 1989; Calow and Sibly, 1990; Hol-
loway et al., 1990; Baveco and De Roos, 1996; Walker et al., 1996). Other
contexts currently lack such models. However, it should be possible to
expand on current hypotheses, for example, on variation in the adrenocorti-
cal response in relation to the ecology of reproduction (see previous discus-
sion), by placing them in a framework of state-dependent life-history evolu-
tion. Models developed for other traits, for example, state-dependent life-
history models for clutch size evolution (McNamara and Houston, 1992),
could be modified to predict how individuals should respond to stress as a
function of their body condition, the amount of parental work they do, and
so on. For instance, a trade-off well known in behavioral ecology is that
between risk of predation and/or starvation and foraging. The optimal
behavioral response to such a trade-off depends on the amount of fat stored
in body tissue and other factors (see, e.g., Lima, this volume; Krebs and
Davies, 1997). As human disturbance is often considered equivalent to a
form of predation risk (Sections IV,D,2, and VII, Sutherland, 1996), this
approach may be fruitful in conservation-oriented studies to explain varia-
tion in alarm reactions, for example, flight distances or aggregation behav-
ior, as a function of body condition, breeding status, and other factors
(Section V,C,2).
6. The Stress Response as an Adaptation
Can the stress response be considered an adaptively tuned trait if stress
is supposed to have detrimental fitness consequences? A stress response
would be considered an evolved adaptation if there was genetic variation
in the stress response (see Parsons, 1988b, 1993c) and natural selection
tuned the stress response to match the challenge posed by the environmental
stimulus (see Sober, 1993, for a more extensive discussion). As Lee and
McDonald (1985) show, Selye (1946) in his original concept of a “general
adaptation syndrome” viewed stages one, alarm, and two, “adaptation”
(the organism adjusts to the presence of a stressor) of the organism’s
response as adaptive in the evolutionary sense. That is, organisms showing
such responses would be favored by natural selection over those that do
not. Stage three, exhaustion and collapse of the system, was viewed by
Selye as maladaptive or nonadaptive in the evolutionary sense. Do such
events imply a nonadaptive situation? Or can we argue that a stress response
is an evolutionary adaptation even though pathological states occur in many
or even all cases? The answers to these two questions are no and yes.
Whether or not adaptive tuning eliminates detrimental fitness conse-
quences depends on a number of factors: (1) If current antropogenic condi-
tions are different from those under which the stress response evolved,
then the stress response may not work well because, for instance, the
magnitude and/or frequency of stress due to human actions is greater than
under natural conditions. However, the stress response would still be consid-
ered an adaptation for the conditions under which it evolved. (2) It is
possible that natural selection favored a stress response that increased the
Darwinian fitness of the genotype but included pathological states. Such a
condition could evolve if fitness benefits outweigh detrimental conse-
quences (costs); focusing only on costs would be inappropriate. A likely
example is the stress-induced mortality of male dasyurid marsupials dis-
cussed earlier. (3) The evolved response may be the optimal option among
available ones in the sense that it minimizes detrimental fitness conse-
quences but does not eliminate them. Selection would favor this response
and the evolved trait would be viewed as adaptive (Williams, 1966; So-
ber, 1993).
Thus, we have to be careful: It is unwise to assume that a response could
not have been adaptively tuned it if fails to protect the organism from all
detrimental fitness consequences. Natural selection maximizes fitness but
not necessarily the well-being of organisms.
C. THEIMPACTOF STRESSON AN ORGANISM

A detailed review of the impact of stress on the organism and the develop-
ment and interactions of the energetic, hormonal, and immunological sub-
systems of a stress response is given by Toates (1995) and several contribu-
tions to this volume, particularly those by von Holst and Apanius. Energetic
costs of avoiding predators (including human disturbance) through in-
creased vigilance or the necessity to shift into safer but poorer microhabitats
and perhaps increased competition for refuges in prey are reviewed by
Lima (this volume). The purpose of this section is therefore t o point to
some impacts that are not yet sufficiently appreciated but relevant for
biological conservation purposes.
1. Stress May Modify Resource Allocation and Individual
Energy Budgets
Coping with stress is usually energetically expensive, favoring genotypes
that have a reduced metabolic rate (see previous discussion). Whereas
numerous measurements of the energetics of foraging or breeding under
natural conditions are available, direct energetic measures of the conse-
quences of anthropogenic disturbances, including handling, intervention by
researchers, and pollution are rare. One such study calculated that each
human approach to an incubating great cormorant, Phalacrocorax carbo,
440 HERrBERT HOFER AND MARION L. EAST
colony that induces incubating birds to fly off requires an additional con-
sumption of 23 g fish per bird or 23 kg per disturbance event for a typical
colony (GrCmillet et a/., 1995). Speakman et a/. (1991) demonstrated that
each event of a tactile stimulus significantly decreased fat stores of several
species of hibernating bats by .05 g, whereas each event of a nontactile
stimulus (head torch, photographic flash, sound, speech, temperature in-
crease) decreased fat stores by merely .001 g. Contamination of sea otters,
Enhydra lutris, with crude oil increased the thermal conductance of the fur
by a factor of 1.8. This caused the otters to increase their average metabolic
rate by a factor of 1.9 through voluntary activity, and shivering, time spent
grooming, and swimming by a factor of 1.7. Metabolism and thermal con-
ductance returned to baseline levels 3-6 days after cleaning (Davis e f al.,
1988). In the mountain crab, Pseudothelphusa garmani, handling doubled
the average increase in aerial oxygen uptake (Innes et a/., 1986).
The impact of social and environmental stress on allocation of resources
during juvenile growth has been experimentally studied. In general, stres-
sors slow down growth. For instance, polychlorinated biphenyl (PCB) in-
take slowed growth of captive-bred mallards, Anus platyrhynchos, but not
that of wild wood duck, Aix sponsa (Brisbin et al., 1986), suggesting that
there are species-specific responses to environmental contaminants. Scott
and Koehn (1990) demonstrated that heterozygous coot clams, Mulinia
lateralis, grew faster than homozygous individuals and that heterozygote
advantage was significantly more important when individuals grew up under
conditions of environmental stress, suggesting that environmental stress
accentuates genetic variation in stress resistance. Correlational evidence
suggests that subordinate juvenile steelhead trout, Salmo gairdneri, experi-
enced higher rates of aggression and grew more slowly than dominant
juveniles even though both categories of individuals received equal food
rations (Abbott and Dill, 1989). Additional experiments would be required
to check that there were no genetically based differences in growth rate
that determined social status in the first place.
Several studies have assessed the increase in energy expenditure caused
by back-mounted data loggers in swimming or flying animals (Obrecht et
a/., 1988; Culik and Wilson, 1991a, 1992; Bannasch et al., 1994; Culik et al.,
1994). Additional energy expenditure because of data loggers may be as
high as 42% (Culik and Wilson, 1991a), but can be substantially reduced
by optimizing size, shape, and location of the device (Bannasch et al., 1994).
Many studies have recorded changes in heart rate in response to anthro-
pogenic disturbance. Their value is limited because an increase in heart
rate does not necessarily constitute evidence of stress. Evidence for stress
would require comprehensive records that compare intensity and frequency
of occurrence of heart rate changes due to anthropogenic stressors with
changes in natural contexts, and a demonstration that the observed changes

in heart rate have detrimental fitness consequences through an increase in
energy expenditure or other mechanisms.
2. Stress May Impair Immunological Competence
Apanius (this volume) has reviewed in detail the relationship between
stress and the immune system. We therefore restrict our comments to
conceptual and factual issues that are still insufficiently appreciated by
many practitioners of biological conservation. Incorrect but widely held
beliefs include: (1) The immune system may be impacted only if there is
chronic stress. (2) Stress does not influence an organism’s capacity to cope
with an infection by a pathogen. (3) Latent viruses (viruses that are inactive
and whose presence inside the host cannot be demonstrated yet after the
appropriate stimulus become active and pathogenic) are rare, are benign,
and are not activated by stress. (4) Conservation activities do not need to
consider the current physical condition of an organism, that is, stress caused
by handling cannot impair the organism’s capacity to cope with current in-
fections.
a. Stress Impairs Immunological Functions. It is increasingly recognized
that stressors modify an organism’s immunological competence. For in-
stance, glucocorticoids inhibit immune function by inhibiting interleukin 1
synthesis (Haour et al., 1995). Careful, detailed measurements of the impact
of stressors relevant in a biological conservation context have been under-
taken only recently. In harbor seals, Phoca vitulina, the consumption of
organochlorine-contaminated prey reduced the cytotoxic activity of periph-
eral blood mononuclear cells (natural killer cells) to a level approximately
25% lower than that in controls (Ross et al., 1996b). Several studies have
recently considered stress associated with social instability and aspects of
housing in captivity, including “enrichment” of housing conditions in pri-
mates .and rodents. In rhesus monkeys, Macaca mulatta, perturbation of
early rearing environment created numerous behavioral abnormalities and
stereotypies, but also reduced the proportion of CD8 cells, natural killer
cell activity, and increased lymphocyte proliferation response to mitogen
stimulation (Lubach et al., 1995). In the same species, “enriched” housing
enhanced phytohemagglutinin-stimulated production of interleukin 4 and
gamma interferon but reduced natural killer cell activity (Schapiro et al.,
1996). In pigtail macaques, Macaca nemestrina, and bonnet macaques, Ma-
caca radiata, separation of infants from their mothers for 2 weeks increased
infant vocalization and caused them to withdraw into a huddle. Changes
in indicators of innate immunity (e.g., natural killer cell activity) paralleled
behavioral changes but were transient (Laudenslager et al., 1996). In long-
tailed macaques, Macaca fascicularis, a change of membership in captive
social groups produced aggression and fear, increased lymphocyte counts,

reduced natural killer cell activity, and decreased lymphocyte proliferation
particularly among individuals showing high levels of fear (Line et al.,
1996). None of these studies has, however, considered or demonstrated
detrimental fitness consequences of the experimental stressor.
b. Stress Impairs the Organism’s Response to Pathogen Infection. There
are some experimental studies that assess the potentially detrimental impact
of stress on an organism’s capacity to respond to pathogen infections.
An experimental study in domestic pigs, Sus scrofa, demonstrated that
fluctuating temperatures enhanced the susceptibility to Aujeszky’s disease
virus (ADV) infection because fewer immunoglobulin-containing cells re-
sponded to ADV infection (Narita et al., 1992). Hybrids of wild boar Sus
scrofa and domestic pigs that were transported with unfamiliar conspecifics
for 5 h in confinement showed a reduction in interferon-a production after
exposure to a standardized monolayer of porcine cells infected with ADV
(Wattrang et al., 1994). Barnard et af. (1996) studied laboratory mice that
were placed in “enriched” housing facilities. In support of previous studies,
enrichment of cages with objects that are defensible increased rates of
aggression, which in turn reduced immunocompetence, as measured by
immunoglobulin G concentrations and resistance to experimental infections
with a protozoan parasite. However, the provision of refuges as part of the
enriched environment helped to alleviate some of the detrimental effects.
Thus, environmental enrichment may start a complex chain of behavioral
and immunological changes. Experimental injection of glucocorticoids in
two species of dasyurid marsupials increased the prevalence of lesions in
the prostate (cytomegalic disease) caused by a herpes virus (Barker et al.,
1981). Several experimental studies of laboratory mice, Mus musculus, have
been able to clarify the precise mechanism by which immobilization and
manual restraint impair the organism’s response to a variety of viruses
(Ozherelkov et al., 1990; Bonneau et al., 1991; Kramskaya et al., 1991; Ben-
Nathan, 1994; Pokhil’ko et al., 1995).
Many observational studies suggest a detrimental impact of stress on
coping with an infection. For instance, synergistic effects between pollution
and pathogen infections that may lead to mass mortalities have been consid-
ered in the context of catastrophic population declines in marine mammals
(Section V,A). A problem with observational studies can be that stress
may be used as an explanation of “last resort.” However, there are some
observational studies in which the evidence is suggestive. Moving a captive
herd of Dall’s sheep, Ovis dafli, to a new exhibit in a zoo may have caused
an outbreak of pneumonia caused by Mycoplasma ovipneumoniae (Black
et al., 1988). Transportation and confinement may have caused immunosup-
pression in yellow baboons, Papio cynocephalus, exacerbating the conse-
quences of mixed helminth infections and causing substantial mortality of

30% (Farah, 1996). Harsh environmental conditions in a marginal popula-
tion of apparently healthy white-tailed deer were held responsible for a
high (53%) seroprevalence of a herpes virus (Lamontagne et al., 1989). Low
water levels and high population density substantially increased mortality
of white sturgeon, Acipenser transrnontanus, infected with white sturgeon
iridovirus (LaPatra et al., 1993). Osmotic stress exacerbated the conse-
quences of infection with the iridovirus-like VEN virus and led to mass
mortality in Pacific herring, Clupea harengus pallasi (Meyers et al., 1986).
An epidemic outbreak of bluetongue virus disease and high mortality was
recorded in a herd of indigenous domestic sheep that walked in strong
sunlight for 5 days. This outbreak was surprising given the relatively mild
nature of bluetongue in indigenous sheep breeds (Eisa et al., 1980).
c. Reactivation of Latent Viruses. The most detailed studies on latency
and reactivation of viruses have been done on herpes viruses, where it is
now possible to experimentally establish latent infections in the gut of mice
(Gesser et al., 1994). Jenkins and Baum (1995) review what is currently
known about herpes simplex virus (HSV) latency and reactivation in hu-
mans and consider mechanisms by which stress-induced changes in the host
immune and nervous system might assist in establishing or reactivating
latent viral infections. Because the existence of latent viruses is hard to
prove, identification of their presence in wildlife is difficult. They may
therefore be much more common than current records suggest. Neverthe-
less, reactivation of latent viruses by stressful conditions and glucocorticoids
has been observed or experimentally demonstrated across a wide range of
vertebrate hosts and viruses (Table I). Reactivation of latent rabies virus
has also been implicated in the extinction of a study population of African
wild dogs in the Serengeti, Tanzania (Burrows, 1992; Burrows et al., 1994).
3. Prenatal and Perinatal Stress Modifies Behavior
and Irnrnunocompetence
Another aspect of stress that is little understood is the impact of environ-
mental stressors on fetuses or infants shortly after birth. Because neuro-
transmitter systems also continue to develop after birth, interactions be-
tween the individual and its environment after birth may affect the
development of these systems and have long-term effects on behavior
(Rogeness and McClure, 1996). Experimental studies on captive rhesus
monkeys, rats, mice, cats, and guinea pigs demonstrated that the impact
of environmental stimuli during these periods may alter many aspects of
neurological structures, physical and behavioral maturation, growth, loco-
motory competence, cellular immune response, the incidence of exploratory
play, aggressive, abnormal, and disturbance behavior, and gonadal activity
TABLE I
EXPERIMENTAL
A N D OBSERVATIONAL
STUDIES
THAT
SUGGEST
OR DEMONSTRATE
REACTIVATION
OF LATENT VIRUSES
BY POTENTIAL
STRESSORS
Virus Study host Stress Results Reference
Lymphocystis American plaice. Reduction of salinity Reactivation of latent iridovirus Berthiaume et a/.
disease virus Hippoglossoides (1993)
platessoides
Herpes virus Several parrot species Treatment of psittacosis Reactivation of latent herpes virus caused Eskens et nl. (1994)
massive mortality
Adenovirus Common murre Oil pollution, handling, Activation of a latent viral infection in the Lowenstine and Fry
confinement kidney (1985)
Infectious Domestic chicken Unfamiliar birds, onset of egg Onset of egg laying reactivated shedding of Hughes ef al. (1989)
laryngotracheitis lay. corticosteroid treatment large amounts of virus; other treatments
(ILT) virus reactivated sometimes or not at all
Herpes virus Brush-tailed Corticosteroid injection Reactivation of latent infection in kidneys Barker et al. (1981)
phascogale, brown
antechinus
B virus Several species of Assemblage of groups of adult The most effective known inducer of latent Zwartouw er a/.
(Herpesvirus macaques strangers to form breeding B virus, in adults with low antibody titers (1984)
simiae) colonies against the virus
Infectious bovine Calf of domestic cattle Dexamethasone injection Reactivation of virus, in a study that Castrucci er a/.
rhinotracheitis attempted to investigate reactivation of (1980)
virus (IBR) latent herpes virus
Rabies virus Raccoon. Procyon loror Cortisone treatment, pregnancy Reactivation of latent virus and subsequent McLean (1975)
mortality
Rabies virus Guinea pig. Cuvia ACTH injection. crowding Reactivation of latent virus and subsequent Soave et ul. (1961):
aperea mortality Soave (1964)
Pseudorabies virus Domestic pig Injection of corticosteroids Reactivation of latent pseudorabies virus in van Oirschot and
(PRV) 8 of 9 piglets with maternal antibodies Gielkens (1984)
born to 2 vaccinated sows at
postinoculation months 3-4
during subsequent adulthood. These alterations persist for months or years

and may be permanent (Insel et al., 1990; Maestripieri et af., 1991; Meaney
et al., 1991; Schneider, 1992c; Clarke and Schneider, 1993; Schneider and
Coe, 1993; Clarke et al., 1994; Gonzalez et al., 1994; McCune, 1995; Lubach
et al., 1996; Sachser and Kaiser, 1996; Coe et al., 1997). Stressful stimuli
examined included cold temperature, handling of the pregnant mother or
the neonatal offspring, injection of ACTH, repeated removal from familiar
housing and exposure to short, unpredictable noises, and unstable social
environments (single or chronic disruption of social relationships) during
pregnancy or lactation. However, the effect of stimuli need not always be
negative (Section VIII). None of these studies has been undertaken on
free-ranging animals, or has investigated fitness consequences of prenatal
or perinatal stress.
D. INDICATORS
OF STRESSED
STATES
Many studies have attempted to develop reliable criteria that indicate
whether an individual or a population is in a stressed state, driven by Selye’s
(1946) recognition of the “general adaptation syndrome” (Section 11,A).
Selye’s insight stimulated hope that general indicators might be found that
reliably signal a stressed state. However, the search for such indicators is
often driven by the implicit assumption that all individuals of a population
or a species will respond to a stressor in the same manner. This is unlikely
to be so because it ignores the possibility of a conditional response that
depended on the individual’s state (Section IV,B, Fig. 3). The impact of a
stressor may also be contingent on the individual’s state in that the same
environmental stressor may have detrimental fitness consequences if the
individual is in bad shape but no impact if it is in good shape. Indicators
should therefore take the individual’s state into consideration.
The search for reliable indicators is also difficult because physiological
processes that are part of a stress response are frequently activated by the
organism in other contexts (Section 11,A). Indeed, it may be parsimonious
to assume that natural selection used physiological systems that already
operated in other contexts and modified them in such a way that they may
also be used as part of a stress response, rather than assume that entirely
new systems are specifically dedicated to dealing with stress (Sibly and
Calow, 1986). Particular physiological processes might therefore be neces-
sary components of a stress response but on their own may sometimes be
insufficient to demonstrate that an organism is stressed.
Indicators may be valuable to answer some questions but not others.
One indicator may be useful to answer the question “How much does
anthropogenic disturbance reduce the effective population size in a given
habitat?,” whereas others may be more useful to answer the question “Does
pollution cause stress?” Other problems may occur: The time course of a
stress response may be captured incompletely by the indicator chosen; or
the stress response may include hormonal, energetic, and immunological
components, that may be invoked to different degrees by different individu-
als, and measuring only one component may therefore provide a misleading
picture (Section 111,A). Thus, the conclusion from this survey is again a
note of caution. Selye’s (1946) “general adaptation syndrome” and more
recent compilations of critical signs that assist in evaluations of stressed
states (e.g., Wiepkema and Kolhaas, 1993) are helpful, but they should
not be applied in an uncritical fashion. In particular, numerous studies
demonstrate that the most common prejudice-lack of “stressful behavior”
indicates lack of stress-is ill founded because behavior is often an unrelia-
ble guide to the extent to which an organism experiences stress.
1. Indicators of an Organism’s Stressed State
Most experimental studies on this have been on vertebrates, particularly
mammals and birds, and asked the question how reliably the organism’s
behavior indicates a stressed state. These studies have been conducted
under both field and captive conditions. Before we turn to behavior, we
briefly mention some other indicators.
a. Molecular Markers. Examples of molecular markers as indicators of
stressed states come from studies of invertebrates. Fang et al. (1991) demon-
strated experimentally that the ATP content of the corals Acropora hya-
cinthus and A. formosa is a reliable indicator of the degree of “stress”
caused by desiccation and bleaching. The ATP content changed rapidly
after exposure and increased during the recovery phase in subsequent
weeks. Cochrane et al. (1994) used nucleic acid probes based on chaperonin
and 70-heat-shock protein (see Section IV,A) to monitor stress-related
changes in mRNA abundance in cells of the rotifer Brachionus plicatilis
after exposure to heat. They suggested that this method may be feasible
for other marine invertebrates.
Because stress increases energy expenditure (Sections IV,B,3 and
IV,C,Z), a measure of the metabolic energy available to an organism may
indicate how stressed the organism is. High levels of available energy would
suggest stress-free conditions, low levels stressful conditions. Ivanovici and
Wiebe (1981) examined the usefulness of a measure of available metabolic
energy, the adenylate energy charge (AEC), expressed as
[ATP] + ;[ADPI
AEC =
[ATP] + [ADP] + [AMP],
where [ATP], [ADP] and [AMP] are the amounts of adenosine triphos-
phate, diphosphate, and monophosphate, respectively. Values of around

.8-.9 were typical for stress-free optimal conditions, whereas stressful condi-
tions with losses of viability usually led to values of the AE C of around .5
(Ivanovici and Wiebe, 1981).
b. Torpor. Many mammals and birds use torpor to minimize energy
expenditure during periods of low temperature (Lyman et al., 1982). This
has led some authors to assume that the occurrence of torpor is a sign of
energy reserve depletion, often called “nutritional stress” (Dawson and
Hudson, 1970). Carpenter and Hixon (1988) used field observations of
the rufous hummingbird, Selasphorus rufus, to argue that torpor in small
endotherms does not necessarily indicate that animals are energetically
stressed. Individuals in good condition may also use torpor to conserve fat
reserves that, in the case of the rufous hummingbird, may be later used on
migration. Because experimental field studies of torpor are rare, this issue
deserves further investigation.
c. Body Mass Changes May Be an Unreliable Indicator of a Stressed
State. Body mass changes during the breeding season have been frequently
debated as an indicator that breeding is “stressful.” We argued in Section
II,B that this controversy really depends on whether body mass losses are
an unexpected deviation from homeostasis or whether they are an adaptive
change in anticipation of the demands of the breeding season. Thus, al-
though body mass changes might suggest a stressed state, such a claim must
be verified.
d. Reproductive Suppression. The suppression of gonadal activity in so-
cially subordinate mammals and birds has been frequently linked to “psy-
chological stress” caused by agonistic interactions (harassment) with so-
cially superior conspecifics (Bowman et ul., 1978; Keverne et ul., 1982;
Wasser and Barash, 1983; Abbott, 1989; Bronson, 1989). If agonistic interac-
tions increased because of environmental changes, for example, decreased
food availability and reduced foraging efficiency caused by anthropogenic
disturbance, could cessation of gonadal activity be a reliable indicator of
environmental stress? Cessation could arise from general effects of raised
glucocorticoid levels that interfere with gonadal activity (see Section IV,B)
or from specific neuroendocrine mechanisms independent of the general
adrenocortical stress response. In the former case, reproductive suppression
would be an indicator of stress relevant to conservation, whereas in the
latter it might not.
Intensive work in recent years suggests that the social organization of a
species probably provides the decisive clue. In cooperatively breeding birds
and mammals a single dominant pair breeds and other adult group members
assist in the rearing of the offspring. Reproductive suppression is a natural
condition experienced by most members of a population at some stage of
their life cycle even in free-ranging populations. In such species, specific

neuroendocrine mechanisms are the prime cause of reproductive suppres-
sion, whereas the general adrenocortical response plays no role or only a
minor one (Mays et al., 1991; Schoech et al., 1991; Wingfield and Lewis,
1993; Faulkes and Abbott, 1997; French, 1997; reviewed by Abbott ef al.,
1997; Sapolsky, 1997). Because little is known about how such specific
neuroendocrine mechanisms may be triggered by environmentally depen-
dent changes in agonistic behavior, it is unclear to what extent the activation
of such mechanisms may be used as an indicator of stressed states.
In contrast, in many other group-living species, most or all females and/
or males attempt to reproduce. Socially subordinate individuals in such
species may experience reproductive failure as a consequence of an in-
creased production of glucocorticoids caused by harassment by dominant
conspecifics (Wasser and Starling 1986; Sapolsky, 1987, 1997; Altmann el
al., 1988; Dunbar, 1989). Note that such harassment has sometimes been
interpreted as an adaptive reproductive strategy of social superiors (e.g.,
Wasser and Starling, 1986). Thus, reproductive suppression may occur be-
cause stress triggers a significant adrenocortical response, but further infor-
mation is required to assess whether the stress derives from environmental
conservation-relevant sources, or whether it is the consequence of adaptive
behavior of conspecifics.
e. The Adrenocortical Response. Among several components of the hor-
monal response system (Section II,A), the adrenocortical response is fre-
quently cited as a key component of a generalized hormonal stress response
that can be conveniently measured even under field conditions. This is
because activation requires several minutes before a rise in plasma glucocor-
ticoid concentration is detectable, and field conditions often require such
time periods before a blood sample can be taken (Wingfield el al., 1996).
It is not, however, appropriate to consider it uncritically as an all-embracing
panacea. Stress responses need not involve the adrenocortical axis (Section
II,A), and glucocorticoid production may serve other functions in the con-
text of reproductive activity, basal metabolism, or mobilization of reserves
(Buttemer et al., 1991; Wilson and Wingfield, 1992, 1994; Saltzman et al.,
1994). However, the adrenocortical response is a reliable indicator in many
cases, and there is evidence that raised levels of glucocorticoids induce
profound changes in immunological competence (see previous discussion)
and are associated with an increase in heart rate and energy expenditure
(see following discussion). Such changes result in detrimental fitness conse-
quences in dasyurid marsupials (see previous discussion) and alpine mar-
mots, Marmota marmota (Arnold and Dittami, 1997).
J The Sympathetic-Adrenal Medrtllary Response. In contrast to the adre-
nocortical response, the sympathetic-adrenal medullary response (produc-
tion of catecholamines such as epinephrine and norepinephrine) has been

neglected in measurements of the hormonal response to stress in wildlife.
This is possibly because its measurement poses greater difficulties than the
measurement of the adrenocortical response. Because the sympathetic-
adrenal medullary response can operate very quickly, handling procedures
for blood sampling may increase plasma concentrations of catecholamines
severalfold within only 2 min (Le Maho et al., 1992). Le Maho et al. (1992)
used remote-controlled blood sampling on completely undisturbed captive
domestic geese, Anser anser to demonstrate that mean basal values for
epinephrine and norepinephrine recorded by their procedure were 90-
fold and 5-fold, respectively, below the lowest values previously recorded,
suggesting that significant handling effects on measurements must have
been widespread. However, because of its potentially devastating effects
on the cardiovascular system, hyperactivity of the sympathetic-adrenal med-
ullary system in response to conservation-relevant stressors may be arguably
as important as the adrenocortical response (Henry and Stephens, 1977).
For instance, catecholamines were shown to be more important than cortisol
or testosterone in the response by guinea pig males to agonistic encounters
with other males (Sachser, 1987). The measurement of catecholamines
therefore deserves closer attention by future studies. Instead of direct mea-
surements of catecholamine concentrations, measurements of indicators
may become feasible. A recent study on pigs demonstrated that the role
of certain vocalizations may be a reliable indicator of increasing epinephrine
concentrations caused by isolation of the individual from group members
(Schrader and Todt, in press).
g. Heart Rate. Heart rate has been repeatedly shown to be a reliable
predictor of oxygen consumption, and thus of metabolic rate and energy
expenditure, under field conditions (e.g., Culik, 1992). At the same time,
heart rate may be closely related to plasma cortisol concentration (Harlow
et al., 1987a,b). Whereas changes in heart rate may reliably track changes
in metabolic rate, they do not necessarily indicate stress without knowing
the behavioral or physiological context of changes in metabolic require-
ments.
h. Behavior Is Often an Unreliable Indicator of a Stressed State.
Maestripieri et al. (1992) reviewed numerous primate studies (mostly from
captive settings) and came to the conclusion that the measurement of
displacement activities might be a useful indicator of “psychosocial stress.”
Escos et al. (1995) and Alados et al. (1996) introduced two parameters as
indicators of stressful behavioral changes that characterize the temporal
pattern of activity and feeding behavior. These are the fractal dimension
of head-lift frequency and the power spectrum of the time distribution of
feeding. Under “stressful conditions” (parasitic infection, pregnancy) the
complexity of temporal behavior patterning (fractal dimension) declines.

Other indices have been suggested that attempt to capture the degree of
disturbance experienced by animals in terms of their activity levels, amount
of exploratory behavior, and other parameters, particularly in the animal
welfare literature (see Jordan and Burghardt, 1986; Broom and Johnson,
1991). However, the relationship of such parameters to cardiovascular,
adrenocortical, energetic, or immunological aspects of the stress response
is frequently unclear, and their value for predicting fitness consequences
is unknown.
Some experimental studies in both captive and field settings have at-
tempted to use physiological measurements to verify behavioral measures
as reliable indicators of a stressed state. The most frequent measures were
heart rate, often recorded by telemetry devices, and plasma cortisol concen-
trations. Behavior was not a reliable guide to elevation in heart rates of
AdClie penguins, Pygoscelis adeliae, as a response to approaches to the
breeding colony by people or aircraft (Culik et af.,1990; Wilson et at., 1991).
Heart rates increased before any change in behavior became apparent and
substantial increases in heart rate occurred even when behavioral changes
were absent. Heart rate was also a better indicator than behavior of the
stress response of free-ranging domestic sheep, Ovis aries, and bighorn
sheep, Ovis canadensis, exposed to a variety of experimental stressors,
including human disturbance (MacArthur et al., 1982; Baldock and Sibly,
1990), and correlated closely with blood cortisol levels, which in turn were
an indicator of immunological competence (Harlow et al., 1987a,b).
Similar results were obtained in captivity studies. In domestic geese,
behavior was an unreliable guide to the physiological response to handling
(Le Maho et al., 1992). Significant cardiovascular and hormonal response
with simultaneous lack of a behavioral response were also recorded in
baboons (Bentson et al., 1996), Goeldi’s monkeys, Callimico goeldii (Dett-
ling and Pryce, 1996), rhesus monkeys (Preston et al., 1996), and squirrel
monkeys, Saimiri sciureus (Martel et al., 1996). Experiments that measured
the behavioral response of domestic pigs to an approaching human demon-
strated that the pig’s behavioral response was not correlated with its re-
sponse in terms of plasma cortisol concentrations or subsequent growth
performance (Paterson and Pearce, 1992). Hurst et al. (1996) demonstrated
that in captive rats Rattus norvegicus subjected to intense social pressure,
activity increased, the sleep phase was reduced, and pathophysiological
states were pronounced.
2. Indicators of a Population’s Stressed State
In population management the question frequently arises whether food
shortage, pollution, or human disturbance reduce foraging success, thereby
decreasing the animal’s nutritional status, reproductive success or survival,

or the effective population size in a given habitat. Several measures have
been proposed and tested that might be used as an early warning system.
The analysis of urine samples collected from snow has been successfully
used to develop measures of the physiological condition of populations of
terrestrial mammals (wolf, Canis lupus: Mech et af.,1987; white-tailed deer:
Delgiudice et uf., 1989; red deer, Cervus elaphus: Delgiudice et al., 1991).
Williams and Rothery (1990) argued that in gentoo penguins, Pygoscefis
papua, the duration of foraging trips was a useful measure of food shortage,
provided the breeding stage, timing of trips, and the possibility of overnight
trips were considered. They suggested that recording coarse patterns of
the duration of foraging trips for many individuals was more valuable than
detailed records for a few. It is difficult to assess population states in many
seabird species because individuals are long lived, forage far at sea, and
breed in colonies. Foraging trips are difficult to record and changes in
population status may not be easily or quickly detected. Successful alterna-
tives t o traditional measures of individual foraging effort or mortality are
the measurement of the number of nonbreeders at breeding colonies and
a change in age at recruitment into the breeding population, as demon-
strated for the kittiwake, Rissa tridactylu, by Porter and Coulson (1987),
and for the great skua, Catharacta skua, by Klomp and Furness (1992). In
many ungulates nutritional stress may decrease growth rate, delay puberty,
and decrease fertility of young females. Williamson (1991) showed that the
number, fertility, and condition of young females worked well as an indica-
tor of population performance in red lechwe, Kobus leche leche.
Gill et al. (1996) investigated if anthropogenic disturbance of foraging
pink-footed geese, Anser brachyrhynchus, reduced their opportunity to
exploit food resources. They developed a general model that described the
trade-off between resource use and risk of disturbance as conceptually
similar to the trade-off between resource use and risk of predation. The
model requires the measurement of the amount of resources in a number
of patches, the proportion of these resources exploited by the animals, the
total number of individuals supported by the resource, and a measure of
disturbance on each patch. In the case of the geese this information was
successfully applied to derive an estimate of how many animals a disturbed
habitat could support if disturbance was reduced.
Population density itself is not necessarily a good indicator of the stressed
state of a population. If it were, density would be positively linked to
habitat quality and thus provide a composite measure of the impact of
environmental stressors. Van Horne (1983) and Reijnen and Foppen (1995)
demonstrated empirically that density is a reasonable measure of habitat
quality in years with low density. However, in years with high density,
a decline in habitat quality as measured by density in that habitat was

substantially underestimated because animals continued to crowd in habi-
tats that had lost quality. In theories of habitat use, density also does not
reflect habitat quality when overall density is high, whereas it might when
overall density is low (Fretwell, 1972; Bernstein et al., 1991; Sutherland,
1996). Studies that attempt to assess the impact of anthropogenic stressors
using population density therefore need to assess whether the population
is at a high or low density.
If an organism experiences environmental stress during its development
it may have inadequate resources to buffer developmental processes against
distortions introduced by environmental stimuli. Measures of develop-
mental stability, most importantly fluctuating asymmetry, could therefore
be used as an early warning system to indicate that a population is in a
stressed state. This topic has been reviewed by Parsons (1990a,b, 1992),
Clarke (1995, 1996), and in detail by Moller (this volume).
Warwick (1986) developed a model of how pollution or other forms of
environmental stress may modify the composition of freshwater and estua-
rine macroinvertebrate or macrozoobenthos communities. It suggests that
stressors would remove large competitive dominant species from the com-
munity. Changes in community composition would be indicated by a com-
parison of abundance (numerical species diversity) with biomass (biomass
species diversity). This approach works well for freshwater systems but not
for estuarine communities (Meire and Dereu, 1990). A computer software
package (RIVPACS) is now available that compares the actual faunal
composition of a river with a predicted composition based on a small
number of environmental features that characterize a site (Wright et al.,
1993; Wright, 1995).
V. EFFECTS
OF ANTHROPOGENIC
STRESSORS
In previous sections we used mainly experimental evidence from studies

using natural or anthropogenic stress to answer questions about the mea-
surement of the impact of stress and evaluate possible indicators of stressed
states. We now summarize the effects of several anthropogenic stressors,
again emphasizing experimental studies. We first introduce pollution and
disturbance by tourism with case studies 2 through 4 (see next section) to
illustrate the complexity of some of the issues involved. We then discuss
the impact of hunting, noise, and climatic warming. Section VI deals with
the impact of conservation management or research-related activities. We
focus on the effect of anthropogenic stressors on individuals and populations
because there is very little information on the impact on community struc-
ture or composition. This is an area that will be of major importance in

the future (Bodini et al., 1994; Knight and Gutzwiller, 1995).
A. POLLUTION
Environmental pollutants or contaminants have long been held responsi-
ble for an impairment of reproductive activity in natural populations (e.g.,
Krylov, 1990). However, pollutants rarely occur in isolation from other
potential stressors. For instance, a Dutch harbor seal, Phoca vitulina, popu-
lation declined from 350 to 17 individuals; suspected causes included high
hunting pressure, environmental pollution, loss of habitat, and disturbance
at resting places (Mees and Reijnders, 1994). Experimental evidence to
separate the contributions of various factors is therefore vital. Research
on environmental pollutants has concentrated on laboratory tests to identify
lethal doses (risk assessment), the physiological effects of pollutants, and
the geographical distribution and taxonomic occurrence of pollutants. There
have been comparatively few experiments that assess the mechanisms by
which pollutants reduce the Darwinian fitness of organisms in their natural
environment. An exception to this was the detailed studies on the effects
of pollutants on eggshell thinning in birds (reviewed by Peakall, 1993) and
lead poisoning in swans (e.g., Sears et al., 1989); other examples are reviewed
by Walker et al. (1996). The main reason seems to be that studies of
individuals in free-ranging populations rarely look at pollution, whereas
studies that look at pollution do not follow individual life histories in the
field. The 1988 mass die-off of North Sea harbor seals is a well-documented
environmental catastrophe that illustrates this and other points.
1. Case Study 2: The Mass Die-Off of North Sea Harbor Seals in 1988
The mass die-off of North Sea harbor seals in 1988 is an interesting
case that demonstrates the difficulties of identifying factors underlying
population catastrophes and separating cause and effect. A massive research
effort that continues to this day has led to the publication of many papers.
Despite this effort, the cause(s) of the 1988 mass die-off have still not yet
been conclusively identified, although a number of contributing factors have
emerged. Progress was directly related to the introduction of experimental
techniques, but historical reviews and comparisons with other populations
and die-offs were also important.
In 1988, around 18,000 harbor seals died in European waters. The first
casualties were noted in Danish and Swedish waters in April 1988, and by
autumn 1988 seals in Norway, Germany, the Netherlands, and the United
Kingdom were affected (Dietz et al., 1989). The immediate cause of death
was identified as phocine distemper virus (PDV) (Cosby et al., 1988). How-
ever, this does not necessarily provide an explanation of the spatial and
temporal pattern of mass mortalities. Where did PDV come from and why
did it cause mass mortalities? Were other environmental factors involved?
Several hypotheses were proposed (Simmonds, 1991). These were not mutu-
ally exclusive and focused on different aspects of the mass die-off:
1. Higher than average monthly temperatures may cause seals to leave
the water and aggregate on land in unusually high densities. Large aggrega-
tions and high densities may have facilitated widespread and rapid transmis-
sion of pathogens (Lavigne and Schmitz, 1990). There was indeed a correla-
tion between temperatures and several seal mass die-offs known to have
occurred in the twentieth century, including the 1988 die-off (Lavigne and
Schmitz, 1990).
2. Individuals that carried the virus may have originated from infected
seal populations that “invaded” a PDV-naive population, thereby triggering
mass mortality (Heide-Jorgensen el al., 1992). These invaders may have
been either harp seals, Phoca groenlandica, from the Barents Sea that were
known to have “invaded” Scandinavian waters in 1987 (Harwood and
Grenfell, 1990), harbor porpoises, Phocoena phocoena (Kennedy et al.,
1988), or a variety of seal species from Greenland, Canada, or the United
States (see Simmonds, 1991). Analysis of serological samples demonstrated
that the North Sea harbor seal population was PDV-naive and that a variety
of other marine mammals carried PDV (Simmonds, 1991; Heide-Jorgensen
etal., 1992). There are problems assigning the origin of PDV to a particular
marine mammal population because the data on the geographic progress
of the seal die-off and possible contact times of putative carrier candidates
do not match well, and there are no data on actual contact rates between
various species (Simmonds, 1991). It was doubted that the virus could have
been effectively transmitted under field conditions because it requires rather
specific conditions to survive transmission, but the virus must have spread
well because Danish farmed mink, Mustela vison, became infected with
PDV in 1989 (Heide-Jorgensen et al., 1992). Thus, the origin and mode of
transmission of PDV in harbor seals is currently not known.
3. An exceptional bloom of the alga Chrysochromulina polylepsis in
early 1988 may have produced high concentrations of toxic substances that
may have stressed seals and suppressed their immune systems, which in
turn may have reduced resistance to pathogen infection and caused quick
mass mortality (Lavigne and Schmitz, 1990). However, there are discrepan-
cies between the timing and spatial spread of the algal bloom and the
initiation of the seal die-off, the potential toxicity of the algal bloom is
unknown, and evidence on links between toxic algal blooms and previous
marine mammal die-offs is contentious (Lavigne and Schmitz, 1990; Sim-

monds, 1991).
4. Marine pollution, particularly organochlorines such as PCBs, may
have exacerbated the consequences of pathogen infection and thereby
initiated mass mortality (Simmonds and Johnston, 1989). When Lavigne
and Schmitz (1990), Simmonds (1991), and Heide-Jgrgensen et al. (1992)
published their papers, evidence of the possible impairment of immunocom-
petence by contaminants such as organochlorines and the fitness conse-
quences of contaminants was largely circumstantial and based on observa-
tional data in other marine mammals. Lavigne and Schmitz (1990) therefore
proposed that this explanation contributed little explanatory power to other
factors such as elevated temperatures and very high population densities.
Similarly, Skaare et al. (1990) concluded that observed organochlorine and
heavy metal concentrations from harbor seals in Norway gave no support
to suggestions that organochlorines and heavy metal pollution may have
been directly involved in the observed seal deaths. However, impairment
of fecundity by organochlorines had been demonstrated by observational
studies in Dutch harbor seals (Reijnders, 1980) and feeding experiments
with Dutch harbor seals fed on PCB-contaminated fish (Reijnders, 1986).
It was already known that exogenous toxins such as PCB and other organo-
chlorines seemed to interfere primarily with the endocrine system causing
changes similar to those present in hyperadrenocorticism in gray seals,
Halichoerus grypus, and ringed seals, Phoca hispida (Bergman and Olsson,
1986). St. Aubin and Geraci (1986) identified the zona glomerulosa as an
organ through which a stress response to contaminants or other environ-
mental stimuli might exhaust adrenal hormone reserves or desensitize the
adrenal cortex to other physiological stimuli.
Data on blood biochemistry and thyroid hormone levels from relatively
uncontaminated individuals became available in 1995 (Schumacher et al.,
1995), studies on the historical progress of stress-related morphological
changes were conducted in 1994 (Olsson et al., 1994), and the first experi-
ments to investigate potential links between contaminants and impairment
of immune function were published in 1994 (de Swart et al., 1994). More
is now also known about historical changes in levels of contaminants in seal
populations and what proportion of contaminants are transferred between
mothers and pups. Basic blood biochemcial parameters of seals changed
between areas containing polluted and unpolluted prey and were not af-
fected by captivity conditions if the animals were held in captivity for a
long time (9 months, Schumacher et al., 1995). Baltic gray seals and ringed
seals suffered from a disease complex described as a primary lesion in the
adrenals causing secondary reactions in various other organs, including
skull bone lesions. The incidence of skull bone lesions has increased in
Baltic seals since World War 11, indicating the presence of unnatural stress
factors. Analytical results and pathological findings suggested a particular
class of contaminants that include PCBs to be the most likely candidate
to instigate the disease complex (Olsson et al., 1994). Finally, long-term
experiments have demonstrated that immune function in harbor seals was
impaired if they were fed fish from polluted waters (de Swart et al., 1994).
Contaminated food not only changed cellular immunity, it also impaired
virus-specific immune responses and thus caused immunosuppression (de
Swart et al., 1993, 1995a; Ross et al., 1995). Exposure to contaminants may
therefore have had an adverse effect on the defense against virus infections,
affecting the severity of viral infections, survival rates, and the spread of
infections during recent epizootics (Ross et al., 1996b). Further studies
produced interesting evidence on vertical transmission of contamination
between mothers and pups. Female Baikal seals, Phoca sibirica, transferred
about 20% of their total DDTs and 14% of their total PCBs to the pup
during lactation (Nakata et al., 1995). Short-term fasting typical for lactating
mother harbor seals did not aggravate immunosuppression in animals with
high burdens of organochlorines (de Swart et al., 1995b), but lymphocyte
functionality and total immunoglobulin G levels were reduced in mothers
at the end of lactation (Ross et al., 1993). Pups at birth and females late
in lactation may therefore be more susceptible to infection by viral and
bacterial agents than other population segments. Perinatal exposure to
environmental contaminants represented a greater immunotoxic threat
than exposure as juvenile or adult (Ross et al., 1996a).
This spate of recent experimental and historical studies suggests that
environmental pollutants may have contributed to the severity and extent
of distemperlike infections in seals and dolphins in recent years (Osterhaus
et al., 1995).
2. Experimental Studies of Physiological Responses to and Fitness
Consequences of Pollution
Detailed measurements under well-defined conditions of the impact of
pollutants have not been undertaken until recently (cf. Dunnet, 1982).
Experimental studies both in captivity and the field now provide evidence
that the physiological response to and the fitness consequences of experi-
mental pollution may be substantial, vary between conditions of application,
species, and type of pollutant, and that short-term measures may misjudge
long-term impacts of pollution on fitness (see following discussion; Sections
III,B and IV,C,I; Table I; Walker et al., 1996).
Fowler et al. (1995) demonstrated that even low levels of oil fouling in
Magellanic penguins significantly elevated plasma corticosterone concen-
trations in females and reduced their fitness because few of the pairs with
an oiled partner later established nests with eggs. PCB-contaminated food
fed to captive mink before and during pregnancy increased urinary cortisol
concentrations during pregnancy and were associated with an increase in
the resorption of fetal tissue, and a reduction in the chance of successful
parturition (Kihlstrom et al., 1992; Madej el al., 1992).
In some cases, pollution may produce pronounced short-term effects on
behavior of adults that may be associated with a reduction of reproductive
success during the current breeding season. However, some studies found
no significant long-term effects on population persistence by the specific
pollutant and application method. Internal and external crude oil contami-
nation of one adult in breeding pairs in Leach’s storm petrel reduced
hatching success and fledging success in a dose-dependent manner, as adults
temporarily abandoned their breeding burrows, but there were no long-
term effects on reproductive success (Butler er al., 1988). In a 5-year study,
the experimental application of the insecticide fenitrothion on Canadian
forests caused temporary changes in habitat utilization in the chestnut-
sided warbler, Dendroica pensylvanica, the magnolia warbler, D. magnolia,
and the white-throated sparrow, Zonorrichia albicollis, but no long-term
changes in time budgets, suggesting that behavior was affected in the short-
term but not in the long-term (Millikin and Smith, 1990). Another study
that also experimentally applied fenitrothion on forests found that white-
throated sparrows abandoned territories, or failed to defend them, that the
application disrupted incubation, and may have been responsible for some
clutch desertion (Busby et al., 1990). However, clutch size and hatching
success were not affected.
BY TOURISM
B. DISTURBANCE AND LEISURE
ACTIVITIES
Recreational activities and other forms of nonconsumptive utilization
of wildlife are being increasingly recommended as a method to conserve
populations, and have become very important for conservation because
tourism is now the largest industry in the world and thereby a major
economic force in many countries (Goodwin, 1996). Whether tourism can
protect both indigenous people and places is an unresolved debate (King
and Stewart, 1996). Understanding of the mechanisms of the impact of
tourism on protected areas, their ecological significance, and the capacity
to manage tourism in protected areas lags behind the growth of tourism
to protected areas (Goodwin, 1996), because recreational, nonconsumptive
use of wildlife does not fit well into the existing paradigm of wildlife manage-
ment (Duffus and Dearden, 1990). Thus, the consequences of tourism-
related or recreation-related disturbance deserve increasing attention.
Some examples illustrate this. Large caves used by bats for roosting in
Mexico are preferentially targeted for tourism development but have the
highest species richness and are the only places where rare, threatened, or
endangered bat species occur (Arita, 1996). Uncontrolled tourist feeding
is a health and welfare threat to the barbary macaques, Macaca sylvanus,
of Gibraltar because they become overweight and their health deteriorates
(O’Leary, 1996). Increased rates of tourists visiting and interacting with
howler monkey, Alouatta pigra, troops led to acclimation to tourists and
an 8%population increase over 10 years, although guides regularly touched,
fed, and howled at one troop (Lash and Horwich, 1996). However, because
ecotourism started only after the area received sanctuary status in order
to protect the population from hunting, these results merely indicate that
interactions with and disturbance by tourism had less impact than hunting,
not that disturbance was not harmful. In one area in Morocco, high tourist
numbers and general human activity were responsible for clumping of
food resources in a low-density population of barbary macaques. Resource
clumping in this population increased the frequency of aggressive behavior,
caused fatalities from intraspecific fights, and had a destructive effect on
the socioecological organization compared to a high-density population
elsewhere where resource clumping did not occur (A.S. Camperio Ciani,
personal communication).
Studies of the effect of tourism disturbance have typically concentrated
on recording behavioral responses of disturbed animals in observational
studies, but have rarely attempted to use experiments to investigate the
behavioral or physiological response, or the fitness consequences of distur-
bance. In this section we look at two case studies and summarize experimen-
tal studies of the behavioral and physiological responses to and the fitness
effects of disturbance by tourism. A key problem is the design of appropriate
control groups. Case study 3 on whale watching introduces these difficulties.
Case study 4 on the potential disturbance of Antarctic penguins by tourism
illustrates some of the most advanced techniques available to modern inves-
tigations, yet the question of how much penguins are stressed by tourism
has not been fully resolved.
1. Case Study 3: Whale Watching
Whale watching is one of the fastest growing forms of ecotourism, worth
annually more than $35 million in direct and indirect benefits in eight
countries, excluding two of the biggest whale-watching nations, the United
States and Australia (International Fund for Animal Welfare [IFAW],
1996). Potential disturbance effects of watching cetaceans (whales and dol-
phins) include the presence of and the noise (see Section V,D) generated
by the engines of whale-watching vessels. These may cause whales to change
durations of dives, respiration patterns, and surface resting periods, reduce

feeding or nursing, destroy herd integrity or disrupt social groupings, and
displace them from preferred feeding habitats (reviewed by Evans, 1987,
1996; Richardson et al., 1995). Reaction to disturbance could thus adversely
affect individual energy budgets, foraging tactics, distribution and habitat
use, predation risk, and the buildup of reserves prior to migration. Data
on the potential impact of whale watching are almost entirely confined to
observations of short-term effects (minutes or hours, sometimes days) on
behavior and have been primarily carried out on baleen whales, whereas
toothed whales have been studied less often. A large number of behavioral
studies, typically unpublished reports and dissertations, have been summa-
rized by Richardson et al. (1995) and Evans (1996), from which the following
examples are taken.
In Hawaii, humpback whale, Megaptera novaeangliae, presence was in-
versely related to the amount of daily boat traffic. Short-term reactions to
small boats at distances of 500-1000 m that suggest avoidance reactions
include: (1) increased frequencies of dives started without raised flukes and
surfacings without blows; (2) reduced time spent surfacing; and
(3) movements out of favored areas. Humpbacks have also been known to
sometimes “threaten” boats by “charging” toward them and “screaming” at
them. In Alaska, humpback whales responded to vessel traffic and ap-
proaching boats at distances of several kilometers by increasing the duration
of their dives, reducing the duration of surfacing, moving away from vessels,
and temporarily abandoning preferred feeding areas. Off Cape Cod, hump-
backs responded little to boats that followed established guidelines for
whale watching by approaching slowly and steadily and keeping a minimum
distance, but rapid and close (within 30 m) approaches induced avoidance
behavior. Bowhead whales, Balaena mysticetus, and right whales Eubalaena
glacialis and E. australis fled from vessels that were noisy or approached
rapidly, but could often be approached by quiet and slowly moving boats.
The same reaction was noted in blue whales, Balaenoptera musculus, and
fin whales, B. physalus, in the St. Lawrence estuary. In the Gulf of Maine,
fin whales shortened dives and reduced the number of surface blows in the
presence of boats. Off New Zealand, some sperm whale, Physeter catodon,
individuals avoided outboard-powered whale-watching boats when ap-
proached to within 2 km by reducing the number of blows per surfacing,
the intervals between blows, and the duration of surfacing. They also
changed dive patterns and surface movements.
These studies identified many factors that influence the response of ceta-
ceans to whale watching. They include season, differences in responses due
to status (residents versus transient animals), age and sex, activity prior to
approach, group size, group composition, population, species, the number
460 HERlBERT HOFER AND MARION L. EAST
of vessels present, the speed, distance, and angle of approach of vessels,

the type of vessel, the general amount of vessel traffic, and the history of
whale watching in a particular area (Richardson et af., 1995; Evans, 1996;
Janik and Thompson, 1996). Other factors include the changing habits of
whale watchers, changes in regulations of whale watching, and a diversity
of regulations across the globe (IFAW et al., 1996). A majority of studies
have found some sort of short-term effect of boats or whale watching on
the behavior of whales, whereas a minority of studies started that they did
not detect any significant effects.
Studies of whale watching are typically shore based (Stone et af., 1992;
Janik and Thompson, 1996) or conducted from aircraft. Typically, these
studies suffer from methodological problems that need to be overcome to
make substantial progress in the future. Shore-based observations cover
only a limited amount of space; the approach and/or presence of aircrafts
as such may have significant effects on cetacean behavior (Richardson et
al., 1995). A key problem is the lack of individual identification of study
animals. Without individual identification, “control” observations of un-
known cetaceans are less valuable than comparisons of known individuals
in the presence and absence of whale-watching vessels. Experiments such
as playbacks of recorded noise (Section V,D) are of limited usefulness if
carried out in captivity (e.g., Thomas ef al., 1990), yet field experiments (e.g.,
Evans et al., 1994) are difficult to implement unless individual identification
ensures adequate “control” and “treatment” observations. If “control”
and “treatment” groups also have small sample sizes, lack of power (Section
111,DJ) will ensure that differences are unlikely to be recognized as signifi-
cant and that conclusions from studies that find no significant effects should
be considered tentative. Lack of power may also prevent a meaningful
comparison of the effect of different classes of boats that may frequent a
whale-watching area, or differences in vessel behavior, because sample sizes
are likely to be even smaller if split between vessel categories. However,
such studies are essential for understanding the impact of variations in
vessel behavior, and hence for a scientific foundation of guidelines for
whale watching.
Moreover, no study has attempted as yet to quantify potential fitness
consequences of whale watching and little is known about the physiological
consequences of anthropogenic disturbance of whales or the issue of vessel
noise masking acoustic channels important for intraspecific communication
or antipredator vigilance (Richardson et al., 1995). Reviewers of the impact
of anthropogenic disturbance on cetaceans frequently suggest that apparent
stability or increases in population size in a particular geographical area
indicate that this population must be tolerating the disturbance well and
hence individuals are unlikely to experience detrimental fitness conse-
quences. In the absence of data on individual fitness, this conclusion is

usually not justified for several reasons:
1. Because of their slow reproductive rates, many whale populations are

still believed to be recovering from excessive hunting earlier this century
(Evans, 1987). In an expanding population, a significant but moderate
negative effect from anthropogenic disturbance may manifest itself as a
reduction in population growth, rather than a population decline. Observed
population stability or increases as such are therefore unreliable indicators
of population effects of whale watching.
2. Significant anthropogenic disturbance might reduce habitat quality.
At overall high population densities low-quality habitats would contain
larger populations than expected, masking any fitness effects of a decline
in habitat quality (Section IV,D,2). Hence, population density or size can be
a misleading indicator of habitat quality, particularly of low-quality habitats.
3. Availability, size, and distance of alternative high-quality habitats and
their degree of occupancy are usually unknown but strongly influence habi-
tat choice and duration of stay (Fretwell, 1972; Bernstein et af., 1991;
Sutherland, 1996).
4. Population estimates based on censuses often have large confidence
limits, making it unlikely that changes in population size can be identified
unless they are substantial.
Caveats also apply to the issues of reactions interpreted as benign and
to behavioral changes after repeated exposure to whale watching. Ap-
proaches of boats by cetaceans are often viewed as a benign reaction, yet
it might be an aggressive response to defend resources (prey or potential
mates), which might also be disturbed by vessel presence, and be associated
with an interruption of essential activities and a physiological stress re-
sponse. Secondary detrimental effects of approaches (e.g., incidental pollu-
tion) may also be possible ( Janik and Thompson, 1996). Behavioral changes
might include habituation, a gradual decline in response to continuous or
discrete repeated stimuli (yet still be associated with detrimental fitness
consequences; see Section II,B and following discussion), and sensitization,
an increase in responsiveness under the same conditions (see any textbook
on animal behavior). Clear evidence of habituation or sensitization is scarce
because of the lack of observations of individually identified animals in
the presence of repeated or continuous stimuli (Richardson et al., 1995;
Evans, 1996).
Another problem could be the continued illegal offtake of protected
species such as humpbacks as recently as 1993 (Baker and Palumbi, 1994).
Populations that continue to experience hunting are likely to react differ-

ently to approaching whale-watching vessels than nonhunted populations
(Richardson et al., 1995; see Section V,C). It will be important to establish
the history of hunting of a population to ensure that adverse reactions t o
the presence of boats are attributed to the appropriate cause.
Some of these problems were recognized by a recent workshop on scien-
tific aspects of managing whale watching, which emphasized the importance
of (1) conducting carefully controlled long-term studies; (2) using experi-
ments to assess the impact of whale watching on the behavioral response
of cetaceans; (3) improving methods to assess the hormonal, immunological,
and physiological response to vessel presence; and (4) assessing the fitness
consequences of whale watching (IFAW et al., 1996). Whale watching is a
case where intensive efforts are under way to establish internationally
standardized guidelines and regulations even though scientific information
on the potential fitness consequences of whale watching is not yet available
(IFAW, 1996; IFAW et af., 1996).
2. Case Study 4: Antarctic Penguins and Tourism
Increasing numbers of tourists visit breeding colonies of several species of
Antarctic penguins every year ever since shipborne tourism to the Antarctic
began in 1958 (Enzenbacher, 1993). Concern about the potential impact
of human presence on breeding performance of penguins increased when
it became clear that colonies in the vicinity of Antarctic research stations
declined during the operation of these stations and recovered when opera-
tions were terminated (Culik and Wilson, 1991b). Wilson et al. (1991) used
external electrocardiogram recording devices or surgically implanted heart
rate transmitters to assess the impact of disturbance on heart rate in AdClie
penguins living in a colony close to a research station. Significant heart rate
increases (50% or more) were recorded from breeding penguins exposed to
a solitary human approaching nests or a major penguin commuter pathway,
and to low-flying aircraft approaching the colony, even when the penguins
showed no obvious behavioral signs of excitement. However, penguins also
responded behaviorally by deviating 70 m from the commuting path for
several hours after the human had left the area, fleeing from the nest, or
refraining from returning to their nest (Wilson et al., 1991). Three days of
helicopter exposure caused 80% of penguins to abandon their nests (Wilson
et af., 1991). Culik and Wilson (1991b) concluded that “tourism does ad-
versely affect breeding penguins, almost irrespective of how ‘well-behaved’
the tourists are.”
Nimon et al. (1994,1995) pointed out that the handling procedures used
by Wilson et af. (1991) may have caused the study animals to be negatively
predisposed toward humans and that the records of heart rate by Wilson
et al. (1991) reflected the effect of handling rather than the effect of distur-
bance by the approaching human or aircraft. They placed an artificial egg
with an infrared heartbeat sensor in nests incubated by gentoo penguins
in order to record changes in heart rate caused by disturbance. They ex-
pressed the hope that their minimal handling procedure would not influence
records. Their preliminary results showed that heart rates were not signifi-
cantly increased when a single experimental visitor approached slowly and
in a nonobtrusive fashion, and they concluded that tourism did not necessar-
ily cause a disturbance.
In reply, Culik and Wilson (1995) pointed out that (1) the heartbeat
records of Nimon et al. (1994,1995) were taken during incubation, whereas
those of Wilson et ul. (1991) were obtained from penguins guarding crkches;
and (2) the reaction of incubating birds are minimal compared with reac-
tions during other stages of the reproductive period and thus cannot be
extrapolated to these other stages. Also, as tourists come in batches rather
than as single visitors, the experimental setup by Nimon et al. (1994, 1995)
was unlikely to represent the situation of tourist visitors (but may well
represent the situation of researchers studying penguin colonies). They
further pointed out that there was considerable intra- and interspecific
variation in responses to disturbance and that the strength of response (e.g.,
in terms of flight distance) may depend on the general level of disturbance
experienced by a colony.
Thus, disturbance of penguin colonies sometimes had significant detri-
mental fitness consequences, behavioral responses varied between repro-
ductive stages and depended on the general level of disturbance, and re-
sponses varied between species. These results suggest that (1) more refined
methods and long-term records are required before interspecific generaliza-
tions can be safely drawn; and (2) nondisruptive behavior of visitors is a
minimal requirement to protect penguins from significant disturbance.
3. Experimental Studies of the Effects of Tourism and Leisure Activities
Several experimental studies have demonstrated that disturbance associ-
ated with tourism may change heart rate, increase energy budgets, and
decrease reproductive performance (case studies 1, 3, and 4; Sections
IV,B-D; see following discussion). However, the potential disturbance
caused by several important types of tourism, notably wildlife viewing on
safaris in eastern and southern Africa, has not been studied in this way.
Disturbance of oystercatchers, Huematopus ostrulegus, by people, artifi-
cial kites, and single-engine planes increased heart rate in incubating individ-
uals (Huppop and Hagen, 1990). Disturbance of ptarmigans, Lagopus mu-
tus, by hikers near incubating birds decreased heart rate associated with
freezing behavior (Ingold et al., 1992). Low-level disturbance during the
incubation and nestling phases increased (1) time and energy expenditure
of marsh harrier, Circus aeruginosus, parents on reproductive and nonrepro-
ductive activities, and (2) plasma urea concentrations in nestlings, suggesting
a detrimental effect on lifetime reproductive success (Fernandez and Az-
kona, 1993). Madsen (1995) reported on the results of a “natural” experi-
ment, in which farmers either disturbed or did not disturb grazing pink-
footed geese, on their subsequent body condition and breeding success.
Disturbed geese had a poorer body condition and were significantly less
likely to breed successfully. The presence and behavior of tourists at nesting
beaches of green turtles, Chelonia rnydas, reduced the arrival of females by
a third but did not influence the proportion of successful nesting, incomplete
nesting, or no nesting attempts (Jackson and Lopez, 1994). Recreational
disturbance of winter denning sites of American black bears, Ursus arneri-
canus, resulted in the abandonment of dens and cubs and a delay in entry
to hibernation (Goodrich and Berger, 1994). Experimental harassment of
mule deer by all-terrain vehicles induced reproductive pauses in the subse-
quent breeding season (Yarmoloy et ul., 1988).
Does wildlife habituate to disturbance? In many cases it does, as evi-
denced by a gradual decline in the behavioral response of disturbed wildlife
to groups of tourists (e.g., Van Heezik and Seddon, 1990). However, as
discussed in Section II,B, disturbance may continue to have detrimental
fitness consequences even when animals are habituated. Habituation cannot
be expected to occur under all circumstances, and is less likely to occur if
the disturbance involves close approaches or unusual or unpredictable
events (Huppop and Hagen, 1990). For instance, bighorn sheep, red deer,
and marmots, Marrnotu rnurrnotu, were less responsive to people on major
roads or hiking trails than to people encountered off-trial or with a domestic
dog as companion (Schultz and Bailey, 1978;MacArthur et al., 1982;Mainini
et al., 1993). However, the experiments reported in these studies in some
cases cannot exclude the possibility that individuals less sensitive to distur-
bance, for example, individuals prepared to take higher risks because they
are of below-average quality, settle closer to major roads or hiking trails.
Some studies demonstrate that animals did not habituate, even after years
of exposure, o r may even become sensitized by repeated disturbance events.
For instance, paragliding in the Swiss Alps initiated flight responses in
chamois, Rupicupru rupicupru, and Capricorn, Capra ibex, and led to changes
in habitat utilization by these species. Experimentally disturbed animals
did not show any evidence of habituation (Schnidrig et al., 1992; Ingold et
al., 1996), although differences in the routine of experimental and usual
paragliders may have prevented habituation to occur. Mountain goats,
Orearnnos urnericus, showed substantial individual differences in response
to, but did not habituate to, intense industrial exploration activity (Foster
and Rahs, 1983).
D o successful education and awareness campaigns always reduce distur-
bance of wildlife? A recent study has suggested that such campaigns may
also initiate new forms of disturbance. Initially, water-based tourism posed
the greatest threat to the survival of manatees, Trichechus manatus, because
of mortality and injuries from boat propellers (Shackley, 1992). After a
highly successful public-awareness campaign, new forms of tourism devel-
oped, including helicopter flights, canoeing, and SCUBA diving, creating
new sources of disturbance. The long-term consequences of these forms of
disturbance are not yet known.
C. HUNTING
Hunting by people is a form of human disturbance that is thought to
closely resemble predation and may sometimes even operate synergistically
with natural predation. For instance, Schauer and Murphy (1996) observed
that the rate of egg loss from nests of colonial cliff-nesting common murres
was high on days when human hunters discharged firearms and shot adults
on or near cliffs. Flushed adults often accidentally dislodged eggs, and
abandoned eggs were taken by natural egg predators such as glaucous gulls,
Larus hyperboreus. Potentially stressful impacts of hunting may be relevant
to conservation efforts for two reasons. Hunting is often considered a
recreational activity, and thus should be considered in the context of distur-
bance of wildlife by recreational activities. A second aspect is that one
school of thought proposes that conservation will work only if potential
conflicts between local communities and conservation activities can be mini-
mized. According to this idea, the minimization of such conflicts requires
the economic exploitation of wildlife because consumptive exploitation,
including sports hunting, pays local communities, and hence ensures the
conservation of wildlife populations (see Taylor and Dunstone, 1996a).
This approach encourages hunting as a method to advance conservation.
In both contexts it might be useful to know how stressful hunting is. Here
we consider three aspects of hunting that may have detrimental fitness
consequences: the physiological consequences of pursuit, modification of
behavior between hunting and nonhunting seasons, or hunting and non-
hunting areas, and the consequences of crippling.
1. Physiological and Fitness Consequences of Pursuit
Some forms of hunting, for instance riding with hounds to hunt red foxes,
Vulpes vulpes, red deer, or hares, Lepus europaeus, consist of driving the
target animals over considerable distances. We call this pursuit hunting.
Proponents of pursuit sports hunting frequently argue that there is little

difference between the pursuit of a red deer by natural predators or the
pursuit of a red deer by a staghunt. However, there has been no quantitative
investigation that compares the “stress” response of the target animal to
pursuits by predators with the response to pursuits by hunts, and there are
comparatively few descriptive or experimental studies that have explored
the physiological consequences of pursuit hunting (see following discus-
sion). Such measurements should include the hormonal, immunological,
and anatomical response to hunts as a function of the duration of a pursuit,
the distance covered and average and maximum travel speeds, and subse-
quent fitness consequences (e.g., delayed mortality). The hypothesized par-
allels between pursuits by natural predators and hunters fail for such prac-
tices as badger, Meles meles, “digging” using terriers in the United Kingdom.
Again, data are few on details of such hunting practices (for an exception,
see Griffiths, 1994) and none are available to assess the physiological re-
sponse or its fitness consequences for the hunted animals.
In an experimental study, Harlow et al. (1992) assessed the consequences
of five to six pursuit chases of cougar, Felis concofor, over the course of
one hunting season by comparing the adrenocortical response of individuals
to ACTH stimulation before and after chases at the beginning and at
the end of the hunting season. These pursuit chases were sufficient to
permanently alter the physiological response of the adrenals. It is unclear,
however, whether such changes enhanced or reduced the ability to cope
with stress and whether they would have detrimental fitness consequences.
Measurements of the adrenocortical, immunological, or anatomical re-
sponse to pursuits have not been undertaken until recently, although they
would have contributed to a more scientifically informed debate on the
welfare implications of sports hunts in the United Kingdom (Taylor and
Dunstone, 1996b). A recent 2-year study compared the adrenocortical and
anatomical response of red deer that were shot, involved in car traffic
accidents, pursued by sports hunts. Red deer that were pursued during the
course of normal staghunts were chased over average distances of 19 km
and at the end of the chase had no blood sugar left, substantial damage to
blood cells and muscles (reminiscent of capture myopathy), and significantly
higher levels of cortisol and beta endorphins than deer that were killed in
other ways (Bateson, 1997). This suggests that pursuits may cause substan-
tial damage and that deer that escaped at the end of the chase may die
from capture myopathy or other debilitating consequences of the pursuit.
Observations on several species indicate that hunts of whales are usually
associated with prolonged struggles of the pursued individual and wounded
animals may take up to 1 hour to finally succumb (Kestin, 1995). A descrip-
tive study of freshly killed fin whales and sei whales, Baluenoptera borealis,
tried to assess whether the pursuit of whales during whale-catching opera-

tions may cause “heat stress” in terms of raised core body temperatures.
It found no evidence for raised core temperatures and concluded that either
high propulsive efficiency or high thermoregulatory capacity prevented heat
stress in whales after an intensive pursuit (Brodie and Paasche, 1985).
However, as details of pursuit intensity were not available, and temperature
measurements took place some time after death and were restricted to a
few sites, generalizations from this study are probably unwarranted without
further evidence.
Pursuits of individuals for the purpose of capture may result in capture
myopathy (Section V1,C). Species probably vary substantially in their toler-
ance of pursuits before capture myopathy occurs. In the dugong, Dugong
dugon, even short pursuit distances by traditional hunting methods are
likely to lead to severe physiological problems including mortality from
capture myopathy (Anderson, 1981). Continuous pursuit of African lions,
Panthera leo, by cattle farmers with rifles, dogs, and gin traps over 2 days
led to severe pathological changes (as subsequently revealed by autopsy),
and capture myopathy with subsequent mortality when the individual was
immobilized on the third day of the hunt (Joubert and Stander, 1990).
2. Effects on Behavior and Habitat Choice
It has been repeatedly argued that hunting has a pronounced influence
on activity budgets and the reaction of wildlife to people. One hypothesis
considers hunting as just one of many forms of human disturbance and
predicts that animals should react to hunter presence and activity in a
manner similar to that for other forms of human disturbance. A second
hypothesis, however, explicitly distinguishes hunting from other forms of
human disturbance and argues that in areas with hunting or during periods
of hunting wildlife tends to be shy and is more likely to flee from any kind
of approaching human being, particularly during daytime. By contrast, in
areas without hunting or during periods of no hunting wildlife habituates
quickly to human presence and can habituate even to high densities of
visitors involved in recreational activities. Some observational, experimen-
tal, and theoretical studies have attempted to quantify the impact of distur-
bance caused by hunters and obtained variable results, but so far no rigorous
experimental tests have been carried out that were designed to distinguish
between these hypotheses.
There are several reports of marine mammals that alter their response
to other forms of anthropogenic disturbance if the population is subjected
to hunting (Richardson et al., 1995). Grizzly bear, Ursus arctos, females
avoided private lands on which they would be hunted (Mace et aL, 1996).
Madsen (1995) reported on the progress of a large-scale experiment that
aims to assess whether hunting is responsible for habitat shifts and other
behavioral changes in waterfowl. Every year experimental refuge areas
(areas where hunting was prohibited for that year) were set up in a new
location and the distribution of individuals over hunting and nonhunting
areas was subsequently observed. Both species that were hunted and species
that were legally protected significantly preferred the refuge areas. In an
observational study, Jeppesen (1987) compared the behavior of a popula-
tion of red deer between seasons of no hunting, moderate hunting, and
intensive hunting. During hunting periods red deer moved greater distances
and enlarged their home ranges. During an intensive hunting period high
tourist numbers led to deer completely abandoning one type of plantation,
a major habitat shift by the deer. In previous years high tourist numbers
were associated with only moderate hunting pressure and habitat shifts
were not observed. Jeppesen (1987) argued that this singled out hunting
as a particularly severe form of disturbance and that it was the increase in
hunting pressure that had led t o habitat shifts by the deer, but he was
unable to conclusively identify cause and effect in this case. Skogland and
Grovan (1988) demonstrated that foraging and aggregation behaviors of
reindeer, Rangifer tarandus, were significantly affected by hunting and
asked whether such behavioral changes depended on the initial body condi-
tion and the foraging needs of individual reindeer (Section IV,B,S). They
found that during the hunting season, well-fed animals aggregated into
larger groups, spent more time alert, and foraged less than during the
nonhunting period. That is, they pursued a risk-minimization tactic. Poorly
fed animals, however, pursued a risky nutrient maximization strategy by
moving more and losing more body mass during the hunting season than
well-fed animals. Frederick et al. (1987) used a stochastic simulation model
to analyze the effect of increasing hunting pressure on population size and
emigration rates of the lesser snow goose, Chen caerulescens. They found
that direct mortality from hunting had a smaller effect on population size
than the reduced energy gains and increased emigration rates of geese
caused by disturbance during feeding by hunters.
Several studies have found no effect of hunting on behavior. In an experi-
mental study, Olsson et al. (1996) compared the movements of radio-
collared willow grouse, Lagopus lagopus, in heavily hunted areas with those
in areas in which hunting was prohibited. No significant differences in
movement distances or rates, or in the likelihood of emigration from the
study area were detected. Olsson et al. (1996) interpreted this as evidence
for a predator-avoidance strategy that relies on utilizing a familiar area
with known escape sites. Kernohan et al. (1996) asked whether 24-hour
habitat use can be predicted from diurnal habitat use in white-tailed deer
during seasons of hunting and no hunting, and found that this was the case
for both types of seasons. This does not, however, preclude the possibility
that changes in activity patterns did occur between the hunting and no
hunting seasons, but if so, they did not affect the diurnal behavior of deer.
3. The Consequences of Crippling
Hunting by shooting or snaring may also cause mutilations that perma-
nently cripple the victim and generally make life more difficult. There are
few data on the proportion of animals crippled as a result of hunting in
wildlife populations, or the fitness consequences of crippling. Crippling such
as amputations and persistent injuries might increase energetic expenditure
during foraging trips or decrease foraging skills, increasing the chance of
mortality or decreasing the ability to successfully raise young.
Wheeler et ul. (1984) estimated that during periods of high-intensity
shooting of ducks in autumn in an American marsh 24-32% of ducks
were crippled, but they did not record fitness consequences of crippling.
In spotted hyenas, Crocuta crocuta, the chance of escaping from snares
set by poachers to capture herbivores was 25-64% (Hofer et al., 1993).
Individuals that successfully escaped from snare locations by biting through
the tethering wire retained snares on their bodies for variable periods.
Snare wounds may become infected and persist until they eventually kill
individuals, skin may grow over the snare and the snare may become a
permanent feature of the animal, or snare wounds lead to an amputation
of parts of an extremity. In several lactating females, snare wounds and
amputations led to an increase in suckling intervals and the loss of the litter
(H. Hofer and M. L. East, unpublished data). Clearly, such impairments
can have detrimental fitness consequences (see Hofer et al., 1996). In forest
chimpanzees, Pun troglodytes, 11 out of 34 individuals (32%)suffered injur-
ies from snares set for bushbuck, Trugeluphus scriptus crippling or amputat-
ing one or both hands a n d o r feet. Observations of foraging groups showed
that disabled individuals were well integrated into the groups (Quiatt et
ul., 1994). Chimpanzees suffering from snaring injuries were typically lower
ranking and fed at greater heights but achieved the same feeding rate in
fig trees (Smith, undated). Fitness consequences of these injuries have not
yet been measured in chimpanzees.
D. Noise
There is an increasing recognition that environmental noise, particularly
noise generated by human activities, is harmful not only to humans (Clark,
1992) but also to wildlife (Fletcher and Busnel, 1978). Sources of environ-
mental anthropogenic noise are flying devices, including hot-air balloons,
low-flying military and civilian aircraft and helicopters, road, boat, and
vessel traffic, terrestrial and marine seismic exploration and geophysical
surveys, explosions, marine sonars, terrestrial and marine resource extrac-

tion including mining, oil and gas drilling and production, dredging and
construction, and recreational activities including wildlife viewing and the
chatter emanating from safari vehicles. Noise can influence animals over
distances of up to 30 km (Cosens and Dueck, 1993) and impact the entire
animal community. Ideally, carefully designed and executed experimental
studies on wildlife assess the response to and identify the fitness conse-
quences of anthropogenic noise, and separate the effect of noise as such
from other aspects (e.g., visual cues) of disturbance. However, compara-
tively few experimental studies have considered these issues and most
information comes from observational studies. These either provide evi-
dence of the detrimental effects of noise or demonstrate that a reduction
of noise levels had a positive effect, but often cannot separate the influence
of noise from that of other factors. For instance, the decrease in breeding
density of hazel hen, Bonasia bonasia, in partly urbanized areas was sus-
pected to be due to a permanent increase in background noise level, which
may have reduced the chance of individuals to hear alarm calls, thereby
increasing mortality from predation (Scherzinger, 1979), but appropriate
experimental tests were not carried out.
Examples of experimental evidence for the beneficial effects of noise
reduction come from studies in captivity. Quiet handling, based on an
understanding of behavior, reduced excitement and the incidence of injuries
in domestic cattle (Grandin, 1987). Fear-related behavior and heart rate of
red deer housed indoors were reduced by sound-proofing housing facilities
(Price et al., 1993). Examples of experimental evidence for the detrimental
effects of noise include short startle noises used to induce “prenatal stress”
in captive, nonhuman, pregnant, female primates after removal from their
home cage. These studies produced significant effects of prenatal stress
on the behavior and immunocompetence of infants (Lubach et af., 1996;
Schneider 1992a,b,c; Clarke & Schneider, 1993; Clarke et al., 1994), but did
not separate the effect of removing the pregnant female from her home
cage from that of noise itself.
The hormonal or physiological response of wildlife to noise is not well
researched. The effect of noise on blood pressure and sleep (known to be
detrimentally affected in humans, Berglund et al., 1990) has to our knowl-
edge not yet been studied, nor has the effect of human disturbance and
noise on heart rate of marine mammals (Richardson et al., 1995). Noise
from car traffic increased heart rate and thus may affect birds breeding
close to highways (Helb and Huppop, 1991). In free-ranging mountain
sheep, low-flying aircraft and vehicle traffic produced no overt behavioral
response and a modest increase in heart rate (modest compared to the
increase observed after the animal was handled for instrumenation) (Mac-
BIOLOGICAL CONSERVATION A N D STRESS 47 1
Arthur et al., 1982, 1986). Harrington and Veitch (1991) noted that the
sound rather than the visual appearance of low-flyingjet aircraft was respon-
sible for startling Alaskan caribou, Rangifer tarandus. In a partly experi-
mental study, Harrington and Veitch (1992) demonstrated a significant
negative correlation between the frequency of exposure of Alaskan caribou
to low-level jet overflights during the calving and postcalving season and
subsequent calf survival. In an observational study of 43 species of
woodland-breeding birds, Reijnen et al. (1995) identified noise emanating
from car traffic along main roads as the key variable that depressed breeding
densities over distances of 100-1500 m away from main roads, when the
results were controlled for other potentially confounding factors. This con-
firmed a single-species study of the willow warbler, Phylloscopus trochilus,
that suggested that the noise emanating from car traffic along main roads
prevented males close to a highway from attracting or keeping females
(Reijnen and Foppen, 1994); as a consequence these males moved away
from the road during the following breeding season. Short-term seismic
exploration and timber harvest in an area with few permanent developments
and low levels of vehicular traffic caused little overt response and minimal
changes in habitat utilization by grizzly bears (McLellan and Shackleton
(1988, 1989). However, in an area with a higher intensity of vehicle traffic,
Mace et al. (1996) found that grizzly bears avoided buffer zones of 500-m
width surrounding roads with a traffic intensity of more than 10 vehicles
per day, whereas their response was neutral to roads with a traffic intensity
below that.
The literature on the effect of anthropogenic noise on marine animals
has been reviewed by Myrberg (1990), Richardson et al. (1995), and Evans
(1996). Most studies recorded the behavioral response of mammals and
fish to offshore petroleum exploration and production, using observations
and playback experiments. Intensity of response depended on noise level,
activity at the time of exposure to human-made noise, and prior experience.
Where such studies combined playbacks with measurements of physiologi-
cal variables, sample sizes are so low that the power (Section II1,DS) to
identify significant effects is very low (e.g., Thomas et al., 1990). However,
few quantitative data are available on the hearing ability of most marine
animals, their physiological response to sounds of different intensities, fre-
quencies, and durations, or the fitness consequences of such noise (Richard-
son et al., 1995). Some evidence on detrimental fitness consequences comes
from several species of finfish and shellfish where high intensity of anthropo-
genic noise caused abnormal growth and reproductive processes (Myrberg,
1990). We conclude that noise is currently suspected to be a potential
stressor, but there is little factual evidence on its impact on animals, or on
how noise management can improve the success of conservation activities.
E. CLIMATIC
WARMING
Climatic warming is likely to have a pronounced effect on many popula-
tions and species (Peters and Lovejoy, 1992; Kareiva et al., 1993) because
(1) temperature fluctuations as well as the mean ambient temperature
are key factors that determine the survival of individuals (Hoffmann and
Parsons, 1991), and (2) climatic warming is also associated with other forms
of environmental stress, particularly atmospheric pollutants, that are likely
to operate in a synergistic fashion (Parsons, 1990~).A rigorous experimental
test of the impact of global warming on fitness is not feasible, so that
predictions of the impact of global warming frequently rely on measure-
ments of temperature and pollutant tolerance in single species or the model-
ing of extinction probabilities of particular populations in response to envi-
ronmental change. Of particular value will be observations of species that
are at the top of the food chain in an ecosystem (Stirling and Derocher,
1993), and of the changes in the structure of interactions in community
food webs (Bodini et af., 1994).
Harsh temperature conditions may have a profound influence on the
survival of populations or whole species (case study 1) because temperatures
may simply be too hot or cold to ensure individual survival. On a more
moderate level, harsh temperatures may be responsible for a decline in
immunocompetence (e.g., Lamontagne et af.,1989) or fertility (Krebs and
Loeschcke, 1994). Experimental evidence is accumulating that changes in
average temperatures of as little as 1-2°C may be sufficient to cause popula-
tion or species extinction (Parsons 1989b; Baur and Baur, 1993), and obser-
vational evidence indicates that short-term regional climatic changes such
as the El Niiio of the southern Pacific Ocean initiate profound changes in
biological communities (Arntz and Fahrbach, 1991; Trillmich and Ono,
1992).
For conservation purposes, the key question will be whether populations
can tolerate climatic changes or whether adaptive evolution in response to
climatic change can be quick enough. There is some evidence from genetic
studies suggesting that fast evolution of stress tolerance is possible (Hoff-
mann and Parsons, 1991), but the available evidence is insufficient to predict
the outcome of climatic change on specific populations or species (Hoff-
mann and Blows, 1993). Genetic studies have demonstrated that there is
a genetic basis for heat resistance, that some populations improve stress
resistance by increasing the plasticity of phenotypes, and that there is
variation in the genetic basis of heat resistance between populations of a
species (Hoffmann and Blows, 1993). The study of the evolutionary genetics
of stress resistance in marginal populations will help answer questions about
the impact of climatic warming (Parsons, 1990~).
VI. CONSERVATION
RESEARCH A N D MANAGEMENT
AS STRESSORS
ACTIVITIES
Both research and management activities may be a source of stress.

Failure to recognize this problem often reduces the value of a study or a
conservation activity, makes the subjects experience more hardship than
necessary, and prevents the design and implementation of procedures that
minimize the occurrence, magnitude, and consequences of stress. Even if
ecological field work is not explicitly designed as a manipulative experiment,
field studies may routinely cause disturbance because of the presence of an
observer or specific interventions involving capture, handling, and marking.
Whether an intervention causes discomfort, distress, or a reduction in fitness
ought to be the subject of careful scientific evaluation. The widespread
tacit assumption that interventions are always benign, or least do not influ-
ence the results of a study, is not justified, as careful experiments have
demonstrated (see following discussion). However, interventions do not
necessarily have detrimental consequences, and as there is currently no
general theory available that predicts the circumstances under which inter-
ventions may be harmful, the impact of interventions ought to be evaluated
on a case-by-case basis. In most studies reviewed later in this chapter that
report the absence of an intervention effect, the power of tests (Section
II,D,5) was not evaluated, and if it was evaluated it was low, so the failure
to reject the null hypothesis of intervention having no detrimental effects
must be considered tentative in each case.
A. LITTLEATTENTION
HASBEENPAIDTO THE IMPACT
OF INTERVENTION
AND CONSERVATION
ACTIVITIES
Many field and laboratory studies and conservation activities use various
forms of intervention in field experiments (capture, handling, blood samp-
ing, radio-collaring, vaccination, and manipulation of populations or aspects
of their environment). Compared with the number of cases where such
procedures are employed, it is surprising how little attention has been
paid to the ethical and conservation implications of interventions and field
experiments (Cuthill, 1991; Putman, 1995). Handling techniques are known
to provoke a strong physiological response, particularly in the case of
nonacclimated animals (Gartner et al., 1980; Pottinger and Calder, 1995),
that might bias or invalidate physiological results. Despite this, little atten-
tion has been paid to such biases in several disciplines, including physiology
and toxicology (Rowan, 1990; Pottinger and Calder, 1995). Stress-sensitive
physiological data from blood samples were analyzed in 58 out of the 397
publications on macaques surveyed by Reinhardt (1991a), yet 81% of the
studies did not provide any information as to how subjects were caught or
immobilized for blood sampling. Common practice suggests that the animals
were physically restrained with squeeze-backs or forced with fear-inducing
techniques to leave their home areas and enter a transport cage. Such
methods of enforced restraint result in significantly increased adrenal activ-
ity and significant changes in other physiological parameters (Reinhardt et
al., 1995), suggesting that the effects of handling may have biased the results
of many of these studies.
B. CENSUSING THE TARGET

MAYDISTURB POPULATION
AND BIAS
CENSUSRESULTS
Most studies that estimate population size or habitat use assume that
the censusing process does not disturb populations and thereby influence
censusing results. Several studies have tried to experimentally test this
assumption. Dufour et al. (1993) tested the idea that baited traps preferen-
tially attract mallard individuals in bad condition and that such an effect may
change with sample size. They found that there was some trap selectivity but
that it was unlikely to vary with sample size. Bleich et al. (1994) investigated
the effect of helicopter surveys on movements of mountain sheep, Ovis
canadensis, across habitats and sampling blocks. Surveys increased move-
ments between sampling blocks and altered habitat use, potentially increas-
ing “nutritional stress” or susceptibility to predation. Bleich et al. (1994)
suggest that with significant disturbance effects censuses are likely to violate
fundamental assumptions of population estimators. Mallet et al. (1987)
assessed disturbance effects from netting or handling of butterflies Helicon-
ius sp. in mark-recapture studies. They concluded that disturbance caused
mark-recapture estimates to be so inaccurate that the capture of almost
all individuals of a population is required, making the use of the Lincoln
Index unnecessary. Similarly, cane toads, Bufo marinus, were highly sensi-
tive to disturbance by trapping or handling, reducing the chance of recap-
ture, and making mark-recapture estimates unreliable (Lampo and Bayliss,
1996). Other forms of population monitoring require visits to breeding or
nesting sites. Such visits may or may not influence breeding success. Check-
ing of nesting burrows of a Canadian population of Atlantic puffins, Frater-
cula arctica, by researchers caused birds to abandon nests and reduced
breeding success by more than a third (Rodway et al., 1996). Visits at
different rates by researchers to nesting sites of rock ptarmigans, Lagopus
mucus, however, did not influence predation rate, clutch size, nesting, fledg-
ing, or hatching success (Cotter and Gratto, 1995).
These experimental studies suggest that population censusing may some-
times stress target populations and that studies should assess the potential
influence of the censusing process on results. As population size estimates

are a vital ingredient of conservation activities, potential biases in these
estimates caused by the censusing procedure may have profound conse-
quences.
A N D CONSERVATION
C. INTERVENTIONS ACTIVITIES
ARE
SOMETIMES
STRESSORS
The possibility that interventions and conservation activities may them-
selves be the source of stress is frequently ignored. Retrospective, descrip-
tive comparisons of handled versus unhandled individuals are often the
only way to analyze data in the observational studies when there is a
suspicion that handling might have a significant impact on fitness. Such
studies may provide important hints and suggest factors responsible for
handling effects. However, descriptive retrospective studies sometimes use
small sample sizes that imply low statistical power and make it unlikely
that a significant difference between handled and unhandled animals could
be identified if there really was one; other caveats may apply (Section
111,D). To overcome these problems, experimental studies of intervention
are required in which subjects are assigned randomly to experimental and
control groups (Section 111,DJ). Observational studies may sometimes
meet these criteria if it can be argued that the selection of individuals
for interventions had proceeded in a random way. The following sections
emphasize results from careful experiments that in many cases measured
fitness consequences of interventions. When considering the results of these
studies it is important to remember that some measures of fitness are short-
term ones that may not necessarily predict long-term effects.
I. Capture, Measurements and Palpation, and Blood Sampling
Because a typical handling event usually consists of capture, restraint,
examination, measurements, and blood sampling of an animal, these compo-
nents of handling will be considered together. Capture, measurements, and
blood sampling may sometimes cause a substantial physiological response.
An experimental study of captive domestic geese demonstrated that weigh-
ing, injecting, and blood sampling disrupted the acid-base balance and
caused a dramatic increase in the level of humoral indexes of stress (cate-
cholamines, corticosterone, and lactate) within 2 min (Le Maho et al., 1992).
Routine restraint for 1 h, handling, and examination in captivity (including
measuring and weighing) provoked substantial increases in the level of
glucocorticoids in several species of bats (Widmaier et al., 1994). “Unpleas-
ant” handling of captive domestic pigs alone or in groups reduced growth
rate and feed conversion efficiency and led to higher glucocorticoid levels
in individually handled animals (Hemsworth and Barnett, 1991). Handling

captive rats and domestic cattle for blood sampling when animals were not
acclimated to the procedure raised plasma concentrations of cortisol and
other hormones, substantially changed heart rate, and depressed tonic lu-
teinizing hormone secretion (Gartner et al., 1980; Echternkamp, 1984).
The potential fitness consequences of capture, basic body measurements,
and blood sampling has been repeatedly examined by retrospective descrip-
tive studies in many species. Research activities that included handling and
ringing are now considered a major factor responsible for high mortality
and the decline in population size in many European species of cave bats
in the 1970s (Gaisler et al., 1981). In birds, where handling effects have been
most intensively studied, detrimental fitness consequences of interventions
were sometimes recorded in retrospective studies; these, and other studies
that found no detrimental consequences, were reviewed by the American
Ornithologists’ Union (1988) and Kania (1992).
Several experimental studies show that capture, measurements, and
blood sampling had no fitness consequences in the wild. In the red-winged
blackbird, Agelaius phoeniceus, capture and blood sampling did not affect
migratory behavior, annual return rates, chance of territory loss, or repro-
ductive success in the field and did not affect mass changes in captive birds
(Hoysak and Weatherhead, 1991). In the endangered Chatham Island black
robin, Petroica traversi, capture in traps, measuring, and blood sampling
had no detrimental effect on behavior and adult survival for one year
(Ardern et al., 1994). In the red-cockaded woodpecker, Picoides borealis,
sampling blood and feather pulp from nestlings had no effect on nestling
survival (Stangel and Lennartz, 1988). Capture, handling, and blood sam-
pling of white-crowned sparrows, Zonotrichia leucophrys, during the breed-
ing season did not affect survival, migration, annual return rates, chance
of territory loss, or reproductive success (Wingfield and Farner, 1976). The
behavioral ecology literature contains many examples in which manipula-
tions of study animals were carried out, and controls that assessed the
effects of handling as such found no effects on fitness.
Some experimental studies have also found significant effects of handling
on fitness. In the semipalmated sandpiper, Calidris pusilla, the spotted
sandpiper, Actitis macularia, the red-necked phalarope, Phalaropus lobatus,
and Wilson’s phalarope, Phalaropus tricolor, capture and blood sampling
during different stages of the breeding season caused little mortality, but
behavioral responses with fitness consequences varied across reproductive
stages and among species with different mating systems (Colwell et al.,
1988). Parental desertion was least common in uniparental species. Wilson’s
phalarope was more likely to desert when captured during the laying stage.
Incubating semipalmated sandpipers were more likely to desert if both
parents were bled rather than if one or neither of the adults was bled. Of
the previously mentioned experimental studies, only Colwell et al. (1988),
Hoysak and Weatherhead (1991), and Ardern et al. (1994) used experiments
to study both changes in behavior and potential fitness consequences of
capture, handling, and blood sampling in birds.
As previously mentioned, a well-known detrimental consequence of cap-
ture and handling of animals is the phenomenon of capture myopathy
(Chalmers and Barrett, 1982). Death due to capture myopathy has been
described for birds (Spraker et al., 1987; Dabbert and Powell, 1993), marine
mammals (Anderson, 198l), and several orders of terrestrial mammals
including marsupials (Shepherd, 1986), ungulates (Kock et al., 1987a; Be-
ringer et al., 1996), primates (Harthoorn, 1976), and carnivores (Joubert
and Stander, 1990). Sometimes the animal may survive but suffer from
chronic pathological consequences for months to years (Kock et al., 1987b).
It is difficult to predict which species are likely to suffer from capture
myopathy. Australian dugongs, Dugong dugon, are large, marine mammals
that are considered to be susceptible to capture myopathy and die within
hours of a pursuit (Anderson, 1981), whereas in a large sample of captured
American manatees, a closely related and ecologically very similar species,
there was no evidence of capture myopathy (Oshea et al., 1985). Capture
myopathy can sometimes be cured or avoided (Harthoorn et al., 1974).
2. Manual Restraint
Data on the impact of manual restraint are largely restricted to physiologi-
cal responses. A comparison of the adrenocortical response to capture by
physical restraint with that to chemical restraint (chemical immobilization)
in 18 mammalian species suggested that manual restraint was less stressful
than chemical immobilization because cortisol levels rose more substantially
after chemical immobilization (Morton et al., 1995). Traditional involuntary
restraint techniques in studies of nonhuman primates are an intrinsic source
of distress causing fear, resulting in significantly increased adrenal activity
and significant changes in a variety of other physiological parameters (Rein-
hardt et al., 1995). In mice, short-term restraint reduces immunocompetence
in response to a herpes simplex virus infection (Bonneau et al., 1991).
Short-term restraint of domestic sheep led to high increases in epinephrine,
norepinephrine, cortisol, glucose, and free fatty acids (Niezgoda et al., 1993).
In Wied’s black-tufted ear marmoset, Callithrix kuhli, isolation in captivity
followed by short mammal restraint provoked a significant adrenocortical
response, as measured by urinary cortisol (Smith and McGreer-Whitworth,
1996). The response was sensitive to subtle changes in stressor severity in
a dose-dependent manner. “Gentle” handling and manual restraint for
brief periods did not appear to cause chronic stress, as measured by plasma
corticosterone levels, in the ball python, Python regius, or the blue-tongued

skink, Tifiqua scincoides. Restraint of pythons in a container, however,
resulted in short-term elevation of corticosterone but not in behavioral
changes (Kreger and Mench, 1993). Manual restraint thus may produce a
significant physiological response, but as yet little is known about its fitness
consequences.
3. Chemical Immobilization and Anesthesia
Chemical immobilization has often been mistakenly thought to cause
little or no stress if an animal remains quiet and shows no behavioral signs
of excitement (e.g., Creel, 1992). We have already pointed out that behavior
is often unreliable as an indicator of a stressed state (Section IV,DJ).
“Stress” in the sense of a significant physiological, hormonal, or immunolog-
ical response to chemical immobilization results from disorientation before
unconsciousness and does not require a display of behavioral excitement
(Sapolsky, 1982). In the gray wolf immobilization and anesthesia caused
significant reduction in heart rate and hypertension (Kreeger et al., 1987).
In captive rhesus monkeys capture, injection, and disorientation prior to
anesthesia were responsible for increased ACTH and cortisol concentra-
tions (Clarke et al., 1994). Tethering, sedation, surgery, and chronic cathe-
terization increased cortisol levels in long-tailed macaques (Crockett et al.,
1993). Handling, ether vapor anesthesia, and blood sampling consistently
increased serum luteinizing hormone and prolactin concentrations in male
rats (Euker et al., 1975). Handling, ether anesthesia, and cardiac puncture
induced significant but variable elevations of serum prolactin in female
golden hamsters but not in males (Matt et al., 1983), suggesting that the
response to chemical immobilization may be sex specific. Because chemical
immobilization can be presented as a standardized stimulus, it has been
extensively used as an experimental paradigm to understand the factors
that mold the adrenocortical response (Sapolsky, 1997).
There is now also considerable evidence that chemical immobilization
and anesthesia may compromise the immune system, causing a deficient
cellular response and a delay and depression in antibody production in
response to an infection (Ozherelkov et af., 1990 Kramskaya et af., 1991;
Pokhil’ko et al., 1995) or vaccination (Mayr et al., 1990), with detrimental
fitness consequences in terms of increased mortality (e.g., Hansbrough et al.,
1985). Even single events of anesthesia may depress cellular and antibody
immunity, increasing susceptibility to infection (Thomas et al., 1982; Hans-
brough et al., 1985; Felsburg et al., 1986). Such effects may be stronger and
more important in wild than in captive individuals, which suggests that
testing a procedure with captive animals may be of limited value. For
instance, in an experimental study with coyotes, Canis Iatrans, Smith and
Rongstad (1980) showed that glucose levels and leukocyte counts after
capture, handling, immobilization, and blood sampling were significantly
higher in wild compared to pen-raised or captive animals.
The effects of chemical immobilization plus radio-tagging may depend
on life-history stage or social competition. African wild dogs that were
immobilized and radio-tagged prior to emigration survived significantly
longer than individuals immobilized and radio-tagged after they had re-
cently immigrated into a new group (Burrows et al., 1994). Dominance
struggles of immigrant African wild dogs with residents or among them-
selves may entail injuries, while they are not overt among animals prior to
dispersal. Immobilization during immigration may have exacerbated the
decrease in immunological competence associated with immigration (Al-
berts et al., 1992; Sapolsky, 1992).
4. Measurement of Body Temperature
Even such a simple procedure as measuring body temperature by a rectal
probe may have a measurable impact on the organism. A comparison of
body temperatures of laboratory rats measured by rectal probe (requiring
handling of the animal) and those measured by implanted telemetric devices
(recorded by remote control) usually reveal average discrepancies of ap-
proximately 1°C. A careful experimental study by Dilsaver et al. (1992)
measured rectal temperature and telemetered core body temperature in
the same individuals. They demonstrated that the higher body temperatures
recorded by rectal probes were due to the handling necessary for the rectal
probe. Other factors that influence temperature measurements in group-
housed laboratory mice include the sequence in which individuals are mea-
sured and the interval between measurements. Mice measured later in the
sequence had higher temperatures and the percentage of mice showing
hyperthermia increased with increasing interval between subsequent mea-
surements (Zethof et al., 1994).
5. Surgery
Surgery, often accompanied by anesthesia, changes the metabolism of
animals in that after the operation urinary and plasma cortisol levels, blood
pH, partial oxygen pressure, glucose conservation, and lipolysis may be
detrimentally affected (e.g., laboratory rat: Schofield et al., 1986; red fox:
Kreeger et al., 1990; Pacific oyster, Crassostrea gigas: Jones et al,, 1993;
long-tailed macaque: Crockett et al., 1993). Anesthesia and surgical trauma
may also be responsible for immunosuppression with its potentially negative
fitness consequences (e.g., Medleau et al., 1983).
The fitness consequences of surgery, including muscle biopsies, for wild
animals have been rarely considered or subjected to experimental investiga-
tion. In part, this is because surgery is often an emergency procedure, as

in the case when animals become entangled in objects, or when it is likely
that the animal will die or suffer if nothing is done. However, some proce-
dures require routine rather than emergency surgery. These include muscle
biopsies for sampling genetic material or the implantation of radio transmit-
ters and other devices that in the course of research projects collect data
from the animal’s body.
In some species, surgery and radio implantation may cause substantial
postsurgery mortality. Trials with implanted American river otters, Lutra
canadensis, illustrated the importance of diet and extended periods of post-
surgery recovery time in appropriate holding pens (Woolf et al., 1984;
Hoover et al., 1985). Even under improved conditions, 30% of otters died
during the postsurgery holding period. Mortality within 12 months of radio
implantation of African wild dogs in Kruger National Park was more than
twice as high as mortality of unhandled African wild dogs (East, 1996).
These results contrast with numerous studies that found no effects of surgery
and implantation of radio transmitters on subsequent survival or behavior
(white-footed mouse, Peromyscus leucopus: Smith, 1980; canids: Green et
al., 1985; yellow-bellied marmot, Marmota flaviventris: Van Vuren, 1989;
armadillo, Dasypus novemcinctus: Herbst, 1991;ring-necked pheasant, Pha-
sianus colchicus: Ewing et al., 1994).
The effect of muscle biopsies on fitness in birds varies between species.
In white-throated sparrows and Indigo buntings, Passerina cyanea, muscle
biopsies did not affect body condition or survival of either overwintering
or breeding individuals (Westneat, 1986; Westneat et al., 1986). In contrast,
muscle biopsies of white ibises, Eudocimus albus, led to complete nest
desertion (Frederick, 1986). Frederick (1986) could show that this effect
was not due to the procedures of capture or handling as such, as ibises
that were captured, handled, and blood sampled, but not biopsied, were
not affected.
6. Tagging and Radio- Tagging
For individual identification purposes, individuals are commonly tagged
with plastic ear tags, neck or foot rings, freeze-branded, or hot-branded.
The fitness consequences of such tags and improvements that may minimize
such consequences have been reviewed repeatedly (see Putman, 1995). One
key issue that is often neglected is that social relationships and social
success may be permanently affected by tagging (because of the associated
disturbance, human smell, etc.) or by the quality of the tags themselves.
For instance, radio-tagging mule deer fawns may result in females rejecting
their fawns, which in turn reduces the fawn’s chance of survival (Goldberg
and Haas, 1978). Individually distinct color-banding may permanently affect
attractivity and reproductive success of individuals and shift female mate

preferences. Males ringed with attractive color-bands were preferred over
males without any color-bands or males with unattractive color-bands, as
experiments on finches have shown (Burley et al., 1982; Burley, 1986,1988;
Johnson et al., 1993). Providing birds with one ring or band or mammals
with one ear tag may render their appearance asymmetrical. Mdler (1992)
and others have shown that asymmetrical traits can reduce individuals’
attractiveness to mates.
The short-term physiological response to tagging an animal with a radio
transmitter may be pronounced. In bighorn sheep, trapping, handling, and
radio instrumentation caused a sustained increase in heart rate for 2 h after
release and a cardiac recovery time ten times longer than the maximum
recovery time recorded for any anthropogenic disturbance to which individ-
uals were subsequently exposed (MacArthur et af., 1986).The fitness conse-
quences of radio transmitters attached by some form of harness or collar
have been more frequently studied by rigorous experiments than any other
aspect of intervention. Numerous studies have demonstrated that transmit-
ters may or may not change behavior or foraging success, or reduce the
chance of breeding, survival, nesting success, or the chance of predation
(reviewed by Kenward, 1987; White and Garrott, 1990. Are there any rules
that predict under what circumstances a significant detrimental effect of
attached transmitters is likely to occur?
Most studies have been concerned with the impact of radio transmitters
on birds, even though there are many more wildlife radio-tracking studies
of mammals than of birds (White and Garrott, 1990). White and Garrott
(1990) argued that studies on the consequences of attaching transmitters
were more likely to be undertaken if the species was small and depended
on flight because of the frequently voiced concern that relative transmitter
weight (RTW) may determine whether or not attachment of a transmitter
has detrimental fitness consequences. Consequently, little attention has
been paid to the effect of transmitters on large carnivores or ungulates
where RTW is well below 1% of body weight (BW) and where it is conven-
tionally assumed that transmitters have no detrimental effects (White and
Garrott, 1990). There has been no literature survey that looked at the
relationship between RTW and of the likelihood of detrimental fitness
consequences. Such a survey would have to tackle the issue that if studies
were more likely to be undertaken if a problem was suspected in the first
place, then the results of the survey might be biased. We therefore prefer
to review experimental studies that looked at the effects of RTW within
a species.
Greenwood and Sargeant (1973) used three different weight classes of
transmitters and showed that captive blue-winged teals, Anus discors, lost
more body weight as RTW increased, whereas there was no relationship

between RTW and body weight loss in captive mallards; all three weight
classes had a significant impact. Houston and Greenwood (1993) also looked
at three weight classes in captive mallards and could discern no effect of
transmitters in general, and no effect of RTW on a variety of fitness mea-
sures including clutch size and nesting interval, but the power (Section
111,DS)of their comparison was low. However, Pietz et af. (1993) and
Rotella et af. (1993) found a significant effect of transmitters on fitness
measures in wild mallards, and Rotella et af. (1993) found that this effect
varied significantly between three different types of transmitters. Amlaner
et af. (1979) also used three weight classes to demonstrate that the survivor-
ship of clutches of herring gulls, Larus argentatus, declined as RTW in-
creased. Warner and Etter (1983) demonstrated that female longevity of
ring-necked pheasants declined with increasing RTW. In AdClie penguins
the length of foraging trips and the incidence of nest desertion increased
with increasing volume of fitted measurement devices (Wilson et af., 1989).
Cotter and Gratto (1995) showed that male rock ptarmigans, Lagopus
mutus, with light transmitters (2.3% BW) had the same survival rate as
unmarked males, whereas males with heavy (3.6% BW) transmitters had
significantly lower survival than unmarked males.
Can these results suggest a rule of thumb for an acceptable RTW? Small
animals, such as the greater horseshoe bat, Rhinolophus ferrumequinum,
at 15-30 g body weight, can carry transmitters weighing 12% BW without
experiencing detrimental fitness consequences (Stebbings, 1982). Larger
animals (above 50 g body weight) should not carry transmitters with an
RTW above 4-6%. A sensible limit for the largest birds and mammals is
probably 1-2% or less (see Kenward, 1987). Even if transmitter weights
stay below these limits, it is unwise to assume that just because a species
may be large transmitters are unlikely to reduce fitness. For instance, in
chinstrap penguins, Pygoscefis antarctica, radio-tagged adults were more
likely to abandon nesting attempts than were controls (Croll et af., 1996),
and in mule deer, instrumentation was observed to reduce fitness (Goldberg
and Haas, 1978; Garrott et af., 1985).
Other aspects of radio-tagging may decide whether detrimental fitness
consequences are likely to occur. An example is the timing of radio-tagging
relative to the reproductive phase of individuals. The general recommenda-
tion is to avoid tagging animals during their reproductive period when they
appear to be sensitive to disturbance (White and Garrott, 1990). However,
wood ducks that were radio-tagged before nesting started were less likely
to start incubation than were controls, whereas wood ducks radio-tagged
during incubation did not differ in fitness measures from controls (although
the power of these comparisons was low; Gammonley and Kelley, 1995).
Radio-tagging can also cause sex-specific fitness effects. Female kangaroo
rats, Dipodomys merriami, reduced excursions from their burrow for the
first few nights after radio-tagging and suffered only 47% of predation-
related mortality compared to radio-tagged males, which did not reduce
their movements (Daly et al., 1992).
7. Vaccination
Several observational and experimental studies in humans and domestic
animals suggest that stress may modify vaccination success and be responsi-
ble for vaccination failure. Anesthesia and surgery after exposure to rabies
virus has been held responsible for vaccination failure (death of the patient)
in humans (Fescharek et aL, 1994). A study of elderly humans demonstrated
that a chronic stressor such as the duty of caregiving for a spouse with a
progressive dementia was responsible for a downregulation of antibody
response to vaccination against influenza virus (Kiecolt-Glaser et al., 1996).
Cattle that were vaccinated with a dead vaccine against the IBWIPV virus
excreted virus after they were subjected to confinement and transport (Frer-
king et a/., 1995). An experimental study showed that injection of chickens
with corticosterone after vaccination against Marek’s disease virus and a
new challenge by the virus significantly increased the incidence of Marek’s
disease in the vaccinated chickens and downregulated their immune re-
sponse (Powell and Davison, 1986). After vaccination against caprivox
virus, first-calf lactating cows suffered a decrease in milk production and
severe generalized skin lesions from which the virus could be isolated,
whereas nonlactating cattle did not develop any reactions (Yeruham et
al., 1994).
There have been no experimental studies of the impact of vaccination
on wildlife populations. However, observational studies of both captive
and free-ranging wildlife populations demonstrate that it is unwise to as-
sume that vaccinations cannot make things worse. Examples are vaccination
failures resulting in the vaccine-induced death of captive, endangered black-
footed ferrets (Carpenter et al., 1976), captive African wild dogs (Durchfeld
et al., 1990), and lesser pandas, Ailurus fulgens (Bush et a/., 1976), all
vaccinated against canine distemper virus. A retrospective observational
study of the life expectancy of African wild dogs in both the Serengeti and
Mara ecosystems, after vaccination against rabies with vaccine-filled dart
syringes, demonstrated that vaccinated individuals survived for a signifi-
cantly shorter period than radio-collared ones, and animals either vacci-
nated or radio-collared were less likely to survive for 12 months than
unhandled animals (Burrows et al., 1994, 1995).
8. Captivity, Housing, and Enrichment
For the purpose of biological conservation, several aspects of captive
housing are important. Housing ought to provide adequate standards of
welfare (Fraser and Broom, 1990; Broom, this volume), and provide an
appropriate environment for successful ex situ conservation activities, prin-
cipally breeding. Poor housing and a boring environment are responsible
for permanent changes in behavior (Mason, 1991) and physiology. For
instance, solitary housing of vervet monkeys, Cercopithecus aethiops, caused
a withdrawal response and was associated with permanent adrenal changes
and an increase in the incidence of gastric ulcers with subsequent mortality
(Tarara et al., 1995). Stress from crowded captive conditions was a possible
primary cause of high chick mortalities in Cape francolins, Francolinus
capensis (Hey1 et al., 1988). In pinnipeds, captivity or environmental stres-
sors may cause a potentially fatal sodium imbalance caused by exhaustion
of adrenal hormone reserves or desensitization of the cortex to other physio-
logical stimuli (St. Aubin and Geraci, 1986). In zoos, one neglected aspect
is the effect of the behavior of visitors on zoo animals when visitors inter-
acted with zoo animals, which could be modified by appropriate housing
facilities (Nimon and Dalziel, 1992).
There have been strong efforts to improve captive housing by methods
summarized as “environmental enrichment.” Environmental enrichment
is the provision of objects in captive housing that increases spatial heteroge-
neity, structural complexity, facilitates an increased variety of activities,
and creates different sites of shelter (Markovitz, 1982; Chamove, 1989). A
key assumption of environmental enrichment is that an improvement of
housing conditions automatically leads to a reduction of stress and an
improvement in terms of animal welfare. The measurement of the hormonal
and immunological response of captive organisms to enriched housing facili-
ties and their fitness consequences is still in its infancy but has already
yielded some surprises (Section IV,C,2). Some studies have demonstrated
that enrichment causes a reduction in cortisol levels (e.g., Chamove, 1988).
However, results of studies of species held in groups suggest that the impact
of enrichment may depend on the social organization of the species, for
example, the presence or absence of territoriality (Section IV,C,2). This
suggests that enriched environments cannot automatically be considered
less “stressful” than conventional environments and that enrichment is
unlikely to achieve its aims unless it takes the social organization of a
species into account.
Good, appropriate housing, however, often ends up as being a relatively
benign environment where access by competitors and predators is usually
prevented. The problem for biological conservation is that such benign
environments (1) may be of limited use to prepare captive animals for the
challenges that await them once they are released into the wild (see the
next section); (2) stifle the development and fine-tuning of complex behav-
iors, for example, antipredator behavior (see Curio, 1996); and (3) may not
reveal whether the captive stock is of poor genetic quality. Animals that
grow up in benign captive environments may lack appropriate antipredator
(Jarvi and Uglem, 1993) or foraging and hunting behavior (Scheepers and
Venzke, 1995), show a reduced ability to adapt to environmental fluctua-
tions (Kohane and Parsons, 1988), or lack exposure to the level of intra-
and interspecific competition necessary to reveal the consequences of in-
breeding depression (Miller, 1994). Finding the optimal trade-off between
welfare concerns and the requirements of biological conservation will be
a major task for the future (Wuichet and Norton, 1995).
9. Transportation and Translocation
Transportation and translocation may be an important stressor, but stud-
ies of the physiological response to and fitness consequences of such activi-
ties have only recently begun (Woodford and Kock, 1991). Transportation
of frequently handled sheep caused a substantial increase in heart rate
(Baldock and Sibly, 1990). In bighorn sheep released after capture and
instrumentation, heart rate was increased for 2 h after release and the
cardiac recovery time was 10 times longer than the maximum recovery
time for any disturbance to which individuals were subsequently exposed
(MacArthur et al., 1986). The physiological response to transportation in-
cludes a reduction in immunocompetence as demonstrated by a reduction in
the production of interferon and other components of the immune response
(Wattrang et al., 1994; Section IV,C,2; subsection on Vaccination in this
section). Fitness consequences of transportation in terms of high mortality
may be substantial and are often associated with a decline in immunocompe-
tence (Section IV,C,2).
Although there have been many translocations or reintroductions, few
studies have reviewed or experimentally explored factors that determine
the success or failure of such projects, or monitored the fate of released
animals in their new environment. In a review of translocation of mussels,
Cope and Waller (1995) found that only 16% of translocated populations
were monitored for five or more consecutive years, mortality in translocated
populations was unreported in 27% of projects, and there was little guidance
on the methods for translocation or for monitoring the subsequent long-
term status of translocated mussels. Wolf er al. (1996) reviewed a large
number of translocations and concluded that they are more likely to succeed
if animals were released into the core of the historical range of a species,
into habitat of good to excellent quality, if the population released was
large, and if the species had an omnivorous diet. Genetic factors may also
play a role because translocated populations and refuge populations show
reduced allozyme diversity, as allozymes rare in the parental population
are lost (Stockwell et al., 1996). Two behavioral hypotheses that underpin
many translocation efforts are that animals acclimated to the novel site
before release do better than those who are not and that wild-caught animals
do better than captive-bred ones (Bright and Morris, 1994). In other words,
there is a suspicion that released animals are “stressed” and fail to cope
because they are not sufficiently acclimated or lack the training to face the
challenges posed by a new unknown environment. There is observational
and experimental evidence that support both hypotheses. An experimental
study by Bright and Morris (1994) was set up to test both hypotheses
and found support for both of them. Stussy et al. (1994) estimated that
translocated female red deer had a lower annual survival rate than resident
females; they suggested that translocation may result in higher survival if
conducted outside winter months when conditions are less severe. A case
study of a released Iberian lynx, Lynx pardinus, by Rodriguez et al. (1995)
suggested that the successful release was based on careful feeding-training
and avoidance of human contact during the captive phase, as well as select-
ing a site with good habitat quality.
VII. THEEQUIVALENCE
OF NATURAL
A N D ANTHROPOGENIC
STRESSORS
An anthropogenic stressor would be considered equivalent to a natural

stressor if an organism’s response to the two was similar or identical. Figure
5 summarizes some hypotheses about such equivalence relationships be-
tween natural and anthropogenic stressors. Because equivalence relation-
ships may vary between species, Figure 5 illustrates potential equivalence
relationships that may apply to some but not all species. This is not an
exhaustive list of all documented links and the links discussed later in this
chapter are supported by evidence to a varying degree. We believe that
this exercise is useful because there is a well-developed body of theory
(behavioral and evolutionary ecology) that predicts how environmental
and social factors, resource availability, population density, pathogens, and
predators mold life-history tactics, behavior, foraging, survival, and repro-
ductive success of wildlife (e.g., Roff, 1992; Steams, 1992; McNamara and
Houston, 1996; Krebs and Davies, 1997). Because there are natural equiva-
lents for most anthropogenic factors, there may be evolved abilities of
coping with anthropogenic factors (Sections IV and V). The predictive
power of stress studies for biological conservation purposes would therefore
be greatly enhanced if this body of theory could be utilized (Section IV,B,S).
The best documented link is that between predation and several anthro-
pogenic stressors. These include visitor disturbance (Section V,B), sports
hunting (Section V,C), and handling (Section V1,C). It could be argued
FIG. 5. Hypothesized equivalence relationships between natural and anthropogenic stres-

son. Evidence from stress response studies suggests that the anthropogenic stressors in the
central unshaded group evoke stress responses that are borrowed from evolved responses to
natural stressors arranged in the outer shaded group.
that housing in captivity may be related to predation risk because housing

in captivity is inevitably associated with handling events.
Anthropogenic stressors that have a similar effect to temperature are
climatic (global) warming (Section V,E), and urbanization or development.
An experimental study demonstrated that local extinction of a land snail
was most likely caused by a small increase in ambient temperature caused
by thermal radiation from an urban area (Baur and Baur, 1993). Climatic
warming may also have effects similar to those of pathogens, because it
may favor pathogens and/or because host energy budgets are often related
to ambient temperature. Pathogens can modify the energy budgets of hosts,
with detrimental fitness consequences for the host (Munger and Karasov,
1989, 1991; Holmes and Zohar, 1990; Forstad et al., 1991).
Pollution and handling are linked to pathogens because both reactivate
latent viruses (Table I) and impair immunocompetence (Sections IV,C,
V,A, V1,C). Natural catastrophes could sometimes be considered the natu-
ral equivalent of pollution events when they are associated with significant
changes in atmospheric, terrestrial, freshwater, or marine contaminants,
and thereby may also contribute to global warming.
Visitor disturbance and development may be functionally equivalent to
a change in patch richness and other aspects of natural resource availability
that influence animals’ decisions on time budgeting and foraging tactics
(Gill et al., 1996; Sutherland, 1996). This may also include disturbance
caused by sports hunting (Section V,C). Housing in captivity often implies
food availability ad libitum, or practices equivalent to clumping of resources

(e.g., Boccia et al., 1988).
We suspect that anthropogenic factors that are equivalent to social insta-
bility may include translocation, which often is functionally equivalent to
migration or dispersal (Bright and Morris, 1994) because it removes individ-
uals or populations from one site and introduces them at another site where
conspecifics may already exist. The links between high population density
and confinement, transportation, translocation, and captivity follow the
hypothesis of Christian (1971, 1978, 1980).
OCCURRENCE
VIII. MINIMIZING AND IMPACTOF STRESS
IN CONSERVATION
RESEARCH
A N D MANAGEMENT
Conservation actions are often initiated to reduce the impact of anthropo-

genic stressors. For instance, guidelines for tourists visiting Antarctic pen-
guin colonies (Wilson et al., 1991) or internationally standardized rules on
whale watching (IFAW, 1996; IFAW et al., 1096) aim to minimize the
potential disturbance effects of tourism. Probably no disturbance is better
than little disturbance, and little disturbance is better than a lot of distur-
bance. Beyond such simple rules, however, it is currently impossible to
make general recommendations based on a sound factual and theoretical
basis because there are few rigorous studies that predict the fitness conse-
quences of stressors in a quantitative, dose-dependent fashion. Predictions
would be further complicated by the diversity of organisms, stress response
systems, and anthropogenic stressors that may occur (Sections V and VI).
As a result, many currently practiced recommendations are essentially
hunches based on rules of thumb. In many cases (e.g., wildlife viewing by
tourists in East African savannas), very different rules are practiced in
different countries even if they share the same ecosystem (e.g., the Serengeti
shared by Tanzania and Kenya) and current information is insufficient to
evaluate which rules are most appropriate. Even an international compari-
son of the implementation and consequences of different guidelines of how
to approach wildlife would be a major step forward.
In this section we concentrate on the issue of minimizing the stressful
consequences of conservation management and research activity itself.
Here, the data are better, alternatives have been frequently developed, and
sometimes their consequences have been rigorously tested. The first step
to minimize the occurrence and impact of stress in conservation research
and management is to recognize that conservation activities may be poten-
tially stressful. The second is to pay attention to principles of study design
(Section 111,D). The third step is to recognize that actions that minimize
the consequences of potentially stressful conservation activities readily fall

into two categories. The first category includes options for minimizing the
occurrence of stressors. The occurrence of stressors is sometimes controlled
by individual conservation or research programs (e.g., the frequency of
handling) and sometimes it is not (e.g., global warming), although concerted
efforts might be able to eventually reduce the impact of large-scale anthro-
pogenic stressors such as global warming. The second category includes
cases in which it is accepted that under certain conditions the occurrence
of a potential stressor is unavoidable and actions aim to maximize an
individual’s ability to cope with a stressor.
There are several general reviews useful for the practical design of conser-
vation research activities. Recommendations for the use of wild birds in
research can be found in the publications by the American Ornithologists’
Union (1988), Hoysak and Weatherhead (1991), and Le Maho et al. (1992).
Putman (1995) developed a framework to assess the costs and benefits of
capture, handling, and marking mammals in ecological field studies. Other
reviews and recommendations are provided by Cuthill (1991) and the Asso-
ciation for the Study of Animal Behaviourkhe Animal Behavior Society
published at regular intervals in the journal Animal Behaviour. Procedures
for the identification of subjects, capture, telemetry, and sampling of urine
and feces in socially living primates have been reviewed, among others, by
Rasmussen (1991), for capture, medical management, and anesthesia of
free-ranging wildlife by Jessup (1992), and for vaccinating wildlife by Hall
and Harwood (1990).
THE OCCURRENCE
A. MINIMIZING OF STRESS
Because interventions can be a major stressor, reducing the incidence of

interventions by replacing standard techniques with noninvasive ones or
at least improving standard research equipment to minimize the impact of
interventions would reduce the occurrence of stress in some instances.
1. Minimal-Invasive and Noninvasive Alternatives to

Standard Procedures
The development of minimal-invasive and noninvasive procedures has
been greatly facilitated by recent advances in molecular techniques. These
techniques can sometimes replace interventions that include blood sampling
as the conventional method of choice to answer many research questions.
As an example of the scope of the new techniques, fecal samples may
now be used to sex animals by analyzing fecal sex steroids (giant panda,
Ailuropoda melanoleuca: Kubokawa, 1993), or prove infection by canine
parvovirus using negative contrast electron microscopy (gray wolf Muneer

et al., 1988).
a. Adrenocortical Response. Blood sampling to determine the magni-
tude of the adrenocortical response (secretion of corticosteroids) has on
occasion been successfully replaced by the analysis of urine samples (domes-
tic dog: Jones et al., 1990; bighorn sheep: Miller et al., 1991; mule deer:
Saltz and White, 1991; mink: Madej et al., 1992; rhesus monkey: Crockett et
al., 1993; domestic cat: Graham and Brown, 1996), saliva samples (humans:
Kirschbaum and Hellhammer, 1994; white rhinoceros, Ceratotherium si-
mum: Schmidt and Sachser, 1996), or fecal samples (bighorn sheep: Miller
et a/., 1991; domestic cat: Graham and Brown, 1996).
6. Reproductive State. Daily monitoring of female reproductive state in
breeding programs may be possible by analyzing urine (Goeldi’s monkey,
Callimico goefdii: Jurke et a/., 1994) or fecal samples (Most1 et al., 1984).
Scrota1 dimensions have been determined noninvasively by conditioning
captive subjects to hold themselves in a standard vertical position on the
mesh walls of their cages so that the maximum width of the scrotum may
be compared to a square paper card (cotton-top tamarin, Saguinus oedipus:
Ginther and Washabaugh, 1996). Vibrostimulation is a much gentler, yet
more reliable method than conventional electro-ejaculation to assess sperm
quality and quantity (squirrel monkey: Yeoman et af., 1996).
c. Conservation Genetics. Protein diversity may be assessed from feather
pulp, obviating the need to blood-sample study animals (Marsden and May,
1984). Hair samples and fecal samples can be used t o identify mitochondrial
and nuclear gene sequences in conservation genetics studies (brown bear,
Ursus arctos: Taberlet and Bouvet, 1992; Kohn et al., 1995). A single plucked
feather may be sufficient for genetic studies that require only small amounts
of DNA, for example, when analyzing mitochondria1 DNA or microsatel-
lites with the help of the polymerase chain reaction (blue tit, Purus caeruleus:
Taberlet and Bouvet, 1991).
d. Monitoring Environmental Contaminants. Instead of killing study ani-
mals, levels and effects of organochlorines (PCB and DDT) may be moni-
tored by drawing blood samples (gray seal: Jenssen et af., 1994, 1995).
Monitoring mixed function oxidase activity and organochlorine content of
body tissue can now be accomplished by remote skin and hypodermic
biopsy (fin whale, Balaenoptera physalus, striped dolphin, Stenelfa coeru-
leoalba: Fossi et a/., 1992). Measurements of contamination with copper,
zinc, mercury, cadmium, or lead has been carried out on hair samples
collected from resting sites of endangered species in the absence of individu-
als (Mediterranean monk seal, Monachus monahus: Yediler et al., 1993).
2. Improving Standard Research Equipment
Sometimes, noninvasive alternatives are not yet available. In such cases,
improving standard research equipment to minimize stress is an option.
Designs that minimize stress may reduce the number of handling events
by avoiding the need to recapture animals. An example is the attachment
of a buoyant pack containing a VHF transmitter and a data recorder to
the pelt of harbor seals, which drops off after a predefined duration (Ellis
and Trites, 1992). Data collection by implantable transmitters can now be
remote-controlled by a computer. These devices are sophisticated enough
to detect acute changes in heart rate and core body temperature and discrim-
inate among several different types of behavior without restraining or han-
dling animals (Diamant et al., 1993; Kramer et al., 1993). The stressfulness
of the handling event may also be reduced. Delgiudice et al. (1990) reported
on the development of a capture collar that includes a syringe with immobili-
zation drugs. Once the capture collar is attached to an animal, chemical
immobilization can be initiated by a remote-controlled radio signal, obviat-
ing the need for darting the animal. Under some experimental conditions,
regular disturbance or handling of captive animals for blood sampling can
be avoided with remote-controlled blood sampling techniques (Le Maho
et al., 1992; Alexander et al., 1996). The shape and point of attachment of
data loggers can be designed so that they match the body contour of
the animal, resulting in substantial energy savings during foraging and
locomotion (penguins: Bannasch et al., 1994; Culik et al., 1994).
THE ABILITY
B. MAXIMIZING To COPESUCCESSFULLY
WITH STRESS
In some cases, interventions or other forms of potential stress may be

unavoidable. How can an organism be prepared to maximize its ability to
cope with stress?
1. Early Experience and Behavioral Training
In captivity, rat pups exposed to brief periods of innocuous handling
early in life showed a reduced adrenocortical response to a wide variety
of potential stressors, and this effect persisted throughout the life of the
animal. This has been interpreted to mean that early experience of handling
can reduce the stressfulness of handling later in life (Meaney et al., 1991;
Gonzalez et al., 1994; Rostene et al., 1995). Recent experiments with cats
and rhesus monkeys confirmed these results (Meaney et al., 1993; McCune,
1995). Apparently this effect occurs because handling modifies the tran-
scription activity of glucocorticoid receptor genes early in life, permanently
altering neuroendocrine responsitivity to stress (Meaney et al., 1993).
Conservation programs often consider the release of captively bred indi-
viduals into the wild where they might encounter environmental stressors
and predators, and the success of translocation or release programs may
be heavily affected by this (e.g., Scheepers and Venzke, 1995). Jarvi (1990)
and Jarvi and Uglem (1993) used a series of elegant experiments to investi-
gate the impact of encounters of salmon smolts that had been hatchery
reared in freshwater with predators under conditions of osmotic stress (in
seawater). Smolts were either naive or previously trained by exposing them
to predators in a contact (freely hunting predator) or noncontact fashion
(predator behind a glass screen). Trained smolts were more likely to re-
spond with appropriate behavior and show a reduction in the physiological
stress response than naive smolts, and contact-exposed smolts did better
than noncontact smolts. Active antipredator training might therefore im-
prove the performance of released animals and enhance the success of
release programs.
A soft release (in which animals are acclimated to a release site and/or
decide on their own when to leave a holding area) may be viewed as a
self-training exercise and thus be predicted to achieve a similar positive
effect. Bright and Morris (1994) experimentally demonstrated that soft
releases were indeed more successful than hard releases (no acclimation
at the release site).
2. Training to Cooperate
Numerous reports demonstrate that nonhuman primates can be trained
to cooperate with rather than resist common handling procedures such as
capture, injection, blood sampling, and veterinary examination (Reinhardt
et al., 1995). The same applies to captive cetaceans trained to drape their
flukes over the edge of the holding pool for blood samples (Thomas et al.,
1990). Animals can also be trained to donate urine or fecal samples in
appropriate containers (Kelley and Bramblett, 1981; Phillippi-Falkenstein
and Clarke, 1992; Anzenberger and Gossweiler, 1993). Cooperative animals
showed reduced or absent behavioral and physiological signs of distress
(Reinhardt et al., 1995). Some of these training procedures invoke aversive
stimulation, at least in primates (Rasmussen, 1991). Several studies demon-
strate that individuals or groups can be trained using positive reinforcement
procedures without aversive stimulation and that the effort required is
much less than skeptics might expect. Voluntary presentation by rhesus
monkeys of their leg for blood collection without mechanical restraint was
easily achieved (Reinhardt, 1991b). Only a minimal time investment was
needed to train a large troop of laboratory rhesus monkeys to cooperate
in a capture procedure, minimizing risk to personnel and distress to the
animals (Luttrell et al., 1994). Positive reinforcement was used to rapidly
train chimpanzees to move to indoor quarters on a verbal cue, allowing
personnel safe access to enclosures for maintenance, to move animals to
other quarters, and facilitate veterinary and research procedures (Stone et
al., 1996). The success of habituation and positive reinforcement procedures
is not restricted to nonhuman primates, as training of nyala, Tragelaphus
angasi, to enter a wooden crate for veterinary examination and blood

sampling has demonstrated (Grandin et al., 1995).
3. Social Support
Social stability and the presence of preferred social partners may assist
to significantly reduce the glucocorticoid stress response in an acute challen-
ging situation (Sachser and Beer, 1995; Sachser et al., 1997; Sapolsky, 1997).
This may be because the presence of preferred social partners improves
the confidence of an individual that it will be successful in meeting the
challenge, or because a potentially stressful situation is experienced as less
challenging if a preferred social partner is present.
4. Tranquilizers and Sedatives
Tranquilizers and sedatives may help to tone down the apparent behav-
ioral response of an animal to a handling event. For instance, placing a
sedative injector in the nest of breeding seabirds and administering an
intramuscular sedative by remote-control to facilitate capture avoided fear
and stress-related behavior during capture of birds (Wilson and Wilson,
1989). However, this does not necessarily imply that the physiological re-
sponse to intervention is muted in a similar way to the behavioral response.
Tranquilizers did not suppress the physiological stress response in impala,
Aepyceros melampus, to repeated capture, handling, and blood sampling,
as measured by increases in lactate, glucose, cortisol, total catecholamines,
osmolality, and hematocrit (Knox et al., 1990).
5. Optimizing Intervention Conditions
Several experimental studies confirm that optimizing intervention condi-
tions can result in substantial downregulation of the physiological response
to intervention. Predictable handling and husbandry routines and the provi-
sioning of appropriate places for concealment reduce the distress displayed
by domestic cats responding to handling by familiar or unfamiliar humans
(Carlstead et al., 1993). Quiet, efficient handling based on an understanding
of behavior reduces apparent levels of excitement and the incidence of
injuries when domestic cattle are tagged with ear implants and injected
(Grandin, 1987). The impact of handling in terms of the adrenocortical
response may also be reduced by handling captive individuals in a familiar
environment rather than taking them to a restraint apparatus (Reinhardt
et al., 1991). Enclosures may be optimized for efficient handling and data
collection, simultaneously reducing the potential of injury to the animal.
An example is the portable, semirigid enclosure with opaque, vertical sides
of vinyl-coated nylon developed by Davis and Allen (1989) to handle
captured waterfowl. Fear-related behavior and heart rate of red deer, either
confined in an unfamiliar holding pen with a stationary human present or

housed indoors, was reduced by darkening the holding area (Pollard and
Littlejohn, 1999, or by sound-proofing housing facilities to minimize noise
(Price et al., 1993).
IX. How IMPORTANT

CONCLUSIONS: Is STRESSIN
BIOLOGICAL
CONSERVATION?
From our review a few interesting rules and trends emerge. It is unwise
to assume that: (1)issues related to stress can be safely ignored in conserva-
tion or research; (2) short-term observations of behavior provide reliable
evidence about whether an organism is stressed; (3) a single-factor explana-
tion of stress-related individual and population responses is sufficient; and
(4) interventions in the course of research or conservation actions are
always benign. There are, however, good reasons to believe that: (1) the
awareness of the importance of stress in biological conservation has in-
creased steadily and will continue to do so; (2) in some cases individuals
and populations are adequately equipped to deal with anthropogenic stress
of moderate intensity because such stress mimics other factors to which
these animals have evolved an adaptive response; ( 3 ) it is possible to mea-
sure the extent to which resource use declines with disturbance and quantify
how much anthropogenic disturbance reduces the effective population size
of a habitat; (4) there are now many noninvasive alternatives to conven-
tional research procedures; and (5) stress associated with conservation and
research activities can be substantially reduced.
Conservation efforts attempt to improve population size, population per-
sistence, habitat distribution and quality, minimize the chance of genetic
depauperation within species, and preserve community structure and inter-
actions as intact as possible. How do stress studies help in this respect? We
believe that stress studies contribute in three major ways.
The first refers to the practicalities of conservation work: Should minimi-
zation of stress be considered an important element of the implementation
of conservation activities? And would the success of conservation efforts
be improved by minimizing stress-related fitness consequences? Our review
suggests that the answer to both questions is yes. Recognizing that stress
is an important factor paves the way for organizing conservation activities
around two paradigms: Minimizing the occurrence of stress, and maximizing
an individual’s or population’s ability to cope successfully with stress (Sec-
tion VIII). In this context it would be helpful if conservation activists and
researchers routinely reported the precise circumstances of interventions
and explain how they attempted to control for the effects of intervention.
The second refers to the recognition that stress must be considered an

important factor in evolution (Section IV and Hoffmann and Parsons,
1991; Parsons, 1988a, 1991, 1993a,b, 1994). Anthropogenic changes of the
environment are likely to create, or have already created, substantial selec-
tion pressures on populations, increasing additive genetic variance and
recombination rates (Parsons, 1988a). If we want to predict the future
course of populations, then understanding the magnitude and effects of
such selection pressures is important.
The third is that recent work on the genetics of stress resistance suggest
that conservation geneticists need to address several points of concern
neglected by standard treatments of evolutionary genetics. Current evolu-
tionary theory is not well geared to understand the effects of fluctuating
and stressful environments, as most work in evolutionary genetics focuses
on equilibrium populations in constant and stable environments (see Hoff-
mann and Parsons, 1991). Maximizing stress resistance or stress tolerance
may require: (1) the preservation of rare alleles rather than that of overall
genetic diversity (Futuyma, 1983); (2) an emphasis on the preservation of
marginal rather than central populations (Parsons, 1989a, 1995); (3) the
prevention of “genetic flooding” of marginal populations caused by the
creation of habitat corridors (Hoffmann and Parsons, 1991); (4) the recogni-
tion that the consequences of potential inbreeding depression in a newly
released captive population may not be predicted from a relatively benign
captive environment (Miller, 1994); ( 5 ) the recognition that heterozygotes
may express superior stress resistance only under conditions of environmen-
tal stress and not in benign environments (e.g., Scott and Koehn, 1990);
(6) the acknowledgment that long-term breeding in benign captive environ-
ments (e.g., in temperature-regulated facilities) may reduce the capacity of
the captive population to adapt to conditions of natural habitats (Kohane
and Parsons, 1988).
What are the consequences of ignoring the possible effects of stress in
conservation work? They may be profound if stress is part of the reason
why populations or species decline, if appropriate activities to stop the
decline are available and yet stress is not identified as a contributing factor.
This problem has been recently highlighted by the “cheetah controversy”
(May, 1995). O’Brien (1994) argued that in cheetahs small litter sizes and
difficulties encountered by captive breeding programs point to inbreeding
depression, which is thought to be caused by abnormally low levels of
genomic diversity. Caughley (1994) and Merola (1994) suggested instead
that low reproductive performance was not an adequate indicator of in-
breeding depression because breeding performance in captivity may have
been impaired by the stress caused by inadequate conditions. With im-
proved conditions, captive breeding programs should, and do, increase

reproductive success (May, 1995).
We conclude that a biologically appropriate conservation effort recog-
nizes potential stress situations and attempts to minimize the stressful conse-
quences of conservation activities.
X. SUMMARY
This chapter reviews why stress has important implications for biological
conservation and considers practical ways in which conservationists can
identify and tackle problems caused by stress. We take an evolutionary
approach that emphasizes links between stress and its consequences and
possible adaptations that permit animals to cope with stress. Using informa-
tion from several scientific disciplines we outline current knowledge of
the kinds of factors known to generate stress and the Darwinian fitness
consequences of stress. The magnitude and nature of the organism’s stress
response can be highly variable between species, populations of the same
species, individuals, and even change with the reproductive state or body
condition of an individual. Detailed knowledge on the sources of this varia-
tion and the rules that govern it is urgently required to predict the likely
impact of anthropogenic stressors. Such knowledge is lacking because the
plethora of retrospective observational studies of natural or anthropogenic
stressors contrasts with a paucity of rigorously designed experimental stud-
ies. Experimental studies have demonstrated that anthropogenic factors
such as environmental pollution, tourism and leisure activities, hunting,
noise, and global warming d o sometimes-but not invariably-cause stress,
with detrimental fitness consequences, in a wide variety of species, contexts,
and dosages. There are currently no general rules available to predict
when anthropogenic factors are likely to generate stress that will result in
detrimental fitness consequences, or which factors are likely to have the
most severe impact. Conservationists are not sufficiently aware of the impact
of natural stress, anthropogenic stress, or stress created by conservation-
related activities and scientific research, which are usually considered be-
nign. We outline a research program for stress in conservation biology and
provide recommendations that can improve the success of conservation
efforts by minimizing the frequency of occurrence of conditions that gener-
ate stress and maximizing the chance of organisms to successfully cope
with stress.
Acknowledgments
We are grateful to numerous colleagues who have discussed some of these issues with us
over the years or provided information. reprints. o r preprints. In particular we would like to
thank D. H. Abbott. G. Anzenberger. F. Aureli, N. Bahr. M. Bekoff. R. Burrows. A. Camperio

Ciani, A. Cockburn. E. Curio, P. Deimer. N. Hillgarth. L. Haas. J. Lamprecht. G. Lubach,
C. Richard-Hansen, R. M. Sapolsky. D. Quiatt, V. Reynolds, N. Sachser. J. Silk, T. Smith,
W. J. Sutherland, and J. Wingfield. We are grateful to R. Klein, B. Knauer, K. Schulz, A.
Turk, W. Wickler, and the Max-Planck-Gesellschaft for assistance and financial support and
to the referees for constructive comments.
References
Abbott, D. H. (1989).Social suppression of reproduction in primates. In “Comparative Socio-

ecology” (V. Standen and R. A. Foley, eds.). pp. 285-304. Blackwell, Oxford.
Abbott, D. H.. Saltzman, W., Schultz-Darken. N. J.. and Smith, T. E. (1997).Specific neuroen-
docrine mechanisms not involving generalized stress mediate social regulation of female
reproduction in cooperatively breeding marmoset monkeys. Ann. N. Y. Acad. Sci. 807,
219-239.
Abbott, J. C..and Dill, L. M. (1989).The relative growth of dominant and subordinate juvenile
steelhead trout (Salrno gairdneri) fed equal rations. Behavioitr 108, 104-1 13.
Alados, C. L.. Escos. J. M., and Emlen. J. M. (1996).Fractal structure of sequential behavior
patterns. Anim. Behav. 51, 437-443.
Alberts. S. C.. Sapolsky. R. M.. and Altmann, J. (1992).Behavioral, endocrine, and immunolog-
ical correlates of immigration by an aggressive male into a natural primate group. Horrn.
Behnv. 26, 167-178.
Alexander, S. L., Irvine, C. H. G., and Donald, R. A. (1996). Dynamics of the regulation of
the hypothalamo-pituitary-adrenal (HPA) axis determined using a nonsurgical method
for collecting pituitary venous blood from horses. Front. Neuroendocrinol. 17, 1-50.
Altmann. J., Hausfater, G., and Altmann. S. A. (1988).Determinants of reproductive success
in savannah baboons, Papio cynocephaliis. In “Reproductive Success” (T. H. Clutton-
Brock, ed.). pp. 403-418. University of Chicago Press, Chicago.
American Ornithologists’ Union. (1988). Report of committee on the use of wild birds in
research. Auk 105, Suppl., la-41a.
Amlaner. C. J., Sibly, R. M., and McCleery. R. (1979). Effects of telemetry transmitter
weight on breeding success in herring gulls. In “Proceedings of the Second International
Conference on Wildlife Biotelemetry” (F. M. Long, ed.), pp. 254-259. University of
Wyoming, Laramie.
Anderson. P. K. (1981).The behavior of the dugong (Dugong dicgon) in relation to conserva-
tion and management. Bull. Mar. Sci. 31, 640-647.
Anzenberger, G., and Gossweiler, H. (1993).How to obtain individual urine samples from
undisturbed marmoset families. Am. J. Prirnatol. 31, 223-230.
Ardern, S. L.. McLean, 1. G . , Anderson, S.. Maloney, R., and Lambert, D. M. (1994).The
effects of blood sampling on the behavior and survival of the endangered Chatham Island
black robin (Petroicn traversi). Conserv. B i d . 8, 857-862.
Arita, H. T. (1996).The conservation of cave-roosting bats in Yucatan. Mexico. Biol. Conserv.
76, 177-185.
Arnold, W., and Dittami, J. (1997).Reproductive suppression in male alpine marmots. Anirn.
Behav. 53, 53-66.
Arntz, W. E., and Fahrbach, E. (1991). “El NiAo. Klimaexperiment der Natur.” Birk-
hauser, Basel.
Arthur, A. Z. (1987).Stress as a state of anticipatory vigilance. Percept. Mot. Skills 64,7545.
Astheimer, L. B., Buttemer, W. A,, and Wingfield, J. C. (1992).Interactions of corticosterone
with feeding activity and metabolism in passerine birds. Ornis Scand. 23, 355-365.
Astheimer, L. B., Buttemer, W. A,, and Wingfield. J. C. (1994). Gender and seasonal differences
in the adrenocortical response to ACTH challenge in an arctic passerine, Zonotrichia
leucophrys gambelii. Gen. Comp. Endocrinol. 94, 33-43.
Astheimer, L. B., Buttemer, W. A,, and Wingfield, J. C. (1995). Seasonal and acute changes
in adrenocortical responsiveness in an arctic-breeding bird. Horm. Behav. 29, 442-457.
Backstrom, L., and Kauffman, R. (1995). The porcine stress syndrome: A review of genetics,
environmental factors, and animal well-being implications. Agri-Practice 16(8), 24-30.
Baker, C. S.. and Palumbi, S. R. (1994). Which whales are hunted? A molecular genetic
approach to monitoring whaling. Science 265, 1538-1539.
Baldock. N. M., and Sibly, R. M. (1990). Effects of handling and transportation on the heart
rate and behavior of sheep. Appl. Anim. Behav. Sci. 28, 15-39.
Bannasch, R., Wilson, R. P., and Culik, B. M. (1994). Hydrodynamic aspects of design and
attachment of a back-mounted device in penguins. J. Exp. Biol. 194,83-96.
Barker, I. K., Carbonell, P. L., and Bradley, A. J. (1981). Cytomegalovirus infection of the
prostate in the dasyurid marsupials. Phascogale tapoatafa and Antechinus stuartii. J. Wildl.
Dis. 17, 433-441.
Barnard, C. J., Behnke, J . M., and Sewell, J. (1996). Environmental enrichment, immunocompe-
tence and resistance to Babesia microri in male laboratory mice. Physiol. Behav. 60,1223-
1231.
Bateson, P. P. G. (1997). “The behavioural and physiological effects of culling red deer.”
Report to the Council of the National Trust, 77p.
Baur, B., and Baur, A. (1993). Climatic warming due to thermal radiation from an urban
area as possible cause for the local extinction of a land snail. J. Appl. Ecol. 30,333-340.
Baveco, J. M., and De Roos. A. M. (1996). Assessing the impact of pesticides on lumbricid
populations: An individual-based modelling approach. J. Appl. Ecol. 33, 1451-1468.
Beletsky, L. D., Orians, G. H., and Wingfield, J. C. (1989). Relationships of steroid hormones
and polygyny to territorial status, breeding experience, and reproductive success in male
red-winged blackbirds. Auk 106, 107-1 17.
Beletsky, L. D., Orians, G. H., and Wingfield. J. C. (1990). Steroid hormones in relation to
territoriality, breeding density, and parental behavior in male yellow-headed blackbirds.
Auk 107, 60-68.
Ben-Nathan, D. (1994). Stress and infectious disease. Isr. J . Vet Med. 49, 105-112.
Bentson, K. L., Astley, C. A,. Miles, F. P., Goldstein, D. S., Holmes, C., and Smith, 0. A.
(1996). Behavioral and physiological aspects of exposure of a group of baboons to an
intruder. Ahstr., Congr. Int. Primatol. Soc., 16th, 92.
Berglund, B., Lindvall, T., and Nordin, S. (1990). Adverse effects of aircraft noise. Environ.
Int. 16,315-338.
Bergman, A,, and Olsson, M. (1986). Pathology of baltic gray seal (Halichoerus grypus) and
ringed seal (Phoca hispida botnica) females with special reference to adrenocortical
hyperplasia: Is environmental pollution the cause of a widely distributed disease syndrome.
Riistatieteellisia Julkaisuja 44,47-62.
Beringer, J., Hansen, L. P., Wilding, W., Fischer, J., and Sheriff, S. L. (1996). Factors affecting
capture rnyopathy in white-tailed deer. J . Wildl. Manage. 60, 373-380.
Bernstein, C., Krebs, J. R., and Kacelnik, A. (1991). Distribution of birds amongst habitats:
Theory and relevance to conservation. In “Bird Population Studies” (C. M. Perrins,
J . D. Lebreton, and G. M. Hirons, eds.), pp. 317-345. Oxford University Press, Oxford.
Berthiaume. L., Heppell. J., Desy, M., Leblanc, L., Lallier, R., Bailey, R., and Dutil,
J. D. (1993). Manifestation of lymphocystis disease in American plaice (Hippoglossoides
platessoides) exposed to low salinities. Can. J. Fish. Aquat. Sci. 50, 430-434.
Black, S. R., Barker, I. K., Mehren, K. G., Crawshaw, G. J., Rosendal, S.. Ruhnke. L., Thorsen.
J., and Carman, P. S. (1988). An epizootic of mycoplasma ovipneumoniae infection in
captive Dall’s sheep (Ovis dalli dalli). J. Wildl. Dis. 24, 627-635.
Bleich. V. C., Bowyer, R. T.. Pauli, A. M., Nicholson, M. C., and Anthes, R. W. (1994).
Mountain sheep Ovis canadensis and helicopter surveys: ramifications for the conservation
of large mammals. Biol. Conserv. 70, 1-7.
Boccia, M. L., Laudenslager, M., and Reite, M. (1988). Food distribution, dominance and
aggressive behaviors in bonnet macaques. Am. J . Primatol. 16, 123-130.
Bodini, A,, Giavelli, G., and Rossi, 0.(1994). The qualitative analysis of community food webs:
Implications for wildlife management and conservation. J . Environ. Manage. 41,49-65.
Bonneau, R. H., Sheridan, J. F., Feng, N., and Glaser, R. (1991). Stress-induced effects on
cell-mediated innate and adaptive memory components of the murine immune response
to herpes simplex virus infection. Brain, Behav., Immun. 5, 274-295.
Bowman, L. A,, Dilley, S. R., and Keverne, E. B. (1978). Suppression of oestrogen-induced
LH surges by social subordination in talapoin monkeys. Nature (London) 275, 56-58.
Boyce, M. S. (1992). Population viability analysis. Annu. Rev. Ecol. Syst. 23, 481-506.
Bradley, A. J. (1987). Stress and mortality in the red-tailed phascogale, Phascogale calura
(Marsupialia: Dasyuridae). Gen. Comp. Endocrinol. 67, 85-100.
Bradley, A. J. (1990). Failure of glucocorticoid feedback during breeding in the male red-
tailed phascogale Phascogale calura (Marsupialia: Dasyuridae). J . Steroid Biochem. Mol.
Biol. 37, 155-163.
Bradley, A. J., McDonald, I. R., and Lee, A. K. (1980). Stress and mortality in a small
marsupial (Antechinus sfuartii).Gen. Comp. Endocrinol. 40, 188-200.
Brereton, R., Bennett, S., and Mansergh, I. (1995). Enhanced greenhouse climate change and
its potential effect on selected fauna of south-eastern Australia: A trend analysis. Biol.
Conserv. 72,339-354.
Bright, P. W.. and Morris, P. A. (1994). Animal translocation for conservation: Performance
of dormice in relation to release methods, origin and season. J. Appl. Ecol. 31,699-708.
Brisbin, I. L., White, G. C., and Bush, P. B. (1986). Polychlorinated biphenyl intake and the
growth of waterfowl: Multivariate analyses based on a reparameterized Richards sigmoid
model. Growrh 50, 1- 11.
Brodie. P.. and Paasche, A. (1985). Thermoregulation and energetics of fin and sei whales
based on postmortem stratified temperature measurements. Can. J. 2001.63,2267-2269.
Bronson, F. H. (1989). “Mammalian Reproductive Biology.” University of Chicago Press,
Chicago.
Broom, D. M., and Johnson, K. G. (1991). “Animal Welfare.” Chapman & Hall, London.
Broome, L. A., and Geiser, F. (1 995). Hibernation in free-living mountain pygmy-possums,
Burramys parvrcs (Marsupialia: Burramyidae). Aust. J. Zool. 43, 373-379.
Burel, C., Mezger, V., Pinto, M.. Rallu, M., Trigon, S., and Morange, M. (1992). Mammalian
heat shock protein families. Expression and functions. Experientia 48, 629-634.
Burley, N. (1986). Comparison of the band-color preferences of two species of estrildid finches.
Anim. Behav. 34, 1732-1741.
Burley, N. (1988). Wild zebra finches have band-color preferences. Anim. Behav. 36, 1235-
1237.
Burley. N., Krantzberg, G., and Radman, P. (1982). Influence of colour-banding on the
conspecific preferences of zebra finches. Anim. Behav. 27,686-698.
Burrows, R. (1992). Rabies in wild dogs. Nature (London) 359, 277-277.
Burrows, R., Hofer, H., and East, M. L. (1994). Demography, extinction and intervention in
a small population: The case of the Serengeti wild dogs. Proc. R. Soc. London, Ser. B
256, 281 -292.
Burrows, R.. Hofer. H.. and East, M. L. (1995). Population dynamics, intervention and survival
in African wild dogs. (Lycaon pictus). Proc. R. Soc. London, Serv. B 262, 235-245.
Busby, D. G.. White, L. M., and Pearce, P. A. (1990). Effects of aerial spraying of fenitrothion
on breeding white-throated sparrows. J . Appl. E d . 27,743-755.
Bush, M.,Montali, R. J., Brownstein, D., James, A. E., and Appel. M. J. G. (1976). Vaccine-
induced canine distemper in a lesser panda. J. Am. Vet. Med. Assoc. 169, 959-960.
Butler, R. G.. Harfenist. A., Leighton, F. A., and Peakall, D. B. (1988). Impact of sublethal
oil and emulsion exposure on the reproductive success of Leach’s storm petrels: Short
and long-term effects. J. Appl. Ecol. 25, 125-143.
Buttemer, W. A., Astheimer. L. B., and Wingfield, J. C. (1991). The effect of corticosterone
on standard metabolic rates of small passerine birds. J . Comp. Physiol. B 161, 427-431.
Calow. P.. and Sibly, R. M. (1990). A physiological basis of population processes: ecotoxicologi-
cal implications. Funct. Ecol. 4, 283-288.
Carlstead. K., Brown. J. L., and Strawn. W. (1993). Behavioral and physiological correlates
of stress in laboratory cats. Appl. Anim. Behau. Sci. 38, 143-158.
Carpenter. F. L., and Hixon, M. A. (1988). A new function for torpor fat conservation in a
wild migrant hummingbird. Condor 90, 373-378.
Carpenter, J. W., Appel, M. J. G., Erickson, R. C.. and Novilla, M. N. (1976). Fatal vaccine-
induced canine distemper virus infection in black-footed ferrets. J. Am. Vet. Merf.Assoc.
169, 961-964.
Castrucci. G., Frigeri, F.. Cilli. V., Tesei, B., Arush, A. M.. Pedini, B., Ranucci. S., and
Rampichini. L. (1980). Attempts to reactivate bovid herpesvirus-2 in experimentally
infected calves. Am. J . Vet. Res. 41, 1890-1893.
Caughley. G. (1994). Directions in conservation biology. J. Anim. Ecol. 63, 215-244.
Chalmers. G. A., and Barrett. M. W. (1982). Capture myopathy. In “Non-infectious Diseases
in Wildlife” (G. L. Hoff and J. W. Davies, eds.), pp. 84-94. Iowa State University
Press, Ames.
Chamove, A. S. (1988). Cage design reduces emotionality in mice. Lab. AnOn. 23, 215-219.
Chamove, A. S. (1989). Environmental enrichment: A review. Anim. Technol. 40, 155-178.
Charlesworth, B. (1994). “Evolution in Age-structured Populations.” Cambridge University
Christian, J. J. (1971). Population density and reproductive efficiency. Biol. Reprod. 4,248-294.
Christian. J. J. (1978). Neurobehavioural endocrine regulation in small mammal populations.
In “Populations of Small Mammals under Natural Conditions” (D. P. Snyder, ed.), pp.
143-158. University of Pittsburgh Press, Pittsburgh.
Christian, J. J. (1980). Endocrine factors in population regulation. In “Biosocial Mechanisms
in Population Regulation” (M. N. Cohen, R. S. Malpass, and H. Klein, eds.), pp. 55-115.
Yale University Press, New Haven, CT.
Chrousos. G. P., Loriaux, D. L., and Gold. P. W. (1988). The concept of stress and its historical
development. In “Mechanisms of Physical and Emotional Stress” (G. P. Chrousos,
D. L. Loriaux, and P. W. Gold. eds.). pp. 3-7. Plenum. New York.
Clark, W. W. (1992). Hearing: The effects of noise. Otolaryngol. Head Neck Surg. 106,669-676.
Clarke, A. S., and Schneider. M. L. (1993). Prenatal stress has long-term effects on behavioral
responses to stress in juvenile rhesus monkeys. Dev. Psychobiol. 26,293-304.
Clarke, A. S., Wittwer, D. J., Abbott. D. H.. and Schneider, M. L. (1994). Long-term effects
of prenatal stress on HPA axis activity in juvenile rhesus monkeys. Deu. Psychobiol.
27,257-269.
Clarke, G. M. (1995). Relationships between developmental stability and fitness: Application
for conservation biology. Conserv. Bid. 9, 18-24.
Clarke, G. M. (1996). Relationships between fluctuating asymmetry and fitness: How good
is the evidence? Pac. Conserv. Biol. 2, 146-149.
Cochran, W. G. (1977). “Sampling Techniques.” Wiley, New York.
Cochrane, B. J., Mattley, Y. D., and Snell, T. W. (1994). Polymerase chain reaction as a tool
for developing stress protein probes. Environ. To.ricol. Chem. 13, 1221-1229.
Cockburn. A. (1997). Living slow and dying young: senescence in marsupials. I n “Recent
Advances in Marsupial Biology” (N. Saunders and L. Hinds, eds.). University of New
South Wales Press, New Kensington, Australia (in press).
Coe, C. L., Lubach, G. R., and Ershler, W. B. (1994). Maternal influences on the development
of immune competence in infancy. Psychol. Beitr. 36, 15-21.
Coe, C. L., Lubach, G. R., Karaszewski, J. W.. and Ershler, W. B. (1996). Prenatal endocrine
activation alters postnatal cellular immunity in infant monkeys. Brain, Behav.. Immun.
10, 221-234.
Cohen, J. (1988). “Power Analysis for the Behavioral Sciences.” 2nd ed., Erlbaum. Phila-
delphia.
Colwell, M. A., Gratto, C. L., Oring, L. W., and Fivizzani, A. J. (1988). Effects of blood
sampling on shorebirds: Injuries, return rates and clutch desrtions. Condor 90, 942-945.
Cope, W. G., and Waller, D. L. (1995). Evaluation of freshwater mussel relocation as a
conservation and management strategy. Regul. Rivers: Res. Manage. 11, 147-155.
Cosby, S. L., McQuaid, S., Duffy, N., Lyons, C., Rima, B. K., Allan, G. M., McCullough.
S. J., Kennedy, S., Smyth, J. A., McNeilly, F., and Craig, C. (1988). Characterization of
a seal morbillivirus. Nature (London) 336, 115-115.
Cosens, S. E., and Dueck, L. P. (1993). Icebreaker noise in Lancaster Sound. Northwest
Territories. Canada: Implications for marine mammal behavior. Mar. Mammal. Sci. 9,
285-300.
Costa, D. P., and Trillmich, F. (1988). Mass changes and metabolism during the perinatal
fast: A comparison between Antarctic (Arcrocephalus gazella) and Galapagos fur seals
(Arctocephalus galapagoensis). Physiol. Zool. 61, 160-169.
Cotter, R. C., and Gratto, C. J. (1995). Effects of nest and brood visits and radio transmitters
on rock ptarmigan. J. Wildl. Manage. 59, 93-98.
Creel, S. (1992). Causes of wild dog deaths. Nature (London) 360, 633-633.
Creel, S. (1996). Conserving wild dogs. Trends Ecol. Evol. 11, 337-337.
Creel, S. R., Creel, N. M., and Monfort, S. (1997). Radiocollaring and stress hormones in
African wild dogs. Conserv. Biol. 11,544548.
Crockett, C. M., Bowers, C. L., Sackett, G. P., and Bowden, D. M. (1993). Urinary cortisol
responses of longtailed macaques to five cage sizes, tethering, sedation, and room change.
Am. J. Primatol. 30,55-14.
Croll, D. A., Jansen, J. K., Goebel, M. F., Boveng, P. L., and Bengtson, J. L. (1996). Foraging
behavior and reproductive success in chinstrap penguins: The effects of transmitter attach-
ment. J. Field Omithol. 67, 1-9.
Culik, B. M. (1992). Diving heart rates in Adtlie penguins (Pygoscelisadeliae).Comp. Biochem.
Physiol. A 102,487-490.
Culik, B. M., and Wilson, R. P. (1991a). Swimming energetics and performance of instrumented
AdClie penguins (Pygoscelis adeliae). J. Exp. Biol. 158,355-368.
Culik, B. M., and Wilson, R. P. (1991b). Penguins crowded out? Nature (London) 351,340-340.
Culik, B. M., and Wilson, R. P. (1992). Field metabolic rates of penguins instrumented penguins
using doubly-labelled water. J. Comp. Physiol. B 162, 567-573.
Culik, B. M., and Wilson, R. P. (1995). Penguins disturbed by tourists. Nature (London)
376,301-302.
Culik, B. M., Adelung, D.. and Woakes, A. J. (1990). The effect of disturbance on the heart
rate and behaviour of Adelie penguins (Pygoscelis adeliae) during the breeding season.
In “Antarctic Ecosystems. Ecological Change and Conservation” (K. R. Kerry, and G.
Hempel, eds.), pp. 171-182. Springer, Berlin.
Culik. B. M., Bannasch, R.. and Wilson, R. P. (1994). External devices on penguins: How
important is shape? Mar. Biol. 118,353-357.
Curio, E.(1996). Conservation needs ethology. Trends Ecol. Evol. 11, 260-263.
Cuthill, I. (1991). Field experiments in animal behaviour: methods and ethics. Anim. Behav.
42, 1007-1014.
Daan. S., Deerenberg. C., and Dijkstra, C. (1996). Increased daily work precipitates natural
death in the kestrel. J. Anim. E d . 65, 539-544.
Dabbert, C. B., and Powell, K. C. (1993). Serum enzymes as indicators of capture myopathy
in mallards Anus platyrhynchos. J. Wildl. Dis. 29, 304-309.
Daly. M., Wilson, M. I., Behrends, P. R.. and Jacobs. L. F. (1992). Sexually differentiated
effects of radio transmitters on predation risk and behaviour in kangaroo rats Dipodoniys
merriami. Can. J. Zool. 70, 1851-1855.
Davis. D. S., and Allen, H. A. (1989). An improved waterfowl enclosure: Considering animal
welfare as a research priority. J. Field Ornifhol. 60, 162-167.
Davis, R. W., Williams, T. M., Thomas, J. A,, Kastelein, R. A,, and Cornell, L. H. (1988). The
effects of oil contamination and cleaning on sea otters (Enhydra lutris). 11. Metabolism,
thermoregulation and behavior. Can. J. Zool. 66, 2782-2790.
Dawson, W. R., and Hudson, J. W. (1970). Birds. In “Comparative Physiology of Thermoregu-
lation I ” (G. C. Whittow, ed.), pp. 223-310. Academic Press, New York.
Delgiudice, G. D., Mech, L. D., and Seal, U. S. (1989). Physiological assessment of deer
populations by analysis of urine in snow. J. Wildl. Manage. 53, 284-291.
Delgiudice, G. D., Kunkel, K. E., Mech, L. D., and Seal, U. S. (1990). Minimizing capture-
related stress on white-tailed deer with a capture collar. J. Wildl. Manage. 54, 299-303.
Delgiudice, G. D., Seal, U. S., and Mech. L. D. (1991). Indicators of severe undernutrition
in urine of free-ranging elk during winter. Wildl. Soc. Bull. 19, 106-110.
den Boer, P. J. (1990). On the stabilization of animal numbers. Problems of testing. 3. What
do we conclude from significant test results? Oecologia 83, 38-46.
den Boer, P. J. (1991). Seeing the trees for the wood: Random walks or bounded fluctuations
of population size? Oecologia 86, 484-491.
Dennis, B., and Taper, M. L. (1994). Density dependence in time series observations of natural
populations: Estimation and testing. Ecol. Monogr. 64, 205-224.
de Swart, R. L., Kluten, R. M. G., Huizing, C. J., Vedder. L. J., Reijnders, P. J. H., Visser.
1. K. G., UytdeHaag, F. G. C. M., and Osterhaus, A. D. M. E. (1993). Mitogen and
antigen induced B and T cell responses of peripheral blood mononuclear cells from the
harbour seal (Phoca vitulina). Vet. Inmunol. Immunopathol. 37, 217-230.
de Swart, R. L.. Ross, P. S., Vedder, L. J., Timmerman. H. H., Heisterkamp, S. H.. van
Loveren, H., Vos, J. G.. Reijnders, P. J. H., and Osterhaus, A. D. M. E. (1994). Impairment
of immune function in harbour seals (Phoca vitulinu) feeding on fish from polluted waters.
Ambio 23, 155-159.
de Swart, R. L., Ross, P. S., Timmermann, H. H., Vos. H. W.. Reijnders, P. J. H., Vos. J . G.,
and Osterhaus. A. D. M. E. (1995a). Impaired cellular immune response in harbour seals
(Phoca vitulina) feeding on environmentally contaminated herring. Clin. Exp. Immunol.
101,480-486.
de Swart, R. L., Ross. P. S., Timmerman, H. H.. Hijman. W. C., de Ruiter. E. M., Liem,
A. K. D., Brouwer, A,, van Loveren, H., Reijnders, P. J. H., Vos. J. G.. and Osterhaus,
A. D. M. E. (199%). Short term fasting does not aggravate immunosuppression in harbour
seals (Phoca vitulina) with high body burdens of organochlorines. Chemosphere 31,4289-
4306.
Dettling, A., and Pryce, C. R. (1996). Social and emotional development in captive Goeldi’s
monkeys. Abstr. Congr. Int. Primatol. Soc., 16th, 088.
Diamant, M., Van Wolfswinkel, L., Altorffer, B., and De Wied, D. (1993). Biotelemetry:
Adjustment of a telemetry system for simultaneous measurements of acute heart rate
changes and behavioral events in unrestrained rats. Physiol. Behav. 53, 1121-1 126.
Dietz, R., Ansen. C. T., Have, P., and Heide-Jflrgensen, M. P. (1989). Mass deaths of harbor
seals (Phoca vitulina) in Europe. Ambio 18, 258-264.
Dilsaver, S. C., Overstreet, D. H., and Peck, J. A. (1992). Measurement of temperature in
the rat by rectal probe and telemetry yields compatible results. Pharmacol., Biochem.
Behav. 42,549-552.
Duffus. D. A., and Dearden, P. (1990). Non-consumptive wildlife-oriented recreation: A
conceptual framework. Biol. Conserv. 53, 21 3-232.
Dufour. K. W., Ankney, C. D., and Weatherhead, P. J. (1993). Nonrepresentative sampling
during waterfowl banding: Emphasis on body condition. J. Wildl. Manage. 57, 741-751.
Dunbar, R. I. M. (1989). Reproductive strategies in female gelada baboons. In “The Sociobiol-
ogy of Sexual and Reproductive Strategies” (0.A. E. Rasa. C. Vogel, and E. Voland,
eds.), pp. 74-92, Chapman & Hall, London.
Dunlap, K. D., and Wingfield, J. C. (1995). External and internal influences on indices of
physiological stress: I. seasonal and population variation in adrenocortical secretion of
free-living lizards, Sceloporus occidentalis. J. Exp. Zool. 271, 36-46.
Dunn. A. J., Powell, M. L.. Moreshead, W. V., Gaskin, J. M., and Hall, N. R. (1987). Effects
of Newcastle disease virus administration to mice on the metabolism of cerebral biogenic
amines, plasma corticosterone. and lymphocyte proliferation. Brain, Behav., Immunol.
1, 216-230.
Dunn, A. J., Powell, M. L., Meitin, C., and Small, P. A. (1989). Virus infection as a stressor:
Influenza virus elevates plasma concentrations of corticosterone, and brain concentrations
of MHPG and tryptophan. Physiol. Behav. 45,591-594.
Dunnet, G. M. (1982). Oil pollution and seabird populations. Philos. Trans. R. Soc. London,
Ser B 297,413-427.
Durchfeld, B., Baumgartner, W., Herbst, W., and Brahm, R. (1990). Vaccine-associated canine
distemper infection in a litter of African hunting dogs (Lycaon pictus). Zentralbl. Veteri-
narmed., Reihe, B 37, 203-212.
East, M. L. (1996). Survivorship in African wild dogs. Conserv. Biol. 10, 313-313.
Echternkamp, S. E. (1984). Relationship between luteinizing hormone and cortisol in acutely
stressed beef cows. Theriogenology 22, 305-312.
Eisa, M., Osman, 0.M., Karrar, A. E., Abdel Rahim, A. H. (1980). An outbreak of bluetongue
in sheep in the Sudan. Vet. Rec. 106,481-482.
Ellis, G. M., and Trites, A. W. (1992). The RAM-packs came back: A method for attaching
and recovering pinniped data recorders. A q u a Mamm. 18, 61-64.
Enzenbacher, D. J. (1993). Tourists in Antarctica: Numbers and trends. Tourism Manage.
14, 142-146.
Escbs, J . M.. Alados, C. L., and Emlen, J . M. (1995). Fractal structures and fractal functions
as disease indicators. Oikos 74, 310-314.
Eskens, U., Kaleta. E. F., and Unger, G. (1994). Eine Herpesvirus-bedingte Enzootie-
Pachecosche Papageienkrankheit-in einem Psittazidenbestand. Tieriirztl. Praxis 22,
542-553.
Etter, R. J. (1988). Physiological stress and color polymorphism in the intertidal snail Nucella
lapillus. Evolution (Lawrence, Kans.) 42, 660-680.
Euker, J. S., Meites, J.. and Riegle, G. D. (1975). Effects of acute stress on serum LH
and prolactin in intact. castrate and dexamethasone-treated male rats. Endocrinology
(Baltimore) 96, 85-92.
Evans, P. G. H. (1987). “The Natural History of Whales and Dolphins.” Christopher
Helm. London.
Evans, P. G. H. (1996). Human disturbance of cetaceans. I n “The Exploitation of Mammal
Populations” (V. J. Taylor and N. Dunstone, eds.), pp. 376-394. Chapman & Hall. London.
Evans, P. G. H., Carson, Q., and Fisher, P. (1994). A study of the reactions of harbour
porpoises to various boats in the coastal waters of S. E. Shetland. I n ”European Research
on Cetaceans 8” (P. G. H. Evans, ed.), pp. 60-64. European Cetacean Society, Cam-
bridge. UK.
Ewing. D. E.. Clark, W. R.. and Vohs, P. A. (1994). Evaluation of implanted radio transmitters
in pheasant chicks. J. Iowa Acad. Sci. lOl(3-4), 86-90.
Fang, L. S., Chen, Y.W., and Chen, C. S. (1991). Feasibility of using ATP as an index for
environmental stress on hermatypic coral. Mar. E d . : Progr. Ser. 70, 257-262.
Farah. I. 0. (1996). Stress related enterotyphylocolitis in baboons. Abstr., Congr. In?. Primatol.
Soc.. 16th p. 642.
Faulkes, C. G., and Abbott. D. H. (1997). The physiology of a reproductive dictatorship:
Regulation of male and female reproduction by a single breeding female in colonies of
naked mole rats. In “Cooperative Breeding in Mammals” (N. J. Solomon and J. A.
French, eds.), pp. 302-334. Cambridge University Press, Cambridge, UK.
Felsburg, P. J., Keyes, L. L., Krawiec, D. R., and Rubin, S. I. (1986). The effect of general
anesthesia on canine lymphocyte function. Vet. Immitnol. Immunopnthol. 13, 63-70.
Fernandez. C., and Azkona, P. (1993). Human disturbance affects parental care of marsh
harriers and nutritional status of nestlings. J. Wildl. Manage. 57, 602-608.
Fescharek, R., Franke. V., and Samuel, M. R. (1994). Do anaesthetics and surgical stress
increase the risk of post-exposure rabies treatment failure? Vaccine 12, 12-13.
Fleming, M. R. (1985). The thermal physiology of the mountain pygmy-possum. Burramys
parvics (Marsupialia: Burramyidae). Aust. Mammal. 8, 79-90.
Fletcher, J . L., and Busnel, R. G. (1978). “Effects of Noise on Wildlife.” Academic Press,
New York.
Forstad, I . , Nilssen. A. C., Halvorsen, O., and Andersen, J. (1991). Parasite avoidance: The
cause of post-calving migrations in Rangifer? Can. J. 2001.69,2423-2429.
Fossi, M. C., Marsili, L., Leonzio, C., Notarbartolo, D. I., Sciara, G . , Zanardeli, M., and
Focardi, S. (1992). The use of non-destructive biomarker in rnediterranean cetaceans:
Preliminary data on MFO activity in skin biopsy. Mar. Pollut. Bull. 24, 459-461.
Foster, B. R., and Rahs, E. Y. (1983). Mountain goat response to hydroelectric exploration
in northwestern British Columbia. Environ. Manage. 7, 189-197.
Fowler, G . S.. Wingtield, J. C., and Boersma, P. D. (1995). Hormonal and reproductive effects
of low levels of petroleum fouling in Magellanic penguins (Spheniscus magellanicus).
Auk 112,382-389.
Frankel. 0. H., and SoulC, M. E. (1981). “Conservation and Evolution.” Cambridge University
Fraser. A. F., and Broom, D. M. (1990). “Farm Animal Behaviour and Welfare.” Baillikre
Tindall, London.
Fraser, D., Ritchie, J. S. D., and Fraser, A. F. (1975). The term ‘stress’ in a veterinary context.
Br. Vet. J. 131, 653-662.
Frederick, P. C. (1986). Parental desertion of nestlings by white ibis (Eudocirnus albus) in
response to muscle biopsy. J. Field Omithol. 57, 168-173.
Frederick, R. B., Clark, W. R., and Klaas, E. E. (1987). Behavior energetics and management
of refuging waterfowl: A simulation model. Wildl. Monogr. 96, 1-35.
Freeman, B. M. (1985). Stress and the domestic fowl: Physiological fact or fantasy? World’s
Poult. Sci. J. 41, 45-51.
French. J. A. (1997). Proximate regulation of singular breeding in callitrichid primates. In
“Cooperative Breeding in Mammals” (N. S. Solomon and J. A. French, eds.), pp. 34-75.
Cambridge University Press, Cambridge, UK.
Frerking, H.. Kramer, R., Schiele, R., and Roder, B. (1995). Klinische Erfahrungen mit
Paramunitats-inducern beim Rind anlablich von Tierschauen. Dsch. Tierarztl. Wochen-
schr. 102, 188-189.
Fretwell. S. D. (1 972). “Populations in a Seasonal Environment.” Princeton University Press,
Princeton. NJ.
Frisch. J. E. (1981). Changes occurring in cattle as a consequence of selection for growth rate
in a stressful environment. J. Agric. Sci. 96, 23-38.
Futuyma, D. J. (1983). Interspecific interactions and the maintenance of genetic diversity. In
“Genetics and Conservation” (C. M. Schonewald-Cox, S. M. Chambers, B. MacBryde,
and L. Thomas, eds.), pp. 364-373. BenjaminKummings, London.
Gaisler. H., Hanak, V., and Horocek, I. (1981). Remarks on the current status of bat popula-
tions in Czechoslovakia. Myotis 18-19, 68-75.
Gammonley, J. H., and Kelley, J. R. (1995). Effects of back-mounted ratio packages on
breeding wood ducks. J . Field Ornithol. 65, 530-533.
Garrott. R. A., Bartmann, R. M.. and White, G. C. (1985). Comparison of radio-transmitter
packages relative to deer fawn mortality. J. Wildl. Manage. 49, 758-759.
Gartner, K., Biittner, D.. Dohler, K., Friedel, R., Lindena, J., and Trautschold, I. (1980).
Stress response of rats to handling and experimental procedures. Lab. Anim. 14,267-274.
Geiser. F., and Broome, L. S. (1991). Hibernation in the mountain pygmy-possum Burramys
parvus (Marsupialia). J . Zool. 223, 593-602.
Geiser. F., and Broome, L. S. (1993). The effect of temperature on the pattern of torpor in
a marsupial hibernator. J. Comp. Physiol. B 163, 133-137.
Geiser, F.. Sink. H. S., Stahl, B., Mansergh, I., and Broome, L. S. (1990). Differences in the
physiological response to cold in wild and laboratory-bred mountain pygmy-possums,
Burramys parvus (Marsupialia). Aust. Wildl. Res. 17, 535-539.
Gesser, R. M., Valyi-Nagy, T.. Altschuler, S. M., and Fraser, N. W. (1994). Oral-oesophageal
inoculation of mice with herpes simplex virus type 1 causes latent infection of the vagal
sensory ganglia (nodose ganglia). J. Gen. Virol. 75, 2379-2386.
Gill, J. A., Sutherland, W. J., and Watkinson, A. R. (1996). A method to quantify the effects
of human disturbance on animal populations. J. Appl. Ecol. 33, 786-792.
Ginther, A. J.. and Washabaugh, K. F. (1996). Non-invasive measurement of scrota1 width
in captive cotton-top tamarins (Suguinus o. oedipus): A tool for evaluating development.
Abstr., Congr. Int. Primatol. Soc. 16th, 114.
Ginzburg, L. R.. Ferson, S., and Akqakaya, H. R. (1990). Reconstructibility of density depen-
dence and the conservative assessment of extinction risks. Conserv. Biol. 4, 63-70.
Goldberg, J. S., and Haas, W. (1978). Interactions between mule deer dams and their radio-
collared and unmarked fawns. J. Wildl. Manage. 42,422-425.
Gonzalez, A. S., Echandia, E. L. R.,Cabrera, R., and Foscolo. M. R. (1994). Neonatal chronic
stress induces subsensitivity to chronic stress in adult rats: 11. Effects on estrous cycle in
females. Physiol. Behav. 56, 591-595.
Goodrich, J. M., and Berger, J. (1994). Winter recreation and hibernating black bears llrsus
umericunus. Biol. Conserv. 67, 105-110.
Goodwin, H. (1996). In pursuit of ecotourism. Biodiversity Conserv. 5, 277-291.
Gottman, J. M.. and Roy. A. K. (1990). “Sequential Analysis. A Guide for Behavioral Re-
searchers.” Cambridge University Press, Cambridge, UK.
Graham, L. H.. and Brown, J. L. (1996). Cortisol metabolism in the domestic cat and implica-
tions for non-invasive monitoring of adrenocortical function in endangered felids. Zoo
Biol. 15, 71-82.
Grandin, T. (1987). Animal handling. Vet. Clin. North Am. Food Anim. Pract. 3, 323-338.
Grandin. T., Rooney, M. B., Phillips, M., Cambre, R. C., Irlbeck. N. A,, and Graffam, W.
(1995). Conditioning of nyala (Tragelaphiis nngasi) to blood sampling in a crate with
positive reinforcement. Zoo B i d . 14, 261-273.
Green, J . S., Golightly, R. T., Lindsey, S. L., and Leamaster, B. R. (1985). Use of radio
transmitter implants in wild canids. Great Basin Nat. 45, 567-570.
Green, R. H. (1979). “Sampling Design and Statistical Methods for Environmental Biologists.”
Wiley, New York.
Greenberg, N.. and Wingfield. J. C. (1987). Stress and reproduction: Reciprocal relationships.
In “Reproductive Endocrinology of Fishes, Amphibians and Reptiles” (D. 0.Norris and
R. E. Jones. eds.), pp. 461-503. Wiley, New York.
Greenwood, R. B., and Sargeant, A. B. (1973). Influence of radio packs on captive mallards
and blue-winged teal. 1. Wildl. Manage. 47, 369-375.
Gremillet, D., Schmid. D.. and Culik, B. M. (1995). Energy requirements of breeding great
cormorants. Phalacrocorax carbo sinensis. Mar. Ecol.: Prog. Ser. 121, 1-9.
Griffiths, H. 1. (1994). The effects upon badgers (Meles meles) of the activities of a single,
persistent poacher. Anim. Welfare 3, 219-225.
Gustafson. A. W.. and Belt. W. D. (1981). The adrenal cortex during activity and hibernation
in the male little brown bat, Myoris hccifugiis htcifilgus: Annual rhythm of plasma cortisol
levels. Gen. Comp. Endocrinol. 44, 269-278.
Hall, A., and Harwood, J. (1990). “The Intervet Guidelines to Vaccinating Wildlife.” Intervet,
Cambridge, UK.
Hannon, S. J., and Wingfield. J. C. (1990). Endocrine correlates of territoriality, breeding stage,
and body molt in free-living willow ptarmigan of both sexes. Can. J. 2001.68,2130-2134.
Hansbrough, J. F., Zapata-Sirvent. R. L., Bartle, E. J., Anderson, J. K., Elliott, L., Mansour,
M. A., and Carter, W. H. (1985). Alterations in spleenic lymphocyte subpopulations and
increased mortality from sepsis following anesthesia in mice. Anesthesiology 63,267-273.
Haour. F., Marquette. C., Ban, E., Crumeyrolle-Arias, M., Rostene, W.. Tsiang, H., and
Fillion, G. (1995). Receptors for interleukin-I in the central nervous and neuroendocrine
systems: Role in infection and stress. Ann. Endocrinol. 56, 173-179.
Hare, S.. and Robertson, 0.H. (1959). Changes in plasma 17-hydroxycorticosteroids accompa-
nying sexual maturation and spawning of the Pacific salmon (Oncorhynchirs tschawyrscha)
and rainbow trout (Salmo gairdneri). Proc. Natl. Acad. Sci. U.S.A. 45, 886-893.
Harlow. H. J., Thorne. E. T., Williams, E. S., Belden, E. L., and Gem, W. A. (1987a). Adrenal
responsiveness in domestic sheep (Ovis aries) to acute and chronic stressors as predicted
by remote monitoring of cardiac frequency. Can. J. Zool. 65,2021-2027.
Harlow, H. J., Thorne, E. T., Williams, E. S., Belden, E. L., and Gern, W. A. (1987b). Cardiac
frequency: A potential predictor of blood cortisol levels during acute and chronic stress
exposure in Rocky Mountain bighorn sheep (Ovis canadensis). Can. J. Zool. 65, 2028-
2034.
Harlow, H. J.. Lindzey, F. G., Van Sickle, W D., and Gem, W. A. (1992). Stress response of
cougars to nonlethal pursuit by hunters. Can. J. 2001.70, 136-139.
Harper, D. G., Tornatzky. W.. and Miczek, K. A. (1996). Stress induced disorganization of
circadian and ultradian rhythms: Comparisons of effects of surgery and social stress.
Physiol. Behav. 59,409-41 9.
Harrington, F. H., and Veitch, A. M. (1991). Short-term impacts of low-level jet fighter training
on caribou in Labrador. Arctic 44,318-327.
Harrington. F. H.. and Veitch, A. M. (1992). Calving success of woodland caribou exposed
to low-level jet fighter overflights. Arctic 45, 213-218.
Harthoorn, A. M. (1976). “The Chemical Capture of Animals.” Bailliere Tindall. London.
Harthoon, A. M., van der Walt, K.. and Young, E. (1974). Possible therapy for capture
myopathy in captured wild animals. Nature (London) 274,577-577.
Harwood, J., and Grenfell, B. (1990). Long term risks of recurrent seal plagues. Mar. Polltit.
Bitll. 21, 284-287.
Hegner, R. E., and Wingfield. J. C. (1987). Effects of brood-size manipulations on parental
investment. breeding success and reproductive endocrinology of house sparrows. Auk
104,470-480.
Heide-Jorgensen, M. P.. Harkonen. T., Dietz, R., and Thompson, P. M. (1992). Retrospective
of the 1988 European seal epizootic. Dis. A q u a . Org. 13,37-62.
Heinsohn, R. (1992). When conservation goes to the dogs. Trends Ecol. Evol. 7, 214-215.
Helb, H. W., and Huppop, 0. (1991). Herzschlagrate als Map zur Beurteilung des Einflusses
von Storungen vei Vogeln. In “Ornithologenkalender 1992” (E. Bezzel, P. H. Barthel.
H. H. Bergmann. H. W. Helb. and K. Witt, eds.), pp. 217-230. A d a , Wiesbaden.
Hemsworth, P. H., and Barnett, J. L. (1991). The effects of aversively handling pigs either
individually or in groups on their behavior, growth and corticosteroids. Appl. Anim.
Behav. Sci. 30, 61-72.
Henry, J. P.. and Stephens, P. M. (1977). “Stress, Health, and the Social Environment.”
Springer, New York.
Herbst, L. (1991). Pathological and reproductive effects of intraperitoneal telemetry devices
on female armadillos. J. Wildl. Manage. 55, 628-631.
Heyl, C. W., Bigalke, R. C., and Pepler, D. (1988). Captive rearing of the Cape francolin and
prospects for stocking. S. Afr. J . Wildl. Res. 18, 22-29.
Higuchi, T., Negoro, H., and Arita, J. (1989). Reduced responses of prolactin and catechola-
mine to stress in the lactating rat. J. Endocrinol. 122, 495-498.
Hofer, H., East. M. L., and Campbell, K. L. I. (1993). Snares. commuting hyaenas, and
migratory herbivores: Humans as predators in the Serengeti. Symp. Zool. Soc. London
65,347-366.
Hofer, H., Campbell, K. L. I.. East, M. L., and Huish. S. A. (1996). The impact of game meat
hunting on target and non-target species in the Serengeti. In “The Exploitation of Mammal
Populations” (V. Taylor and N. Dunstone, eds.). pp. 117-146. Chapman & Hall, London.
Hoffmann, A. A., and Blows, M. W. (1993). Evolutionary genetics and climate change: Will
animals adapt to global warming? In “Biotic Interactions and Global Change” (P. M.
Kareiva, J. G. Kingsolver, and R. B. Huey, eds.), pp. 165-178. Sinauer, Sunderland, MA.
Hoffmann. A. A., and Parsons, P. A. (1989). Selection for increased desiccation resistance
in Drosophila melanogaster: additive genetic control and correlated responses for other
stresses. Generics 122, 837-846.
Hoffmann, A. A.. and Parsons, P. A. (1991). “Evolutionary Genetics and Environmental
Holloway, G. J., Sibly, R. M., and Povey, S. R. (1990). Evolution in toxin-stressed environments.
Funct. Ecol. 4, 289-294.
Holmes, J. C.. and Zohar, S. (1990). Pathology and host behaviour. In “Parasitism and Host
Behaviour” (C. J. Barnard and J. M. Behnke, eds.). pp. 34-63. Taylor & Francis, London.
Hoover, J. P., Bahr. R. J., Nieves, M. A., Doyle, R. T., Zimmer, M. A,, and Lauzon, S. E.
(1985). Clinical evaluation and prerelease management of American river otters in the
second year of a reintroduction study. J. Am. Vet. Med. Assoc. 187, 1154-1161.
508 HERIBERT HQFER A N D MARION L. EAST
Houston, A. I.. and McNamara. J. M. (1993). A theoretical investigation of fat reserves and
mortality levels in small birds in winter. Ornis Scand. 24, 205-219.
Houston, R. A., and Greenwood, R. J. (1993). Effects of radio transmiters on nesting captive
mallards. J. Wildl. Manage. 57, 703-709.
Hoysak. D. J. and Weatherland, P. J. (1991). Sampling blood from birds: A technique and
assessment of its effect. Condor 93, 746-752.
Huffaker, C. B. (1958). Experimental studies of predation: Dispersion factors and predator-
prey interactions. Hilgardia 27, 343-383.
Hughes, C. S., Gaskell, R. M., Jones, R. C., Bradbury, J. M., and Jordan. F. T. (1989). Effects
of certain stress factors on the re-excretion of infectious laryngotracheitis virus from
latently infected carrier birds. Res. Vet. Sci. 46, 274-276.
Huppop. O., and Hagen, K. (1990). Der Einfluss von Storungen auf Wildtiere am Beispiel der
Herzschlagrate brutender Austernfischer Haematopus ostralegus. Vogelwarfe35,301-310.
Hurlbert, S. H. (1984). Pseudoreplication and the design of ecological field experiments. Ecol.
Monogr. 54,187-21 1.
Hurst, J. L., Barnard, C. J.. Hare. R., Wheeldon, E. B., and West, C. D. (1996). Housing and
welfare in laboratory rats: Time-budgeting and pathophysiology in single-sex groups.
Anim. Behav. 52,335-360.
Ingold, P., Huber, B., Mainini, B., Marbacher, H., Neuhaus, P., Rawyler, A., Roth, M.,
Schnidrig, R., and Zeller, R. (1992). Freizeitaktivitaten-ein gravierendes Problem fur
Tiere? Ornithol. Beob. 89, 205-216.
Ingold, P., Schnidrig-Petrig, R., Marbacher, H., Pfister, U., and Zeller, R.(1996). “Tourismusl
Freizeitsport und Wildtiere im Schweizer Alpenraum.” Schriftenr. Umwelt 262, Bunde-
samt fur Umwelt, Wald iind Landschaft, Bern, Switzerland.
Innes, A. J., El Haj, A. J., and Gobin, J. F. (1986). Scaling of the respiratory cardiovascular
and skeletal muscle systems of the freshwater-terrestrial mountain crab Pseudorhelphusa
garmani garmani. J. Zool., Ser. A 209,595-606.
Insel, T. R., Kinsley, C. H., Mann, P. E., and Bridges, R. S. (1990). Prenatal stress has long-
term effects on brain opiate receptors. Brain Res. 511, 93-97.
International Fund for Animal Welfare (IFAW). (1996). “Report of the International Work-
shop on the Special Aspects of Watching Sperm Whales, Roseau, Commonwealth of
Dominica, East Caribbean 8th January-11th January 1996.” Int. Whal. Comm., Sci. Comm.
Doc. SC/48/0 26. IFAW, Crowborough, East Sussex, UK.
International Fund for Animal Welfare (IFAW), Tethys Research Institute and Europe
Conservation. (1996). “Report of the Workshop on the Scientific Aspects of Managing
Whale Watching, Montecastello di Vibio, Italy.” IFAW, Crowborough, East Sussex, UK.
Ivanovici, A. M., and Wiebe, W. J. (1981). Towards a working “definition” of ‘stress’: A
review and critique. I n “Stress Effects on Natural Ecosystems” (G. W. Barrett and R.
Rosenberg, eds.), pp. 13-27. Wiley, New York.
Jacobson, S. K., and Lopez, A. F. (1994). Biological impacts of ecotourism: Tourists and
nesting turtles in Tortuguero National Park, Costa Rica. Wildl. SOC. Bull. 22, 414-419.
Janik, V. M., and Thompson, P. M. (1996). Changes in surfacingpatterns of bottlenose dolphins
in response to boat traffic. Mar. Mammal. Sci. 12, 597-602.
Jarvi, T. (1990). Cumulative acute physiological stress in Atlantic salmon smolts: The effect
of osmotic imbalance and the presence of predators. Aquaculfure 89, 337-350.
Jarvi, T., and Uglem, I. (1993). Predator training improves the anti-predator behaviour of
hatchery reared Atlantic salmon (Salmo salar) smolt. Nord. J. Freshwater Res. 68,63-71.
Jenkins. F. J., and Baum, A. (1995). Stress and reactivation of latent herpes simplex virus-a
fusion of behavioral medicine and molecular biology. Ann. Behav. Med. 17, 116-123.
Jenssen, B. M., Skaare, J. U., Ekker. M., von Graven, D., and Silverstone, M. (1994). Blood
sampling as a non-destructive method for monitoring levels and effects of organochlorines
(PCB and DDT) in seals. Chemosphere 28, 3-10.
Jenssen, B. M., Skaare, J. U., Woldstad, S., Nastad, A. T.. Haugen, 0..Kloven, B., and Sormo,
E. G. (1995). Biomarkers in blood to assess effects of polychlorinated biphenyls in free-
living grey seal pups. Dev. Mar. Bid. 4, 607-615.
Jeppesen, J. L. (1987). Impact of human disturbance on home range, movements and activity
of red deer Cervus eluphus in a Danish environment. Dan. Rev. Game Biol. 13(2), 1-38.
Jessup, D. A. (1992). Veterinary contributions toward improving capture, medical manage-
ment, and anesthesia of free-ranging wildlife. J. Am. Vet. Med. Assoc. 200, 653-658.
Johnson, K., Dalton. R., and Burley, N. (1993). Preferences of female American goldfinches
Carduelis tristis for natural and artificial male traits. Behuv. Ecol. 4, 138-143.
Jones, C. A,, Refsal, K. R., Lippert, A. C., Nachreiner, R. F., and Schwacha, M. M. (1990).
Changes in the adrenal cortisol secretion as reflected in the urinary cortisollcreatinine
ratio in dogs. Domest. Anim. Endocrinol. 7, 559-572.
Jones, T. O.,Bourne, N. F., Bower, S. M., and Iwama, G. K. (1993). Effect of repeated
sampling on haemolymph pH, partial pressure of oxygen and haemocyte activity in the
pacific oyster Crussostreu gigus Thunberg. J. Exp. Mar. Biol. Ecol. 167, 1-10,
Jordan, R. H., and Burghardt, G. M. (1986). Employing an ethogram to detect reactivity of
black bears (Ursus americanus) to the presence of humans. Ethology 73,89-115.
Joubert. F. G., and Stander, P. E. (1990). Capture myopathy in an African lion. Madoquu
17,51-52.
Jurke, M. H., Pryce, C. R., Doebeli, M., and Martin, R. D. (1994). Non-invasive detection
and monitoring of pregnancy and the postpartum period in Goeldi’s monkey (Cullimico
goeldii) using urinary pregnanediol-3-alpha-glucuronide. Am. J. Primutol. 34, 319-331.
Kania, W.(1992). Safety of catching adult European birds at the nest. The Ring 14,5-50.
Kareiva, P. M., Kingsolver, J. G., and Huey, R. B. (1993). “Biotic Interactions and Global
Change.” Sinauer, Sunderland, MA.
Karr, J. R. (1990). Avian survival rates and the extinction process on Barro Colorado Island.
Panama. Conserv. Biol. 4, 391-397.
Kaufmann, J. H. (1983). On the definitions and functions of dominance and territoriality.
Biol. Rev. Cambridge Philos. Soc. 58, 1-20.
Keller, V. (1989a). Variations in the response of great crested grebes Podiceps cristutus to
human disturbance: A sign of adaptation? Biol. Conserv. 49, 31-45.
Keller, V. (1989b). Egg-covering behaviour by great crested grebes Podiceps cristutus. Ornis
Scand. 20, 129-131.
Kelley, T. M., and Bramblett, C. A. (1981). Urine collection from vervet monkeys by instrumen-
tal conditioning. Am. J. Primutol. 1, 95-97.
Kennedy. S., Smyth, J. A,. Cush. P. F., McCullough, S. J., Allan, G. M., and McQuaid. S.
(1988). Viral distemper now found in porpoises. Nature (London) 336, 21-21.
Kenward, R. (1987). “Wildlife Radio Tagging.” Academic Press, London.
Kernohan. B. J., Jenks, J. A,, Naugle. D. E.. and Millspaugh. J. J. (1996). Estimating 24-h
habitat use patterns of white-tailed deer from diurnal use. J. Environ. Munuge. 48,299-303.
Kestin, S. C. (1995). Welfare aspects of the commercial slaughter of whales. Anim. Welfare
4, 11-27.
Keverne, E. B., Meller, R. E., and Eberhardt. J. A. (1982). Dominance and subordination:
Concepts or physiological states? I n “Advanced Views on Primate Biology” (0.Chiarelli,
ed.), pp. 81-94. Springer, New York.
Kiecolt-Glaser. J. K., Glaser, R., Gravenstein, S., Malarkey, W. B., and Sheridan, J. (1996).
Chronic stress alters the immune response to influenza virus vaccine in older adults. Proc.
Nutl. Acud. Sci. U.S.A.93, 3043-3047.
Kihlstrom, J . E.. Olsson, M., Jensen, S., Johansson, A,. Ahlborn, J., and Bergman, A. (1992).
Effect of PCB and different fractions of PCB on the reproduction of the mink (Mustela
vison). Ambio 21, 563-569.
King, D. A.. and Stewart, W. P. (1996). Ecotourism and comodification: Protecting people
and places. Biodiversity Conserv. 5, 293-305.
Kirschbaum, C., and Hellhammer, D. H. (1994). Salivary cortisol in psychoneuroendocrine
research: Recent developments and applications. Psychoneuroendocrinolo~y19,313-333.
Klomp, N. I., and Furness, R. W. (1992). Non-breeders as a buffer against environmental
stress: Declines in number of great skuas on Foula, Shetland, and prediction of future
recruitment. J. Appl. Ecol. 29, 341-348.
Knight, R. L., and Gutzwiller. K. J. (1995). “Wildlife and Recreationists.” Island Press,
Washington, DC.
Knox, C. M., Hattingh, J.. and Raath, J. P. (1990). The effect of tranquilizers on the immediate
responses to repeated capture and handling of boma-kept impala. Comp. Biochem. Phys-
iol. C. 95C, 247-251.
Kock, M. D.. Jessup, D. A., Clark, R. K., Franti, C. E., Weaver, R. A. (1987a). Capture
methods in five subspecies of free-ranging bighorn sheep: An evaluation of drop-net,
drive-net, chemical immobilization and the net-gun. J. Wildl. Dis. 23, 634-640.
Kock. M. D., Clark, R. K., Franti, C. E., Jessup, D. A,, and Wehausen, J. D. (1987b).
Effects of capture on biological parameters in free-ranging bighorn sheep Ovis canadensis:
Evaluation of normal. stressed and mortality outcomes and documentation of postcapture
survival. J. Wildl. Dis. 23, 652-662.
Kock, M. D., Morkel, P.. Atkinson. M., and Foggin, C. (1995). Chemical immobilization of
free-ranging white rhinoceros (Ceratotherium simum simum) in Hwange and Matobo
National Parks. Zimbabwe, using combinations of etorphine (M99), fentanyl, xylazine,
and detomidine. J. Zoo Wildl. Med. 26, 207-219.
Kohane, M. J.. and Parsons. P. A. (1988). Domestication. Evolutionary change under stress.
Evol. Biol. 23, 31-48.
Kohn, M., Knauer, F., Stoffella, A,, Schroder, W., and Paabo, S. (1995). Conservation genetics
of the European brown bear: A study using excremental PCR of nuclear and mitochondria1
sequences. Mol. Ecol. 4, 95-103.
Kooijman, S. A. L. M., van der Hoeven. N., and van der Werf, D. C. (1989). Population
consequences of a physiological model for individuals. Funct. Ecol. 3, 325-336.
Korpimaki, E. (1990). Body mass of breeding Tengmalm’s owls Aegolius funereus: Seasonal,
between-year, site and age-related variation. Ornis Scand. 21, 169-178.
Kortner, G., and Geiser, F. (1995). Effect of photoperiod and ambient temperature on activity
patterns and body weight cycles of mountain pygmy-possums, Burramysparvus (Marsupi-
aha). J. Zool. 235, 311-322.
Kramer, K., van Acker, S. A,, Voss, H. P., Grimbergen, J. A,, and van der Vijgh, W. J., and
Bast. A. (1993). Use of telemetry to record electrocardiogram and heart rate in freely
moving mice. J. Pharmacol. Toxicol. Methods 30, 209-215.
Kramskaya,T. A,, Chetverikova, L. K., Frolov, B. A,, and Polyak, R. Y. A. (1991). Disturbance
of resistance to a repeated infection by homological influenza virus in mice after stress.
Vestn. Akad. Med. Nauk SSSR 4, 47-50.
Krebs, J. R., and Davies. N. B. (1997). “Behavioural Ecology.” Blackwell, Oxford.
Krebs, R. A.. and Loeschcke, V. (1994). Response to environmental change: Genetic variation
and fitness in Drosophila buzzarii following temperature stress. In “Conservation Genet-
ics” (V. Loeschcke, J. Tomiuk, and s. K. Jain, eds.). pp. 309-321. Birkhauser, Basel and
New York.
Kreeger, T. J., Faggella, A. M.. Seal, U. S., Mech, L. D., Callahan, M., and Hall, B. (1987).
Cardiovascular and behavioral responses of gray wolves to ketamine-xylazine immobiliza-
tion and antagonism by yohimbine. J. Wildl. Dis. 23, 463-470.
Kreeger, T. J., Seal, U. S., Tester, J. R., Callahan, M., and Beckel, M. (1990). Physiological
responses of red foxes Vulpes viclpes to surgery. J. Wildl. Dis. 26, 219-224.
Kreger, M. D., and Mench, J. A. (1993). Physiological and behavioral effects of handling and
restraint in the ball python (Pyrhon regius) and the blue-tongued skink (Tiliqua scin-
coides). Appl. Anim. Behav. Sci. 38, 323-336.
Krylov, V. 1. (1990). Ecology of the Caspian seal. Finn. Game Res. 47, 32-36.
Kubokawa, K. (1993). Field endocrinology in Japan. J. Reprod. Dev. 39, j117-jl25 (in Jap-
anese).
Lamontagne, L., Sadi, L., and Joyal, R. (1989). Serological evidence of bovine herpesvirus
1-related virus infection in the white-tailed deer population on Anticosti Island Quebec,
Canada. J. Wildl. Dis. 25, 202-205.
Lampo, M., and Bayliss, P. (1996). Density estimates of cane toads from native populations
based on mark-recapture data. Wildl. Res. 23, 305-315.
Lamprecht, J., and Hofer, H. (1994). Cooperation among sunfish: Do they have the cognitive
abilities? Anim. Behav. 47, 1457-1458.
LaPatra. S. E., Groff, J. M., Jones, G. R., Munn, B., Patterson. T. L., Holt, R. A,, Hauck,
A. K., and Hedrick, R. P. (1993). Occurrence of white sturgeon iridovirus infections
among cultured white sturgeon in the pacific northwest. Aquaculture 126.
Lash, G. Y. B., and Horwich, R. H. (1996). Successes and problems of community-managed
ecotourism in protecting primates: A case study in Belize. Abstr., Congr. Int. Primatol.
Soc.. 16th, 470.
Laudenslager, M., Berger, C., Bocchia, M., Worlein, J., and Reite. M. (1996). Innate immunity
in young macaques during brief social separation experiences. Absrr., Congr. Inf.Primatol.
SOC., 16th, 757.
Laurenson, M. K., and Caro, T. M. (1994). Monitoring the effects of non-trivial handling in
free-living cheetahs. Anim. Behav. 47, 547-557.
Lavigne, D. M., and Schmitz, 0.J. (1990). Global warming and increasing population densities:
A prescription for seal plagues. Mar. Pollur. Bull. 21, 280-284.
Lee, A. K., and Cockburn, A. (1985). “Evolutionary Ecology of Marsupials.” Cambridge
University Press, Cambridge, UK.
Lee, A. K., and McDonald, I. R. (1985). Stress and population regulation in small animals.
Oxford Rev. Reprod. Biol., pp. 261-304.
Lee, A. K., Bradley, A. J., and Braithwaite, R. W. (1977). Corticosteroid levels and male
mortality in Antechinus stuarfii. In “The Biology of Marsupials” (B. Stonehouse and D.
Gilmore, eds.), pp. 209-220. Macmillan, London.
Le Maho, Y., Karmann, H., Briot, D., Handrich, Y., Robin, J. P., Mioskowski, E., Cherel,
Y., and Farni, J. (1992). Stress in birds due to routine handling and a technique to avoid
it. Am. J. Physiol. 263, R775-R781.
Lemaire, V., and Mormkde, P. (1995). Telemetered recording of blood pressure and heart
rate in different strains of rats during chronic social stress. Physiol. Behav. 58,1181-1 188.
Lima, S. L. (1986). Predation risk and unpredictable feeding conditions: Determinants of
body mass in birds. Ecology 67, 377-385.
Line, S. W., Kaplan, J. R., Heise, E. R., Hilliard, J. K.. Cohen, S., Rabin, B. S.. and Manuck,
S. B. (1996). Effects of social reorganization on cellular immunity in male cynomolgus
monkeys. Am. J. Primatol. 39,235-249.
Logan, C . A., and Wingfield, J. C. (1995). Hormonal correlates of breeding status, nest
construction, and parental care in multiple-brooded northern mockingbirds, Mimum
polyglotfos. Horm. Behav. 29, 12-30.
Lombardi, C. M.. and Hurlbert. S. H. (1996). Sunfish cognition and pseudoreplication. Anini.
Behav. 52,419-422.
Lowenstine, L. J.. and Fry, D. M. (1985). Adenovirus-like particles associated with intranuclear
inclusion bodies in the kidney of a common murre (Uria aalge). Avian Dis. 29, 208-213.
Lubach. G. R., Coe. C. L., and Ershler, W. B. (1995). Effects of early rearing environment
on immune responses of infant rhesus monkeys. Brain. Behav., imnirtn. 9, 31-46.
Lubach, G. R., Karaszewski, J. W., and Schneider. M. L. (1996). Pregnancy conditions influence
immune and behavioral development of rhesus monkey infants. Abstr., Congr. Int. Prinza-
to/. Soc., 16th. 103.
Luttrell. L., Acker, L., Urben. M., and Reinhardt, V. (1994). Training a large troop of rhesus
macaques to co-operate during catching: Analysis of the time investment. Anim. Welfare
3, 135-140.
Lyman, C. P., Willis, J. S., Malan, A., and Wang, L. C. H. (1982). “Hibernation and Torpor
in Mammals and Birds.” Academic Press, London.
Macario, A. J. L. (1995). Heat-shock proteins and molecular chaperones: Implications for
pathogenesis, diagnostics, and therapeutics. Itit. J . Clin. Lab. Res. 25, 59-70.
MacArthur, R. A., Geist. V.. and Johnson, R. H. (1982). Cardiac and behavioural responses
of mountain sheep to human disturbance. J . W i l d . Manage. 46,351-358.
MacArthur, R. A., Geist, V., and Johnson, R. H. (1986). Cardiac responses of bighorn sheep
to trapping and radio instrumentation. Can. J. 2001.64, 1197-1200.
Mace, R. D., Waller. J. S., Manley, T. L.. Lyon, L. J., and Zuuring, H. (1996). Relationships
among grizzly bears, roads and habitat in the Swan Mountains, Montana. J . Appl. Ecol.
33, 1395-1404.
Machlis, L., Dodd, P. W. D., and Fentress. J. C. (1985). The pooling fallacy: Problems arising
when individuals contribute more than one observation to the data set. Z . Tierpsychol.
68,201-214.
Madej, A,, Forsberg. M., and Edqvist, L. E. (1992). Urinary excretion of cortisol and oestrone
sulfate in pregnant mink fed PCB and fractions of PCB. Anzbio 21, 582-585.
Madsen, J. (1995). Impacts of disturbance on migratory waterfowl. ibis 137, Suppl, S67-S74.
Maestripieri D.. Badiani, A,, and Puglisiallegra S. (1991). Prepartal chronic stress increases
anxiety and decreases aggression in lactating female mice. Behav. Neurosci. 105,663-668.
Maestripieri, D., Schino. G., Aureli. F., and Troisi, A. (1992). A modest proposal: Displacement
activities as an indicator of emotions in primates. Anim. Behav. 44,967-979.
Mainini, B., Neuhaus, P., and Ingold, P. (1993). Behavior of marmots Marmota marmota
under the influence of different hiking activities. B i d . Conserv. 64, 161-164.
Mallet, J., Longino, J . T., Murawski, D., Murawski. A,. Simpson de Gamboa. A. (1987).
Handling effects in Heliconiiis: Where do all the butterflies go? J. Aninz. Ecol. 56,377-386.
Manly, B. F. J. (1992). “The Design and Analysis of Research Studies.” Cambridge University
Mansergh, I. M., and Broome, L. (1994). “The Mountain Pygmy-possum of the Australian
Alps.” New South Wales University Press, New Kensington. Australia.
Mansergh, I. M.. and Scotts, D. J . (1989). Habitat continuity and social organization of the
mountain pygmy-possum restored by tunnel. J . Wildl. Manage. 53,701-707.
Markovitz, H. (1982). “Behavioral Enrichment in Zoos.” Van Nostrand-Reinhold, New York.
Marsden. J. E., and May, B. (1984). Feather pulp: A non-destructive sampling technique for
electrophoretic studies of birds. Artk 101, 173-175.
Martel, F. L., Lyons, D. M., and Levine, S. (1996). Long-term consequences of postnatal
stress: Behavioral and pituitary-adrenal responses in squirrel monkeys. Abslr., Congr.
Int. Primarol. Soc., 16th, 128.
Masman. D.. Dijkstra. C., Daan, S.. and Bult, A. (1989). Energetic limitation of avian parental
effort: Field experiments in the kestrel Falco rinnuncubcs. J. Evol. Biol. 2,435-455.
Mason, G. J. (1991). Stereotypes: A critical review. Anim. Behav. 41, 1015-1037.
Matt, K. S.. Soares. M. J., Talamantes. F., and Bartke, A. (1983). Effects of handling and
ether anesthesia on serum prolactin levels in the golden hamster. Proc. Soc. Exp. B i d .
Med. 173,463-466.
May. R. M. (1995). The cheetah controversy. Nariire (London) 374, 309-310.
Mayr. B.. Honig, A.. and Gutbrod, F. (1990). Untersuchungen iiber die Wirksamkeit und
Unschadlichkeit einer Schutzimpfung gegen Parvovirus bzw. Tollwut bei narkotisierten
Hundewelpen. Tieriirztl. Praxis 18, 165-169.
Mays, N. A., Vleck. C. M.. and Dawson. J. W. (1991). Plasma luteinizing hormone, steroid
hormones, behavioral role, and nest stage in the cooperatively breeding Harris’ hawk
(Pambuteo itnicincrus). Auk 108, 619-637.
McCarty. R. (1983). Stress, behaviour and the sympathetic-adrenal medullary system. In
“Stress and Alcohol Use” (L. A. Pohorecky and J. Brick, eds.), pp. 7-22. Elsevier,
New York.
McConway. K. (1992). The number of subjects in animal behavior experiments: is Still right?
In “Ethics in Research on Animal Behaviour” (M. S. Dawkins and L. M. Gosling, eds.),
pp. 35-58. Academic Press, London.
McCune. S. (1995). The impact of paternity and early socialization on the development of
cats’ behavior to people and novel objects. Appl. Anim. Behav. Sci. 45, 109-124.
McDonald. I. R.. Lee. A. K.. Than, K. A,. and Martin, R. W. (1986). Failure of glucocorticoid
feedback in males of a population of small marsupials (Antechinus swainsonii) during
the period of mating. J. Endocrinol. 108, 63-68.
McLean, R. G. (1975). Racoon rabies. I n “The Natural History of Rabies 2” (G. M. Baer.
ed.). pp. 53-77. Academic Press, New York.
McLellan, B. N.. and Shackleton, D. M. (1988). Grizzly bears and resource-extraction indus-
tries: Effects of roads on behaviour. habitat use and demography. J. Appl. Ecol. 25,
45 1-460.
McLellan, B. N., and Shackleton, D. M. (1989). Grizzly bears and resource-extraction indus-
tries: Habitat displacement in response t o seismic exploration, timber harvesting and
road maintenance. J . Appl. Ecol. 26,371-380.
McNamara, J. M., and Houston, A. I. (1992). State-dependent life history theory and its
implications for optimal clutch size. Evol. Ecol. 6, 170-185.
McNamara, J . M., and Houston, A. I. (1996). State-dependent life-histories. Nature (London)
380, 215-221.
Mead, R. (1991). “The Design of Experiments. Statistical Principles for Practical Application.”
Cambridge University Press, Cambridge, UK.
Meaney. M. J., Aitken, D. H., Bhatnagar, S., and Sapolsky, R. M. (1991). Postnatal handling
attenuates certain neuroendocrine, anatomical. and cognitive dysfunctions associated with
aging in female rats. Neurohiol. Aging 12, 31-38.
Meaney, M. J., Bhatnagar. S., Diorio, J.. Larocque, S.. Francis. D., O’Donnell. D.. Shanks,
N.. Sharma. S., Smythe, J., and Viau. V. (1993). Molecular basis for the development of
individual differences in the hypothalamic-pituitary-adrenal stress response. Cell. Mol.
Neurobiol. 13, 321 -347.
Mech, L. D.. Seal. U. S.. and Delgiudice. G. D. (1987). Use of urine in snow to indicate
condition of wolves. J . Wildl. Manage. 51, 10-13.
Medleau. L., Crowe, D. T., and Dawe, D. L. (1983). Effect of surgery on the in vitro response of
canine peripheral blood lymphocytes t o phytohemagglutinin. Am. J. Vet. Res. 44,859-860.
Mees, J., and Reijnders. P. J. H. (1994). The harbour seal, Phoca vitulina, in the Oosterschelde:
Decline and possibilities for recovery. Hydrohiolagy 282-283, 547-555.
Meire, P. M., and Dereu, J. (1990). Use of the abundancelbiomass comparison method for
detecting environmental stress: Some considerations based on intertidal macrozoobenthos
and bird communities. J. Appl. Ecol. 27, 210-223.
Mendoza, S. P., and Mason, W. A. (1986). Contrasting responses t o intruders and to involuntary
separation by monogamous and polygynous New World monkeys. Physiol. Behav. 38,
795-801.
Merkle, M. S., and Barclay, R. M. R. (1996). Body mass variation in breeding mountain
bluebirds. Sialia currucoides: Evidence of stress or adaptation for flight? J . Anim. Ecol.
65,401-413.
Merola, M. (1994). A reassessment of homozygosity and the case for inbreeding depression
in the cheetah. Acinonyxjrbarus: Implications for conservation. Conserv. Biol. 8,961-971.
Meyers, T. R.. Hauck, A. K., Blankenbeckler, W. D., and Minicucci, T. (1986). First report
of viral erythrocytic necrosis in Alaska, USA, associated with epizootic mortality in pacific
herring Chipea harengus pallasi Valenciennes. J. Fish Dis. 9, 479-491.
Miller, M. W., Thompson, Hobbs. N., and Sousa, M. C. (1991). Detecting stress responses
in Rocky Mountain bighorn sheep (Ovis canadensis canadensis): Reliability of cortisol
concentrations in urine and feces. Can. J. 2001.69, 15-24.
Miller, P. S. (1994). Is inbreeding depression more severe in a stressful environment? Zoo
Biol. 13, 195-208.
Millikin, R. L., and Smith, J. N. M. (1990). Sublethal effects of fenitrothion on forest passerines.
J . Appl. Ecol. 27,983-1000.
Moberg, G. P. (1985). Biological response to stress: Key to assessment of animal well-being?
In “Animal Stress” (G. P. Moberg, ed.), pp. 27-49. Am. Physiol. SOC.. Bethesda, MD.
Moberg, G. P. (1987). Problems in defining stress and distress in animals. J . Am. Vet. Med.
Assoc. 191, 1207-1211.
Moller, A. P. (1992). Female choice selects for male sexual tail ornaments in the monogamous
swallow. Nature (London) 357,238-240.
Moreno, J . (1989). Strategies of mass changes in breeding birds. B i d . J . Linn. Soc. 37,297-310.
Morimoto, R. I., Tissitres, A,, and Georgopoulos. C. (1990). “Stress Proteins in Biology and
Medicine.” Cold Spring Harbor Lab. Press, Cold Spring Harbor, NY.
Morton, D. J.. Anderson, E., Foggin, C. M., Kock, M. D., and Tiran, E. P. (1995). Plasma
cortisol as an indicator of stress due to capture and translocation in wildlife species. Vet.
Rec. 136, 60-63.
Mostl, E., Choi. H. S., Wurm, W.. Ismail, N.. and Bamberg, E. (1984). Pregnancy diagnosis
in cows and heifers by determination of oestradiolL17cy in faeces. Br. Vet. J . 140,287-291.
Muneer. M. A.. Farah, I. O., Pomeroy, K. A., Goyal, S. M., and Mech, L. D. (1988). Detection
of parvoviruses in wolf feces by electron microscopy. J. Wildl. Dis. 24, 170-172.
Munger, J . C., and Karasov, W. H. (1989). Sublethal parasites and host energy budgets:
Tapeworm infection in white-footed mice. Ecology 70, 904-917.
Munger, J . C., and Karasov, W. H. (1991). Sublethal parasites in white-footed mice: Impact
on survival and reproduction. Can. J. Zoo1 69, 398-404.
Myrberg, A. A. (1990). The effects of man-made noise o n the behavior of marine animals.
Environ. Int. 16, 575-586.
Nakata, H.. Tanabe, S., Tatsukawa, R.. Amano, M., Miyazaki, N., and Petrov, E. A. (1995).
Persistent organochlorine residues and their accumulation kinetics in Baikal seal (Phoca
sibirica) from Lake Baikal. Russia. Environ. Sci. Technol. 29, 2877-2885.
Narita, M., Nanba, K., Haritani, M., and Kawashima, K. (1992). lmmunopathology in Aujesz-
ky’s disease: Virus-infected pigs exposed to fluctuating temperatures. J. Comp. Pathol.
107, 221-229.
Nevo, E., Beiles, A,. and Ben-Shlomo. R. (1984). The evolutionary significance of genetic
diversity: Ecological, demographic and life history correlates. In “Evolutionary Dynamics
of Genetic Diversity” (G. S. Mani, ed.), pp. 13-213. Springer, Berlin.
Niezgoda, J.. Bobek, S.. Wronska-Fortuna, D., and Wierzchos. E. (1993). Response of
sympatho-adrenal axis and adrenal cortex to short-term restraint stress in sheep. J. Vet.
Med. AM, 63 1-638.
Nimon. A. J., and Dalziel, F. R. (1992). Cross-species interaction and communication: A study
method applied to captive siamang Hylobates syndactylus and long-billed corella Cacatua
tenuirostris contacts with humans. Appl. Anim. Behav. Sci. 33, 261-272.
Ninion, A. J., Oxenham, R. K. C., Schroter, R. C.. and Stonehouse, B. (1994). Measurement
of resting heart rate and respiration in undisturbed and unrestrained, incubating gentoo
penguins (Pygoscelzs papua). J. fhysiol. (Cambridge) 481,57P-58P.
Nimon, A. J., Schroter, R. C., and Stonehouse, B. (1995). Heart rate of disturbed penguins.
Nature (London) 374, 415-415.
Nover, L. (1991). “The Heat Shock Response.” CRC Press, Boca Raton, FL.
Obrecht. H. H., Pennycuick, C. J., and Fuller, M. R. (1988). Wind tunnel experiments to
assess the effect of back-mounted radio transmitters on bird body drag. J. Exp. B i d .
135,256-273.
O’Brien, S. J. (1994). The cheetah’s conservation controversy. Conserv. Biol. 8, 1153-1 155.
O’Leary, H. (1996). Contrasts in diet amongst barbary macaques on Gibraltar: Human influ-
ences. Anim. Welfare 5, 177-188.
Olsen, N. J., Nicholson, W. E., Debold, C. R., and Orth, D. N. (1992). Lymphocyte-derived
adrenocorticotropin is insufficient to stimulate adrenal steroidogenesis in hypophysecto-
mized rats. Endocrinology (Bakimore) 130, 2113-2119.
Olsson, G. E., Willebrand, T., and Smith, A. A. (1996). The effects of hunting on willow
grouse Lagopus lagopus movements. Wildl. B i d . 2, 11-15.
Olsson, M., Karlsson, B.. and Ahnland, E. (1994). Diseases and environmental contaminants
in seals from the Baltic and the Swedish west coast. Sci. Total Environ. 154, 217-227.
Oppliger, A,. Christe, P., and Richner, H. (1996a). Clutch size and malaria resistance. Nature
(London) 381,565.
Oppliger, A.. Christe. P., and Richner, H. (1996b). Nature (London) 382, 502.
O’Reilly, K. M.. and Wingfield, J. C. (1995). Spring and autumn migration in arctic shorebirds:
Same distance, different strategies. Am. Zool. 35, 222-233.
Oshea, T. J.. Rathbum, G. B., Asper, E. D., and Searles, S. W. (1985). Tolerance of West-
Indian manatees Trichechus manatus to capture and handling. Biol. Conserv. 33,335-350.
Osterhaus. A. D. M. E., de Swart, R. L.. Vos. H. W., Ross, P. S., Kenter, M. J., and Barrett,
T. (1995). Morbillivirus infections of aquatic mammals: Newly identified members of the
genus. Vet. Microbial. 44,219-227.
Ozherelkov, S. V., Khozinsky. V. V.. and Semenov, B. F. (1990). Replication of Langat
virus in immuno-competent cells of mice subjected to immobilization stress. Acfa Virol.
34,291 -294.
Parmar. M. K. B.. and Machin, D. (1995). “Survival Analysis. A Practical Approach.” Wi-
ley. Chichester.
Parsell. D. A., Kowal, A . S.. Singer. M. A,, and Lindquist, S. (1994). Protein disaggregation
mediated by heat-shock protein hsp 104. Nature (London) 372,475-478.
Parsons. P. A. (1988a). Evolutionary rates: Effects of stress upon recombination. B i d . J. Linn.
Soc. 35,49-68.
Parsons. P. A. (1988b). Behavior, stress and variability. Behav. Genet. 18, 293-308.
Parsons. P. A. (1989a). Environmental stresses and conservation of natural populations. Annu.
Rev. Ecol. Syst. 20, 29-50.
Parsons, P. A. (1989b). Conservation and global warming: A problem in biological adaptation

to stress. Ambio 18, 322-325.
Parsons, P. A. (1990a). Fluctuating asymmetry and stress intensity. Trends Ecol. Evol. 5,97-98.
Parsons, P. A. (1990b). Fluctuating asymmetry: An epigenetic measure of stress. B i d . Rev.
Cambridge Philos. SOC.65, 131-145.
Parsons, P. A. (1990~).The metabolic cost of multiple environmental stresses: Implications
for climatic change and conservation. Trends Ecol. Evol. 5, 315-317.
Parsons, P. A. (1991). Evolutionary rates, stress and species boundaries. Annu. Rev. Ecol.
Syst. 22, 1-18.
Parsons, P. A. (1992). Fluctuating asymmetry: A biological monitor of environmental and
genomic stress. Heredity 68, 361-364.
Parsons, P. A. (1993a). Stress, extinctions and evolutionary change: From living organisms
to fossils. B i d . Rev. Cambridge Philos. Soc. 68, 313-333.
Parsons, P. A. (1993b). The importance and consequences of stress in living and fossil popula-
tions from life-history variation to evolutionary change. Am. Nut. 142, S5-S20.
Parsons, P. A. (1993~).Evolutionary adaptation and stress: Energy budgets and habitats
preferred. Brhav. Genet. 23, 231-238.
Parsons, P. A. (1994). Habitats, stress, and evolutionary rates. J. Evol. B i d . 7, 387-397.
Parsons, P. A. (1995). Evolutionary response to drought stress: Conservation implications.
B i d . Conserv. 74, 21-27.
Paterson, A. M., and Pearce, G. P. (1992). Growth response to humans and corticosteroids
in male pigs housed individually and subjected to pleasant. unpleasant or minimal handling
during rearing. Appl. Anim. Behav. Sci. 34, 315-328.
Peakall, D. B. (1993). DDE-induced eggshell thinning: An environmental detective story.
Environ. Rev. 1, 13-20.
Peters, R. L., and Lovejoy. T. E. (1992). “Global Warming and Biological Diversity.” Yale
University Press, New Haven, CT.
Phillippi-Falkenstein, K., and Clarke, M. R. (1992). Procedure for training corral-living rhesus
monkeys for fecal and blood sample collection. Lab. Anim. Sci. 42, 83-85.
Pietz, P. J., Krapu, G. L., Greenwood, R. J.. and Lokemoen, J. T. (1993). Effects of harness
transmitters on behavior and reproduction of wild mallards. J. Wildl. Manage. 57,696-703.
Pokhil’ko, A. V., Kramskaia, T. A,. and Poliak, R. I. A. (1995). Stress and viral infection:
Dynamics of expression of influenza A viral antigens during immobilization stress. Vopr.
Virusol. 40, 76-79.
Pollard, J. C., and Littlejohn, R. P. (1995). Effects of lighting on heart rate and positional
preferences during confinement in farmed red deer. Anim. Welfare 4,329-337.
Porter, J. M., and Coulson, J. C. (1987). Long-term changes in recruitment to the breeding
group, and the quality of recruits at a kittiwake ROsu tridactyla colony. J. Anim. Ecol.
56, 675-689.
Pottinger, T. G., and Calder, G. M. (1995). Physiological stress in fish during toxicological
procedures: A potentially confounding factor. Environ. Toxicol. Water Qual. 10,135-146.
Powell, P. C., and Davison, T. F. (1986). Induction of Marek’s disease in vaccinated chickens
by treatment with betamethasone or corticosterone. Isr. J. Vet. Med. 42, 73-78.
Preston. S. D., Aureli, F., and Insel, T. R. (1996). Heart rate telemetry as a window to the
emotionality of freely-moving rhesus macaques: a pilot study. Abstr., Congr. Int. Primatol.
Soc., 16th, 350.
Price. S., Sibly, R. M., and Davies, M. H. (1993). Effects of behaviour and handling on heart
rate in farmed red deer. Appl. Anim. Behav. Sci. 37, 111-123.
Primack, R. B. (1993). “Essentials of Conservation Biology.” Sinauer, Sunderland. MA.
Putman, R. J. (1995). Ethical considerations and animal welfare in ecological field studies.
Eiodiversify Conserv. 4, 903-915.
Quiatt, D., Rutan, B., Stone, T., and Reynolds, V. (1994). “Bundongo Forest Chimpanzees:
Composition of Feeding Groups During the Rainy Season, with Attention t o the Spatial
Integration of Disabled Individuals.” Budongo For. Proj. Rep. 21. Institute of Biological
Anthropology, Oxford.
Rasmussen, K. L. R. (1991). Identification, capture and biotelemetryof socially living monkeys.
Lab. Anim. Sci. 41,350-354.
Reijnders, P. J. H. (1980). Organochlorine and heavy metal residues in harbour seals from
the Wadden Sea and their possible effects on reproduction. Nefh. J. Sea Res. 14, 30-65.
Reijnders, P. J . H. (1986). Reproductive failure in common seals feeding on fish from polluted
coastal waters. Nature (London) 324,456-457.
Reijnen, R., and Foppen, R. (1994). The effects of car traffic on breeding bird populations
in woodland. I. Evidence of reduced habitat quality for willow warblers (Phylloscopw
trochilus) breeding close to a highway. J . Appl. Ecol. 31,235-94.
Reijnen, R., and Foppen, R. (1995). The effects of car traffic on breeding bird populations
in woodland. 1V. Influence of population size on the reduction of density close to a
highway. J . Appl. Ecol. 32, 481-491.
Reijnen, R., Foppen, R., ter Braak, C., and Thissen, J. (1995). The effects of car traffic on
breeding bird populations in woodland. 111. Reduction of density in relation to the
proximity of main roads. J. Appl. Ecol. 32, 187-202.
Reinhardt, V. (1991a). Impact of venipuncture on physiological research conducted in con-
scious macaques. J. Exp. Anim. Sci. 34,212-217.
Reinhardt, V. (1991b). Training adult male rhesus monkeys to actively cooperate during in-
homecage venipuncture. Anim. Technol. 42, 11-17.
Reinhardt, V., Cowley, D., Eisele, S., and Scheffler, J. (1991). Avoiding undue cortisol re-
sponses to venipuncture in adult male rhesus macaques. Anim. Technol. 42,8346.
Reinhardt, V., Liss, C., and Stevens, C. (1995). Restraint methods of laboratory non-human
primates: A critical review. Anim. Welfare 4, 221-238.
Renshaw, E. (1991). “Modelling Biological Populations in Space and Time.” Cambridge
University Press, Cambridge, UK.
Richardson, W. J., Greene, C. R., Malme, C. I., and Thomson, D. H. (1995). “Marine Mammals
and Noise.” Academic Press, San Diego, CA.
Rideout, B. A., Gause, G. E., Benirschke, K., and Lasley, B. L. (1985). Stress-induced adrenal
changes and their relation to reproductive failure in captive nine-banded armadillos
(Dasypus novemcinctus). Zoo Biol. 4, 129-137.
Rivier, C. (1989). Involvement of endogenous corticotropin-releasing factor (CRF) in modulat-
ing ACTH and LH secretion function during exposure to stress, alcohol or cocaine in
the rat. In “Molecular Biology of Stress” (S. Breznitz and 0. Zinder, eds.), pp. 31-47.
Alan R. Liss, New York.
Robertson, 0.H., and Wexler, B. C. (1957). Pituitary degeneration and adrenal tissue hyperpla-
sia in spawning Pacific salmon. Science 125, 1295-1296.
Robertson, 0. H., and Wexler. B. C. (1960). Histological changes in the organs and tissues
of migrating and spawning Pacific salmon (genus Oncorhynchus). Endocrinology (Ealri-
more) 66, 222-239.
Rodriguez, A., Barrios, L., and Delibes, M. (1995). Experimental release of an Iberian lynx
(Lynx pardina). Eiodiv. Conserv. 4, 382-394.
Rodway, M. S., Montevecchi, W. A,, and Chardine, J. W. (1996). Effects of investigator
disturbance on breeding success of Atlantic puffins. Fratercula arctica. Biol. Conserv.
76,311-319.
518 H E R B E R T HOFER AND MARION L. EAST
Roff. D. A. (1992). “The Evolution of Life Histories.” Chapman & Hall, London.
Rogeness, G. A,. and McClure, E. B. (1996). Development and neurotransmitter-environmen-
tal interactions. Dev. fsychoparhol. 8, 183-199.
Rogers, C. M.. Ramenofsky, M., Ketterson, E. D.. Nolan, V., and Wingfield, J. C. (1993). Plasma
corticosterone. adrenal mass. winter weather, and season in nonbreeding populations of
dark-eyed juncos (Junco hyernalis hyernalis). Auk 110, 279-285.
Rollo, C. D. (1995). “Phenotypes. Their Epigenetics, Ecology and Evolution.” Chapman &
Hall, London.
Root, T. ( 1988). Energetic constraints on avian distributions and abundances. Evolufion
(Lawrence, Kuns.) 69, 330-330.
Ross, P. S., Pohajdak, B., Bowen, W. D., and Addison, R. F. (1903). Immune function in
free-ranging harbor seal (Phoca virulina) mothers and their pups during lactation.
J . Wild. Dis. 29, 21-29.
Ross, P. S.. de Swart. R. L., Reijnders, P. J . H., van Loveren, H.. Vos, J. G., and Osterhaus.
A. D. M. E. (1995). Contaminant-related suppression of delayed-type hypersensitivity
and antibody responses in harbor seals fed herring from the Baltic Sea. Environ. Health
Perspect. 103, 162-167.
Ross. P. S.. de Swart, R. L., Addison, R., van Loveren, H.. Vos, J . G., and Osterhaus,
A. D. M. E. (1996a). Contaminant-induced immunotoxicity in harbour seals-wildlife at
risk. Toxicology 112, 157-169.
Ross, P. S., de Swart, R. L., Timmerman. H. H., Vedder. L. J., van Loveren, H., Vos, J . G.,
Reijnders, P. J. H., and Osterhaus, A. D. M. E. (1996b). Suppression of natural killer
cell activity in harbour seals (Phoca virulina) fed Baltic Sea herring. Aquat. Toxicol.
34, 71-84.
Rostene, W., Sarrieau. A., Nicot. A,, Scarceriaux. V.. Betancur, C., Gully, D., Meaney, M.,
Rowe, W.. de Kloet. R.. Pelaprat, D., and Berod. A. (1995). Steroid effects on brain
functions: An example of the action of glucocorticoids on central dopaminergic and
neurotensinergic systems. J. Psychiafr. Neurosci. 20, 349-356.
Rotella, J . J., Howerter, D. W.. Sankowski, T. P.. and Devries, J. H. (1993). Nesting effort
by wild mallards with three types of radio transmitters. J. Wildl. Manage. 57, 690-695.
Rowan, A. N. (1990). Refinement of animal research technique and validity of research data.
Fundam. Appl. Toxicol. 15, 25-32.
Rushen, J. (1986). Some problems with the physiological concept of ‘stress.” Ausf. Ver. J .
63, 359-361.
Sachser, N. (1987). Short-term responses of plasma norepinephrine. epinephrine, glucocorti-
coid and testosterone titers to social and non-social stressors in male guinea pigs of
different social status. Physiol. Behav. 39, 11-20.
Sachser. N. (1994). Social stratification and health in non-human mammals-a case study in
guinea pigs. In “Social Stratification and Socioeconomic Inequality” (L. Ellis, ed.), Vol.
2, pp. 113-121. Praeger, Westport, CT.
Sachser. N.. and Beer. R. (1995). Long-term influences of social situation and socialization
on adaptability in behaviour. In “Research and Captive Propagation” (U. Ganslosser.
J. K. Hodges, and W. Kaumanns, eds.). pp. 207-214. Filander, Furth. Germany.
Sachser. N., and Kaiser. S. (1 996). Prenatal social stress masculinizes the females’ behaviour
in guinea pigs. Physiol. Behav. 60, 589-594.
Sachser, N., and Lick, C. (1989). Social stress in guinea pigs. Physiol. Behav. 46, 137-144.
Sachser. N.. and Lick, C. (1991). Social experience, behavior, and stress in guinea pigs. Physiol.
Behuv. 50, 83-90.
Sachser. N., Lick, C., and Stanzel, K. (1994). The environment, hormones, and aggressive
behavior: A 5-year-study in guinea pigs. Psychoneuroendocrinvlo~y19, 697-707.
Sachser, N., Diirschlag. M., and Hirzel, D. (1997). Social relationships and the management
of stress. Psychoneuroendocrinology (in press).
Sage, G., Khawplod, P., Wilde, H., Lobaugh, C., Hemachudha, T., Tepsumethanon. W., and
Lumlertdaecha, B. (1993). Immune response to rabies vaccine in Alaskan dogs: Failure
to achieve a consistently protective antibody response. Trans. R. SOC. Trop. Med. Hyg.
87, 593-595.
Saltz, D., and White, G. C. (1991). Urinary cortisol and urea nitrogen responses to winter
stress in mule deer. J. Wildl. Manage. 55, 1-16.
Saltzman. W., Schultz-Darken, N. J., Scheffler. G., Wegner, F. H.. and Abbott, D. H. (1994).
Social and reproductive influences on plasma cortisol in female marmoset monkeys.
Physiol. Behav. 56, 801-810.
Sapolsky, R. M. (1982). The endocrine stress-response and social status in the wild baboon.
Horm. Behav. 16,279-292.
Sapolsky. R. M. (1985). Stress-induced suppression of testicular function in the wild baboon:
Role of glucocorticoids. Endocrinology 116, 2273-2278.
Sapolsky, R. M. (1987). Stress, social status, and reproductive physiology in free-living baboons.
In “Psychobiology of Reproductive Behavior” (D. Crews, ed.), pp. 291-322. Prentice
Hall, Englewood Cliffs. NJ.
Sapolsky. R. M. (1992). Neuroendocrinology of the stress response. In “Behavioral Endocrinol-
ogy” ( J . B. Becker, S. M. Breedlove. and D. Crews, eds.). pp. 287-324. MIT Press,
Cambridge, MA.
Sapolsky, R. M. (1994). “Why Zebras Don’t Get Ulcers.” Freeman, New York.
Sapolsky, R. M. (1997). The physiological and pathophysiological implications of social stress
in mammals. In “Handbook of Physiology” (in press).
Sapolsky, R. M., and Ray, J . C. (1989). Styles of dominance and their endocrine correlates
among wild olive baboons (Papio anribis). Am. J. Primntol. 18, 1-13.
Satterlee, D. G., Jones, R. 9.. and Ryder, F. H. (1993). Effects of vitamin C supplementation
on t h e adrenocortical and tonic immobility fear reactions of japanese quail genetically
selected for high corticosterone response to stress. Appl. Anim. Behav. Sci. 35,347-357.
Schapiro, S. J., Nehete, P. N., Perlman, J. E., and Sastry, K. J. (1996). Effects of environmental
enrichment on cellular immune function of rhesus macaques. Ahstr., Congr. Int. Primatol.
Soc.. 16th, 353.
Schauer, J. H. S., and Murphy, E. C. (1996). Predation on eggs and nestlings of common
murres (Uria cralge) at Bluff. Alaska. Colon. Waterbirds 19, 186-198.
Scheepers. J. L.. and Venzke, K. A. E. (1995). Attempts to reintroduce African wild dogs
Lycaon pictics into Etosha National Park, Namibia. S. Afr. J. Wildl. Res. 25, 138-140.
Scherzinger. W. (1979). Zum Feindverhalten des Haselhuhnes (Bonasa bonasia). Vogelwelt
100, 205-217.
Schmidt. C.. and Sachser, N. (1996). Auswirkungen unterschiedlicher Futterverteilungen auf
Verhalten und Streflhormonkonzentrationene von Breitmaulnashornern (Ceratotherium
siniirm simum) im Allwetterzoo Munster. Verhl. Dsch. Zoo/. Ges. 89(1), 294-294.
Schneider, M. L. (1992a). The effect of mild stress during pregnancy on birthweight and
neuromotor maturation in rhesus monkey infants (Macaca mulafta). Infant Behav. Dev.
15,389-403.
Schneider. M. L. (1992b). Delayed object permanence development in prenatally stressed
rhesus monkey infants (Macaca m d a t m ) . Occup. Ther. J. Res. 12,96-110.
Schneider, M. L. (1992~).Prenatal stress exposure alters postnatal behavioral expression under
conditions of novelty challenge in rhesus monkey infants. Dev. Psychohiol. 25,529-540.
Schneider. M. L., and Coe, C. L. (1993). Repeated social stress during pregnancy impairs
neuromotor development of the primate infant. J. Dev. Behav. Pediatr. 14, 81-87.
Schnidrig, R., Marbacher. H., and Ingold. P. (1992). How is paragliding affecting the behaviour
and general condition of chamois Rupicapra rupicapra: A study in the Swiss alps. In
“Ongules/Ungulates 91” (F. Spitz, ed.). pp. 609-610. Inst. Rech. Grands Mammifkres,
Toulouse, France.
Schoech. S. J., Mumme, R. L., and Moore, M. C. (1991). Reproductive endocrinology and
mechanisms of breeding inhibition in cooperatively breeding Florida scrub jays. (Aphelo-
coma c. coerulescens). Condor 93, 354-364.
Schofield, P. S., McLees, D. J., Kerbey, A. L. and Sugden, M. C. (1986). Activities of cardiac
and hepatic pyruvate dehydrogenase complex are decreased after surgical stress. Biochem.
Int. 12, 189-197.
Schrader. L. and Todt, D. (in press). Vocalisations can indicate an endocrine stress response
in domestic pigs (Sus scrofa domesticn). Ethology, in press.
Schultz. R. D.. and Bailey, J. A. (1978). Response of white-tailed deer to snowmobiles and
snowmobile trails in Maine. Can. Field Nat. 92, 334-344.
Schumacher, U., Heidemann, G . ,Skirnisson, K., Schumacher, W., and Pickering, R. M. (1995).
Impact of captivity and contamination level on blood parameters of harbour seals (Phoca
vitulina). Comp. Biochem. Physiol. A 112A, 455-462.
Scott, T. M.. and Koehn, R. K. (1990). The effect of environmental stress on the relationship
of heterozygosity to growth in the coot clam Mulinia lateralis (Say). J. Exp. Mar. B i d .
Ecol. 135, 109-116.
Sears, J., Cooke, S. W., Cooke, Z . R., and Heron, T. J. (1989). A method for the treatment
of lead poisoning in the mute swan (Cygnus olor) and its long-term success. Brit. vet. J.
145, 586-595.
Selye, H. (1946). The general-adaptation syndrome and disease of adaptation. J. Clin. Endocri-
nol. Metab. 6, 117-230.
Selye, H. (1973). The evolution of the stress concept. Am. Sci. 61, 692-699.
Shackley, M. (1992). Manatees and tourism in southern Florida: opportunity or threat?
J. Environ. Manage. 34, 257-265.
Shepherd, N. (1986). Capture myopathy. I n “Care and Handling of Australian Native Mam-
mals” (S. J. Hand, ed.), pp. 143-148. Surrey Beatty, Chipping Norton, Australia.
Sibly. R. M., and Calow, P. (1983). An integrated approach to life-cycle evolution using
selective landscapes. J. Theor. Biol. 102, 527-547.
Sibly, R. M., and Calow, P. (1986). “Physiological Ecology of Animals.” Blackwell, Oxford.
Sibly, R. M., and Calow, P. (1989). A life-cycle theory of responses to stress. Biol. J . Linn.
Soc. 37, 101-116.
Simmonds, M. (1991). The involvement of environmental factors in the 1988 european seal
epidemic. Rev. Int. Oceanogr. Med. 101-104, 109-1 14.
Simmonds, M., and Johnston, P. A. (1989). Seals, sense and science. Mar. Pollut. Bull. 20,
580-583.
Sinclair. A. R. E., and Pech, R. P. (1996). Density dependence, stochasticity, compensation
and predator regulation. Oikos 75, 164-173.
Skaare. J. U., Markussen, N. H., Norheim, G.. Haugen, S., and Holt, G. (1990). Levels of
polychorinated biphenyls, organochlorine pesticides, mercury, cadmium, copper, sele-
nium, arsenic and zinc in the harbor seal (Phoca vitulina) in Norwegian waters. Environ.
Poll~it.66, 309-324.
Skogland, T., and Grovan, B. (1988). The effects of human disturbance on the activity of wild
reindeer in different physical condition. Rangifer 8, 11-19,
Smith, G. J., and Rongstad, 0. J. (1980). Serologic and hematologic values of wild coyotes
in Wisconsin. J . Wildl. Dis. 16, 491-497.
Smith. G. T., Wingfield, J. C.. and Veit. R. R. (1994). Adrenocortical response to stress in
the common diving petrel. Pelernnoides urinafrix. Physiol. Zoo/. 67, 526-537.
Smith, H. R. (1980). Growth. reproduction and survival in Peromyscus leucopus carrying
intraperitoneally implanted transmitters. In “ A Handbook of Biotelemetry and Radio-
tracking’’ (C. A. Amlaner and D. W. Macdonald. eds.), pp. 367-374. Pergamon. Oxford.
Smith. R. J . (undated). Some effects of limb injuries on the chimpanzees of the Budongo
Forest. Ms, Institute of Biological Anthropology, Oxford.
Smith. T. E., and McGreer-Whitworth, B. (1996). Psychosocial stress and urinary cortisol
excretion in marmoset monkeys (Callirhrix kuhli). Abstr. Congr. Int. Primatol. Soc.
I6rh, 086.
Soave, 0.A . (1964). Reactivation of rabies virus in a guinea pig due to the stress of crowding.
Am. J . Vet. Rex 25, 268-269.
Soave, 0. A,, Johnson, H. N.. and Nakamura, K. (1961). Reactivation of rabies virus infection
with adrenocorticotropic hormones. Science 133, 1360-1360.
Sober, E. (1 993). “Philosophy of Biology.” Oxford University Press. Oxford.
Soule. M. E. (1986). “Conservation Biology.” Sinauer, Sunderland, MA.
SoulC, M. E. (1987). “Viable Populations for Conservation.” Cambridge University Press.
Cambridge, UK.
Southwood, T. R. E. (1988). Tactics, strategies and templets. Oikos 52, 3-18.
Speakman. J. R., Webb. P. I.. and Racey. P. A. (1991). Effects of disturbance on the energy
expenditure of hibernating bats. J. Appl. Ecol. 28, 1087-1095.
Spraker. T. R., Adrian, W. J . , and Lance, W. R. (1987). Capture myopathy in wild turkeys
Meleagri.s gnllopavo following trapping, handling and transportation in Colorado. USA.
J . Wildl. Dis. 23, 447-453.
Stangel, P. W.. and Lennartz. M. R . (1988). Survival of red-cockaded woodpecker nestlings
unaffected by sampling blood and feather pulp for genetic studies. J. Field Ortiithol.
59,389-394.
St. Aubin. D. J., and Geraci, J . R. (1986). Adrenocortical function in pinniped hyponatremia.
Mur. Mammol. Sci. 2, 243-250.
Steams. S. C. (1992). “The Evolution of Life Histories.” Oxford University Press. Oxford.
Stebbings, R. E. (1982). Radio tracking greater horseshoe bats with preliminary observations
on flight patterns. Symp. Zoo/. Soc. London 49, 161-173.
Stefano. G. B., and Smith, E. M. (1996). Adrenocorticotropin-a central trigger in immune
responsiveness: Tonal inhibition ‘of immune activation. Med. Hypotheses 46, 471-478.
Still, A. W. (1982). O n the number of subjects used in animal behaviour experiments. Anim.
Behav. 30, 873-880.
Stirling, I.. and Derocher. A. E. (1993). Possible impacts of climatic warming on polar bears.
Arctic 46, 240-245.
Stockwell, C. A.. Mulvey. M.. and Vinyard, G. L. (1996). Translocations and the preservation
of allelic diversity. Conserv. B i d . 10, 1133-1 141.
Stone. A. M., Bloomsmith, M. A,, Laule. G. E.. and Alford. P. L. (1996). Positive reinforcement
training for voluntary movement of group-housed chimpanzees. Abstr., Cungr. fnt. Primcr-
tol. Soc.. 16th, 679.
Stone, G. S.. Katona, S. L., Mainwaring, A,. Allen. J. M., and Corbett, H. D. (1992). Respiration
and surfacing rates of fin whales (Balaenopferu physalus) observed from a lighthouse
tower. Rep. Inr. Whal. Cornm. 42, 739-745.
Stussy. R. J., Edge. W. D., and O’Neil, T. A. (1994). Survival of resident and translocated
female elk in the Cascade Mountains of Oregon. Wildl. Soc. Birll. 22, 242-247.
Sutherland. W. J. (1996). “From Individual Behaviour to Population Ecology.” Oxford Univer-
sity Press. Oxford.
Taberlet, P.. and Bouvet. J . (1991). A single plucked feather as a source of DNA for bird
genetic studies. Auk 108, 959-960.
Taberlet. P., and Bouvet, J. (1992). Bear conservation genetics. Nature (London)358,197-197.
Tarara, E. B., Tarara, R. P.. and Suleman, M. A. (1995). Stress-induced gastric ulcers in vervet
monkeys (Cercopitlzecus aethiops): The influence of life history factors. Part 11. J . Zoo
Wildl. Med. 26, 72-75.
Taylor, V. J.. and Dunstone, N., eds. (1996a). “The Exploitation of Mammal Populations.”
Chapman & Hall. London.
Taylor, V. J., and Dunstone. N. (1996b). The exploitation, sustainable use and welfare of wild
mammals. In “The Exploitation of Mammal Populations” (V. J. Taylor and N. Dunstone,
eds.), pp 3-15. Chapman & Hall, London.
Thirgood. S. J.. Redpath, S. M.. Hudson, P. J.. Hurley, M. M.. and Aebischer, N. J. (1995).
Effects of necklace radio transmitters on survival and breeding success of red grouse
Lagopus lagopi~sscoficus. Wildl. B i d . 1, 121-126.
Thomas, J., Carver, M., Haische, C., Thomas, F.. Welch, J., and Carchman, R. (1982). Differen-
tial effects of intravenous anaesthetic agents on cell-mediated immunity in the rhesus
monkey. Clin. Exp. lmmrinol. 47, 457-466.
Thomas, J. A,, Kastelein. R. A,, and Awbrey. F. T. (1990). Behavior and blood catecholamines
of captive belugas during playbacks of noise from an oil drilling platform. Zoo Biol.
9, 393-402.
Thomas, L.. and Juanes. F. (1996). The importance of statistical power analysis: An example
from Animal Behaviour. Anim. Behav. 52,856-859.
Thompson, D. R.. Hamer, K. C.. and Furness, R. W. (1991). Mercury accumulation in great
skuas Catharacta skuci of known age and sex and its effects upon breeding and survival.
J . Appl. Ecol. 28,672-684.
Thorne, E. T.. and Williams, E. S. (1988). Disease and endangered species: The black-footed
ferret as a recent example. Conserv. B i d . 2, 66-74.
Toates, F. (1995). “Stress. Conceptual and Biological Aspects.’’ Wiley, Chichester.
Tranvik. L., Bengtsson. G.. and Rundgren, S. (1993). Relative abundance and resistance traits
of two Collemhola species under metal stress. J. Appl. Ecol. 30, 43-52.
Trillmich, F., and Ono. K. A. (1992). “Pinnipeds and El Niiio. Responses to Environmental
Stress.” Springer. Berlin.
Underwood, A. J., and Kennelly. S. J. (1990). Pilot studies for designs of surveys of human
disturbance of intertidal habitats in New South Wales. Australia. Auut. J . Mar. Freshwater
Res. 41, 165-174.
van Eden, W., and Young. D. B. (1996). “Stress Proteins in Medicine.” Dekker. New York.
Van Heezik, Y.. and Seddon. P. J . (1990). Effect of human disturbance on beach groups of
jackass penguins. S. A,fr. J . Wildl. Rrs. 20, 89-93,
Van Horne, B. (1983). Density as a misleading indicator of habitat quality. J. W i l d Manage.
47, 893-90 1.
van Oirschot. J. T.. and Gielkens, A. L. J. (1984). In vivo and in vitro reactivation of latent
pseudorabies virus in pigs born to vaccinated sows. Am. J . Vet. Res. 45, 567-571.
Van Vuren, D. (1989). Effects of intraperitoneal transmitter implants on yellow-bellied mar-
mots. J. Wildl. Manage. 53, 320-323.
von Holst, D. (1986). Psychosocial stress and its pathophysiological effects in tree shrew
(Tupaia belangeri). In “Biological and Psychological Factors in Cardiovascular Disease”
(T. H. Schmidt, T. M. Dembroski. and G. Blumchen. eds.), pp. 476-489. Springer.
Heidelberg.
Walker. C. H.. Hopkin. S. P.. Sibly, R. M.. and Peakall, D. B. (1996). “Principles of Ecotoxicol-
ogy.” Taylor & Francis, London.
Walter, H. S. (1990). Small viable population: The red-tailed hawk of Socorro Island. Conserv.
B i d . 4, 441-443.
Warner. R. E., and Etter, S. L. (1983). Reproduction and survival of radio-marked hen ring-
necked pheasants in Illinois. J. Wildl. Mange. 47, 369-375.
Warwick. R. M. (1986). A new method for detecting pollution effects on marine macrobenthic
communities. Mar. Biol. 92, 557-562.
Wasser. S. K., and Barash, D. P. (1983). Reproductive suppression among female mammals:
Implications for biomedicine and sexual selection theory. Q. Rev. Biol. 58, 513-538.
Wasser. S. K.. and Starling, A. K. (1986). Reproductive competition among female yellow
baboons. In “Primate Ontogeny. Cognition and Social Behaviour” ( J . G. Else and P. C.
Lee. eds.). pp. 343-354. Cambridge University Press, Cambridge, UK.
Watson, K. (1990). Microbial stress proteins. A h . Microb. Physiol. 31, 183-223.
Wattrang, E.. Edfors-Lilja, I., Anderson. L.. and Fossum, C. (1994). Genetic and stress-
mediated influence on Aujeszky’s disease virus induced interferon-alpha production in
porcine leukocytes. Acfa Vet. Hung. 42, 331-336.
Westneat. D. F. (1986). The effects of muscle biopsy on survival and condition in white
throated sparrows. Wilson Bull. 98, 281-285.
Westneat, D. F., Payne. R. B.. and Doehlert, S. M. (1986). Effects of muscle biopsy on survival
and breeding success in Indigo buntings. Condor 88, 220-227.
Wheeler. W. E., Gatti, R. C.. and Bartelt. G. A. (1984). Duck breeding ecology and harvest
characteristics on Grand River Marsh Wildlife Area, Wisconsin, USA. Wis. Dep. Nar.
Res. Tech. Bull. 145, 1-49,
White, G. C., and Garrott. R. A. (1990). “Analysis of Wildlife Radio-tracking Data.” Academic
Widmaier, E. P.. Harmer, T. L.. Sulak, A. M., and Kunz, T. H. (1994). Further characterization
of the pituitary-adrenocortical responses to stress in chiroptera. J . Exp. Zool.269,442-449.
Wiepkema, P. R., and Kolhaas, J. M. (1993). Stress and animal welfare. Anim. Welfare
2, 195-218.
Williams, G. C . (1966). “Adaptation and Natural Selection.” Princeton University Press.
Princeton, NJ.
Williams. T. D., and Rothery. P. (1990). Factors affecting variation in foarging and activity
patterns of gentoo penguins (Pygoscelis papitu) during the breeding season at Bird Island,
South Georgia. J. Appl. Ecol. 27, 1042-1054.
Williamson. D. T. (1991). Condition, growth and reproduction in female red lechwe (Kobiis
leche leche Gray 1850). Afr. J. Ecol. 29, 105-117.
Wilson, B. S.. and Wingfield, J. C. (1992). Correlation between female reproductive condition
and plasma corticosterone in the lizard Ura stansburiana. Copeia, pp. 691-697.
Wilson, B. S.. and Wingfield, J. C. (1994). Seasonal and interpopulational variation in plasma
levels of corticosterone in the side-blotched lizard (Ura sransburiana). Physiol. 2001.
67, 1025-1049.
Wilson. R. P., and Wilson, M. P. T. J. (1989). A minimal-stress bird-capture technique.
J. Wildl. Manage. 53, 77-80.
Wilson. R. P., Coria. N. R.. Spairani, H. J.. Adelung. D., and Culik, B. M. (1989). Human-
induced behavior in AdClie penguins (Pygoscelis adeliae). Polar Biol. 10, 77-80.
Wilson, R. P., Culik, B. M., Danfeld. R.. and Adelung, D. (1991). People in Antarctica: How
much do AdClie penguins (Pygoscrlis adeliae) care? Polar B i d . 11, 363-370.
Wingfield. J. C. (1985). Influences of weather on reproductive function in female song sparrows
(Melospiza melodia). J. Zool. 205, 545-558.
Wingfield. J. C. (1988). Changes in reproductive function of free-living birds in direct response
to environmental perturbations. In “Processing of Environmental Information in Verte-
brates” (M. H. Stetson. ed.), pp. 121-148, Springer, New York.
Wingfield. J. C.. and Farner. D. S. (1976). Avian endocrinology-field investigations and

methods. Condor 78, 570-573.
Wingfield, J. C., and Farner, D. S. (1980). “Endocrinologic and Reproductive States of Bird
Populations under Environmental Stress,” U S . EPA Report Contract No. 699095. U S .
Environ. Prot. Agency, Washington, DC.
Wingfield, J. C.. and Grimm. A. S. (1977). Seasonal changes in plasma cortisol, testosterone
and oestradiol-17-beta in the plaice, Pleuronectes platessa L. Gen. Comp. Endocrinol.
31, 1-11.
Wingfield, J. C.. and Lewis. D. M. (1993). Hormonal and behavioural responses to simulated
territorial intrusion in the cooperatively breeding white-browed sparrow weaver. Ploce-
passer mahali. Anim. Behav. 45, 1-1 1.
Wingfield, J. C., Hahn. T. P.. Levin, R.. and Honey, P. (1992a). Environmental predictability
and control of gonadal cycles in birds. J. Exp. Zoo/. 261,214-231.
Wingfield. J. C., Vleck. C. M., and Moore. M. C. (1992b). Seasonal changes of the adrenocortical
response to stress in birds of the Sonoran desert. J. Exp. Zoo/. 264, 419-428.
Wingfield, J. C., Deviche, P.. Sharbaugh, S.. Astheimer. L. B.. Holberton, R.. Suydam, R.,
and Hunt, K. (1994a). Seasonal changes of the adrenocortical responses to stress in
redpolls, Acanthis fiammea, in Alaska. J. Exp. Zoo/. 270, 372-380.
Wingfield. J. C.. Suydam, R.. and Hunt, K. (1994b). The adrenocortical responses to stress
in snow buntings (Plectrophenux nivn/is) and Lapland longspurs (Calcnriirs iapponicus)
at Barrow, Alaska. Comp. Biochem. Physiol. C 108,299-306.
Wingfield, J. C., Kubokawa, K., Ishida, K., Ishii, S., and Wada. M. (1995a). The adrenocortical
response to stress in male bush warblers, Cettia diphone: A comparison of breeding
populations in Honshu and Hokkaido, Japan. Zoo/. Sci. 12, 615-621.
Wingfield, J. C.. O’Reilly, K. M., and Astheimer, L. B. (1995b). Modulation of the adrenocorti-
cal responses to acute stress in Arctic birds: a possible ecological basis. Am. Zoo/. 35,
285-294.
Wingfield, J. C.. Hunt, K.. Breuner. C., Dunlap, K., Fowler. G. S., Freed, L.. and Lepson, J.
(1 996). Environmental stress, field endocrinology, and conservation biology. I n “Behav-
ioral Approaches to Conservation in the Wild” ( J . R. Clemmons and R. Buchholz, eds.),
pp. 95-131. Cambridge University Press, Cambridge, UK.
Wingfield. J. C., Breuner, C., and Jacobs. J. (1997). Corticosterone and behavioral responses
to unpredictable events. I n “Perspectives in Avian Endocrinology” (S. Harvey and
R. J. Etches. eds.), pp. 1-12. Journal of Endocrinology Ltd, Bristol.
Wingfield. J. C., Maney, D. L., Breuner, C. W.. Honey, P. K., Jacobs, J. D., Lynn, S.. Ramenof-
sky, M., and Richardson, R. D. (in press). Ecological bases of hormone-behavior interac-
tions: The “emergency life history stage”. Am. Zool., (in press).
Wolf, C. M., Griffith, B., Reed, C.. and Temple, S. A. (1996). Avian and mammalian transloca-
tions: Update and reanalysis of 1987 survey data. Consrrv. B i d . 10, 1142-1 154.
Woodford. M. H., and Kock, R. A. (1991). Veterinary considerations in re-introduction and
translocation projects. Symp. Zool. Soc. London 62, 101-1 10.
Woolf, A,, Curl. J. L., and Anderson, E. (1984). Inanitation following implantation of a
radiotelemetry device in a river otter. J. Am. Vet. Med. Assoc. 185, 1415-1415.
Wright, J . F. (1995). Development and use of a system for predicting the macroinvertebrate
fauna in Howing waters. Atrst. J. Ecol. 20, 181-197.
Wright, J. F., Furse. M. T., Armitage, P. D., and Moss, D. (1993). New procedures for
identifying running-water sites subject to environmental stress and for evaluating sites
for conservation based on the macroinvertebrate fauna. Arch. Hvdrobiol. 127, 3 19-326.
Wuichet. J.. and Norton, B. (1995). Differing conceptions of animal welfare. In “Ethics on
the Ark: Zoos, Animal Welfare. and Wildlife Conservation” (B. G . Norton. M. Hutchins,
E. F. Stevens. and T. L. Maple. eds.). pp. 235-250. Smithsonian Institution Press. Washing-
ton, D C and London.
Wynn. P. C., Shahneh. A. Z.. Ribgy. R. D. G . . Behrendt, R.. Giles, L. R., Gooden, J. M.,
and Jones. M. R. (1995). Physiological consequences of the induction of auto-immunity
to adrenocorticotropin (ACTH). Livest. Prod. Sci. 42, 247-254.
Yarmoloy, C.. Bayer, M., and Geist, V. (1988). Behavior responses and reproduction of mule
deer. Odocoileus hentionus, does following experimental harassment with an all-terrain
vehicle. Can. Field Nat. 102, 425-429.
Yediler. A.. Panou, A., and Schramel, P. (lY93). Heavy metals in hair samples of the mediterra-
nean monk seal (Monachrrs monc7hus). Mar. Polhir. B d l . 26, 156-159.
Yeoman. R. R.. Ricker. R. B.. Williams. L. E.. Sonksen. J.. and Abee. C. R. (1996). Vibrostimu-
lation (VS) is superior to electro-ejaculation (EE) in squirrel monkeys (Suirniri). Ahstr..
Congr. Int. Primatol. Soc.. 16th, 222.
Yeruham, I.. Perl. S., Nyska. A,, Abraham. A., Davidson, M.. Haymovitch, M., Zamir, 0..
and Grinstein, H. (1994). Adverse reactions in cattle to a capripox vaccine. Vet. Rec.
135,330-332.
Zethof, T. J. J . , van der Heyden. J . A. M., Tolboom, J. T. B. M., and Olivier. B. (1994).
Stress-induced hyperthermia in mice: a methodological study. Physiol. Behav. 55,109-1 15.
Zwartouw. H. T., MacArthur. J. A,. Boulter, E. A,, Seamer. J. H.. Marston, J . H.. and
Chamove, A . S. (1984). Transmission of B virus infection between monkeys especially
in relation to breeding colonies. Lob. Anirri. 18, 125-130.
Index
A weight measurements
psychosocial stress, 48-49
Abiotic stress usefulness, 22-23
adaptation. 165-166 Adrenal medulla, see Sympathetico-
energy balance, 155-156 adrenomedullary system
Achievement pleasure, 392 Adrenocortical hormones, 19-22
Acoustics, cry, offspring condition, 354-355 Adrenocortical response, see also
ACTH (adrenocorticotrophic hormone) Pituitary-adrenocortical system; Stress
central effects, 20-21 response
glucocorticosteroid release. 19-20 aggressive behavior, 45-47, 59
Active stress response, 107-108 assessment, 22-29, 490
Cannon’s, 3-5. 62 bidirectional nature, 63
chronic stress, 10 challenge test, 25-29
dominance relationships, 62, 63, 88-89 chemical immobilization and anesthesia,
gonadal activity, 14
478
tree shrews, 54-58
dominance relationships
Activity, as indicator of stressed state.
rabbits, 64-65
449-450
tree shrews, 55-57
Adaptation
emotionally induced, 6-7
abiotic stress, 165-166
energetics, 435-436
chronic challenges, 5-10
critical situations. 3 fitness effects, 431
energy limits. 158-164 gonadal activity
evolutionary, relationship to stress, 414 negative coupling, 432-434
extending limits. 165-169 uncoupling, 434-435
learning, 168-169 manual restraint, 477
resource heterogeneity, 166-167 metabolic rate, 435-436
resource polymorphisms, 167-168 passive chronic stress, 10-11
as solution to environmental stress, 320 predictability and control, 7. 8-10. 63.
stress response as. 438-439 88-89, 437
Adaptation diseases, 6 psychosocial stress. 48-49
Adaptation syndrome, general. 5-6. 62 reproduction, 431 -436
Adaptive sociality. predation. 239-241 Selyean concept, 5-6. 62
Adenine energy charge, 134, 446-447 sensitivity, environmental stress, 431-432.
Adrenal cortex 433-434
hormonal activation, 19-20 social bonds, disruption. 90, 93
sympathetic activation, 20 social factors, 436-437
zones, 19 tuning options, 431-435
Adrenal glands variability, 430-431
enlargement. long-term stress. 22 Adrenocorticotrophic hormone
hormone measurements, usefulness. central effects, 20-21
22-29 glucocorticosteroid release, 19-20
527
528 INDEX
AEC, see Adenine energy charge Anfechinus, stress reactions, males, 1-2.
African lions (fandirra leo), pursuit 434-435
hunting, 467 A nthropogenic st ressors
African talapoin (Miopifhwrrs frilqmin), effects. 452-472
stress response. dominance energetic measures, 439-441
relationships, 71 equivalence to natural strcssors, 486-488
African wild dogs (Lvcaon picfirs). stress examples. 415
response, dominance relationships. 72 fitness consequences. 413
Agdaiiis phoeniceiis. food aversions. 307 Antibacterial foliage. 296-297
Aggressive behavior Antibodies. 38. 41
adrenocortical response, 45-47. 59 Anticipation. stress concept. 408
cardiovascular response. 67-69 Anting, birds, 297-298
mortality. 1-2, 434-435 Antipredator decision making. see NISO
social stress. 45-47
Predator-induced stress
testosterone, 75. 86-87
developmental instability, 189-190. 192
Alarm reaction. general adaptation
ecological effects, 248-261
syndrome. 5
foraging behavior. 217-225
Aldosterone, stress response. 22
future research. 264-265
Algal bloom, 454-455
long-term consequences. 245-248
American black bears (Ursrrs rinirriconirs),
overview, 215-216
tourism-related disturbance. 464
patterns of activity. 225-235
Amphibians. use of space, antipredator
levels, 226-23 1
decision making, 256
temporal. 231-235
Androgens
adrenal. stress response, 22 types, 226
physiological effects. 3.5 population dynamics. 257-259
production and release. 36 postencounter, 235-239
Anesthesia. as stressor. 478-479. 479 approaching and inspecting predators.
Anger. function. 390 238-239
Animal breeding, developmental instability, escape behavior. 237-238
191 flight initiation. 237
Animal handling, see Handling pursuit-deterrence signals, 236-237
Animal housing resumption of activity. 236
equivalence to natural stressors. 487-488 predator reaction, 263
as stressor. 16. 483-485 reproductive, 241-245, 259
Animal models. “knock out,” immune risk assessment. 262-263
system. 138 role of modeling. 225
Animal welfare scaling to real world. 264
assessment, 396-397 social situations. 239-241
coping systems. 400 species interactions. 259-261
preference tests. 397-398 stress response. 261 -262
time. 398 use of space. 248-257
definition. 394-395 Antipredator training. 492
devclopmental instability, 197-199 Antischistosomal drug use. baboons. 296
relationship to health. 396. 397 Antisymmetry, 181, 183
relationship to stress. 396. 397 Ants, adaptation. abiotic stresses. 166
relevance of feclings. 371, 395. 399-400 Anxiety. function, 385
stress concept, 407-408 Aphids, escape behavior, 238
Antarctic penguins, tourism-related Arctic charr (Srrlvelinrrs alpinus). resource
disturbance, case study. 462-463 polymorphisms. 167
INDEX 529
Aspilia. self-medication. chimpanzees. Badgers

298-301 escape behavior. 238
Asymmetry hunting effects. 466
behavioral. 204-205 Baltic gray seals, skull bone lesions.
correlations. 185 455-456
directional. 183 Barnacles. tidal activity patterns. 235
facial, 326-327 Behavior
fitness correlates. 192-193 asymmetrical, 204-205
fluctuating. see Fluctuating asymmetry as indicator of stressed state, 44Y-450
morphological invariance. 201-204
behavioral effects, 204 observation. recognition of feelings,
types, 181, 183 375-376
performance costs, 184 sequences. fractal dimension, 203-204
secondary sexual characters. 190-191, stress concept. 409-410
323-324 Behavioral conditioning. immune function,
selection pressure. 184 141, 144
types. 181-183 Behaviorally transmitted indirect effects
Athleticism. immunocompetence. activity level, 259-260
327-328 top-down ecosystem regulation. 260
ATP. as indicator of stressed state. use of space. 260
446-447 Behavioral research, stress concept. 14-15.
Attraction and attractiveness. humans, 44-45
321-332,356-357 Behavioral traits. repeatability. 202-203
developmental stability, 321-323 Betfo splendms. sexual behavior, energy
secondary sexual characters. 323-324 costs, 160
stress resistance Bighorn sheep (Ovis canadensis),
nonsexual social behavior, 330-331 heterozygosity and fitness, 164
sexual selection, 324-330 Biological conservation. see Conservation
Audiogenic stress, developmental biology
instability, 189 Biomedicine, behavioral stress research.
Autonomic nervous system. 29 44-45
innervation of lymphoid organs, 140-141 Biopsies, muscle. as stressor, 479-480
Aves, pursuit-deterrence signals. 237 Birds
Awareness, definition, 374 activity patterns, 234
anting, 297-298
body mass loss. breeding season.
B 414
conservation research recommendations.
Baboons 489
antischistosomal drug use. 296 handling effects, 476-477
coronary artery disease, chronic social parental effort. immune effects. 145
stress, 70 plant use, 296-297
dominance relationships radio transmitter effects, 481-482
female reproductive success. 81 use of space, antipredator decision
immunological consequences, 83. 85 making, 256-257
stress reactions, 69, 71 Blood pressure
stimulant use. 295-296 aggressive behavior, 67-69
Baby blues, 344 measurement techniques. 34-35
Bacterial infection. immune response, psychosocial stress, 49
137-138 stress response, 14, 33. 34-35
530 INDEX
Blood sampling Cardiovascular system, see also Heart rate

challenge test, 26-29, 42 chronic stress effects, 34, 70
chemical contamination monitoring, stress response, dominance relationships,
490 55, 66. 67-69
immunological parameters. 40-42 Cotaglyphis bombycina, adaptation, abiotic
methodological problems, 16-18, stresses. 166
23-24 Catastrophes. equivalence to anthropogenic
remote-controlled, 491 stressors. 487
as stressor, 476-477 Catecholamines
Blowfly (Phormia regina). specific hungers, biosynthesis, 31
306 challenge tests. 32-33
Bluegill sunfish (Lepomis macrochinis). circuits, activation, 31
learning, 169 immunosuppressive effect, 39
Blushing, function, 387 as indicator of stressed state, 449
B lymphocytes, 38, 40, 41 Cave bats. handling effects, 476
Body mass Cavia crperea f. poreellus, see Guinea pigs
birds, breeding season, 414 Cellular factors. immune system. 38, 137
as indicator of stressed state, 447 Cellular stress response. 429
male, immunocompetence, 327-328 Censuring disturbance, 474-475
Body temperature, measurement, as Central nervous system
stressor, 479 ACTH effects, 20-21
Bonnet macaques (Macnca radiata), social immune system connections, 38-39
stress, immunological consequences, Centrocercus iirophasianus, sexual behavior,
441-442 energy costs, 160
Boredom, function, 388-389 Challenge test
Brain asymmetry, 205-206 catecholamines, 32-33
Breast asymmetry, 199-200, 328-329 glucocorticosteroids, 26-29
Breeding suppression. predation, 244-245, Chamois (Riipicapra rupicapra),
259 tourism-related disturbance, 464
Bumblebees, escape behavior, 238 Cheetah, inbreeding depression. 495-496
Burramys parvrts, tourism-related stress, Chelonia mynas, tourism-related
415, 417-420 disturbance, 464
Butterflies, wing asymmetry, 197 Chemical contamination, see Pollution
Chemical detection, predators, 263
Chemical immobilization
C remote-controlled, 491
as stressor. 477, 478-479
Calopteryx tnaciilara, sexual behavior. Chen caerrrlescens, hunting effects, 468
energy costs, 159 Chickens
Cannon's fight or flight response, 3-5. skeletal asymmetry, 197-198
10, 62 specific hungers, 305-306
Capricorn (Capra ibex), tourism-related Chimpanzees
disturbance, 464 self-medication, 298-303, 304
Captivity snare wounds, 469
equivalence to natural stressors. Chronic stress
487-488 active vs. passive coping, 10-11
as stressor. 483-485 adaptation, 5-10
Capture, as stressor, 476-477 blood pressure, 67-69
Capture collar, 491 cardiovascular effects, 34, 70
Capture myopathy. 422, 467, 477 glucocorticosteroid-induced effects, 19. 22
INDEX 531
immune response, 138 relationship to physiological stress

mortality. 70-71 response, 7, 8-10, 63, 88-89, 437
Selye's general adaptation, 5-6, 62 testosterone level, 75-76
social experience. 64 Control groups, 42.5
Circus aencginosics, tourism-related Cooperation training, 492-493
disturbance, 464 Coping strategies
Cliff swallow (Hiruntlo pyrrhonala), active. 10
parasite stress, energy costs, 160-161 appraisal, 12-13
Climatic warming. 472, 487 passive, 3, 5 , 10-1 1
Colicmba livia, parasite stress, 160 relationship to social status and
Community composition, as indicator of physiological stress response, 88-89
stressed state, 452 stress concept, 411, 412-413
Complement system, assessment, 41 Coping systems
Confinement, equivalence to natural relevance of feelings, 393-394, 395
stressors, 488 welfare assessment, 400
Confounding factors, experimental design, Copulatory orgasm, females
425-426 female choice, 335-337
Consciousness with symmetrical men, 328
definition. 373. 374 Coronary artery disease, chronic social
evolution, 377-379 stress, 70
Conservation biology, 405-491 Corticosteroid binding proteins. 432. 434
activity-related stressors Corticosterone
attention given. 473-474 dominance relationships, 21-22, 67
examples, 415. 474-486 species variation, 19
minimizing consequences. 488-494 Corticotrophin-releasing hormone
case studies glucocorticosteroid release, 19
Antarctic penguin, 462-463 physiological effects, 140
mountain pygmy-possum. 417-418 postpartum depression, 345
North Sea harbor seal, 453-456 Cortisol, species variation, 19
whale watching, 458-462 Cougar (Fehs concolor), hunting effects.
developmental instability, 195-197 466
equipment, improvements, 490-491 Courtship, predation, 243
experimental design issues, 423-428. Cows, asymmetry, milk production, 199
489 Crayfish, nocturnal activity pattern, 234
importance of stress, 494-496 CRH, see Corticotrophin-releasing
stress concept, 407-414 hormone
stress research Crickets, mate choice, predation, 242
contributions, 494-495 Crippling, consequences, 469
experimental design issues, Crocuta crocuta
423-428 food aversions, 307
framework, 414-415 snare wounds, 469
measurement issues, 421-423 Crowding, experimental
questions about stress, 41.5, 417-420 immunological consequences, 81-82
Conservation ecology, self-medication, population regulation, 43-44
implications, 308 Crying
Conservation genetics, molecular adult women, fitness benefits, 350
techniques. 490 infant. see Infant crying
Control, see also Coping strategies Cry pitch, offspring condition, 354-355
definition, 8 Cryptomys damarensis, energy limits,
developmental. energy costs, 184-185 nonsexual behavior, 158-159
532 INDEX
Culture. postpartum depression. 351 Development rate, mate choice, 170

Cyprinodon pecosensis. sexual behavior, Die1 drift periodicity, stream insects.
energy costs, 160 234
Cytokines, 38. 41. 137. 141-142 Die1 migration
fish, 233
zooplankton, 23 1-233
D Diet selection
learned aversions, 306-307
Daily activity patterns, birds, 234 parasitism. 292, 308-309
Damara mole rats (Cryptornys rlamnrmsis), predation. 224
energy limits, nonsexual behavior, relationship to stress, 147
158-159 self-medication, 309
Damsel flies (Calopteryx maculntn). sexual social learning, 307
behavior, energy costs, 159 specific hungers, 305-306
Dasyurid marsupials (Antechinus). Differential parental solicitude
stress-induced male mortality, 1-2. infant crying, 353
434-43s postpartum depression, 345-346
Data loggers, 491 Directional asymmetry. 183
Death, see Mortality Directional selection. 190-191, 324
Delichon urbicn, energy limits. nonsexual Disease
behavior. 159 adaptation, 6
Dendritic atrophy, glucocorticosteroid- equivalence to anthropogenic stressors.
induced, 19 4x7
Depression induction, 40
females vs. males, 341 relationship to stress. 39-40
function, 388 resistance to
postpartum, 341-352 behavioral, 137
Despair, function. 389. 394 definition, 136
Development, environmental. equivalence genetically based. 136-137
to natural stressors, 487 immune-mediated, 136-137
Developmental control. energy costs. secondary sexual characters. 323
184- 185 social stress, 81-8.5
Developmental stabilityhstability stress-induced impairment, 442-443
animal welfare. 197-199 Displacement activities, as indicator of
attractiveness, 328 stressed state, 449
behavioral, 201-206 Diurnal activity patterns. 233
conservation biology. 195-197 Diurnal rhythms. physiological
definition, 181 parameters. 17
directional selection. 190-191 Divergence
environmental causes, 184. 188-190 resource heterogeneity. 167
fitness correlates. 192-193 resource polymorphisms. 167-168
future studies. 206-207 Dogs, stress reactions, predictability. 7
genetic causes, 185-188 Dominance relationships, see also Social
human and veterinary medicine, 199-201 stress
as indicator of stressed state, 452 female. reproductive success, 76-81
measurement, 181-1 84, 201-206 male
monitoring, 194- 195 female choice. 85-86
morphological, 204-206 physiological costs, 85-88
practical uses, 193-206 role of ACTH. 20-21
types. 321-323 role of corticosterone, 21-22. 67
INDEX 533
stress responses, 45-89 Energetics
active. 62. 63. 88-89 adrenocortical response. 435-436
assessment of rival, 58-61 anthropogenic stressors, 439-441
cardiovascular, 55. 66, 67-69 predator avoidance, 223
establishment of dominance, 51-58 Energetic stress
female, 72 postencounter resumption of activity, 236
gonadal, 73-81 state-dependent risk taking, 217-221
guinea pigs, 58-59 Energy balances, 155-156,439-441
immune, 57. 60-61, 81-85 Energy charge, adenine. 134, 446-447
mice, 62-64
Energy costs
passive. 62, 63, 88-89
developmental control, 184-185
pituitary-adrenocortical, 61 -73
learning, 168-169
primates. 69-73, 83
nonsexual behavior, 158-159
rabbits, 64-66
resource polymorphisms, 167-168
rats, 66-69
sexual behavior, 159-162
relationship to control and
predictability, 88-89 speciation, 165-166
sympathetico-adrenomedullary, 61 -73 species boundaries. 162-164
tree shrews, 51-58, 59-61. 76 Energy metabolism. immunocompetence.
testosterone, 63, 67. 73-76 I44
Dopamine, stress response, 140 Environmental contamination, sre Pollution
Dragonflies, ovipositional behavior, Environmental enrichment
predation, 243 immunological consequences, 442
Drosophila as stressor, 484
energy limits. nonsexual behavior, 158 Environmental factors, developmental
habitat preference, 157, 158 instability, 184, 188-190, 194-195
species boundaries, 162, 163 Environmental stress, 41 1
Ducks, crippling consequences, 469 adaptation, 320
Dugongs (Dngong dugon) adrenocortical response sensitivity,
handling effects, 477 431-432, 433-434
pursuit hunting, 467 prenatal and perinatal impact, 443, 445
Dwarf mongooses (Hrlogule purvula), resistance, genetic variation, 440, 495
dominance relationships resource allocation, 440
reproductive success, 78-79 Envy, function, 390
stress responses, 72 Enzyme activities, sympathetico-
adrenomedullary activation, 33
Epidemiological studies, 15, 39-40
E Epinephrine
biosynthesis. 31
Eating pleasure, function, 391 physiological effects, 30-31
Echinoderms, escape behavior, 238 secretion, 29
Ecological stress, 31 9-320 stress response, 10. 140
Ecosystem regulation, top-down, indirect Error
effects, 260 measurement. fluctuating asymmetry,
El Niiio, 472 183-184
Emotions statistical tests, 426-427
functional significance, 342 Escape behavior, postencounter. 237-238
initiation of feelings. 372 Estrogen
stress response, 6-7, 10, 50-51 facial secondary sexual traits. 325
Endotoxin. 147 physiological effects, 36
534 INDEX
European wild rabbits (Orvcrolagus pleasant vs. unpleasant. 374-375

cuniculus) relevance to welfare, 395. 397. 398-400
female, social rank, 78 Female choice
physiological stress responses. 46-47 costs. 161-162
social bonds. stress-reducing effects, dominance relationships, 85-86
94-98 orgasm, 328, 335-337
Exhaustion stage, general adaptation sexual arousal, 335-337
syndrome. 6 Females
Exhilaration, function. 391-392 dominance relationships, 72. 76-81.
Experience 85-86
early, handling. 491 mate choice, see Female choice
social secondary sexual characters,
role in stress responses, 12-13. 64 immunocompetence. 328
self-medication, 307 sexual behavior, humans, 332-341,
stress concept, 407-408 357-358
Experimental design. conservation biology. Fenitrothion contamination, 457
423-428.489 Ficediila hypoleuca, 159-160
Fight or flight response. 3-5, 10, 62
Fights, see Aggressive behavior
F Finch (Pinaroloxias inornoto), learning. 168
Fish
Face courtship. predation, 243
secondary sexual traits, 324-326 die1 migration, 233
symmetry, 326-327 flight initiation distance. 237
Falconiformes. plant use. 296 predator inspection. 238-239
Father-daughter relations predator-prey population dynamics, 258
postpartum depression, 348-350 resource polymorphisms, 167-1 68
women’s sexual behavior, 334, 337-339 use of space, antipredator decision
Fear making, 255-256
definition, 384 Fitness
function, 384-385 consequences
initiation, 373 adrenocortical response, 431
responses, 384 examples, 423
Feather pulp sample. protein assessment, importance of measuring, 423
490 stress response, 438-439
Fecal samples, 25, 489-490 timescale. 421-422
Fecundity. developmental instability. 193 developmental instability, 192-193
Feeding behavior, see Foraging behavior long-term monitoring. 422
Feelings mating. relationship to energy
adaptive effects. 378-379 consumption, 159-160
characteristics. 372 measurement, 172-173,412. 423
consciousness, 373-374 metabolic efficiency, 171-173
definition, 374 offspring, assessment, 346-347. 353-354,
evidence, 375-376 354-355
evolution, 376-379 psychological pain. 342
examples, 372 reduction. stress concept, 411. 412-413
function. 393-394 relationship to stress level, 156
initiation, 372-373 secondary sexual characters, 325-326
as part of coping systems, 393-394. stress-resistance. 170- 171. 319
395 Flight initiation distance, 237
INDEX 535
Fluctuating asymmetry, 321-323 Genetic factors

definition, 181 developmental instability, 185-188
etiology. 322 resistance to disease, 136-137
heterozygosity, 173 stress response, magnitude, 135
as indicator of stressed state, 452 Genetic variability, 164-165, 202
individual, 181, 185. 192-193, 322 Genotypes. stress-resistant, 169-171. 440,
measurement, 173, 183-184 495
secondary sexual characters, 328-329 Geophagy, primates, 295
sexual selection, 322. 323-324 Gerbils, moonlight avoidance, 235
Follicle-stimulating hormone, 36 Global warming, 472, 487
Foraging behavior Glucocorticosteroids, 19-22
energetic stress and state-dependent risk acute effects, 19
taking, 217-221 challenge tests, 26-29
food selection, see Diet selection circadian oscillation, 143
as indicator of stressed slate, 449-451 immune effects, 142-144
location. 224 long-term effects, 19, 22
mortality/growth rate rule. 221-222 measurement. 23-29
patch choice, 222-223 ratio of free to bound, 24
role of modeling, 225 synthesis and release. 22
time in patches, 223 transport proteins, 22
Foraging theory. self-medication, Gobies, nest building and defense.
308-309 predation, 244
Fractal dimension, as measure of Gonadal activity, see also Pituitary-gonadal
developmental instability, 203-204 system
Freezing response. function, 384 adrenocortical activity
Freudian theory, female sexual arousal, negative coupling, 432-434
337 uncoupling, 434-435
Frogs dominance relationships, 73-81
courtship, predation. 243 as indicator of stressed state, 447-448
seasonal activity patterns, 235 psychosocial stress, reproductive success,
sexual selection. energy costs, 159 80-81
Frustration stress response
function. 386-387 active vs. passive, 14
initiation, 373 assessment, 35-37
jealousy, 390 Gonadotropin. suppression, psychosocial
Fur rubbing, mammals. 298 stress, 80
Gonadotropin-releasing hormone, 36
Good genes hypothesis
G good genotype vs., 172
stress-resistance, 170-171
Gastric ulceration. rats, physiological stress Gorillas (Gorilla gorilla), dental
response, 8-9 asymmetry, 196
Gemsbok (Oryx gnzella), horn asymmetry, Great crested grebes (Podiceps crisratus),
196-1 97 boating-induced stress, 413
Gene coadaptation, 188 Great tits (Parus major)
General adaptation syndrome, 5-6, 62 sexual behavior, energy costs,
Genetic diversity 159-160
measurement, 419 song drift, 203
preservation. 419, 495 Green turtles (Chelonia mydas),
under stress, 164-165, 419-420 tourism-related disturbance, 464
536 INDEX
Grief Heart rate. see rrlso Cardiovascular system

displays, fitness benefits. 349-350 anthropogenic stress. 440-441
function. 385-386 dominance relationships
recognition. 375 rabbits, 66
Grizzly bear (Ursus arcfos), hunting effects, tree shrews. 55
467 as indicator of stressed state. 449
Group choice, predation, 239 as indicator of sympathetico-
Growth hormone, immune effects. 144 adrenomedullary activity. 33-34
Growth performance, developmental measurement techniques, 34
instability, 193 Heat shock proteins. 166. 429
Guilt, function, 387 Helogale parvulm. 72, 78-79
Guinea pigs Helplessness
maternal-infant separation. 92 learned. 388
social bonds, stress-reducing effects. 94 passive chronic stress, 10-1 1
stress responses Hemifaces. attractiveness, 326-327
dominance relationships, 58-59 Hemipteran insects. mating dynamics.
role of social experiences, 12- 13 predation. 242
Guppies. mate choice, predation, Herpes viruses. stress-induced reactivation,
242 443
Heterocephahis glcrber, reproductive
success. social rank, 80
H He terozygosi ty
developmental instability. 173. 187
Habitat preference facial features, averageness. 327
energy balance, 156-158 relationship to fitness, 164-165. 171-173
hunting effects. 467-469 Hippocampus, glucocorticosteroid
Habitat quality, 451-452 effects. 19
Haemutopits ostralegus. tourism-related Homeostasis
disturbance, 463 measurement. 173
Hair samples, 490 stress concept, 407. 410-41 1
Hamadryas baboons (Papio hrmiotiryus). Homozygosity. developmental instability,
cardiovascular stress responses. 186
dominance relationships. 69 Hormone concentrations
Handling catecholamines, 31-33
early experience, 491 fecal, 25
equivalence to natural stressors, 486. glucocorticosteroids, 25-29
487 salivary, 25-26
fitness consequences, 476-477 serum o r plasma. 23-24
physiological response. 16- 17. 473-474. sex hormones, 37
475-476 urinary. 24-25
predictable. 493 Hormones
Happiness. function, 392 chemical immobilization and anesthesia
Harbor seals effects, 478
North Sea. case study. 453-456 initiation of feelings. 372
stress-induced immunosuppression, linkage to stress and
44 1 immunocompetence, 142-145
Health psychological influences, 7
developmental, metabolic rate, 328 rhythms, 17
infant, postpartum depression. 347-348 House martin (Delichon iirhica), energy
relationship to animal welfare. 396, 397 limits, nonsexual behavior, 15Y
INDEX 537
House mice (Mus rniucuhcs), stress behavioral conditioning, 141, 144

responses. 12 cellular factors, 38. 137
Housing, animal central connections, 38-39
equivalence to natural stressors. 487-488 chemical immobilization effects. 478-479
as stressor. 16, 483-485 compensation, 146
Humans components, 38
attraction and attractiveness, 321-332. costs, 139. 146
356-357 energy metabolism, 144
developmental stability. 321-323 enhancement
secondary sexual characters, 323-324 socially induced. 100
stress resistance, 324-331 stress-induced, 138, 143-144
infant crying, 352-356, 358-359 environmental enrichment effects, 442
postpartum depression, 341-352.358 handicap hypothesis, 323, 350
sexual behavior. women, 332-341. humoral factors, 38, 40-41. 137
357-358 innatelnatural immunity, 37-38. 137
social bonds, stress-reducing effects, 98 prenatal and perinatal stress, 443, 445
stress response regulation. 137- 138
adrenocortical, 63, 64 resistance to disease. 136-137
sympathetico-adrenomedullary. 64 robustness, 138
Humoral factors, immune system, 38, secondary sexual characters, 86, 323,
40-41. 137 327-328
Humpback whale (Megoprero stress response, 14, 38-39
novueungliue), response to tourism, 459 assessment. 37-42
Hunger, function, 383 compensatory, 138
Hunting dominance relationships, 57, 60-61,
behavioral effects, 467-469 81-85
crippling. consequences, 469 hormonal linkage, 142-145
physiological and fitness consequences, neurological linkage, 140-142
465-467 relationship to disease, 39-40
sports, 486. 487 testosterone, 87-88
Hybridization, developmental instability, suppression
187-188 catecholamines, 39
Hypothalamo-pituitary-adrenocorticalaxis, sex hormones, 87-88, 325
see Adrenocortical response; Pituitary- socially induced, 90-91, 92-93, 96-98
adrenocortical system stress-induced, 136, 145-149, 441-443
Hypothalamo-pituitary-gonadal axis, see tearfulness, 350
Pituitary-gonadal system surgery effects. 479
transportation effects, 485
Inbreeding
I cheetahs, 495-496
developmental instability, 186
Ideal free distribution model, patch choice. Inclusive fitness, definition of stress,
222-223 134-135
Immobilization, chemical Infant
remote-controlled, 491 health, postpartum depression, 347-348
as stressor, 477, 478-479 maternal bonds, 90-92, 346
Immune system lnfant crying
adaptivelacquired immunity, 38. 137 acoustics, offspring condition, 354-355
anesthesia effects, 478-479 honest signal model, 354
assessment problems, 139 offspring need model, 353-354
538 INDEX
parental reactions, 355 energy expenditure. 168-169

phenotypic quality, 352-356. 358-359 self-medication. 305-307
postpartum depression, 358-359 Lepomis macrochinis, learning, 169
Infanticide. 346. 350-351 Lepus americanics, predator-prey
Infection, see also Disease population dynamics, 258-259
assessment, fecal samples, 489 Lesser snow geese (Chen caerulescens),
susceptibility. stress-induced, 42, hunting effects, 468
147- I48 Lethality of effects, stress concept, 41 1
Intelligence test, body asymmetry, 206 Leukocytes, assessment, 40-41
Interleukin 1 Leydig cells, 36
circadian oscillation, 143 Life history models, stress response traits,
linkage to stress and immunocompetence. 437-438
141-142, 143 Life-span. reproductive success, 170
Internal changes, stress concept, 409 Lizards
Interspecific competition courtship, predation, 243
behaviorally transmitted indirect effects, flight initiation distance, 237
259-261 parasite stress, energy costs, 160
developmental instability, 192 species boundaries, thermal environment.
Interventions 163-164
optimizing. 493-494 Loneliness, function, 389
stressful, see Conservation biology, Lunar activity patterns, 234-235
activity-related stressors Lust, function. 390
Intraspecific competition, developmental Luteinizing hormone, 36
instability, 192 Lycaon pictus, dominance relationships,
Invasive procedures. alternatives, stress responses. 72
489-490 Lymphocytes. 38. 40, 41, 138
Invertebrates, use of space, antipredator Lymphoid organs, sympathetic innervation.
decision making, 255 140-141
M
J
Macaca fasciciilrzris. 70, 71, 93
Java monkeys (Maraca fascicularis) Macaques
coronary artery disease, chronic social maternal-infant separation, 90-91
stress, 70 social stress. immunological
dominance relationships, stress consequences, 441-442
reactions, 71 Malaise, function, 381-382
social bonds, disruption, 93 Males
Jealousy, function, 390 dominance relationships
female choice, 85-86
physiological costs, 85-88
father-daughter relations, 334, 337-339.
L 348-350
mating tactics, predation. 242
Laboratory selection experiments, 191 secondary sexual characters,
Lagopiis lagoptis. hunting effects, immunocompetence, 327-328
468 stress-induced mortality, 1-2,
Lagopus muhis, tourism-related 434-435
disturbance, 463 Mammals
Learning, see also Social experience active vs. passive stress responses.
developmental instability, 207 62. 63
INDEX 539
conservation research recommendations, Mice

489 crowding stress. 165
female, social rank. reproductive success, dominance relationships
76-81 gonadal endocrine activity, 74-75
fur rubbing, 298 stress responses, 62-64
marine. hunting effects. 467 stress-induced immunosuppression,
self-medication, 293. 298 81-82. 84-85,442
social stress, 42-106 stress responses. 12. 62-64
specific hungers. 306 Microtiis ochrogaster, social bonds.
use of space, antipredator decision stress-reducing effects, 100, 103
making, 256-257 Mineralocorticosteroids. 22
Manatees (Trichechus manatus). Molecular markers, stressed state, 446-447
tourism-related disturbance. 465 Mole rat (Spalax ehrenhergi), energy limits,
Manual restraint. as stressor, 477-478 nonsexual behavior, 159
Marine animals Molluscs, escape behavior. 238
hunting effects, 467 Monkeys, maternal-infant separation, 90-91
noise disturbance, 471 Moonlight avoidance, 235
Marmoset monkeys, social rank, Mortality
reproductive success, 79-80 aggressive behavior, 1-2, 434-435
Marrnota monax, flight initiation distance, chronic social stress. 70-71
237 parasitism. 147-149
Marsh harrier (Circus aeruginosirs), psychosocial stress, 49-51
tourism-related disturbance, 464 stress-induced emotional arousal, 50-51
Mate choice Mortality/feeding rate (pff) rule. 221-222
costs, female preference, 161-162 Mortality/growth rate ( d g ) rule
development rate, 170 diet selection, 224
female, dominance relationships, 85-86 optimal behavior, 221 -222
predation, 241-242 patch choice, 223
Mating Mother-daughter relations
predation, 242-243 postpartum depression. 338
stress-resistance, 169-170 women's sexual behavior, 334-335
Mayflies Mother-infant bonds. 90-92, 346
nocturnal activity pattern, 234 Mountain goats (Oreamnos americus),
predator-induced changes, 245, 248 tourism-related disturbance. 464-465
Measurement issues Mountain pygmy-possum (Burramys
conservation biology research, 421-423 parvus), tourism-related stress, case
stress induction. 476-477, 479 study, 415,417-420
Medicine, see also Self-medication Movement, prey activity, 226
Darwinian. 199-201 MRNA, as indicator of stressed state, 446
developmental instability, 199-201 Mule deer, tourism-related disturbance, 464
Megapiera novaeangliae, response to Muscle biopsies, as stressor, 479-480
tourism, 459 MUSmusculus, stress responses, 12
Metabolic energy. as indicator of stressed Mutations, developmental instability, 188
state. 446-447 Myopathy, capture, 422, 467, 477
Metabolism
adrenocortical response, 435-436 N
developmental health. 328
heterozygosity, I71 -173 Naked mole rats (Heterocephalus glaher),
stress response, 14 social rank, reproductive success, 80
Metal pollution, developmental instability. Natural catastrophes, equivalence to
19s anthropogenic stressors, 487
540 INDEX
Natural experiments, design issues, 424 Orgasm, female

Natural killer cells, 38, 40, 41 costs, 336
Natural stressors female choice, 328. 335-337
equivalence to anthropogenic stressors. frequency
486-488 father-daughter relations, 338-339
examples, 415 female choice, 335-337
Nausea, function. 382 value judgments, 340-341
Nepotism, physical attractiveness, 331 Ornaments, see also Secondary sexual
Nervous system characters
asymmetry, 205-206 energy costs, 161-162
autonomic, 29, 140-141 Oryctolagus cuniculus, 46-47, 78, 94-98
central, 20-21, 38-39 Oryx gazella, horn asymmetry, 196-197
initiation of feelings, 372 Osmoregulation, stress response, 14
sympathetic, see Sympathetico- Ovaries, stress response, 36, 37
adrenomedullary system Ovipositional behavior, predation, 243
Nest material, antibacterial, 296-297 Ovis canadensis, heterozygosity and fitness.
Neuroendocrine stress response, see Stress 164
response Oxytocin
Nociceptive systems, 380 female orgasm, 336
Nocturnal activity patterns, 234 social bonds, 103
Noise stress, 189, 469-471 stress-reducing effects, 103
Noninvasive procedures, 489-490 Oystercatchers (Haematopus ostralegus),
Norepinephrine tourism-related disturbance, 463
biosynthesis, 31
secretion, 29
stress response. 10. 140
P
North Sea harbor seals, case study, 453-456
Nutrition, see Diet selection
Pain
Nutritional stress, 447, 451
definition. 373, 379-380
developmental instability, 188-189
extreme, 381
immune effects. 139
function, 381
initiation, 373
responses, 380-381
0 system, 380
Pair bonding
Observational studies, design issues, 424 costs of orgasm, 336
Odocoileris virginianus, 164, 237, 468-469 stress-reducing effects, 98-100. 103
Offspring fitness, parental assessment, Pantherrr leo. pursuit hunting, 467
346-347. 353-354 Papio anribis
Oil fouling. consequences, 456-457 chronic social stress, coronary artery
Olive baboons (Papio anubis) disease, 70
coronary artery disease, chronic social dominance relationships
stress, 70 female reproductive success, 81
dominance relationships immunological consequences, 83
female reproductive success, 81 Papio hamadryas, dominance relationships,
immunological consequences, 83 cardiovascular stress responses, 69
Omegas, 69 Paragliding disturbance, 464
Orerimnos umericus, tourism-related Parasitism
disturbance, 464-465 contribution to mortality, 147-149
Organochlorines, 455. 490 definition, 291-292
INDEX 541
developmental instability, 192, 200 sympathe tico-adrenomedullary

diet selection, 292, 308-309 activity. 33
effects on host behavior, 292 Phoca vitulina, stress-induced
energy costs, 161 immunosuppression, 441
human attractiveness, 326 Phocine distemper virus, 453-454
sexual behavior. 160-161 Physical activity, immune effects, 147
susceptibility, stress-induced. 42, 147-148 Physical attractiveness, see Attraction and
Parental investment attractiveness
physical attractiveness. 331 Physiological parameters
postpartum depression, 345-351 assessment of stress, 15-16
Parental solicitude evidence of feelings, 376
infant crying, 353 rhythms, 17
postpartum depression, 345-346 Pied flycatchers (Ficedula hypoleiica).
Parent-daughter relations, women’s sexual sexual behavior, energy costs. 159-160
behavior, 332-341, 337-339 Pigs. stress-induced immunosuppression,
Parenting, predation, 243-244 82-83, 442
Parent-offspring conflict theory, 345-346 Pigtail macaques (Macaca nemestrina).
Pariis miijor social stress, immunological
sexual behavior, energy costs, 159-160 consequences, 83,441-442
song drift, 203 Pinaroloxias inornata, learning, 168
Passerines, plant use, 296 Pink-footed geese, tourism-related
Passive stress response, 108 disturbance, 464
chronic stress, 10-11 Pipefish (Syngnathus typhle). mating
dominance relationships, 62. 63, 88-89 dynamics. predation, 243
Pituitary-adrenocortical system
gonadal activity, 14
assessment, 19-29, 448
maternal separation, 90, 91
dominance relationships, 61-73
Selyean. 5-6. 62
immune system connections, 39
tree shrews, 54-58
regulation, 140
Patch choice. 222-223
stress response, see Adrenocortical
Patch location, 224
response
Patch residence times, 223
Pituitary-gonadal system. see also Gonadal
Paternal investment
activity
postpartum depression, 348-350
psychosocial stress. reproductive success,
women’s sexual behavior, 334. 337-339
80-81
Pathogens. see Disease stress response, assessment, 35-37,
PCB contamination, 455. 457, 490 447-448
Penguins. tourism-related disturbance, case Plant breeding, developmental instability,
study. 462-463 I91
Perinatal stress, 443, 445 Plant use. see also Self-medication
Petaurus breviceps, dominance birds, 296-297
relationships, testosterone, 73-74 secondary metabolites, 293
Phagocytes. 38, 41 Pleasure. 374-375
Phenodeviant behavior achievement, 393
morphological basis, 204-206 eating, 391
repeatability. 202-203 sensory, 392-393
Phenotype deviance, 183.321, 352-356 sexual, 391-392
Pheny lethanolamine-N-methyltransfcrase Plecoptera, species boundaries, 163
epinephrine conversion, 31 PNMT, see Phenylethanolamine-N-
regulation, variability, 31 methyltransferase
542 INDEX
Podiceps crisrarus. boating-induced stress. Prenatal stress

413 impact, 443.445
Pollution primates, noise disturbance. 470
case study, 453-456 Preputial glands, social rank, 75
developmental instability, 189, 194-195 Primates
equivalence to natural stressors. 487 conservation research recommendations,
experimental studies, 456-457 48’)
monitoring techniques, 490 dominance relationships
Population censuring, disturbance, 474-475 immunological consequences, 83
Population density reproductive success, 77-78
developmental instability, 189 stress responses, 69-73
equivalence to anthropogenic stressors. geophagy. 295
488 noise disturbance, 470
as indicator of stressed state, 451-452 Prisoner’s dilemma, fish, 239
social stress, 42-44 Progesterone, physiological effects, 36
Population dynamics Prolactin, immune effects, 145
predator-prey interactions, 257-259 Proteins
stress response, 430-431 assessment, feather pulp sample, 490
Population persistence, 41 1-412, 420 corticosteroid binding, 432, 434
Postpartum depression, 341-352, 358 heat shock, 166, 429
continuum, 344-345 transport, 22. 24
cultural traditions, 351 Pseudoreplication, 427-428
infant crying, 358-359 Psychical stress, 6-7
infant health, 347-348 Psychological pain
infanticidal ideation, 350-351 as evolutionary adaptation, 341-343
parental investment behaviors, 345-351 postpartum, 343-345
paternal investment, 348-350 Psychology, stress concept, 407-408
prevalence, 343-344 Psychoneuroimmunological research, 45
resource base, 350 Psychosocial stress, 48-50
social support, 347-348, 350 mortality, 49-51
tearfulness, fitness benefits, 350 reproductive success, 80-81
Postpartum psychosis. 344 Ptarmigan (Lagopus nzutus),
Power, statistical tests, 426-427 tourism-related disturbance, 463
PPD, see Postpartum depression Pupfish (Cyprinodon pecosensis). sexual
Prairie vole (Microrus ochrogaster), social behavior, energy costs, 160
bonds, stress-reducing effects, 100, 103 Pursuit-deterrence signals, 236-237
Predator-induced stress, see also Pursuit hunting, 465-466
Antipredator decision making
definition. 215-216 R
equivalence to anthropogenic stress,
486-487 Rabbits, see European wild rabbits
long-term costs, 245, 248 (Oryctolagus cuniculus)
physiological response, 261-262 Radiation, developmental instability,
role of modeling, 225 194-1 95
Predictability Radio implantation, as stressor. 480
handling effects, 493 Radio-tagging. as stressor, 480-483
relationship to stress response. 7. 8-10, Radiotelemetry
63.88-89,437 blood pressure, 35
testosterone level, 75-76 heart rate, 34
Preference tests, role, 397-398 improvements, 49 1
INDEX 543
Rails (Aves). pursuit-deterrence signals, Reptiles, use of space, antipredator
237 decision making, 256
Random selection. 423-425 Resistance stage, general adaptation
Rangifer tarandus, 468 syndrome, 5-6
Rape, victims’ anguish, 342-343 Resources
Rats heterogeneity, adaptation, 166-167
diurnal activity pattern. 233 polymorphisms, energy expenditure,
food selection, 305. 306-307 167-168
maternal-infant separation. 92 Resource stress
social stress. immunological allocation differences, 439-441
consequences, 82, 84 equivalence to anthropogenic stressors,
stress response 487-488
dominance relationships, 66-69 postpartum depression, 350
predictability and control, 8-9 sexual behavior, 160,333-334.338-339,
Razorfish (Xvrichtys splendens), mating 340
dynamics, predation, 242-243 Restraint, as stressor. 477-479
Reciprocity. physical attractiveness, Rhesus monkeys (Macaca mularta)
330-331 dominance relationships, 70-71
Recreation disturbance, see Tourism/ immunological consequences, 83
recreation-related disturbance stress responses, 69. 72
Red deer maternal-infant separation, 90
hunting effects, 466, 468 social bonds, disruption, 92-93
noise disturbance, 470 stress-induced immunosuppression, 441
Red-winged blackbirds (Agelaius Rhythms
phoeniceus), food aversions. 307 die1 drift periodicity, 234
Refuging, prey activity, 226 disturbance, timescafe, 421
Reindeer (Rongifer tarandus), hunting physiological parameters, 17
effects, 468 Risk assessment
Release, soft, 492 antipredator decision making, 262-263
Repeatability coping style, 12-13
individual fluctuating asymmetry, 185 dominance relationships, stress responses,
as measure of developmental stability, 58-61
202-203 Rock doves (Columbia livia), parasite
Reproduction. see also Gonadal activity stress. energy costs, 160
adrenocortical response, 431-436 Rodents, moonlight avoidance, 235
costs, parasitism, 148 Rubia, self-medication, chimpanzees,
decision making, predation, 241-245 302-303,304
dominance relationships, females, 76-81
parental effort, immune effects, 145
phenotypic variation, stress resistance, 5
319, 324-330
prenatal stress, 43-44, 73 Sage grouse (Cenfrocercus urophasianus),
social rank, female mammals, 76-81 sexual behavior. energy costs, 160
social stress, 43-44, 73 Saharan silver ant (Cataglyphis bombycina),
Reproductive system adaptation, abiotic stresses, 166
assessment, molecular techniques, 490 Saimiri sciureus, see Squirrel monkeys
state. as indicator of stressed state. Saliva samples, 25-26, 490
447-448 Salt, specific hunger, 306
stress response, 14 Sample size. 426
544 INDEX
Sceloporirs occidentalis. stress response, Sex steroids

variability. 430-431 stress response, 22
Schizophrenia, phenodeviance, 203 transport proteins, 22
SDP, see Stochastic dynamic programming Sexual arousal, father-daughter relations,
Seasonal activity patterns, 235 337-339
Secondary sexual characters, see also Sexual behavior, women
Sexual selection arousal
asymmetry, 190-191,323-324,328-329 father-daughter relations, 337-339
developmental instability. 192-193, female choice, 335-337
323-324 flexibility. 334
facial, 324-326, 327 parent-daughter relations, 332-341,
immunocompetence. 86, 323, 357-358
327-328 resource level. 333-334. 338-339
nonfacial, 327-330 Sexual pleasure, function, 391-392
stress, 191. 323-324 Sexual selection, see ulso Secondary sexual
Sedatives, 493 characters
Selection developmental instability. 190-191,
directional. 324 192-193
sexual, see Sexual selection energy costs. 159-162
stress-related, 31Y-320 fluctuating asymmetry, 322, 323-324
Self-medication parasite stress, 86, 161
behavioral mechanisms, 304-307 self-medication, 293
chimpanzees, 298-303 stress resistance. 170-171
contexts. 292 humans, 324-330
forms. 294 Sheep, parental effort, immune effects,
future research, 310 145
implications. 308-310 Signal evolution
mammals, 293. 298 fitness impacts, 353
prophylactic, 294-298 infant crying. 350, 354
baboons, 295-296 Sinusitis. facial secondary sexual characters.
birds. 296-298 326
primates, 295 Size-assortative grouping, predation, 240
therapeutic vs.. 294 Sleep, function, 383
secondary plant metabolites, 293 Sneaky copulations, predation, 242
sexual selection, 293 Snowshoe hare (Lepus americnni~s),
skepticism, 303-304 predator-prey population dynamics,
therapeutic, 294, 298-303 258-259
individual learning, 305-307 Social behavior
social learning. 307 nonsexual. stress resistance, 330-331
Selye’s general adaptation syndrome, sexual, see Sexual behavior
5-6, 62 Social b o n d s h p p o r t
Sensation costs of orgasm, 336
initiation of feelings, 372 disruption, 89-93
pleasurable. function. 392-393 oxytocin. 103
Separation. mother-infant, 90-92 postpartum depression, 347-348
Serotonin, stress response, 140 stress-reducing effects. 93-103, 493
Sex determination, fecal samples, 489 Social density. 44
Sex hormones Social experience. see also Learning
measurement. 37 role in stress responses, 12-13, 64
stress response, assessment, 35-37 self-medication, 307
INDEX 545
Social factors social bonds

adrenocortical response. 436-437 disruption, 90, 92
antipredator decision making. 239-241 stress-reducing effects, 93-94
stress response, 89-103, 436-437 Stability, see also Control
Sociality, adaptive. 239-241 developmental, see Developmental
Social rank, see Dominance relationships stability/instability
Social stress, 42-105 relationship to stress response. 7. 8-10,
aggressive behavior. 45-47 63, 88-89,437
behavioral research. 44-45 social systems, 88-89
disruption of social bonds, 89-93 Starlings (Stirrnus vulgaris). plant use.
dominance relationships. see Dominance 296-297
relationships State-dependent life history models, stress
population regulation, 42-44, 73 response traits, 437-438
psychological vs. physical processes. State-dependent risk taking
48-50 empirical studies, 217. 218-219
reproductive success. 80-81 stochastic model. 217, 220-221. 224, 236
resistance to disease. 81-85 Statistical tests, power, 426-427
resource allocation, 440 Stimulant use, baboons, 295-296
social conflict. 45-61 Stochastic dynamic programming. 217,
social support effects, 93-103 220-221.224. 236
tree shrews, 50-58 Stoneflies (Plecoptera). species boundaries,
Social systems, stability, 88-89. 488 163
Social vigilance, 240-241 Stream insects, die1 drift periodicity, 234
Socioeconomic level Stress, see also specific type
postpartum depression. 350 definition, 2. 14-15, 395-396, 407
women's sexual behavior. 340 biological, 291
Sodium, specific hunger, 306 classical. 216
Soft release. 492 evolutionary. 215-216. 412-414
Space, use. predator-induced stress. operational. 134-135
248-257 physical, 291
Spnlm ehrenhergi. energy limits. nonsexual Selyean, 5
behavior, 159 natural history. 414-415, 428-452
Spatial positioning, predation, 239 relationship to disease. 39-40
Speciation. energy costs, 165-166 relationship to welfare, 396. 397
Species boundaries, energy limits, 162-164 systems view. 407
Species interactions. antipredator decision types. 62
making. 259-261 Stress concepts
Species variation. stress response, 430 development. 3-15
Spider mite (Tetraqvchus urticrie). habitat distinguishing criteria, 407-41 1
preference, 1.57 evolutionary. 41 1-414.437-438
Sports hunting, equivalence to natural Stress-odor exposure. mice, immunological
stressors. 486. 487 consequences, 84
Spotted hyenas (Crocirtn crociirri) Stressors
food aversions. 307 anthropogenic, see Anthropogenic
snare wounds. 469 stressors
Squirrel monkeys (Sairniri sciureus) conservation activity-related
dominance relationships attention given, 473-474
stress reactions. 71 examples. 41.5. 474-486
testosterone. 74 minimizing consequences, 488-494
maternal-infant separation. YO definition, 407
546 INDEX
natural, 415, 486-488 Strrrnrrs vulgoris. plant use. 296-297

physical vs. psychical, 6-7 Submission, see Dominance relationships
temporal characteristics, 423 Suffering
Stress proteins, 166, 429 definition, 374-375
Stress resistance function, 389, 394
genotypes, 169-171,440,495 Sugar. specific hunger, 306
metabolic rate, 435-436 Sugar gliders (Petourus hreviceps),
Stress response dominance relationships, testosterone.
active, see Active stress response 73-74
as adaptation, 438-439 Surgery, as stressor, 479-480
adrenocortical, see Adrenocortical Survival aids, 378
response Survival handicaps. 323. 350
adrenomedullary, see Sympathetico- Survival rates, developmental instability,
adrenomedullary system, stress 193
response Survival variants, 165
antipredator decision making. 261-262 Swine, social bonds, disruption, 93
assessment, 445-452 Symmetry, see also Asymmetry
epidemiological approach, 15, 39-40 facial, 326-327
methodological issues. 16-18. 420- selection pressure, 184
428 Sympathetico-adrenomedullary system, 20
molecular markers. 446-447 active chronic stress, 10
organisrnal markers, 446-450 fight or flight response. 3
physiological approach, 15-16 immune system connections, 39
physiological markers, 18-42 physiological effects, 29-31
Cannon’s, 3-5, 62 regulation, 140
cellular, 429 in Selyean concept, 5
conceptual issues. see Stress concepts stress response
definition, 407 assessment, 29-35, 31-35, 448-449
emotions, 6-7, 10, 50-51 dominance relationships, 54-55, 56-57.
magnitude, genetic component, 135 61-73
modulating factors, 416, 429. 430-431 rabbits, 46
multiple components. 421 tree shrews, 54-55, 56-57
nonspecific. 7-8, 14 Syngnrrthits typhle, mating dynamics,
passive. see Passive stress response predation, 243
physiological pattern, 13-14 Syrian hamsters, dominance relationships,
predictability and control, 7, 8-10, 63, immunological consequences, 82
88-89.437
Selye’s, 5-6, 62
social factors, 89-103. 436-437 T
sympathetic, see Sympathetico-
adrenomedullary system. stress Tagging, as stressor, 480-483
response Tamarins, social rank, reproductive success,
theories. see Stress concepts 79-80
timescale, 421-422 Tearfulness. fitness benetits, 350
variability, 429-439. 438 Temperature, measurement. as stressor, 479
Stress state Temperature deviations
definition. 407 climatic, 472. 487
indicators, 445-452 developmental instability. 188
organismal, 446-450 equivalence to anthropogenic stressors,
population. 450-452 487
INDEX 547
Temporal patterns Translocation

as indicator of stressed state. 449-450 equivalence to natural stressors. 488
prey activity, 231-235 soft release, 492
Testes. stress response, 35-36, 37 as stressor, 485-486
Testosterone Transmitters
aggressive behavior, 75. 86-87 improvements, 491
dominance relationships, 63, 67, 73-76 radiotelemetry, 34, 35
facial secondary sexual traits, 325 weight, 481-482
immunological consequences, 87-88, 325 Transportation
physiological effects, 35 equivalence to natural stressors. 488
predictability and control, 75-76 as stressor, 485-486
stress effects, 36-37 Transport proteins, 22, 24
Tefranychus urficae, habitat preference. Tree shrews
157 challenge test, 27-29
TH, see Tyrosine hydroxylase social bonds, stress-reducing effects,
Thermal discomfort, function. 383-384 98- 100
Thiarubrine-A, 300, 301 stress response
Thirst active and passive, 54-58
function. 383 dominance relationships, 51-58,
initiation, 373 59-61, 76
Thymus, noradrenergic innervation, new housing, 16
140-141 social stress, 50-58
Thyroid hormone, immune effects, 144 Trichechus manatus, tourism-related
Tidal activity patterns, 235 disturbance, 465
Time Tyrosine hydroxylase
patch residence, 223 adrenal concentration, 31
stress concept, 408-409 catecholamine biosynthesis, 31
Timescale as measure of sympathetico-
animal welfare assessment, 398 adrenomedullary activity, 33
conservation biology research, 421 -422
Tiredness, function, 383
Tit-for-tat strategy, fish, 239 U
Titi monkeys (Cdlicehus moloch), social
bonds, stress-reducing effects, 94 Ulcers, rats, predictability and control, 8-9
T lymphocytes, 38, 40. 41 Ungulates, pursuit-deterrence signals, 237
Top-down ecosystem regulation, indirect Urbanization, equivalence to natural
effects. 260 stressors, 487
Torpor, as indicator of stressed state, 447 Urine samples, 24-25, 451, 490
Tourism/recreation-related disturbance, Urscts americanus, tourism-related
457-465 disturbance, 464
Antarctic penguins. case study, 462-463 Ursus arctos, hunting effects, 467
development-related stress, case study,
415. 417-420
experimental studies. 463-465 V
hunting, 465-471
whale watching, case study, 458-462 Vaccination
Training long-term monitoring, 422
antipredator, 492 as stressor, 483
cooperation, 492-493 Variability, genetic, 164-165. 202
Tranquilizers. 493 Variants, survival, 165
548 INDEX
Vernonia, self-medication, chimpanzees, Wildlife

301-302 conservation research recommendations,
Vervet monkeys (Cercopifhecus nethiops 489
sohoeus), dominance relationships, hunting effects, 465-469
stress reactions, 71 noise disturbance, 469-471
Veterinary medicine, developmental tourism-related disturbance, see Tourism/
instability, 199-201 recreation-related disturbance
Vigilance. social, 240-241 vaccination effects, 483
Viruses. stress-induced reactivation, 443 Willow grouse (Lagopus Iugopirs). hunting
Visitor disturbance, see also Tourismi effects, 468
recreation-related disturbance Woodchucks (Morninfa monux), night
equivalence to natural stressors, 486. 487 initiation distance. 237
Vocalizations, behavioral asymmetry,
204-205
Voles, breeding suppression, predation. X
244-245. 259
Xvrichfys splendens, mating dynamics.
predation. 242-243
W
Waist-to-hip ratio, phenotypic quality, Y

328-329
Water, specific hunger, 305-306 Yellow baboons, dominance relationships,
Waterfowl, hunting effects, 467-468 immunological consequences, 85
Water striders
mate choice, predation, 242
mating dynamics, predation, 242 2
Welfare, see Animal welfare
Western fence lizard (Sceloporcrs Zahavian paradigm, signal evolution, 350,
occidentrilis),stress response, 353,354
variability, 430-431 Zona fasciculata
Whales enlarged, long-term stress. 19
hunting effects, 466-467 glucocorticosteroid production, 19
tourism, case study, 458-462 Zona glomerulosa. mineralocorticosteroid
White-tailed deer (Odocoileus virginionus) production, 22
heterozygosity and fitness. 164 Zooplankton. die1 vertical migration,
hunting effects. 468-469 23 1-233
pursuit-deterrence signals, 237 Zoos, housing stress, 484
~~ ~~ ~~ ~
Contents of Previous Volumes
Volume 16 Behavioral Ecology: Theory into Practice

NEIL B. METCALFE AND PAT
Sensory Organization of Alimentary Behav- MONAGHAN
ior in the Kitten
The Dwarf Mongoose: A Study of Behav-
K. V. SHULEIKINA-TURPAEVA
ior and Social Structure in Relation to Ecol-
Individual Odors among Mammals: Origins ogy in a Small, Social Carnivore
and Functions 0. ANNE E. RASA
ZULEYMA TANG HALPIN
Ontogenetic Development of Behavior:
The Physiology and Ecology of Puberty The Cricket Visual World
Modulation by Primer Pheromones RAYJOND CAMPAN, GUY BEUGNON.
JOHN G. VANDENBERGH AND AND MICHEL LAMBIN
DAVID M. COPPOLA
Relationships between Social Organization
Volume 18
and Behavioral Endocrinology in a
Monogamous Mammal
C. SUE CARTER, LOWELL L. GETZ, Song Learning in Zebra Finches (Taenio-
AND MARTHA COHEN-PARSONS pygia guttata): Progress and Prospects
PETER J. B. SLATER, LUCY A.
Lateralization of Learning Chicks EALES, AND N. S. CLAYTON
L. J. ROGERS
Behavioral Aspects of Sperm Competition
Circannual Rhythms in the Control of in Birds
Avian Migrations T. R. BIRDHEAD
EBERHARD GWINNER
Neural Mechanisms of Perception and Mo-
The Economics of Fleeing from Predators tor Control in a Weakly Electric Fish
R. C. YDENBERG AND L. M. DILL WALTER HEILIGENBERG
Social Ecology and Behavior of Coyotes Behavioral Adaptations of Aquatic Life in
MARC BEKOFF AND MICHAEL C. Insects: An Example
WELLS ANN CLOAREC
The Cicadian Organization of Behavior:
Timekeeping in the Tsetse Fly, A Model
Volume 17
System
JOHN BRADY
Receptive Competencies of Language-
Trained Animals
LOUIS M. HERMAN
Volume 19
Self-Generated Experience and the Devel-
opment of Lateralized Neurobehavioral Or- Polyterritorial Polygyny in the Pied Fly-
ganization in Infants catcher
GEORGE F. MICHEL P. V. ALATALO AND A. LUNDBERG
549
550 CONTENTS OF PREVIOUS VOLUMES
Kin Recognition: Problems, Prospects. and Lekking in Birds and Mammals: Behavioral
the Evolution of Discrimination Systems and Evolutionary Issues
C. J. BARNARD R. HAVEN WILEY
Maternal Responsiveness in Humans: Emo-
tional, Cognitive. and Biological Factors Volume 21
CARL M. CORTER AND ALISON S.
FLEMING Primate Social Relationships: Their Deter-
minants and Consequences
The Evolution of Courtship Behavior in
ERIC B. KEVERNE
Newts and Salamanders
T. R. HALLIDAY The Role of Parasites in Sexual Selection:
Ethopharmacology: A Biological Approach Current Evidence and Future Directions
to the Study of Drug-Induced Changes in MARLENE ZUK
Behavior Conceptual Issues in Cognitive Ethology
A. K. DIXON, H. U. FISCH, AND K. H. COLIN BEER
MCALLISTER
Responses in Warning Coloration in Avian
Additive and Interactive Effects of Geno- Predators
type and Maternal Environment W. SCHULER AND T. J. ROPER
PIERRE L. ROUBERTOUX, MARIKA
NOSTEN-BERTRAND, AND Analysis and Interpretation of Orb Spider
MICHELE CARLIER Exploration and Web-Building Behavior
FRITZ VOLLRATH
Mode Selection and Mode Switching in For-
aging Animals Motor Aspects of Masculine Sexual Behav-
GENE S. HELFMAN ior in Rats and Rabbits
GABRIELA MORAL^ AND CARLOS
Cricket Neuroethology: Neuronal Basis of BEYER
Intraspecific Acoustic Communication
FRANZ HUBER On the Nature and Evolution of Imitation
in the Animal Kingdom: Reappraisal of a
Some Cognitive Capacities of an African Century of Research
Grey Parrot (Psirrnclts erirhncits) A. WHITEN AND R. HAM
IRENE MAXINE PEPPERBERG
Volume 22
Volume 20
Male Aggression and Sexual Coercion of
Social Behavior and Organization in the Females in Nonhuman Primates and Other
Macropodoidea Mammals: Evidence and Theoretical Impli-
PETER J. JARMAN cations
The 1 Complex: A Story of Genes. Behav- BARBARA B. SMUTS AND ROBERT
ior. and Population W. SMUTS
SARAH LENINGTON * Parasites and the Evolution of Host Social
The Ergonomics of Worker Behavior in So- Behavior
cial Hymenoptera ANDERS PAPE M0LLER, REIJA
PAUL SCHMID-HEMPEL DUFVA, AND KLAS ALLANDER
“Microsmatic Humans” Revisited: The Gen- The Evolution of Behavioral Phenotypes:
eration and Perception of Chemical Signals Lessons Learned from Divergent Spider
BENOIST SCHAAL AND RICHARD H. Populations
PORTER SUSAN E. RIECHERT
CONTENTS OF PREVIOUS VOLUMES 551
Proximate and Developmental Aspects of The Behavioral Diversity and Evolution of

Antipredator Behavior Guppy, Poecilin rericrtlata, Populations in
E. CURIO Trinidad
A. E. MAGURRAN. B. H. SEGHERS.
Newborn Lambs and Their Dams: The In-
P. W. SHAW, AND G. R. CARVALHO
teraction That Leads to Sucking
MARGARET A. VINCE Sociality, Group Size, and Reproductive
The Ontogeny of Social Displays: Form De- Suppression among Carnivores
velopment. Form Fixation, and Change in SCOTT CREEL AND DAVID
Context MACDONALD
T. G. GROOTHUIS Development and Relationships: A Dy-
namic Model of Communication
Volume 23 ALAN FOGEL
Why Do Females Mate with Multiple
Sneakers. Satellites. and Helpers: Parasitic Males? The Sexually Selected Sperm Hy-
and Cooperative Behavior in Fish Repro- pothesis
duction LAURENT KELLER AND HUDSON K.
MICHAEL TABORSKY REEVE
Behavioral Ecology and Levels of Selec-
Cognition in Cephalopods
tion: Dissolving the Group Selection Con-
JENNIFER A. MATHER
troversy
LEE ALAN DUGATKIN AND
HUDSON KERN REEVE
Volume 25
Genetic Correlations and the Control of Be.
havior. Exemplified by Aggressiveness in
Sticklebacks Parental Care in Invertebrates
T H E 0 C. M. BAKKER .STEPHEN T. TRUMBO
Territorial Behavior: Testing the Assump- Cause and Effect of Parental Care in
tions Fishes: An Epigenetic Perspective
JUDY STAMPS STEPHEN S. CRAWFORD AND
EUGENE K. BALON
Communication Behavior and Sensory
Mechanisms in Weakly Electric Fishes Parental Care among the Amphibia
BERND KRAMER MARTHA L. CRUMP
An Overview of Parental Care among the

Volume 24 Reptilia
CARL CANS
Is the Information Center Hypothesis a
Flop? Neural and Hormonal Control of Parental
HEINZ RICHNER AND PHILIPP HEEB Behavior in Birds
JOHN D. BUNTIN
Maternal Contributions to Mammalian Re-
productive Development and the Diver- Biochemical Basis of Parental Behavior in
gence of Males and Females the Rat
CELlA L. MOORE ROBERT S. BRIDGES
Cultural Transmission in the Black Rat: Somatosensation and Maternal Care in Nor-
Pine Cone Feeding way Rats
JOSEPH TERKEL JUDITH M. STERN
552 CONTENTS OF PREVIOUS VOLUMES
Experiential Factors in Postpartum Regula- Volume 26

tion of Maternal Care
ALISON S. FLEMING. HYWEL D. Sexual Selection in Seawood Flies
MORGAN, A N D CAROLYN WALSH THOMAS H. D A Y A N D A N D R E S.
Maternal Behavior in Rabbits: A Historical G ILB U RN
and Multidisciplinary Perspective Vocal Learning in Mammals
GABRIELA GONZALEZ-MARISCAL VINCENT M. JANIK A N D PETER J. B.
A N D J A Y S . ROSENBLATT SLATER
Parental Behavior in Voles Behavioral Ecology and Conservation Biol-
ZUOXIN WANG A N D THOMAS R. ogy of Primates and Other Animals
INSEL KAREN B. STRIER
Physiological, Sensory, and Experiential
How to Avoid Seven Deadly Sins in the
Factors of Parental Care in Sheep
Study of Behavior
F. LEVY, K. M. KENDRICK, E. B.
MANFRED MILINSKI
KEVERNE, R. H. PORTER, A N D
A. ROMEYER Sexually Dimorphic Dispersal in Mammals:
Patterns, Causes, and Consequences
Socialization, Hormones, and the Regula-
L A U R A SMALE. SCOTT NUNES, A N D
tion of Maternal Behavior in Nonhuman
K A Y E. HOLEKAMP
Simian Primates
CHRISTOPHER R. PRYCE Infantile Amnesia: Using Animal Models
Field Studies of Parental Care in Birds: to Understand Forgetting
New Data Focus Questions on Variation H. M O O R E A R N O L D A N D NORMAN
among Females E. SPEAR
PATRICIA A D A I R GOWATY Regulation of Age Polyethism in Bees and
Parental Investment in Pinnipeds Wasps by Juvenile Hormone
FRITZ TRILLMICH SUSAN E. F A H R B A C H
Individual Differences in Maternal Style: Acoustic Signals and Speciation: The Roles
Causes and Consequences of Mothers and of Natural and Sexual Selection in the Evo-
Offspring lution of Cryptic Species
LYNN A. FAIRBANKS G A R E T H JONES
Mother-Infant Communication in Primates Understanding the Complex Song of the

D A R I O MAESTRIPIERI A N D European Starling: An Integrated Ethiologi-
JOSEP CALL cal Approach
M A R C E L EENS
Infant Care in Cooperatively Breeding
Species Representation of Quantities by Apes
CHARLES T. SNOWDEN S A R A H T. BOYSEN
ISBN 0-12-004527-3

(Advances in The Study of Behavior 27) Anders Pape Møller, Manfred Milinski and Peter J.B. Slater (Eds.) - Stress and Behavior-Elsevier, Academic Press (1998)

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

(Advances in The Study of Behavior 27) Anders Pape Møller, Manfred Milinski and Peter J.B. Slater (Eds.) - Stress and Behavior-Elsevier, Academic Press (1998)

Hochgeladen von

Copyright:

Verfügbare Formate

Advances in

THE STUDY OF BEHAVIOR

Acadeniic Press Limited

International Standard Book Number: 0-1 2-004527-3

PRINTED IN THE UNITED STATES OF AMERICA

The Concept of Stress and Its Relevance for Animal Behavior

Stress and Immune Response

Behavioral Variability and Limits to Evolutionary Adaptation

Developmental Instability as a General Measure of Stress

Stress and Decision-Making under the Risk of Predation:

Parasitic Stress and Self-Medication in Wild Animals

VII . Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

Stress and Human Behavior: Attractiveness. Women’s Sexual

Welfare. Stress. and the Evolution of Feelings

Biological Conservation and Stress

VIII . Minimizing Occurrence and Impact of Stress in

Contents of Previous Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . 549

VICTOR APANIUS (133),Department of Biological Sciences, Florida International

DONALD M. BROOM (371), Department of Clinical Veterinary Medicine, Univer-

MARION L. EAST (405), Max -Planck -1nstitut f i r Verhaltensphysiologie, 0-82319

BRYANT FURLOW (319). Department of Biology, University of New Mexico,

HERIBERT HOFER (405). Max -P/anck -1nstitut far Verhaltensphysiologie, D -

DIETRICH VON HOLST (1). Department of Animal Physiology, University of

STEVEN L. LIMA (215), Department of Life Sciences, Indiana State University,

G. A. LOZANO (291), Department o,f Biology, University of California, Riverside,

ANDERS PAPE M0LLER (181), Laboratoire d’Ecologie, UniversitP Pierre et

P. A. PARSONS (155), School of Genetics and Human Variation, La Trobe Univer-

RANDY THORNHILL (31Y), Department of Biology, University of New Mexico,

The aim of the Advances in the Study of Behavior series remains as it

All organisms suffer from a deficiency of one or more resources during

Laboratory studies are generally performed under benign conditions with

We believe that a concerted research effort concerning behavior of ani-

The Concept of Stress and Its Relevance for

1. Cannon’s Fight or Flight Syndrome

Increased blood flow to

“The Expression of the Emotions in Man and Animals,” in which he spoke

resistance is of particular importance. The adrenal cortex adapts to the

17-hydroxycorticosteroid plasma levels in rhesus monkeys in an emotionally

organism to many different stimuli or agents (stressors). Instead, he consid-

d. Predictability and Control. As mentioned above, the typical Selyean

Signal Signal + punishment

FIG.2. Length of gastric lesions (medians) of nonshock, avoidance-escape, and yoked

and symptoms indicative of depression. This passive stress response is

Threat to cont ss of control

Behavioral arousal with Inhibition of spatially organized

Defense reaction Defeat reaction

Control Striving Loss of control

c. Coping Behavior and Appraisal. These two styles of coping depend

700 1 I I Males raised

glands, examples are to be found in the modification of neuroendocrine

conspecifics trigger central nervous processes, which themselves induce

1. General Methodological Considerations

knowledge, which makes interpretation of apparently contradictory results

150 12 males per

seconds (catecholamines), minutes (glucocorticosteroids, thyroid and go-

Body temperature ("C) Heart rate (beatshin)

Triiodothyronine (nglrnl serum) Glucose (mg/100 ml blood)

Cortisol (ng/ml serum) Testosterone (ng/ml serum)

Elevated levels of glucocorticosteroid hormones

Mobilization of energy (Gluconeogenesis) Loss of muscle mass, fatigue, steroid

ticosteroids through other mechanisms (such as through direct action of

conclude that, whereas ACTH may be important in the regulation of aggres-

extent of adrenocortical activity by Selye and other scientists up to the end