Sie sind auf Seite 1von 14

Biodivers Conserv (2007) 16:27152728

DOI 10.1007/s10531-006-9100-3
ORIGINAL PAPER

Comprehensive criteria for biodiversity evaluation


in conservation planning
Helen M. Regan Frank W. Davis
Sandy J. Andelman Astrid Widyanata
Mariah Freese

Received: 23 September 2005 / Accepted: 5 July 2006 / Published online: 27 October 2006
Springer Science+Business Media B.V. 2006

Abstract In this paper we present the results of a multi-criteria decision analysis


used to identify a comprehensive set of criteria for assigning biodiversity value to
sites for conservation planning. For effective conservation management, biodiversity
value needs to be a composite of biotic and abiotic factors. However, in the reserve
design literature, conservation value is assigned with a limited set of metrics usually
based on comprehensiveness, representativeness and persistence which may be
insufficient at fully capturing biodiversity value. A group of conservation specialists
in California, USA, used a multi-criteria decision making framework to elucidate
and weight criteria for scoring biodiversity value at sites. A formal model for consensus and negotiation was applied to aggregate individuals criteria weights across
all group members. The group identified ecological condition, followed by biotic
composition as the most important contributors to site conservation value. Longand short-term threats causing fragmentation and degradation are also important
criteria to consider. Key criteria are identified for which further data collection
would serve the greatest purpose in prioritizing sites and the role of prioritization
criteria in the larger context of systematic conservation planning is discussed. With
the recognition that biodiversity value plays an important role in conservation

H. M. Regan (&) A. Widyanata M. Freese


Department of Biology, San Diego State University, 5500 Campanile Drive, San Diego,
CA 92182-4614, USA
e-mail: hregan@sciences.sdsu.edu
F. W. Davis
Donald Bren School of Environmental Science and Management, University of California
Santa Barbara, Santa Barbara, CA 93106-4060, USA
S. J. Andelman
Conservation International, 1919 M Street, NW Suite 600, Washington, DC 20036, USA
H. M. Regan S. J. Andelman
National Center for Ecological Analysis and Synthesis, University of California Santa Barbara,
735 State St, Suite 300, Santa Barbara, CA 93101, USA

123

2716

Biodivers Conserv (2007) 16:27152728

decisions, the criteria presented here represents a comprehensive suite of factors to


consider when assigning biodiversity value to sites for conservation planning. These
can serve as an encompassing list which other groups can customize for the purpose
of biodiversity evaluation for alternative conservation planning contexts.
Keywords Biodiversity value Conservation planning Group decision making
Multi-criteria decision making
Introduction
The goal of systematic conservation planning is to provide a structured, scientifically
defensible and effective framework for protecting biodiversity. The assignment of
biodiversity value to terrestrial sites is a key component of this process. However,
examples abound of areas designated for conservation because of low agricultural,
production or development value, rather than their ability to meet specific conservation goals (Pressey et al. 1996; Margules and Pressey 2000; Scott et al. 2001;
Mendel and Kirkpatrick 2002; Groves 2003). Areas with the greatest biodiversity
and the highest number of endangered and threatened species and ecosystems tend
to coincide with locations of high human land-use potential (Dobson et al. 1997;
Groves 2003). Moreover, the most imperiled species are largely unrepresented
within protected areas (Andelman and Willig 2003; Rodrigues et al. 2004).
Meir et al. (2004) show that reserve network design works best when the entire
network can be implemented immediately. However, in the context of biodiversity
loss, uncertainty, and dynamic site and budget availability, current reserve design
methods can lead to sub-optimal conservation planning decisions. They conclude
that prioritizing sites according to biodiversity value and relative importance may be
a better investment of conservation resources than a comprehensive reserve network
design. This raises the question of how to assign biodiversity value to sites. In a study
comparing the effectiveness of reserve design using different criteria, Bonn and
Gaston (2005) assert that priority areas for nature conservation should be protective
of not just genotypes and species, but also the non-living environment. They conclude that focusing on any single biodiversity component alone is insufficient to
protect other components and promote the inclusion of criteria related to ecosystem and environmental diversity for capturing biodiversity value (Bonn and Gaston
2005). The assignment of biodiversity value to sites for conservation planning,
however, is somewhat subjective and non-trivial. Currently in the reserve design
literature, conservation value is assigned with a limited set of metrics usually based
on comprehensiveness, representativeness and persistence (Margules and Pressey
2000; Cabeza and Moilanen 2001). Sites may be further prioritized based on economic factors and land availability (Costello and Polasky 2004; Haight et al. 2005).
However, conservation value for land-use planning purposes depends on a broad
range of features, not just the number, type and spatial distribution of species. Both
biotic and abiotic factors will be important in assessing the biodiversity value of a
site for conservation purposes (Wilson 1992). Furthermore, due to the complexities
in defining biodiversity value it is necessary to elicit the opinions of multiple experts.
MCDM is a longstanding technique that has been applied extensively to a wide
variety of problems including resource allocation (Pohekar and Ramachandran
2004), capital planning (Steuer and Na 2003), human resources (Caballero et al.

123

Biodivers Conserv (2007) 16:27152728

2717

2004), strategic planning (Kasanen et al. 2000), risk assessment (Haimes 2000), and
natural resource conflict resolution (Redpath et al. 2004). Applications do exist in
Conservation Area Networks (CANs) where multiple criteria are considered.
Moffett and Sarkar (2006) provide a review of criteria used in MCDM for the design
of CANs. Criteria considered in the design of CANs include: biodiversity surrogates
(e.g. indicator species), site size, shape and arrangement, connectivity, economic
cost, recreational value, human population effects, future economic value, scenic
beauty and cultural heritage (Moffett and Sarkar 2006). However, no one study
incorporates all of these criteria. To date, there has not been an attempt to compile
an exhaustive list of criteria, with relative weights of importance, for assigning
biodiversity value to sites for conservation planning.
The main purpose of this paper is to present a comprehensive set of criteria for
the purpose of assigning biodiversity value for conservation planning. This set of
criteria was derived through a formal multi-criteria decision making exercise applied
to a group of conservation planning experts. In conservation planning there is a
critical need for systematic approaches for ascribing biodiversity value to sites.
Currently the assignment of biodiversity value is limited to a few features and does
not routinely involve the opinions of a broad group of experts. We apply a recently
developed method, based on consensus convergence modeling, to aggregate values
(or, in this case, criterion weights) across group members (Regan et al. 2006). Unlike
many aggregation methods in the literature, this method relies on philosophical
foundations in consensus and negotiation to provide a formal and objective procedure to aggregate values across a group.

Methods
Identification of criteria
Criteria for biodiversity evaluation were set in the context of conservation planning
in California. California is a suitable case study for deriving general biodiversity
criteria due to its extreme geological and biological heterogeneityit is ecologically
diverse and contains several distinct bioregions. California supports a vast proportion of biodiversity in the US, with more native plant and animal species, and more
imperiled native species, than any other state (Stein et al. 2000). Furthermore,
California is experiencing unprecedented population increase. Population forecasters predict that Californias human population will reach 45 million by the year 2020,
an increase of about 8 million from year 2006 estimates, and could reach 60 million
by 2050 (The Resources Agency 2001). California has already suffered substantial
habitat loss. It has lost over 90% of its original wetlands acreage, and almost 20,000
acres of forest is fragmented each year as a result of urban development (The
Resources Agency 2001). A burgeoning human populations reliance on natural
resources ensures that habitat conversion, loss and/or degradation is unlikely to
subside significantly in the near future.
Eight scientists and consultants from The California Resources Agency (including
the California Department of Fish and Game and the California Coastal Conservancy) and the University of California Santa Barbara developed a set of criteria to
identify high priority lands that would be protective of terrestrial biodiversity. A
separate MCDM exercise was performed to select the scale of the planning units.

123

2718

Biodivers Conserv (2007) 16:27152728

The choice of planning units for conservation planning is critically important


because the measured conservation value of any location depends on the size, shape
and logic of the planning unit in which it is embedded (Austin and Margules 1986;
Stoms 1994). Hydrologically-defined planning units, namely Calwater Super Planning Watersheds (http://www.ca.nrcs.usda.gov/features/calwater/) and Major River
Basins, ranked the highest of the alternatives considered and criteria for biodiversity
value were identified with this spatial scale in mind. Throughout this paper site
refers to a project within a planning unit and only private lands are considered since
public lands already undergo some management or stewardship for biodiversity.
MCDM techniques are usually based on three main components: a decision goal,
a list of criteria that are considered to be desirable to meet that goal, and a list of
alternatives from which the best option is selected with respect to the decision goal.
The goal here was to identify lands for acquisition, stewardship or management that
are of statewide significance for conserving all terrestrial native species, communities
and habitats and the ecological processes that maintain them. In this paper we
restrict our attention to an important step towards that goalthe elicitation of
criteria to identify lands important for the protection of terrestrial biodiversity.
Three over-riding criteria emerged from a brainstorming session as the most
relevant for terrestrial biodiversity. These were: Current Biological Value; Fully
Restored Biological Value; and Threat. Current Biological Value pertains to the
current ecological state of candidate sites and how it contributes to overall terrestrial
biodiversity, whereas Fully Restored Biological Value refers to the state that degraded sites could achieve with respect to terrestrial biodiversity if they were restored. It should be noted that what constitutes a fully restored site is highly
contentious and in most cases full restoration to original condition is uncertain at
best (Wilkins et al. 2003). Nevertheless, the potential of restoration and the resulting
effects on terrestrial biodiversity were considered important in identifying relevant
sites. If land is acquired with the intention of full restoration then that site may
contribute significantly to the terrestrial biodiversity goal in the future. It was considered as a criterion in addition to Current Biological Value because the underlying
issues are essentially different. Threat refers to the immediate (within 510 years)
and long-term potential changes to a site from adverse impacts including stresses
from land use and environmental events (e.g. weather, wildfires, pollution, invasion
by exotic species, climate change) and how these impacts will affect terrestrial
biodiversity at that site. Sites that are threatened are to be given priority over nonthreatened sites for conservation management.
Considerations of the costs of acquisition, management and restoration are
important to the conservation value and feasibility of acquiring and managing lands.
It was decided, however, that they should not influence identification of sites with
terrestrial biodiversity value in the initial stages of the decision process. These factors should come into play in a major way in the allocation stages where the costs
and benefits would be weighed and optimized across all of the conservation objectives. The aim here is to explicitly identify and define criteria that can be used in
valuing sites for terrestrial biodiversity.
The ideas from the brainstorming session were converted into a decision tree,
where the criteria are considered in terms of their importance to the decision goal.
Criteria at Level 1 are the three main overriding criteria discussed above. These are
split into relevant sub-criteria at Level 2 which are split further into sub-criteria at

123

Biodivers Conserv (2007) 16:27152728

2719

higher levels and so on. Figures 13 show the sub-criteria for each of the three overriding criteria. Table 1 lists criteria definitions.
In the criteria in Figs. 13, the area or size of a site is not explicitly considered.
The issue of area, however, is implicit in many of the criteria. For instance, a site will
satisfy the Richness and Representation criteria only if there is sufficient area to
support these characteristics. A site will Contribute to watershed values only if it is
large enough to do so. In this way many of the criteria rely on a site being of a
reasonable size to achieve those aspects related to the terrestrial biodiversity goal.
Assignment of weights to criteria
The Analytic Hierarchy Process (AHP), a special case of MCDM developed by
Saaty (1980), was the forum in which weights were assigned to criteria. AHP uses
numerical pair-wise comparison of the relative importance of one criterion over
another. The software package Criterium DecisionPlus 3.0 was used to facilitate
the assignment of pair-wise preference values at each level in the decision tree. Each
group member specified scores to indicate the relative importance of each criterion
with respect to all other co-criteria. Here, we define co-criteria as those criteria that
branch from the same lower level criterion. For instance, in Fig. 1 Low Fragmentation and Low Degradation are co-criteria because they both branch from the Level
2 criterion Ecological Condition. In cases where there was wide discrepancy between
the lowest and highest preference values for the pair of criteria, the proponents
discussed the basis for their differences. In some cases the preference scores were
amended based on the discussion. In these cases, the discrepancy among original
AHP scores was usually the result of different criterion definitions used across the

Communities
0.30

Richness 0.19
Rarity 0.50
Representation 0.31
Trust species
0.46

Biotic Composition
0.33

Species
0.31

Vertebrates
0.18
Invertebrates
0.18
Plant species
0.18

Ecological context
0.19

Ecological condition
0.48

Representation of biophysical
environments
Site connects important areas
Site buffers important areas
Site contributes to watershed
values
Site contributes to network of
sites

Low fragmentation
0.45

Low degradation
0.55

Rare Threatened and Endangered species


Game species 0.11
Migratory species 0.21
CEQA species 0.27
Representation 0.41
Richness 0.21
Rarity 0.38
Representation 0.36
Richness 0.22
Rarity 0.41
Representation 0.36
Richness 0.22
Rarity 0.42

0.41

0.39
0.30
0.17
0.21
0.32
Low % agriculture and
forestry
Low % urban and residential
Low % roaded
Vegetation structure
Not invaded by exotics
High soil quality
High air quality
High water quality
Maintenance of natural
disturbance regimes

0.24
0.52
0.24
0.26
0.18
0.17
0.09
0.08
0.22

Fig. 1 List of sub-criteria and their weights for the over-riding criterion Current Biological Value.
Current Biological Value refers to the current state of potential sites and how it contributes to the
overall terrestrial biodiversity. Here, site refers to a project within a planning unit. Current
Biological Value received a weight of 0.63

123

2720

Biodivers Conserv (2007) 16:27152728

Land conversion 0.83


Immediate (5 10 years)
0.81

Land degradation
0.17

Low soil quality 0.47


Low water quality 0.17
Low air quality 0.15
Invasion by exotics 0.21

Land conversion 0.79

Long term (> 10 years)


0.19

Land degradation
0.21

Low soil quality 0.24


Low water quality 0.09
Low air quality 0.10
Invasion by exotics 0.16
Loss of vegetation structure 0.15
Altered natural disturbance regimes

0.26

Fig. 2 List of sub-criteria and their weights for the over-riding criterion Threats. Threat refers to the
threatening processes that could adversely impact terrestrial biodiversity at a site; acquisition and
protection of a site that is subject to threatening processes or is impacted by them to a higher degree,
is considered to be more important to meeting the terrestrial biodiversity goal than for sites that have
no or few associated threats with low impacts. Here, site refers to a project within a planning unit.
Threats received a weight of 0.27

Communities
0.30
Biotic composition
0.74

Species 0.31

Ecological context
0.26

Representation of biophysical
environments 0.39
Site connects important areas 0.80
Site buffers important areas 0.20

Richness 0.19
Rarity 0.50
Representation 0.31
Richness 0.19
Rarity 0.50
Representation 0.31

Fig. 3 List of sub-criteria and their weights for the over-riding criterion Fully Restored Biological
Value. Here, site refers to a project within a planning unit. Fully Restored Biological Value
received a weight of 0.10

group. The process of revisiting group members scores to extract the underlying
rationale ensured that consensus was reached on the criteria definitions so that all
group members were comparing essentially the same criteria. It also provided an
opportunity for participants to explain the rationale behind their preferences.
In the AHP, preference scores are converted to weights by calculating the
eigenvalues of the pair-wise comparison matrices. The weights are then the entries of
the normalized eigenvector associated with the dominant eigenvalue. Higher weights
indicate a greater relative importance of the criterion in comparison with its cocriteria. Each group members AHP scores were converted to a set of criterion
weights, resulting in eight sets of weights, one set for each group member.
In most group decision making contexts, it is usually necessary to reach consensus
on group values (or in this case weights). Here we use consensus convergence
modeling to achieve consensual criterion weights that are representative of the
group (Lehrer and Wagner 1981; Regan et al. 2006). Consensus convergence models
are applicable when all parties agree to negotiate in order to resolve conflict. An
iterative mathematical model for achieving consensus through negotiation has been
developed in Lehrer and Wagner (1981). It is assumed at the outset that the group
has engaged in extended discussion of the issues so that all empirical data and
theoretical ratiocination has been communicated (Lehrer and Wagner 1981). As a
result of these discussions, the method assumes that a group of agents with varying
expertise start off with preferences (weights for criteria in a decision tree, for
example). The first step in the negotiation model involves each agent assigning a
numerical degree of respect for the other agents positions. The second step

123

Biodivers Conserv (2007) 16:27152728

2721

Table 1 Definitions of criteria used in the multi-criteria decision trees for identifying areas
important to terrestrial biodiversity. Here, site refers to a project within a planning unit
Biotic composition: the ecological assets of a site that contribute favorably to biodiversity, including
species, communities, and biophysical environments; components that contribute to biodiversity to a
greater degree than others are considered to be more important in achieving the overall goal
Ecological Context: the value of a site for providing connectivity or for buffering adjacent lands is
considered especially important for meeting biological conservation goals. There is very little legislative authority for acquisition of such lands. The Federal Land and Waters Conservation Act
includes purchase of inholdings in national parks. The migratory waterfowl act includes migration
routes
Ecological Condition: the physical condition of a site in terms of disturbances such as fragmentation
and degradation; the higher the ecological condition at a site (or less fragmentation and degradation
at a site), the more important the site is to terrestrial biodiversity
Site connects important areas: sites that might not be important in and of themselves to terrestrial
biodiversity, but improve the conservation value of the area by providing connections (e.g. corridors)
between sites with high terrestrial biodiversity
Site buffers important areas: sites that might not be important to terrestrial biodiversity in and of
themselves, but improve the conservation value of the area they surround by providing a buffer
against adverse influences
Site contributes to a network of sites: sites that might not be important to terrestrial biodiversity in and
of themselves, but their acquisition improves the conservation value of the area because they contribute to a network of sites in the area
Land conversion: includes the threat of conversion to production lands, urban development, or for
any other purpose that would be detrimental to terrestrial biodiversity within the relevant time frame
Richness: the number of communities or species (whichever is relevant) within a site; the greater the
richness at a site the more important it is to the terrestrial biodiversity goal
Rarity: refers to how rare a community or species (whichever is relevant) is within California; the
rarer the community or species at a site, or the more rare communities or species within a site, the
more important the site is to terrestrial biodiversity
Representation: refers to goals for community or species representation (whichever is relevant), both
amount and/or number of examples; the more under-represented the community or species is within
California, the higher the complementarity; this could be measured as irreplaceability
Low % agriculture and forestry: proportion of the site that is used for agriculture and timber harvesting; the lower the percentage of land used for production purposes, the more important the site is
for terrestrial biodiversity
Low % urban/residential: proportion of the site that has been converted to urban and residential
areas; the lower the percentage of urban and residential development, the more important the site is
to terrestrial biodiversity
Low % roaded: proportion of the site that has been converted to roads; the lower the percentage of
road conversion, the more important the site is to terrestrial biodiversity
Vegetation structure: amount of disturbance to the vegetation structure at a site; sites that have less
disturbance to vegetation structure are more important to terrestrial biodiversity than those with
significant disturbance
High soil quality: soil that has experienced low contamination, low disturbance and has intact soil
profiles
High air quality: ambient air that has low contamination and for which there are low impacts from
ozone
High water quality: water sources that have low contamination and have natural sediment loads
RTE species: rare, threatened and endangered species considered under the federal and California
Endangered Species Act, and Marine Mammal Protection Act; sites with RTE species are considered
to be of greater biodiversity value than those without
Migratory species: sites that are used by migratory species as stopovers are considered more
important in meeting the terrestrial biodiversity goal than those that are not

123

2722

Biodivers Conserv (2007) 16:27152728

Table 1 continued
CEQA species: site contains species considered under the California Environmental Quality Act
ThreatsLow soil quality: refers to sites that are subject to (short- or long-term) threats that will
result in poor soil quality (evidenced by contamination, disturbance and disruption to soil profiles)
ThreatsLow water quality: refers to sites that are subject to (short- or long-term) threats that will
result in poor water quality (evidenced by contamination and disruption to natural sediment loads)
ThreatsLow air quality: refers to sites that are subject to (short- or long-term) threats resulting in
poor air quality (evidenced by contamination of the ambient air and impacts from ozone)
ThreatsInvasion by exotics: refers to sites that are subject to the threat of invasion by exotic species
in the short- or long-term

aggregates and modifies agents criterion weights based on the degree of respect and
their initial preferences. The process is repeated until consensus is reached. Lehrer
and Wagner (1981) have shown that this model guarantees consensus under fairly
loose constraints. The advantage of this method is that it does not require that all
members of the group reach agreement, often an impossible task in group decision
making. Rather, it aggregates group members values in such a way that convergence
towards a single consensual value is obtained that summarizes the group position.
Regan et al. (2006) constructed a metric to calculate degrees of respect based on the
strength of the difference between the weights individuals place on criteria and not
on an individuals personal opinion of group members. It has the advantage of being
a somewhat more objective assessment of differences of opinion within the group,
than directly assigning a degree of respect to an individual. The reader is referred to
Regan et al. (2006) for full details of this model and a discussion of its implications in
group decisions. The consensus convergence model was applied across the eight sets
of criterion weights to calculate a consensual weight for each criterion in the decision
tree. Figures 13 display the consensual criterion weights derived from this process.

Results
The weights for Level 1 criteria indicate that the Current Biological Value (with a
weight of 0.63) of sites is regarded as the most important criterion for meeting the
biodiversity goal. Next in importance is Threats (weight of 0.27) followed by Fully
Restored Biological Value (weight of 0.10). The very low relative weight assigned to
this last criterion reflects the view that there is uncertainty that full restoration could
ever be achieved. Below we summarize the weights and the rationale for the subcriteria of the overriding criteron, Current Biological Value. Parts of this summary
are relevant for the sub-criteria and weights for Fully Restored Biological Value. The
sub-criteria and their weights for Threats are self-explanatory from Fig. 2 and
Table 1.
Of the Level 2 sub-criteria of Current Biological Value, Ecological Condition was
given the greatest importance weighting followed by Biotic Composition and Ecological Context. This preference ordering reflected the view that it was more
important to protect relatively pristine and intact sites as protection of biodiversity
would be implicit. This has been the rationale behind the development and use of

123

Biodivers Conserv (2007) 16:27152728

2723

indices for biotic integrity for aquatic conservation purposes (Moyle and Randall
1998; Griffith 2003).
Of the sub-criteria to Ecological condition, sites with low degradation were
considered to be slightly more important than sites with low fragmentation. The
rationale here was that most sites on private lands experience the effects of fragmentation and that low degradation may be a better distinguishing feature. Within
the fragmentation sub-criteria, the protection of sites that experienced lower urban
and residential fragmentation was considered to be more important than protecting
sites experiencing a similar degree of fragmentation from agricultural and forestry
practices or from roads. Fragmentation due to urban and residential growth carries
with it a myriad of other disturbances due to close proximity to human populations
(Roman-Cuesta et al. 2003; Stenhouse 2004), whereas agricultural and forestry lands
are not necessarily adjacent to large urban centers. Fragmentation due to roads is
also considered to reduce the biodiversity value of a site (Stoms 2000; Wilkie et al.
2000) but less so than fragmentation due to urban development.
For the Degradation sub-criteria, intact vegetation structure was the most
important feature when assessing sites. Long-term changes in vegetation structure
have been shown to reduce biodiversity (Sekercioglu 2002). This criterion was followed by maintenance of natural disturbance regimes, high soil quality, high water
quality, an absence of invasive species and high air quality. There is some overlap of
criteria herean intact vegetation structure relies on all of the other sub-criteria to
some extent. For instance, changes in vegetation structure will be correlated with the
presence of invasive species (Fleishman et al. 2003; Rossiter et al. 2003). However,
all else being equal, a site with high soil quality, for example, should be ranked above
a site with low soil quality since soil composition, in its own right, can affect the
biodiversity at a site (Gardi et al. 2002; Hossack et al. 2004).
Of the sub-criteria to Biotic Composition, it is almost equally important to protect
communities and species. Sites with diverse biophysical elements were regarded as
slightly more important. In fact, all group members weighted communities, species
and biophysical elements equally except for two members who believed biophysical
elements were substantially more important to consider when evaluating biodiversity than communities and species. Sites with rare communities were regarded as the
most important community attribute for the conservation of biodiversity followed by
sites with high community representation and richness. It is unlikely that many sites
within a planning unit would contain a richness of communities, i.e. many diverse
communities. Hence, when considering communities, it is more important to ensure
that rare communities are given priority in a conservation plan followed by communities that are underrepresented.
For species, sites with trust species were ranked the highest. Trust species include
rare, threatened and endangered (RTE) species, game species, migratory species
and species listed under the California Environmental Quality Act (CEQA). Not
surprisingly RTE species ranked the highest of these, whereas game species received
a very low weight. Representation, rarity and richness were ranked similarly across
vertebrate, invertebrate and plant species. For invertebrate and plant species, representation and rarity were given equal weights, whereas for vertebrate species
representation was weighted slightly higher than rarity. This probably reflects the
groups greater familiarity with vertebrate species rather than any serious difference
in the guilds. For all species, species richness within a site was ranked the lowestit
is more important for a conservation plan to capture rare species or species that are

123

2724

Biodivers Conserv (2007) 16:27152728

not represented broadly across the region than it is to capture common species in
multiple sites.
Ecological Context refers to the value of a site for providing connectivity or for
buffering adjacent lands. In California, there is little legislative authority for
acquisition of such lands: the Federal Land and Waters Conservation Act includes
purchase of inholdings in national parks (http://www.thecre.com/fedlaw/legal3/lwcfact.htm); the Migratory Bird Conservation Act includes migration routes (http://
www.ipl.unm.edu/cwl/fedbook/mbca.html). This criterion was given the lowest
weight of the Level 2 sub-criteria but is nevertheless important to consider because it
supports the creation and maintenance of an effective reserve system. The most
important sub-criterion of Ecological Context is the degree to which a site contributes to a network of protected sites. Sites that add biodiversity value to a currently
existing reserve network will be ranked higher under this criterion than sites that do
not. Connectivity is an important criterion since contiguous reserves have a higher
chance of being effective, especially for species that disperse or migrate (Moilanen
and Nieminen 2002). Sites that contribute to watershed values are also important for
conservation. Watershed issues are often publicly volatile and many land management decisions are driven by water considerations. Protection of watersheds often
goes hand in hand with protecting biodiversity.

Conclusions and discussion


The MCDM exercise presented here provides a relatively comprehensive list of
features important for establishing terrestrial biodiversity value at sites. The set of
criteria outlined here represents the values and judgment of decision makers in
California. Even though the multi-criteria decision trees presented here were originally intended for the purpose of systematic conservation planning in California, the
resulting set of criteria is portable and, with the exception of two criteria (CEQA
and Game species in Figs. 1, 3), could be applied almost anywherecriteria were
identified to address terrestrial biodiversity value in a general context and were
justified using general principles of conservation biology. Hence, the multi-criteria
decision tree produced here can serve as a template, to be modified and customized,
for the identification of important landscapes for conservation management for regions beyond California and the US.
We acknowledge that currently there are insufficient data to evaluate sites across
the full suite of criteria. The decision tree was not constructed to coincide with data
availability, although in California much of the information is increasingly available
at county and regional scales to support ongoing habitat conservation planning efforts (Smallwood et al. 1998; see also the new online digital conservation atlas and
associated links at http://www.legacy.ca.gov/new_atlas.epl). The original intent of
this decision tree was to articulate explicitly and scientifically what it means for a site
to have high terrestrial biodiversity value given the current state of knowledge.
While we believe the criteria and the preference ordering presented in this paper
is comprehensive and scientifically defensible, another groups weights might lead to
a different ordering of criteria. This will particularly be the case for application in
different regions. Identification of criteria and weight assignments is somewhat
subjective, value driven and dependent on the experience and background of the
group. Selection of an appropriate and representative group is one of the most

123

Biodivers Conserv (2007) 16:27152728

2725

important tasks for effective decision-making. If the group does not consist of
credible members, or omits relevant backgrounds and expertise in the process, then
the final decision will not have the support or credibility necessary for implementation. This decision tree can serve as a template for modification and customization
by alternative groups. Also, weights can be assigned directly using elicitation
methods rather than the AHP, or by other criteria weighting schemes (Moffett and
Sarkar 2006; Figueira et al. 2004). The AHP was used here for convenience in a
group setting but other methods apply equally well.
We used a novel consensus convergence modeling method to aggregate weights
across group members. It could be argued though that there is merit to explicitly
representing the diversity of opinions of all group members rather than requiring
consensual weights (although see Regan et al. 2006 for the advantages of formal
consensus models when group members agree to negotiate to reach consensus). Such
diversity highlights differences of opinion in a group and can lead to creative and
more equitable problem solving (Peterson et al. 2005). In general however, prioritization processes such as these require single numbers in order to proceed. The
most important factor is that participants agree to its outcome and believe their
views were incorporated into the process and this was the case here.
The aim of this exercise was to compile an explicit and comprehensive list of
criteria pertaining to site-specific terrestrial biodiversity value that could be modified
or adapted and critiqued for feasibility across a larger group of stakeholders. Ultimately though, a decision needs to be made, i.e. different alternatives need to be
assessed against the final set of criteria and ranked according to which satisfies the
most (important) criteria to the highest degree. In this case the alternatives would be
potential projects within planning units. There are a number of ways of assessing
alternatives against criteria. Once the available data have been ascertained, monotonically increasing metrics could be devised for each criterion that give numerical
scores denoting how well features of the site satisfy the criteria. For instance, RTE
species can be represented as a function of the number of such species found within
the site. Representation can be measured in the same way irreplaceability is measured for reserve design algorithms. Low % fragmented might be measured as the
inverse of edge-to-area ratio of remnant fragments. Some criteria will simply require
a yes or no answer converted to a 0 or 1 score (or on some other appropriate scale).
Other criteria are less amenable to measurement and these might be omitted altogether or substituted for surrogates that can be measured.
Multi-criteria evaluation is primarily concerned with how to combine the information from several criteria to form a single index of evaluation. Once the alternatives have been assessed against criteria, the usual procedure is to calculate a
weighted sum of attributes, also known as a weighted linear combination (although
see below for a discussion on alternative procedures for calculating decision scores).
The decision score could then be used in a variety of ways to assist in conservation
planning decisions. Sites may be ranked for conservation priority in descending
order of their final biodiversity score (as suggested in Meir et al. 2004) or biodiversity value of sites can be incorporated into reserve design allocation algorithms
that include multiple values and objectives (e.g. McDonnell et al. 2002).
The calculation of the weighted sum of attributes relies on an assumption of
independence between criteria. For many of the criteria identified in Figs. 13 this
assumption is not met. The consequence of dependencies between criteria in the
decision tree is a double (or multiple) counting of attributes in the weighted sum.

123

2726

Biodivers Conserv (2007) 16:27152728

The result is that some site attributes will effectively receive a higher overall weight
than the weights presented at each node in Figs. 13. For instance, in Fig. 1 the
criteria Not invaded by exotics and Maintenance of natural disturbance regimes appear as sub-criteria of the over-riding criterion Ecological Condition. However, in
some cases exotic species disrupt natural disturbance regimes such as fire
(DAntonio and Vitousek 1992) and create a dependency between the existence of
exotic species and the disruption of natural disturbances. Hence, the existence of
exotic species could count twice against a site, i.e. the site would receive low scores
when assessed against both criteria for essentially the same reason. Group members
were aware of the dependency issue when identifying criteria and combined
dependent criteria into an appropriate single criterion when it made sense to do so.
Although the MCDM methods usually calculate a decision score for alternatives
using a weighted linear combination (Voogd 1983), there are many decision rules
that could be used in MCDM processes. For instance, Boolean constraints could be
imposed where decision scores are calculated as the union (logical OR) or intersection (logical AND) of criteria. In this way, depending on the logical structure of
the decision rule, a site could be ranked highly if only a subset of criteria are met. We
do not promote any particular decision rule here (but we refer the reader to Figueira
et al. 2004 for a full survey of methods). The most appropriate decision rule will
depend on the dependencies between criteria and the features of biodiversity that
are important to emphasize for the case at hand.
In conclusion we found this process of group elicitation for the purpose of
addressing conservation planning for terrestrial biodiversity to be highly effective.
Group members, when interviewed afterwards, responded enthusiastically to the
process even though they had initially approached it with skepticism. The overwhelming feeling was that the process was extremely inclusive of all group members
views, it enabled efficient elicitation of views and values (the workshop was held
over 2 days), and resulted in a consensually derived and comprehensive decision tree
for biodiversity evaluation in which the important over-riding criteria were highlighted. Furthermore, group members could retrospectively determine the logic and
reasoning behind the criteria ranks because a formal decision making framework
was used. This goes a long way to making the process of biodiversity evaluation
repeatable, transparent, and inclusive of multiple experts.
Acknowledgments We wish to thank Marc Beyeler (California Coastal Conservancy), Pete Dangermond (Dangermond Associates), Greg Greenwood (The Resources Agency of California), Diana
Hickson (California Department of Fish and Game), and Mark Hoshovsky (California Department
of Fish and Game) for participating in the workshops and offering their knowledge and experience
on conservation management. We also wish to thank two anonymous reviewers for their helpful
comments. AW and MF were funded by an SDSU RSCA grant to HMR. Parts of this work were
conducted at the National Center for Ecological Analysis and Synthesis, a Center funded by NSF
(Grant #DEB-0072909), the University of California, and the Santa Barbara campus.

References
Andelman SJ, Willig MR (2003) Present patterns and future prospects for biodiversity in the
Western Hemisphere. Ecol Lett 6(9):818824
Austin MP, Margules CR (1986) Assessing representativeness. In: Usher MB (ed) Wildlife conservation evaluation. Chapman and Hall, London, pp 4567

123

Biodivers Conserv (2007) 16:27152728

2727

Bonn A, Gaston KJ (2005) Capturing biodiversity: selecting high priority areas for conservation
using different criteria. Biodivers Conserv 14:10831100
Caballero R, Galache T, Gomez T, Molina J, Torrico A (2004) Budgetary allocations and efficiency
in the human resources policy of a university following multiple criteria. Econ Educ Rev
23(1):6774
Cabeza M, Moilanen A (2001) Design of reserve networks and the persistence of biodiversity.
Trends Ecol Evol 16(5):242247
Costello C, Polasky S (2004) Dynamic reserve site selection. Resour Energy Econ 26(2):157174
DAntonio CM, Vitousek PM (1992) Biological invasions by exotic grasses, the grass/fire cycle, and
global change. Annu Rev Ecol Syst 23:6387
Dobson AP, Rodriguez JP, Roberts WM, Wilcove DS (1997) Geographic distribution of endangered
species in the United States. Science 275:550553
Figueira J, Greco S, Ehrgott M (eds) (2004) Multiple criteria decision analysis: state of the art
surveys. International Series in Operations Research & Management Science, vol. 78. Springer,
Berlin Germany
Fleishman E, Mcdonal N, Mac Nally R, Murphy DD, Walters J, Floyd T (2003) Effects of floristics,
physiognomy and non-native vegetation on riparian bird communities in a Mojave Desert watershed. J Anim Ecol 72(3):484490
Gardi C, Tomaselli M, Parisi V, Petraglia A, Santini C (2002) Soil quality indicators and biodiversity
in northern Italian permanent grasslands. Eur J Soil Biol 38(1):103110
Griffith JA (2003) Regions of high fish diversity for conservation concern within the US Central
Plains. J Freshwat Ecol 18(3):451457
Groves CR (2003) Drafting a conservation blueprint: a practitioners guide to planning for biodiversity. Island Press, Washington D.C
Haight RG, Snyder SA, Revelle CS (2005) Metropolitan open-space protection with uncertain site
availability. Conserv Biol 19(2):327337
Haimes YY (2000) MCDM and its worldwide role in risk-based decision making: opening remarks to
the XIVth International Conference on MCDM. Res Pract Mult Crit Decis Mak: Lect Notes
Econ Math Syst 487:IXXI
Hossack I, Robertson D, Tucker P, Hursthouse A, Fyfe C (2004) A GIS and web-based decision
support tool for the management of urban soils. Cybern Syst 35(56):499509
Kasanen E, Wallenius H, Wallenius J, Zionts S (2000) A study of high-level managerial decision
processes, with implications for MCDM research. Eur J Oper Res 120(3):496510
Lehrer K, Wagner C (1981) Rational consensus in science and society. Philosophical studies series in
philosophy 24. Reidel Publishing Company, Dordrecht, Holland
Margules CR, Pressey RL (2000) Systematic conservation planning. Nature 405:243253
McDonnell MD, Possingham HP, Ball IR, Cousins EA (2002) Mathematical methods for spatially
cohesive reserve design. Environ Model Assess 7(2):107114
Meir E, Andelman S, Possingham HP (2004) Does conservation planning matter in a dynamic and
uncertain world? Ecol Lett 7:615622
Mendel LC, Kirkpatrick JB (2002) Historical progress of biodiversity conservation in the protected
area system of Tasmania, Australia. Conserv Biol 16:15201529
Moffett A, Sarkar S (2006) Incorporating multiple criteria into the design of conservation area
networks: a minireview with recommendations. Divers Distrib 12(2):125137
Moilanen A, Nieminen M (2002) Simple connectivity measures in spatial ecology. Ecology
83(4):11311145
Moyle PB, Randall RJ (1998) Evaluating the biotic integrity of watersheds in the Sierra Nevada,
California. Conserv Biol 12(6):13181326
Peterson MN, Peterson MJ, Peterson TR (2005) Conservation and the myth of consensus. Conserv
Biol 19(3):762767
Pohekar SD, Ramachandran M (2004) Application of multi-criteria decision making to sustainable
energy planning A review. Renew Sust Energy Rev 8(4):365381
Pressey RL, Ferrier S, Hager TC, Woods CA, Tully SL, Weinman KM (1996) How well protected
are the forests of north-eastern New South Wales analyses of forest environments in relation to
formal protection measures, land tenure and vulnerability to clearing. For Ecol Manage 85:311
333
Redpath SA, Arroyo BE, Leckie EM, Bacon P, Bayfield N, Gutierrez RJ, Thirgood SJ (2004) Using
decision modeling with stakeholders to reduce human-wildlife conflict: a Raptor-Grouse case
study. Conserv Biol 18(2):350359

123

2728

Biodivers Conserv (2007) 16:27152728

Regan HM, Colyvan M, Markovchick-Nicholls L (2006) A formal model for consensus and negotiation in environmental management. J Environ Manage 80:167176
The Resources Agency (2001) First draft report on the methodology to identify state conservation
priorities, April (2001) Available from http://www.legacy.ca.gov/pub_docs/CCRISP_Methodology.pdf (accessed May 2006)
Rodrigues ASL, Andelman SJ, Bakarr MI, Boitani L, Brooks TM, Cowling RM, Fishpool LDC, da
Fonseca GAB, Gaston KJ, Hoffmann M, Long JS, Marquet PA, Pilgrim JD, Pressey RL,
Schipper J, Sechrest W, Stuart SN, Underhill LG, Waller RW, Watts MEJ, Yan X (2004)
Effectiveness of the global protected area network in representing species diversity. Nature
428(6983):640643
Roman-Cuesta RM, Gracia M, Retana J (2003) Environmental and human factors influencing fire
trends in enso and non-enso years in tropical Mexico. Ecol Appl 13(4):11771192
Rossiter NA, Setterfield SA, Douglas MM, Hutley LB (2003) Testing the grass-fire cycle: alien grass
invasion in the tropical savannas of northern Australia. Divers Distrib 9(3):169176
Saaty TL (1980) The analytic hierarchy process: planning, setting priorities, resource allocation.
McGraw-Hill International Book Company, London
Scott JM, Davis FW, McGhie G, Groves C (2001) Nature reserves: do they capture the full range of
Americas biodiversity? Ecol Appl 11:9991007
Sekercioglu CH (2002) Effects of forestry practices on vegetation structure and bird community of
Kibale National Park, Uganda. Biol Conserv 107(2):229240
Smallwood KS, Wilcox B, Leidy R, Yarris K (1998) Indicators assessment for habitat conservation
plan of Yolo County, California, USA. J Environ Manage 22:947958
Stein BA, Kutner LS, Adams JS (2000) Precious heritage: the status of biodiversity in the United
States. Oxford University Press, Oxford
Stenhouse RN (2004) Fragmentation and internal disturbance of native vegetation reserves in the
Perth metropolitan area, Western Australia. Landsc Urban Plan 68(4):389401
Steuer RE, Na P (2003) Multiple criteria decision making combined with finance: a categorized
bibliographic study. Eur J Oper Res 150(3):496515
Stoms DM (1994) Scale dependence of species richness maps. Prof Geogr 46(3):346358
Stoms DM (2000) GAP management status and regional indicators of threats to biodiversity. Landsc
Ecol 15(1):2133
Voogd H (1983) Multicriteria evaluation for urban and regional planning. Pion, London
Wilkie D, Shaw E, Rotberg F, Morelli G, Auzel P (2000) Roads, development, and conservation in
the Congo basin. Conserv Biol 14(6):16141622
Wilkins S, Keith DA, Adam P (2003) Measuring success: evaluating the restoration of a grassy
eucalypt woodland on the Cumberland Plain, Sydney, Australia. Restor Ecol 11(4):489503
Wilson EO (1992) The diversity of life. Belknap, ambridge Massachusetts

123

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and rea

Das könnte Ihnen auch gefallen