Quaternary Science Reviews 112 (2015) 17e32

Contents lists available at ScienceDirect

Quaternary Science Reviews

journal homepage: www.elsevier.com/locate/quascirev

Invited review

The role of palaeoecological records in assessing ecosystem services

Elizabeth S. Jeffers a, *, Sandra Nogue

a, b, Katherine J. Willis a, b, c
a
Long Term Ecology Laboratory, Biodiversity Institute, Oxford Martin School, Department of Zoology, University of Oxford, Oxford OX1 3PS, UK
b
Department of Biology, University of Bergen, All

egaten 41, N-5007 Bergen, Norway
c
Royal Botanic Gardens, Kew, Richmon, TW9 3AE, UK

a r t i c l e i n f o a b s t r a c t

Article history: Biological conservation and environmental management are increasingly focussing on the preservation
Received 10 July 2013 and restoration of ecosystem services (i.e. the benefits that humans receive from the natural functioning
Received in revised form of healthy ecosystems). Over the past decade there has been a rapid increase in the number of palae-
12 December 2014
oecological studies that have contributed to conservation of biodiversity and management of ecosystem
Accepted 19 December 2014
Available online
processes; however, there are relatively few instances in which attempts have been made to estimate the
continuity of ecosystem goods and services over time. How resistant is an ecosystem service to envi-
ronmental perturbations? And, if damaged, how long it does it take an ecosystem service to recover?
Keywords:
Ecosystem services
Both questions are highly relevant to conservation and management of landscapes that are important for
Palaeoecology ecosystem service provision and require an in-depth understanding of the way ecosystems function in
Long-term ecological records space and time. An understanding of time is particularly relevant for those ecosystem services e be they
Final ecosystem services supporting, provisioning, regulating or cultural services that involve processes that vary over a decadal
Ecosystem goods (or longer) timeframe. Most trees, for example, have generation times >50 years. Understanding the
Proxies response of forested ecosystems to environmental perturbations and therefore the continuity of the
Dendroecology ecosystem services they provide for human well-being e be it for example, carbon draw-down (regu-
lating service) or timber (provisioning service) e requires datasets that reflect the typical replacement
rates in these systems and the lifecycle of processes that alter their trajectories of change. Therefore, data
are required that span decadal to millennial time-scales. Very rarely, however, is this information
available from neo-ecological datasets and in many ecosystem service assessments, this lack of a tem-
poral record is acknowledged as a significant information gap.
This review aims to address this knowledge gap by examining the type and nature of palaeoecological
datasets that might be critical to assessing the persistence of ecosystem services across a variety of time
scales. Specifically we examine the types of palaeoecological records that can inform on the dynamics of
ecosystem processes and services over time e and their response to complex environmental changes.
We focus on three key areas: a) exploring the suitability of palaeoecological records for examining
variability in space and time of ecosystem processes; b) using palaeoecological data to determine the
resilience and persistence of ecosystem services and goods over time in response to drivers of change;
and c) how best to translate raw palaeoecological data into the relevant currencies required for
ecosystem service assessments.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction
There is growing appreciation for the many goods and services
provided to people by well-functioning ecosystems, including food,
fuel, climate regulation and spiritual values. These goods and
* Corresponding author. Tel.: þ44 1865 281 851.
E-mail address: elizabeth.jeffers@zoo.ox.ac.uk (E.S. Jeffers).
services are derived from ecosystem processes i.e. the physical,
chemical and biological interactions between organisms and their
environment (Fig. 1a). The Millennium Ecosystem Assessment (MA)
evaluated the current state of ecosystem service provision world-
wide and found that the majority of ecosystems are becoming
increasingly degraded, which threatens the long-term supply of
ecosystem service delivery (MA, 2005). Preserving (Chan et al.,
2006) and restoring (Palmer and Filoso, 2009) ecosystems for the
services they provide to people requires informed land and natural

http://dx.doi.org/10.1016/j.quascirev.2014.12.018
0277-3791/© 2014 Elsevier Ltd. All rights reserved.
18 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32

Fig. 1. Schematic of the ecosystem services framework used by the UK National Ecosystem Assessment to demonstrate the links between ecosystem processes, final ecosystem
services and goods (a). Copyright 2011 UK National Ecosystem Assessment. Environmental change (e.g. climate warming) will alter the ability of ecosystems to provide goods and
services. Predicting the impacts of environmental change on human well-being requires knowledge of how ecosystem processes respond to direct drivers of change and how these
changes will cascade through each step in the production of ecosystem services and goods (b).

resource decision making, particularly in landscapes beyond pro-
tected areas, where human activities are reducing biodiversity and
impeding natural ecosystem processes at increasing rates
(Balmford and Bond, 2005).
Managing ecosystems for the continued supply of goods and
services they provide for human well-being depends upon the
availability of information about their variation across space (de
Groot et al., 2010) and over time as systems respond to on-going
environmental change and short-term environmental perturba-
tions (Carpenter et al., 2009). Ecosystem management efforts are
often being aimed at a moving target (Dawson et al., 2011) and
therefore require an understanding of how ecosystem components
and processes respond to direct and indirect drivers of change (Diaz
et al., 2007; Tylianakis et al., 2008) and of the cascading effects
these changes have on the supply of ecosystem services and goods
to humans (Fig. 1b).
Since the publication of the MA, the field of ecosystem service
research has made significant progress in developing methodolog-
ical approaches for determining the variation over space in
ecosystem service provision and how this might change in the future
under particular management scenarios (e.g. UK National Ecosystem
Assessment, UKNEA). In addition, attempts have been made to
provide spatial displays of ecosystem service provision (Eigenbrod
et al., 2010; Lavorel et al., 2011). Output from these exercises vary
in complexity from simple land cover maps (e.g. Naidoo et al., 2008)
to complex models incorporating interactions between ecosystem
components and processes (e.g. Goldstein et al., 2012). The aim of
these approaches is to identify hotspots of ecosystem service pro-
vision that should be protected from development (Fig. 2). However,
these approaches represent ecosystem service provision at a ‘fixed’,
static point in time. What is still lacking, and represents a significant
knowledge gap is the continuity of these services over time (Dawson
et al., 2011; Mace et al., 2012) particularly in response to drivers of
change (e.g. species introductions, climate change, land-use change
Nelson et al., 2005).
To understand ecosystem service provision over time and re-
sponses of different services to environmental perturbations re-
quires knowledge of the baseline context of an ecosystem (if such a
state exists), information on the alternative stable states of that
ecosystem, and an understanding of what happens when the sys-
tem is perturbed. Mapping dynamical ecosystem response to
change therefore demands ecological records that span intervals in
time where such responses can be observed. It has been acknowl-
edged a number of times that palaeoecological records can provide
some of these data (e.g. Dawson et al., 2011; Dearing et al., 2012)
and some excellent case-studies have demonstrated the utility of
palaeoecological records in this respect (e.g. Dearing et al., 2012;
Colombaroli and Tinner, 2013; Gosling and Williams, 2013;
McLauchlan et al., 2013a). But in order for palaeoecological re-
cords to become more widely used in the determination of
ecosystem service provision over time, there are fundamental
questions that the palaeo-ecological community at large need to
ask. In some ways, these are similar to those asked when consid-
ering the use of palaeoecological records in biodiversity conserva-
tion and management (e.g. Willis and Birks, 2006; Froyd and Willis,
2008) and include i) what length of temporal record is needed? ii)
what proxies should be used to reconstruct ecosystem processes?
iii) what datasets should be utilised to reconstruct ecosystem ser-
vice provision? iv) what is the relevant spatial scale at which to
Fig. 2. Spatially-explicit biodiversity and ecosystem data are synthesized in order to
generate maps of ecosystem service provision. These maps identify hotspots of
ecosystem service provision that should be prioritized for conservation and ecosystem
management action, modified from de Groot et al. (2010). Copyright 2010 Ecological
Complexity.
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
reconstruct the ecosystem changes? v) what are the most useful
ways to analyse and display the data to make them accessible to
(and useful for) environmental managers and policy makers alike?
In this review, we highlight the different types of proxy records
that appear to be most appropriate, and the spatial and temporal
scales over which these records provide information relevant for
ecosystem management (Table 1). We then move on to assess the
utility of palaeoecological records in determining the persistence of
ecosystem services in response to external drivers of change.
Finally, we consider how this information should be analysed and
displayed in order to be of greatest use in natural resource planning
and how palaeoecological reconstructions are translated into the
currencies required for ecosystem service assessments.
The ecosystem services framework that we have adopted in this
review is that used by the UK National Ecosystem Services
Assessment (UKNEA, 2011 and described in Mace et al., 2012)
which is itself a modification of the MA framework (MA, 2005). The
NEA framework classifies ecosystems (Fig. 1a) into i) Ecosystem
Processes (i.e. the interactions between biota and their
Table 1
Examples of dynamical indicators of ecosystem processes, final ecosystem services and goods provided by palaeoecological proxies.
Proxy
Nutrient cycling d15N of foliage, wood and lake sediments
Soil formation and Elemental (e.g. ICP-MS) and magnetic susceptibility
stabilization analysis of lake sediments
Biomass (a) Pollen accumulation rates in lake sediments
(b) Landscape Reconstruction Algorithm applied to
proportional pollen data
(c) Bayesian inference model of proportional pol-
len data and modern plant occurrence data
Crops (a) Species distribution model constrained by pol-
len assemblages and palaeo-climate data
(b) Size measurements (length, breadth and thick-
ness) of fossilized crop grains
Water supply (a) Assemblages of invertebrate (Coleoptera, Tri-
choptera and Chironomidae) species and their
known association with modern river flow
regimes
(b) Non-parametric analysis (K nearest neighbour)
of tree ring chronologies and historic river or
stream flow records
(c) Diatom-inferred total phosphorus in lakes
Climate regulation Ecosystem process model (e.g. Land Processes and
eXchanges) of terrestrial carbon storage
constrained by pollen assemblage data
Timber (a) Average cumulative radii of each tree derived
from wood ring analysis of living and sub-fossil
trees
(b) Tree age and lifetime growth trajectory inferred
from dendrochronological analysis of incre-
ment bores of living trees
(c) Severity and timing of pest outbreaks for timber
stands from tree ring width chronologies
(d) Fire return interval and intensity from fire scar
records in tree rings and charcoal accumulation
rates in lake sediments
Soil erosion protection (a) Radiometric (210Pb and 137Cs) dating and
palaeo-magnetic analysis (from magnetic
susceptibility) of lake sediments
(b) Cellular automata model of coupled hydro-
geomorphological processes constrained by
lake sediment influx rates and grain size
Coastal protection Amount of time following a disturbance before
mangrove communities (inferred from fossil pollen)
return to a previous state of cover (in % of total
pollen sum)
19
environment which support the provision of ecosystem services),
including soil formation, nutrient cycling, primary production,
pollination and biomass production; ii) Final Ecosystem Service (i.e.
the outputs from ecosystems that directly deliver gains and losses
in human welfare), which includes crops and clean water supply;
and iii) Goods (i.e. the outcomes of ecosystem services that are
valued by humans) such as timber, storm protection, soil erosion.
The research reviewed here demonstrates that recent analytical
and methodological advances, in combination with the growing
number of palaeoecological databases (which provide ecological
data at high spatial and temporal resolution), can yield the infor-
mation required to sustainably manage landscapes for ecosystem
service provision given expected future environmental changes.
2. Ecosystem processes
Ecosystem processes underpin the provision of final ecosystem
services and an understanding of how they vary in time and space
is critical for managing current and future ecosystem services.
Dynamical indicator(s) Citation
Tree- and ecosystem-scale variation in Elliot and Brush, 2006;
N availability, at annual to millennial McLauchlan et al., 2007, 2010,
scale resolution (respectively). 2013b
Rate (accumulation year1) of soil (a) Willis et al., 1997
evolution (a) and stabilization within a (b) Hu et al., 2001
catchment (b)
Rate of change in biomass per year for (a) Magri, 1994; Sepp€
a et al.,
each taxa (a), the proportion of taxa on 2009
the landscape (b) and proportion of taxa (b) Sugita, 2007a, b
within each 12 km2 cell (c), at annual, (c) Paciorek and McLachlan,
centennial, and millennial resolution 2009
from stand to catchment scale
Probability of occurrence (%) of plant (a) Alba-Sanchez et al., 2010;
crop taxa (a) and rate of change in grain Macias-Fauria and Willis,
weight (mg year1) (b), at centennial 2013
and millennial resolution from local to (b) Ferrio et al., 2004, 2005
continental scales
Surface flow rates in rivers and streams (a) Howard et al., 2009, 2010
(m s1) (a), million acre feet (MAF) yr1 (b) Gangopadhyay et al., 2009
(b) where MAF ¼ 1.233  109 m3, and (c) Bennion et al., 2004
total phosphorus concentrations (mg
l1) in lake water at annual to
centennial resolution (c)
Terrestrial carbon storage (Pg C) within Prentice et al., 2011
biomes for target time periods
Volume (m3 yr1) of standing and dead (a) Metsaranta and Lieffers,
trees (a) and minimum cutting cycle (b), 2009
at the stand scale with annual (b) Rozendaal and Zuidema,
resolution; the % reduction in average 2011
maximum and periodic radial growth
per pest outbreak, average interval
between outbreaks (years) and
duration of reduced growth (years); (c) Swetnam and Lynch, 1993
and the occurrence of individual fires
(year), time in between occurrences
(years) and spatial extent of the
disturbance (ha) (d) Higuera et al., 2011
Sediment loss (t ha1 yr1) through (a) Dearing et al., 1987, 2012
time (a) and sediment discharge rates (b) Coulthard et al., 2002
(m3 yr1) into the lake from each 2
e50 m2 cell within the catchment (b)
Recovery rate (years) of mangroves Gonz
alez et al., 2010
following disturbances at the local to
regional scale
20 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Within the ecosystem service framework the processes themselves
are not directly valued in monetary terms; instead their contribu-
tion is assessed in terms of their impact on the quality and quantity
of the final services they support. Key questions in relation to these
ecosystem processes is how do they vary over time, what is the
most appropriate timescale to consider for obtaining data and how
do they respond to environmental perturbations? There are at least
three ecological processes where palaeoecological records have an
important contribution to make in terms of understanding these
questions namely nutrient cycling, soil formation and biomass
change. Each will be considered in turn.
2.1. Nutrient cycling
A key research challenge is to determine how changes in
nutrient cycling in space and time influence ecosystem services.
Nutrients are essential for the growth of all living organisms and
the growth rates of plants are typically limited by the availability of
either nitrogen (N) or phosphorus (P). As such, the availability and
cycling of nutrients underpins many other ecosystem services such
as food, timber and fuel provision (Lavelle et al., 2005). For much of
human history, people have attempted to overcome nutrient limi-
tation of beneficial plant species, particularly crops, by applying
manure and bones to soils. In the early 1900's the invention (and
mass industrialisation) of the HabereBosch process enabled the
large-scale production of synthetic fertilizers. This revolution in
agriculture allowed humans to overcome persistent nutrient limi-
tation of food crops; however, since nutrients are highly mobile in
the soil profile, much of the N and P added to soils through syn-
thetic fertilizers are lost to surface water (i.e. leaching) and the
atmosphere (i.e. gaseous losses). This occurs when the amount
applied exceeds the amount immediately required by the target
plants (or micro-organisms in the soil). Greater amounts of readily
available N in the environment as a result of excess fertilizer use are
now a major source of pollution in some systems; this can lead to
changes in ecosystem structure and functioning, and reduce the
provision of multiple ecosystem services (Compton et al., 2011;
Sutton, 2011). Where does excess N applied to agricultural land-
scapes go? How do increased supplies of reactive N (i.e. the forms of
N that can be directly utilized by plants) affect terrestrial and
aquatic ecosystems? And how long do the effects of N pollution
persist in the system?
Answering these questions requires information on the climatic,
biotic and geologic controls on the long-term fluxes of N through
ecosystems (McLauchlan et al., 2013a). The response of ecosystems
to additional supplies of N is determined to a large extent by the
balance between the supply of reactive N and the amount
demanded by plants and soil organisms. While nutrients can limit
productivity, plants are also known to exert high levels of control
over the availability and cycling of a number of nutrients (Hobbie,
1992; Chapman et al., 2006). Soil microbes can also lock away
(i.e. immobilize) N, reducing the amount returned to soils for up-
take by plants (Chapman et al., 2006). Furthermore, these re-
lationships vary through time in response to changes in
temperature, precipitation, N deposition and CO2 concentrations on
timescales ranging from a few years to decades (Wardle et al.,
2004). Thus, understanding variability in nutrient cycling in
response to climate change requires indices of nutrient availability
that integrate processes acting on the micro- (i.e. seconds to days)
to landscape (i.e. years to centuries) and regional (i.e. ~1 million
years) scales (Walker and Wardle, 2014).
Stable isotope analysis of N (d15N) in organic matter provides a
reliable way to track the availability of N across space (e.g. from
foliage, Craine et al., 2009) and time (e.g. from plant macrofossils,
Wolfe et al., 2013; from wood and lake sediments, McLauchlan
et al., 2007) at each of these relevant scales. Sedimentary d15N
values in particular reflect the full array of processes acting on N
molecules as they cycle through the terrestrial environment,
especially soil denitrification (Houlton and Bai, 2009), and enter the
lake basin through surface runoff (McLauchlan et al., 2013b). In
some cases, anthropogenically-derived N molecules (e.g. from
agricultural and human waste), which tend to have largely enriched
(i.e. higher) d15N values relative to natural sources of N, can leave a
unique ‘signature’ in stream runoff (Harrington et al., 1998; Burns
and Kendall, 2002) which is ultimately recorded in sediments.
Thus, sedimentary d15N records are able to reliably track the
transfer of N from landscapes to aquatic ecosystems (i.e. leaching)
and the atmosphere (i.e. gaseous losses through denitrification).
When N is lost from the terrestrial landscape through leaching
(i.e. nitrates), it enters aquatic ecosystems, typically through small
streams. It is widely known that surplus N can result in eutrophi-
cation in aquatic systems causing hypoxia and anoxia, leading to
the creation of dead zones within coastal ecosystems (Diaz and
Rosenberg, 2008). However, what is unclear from short-term
(<10 years) records, is how long this impact persists within the
ecosystem. In a study by Elliott and Brush (2006), N isotope analysis
of sedimented organic matter retrieved from a wetland adjacent to
the Chesapeake Bay (Maryland, USA), was used to trace the impact
of increasing N inputs from the surrounding agricultural and sub-
urban landscape on the freshwater system over the last 350 years.
They found progressive enrichment of sedimentary d15N over the
last few centuries was associated with forest clearing for agricul-
ture and increasing nutrient inputs from human waste, particularly
in the last 50 years (Elliott and Brush, 2006). This historical
perspective on N cycling was used to provide a sound evidence-
base to argue for the restoration of wetlands due to their ability
to reduce N pollution in coastal ecosystems by increasing rates of
denitrification (i.e. the return of reactive N compounds to inert N2
gas in the atmosphere) on the landscape (Brush, 2009), thereby
providing a key ecosystem service to people in terms of pollution
control.
Ammonia (NH3) is lost to the atmosphere from agricultural
landscapes through the process of volatilization, particularly when
N supply exceeds plant and soil microbe demand. The redeposition
of this reactive form of N on landscapes can have cascading effects
on the structure and function of terrestrial and aquatic ecosystems
(Sutton, 2011). The ability of plants and soil organisms to retain
surplus N is controlled by their requirements (or demands) for N at
the time of deposition. Under enriched atmospheric CO2 condi-
tions, plant demand for N is increased in order to support higher
rates of photosynthesis. The progressive N limitation (PNL) hy-
pothesis in particular posits that as atmospheric CO2 concentra-
tions increase, plant productivity and the ability of plants to
assimilate CO2 could become limited by available N (Luo et al.,
2004; Reich et al., 2006). Thus, the deposition of reactive N could
in fact be supporting higher rates of primary productivity and
increased carbon storage in terrestrial ecosystems, which in turn
reduces the amount of N lost from terrestrial ecosystems. For
example, an estimated 11.2 Pg of carbon stored in terrestrial eco-
systems since 1860 can be attributed to 1.3 Pg of N created by
anthropogenic activities since that time (Zaehle, 2013). Deter-
mining whether terrestrial ecosystems provide a net terrestrial sink
or source for the additional N created by humans, and the time-
scale over which this occurs, is a key research priority for palaeo-
and neo-ecosystem ecologists (Lewis et al., 2014).
To address this question, McLauchlan et al. (2007) conducted an
analysis of tree wood d15N sampled from 857 tree ring segments
(22 trees), which provided a ~170 year record of N cycling in the
northern hardwood forests surrounding Mirror Lake (northeastern
USA). Results from this study intriguingly indicated that trees (and
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
lake sediments) showed declining d15N values from ~1920 to the
present and thus reduced N losses from the terrestrial environ-
ment, supporting the N limitation hypothesis described above. A
study of herbarium specimens of grass species collected across
eastern Kansas between 1876 and 2008 also showed progressive
declines in foliar d15N from 1940 and declining concentrations of N
in foliage from 1926 (Fig. 3), despite relatively high levels of N
deposition in the region in recent decades (McLauchlan et al., 2010).
In contrast, Hietz et al. (2011) presented two series of wood d15N
studies from Barro Colorado Island, Panama and the Huai Kha
Khaeng Reserve in Thailand which both demonstrate rising d15N
from 1968 to 1920 respectively. Thus, the trajectory of change in N
availability to terrestrial plants as indicated by wood d15N appears
to vary by biome.
In order to understand the nature of the relationship between
the dynamics of N cycling and climatic change (including altered
atmospheric concentrations), a much longer record is needed,
ideally pre-dating large scale human alteration of the N cycle and
spanning contrasting biomes and ecosystems. A global synthesis of
86 lake sediment d15N records spanning the last 15,000 years was
therefore undertaken by McLauchlan et al. (2013b). Interestingly
this study indicated a more complex picture than had been previ-
ously assumed. In the early postglacial period when atmospheric
CO2 and temperatures both increased, N availability declined
globally and this phenomenon continued for nearly 8000 years
while terrestrial environments accumulated carbon, which sup-
ports the progressive N limitation hypothesis. However, over the
past 500 years, no globally consistent trends are apparent in d15N
suggesting that local conditions have had an overriding influence
on the ability of terrestrial ecosystems to absorb and retain N. These
findings together demonstrate the ability of terrestrial ecosystems
Fig. 3. Scatter plot of stable isotopic values of nitrogen (d15N, ‰ relative to air), a proxy
for ecosystem N availability, and total N (mg g1) in leaf tissue from 545 herbarium
specimens collected in Kansas (USA) between 1876 and 2008. Decreasing values of
d15N after 1940 show that N availability in plants declined concurrently with rising
fossil fuel burning and thus concentrations of atmospheric carbon dioxide. Piecewise
linear regression identified inflection points in the trajectories of foliar d15N at 1940 (a)
and foliar N at 1926 (b). It is likely that these observations reflect progressive re-
ductions in N availability to plants despite anthropogenic N deposition in recent de-
cades. Long-term declines in N availability could result in subsequent declines in the
ecosystem services and goods that depend upon regular nutrient availability. Repro-
duced with permission from McLauchlan et al. (2010). Copyright 2010 New Phytologist.
21
to provide a long-term sink for increasing supplies of
anthropogenically-derived N, although the persistence of this
process is largely determined by local or regional factors. It also
makes clear the importance of considering the links between the
carbon and N cycles when attempting to predict future levels of
plant productivity and the provision of key services that are derived
from this (e.g. crops, quality of herbage for grazing and timber
production and climate regulation).
2.2. Soil formation
Soils regulate a number of ecosystem processes (e.g. nutrient
cycling) and support the provision of final services such as crops
and clean water supply (Dominati et al., 2010); as such, it has been
suggested that they can be an essential determinant of the eco-
nomic status of countries (Daily, 1997). In fact, the United Nations
General Assembly declared in 2013 the creation of World Soil Day
in recognition of the important contribution they make to human
wellbeing. Despite the clear importance of soils, current under-
standing of their natural capital is incomplete (Dominati et al.,
2010) mostly due to the static nature of existing global soil maps
(Grunwald et al., 2011). Soil formation is a function of parent ma-
terial, climate, biota, topography and time; with recognition that
soils can take 1000s of years to develop (Jenny, 1941). Existing
global soil maps provide only a snapshot of current soil type and
lack information on the processes regulating soil formation; yet this
information is critical for assessing both the impacts of climate and
also land-use change on soil losses, the amount of time it takes to
replenish soil stocks and the impact of changing vegetation type in
determining the type of soil that develops. Efforts are currently
being made to develop ecosystem process models (e.g. the STEP-
AWBH space-time model for digital soil mapping) that can pro-
duce global soil maps which account for these dynamic processes
(Grunwald et al., 2011); this is an area where palaeoecological re-
cords have much to offer.
Palaeoecological proxies can both inform and validate soil for-
mation models by demonstrating variation in the rate of soil evo-
lution (i.e. pedogenic change over time, sensu Huggett, 1998) in
response to intrinsic (e.g. vegetation cover) and extrinsic (e.g.
climate) changes. Pennington (1971, 1986) presented the first at-
tempts at reconstructing soil processes from elemental analysis of
postglacial lake sediments in the English Lake District. More
recently, Willis et al. (1997) used inductively coupled plasma
emission atomic spectrometry of bulk sediments to reconstruct the
timing of a transition from an acidic podzol to an alkaline brown
earth soil in northeastern Hungary. Their interpretation of the
elemental concentrations as indicators of shifting soil types on the
catchment was validated by a comparable analysis of extant leaf
litter and soils from the basin. The results demonstrated that the
development of a brown earth soil from podzol occurred at least
600 years after a shift from a coniferous to an oak-dominated forest
(Willis et al., 1997). This study provided the first clear and direct
evidence of plant modulation of soil chemistry occurring over
millennia from lake sediment records.
Geochemical data can also be used to infer changes in soil
erosion and influx of mineral matter from the catchment into the
lake as well as soil acidification associated with vegetation and
climatic changes (Hu et al., 2001). For example, in a study of Ho-
locene ecosystem changes in Southwestern Alaska, Hu et al. (2001)
used allogenic concentrations of elements (measured with induc-
tively coupled argon plasma/atomic emission spectroscopy) in
sediments to demonstrate progressive soil stabilization from
10,000 yrs BP associated with widespread colonization of the her-
baceous tundra by Betula following postglacial warming. The
integration of such temporal records across a region can therefore
22 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
provide a wealth of information on rates and directions of change in
soils and their variation over space, particularly in light of antici-
pated climate changes.
2.3. Biomass
Plant biomass, a product of primary productivity, is treated as an
ecosystem process (following Mace et al., 2012) because it is
fundamentally linked to the provision of many key ecosystem
services and goods (e.g. climate regulation, soil erosion protection,
clean water provision and timber and crop yields). Tracking the
response rates of plant biomass to climatic perturbations is
essential for predicting how the provision of these services may
vary in the future; doing so requires information on the current
amount of biomass (i.e. the ‘standing crop’), its spatial distribution
and how this varies over annual to millennial time scales. Below we
present three areas where palaeoecological data can provide
important information on temporal and spatial changes in plant
biomass:
2.3.1. Determining changes in amount of plant biomass over time
When determining changes in the amount of plant biomass over
time, fossil pollen accumulation rates (PAR) in sedimentary basins
can provide a useful proxy (Magri, 1994). For example, PAR has been
utilised to demonstrate changes in overall plant biomass on gla-
cialeinterglacial timescales (Magri, 1994), Holocene changes in
biomass for individual plant species (Giesecke and Fontana, 2008;
Seppa € et al., 2009), and Holocene biomass changes associated
with changing community and population dynamics over time
(Sepp€ €ppa and colleagues (2009)
a et al., 2009). In the latter study, Se
demonstrated the potential of using PAR to determine the changes
that would occur in standing biomass with a climate-driven change
in community composition. To do this they first compared varia-
tions in current biomass (tons ha1) of Pinus, Picea and Betula
populations at varying distances (0e4500 m, 0e1000 m, 0e500 m,
and 0e100 m) from the shores of study lakes in Finland with PAR
values measured from surface samples from those lakes and from
others where these taxa were not present. Results indicated a close
correspondence between PAR and local biomass at regional scales
(i.e. up to 4500 m from the lake shore); therefore lake sites with
high biomass of Pinus, Picea and Betula in the surrounding catch-
ment had correspondingly high influx of pollen from these tree taxa
and conversely those sites with low or no biomass, had similarly
low PAR. This relationship was then used to examine the variability
of biomass of different populations (in particular Picea abies) during
the Holocene in response to climate change (Sepp€ a et al., 2009).
Results indicated that maximum Picea PAR values observed in
Finland during the mid-Holocene are indicative of Picea biomass of
about 50e60 tonnes per hectare, which suggests that modern Picea
biomass in the region is roughly 35e40% of its pre-anthropogenic
extent. Thus, using fossil pollen data to assess biomass responses
to climate change provides information on a greater extent of
environmental conditions than is possible with modern observa-
tions alone.
2.3.2. Spatial distribution of plant biomass
Fossil pollen data e in their raw form e have often been criti-
cised for their inability to represent changes in land cover (Gaillard
et al., 2008). In the past decade, a number of approaches have been
developed to translate pollen percentage and accumulation rate
data into estimates of relative abundance of individual plant taxa in
the surrounding catchment and thus provide changes in the spatial
distribution of plant biomass over time. Examples of two Landscape
Reconstruction Algorithms (LRA) that have been developed for this
purpose include REVEALS which provides a regional estimate of
vegetation abundance and LOVE (LOcal Vegetation Estimates)
(Sugita, 2007a, b). These models translate raw pollen counts into
vegetation abundance by taxa at the regional and local scales,
respectively (Sugita, 2007a, b). Testing of these models has
demonstrated their vast potential for land cover reconstruction.
Overballe-Petersen et al. (2013), for example, tested the ability of
the LRA models to reconstruct vegetation cover surrounding a
Danish forest hollow by comparing the model output with
observed vegetation data from historical records (vegetation maps
and inventory data from management plans) at selected points in
time over the last 150 years. The LRA output corresponded very
well with inventory data up to 200 m from the edge of the forest
hollow and provided a more accurate representation of vegetation
cover than pollen percentage data (Overballe-Petersen et al., 2013).
Bayesian hierarchical models of forest composition have also
been built to reconstruct historic variations in vegetation abun-
dance from fossil pollen percentage data (Paciorek and McLachlan,
2009). This approach estimates the relationship between modern
vegetation cover and the representation of these taxa within fossil
pollen assemblages in nearby ponds. Paciorek and McLachlan
(2009) demonstrated the utility of this approach to generate pre-
dicted abundances of forest taxa across central New England at
different points in time. The authors applied this modelling
approach to pollen data collected from 23 ponds where the sur-
rounding vegetation was known for both the present and colonial
era (1635e1800 AD). Cross-validation analysis of their results
showed that the model provided good a prediction of actual forest
composition at regional (i.e. >50 km) scales (Paciorek and
McLachlan, 2009).
2.3.3. Mapping biomass response to climate change over space and
time
Currently, species niche or bioclimatic envelope modelling is the
most commonly used approach to predict the expected future
distribution and abundance of plant species in response to climate
change (Araújo and Peterson, 2012). This approach correlates the
present day distribution of target species with a number of climatic
parameters associated with their distribution to create a climate
envelope model. The climate envelope model is then run again
using estimated future climatic conditions to predict the potential
distributional changes due to climate change. However, it is widely
acknowledged that many of these climate envelope models are
highly sensitive to the algorithms used (Hijmans and Graham,
2006; Araújo and Peterson, 2012) and a full knowledge of the cli-
matic sensitivity of the plant taxa involved is required in order to
make realistic predictions about future range shifts (Clark et al.,
2011). In particular, the distribution of many European plant taxa
is strongly influenced by historical factors including past human
and climate activities (Willis and Birks, 2006; Froyd and Willis,
2008; Willis et al., 2010). Furthermore, the species occurrence re-
cords that are often used to build climate niche models are static
and may not represent the true range of climatic conditions that
species can tolerate. Quantifying the speed at which species will
respond to climate change, and projecting the potential future
ranges of key plant species requires occurrence records that
correspond with the generation time of the major organisms under
investigation. Without this, climate envelope models are unable to
capture the full range of variability of the species under
investigation.
Fossil pollen data therefore have an important role in testing
these models through the process of hindcasting. This is where the
species envelope model is run back in time (using palaeolimatic
data) and the predicted distribution of the species is examined
against the actual distribution as evidenced in the fossil pollen
records. A good example of this was provided by Alba-Sanchez et al.
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
(2010) in their use of the species distribution model MaxEnt
(maximum entropy machine-learning algorithm) to estimate past
distribution changes in Iberian Abies (fir) under changing climate
scenarios including the Last Glacial Maximum (21 kya) and the
mid-Holocene (6 kya) across the Iberian Peninsula. From this study
they concluded that the integration of modern occurrence records
of Abies species, current ecological-niche characteristics associated
with these occurrences, and palaeoecological evidence of past
occurrence provided a more robust bioclimatic niche model,
particularly for identifying the locations of climatic refugia.
3. Final ecosystem services
The outputs of ecosystems from which people derive direct
benefits in terms of welfare gains and/or losses, such as crops, a
clean water supply and an amiable climate, are classified as
ecosystem services (UKNEA, 2011). However, the knowledge
required to understand and manage these ecosystem services
ranges from local scale considerations (e.g. farm fields for crops) to
whole river basin catchments (e.g. clean water provision), to global
scale climatic processes. Each of these processes range over a
similar variety of temporal scales (i.e. seasonal to millennial). It is
extremely difficult to predict the complex ways in which current
and expected future environmental changes will shift the spatial
and temporal mosaic of ecosystem system service provision
(Mooney et al., 2009) and it is a research priority to find datasets
that can aid the development of a more detailed understanding of
environmental change impacts on ecosystem service provision and
how this will affect human well-being into the future (Pereira et al.,
2010). Examples of three key provisioning services where palae-
oecological records provide important temporal data to enhance
this understanding are crops, water supply and quality, and climate
regulation.
3.1. Crops
The provision of food is one of the most direct ways in which
ecosystems provide services to people and a major global challenge
is to determine current and future impacts of environmental
change on food availability e particularly crops (Godfray et al.,
2010). In order to meet expected food demand in 2050, scientists
need to be able to predict the areas for optimal crop growth given
likely climate scenarios (Foresight, 2011). This requires dynamic
data on crop yields and their relationship to changes in climate as
well as nutrient and moisture availability. Currently, the temporal
extent of global data on crop production and climate change used in
ecosystem assessments is limited to ~30 years (Lobell et al., 2011). A
number of recent palaeoecological and archaeological studies have
started to demonstrate the utility of fossil pollen and stable isotope
analysis in extending this record much further back in time.
Similar to modelling biomass distributions through time, the
typical method for determining future distribution of crops in
response to climate change, is through the use of species envelope
models. However, as mentioned above, these are strongly influ-
enced by the algorithms used to create the models and therefore
independent validation under varying conditions is essential. Fossil
records have great potential to do this through the comparison of
species distribution model outputs (run backwards in time) with
actual distribution as displayed in fossil datasets (Fig. 4). For
example, Macias-Fauria and Willis (2013) aggregated data from
fossil pollen assemblages, landscape features (i.e. relief heteroge-
neity) and global palaeo-climate model output in order to generate
the probability of occurrence of seven economically valuable spe-
cies, including the crops Olea europea and Vitis vinifera, across
Europe for three climatically distinct time periods (the medieval
23
warming period, Little Ice Age and present). While the crop species
were the most difficult to model e due to human management
effects on plant occurrence e the modelled distributions yielded
highly accurate predictions (AUC > 0.90) when compared to the
fossil output. However, the authors also demonstrated that a key
factor affecting the predictability of crop taxa distribution was the
cultural aspects driving cultivation patterns, such as changes in
V. vinifera planting under Muslim versus Christian rule in southern
Europe around the 10th century (CE) (Macias-Fauria and Willis,
2013). Thus, historical e as well as the environmental e contexts
can decouple species responses to climatic variables (Jackson et al.,
2009) and future advances in species distribution modelling of long
term ecological data should be focused on the development of new
approaches that can incorporate the impacts of discrete historical
events within the prediction of species ranges.
While species distribution modelling can provide information
on the areas that are likely to provide the optimal climatic condi-
tions for crop growth, novel methods in fossil crop grain research
can provide additional information about changes in the yields of
species within that range. In cereal crops, grain weight is an
important indicator of grain yield, since larger grain weights
directly correspond with higher yields in the end-products of
wheat, such as flour (Henry and Kettlewell, 1996). Cereal grains are
often found in archaeological sites; however the process of pres-
ervation (i.e. carbonisation) acts to reduce grain mass and dimen-
sion (Ferrio et al., 2004). In order to compare the weight of ancient
grains to those grown in modern conditions, it is necessary to
correct for this effect. Ferrio and colleagues (2004) used morpho-
logical analysis of grains obtained from archaeological deposits
across Catalonia, Spain, dated to 3900e2200 cal. yrs BP to develop a
regression model that provides an estimate of the original grain
weight (mg) at the time of burial given data on the length (mm),
breadth (mm) and thickness (mm) of experimentally carbonized
modern grains of wheat and barley (Ferrio et al., 2004). By
comparing the reconstructed grain weights with independent
sources of environmental data (i.e. temperature or moisture), the
authors were able to tease out the effects of increased water
availability due to beneficial climate change versus improvements
in grain weight from human management activities such selective
breeding for heavier grains by early farmers.
Another approach for measuring the direct impacts of climate
change, in particular water availability, on grain size is to use stable
isotope analysis of carbon (d13C) from individual grains. In a study
by Araus et al. (1997) on fossil cereal grains from Catalonia and
Andalusia (SE Spain) dated between 7000 and 2200 cal. yrs BP, for
example, the authors demonstrated that carbon isotope discrimi-
nation (D13C) in individual grains (i.e. the difference between the
d13C values of air and that of the grains) corresponds well with
water availability to ancient crops (Araus et al., 1997). Thus carbon
isotope discrimination can be used as a proxy reflecting both
changing climatic conditions and irrigation practices (Ferrio et al.,
2005). Such insights can provide a clearer understanding of early
agricultural practices, and their long-term impacts on biodiversity
and ecosystem processes (Whitehouse and Kirleis, 2014) which
have relevance for assessing and potentially improving modern
pastoral crop management practices particularly in the face of cli-
matic changes.
Stable isotopic values of N within fossilized grains can also
provide a reliable record of variations in soil fertility at the time of
plant growth, a factor that can have a major impact on crop yield.
For example, in a study of grain weight from charred grain remains
of wheat and barley in Granada, SE Spain spanning 1500 years
following the establishment of agriculture in the region around
6000 cal. yrs BP, Aguilera et al. (2008) found that grain weight
declined dramatically while D13C values (indicating water
24 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Fig. 4. Species distribution maps of important crop species Olea europea and Vitis vinifera describe the past ranges of these taxa during historic periods of climatic warming (i.e.
Medieval Warm Period) and cooling (Little Ice Age) episodes relative to their modern distributions (i.e. 20th Century Warming). Fossil pollen data for these species were used to
constrain climate- and landscape feature-based models of these crop species' range in each period of time. Copyright 2012 Global Ecology and Biogeography.
availability) remained stable. They found that the observed decline
in grain weight was linked with declining soil fertility (as indicated
by decreasing values of grain d15N), not changing climate. Thus, a
progressive loss of soil fertility resulted in a halving of cereal grain
weights over 1000 years despite the presence of favourable climatic
conditions for growth (Aguilera et al., 2008). By reconstructing the
changes through time in the N and water status of crops along with
measures of yield (via grain weight), the authors were therefore
able to identify the long-term effects of early unsustainable farming
practices on the provision of important food crops (Aguilera et al.,
2008). Such data have the potential to provide an empirical basis
for future scenario models of food supply changes in response to
climate change and changing land management techniques
(Pereira et al., 2010).
3.2. Water supply and quality
Access to clean, reliable freshwater supplies is a fundamental
requirement for human life, yet only 25% of global total river runoff
and groundwater recharge is available for human use and 40% of
this amount is already being removed solely for this purpose
(Heathwaite, 2010). Demand for water will increase with popula-
tion growth, yet climate warming and pollution are reducing
existing fresh water supplies (Vorosmarty et al., 2000; Heathwaite,
2010). Water resource planners need to know how water supply
varies under different environmental conditions relative to current
and projected demands. However, historical records of river flow
(in selected regions where such records were regularly collected)
are typically limited to the last 30e50 years (IPCC, 2001). The
temporal and spatial extent of the instrumental record is therefore
too short to capture natural cycles of drought and recovery over
millennia within the hydrological system, particularly in light of the
anticipated effects on water availability by climate change
(Kundzewicz et al., 2008).
There are a number of palaeoecological proxies that are partic-
ularly useful for indicating vulnerability in water supply under
potential climate change scenarios including using tree-ring
chronologies to extend the temporal extent of past flow regimes
(Jain et al., 2002). A good example of this approach can be found in a
project that determined long-term variability in the flow of the
Upper Colorado River (Gangopadhyay et al., 2009). In this study, the
authors compared the instrumental record of past river flows
available from the Lees Ferry Gauge data (spanning the interval
1906e2005) with tree-ring widths spanning the same interval of
time from a number of living trees within the catchment. A clear
relationship between actual-flow data and tree-ring width was
demonstrated (Fig. 5), which enabled a transfer function to be
developed that translated tree-ring chronologies into estimated
rates of river flow (million acre-feet, MAF, yr1). This transfer
function was then applied to tree-ring records dating back to 1482
to determine the long-term variability in the river flow. This is in
itself an interesting study on baseline conditions; however, the
authors also used the data to build a river flow regime model
capable of predicting the probability of a regime shift between
persistently wet or dry states in the river given the number of years
since the previous shift (Gangopadhyay and McCabe, 2010). This
provides an exceptionally useful tool for managers aiming to allo-
cate water to the many users across the basin.
Palaeoecological data from sediment buried in rivers and lakes
can also provide important baseline information on variation in
water supply and quality. This is particularly important evidence for
government agencies tasked with the implementation of clean
water legislation. The European Water Framework Directive (WFD),
for example requires that biological, hydromorphological and
chemical indicators of water quality should be based on the degree
to which present-day conditions deviate from those expected in the
absence of human influence. The Lotic invertebrate Index for Flow
Evaluation (LIFE) and the Palaeo-LIFE indices, for example, have
been developed which use extant and fossil macroinvertebrate
species assemblages (from the Coleoptera, Trichoptera and Chiro-
nomidae orders) to characterise past river flow regimes (Extence
et al., 1999; Greenwood et al., 2006; Monk et al., 2006; Howard
et al., 2009, 2010).
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Fig. 5. Scatter plot of reconstructed and measured flow rates (million acre feet, MAF,
year1) of the Colorado River at the Lees Ferry, Arizona (USA) gauge for the years
1906e2005. Annual stream flow was reconstructed from a network of tree ring
chronologies across the Upper Colorado River Basin. The least squares fit line shows
that a very strong relationship exists between the reconstructed and observed values
of stream flow, which allows researchers to infer palaeo-stream flows from dendro-
chronologies that pre-date instrumental records. Reproduced from Gangopadhyay
et al. (2009). Copyright 2009 Water Resources Research.
In a study by Monk et al. (2006), for example, ten years
(1990e2000) of stream flow measurements taken from 83 rivers in
England and Wales were used to establish classes of river flow
velocities (e.g. <20 cm s1, 20e100 cm s1 and >100 cm s1). The
macroinvertebrate species associated with these different veloc-
ities were measured and then a stepwise, multiple linear regression
model of the data was used to translate the presence of macro-
invertebrate species within each flow regime into an index of flow
velocity. The resulting index provides a reliable approach for
reconstructing previously unknown river flow velocities from
known invertebrate assemblages. The method was also used by
Howard et al. (2009, 2010) to reconstruct palaeo-flow regimes
(PalaeoLIFE index) from sub-fossil assemblages of Coleoptera, Tri-
choptera and Chironomidae species buried in exposed fluvial se-
quences of the River Trent in England during the middle to late
Holocene. In this study Howard et al., found a shift from moderate
to slow flow velocities that was coincident with increased erosion
from Neolithic clearing and animal grazing (Howard et al., 2010).
In addition, the use of palaeoecological data to reconstruct water
quality baselines for standing water-bodies is illustrated in a study
by Bennion et al. (2004) that used fossil diatom assemblages to
assess eutrophication impacts and baseline reference conditions for
Scottish freshwater lochs. In this, they collected sedimentary cores
from 26 Scottish freshwater basins (lochs) spanning the time in-
terval c 1850 (i.e. pre-Industrial) to the present and reconstructed
the changes that occurred in the fossil diatom assemblages. By
comparing the base and top samples for each core they were able to
determine the amount of turnover and floristic change in the lochs,
a factor which would have been strongly influenced by the nutrient
status of the lake. They found that periods of time with high
turnover rates were indicative of high nutrient enrichment and
thus low water quality. Given the tight coupling between diatom
compositional changes and levels of phosphorus concentrations in
the lake, the authors used a transfer function to translate diatom
species assemblages into concentrations of total phosphorus (TP) in
the water (mg l1). This latter measure is particularly relevant for
management plans because the results indicated that although 19
lochs have had increases in TP, 6 lochs had no change or negligible
25
increases in TP and one loch had a decline in TP since pre-Industrial
times (Bennion et al., 2004). Thus the individual nature and history
of land-management practices around each site has strongly
influenced the changes in water quality in response to nutrient
loading. It was concluded that restoration to pre-Industrial condi-
tions, as required by the WFD, must proceed based on loch-specific
conditions, not a simplistic relationship between TP loading and
water quality.
3.3. Climate regulation
The ability of woody plants to take up carbon dioxide (CO2) from
the atmosphere and store it within woody biomass provides a
benefit to people in the form of climate regulation (Sedjo and
Sohngen, 2012). Payment for ecosystem service (PES) schemes are
being developed (e.g. Reducing Emissions from Deforestation and
forest Degradation, REDD) that aim to assign a monetary value to
forests that are left intact. The value of forests is determined by
their ability to remove and store anthropogenic emissions of CO2
(Kindermann et al., 2008). The success of PES schemes for carbon
storage, therefore, depends upon obtaining accurate assessments of
the current and future ability of forest plots across the globe to store
carbon (Asner et al., 2010). Since the carbon storage ability of
woody plants is affected by dynamic environmental factors such as
climate, nutrients, and fire, assigning a carbon storage value to a
forest is particularly challenging and presents an important
research challenge (Bonan, 2008).
Recent work, for example, has indicated that although between
2000 and 2007 the global forest carbon sink removed 2.5 billion
tonnes of carbon per year from the atmosphere (Pan et al., 2011),
the amount drawn down by different forest types varies greatly.
The most draw-down was contributed by tropical forests (1.3
billion tonnes per year), followed by temperate forests (0.8 billion
tonnes per year) and boreal forests (0.5 billion tonnes per year).
Probably most surprising was the finding that within the tropical
zone secondary forests (i.e. those that are recovering from past
deforestation and logging) provide the largest sink, ~1.7 billion
tonnes of carbon per year due to their relatively rapid early forest
growth (Pan et al., 2011). An understanding of how these different
forest types respond to climatic perturbations and how quickly
they recover from previous disturbance are therefore important
factors that need to be taken into consideration when modelling
potential draw-down.
Palaoecological records have much to offer in this respect e not
only providing evidence of the type of forest which is likely to re-
turn following disturbance (be it human activities or climatic per-
turbations) but also the amount of time required for a particular
forest type to recover to previous levels of biomass (and thus car-
bon storage capacity). In a study by Cole et al. (2014) for example,
which examined rates of recovery of tropical forest following pre-
vious disturbance events using fossil pollen sequences from the
four major areas of tropical forest globally (South America, Central
America, Africa and Southeast Asia), results indicated huge varia-
tion in recovery rates. In some tropical regions, recovery of the
forests occurred within 30 years while in other regions it took up to
500 years. There were also significant differences in rates of re-
covery between regions. Forests in Central America recovered the
quickest whereas South American forests recovered the slowest.
The type of disturbance also affected recovery rates. Recovery was
the quickest following natural disturbances (e.g. hurricanes, fires)
compared to human disturbance (Cole et al., 2014). This global
analysis provides important baseline information for modelling
variation in carbon budgets in response to perturbations.
Palaeoecological data has also been used to constrain global
models of past amounts (Pg C) of carbon storage in the terrestrial
26 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32

biosphere (Peng et al., 1998). For example, Prentice and colleagues

used a dynamic global vegetation model called Land Processes and
eXchanges (LPX) to estimate global terrestrial carbon uptake at the
Late Glacial Maximum (LGM) relative to the pre-Industrial Holo-
cene period (Prentice et al., 2011). Pollen-based reconstructions of
biome distributions derived from BIOME 6000 (Prentice et al.,
2000) at each time period showed overall agreement with the
LPX simulated biome distributions, thus validating the ability of the
model to reconstruct biome shifts. Then the validated LPX model
was used to show that carbon storage in vegetation and soils at the
LGM was 620 Pg C less than in the pre-industrial Holocene period
and this was largely due to reduced carbon density in forest biomes
at that time. The model-based difference in terrestrial carbon
storage between the LGM and the pre-industrial Holocene was
found to fall within the range of values (i.e. 300e700 Pg C) inferred
from stable isotope analysis of carbon d13C in marine sediments
(Prentice and Harrison, 2009).
Forest fires are also known to affect the ability of terrestrial
ecosystems to store carbon in forest soils (Nave et al., 2011), yet
existing PES schemes such as the Reducing Emissions from Defor-
estation and Degradation (REDDþ) strategy have not yet incorpo-
rated the effect of forest fires on carbon storage (Barlow et al.,
2012). Given the evidence for recent increases in wildfires
(Westerling et al., 2006), it is essential to understand how long-
term changes in fire regime may affect the ability of forests to
provide this climate regulation service. Estimates of carbon release
from terrestrial ecosystems due to fire (derived from charcoal
accumulation rates, see more about this in Section 4.1) show that
biomass burning in the Holocene has had significant impacts on
global carbon dynamics (Carcaillet et al., 2002). This is an important
area of future research and resources such as the Global Paleofire
Database (Power et al., 2010) have the potential to provide an
invaluable resource for an integrated analysis of the complex re-
lationships between fire regime and carbon storage capacity of Fig. 6. Logs of Diptocarpaceae tree species harvested from natural forests in Sabah,
Malaysia (a). Tree ring chronologies can be reconstructed from cores taken from living
terrestrial ecosystems. or fallen trees (b), which provides high temporal resolution data on wood growth rates
(volume m3 year1) and how this varies with respect to intrinsic and extrinsic drivers
4. Goods of change. Images reproduced, with permission, from (a) Jake Snaddon and (b) Sandra
Nogue
.

Ecosystem ‘goods’ are the outcomes that final ecosystem ser-

vices produce and the values that these generate for people. The
value of ecosystem goods can be based on a monetary measure (e.g.
the market cost of timber) or a measurable improvement in health
(e.g. reduction in mortality rate) or personal security (e.g. reduced
loss of property). The production of ecosystem goods typically in-
volves some form of human investment (Mace et al., 2012), such as
a production process that transforms trees into timber for building
purposes. Three examples of ecosystem ‘goods’ where the long-
term perspective provided by palaeoecological records can
greatly enhance knowledge of its value and variation over time are
timber, soil erosion protection and coastal erosion protection.
4.1. Timber
Human demand for timber resources is increasing and much of
this demand is being met from managed plantations (i.e. 35% of
harvested roundwood), although some premium woods continue
to be harvested from natural, old-growth forests (Fig. 6a) (Sampson
et al., 2005). In order to set sustainable harvest rates for these high-
value forests, managers typically develop timber harvest plans that
require information on current stocks (i.e. standing volume) and
predictions of future timber availability given the expected growth
rate of existing trees and new recruitment. Since individual tree
growth is affected by intrinsic factors such as age and competition
with neighbours, as well as extrinsic factors including climate and
disturbances from fire and pests, predictions need to take into
account the variability in wood production associated with these
drivers. Foresters have traditionally relied upon repeated mea-
surements of tree diameter at the breast-height (DBH) of the
researcher as a proxy for increases in timber volume over time
because this information can be retrieved quickly from a number of
trees in a research sample plot. However, repeated DBH measure-
ments over time taken from experimental forest plots have known
limitations namely: 1) observations are limited to sampling periods
and miss the variability in annual and sub-annual growth rates that
occur between these (Metsaranta and Lieffers, 2009) and 2) the
temporal extent of sampling efforts tend to be short relative to the
lifespan of the study trees (Rozendaal and Zuidema, 2011).
Wood ring growth analysis of increment bores taken from living
trees (Fig. 6b) provide precise, annualized data on variability in
wood growth which is not captured in DBH measurements.
Metsaranta and Lieffers (2009) for example, used dendrochrono-
logical analyses of radial growth in trees to assess differences in
estimates of stand volume growth, mortality and stem wood
biomass between DBH measurements taken at 10-year intervals
and wood cores and disks taken from living and dead trees
(respectively) from an unmanaged jack pine (Pinus banksiana)
stand in the Canadian boreal forest. They found that DBH mea-
surements taken on 10-year intervals miss significant annual
variation in volume growth, mortality and snag-fall, and thus do
not provide an adequate measure of the stochastic processes acting
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
on wood growth in the boreal forest which are recorded in annual
growth rings.
Dendroecological studies can also be used to predict the amount
of time required for high-value forest trees to reach the minimum
diameter for cutting (i.e. to set the minimum cutting cycle); this is
essential baseline information for forest managers aiming to opti-
mise their harvests. In a study by Therrell et al. (2007), for example,
the researchers obtained tree ring measurements and tree age
determinations from increment bores to show that Pterocarpus
angolensis requires 100 years of growth before it will reach the
minimum cutting diameter in southern Africa (Therrell et al., 2007).
What has also been demonstrated is that these species-specific
measurements vary strongly with respect to site conditions
(Rozendaal and Zuidema, 2011).
While climate change may affect the ability of valuable timber
species to persist in their current locations over the long-term
(Macias-Fauria and Willis, 2013), more immediate threats to
standing stocks of timber may be provided by disturbances such as
pests (e.g. bark beetle, Raffa et al., 2008), pathogens (e.g. ash dieback,
Kowalski, 2006) and increased wildfires (e.g. in boreal forest, Macias
Fauria and Johnson, 2008) as well as the interactive effects of these
disturbances on nutrient supplies to forests (McLauchlan et al.,
2014). Pest outbreaks in particular can greatly reduce timber
yields; therefore, forest managers need to know the extent to which
timber volume is reduced by outbreaks of varying intensity and how
this varies under different environmental conditions.
There are now a number of on-line tools available that can
predict the impact of an infestation on timber yield. The Spruce
Budworm Decision Support System (SBW DSS), for example, pro-
vides forest managers with an estimated benefit (in terms of
marginal timber supply, m3/ha) of protecting stands against spruce
budworm (Choristoneura occidentalis) defoliation (MacLean et al.,
2001). In this model, the predicted future benefit of management
interventions is determined by historical defoliation impacts.
However, these estimates are derived from empirical records of
previous defoliation impacts spanning a period of only five years.
Incorporation of longer-term records would greatly improve the
predictive ability of such tools. Dendroecological records in
particular can provide information on past outbreak impacts on
tree growth (Swetnam and Lynch, 1993) and the effect of infesta-
tion on the radial growth of trees (Swetnam et al., 1985). Swetnam
and Lynch (1993) for example, used this approach to show that past
western spruce budworm outbreaks in the southern Rocky
Mountains led to a 50% decrease in the average maximum radial
growth of host trees during each outbreak that occurred between
1700 and 1983 (with an average duration of 12.9 years). Further
work showed that outbreaks in this region were associated with
wet periods, not drought years as was previously believed
(Swetnam and Betancourt, 1998).
In another study, Simard et al. (2002, 2006) used macrofossil
remains of insect pests such as spruce budworm faeces and Lepi-
doptera head capsules buried in thick humus deposits and mires in
Quebec, Canada to extend the chronology of past pest outbreaks
beyond the lifespan of living trees. They demonstrated that the
abundance of fossil insect remains corresponded in time with
changes in local plant species composition (inferred from fossil
pollen assemblages); however, intense budworm activity as seen in
the 20th century was found to be a rare occurrence during the
Holocene and these large outbreak events tended to coincide with
periods of low fire recurrence and warm climatic conditions
(Simard et al., 2006). The mechanisms underlying the interaction
between fire recurrence and insect outbreaks in forests however
remains unclear, and is a priority area for research into the factors
determining forest resistance and resilience to disturbance (Flower
et al., 2014).
27
Where insect remains are poorly preserved, such as in alkaline
lakes, fossil pollen assemblages and geochemical analysis of bulk
material in lake sediment sequences can be used to indicate the
timing of past outbreaks of insect defoliators (Morris et al., 2010,
2013). Morris and Brunelle (2012), for example, showed that ra-
tios of host to non-host taxa pollen accumulation rates can provide
a conservative estimate of past, large magnitude outbreaks of
spruce bark beetles (Dendroctonus rufipennis Kirby). Furthermore, a
key strength of lake sediment-based reconstructions of past pest
outbreaks is the ability to reconstruct the cascading effects of in-
festations on ecosystem processes such as soil erosion (and the soil
erosion protection service provided by this process) and N leaching
(and the service of nutrient pollution control) when combined with
additional biogeochemical proxies such as sedimentary geochem-
istry and stable isotope analysis (see Sections 2.1 and 2.2). These
factors in turn determine forest response to disturbance (Morris
et al., 2013; McLauchlan et al., 2014); therefore unpicking the
feedback mechanisms between them is an important area of cur-
rent research (see the Novus Research Coordination Network for
more information: https://novusrcn.wordpress.com/tag/nitrogen/).
Wildfires, like pest outbreaks, reduce tree growth, destroy the
commercial value of forest stands and can limit the future yields of
timber species (MacLean, 1990). What is also known however, is
that many forests have a regular cycle of burning related to both
abiotic drivers (e.g. seasonality and lightening strikes), and biotic
factors (e.g. the amount of standing dry mass e and thus flamma-
bility of the system). Understanding this burning regime is critical
to the management of forests e especially if burning is used as a
tool to create forest gaps and encourage more plant diversity
(Turner et al., 1994; Colombaroli and Tinner, 2013). Dendroeco-
logical and micro-fossil data recovered from lake sediments can
therefore provide important temporal information on the timing
and intensity of past fire events. Higuera and colleagues, for
example, have shown that information from fire scars on extant
trees can be integrated with charcoal accumulation rates (CHAR) in
forest hollows (Higuera et al., 2005) and lake sediments (Higuera
et al., 2011) in order to establish precise fire return intervals
(years), estimates of area (ha) burned per fire (a proxy for fire in-
tensity) and how this varies with respect to stand age and pre-
vailing climate over centennial to millennial timescales. A
validation study conducted on a 300-yr tree-ring based fire history
spanning a 128,840 ha area within central Yellowstone National
Park showed that peaks in CHAR represented known fire occur-
rences within 1.2e3.0 km of their study lakes while total CHAR
values provided a reliable prediction of the area burned per fire
within 6e51 km of the lakes. This is a particularly good example
where multiple lines of palaeoecological evidence were combined
with statistical analysis in order to translate fossil data into the
values or currencies required by ecosystem managers. In fact, fossil
records of past fire regime dynamics (i.e. pollen and charcoal) were
the only palaeoecological indicators included in the Millennium
Ecosystem Assessment (MA, 2005).
4.2. Soil erosion protection
Soil erosion is a natural process that can be accelerated by
climate change and human activities; this can lead to significant
economic costs to people due to the loss of productive land for
agriculture (de Groot et al., 2002) and impacts on water quality and
supply (see Section 3.2). Ecosystem components (i.e. vegetation
cover) and processes (i.e. water and sediment retention by plants)
act to reduce the flow of sediments off of the landscape and into
water bodies, where additional losses can be incurred in terms of
water quality changes. Afforestation policies are widely promoted
in order to reduce erosion; however this management approach is
28 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
not appropriate for all ecosystems. For example, expensive affor-
estation projects have been undertaken in arid and semi-arid areas
in China (i.e. 100 billion USD since 2000), particularly in natural
grass and shrublands, in an effort to reduce erosion and stop the
effects of desertification. However, these projects have instead had
undesirable feedback effects on natural vegetation cover, soil
quality and water availability (Farley et al., 2005; Cao et al., 2011),
such that in many places the water table has dropped 30e50%,
resulting in increased rates of erosion and desertification due to the
poor selection of tree species for afforestration projects.
What has been learnt from these studies is that there is a critical
knowledge gap on how changes in climate and vegetation cover
relate to changes in the ability of the landscape to retain sediment.
Filling this gap requires information on the flows of soil from land
into sedimentary basins over decadal-to-millennial timescales. A
palaeoecological technique that is particularly relevant in this
respect is the measurement of magnetic susceptibility of sedi-
mentary sequences where the ability of a landscape to regulate
sediment flows through time can be reconstructed from frequency
dependent magnetic susceptibility (MS) values of lake sediments
(Dearing et al., 1987, 2012); high values of MS equate to high rates of
sediment loss (measured as tons ha1 year1) within the catchment
and thus low levels of erosion regulation. A good example of using
this approach can be found in Dearing et al. (2012) where they
mapped the provision of this service from 1800 to 2006 to show
that agricultural intensity in the Lower Yangtze basin was coinci-
dent with a progressive loss of erosion protection and other
services.
Variation in soil erosion protection can be visualized over space
as well as time with the use of a cellular automaton model.
Coulthard et al. (2002) describe a hydro-geomorphological model
called CAESAR that can simulate changes in the hydrological and
sediment regime of a catchment given input data on detrital sedi-
ment influx and sedimentation rates obtained from lake sediments.
Model outputs provide spatially explicit information on the de-
livery of sediments into a lake and how these vary with respect to
past land use and climate changes. The CAESAR model is able to
simulate sediment discharge rates (m3 year1) over short and long
time scales (i.e. 10e100,000 years) at a fine temporal (i.e. hourly to
annually) and spatial (i.e. 2e50 m2 cells) resolution. It can also
separate the effect of multiple drivers such as rainfall and land use
change on the movement of sediment through the catchment
(Welsh et al., 2009). This model can be used to anticipate potential
future changes in sediment transport given different management
strategies and potential future climate change scenarios, which will
be invaluable for informing restoration activities before costly
mistakes are made.
4.3. Coastal protection
Coastal ecosystems produce disproportionately more ecosystem
services than most other systems yet these valuable resources are
experiencing the most rapid environmental changes (Agardy et al.,
2005). Intact mangrove habitats in particular are very important for
marine productivity, local shrimp and fish production, coastal land
stabilization and storm protection; however they are also cut down
to provide wood to meet local fuel and construction demands
(Walters et al., 2008). There is thus an inherent trade-off between
the short-term benefits of mangrove clearing (i.e. fuel and con-
struction material provision) and the long-term benefits of
mangrove protection in terms of increased shrimp and fish harvest
income and increased protection to coastal properties (McNally
et al., 2011). However, to fully understand this trade-off, it is
essential to determine the amount of time required for mangroves
to regenerate. Gonza lez et al. (2010) used fossil pollen data taken
from a Caribbean mangrove to assess the time required for
mangrove species to recover (i.e. to be present in the fossil pollen
record) following disturbances from sea level change, hurricanes,
climate change and human disturbances. They showed that re-
covery of mangrove species (e.g. Rhizophora, Laguncularia)
following a major hurricane took up to 100 years while recovery
from clearance due to changes in land use (from coconut planta-
tions to commerce) took less than 10 years (Gonza lez et al., 2010).
The management benefit of this information is clear: the data show
that the costs of mangrove clearing (i.e. loss of fishing income and
land stabilization services) will persist for at least a decade,
therefore the short-term benefits of clearance must exceed the net
present value of the expected costs in terms of lost ecosystem
services over this period.
5. Future directions of palaeo-ecosystem service research
Human impacts and natural environmental change are altering
the ability of ecosystems to provide essential goods and services
(MA, 2005). The ability of land use planning and ecosystem man-
agement tools to deliver sustainable management solutions in the
face of on-going environmental change is dependent upon the in-
clusion of information on the long-term variation in ecosystem
service provision; this has yet to be achieved. This review has
illustrated that (1) palaeoecological data are capable of being used
to reconstruct the baseline dynamics of a large number of essential
ecosystem processes, final services and goods, and (2) these data
are available at the high spatial and temporal resolution that is
required for managing biodiversity and ecosystem services at the
landscape and regional scales. However, three major developments
are now required in palaeoecological research in order to underpin
the utility of long-term data for understanding and mapping vari-
ation in ecosystem service provision over time and space (Fig. 7) as
follows:
1) Firstly, it needs to be recognised that novel analytical tech-
niques in palaeoecology and their validation in modern and
experimental studies have greatly advanced our ability to recon-
struct changes in ecosystem processes and function over time. A
few good examples of where this has recently been accomplished is
with sedimentary d15N as a proxy for N availability (McLauchlan
et al., 2007), dung fungal spore accumulation rates (Gill et al.,
2013) and beetle assemblages (Smith et al., 2014) in sediments as
proxies for large herbivore biomass, and PAR as a proxy for plant
biomass dynamics (Sepp€ a et al., 2009). These developments are
rapidly increasing the amount of information about ecosystem
processes and services that can be derived from palaeoecological
studies. Furthermore, by demonstrating precisely how modern
measurements equate to their fossil proxies, these developments
are enabling researchers to cross boundaries between neo- and
palaeo-ecological research, which is essential in order for palae-
oecological data to be included in ecosystem assessments. As new
palaeoecological proxies are developed, it is essential that they are
validated against modern ecological measures used in ecosystem
assessments.
2) Secondly, recent developments in modelling techniques have
also advanced our ability to translate raw palaeoecological data into
units that are comparable with neo-ecological observations and/or
compatible with ecosystem service mapping tools (Table 1). For
example, statistical techniques allow us to translate raw pollen data
into changes in vegetation density over time and space (Sugita,
2007a, b; Paciorek and McLachlan, 2009). Another advancement
is in the use of palaeoecological data to build and validate
ecosystem process models, which simulate complex ecosystem
dynamics at a variety of spatial scales and allow us to determine
how changes in one or more variables cascades through the system
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Fig. 7. Conceptual diagram showing how palaeoecological data taken from individual study sites can be integrated across space and time to generate maps that highlight areas that
have sustained a relatively high level of ecosystem service provision over time.
and leads to changes in plant community dynamics and ecosystem
functions (Anderson et al., 2006). Models such as these are used in
ecosystem assessments (e.g. the MA and UKNEA) to generate
possible future trajectories of ecosystem change given alternative
management and environmental change scenarios (MA, 2005).
When this is done within a Bayesian Belief Network model (BBN)
(Chen and Pollino, 2012), the predictions can be generated along
with an estimate of probability (UKNEA, 2011). These BBNs are also
able to utilize a mix of qualitative and quantitative data from
multiple sources as inputs for ecosystem model development
(McCann et al., 2006). Thus, palaeo- and neo-ecological data (either
as continuous data or discrete events) are able to be integrated
within a holistic model of current and future ecosystem dynamics.
In order for long-term data to be utilised within the tools and
modelling approaches used by landscape and ecosystem managers,
the palaeoecological community must take the lead in demon-
strating the benefits (in terms of improved ecosystem service
outcomes) that can be derived from including these data within
existing ecosystem modelling frameworks. Achieving this requires
increased training in the use of such modelling methods by
palaeoecologists and wider inclusion of such modelling approaches
within traditional palaeoecological studies.
3) Lastly, sustainable land use planning requires the integration
of ecological information over large spatial scales (de Groot et al.,
2010). The synthesis of palaeoecological records from palae-
oecological databases allows for the reconstruction of ecosystem
dynamics over space as well as time. The recent proliferation of
palaeoecological databases (NEOTOMA, Global Pollen Database,
Global Charcoal Database, NCDC etc.) has facilitated the use of past
reconstructions to address pressing questions about the impacts of
regional, continental and global environmental change on
ecosystem service provision (McMahon et al., 2011; Brewer et al.,
2012). However, the usefulness of these datasets for managing
ecosystems could be constrained by their relative inaccessibility to
non-specialists. What is required is an ability to produce maps of
ecosystem characteristics (e.g. forest composition, soil type, N
29
availability, etc.) at specific time slices (e.g. every 50 or 100 years)
from the data stored in these databases (Fig. 7). Here we have
demonstrated the availability of palaeoecological data to meet this
need; what remains to be achieved is the translation of these data
into map-able information layers. Achieving this requires the
development of a common framework for the interpretation of
palaeoecological proxy data, their chronologies and spatial
boundaries.
This review has illustrated a small sub-section of the cutting-
edge research that is already being done in the developing field
of palaeo-ecosystem service research; there is a huge amount of
relevant data e but the task is now to illustrate its usefulness to
non-palaeoecologists. This can be achieved by utilising current (and
future) site-scale records to generate spatially and temporally
explicit tools for managing biodiversity and ecosystem services
across landscapes so that they might be sustained into the future
despite impending global environmental change.
Acknowledgements
ESJ was funded by a James Martin Research Fellowship from the
Oxford Martin School. SN acknowledges financial support from a
Postdoctoral fellowship from the Spanish Ministry of Education
(EX2009-0669) and the Norwegian Academy of Science and Letters
through the VISTA programme. Thanks to Jake Snaddon for giving
advice on figure design and providing the picture of harvested trees
in Sabah.
References
Agardy, T., Alder, J., Dayton, P., Curran, S., Kitchingman, A., Wilson, M., Catenazzi, A.,
Restrepo, J., Birkeland, C., Blaber, S., Saifullah, S., Branch, G., Boersma, D.,
Nixon, S., Dugan, P., Davidson, N., Vo € ro
€ smarty, C., 2005. Coastal systems. In:
Hassan, R., Scholes, R., Ash, N. (Eds.), Ecosystems and Human Well-being:
Current State and Trends. Island Press, Washington, DC, pp. 513e549.
Aguilera, M., Araus, J.L., Voltas, J., Rodriguez-Ariza, M.O., Molina, F., Rovira, N.,
Buxo, R., Ferrio, J.P., 2008. Stable carbon and nitrogen isotopes and quality traits
30 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
of fossil cereal grains provide clues on sustainability at the beginnings of
Mediterranean agriculture. Rapid Commun. Mass Spectrom. 22, 1653e1663.
Alba-Sanchez, F., Lopez-Saez, J.A., Benito-de Pando, B., Linares, J.C., Nieto-Lugilde, D.,
Lopez-Merino, L., 2010. Past and present potential distribution of the Iberian
Abies species: a phytogeographic approach using fossil pollen data and species
distribution models. Divers. Distrib. 16, 214e228.
Anderson, N.J., Bugmann, H., Dearing, J.A., Gaillard, M.J., 2006. Linking palae-
oenvironmental data and models to understand the past and to predict the
future. Trends Ecol. Evol. 21, 696e704.
Araújo, M.B., Peterson, A.T., 2012. Uses and misuses of bioclimatic envelope
modeling. Ecology 93, 1527e1539.
Araus, J.L., Febrero, A., Buxo, R., Camalich, M.D., Martin, D., Molina, F.,
RodriguezAriza, M.O., Romagosa, I., 1997. Changes in carbon isotope discrimi-
nation in grain cereals from different regions of the western Mediterranean
Basin during the past seven millennia. Palaeoenvironmental evidence of a
differential change in aridity during the late Holocene. Glob. Change Biol. 3,
107e118.
Asner, G.P., Powell, G.V.N., Mascaro, J., Knapp, D.E., Clark, J.K., Jacobson, J., Kennedy-
Bowdoin, T., Balaji, A., Paez-Acosta, G., Victoria, E., Secada, L., Valqui, M.,
Hughes, R.F., 2010. High-resolution forest carbon stocks and emissions in the
Amazon. Proc. Natl. Acad. Sci. 107, 16738e16742.
Balmford, A., Bond, W., 2005. Trends in the state of nature and their implications for
human well-being. Ecol. Lett. 8, 1218e1234.
Barlow, J., Parry, L., Gardner, T.A., Ferreira, J., Aragao, L.E.O.C., Carmenta, R.,
Berenguer, E., Vieira, I.C.G., Souza, C., Cochrane, M.A., 2012. The critical impor-
tance of considering fire in REDDþ programs. Biol. Conserv. 154, 1e8.
Bennion, H., Fluin, J., Simpson, G.L., 2004. Assessing eutrophication and reference
conditions for Scottish freshwater lochs using subfossil diatoms. J. Appl. Ecol. 41,
124e138.
Bonan, G.B., 2008. Forests and climate change: forcings, feedbacks, and the climate
benefits of forests. Science 320, 1444e1449.
Brewer, S., Jackson, S.T., Williams, J.W., 2012. Paleoecoinformatics: applying geo-
historical data to ecological questions. Trends Ecol. Evol. 27, 104e112.
Brush, G.S., 2009. Historical land use, nitrogen, and coastal eutrophication: a
paleoecological perspective. Estuaries Coasts 32, 18e28.
Burns, D.A., Kendall, C., 2002. Analysis of d15N and d18O to differentiate NO3
sources in runoff at two watersheds in the Catskill Mountains of New York.
Water Resour. Res. 38, 9-1e9-11.
Cao, S.X., Chen, L., Shankman, D., Wang, C.M., Wang, X.B., Zhang, H., 2011. Excessive
reliance on afforestation in China's arid and semi-arid regions: lessons in
ecological restoration. Earth-Sci. Rev. 104, 240e245.
Carcaillet, C., Almquist, H., Asnong, H., Bradshaw, R.H.W., Carrion, J.S., Gaillard, M.J.,
Gajewski, K., Haas, J.N., Haberle, S.G., Hadorn, P., Muller, S.D., Richard, P.J.H.,
Richoz, I., Rosch, M., Goni, M.F.S., von Stedingk, H., Stevenson, A.C., Talon, B.,
Tardy, C., Tinner, W., Tryterud, E., Wick, L., Willis, K.J., 2002. Holocene biomass
burning and global dynamics of the carbon cycle. Chemosphere 49, 845e863.
Carpenter, S.R., Mooney, H.A., Agard, J., Capistrano, D., DeFries, R.S., Diaz, S., Dietz, T.,
Duraiappah, A.K., Oteng-Yeboah, A., Pereira, H.M., Perrings, C., Reid, W.V.,
Sarukhan, J., Scholes, R.J., Whyte, A., 2009. Science for managing ecosystem
services: beyond the millennium ecosystem assessment. Proc. Natl. Acad. Sci. U.
S. A. 106, 1305e1312.
Chan, K.M.A., Shaw, M.R., Cameron, D.R., Underwood, E.C., Daily, G.C., 2006. Con-
servation planning for ecosystem services. PLoS. Biol. 4, 2138e2152.
Chapman, S.K., Langley, J.A., Hart, S.C., Koch, G.W., 2006. Plants actively control
nitrogen cycling: uncorking the microbial bottleneck. New Phytol. 169, 27e34.
Chen, S.H., Pollino, C.A., 2012. Good practice in Bayesian network modelling. En-
viron. Model. Softw. 37, 134e145.
Clark, J.S., Bell, D.M., Hersh, M.H., Nichols, L., 2011. Climate change vulnerability of
forest biodiversity: climate and competition tracking of demographic rates.
Glob. Change Biol. 17, 1834e1849.
Cole, L.E.S., Bhagwat, S.A., Willis, K.J., 2014. Recovery and resilience of tropical for-
ests after disturbance. Nat. Commun. 5.
Colombaroli, D., Tinner, W., 2013. Determining the long-term changes in biodiver-
sity and provisioning services along a transect from Central Europe to the
Mediterranean. Holocene 23, 1625e1634.
Compton, J.E., Harrison, J.A., Dennis, R.L., Greaver, T.L., Hill, B.H., Jordan, S.J.,
Walker, H., Campbell, H.V., 2011. Ecosystem services altered by human changes
in the nitrogen cycle: a new perspective for US decision making. Ecol. Lett. 14,
804e815.
Coulthard, T.J., Macklin, M.G., Kirkby, M.J., 2002. A cellular model of Holocene up-
land river basin and alluvial fan evolution. Earth Surf. Process. Landf. 27,
269e288.
Craine, J.M., Elmore, A.J., Aidar, M.P.M., Bustamante, M., Dawson, T.E., Hobbie, E.A.,
Kahmen, A., Mack, M.C., McLauchlan, K.K., Michelsen, A., Nardoto, G.B.,
Pardo, L.H., Penuelas, J., Reich, P.B., Schuur, E.A.G., Stock, W.D., Templer, P.H.,
Virginia, R.A., Welker, J.M., Wright, I.J., 2009. Global patterns of foliar nitrogen
isotopes and their relationships with climate, mycorrhizal fungi, foliar nutrient
concentrations, and nitrogen availability. New Phytol. 183, 980e992.
Daily, G.C., 1997. Nature's Services: Societal Dependence on Natural Ecosystems.
Island Press, Washington, DC.
Dawson, T.P., Jackson, S.T., House, J.I., Prentice, I.C., Mace, G.M., 2011. Beyond pre-
dictions: biodiversity conservation in a changing climate. Science 332, 53e58.
de Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in
integrating the concept of ecosystem services and values in landscape planning,
management and decision making. Ecol. Complex 7, 260e272.
de Groot, R.S., Wilson, M.A., Boumans, R.M.J., 2002. A typology for the classification,
description and valuation of ecosystem functions, goods and services. Ecol.
Econ. 41, 393e408.
Dearing, J.A., Håkansson, H., Liedberg-Jo € nsson, B., Persson, A., Skansjo €, S.,
Widholm, D., El-Daoushy, F., 1987. Lake sediments used to quantify the
erosional response to land use change in southern Sweden. Oikos 50, 60e78.
Dearing, J.A., Yang, X., Dong, X., Zhang, E., Chen, X., Langdon, P.G., Zhang, K.,
Zhang, W., Dawson, T.P., 2012. Extending the timescale and range of ecosystem
services through paleoenvironmental analyses, exemplified in the lower
Yangtze basin. Proc. Natl. Acad. Sci. 109, E1111eE1120.
Diaz, R.J., Rosenberg, R., 2008. Spreading dead zones and consequences for marine
ecosystems. Science 321, 926e929.
Diaz, S., Lavorel, S., de Bello, F., Quetier, F., Grigulis, K., Robson, M., 2007. Incorpo-
rating plant functional diversity effects in ecosystem service assessments. Proc.
Natl. Acad. Sci. U. S. A. 104, 20684e20689.
Dominati, E., Patterson, M., Mackay, A., 2010. A framework for classifying and
quantifying the natural capital and ecosystem services of soils. Ecol. Econ. 69,
1858e1868.
Eigenbrod, F., Armsworth, P.R., Anderson, B.J., Heinemeyer, A., Gillings, S., Roy, D.B.,
Thomas, C.D., Gaston, K.J., 2010. The impact of proxy-based methods on map-
ping the distribution of ecosystem services. J. Appl. Ecol. 47, 377e385.
Elliott, E.M., Brush, G.S., 2006. Sedimented organic nitrogen isotopes in freshwater
wetlands record long-term changes in watershed nitrogen source and land use.
Environ. Sci. Technol. 40, 2910e2916.
Extence, C.A., Balbi, D.M., Chadd, R.P., 1999. River flow indexing using British
benthic macroinvertebrates: a framework for setting hydroecological objec-
tives. Regul. Rivers Res. Manag. 15, 545e574.
Farley, K.A., Jobbagy, E.G., Jackson, R.B., 2005. Effects of afforestation on water yield:
a global synthesis with implications for policy. Glob. Change Biol. 11,
1565e1576.
Ferrio, J.P., Alonso, N., Voltas, J., Araus, J.L., 2004. Estimating grain weight in
archaeological cereal crops: a quantitative approach for comparison with cur-
rent conditions. J. Archaeol. Sci. 31, 1635e1642.
Ferrio, J.P., Araus, J.L., Buxo, R., Voltas, J., Bort, J., 2005. Water management practices
and climate in ancient agriculture: inferences from the stable isotope compo-
sition of archaeobotanical remains. Veg. Hist. Archaeobot. 14, 510e517.
Flower, A.G., Gavin, D., Heyerdahl, E.K., Parsons, R.A., Cohn, G.M., 2014. Western
spruce budworm outbreaks did not increase fire risk over the last three cen-
turies: a dendrochronological analysis of inter-disturbance synergism. Plos One
9, e114282.
Foresight, 2011. The Future of Food and Farming: Challenges and Choices for Global
Sustainability. Final Project Report. The Government Office for Science, London.
Froyd, C.A., Willis, K.J., 2008. Emerging issues in biodiversity & conservation
management: the need for a palaeoecological perspective. Quat. Sci. Rev. 27,
1723e1732.
Gaillard, M.-J., Sugita, S., Bunting, M.J., Middleton, R., Brostro €m, A., Caseldine, C.,
Giesecke, T., Hellman, S.V., Hicks, S., Hjelle, K., Langdon, C., Nielsen, A.-B.,
Poska, A., von Stedingk, H., Veski, S., 2008. The use of modelling and
simulation approach in reconstructing past landscapes from fossil pollen
data: a review and results from the POLLANDCAL network. Veg. Hist.
Archaeobot. 17, 419e443.
Gangopadhyay, S., McCabe, G.J., 2010. Predicting regime shifts in flow of the Colo-
rado River. Geophys. Res. Lett. 37, L20706.
Gangopadhyay, S., Harding, B.L., Rajagopalan, B., Lukas, J.J., Fulp, T.J., 2009.
A nonparametric approach for paleohydrologic reconstruction of annual
streamflow ensembles. Water Resour. Res. 45, W06417.
Giesecke, T., Fontana, S.L., 2008. Revisiting pollen accumulation rates from Swedish
lake sediments. Holocene 18, 293e305.
Gill, J.L., McLauchlan, K.K., Skibbe, A.M., Goring, S., Zirbel, C.R., Williams, J.W., 2013.
Linking abundances of the dung fungus Sporormiella to the density of bison:
implications for assessing grazing by megaherbivores in palaeorecords. J. Ecol.
101, 1125e1136.
Godfray, H.C.J., Beddington, J.R., Crute, I.R., Haddad, L., Lawrence, D., Muir, J.F.,
Pretty, J., Robinson, S., Thomas, S.M., Toulmin, C., 2010. Food security: the
challenge of feeding 9 billion people. Science 327, 812e818.
Goldstein, J.H., Caldarone, G., Duarte, T.K., Ennaanay, D., Hannahs, N., Mendoza, G.,
Polasky, S., Wolny, S., Daily, G.C., 2012. Integrating ecosystem-service tradeoffs
into land-use decisions. Proc. Natl. Acad. Sci. 109, 7565e7570.
Gonz alez, C., Urrego, L.E., Martinez, J.I., Polania, J., Yokoyama, Y., 2010. Mangrove
dynamics in the southwestern Caribbean since the ‘Little Ice Age’: A history of
human and natural disturbances. Holocene 20, 849e861.
Gosling, W.D., Williams, J.J., 2013. Ecosystem service provision sets the pace for pre-
Hispanic societal development in the central Andes. Holocene 23, 1619e1624.
Greenwood, M.T., Wood, P.J., Monk, W.A., 2006. The use of fossil caddisfly assem-
blages in the reconstruction of flow environments from floodplain paleo-
channels of the River Trent, England. J. Paleolimnol. 35, 747e761.
Grunwald, S., Thompson, J.A., Boettinger, J.L., 2011. Digital soil mapping and
modeling at continental scales: finding solutions for global issues. Soil Sci. Soc.
Am. J. 75, 1201e1213.
Harrington, R.R., Kennedy, B.P., Chamberlain, C.P., Blum, J.D., Folt, C.L., 1998. 15N
enrichment in agricultural catchments: field patterns and applications to
tracking Atlantic salmon (Salmo salar). Chem. Geol. 147, 281e294.
Heathwaite, A.L., 2010. Multiple stressors on water availability at global to catch-
ment scales: understanding human impact on nutrient cycles to protect water
quality and water availability in the long term. Freshw. Biol. 55, 241e257.
 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Henry, R.J., Kettlewell, P.S., 1996. Cereal Grain Quality. Chapman & Hall, London.
Hietz, P., Turner, B.L., Wanek, W., Richter, A., Nock, C.A., Wright, S.J., 2011. Long-term
change in the nitrogen cycle of tropical forests. Science 334, 664e666.
Higuera, P.E., Sprugel, D.G., Brubaker, L.B., 2005. Reconstructing fire regimes with
charcoal from small-hollow sediments: a calibration with tree-ring records of
fire. Holocene 15, 238e251.
Higuera, P.E., Whitlock, C., Gage, J.A., 2011. Linking tree-ring and sediment-charcoal
records to reconstruct fire occurrence and area burned in subalpine forests of
Yellowstone National Park, USA. Holocene 21, 327e341.
Hijmans, R.J., Graham, C.H., 2006. The ability of climate envelope models to predict
the effect of climate change on species distributions. Glob. Change Biol. 12,
2272e2281.
Hobbie, S.E., 1992. Effects of plant species on nutrient cycling. Trends Ecol. Evol. 7,
336e339.
Houlton, B.Z., Bai, E., 2009. Imprint of denitrifying bacteria on the global terrestrial
biosphere. Proc. Natl. Acad. Sci. 106, 21713e21716.
Howard, L., Wood, P., Greenwood, M., Rendell, H., 2009. Reconstructing riverine
paleo-flow regimes using subfossil insects (Coleoptera and Trichoptera): the
application of the LIFE methodology to paleochannel sediments. J. Paleolimnol.
42, 453e466.
Howard, L.C., Wood, P.J., Greenwood, M.T., Rendell, H.M., Brooks, S.J.,
Armitage, P.D., Extence, C.A., 2010. Sub-fossil Chironomidae as indicators of
palaeoflow regimes: integration into the PalaeoLIFE flow index. J. Quat. Sci.
25, 1270e1283.
Hu, F.S., Finney, B.P., Brubaker, L.B., 2001. Effects of holocene Alnus expansion on
aquatic productivity, nitrogen cycling, and soil development in southwestern
Alaska. Ecosystems 4, 358e368.
Huggett, R.J., 1998. Soil chronosequences, soil development, and soil evolution: a
critical review. CATENA 32, 155e172.
IPCC (Intergovernmental Panel on Climate Change), 2001. Climate Change 2001:
Impacts, Adaptation, and Vulnerability, Technical Summary. IPCC Secretariat,
World Meteorological Organization, Geneva, 56 pages.
Jackson, S.T., Betancourt, J.L., Booth, R.K., Gray, S.T., 2009. Ecology and the ratchet of
events: climate variability, niche dimensions, and species distributions. Proc.
Natl. Acad. Sci. U. S. A. 106, 19685e19692.
Jain, S., Woodhouse, C.A., Hoerling, M.P., 2002. Multidecadal streamflow regimes in
the interior western United States: implications for the vulnerability of water
resources. Geophys. Res. Lett. 29, 2036.
Jenny, H., 1941. Factors of Soil Formation: a System of Quantitative Pedology.
McGraw-Hill, New York; London.
Kindermann, G., Obersteiner, M., Sohngen, B., Sathaye, J., Andrasko, K.,
Rametsteiner, E., Schlamadinger, B., Wunder, S., Beach, R., 2008. Global cost
estimates of reducing carbon emissions through avoided deforestation. Proc.
Natl. Acad. Sci. U. S. A. 105, 10302e10307.
Kowalski, T., 2006. Chalara fraxinea sp. associated with dieback of ash (Fraxinus
excelsior) in Poland. For. Pathol. 36, 264e270.
Kundzewicz, Z.W., Mata, L.J., Arnell, N.W., Do € ll, P., Jimenez, B., Miller, K., Oki, T.,
ŞEn, Z., Shiklomanov, I., 2008. The implications of projected climate change for
freshwater resources and their management. Hydrol. Sci. J. 53, 3e10.
Lavelle, P., Dugdale, R., Scholes, R., Berhe, A.A., Carpenter, E., Codispoti, L., Izac, A.-M.,

guer, P., Ward, B., 2005. Nutrient cycling.
Lemoalle, J., Luizao, F., Scholes, M.C., Tre
In: Hassan, R., Scholes, R., Ash, N. (Eds.), Ecosystems and Human Well-being:
Current State and Trends. Island Press, Washington, pp. 331e353.
Lavorel, S., Grigulis, K., Lamarque, P., Colace, M.P., Garden, D., Girel, J., Pellet, G.,
Douzet, R., 2011. Using plant functional traits to understand the landscape
distribution of multiple ecosystem services. J. Ecol. 99, 135e147.
Lewis, D.B., Castellano, M.J., Kaye, J.P., 2014. Forest succession, soil carbon accu-
mulation, and rapid nitrogen storage in poorly remineralized soil organic
matter. Ecology 95, 2687e2693.
Lobell, D.B., Schlenker, W., Costa-Roberts, J., 2011. Climate trends and global crop
production since 1980. Science 333, 616e620.
Luo, Y., Su, B.O., Currie, W.S., Dukes, J.S., Finzi, A., Hartwig, U., Hungate, B., Mc
Murtrie, R.E., Oren, R.A.M., Parton, W.J., Pataki, D.E., Shaw, M.R., Zak, D.R.,
Field, C.B., 2004. Progressive nitrogen limitation of ecosystem responses to
rising atmospheric carbon dioxide. Bioscience 54, 731e739.
Mace, G.M., Norris, K., Fitter, A.H., 2012. Biodiversity and ecosystem services: a
multilayered relationship. Trends Ecol. Evol. 27, 19e26.
Macias-Fauria, M., Willis, K.J., 2013. Landscape planning for the future: using fossil
records to independently validate bioclimatic envelope models for economi-
cally valuable tree species in Europe. Glob. Ecol. Biogeogr. 22, 318e333.
Macias Fauria, M., Johnson, E.A., 2008. Climate and wildfires in the North American
boreal forest. Philos. Trans. R. Soc. B Biol. Sci. 363, 2315e2327.
MacLean, D.A., 1990. Impact of forest pests and fire on stand growth and timber
yield: implications for forest management planning. Can. J. For. Res. 20,
391e404.
MacLean, D.A., Erdle, T.A., MacKinnon, W.E., Porter, K.B., Beaton, K.P., Cormier, G.,
Morehouse, S., Budd, M., 2001. The Spruce Budworm Decision Support System:
forest protection planning to sustain long-term wood supply. Can. J. For. Res. 31,
1742e1757.
Magri, D., 1994. Late-quaternary changes of plant biomass as recorded by pollen-
stratigraphical data: a discussion of the problem at Valle di Castiglione, Italy.
Rev. Palaeobot. Palynol. 81, 313e325.
McCann, R.K., Marcot, B.G., Ellis, R., 2006. Bayesian belief networks: applications in
ecology and natural resource management. Can. J. For. Res. Rev. Can. Rech. For.
36, 3053e3062.
31
McLauchlan, K.K., Craine, J.M., Oswald, W.W., Leavitt, P.R., Likens, G.E., 2007.
Changes in nitrogen cycling during the past century in a northern hardwood
forest. Proc. Natl. Acad. Sci. U. S. A. 104, 7466e7470.
McLauchlan, K.K., Ferguson, C.J., Wilson, I.E., Ocheltree, T.W., Craine, J.M., 2010.
Thirteen decades of foliar isotopes indicate declining nitrogen availability in
central North American grasslands. New Phytol. 187, 1135e1145.
McLauchlan, K.K., Williams, J.J., Engstrom, D.R., 2013a. Nutrient cycling in the
palaeorecord: fluxes from terrestrial to aquatic ecosystems. Holocene 23,
1635e1643.
McLauchlan, K.K., Williams, J.J., Craine, J.M., Jeffers, E.S., 2013b. Changes in global
nitrogen cycling during the Holocene epoch. Nature 495, 352e355.
McLauchlan, K.K., Higuera, P.E., Gavin, D.G., Perakis, S.S., Mack, M.C., Alexander, H.,
Battles, J., Biondi, F., Buma, B., Colombaroli, D., Enders, S.K., Engstrom, D.R.,
Hu, F.S., Marlon, J.R., Marshall, J., McGlone, M., Morris, J.L., Nave, L.E., Shuman, B.,
Smithwick, E.A.H., Urrego, D.H., Wardle, D.A., Williams, C.J., Williams, J.J., 2014.
Reconstructing disturbances and their biogeochemical consequences over
multiple timescales. Bioscience 64, 105e116.
McMahon, S.M., Harrison, S.P., Armbruster, W.S., Bartlein, P.J., Beale, C.M.,
Edwards, M.E., Kattge, J., Midgley, G., Morin, X., Prentice, I.C., 2011. Improving
assessment and modelling of climate change impacts on global terrestrial
biodiversity. Trends Ecol. Evol. 26, 249e259.
McNally, C.G., Uchida, E., Gold, A.J., 2011. The effect of a protected area on the
tradeoffs between short-run and long-run benefits from mangrove ecosystems.
Proc. Natl. Acad. Sci. U. S. A. 108, 13945e13950.
Metsaranta, J.M., Lieffers, V.J., 2009. Using dendrochronology to obtain annual data
for modelling stand development: a supplement to permanent sample plots.
Forestry 82, 163e173.
Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being:
Current State and Trends. Island Press, Washington, DC, 917 pp.
Monk, W.A., Wood, P.J., Hannah, D.M., Wilson, D.A., Extence, C.A., Chadd, R.P., 2006.
Flow variability and macroinvertebrate community response within riverine
systems. River Res. Appl. 22, 595e615.
Mooney, H., Larigauderie, A., Cesario, M., Elmquist, T., Hoegh-Guldberg, O.,
Lavorel, S., Mace, G.M., Palmer, M., Scholes, R., Yahara, T., 2009. Biodiversity,
climate change, and ecosystem services. Curr. Opin. Environ. Sustain. 1,
46e54.
Morris, J.L., Brunelle, A.R., 2012. Pollen accumulation in lake sediments during
historic spruce beetle disturbances in subalpine forests of southern Utah, USA.
Holocene 22, 961e974.
Morris, J.L., Brunelle, A.R., Munson, A.S., 2010. Pollen evidence of historical
forest disturbance on the Wasatch Plateau, Utah. West. North Am. Nat. 70,
175e188.
Morris, J.L., le Roux, P.C., Macharia, A.N., Brunelle, A., Hebertson, E.G., Lundeen, Z.J.,
2013. Organic, elemental, and geochemical contributions to lake sediment de-
posits during severe spruce beetle (Dendroctonus rufipennis) disturbances. For.
Ecol. Manag. 289, 78e89.
Naidoo, R., Balmford, A., Costanza, R., Fisher, B., Green, R.E., Lehner, B., Malcolm, T.R.,
Ricketts, T.H., 2008. Global mapping of ecosystem services and conservation
priorities. Proc. Natl. Acad. Sci. 105, 9495e9500.
Nave, L.E., Vance, E.D., Swanston, C.W., Curtis, P.S., 2011. Fire effects on temperate
forest soil C and N storage. Ecol. Appl. 21, 1189e1201.
Nelson, G.C., Bennett, E., Berhe, A.A., Cassman, K.G., DeFries, R., Dietz, T., Dobson, A.,
Dobermann, A., Janetos, A., Levy, M., Marco, D., Nakicenovic, N., O'Neill, B.,
Norgaard, R., Petschel-Held, G., Ojima, D., Pingali, P.L., Watson, R., Zurek, M.B.,
2005. Drivers of change in ecosystem condition and services. In: Carpenter, S.R.,
Pingali, P.L., Bennett, E., Zurek, M.B. (Eds.), Ecosystems and Human Well-being:
Scenarios. Island Press, Washington, pp. 173e222.
Overballe-Petersen, M.V., Nielsen, A.B., Bradshaw, R.H.W., 2013. Quantitative
vegetation reconstruction from pollen analysis and historical inventory data
around a Danish small forest hollow. J. Veg. Sci. 24, 755e771.
Paciorek, C.J., McLachlan, J.S., 2009. Mapping ancient forests: Bayesian inference for
spatio-temporal trends in forest composition using the fossil pollen proxy re-
cord. J. Am. Stat. Assoc. 104, 608e622.
Palmer, M.A., Filoso, S., 2009. Restoration of ecosystem services for environmental
markets. Science 325, 575e576.
Pan, Y., Birdsey, R.A., Fang, J., Houghton, R., Kauppi, P.E., Kurz, W.A., Phillips, O.L.,
Shvidenko, A., Lewis, S.L., Canadell, J.G., Ciais, P., Jackson, R.B., Pacala, S.W.,
McGuire, A.D., Piao, S., Rautiainen, A., Sitch, S., Hayes, D., 2011. A Large and
Persistent Carbon Sink in the World's Forests. Science 333, 988e993.
Peng, C.H., Guiot, J., van Campos, E., 1998. Estimating changes in terrestrial vege-
tation and carbon storage: using palaeoecological data and models. Quat. Sci.
Rev. 17, 719e735.
Pennington, W., Tutin, M.T.G., Lishman, J.P., 1971. Iodine in lake sediments in
northern England and Scotland. Biol. Rev. 46, 279e313.
Pennington, W., 1986. Lags in adjustment of vegetation to climate caused by the
pace of soil development - evidence from Britain. Vegetatio 67, 105e118.
Pereira, H.M., Leadley, P.W., Proenca, V., Alkemade, R., Scharlemann, J.P.W., Fer-
nandez-Manjarres, J.F., Araujo, M.B., Balvanera, P., Biggs, R., Cheung, W.W.L.,
Chini, L., Cooper, H.D., Gilman, E.L., Guenette, S., Hurtt, G.C., Huntington, H.P.,
Mace, G.M., Oberdorff, T., Revenga, C., Rodrigues, P., Scholes, R.J., Sumaila, U.R.,
Walpole, M., 2010. Scenarios for global biodiversity in the 21st Century. Science
330, 1496e1501.
Power, M.J., Marlon, J.R., Bartlein, P.J., Harrison, S.P., 2010. Fire history and the Global
Charcoal Database: a new tool for hypothesis testing and data exploration.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 291, 52e59.
32 E.S. Jeffers et al. / Quaternary Science Reviews 112 (2015) 17e32
Prentice, I.C., Harrison, S.P., 2009. Ecosystem effects of CO2 concentration: evidence
from past climates. Clim. Past 5, 297e307.
Prentice, I.C., Harrison, S.P., Bartlein, P.J., 2011. Global vegetation and terrestrial
carbon cycle changes after the last ice age. New Phytol. 189, 988e998.
Prentice, I.C., Jolly, D., participants, B., 2000. Mid-Holocene and glacial-maximum
vegetation geography of the northern continents and Africa. J. Biogeogr. 27,
507e519.
Raffa, K.F., Aukema, B.H., Bentz, B.J., Carroll, A.L., Hicke, J.A., Turner, M.G.,
Romme, W.H., 2008. Cross-scale drivers of natural disturbances prone to
anthropogenic amplification: the dynamics of bark beetle eruptions. Bioscience
58, 501e517.
Reich, P.B., Hobbie, S.E., Lee, T., Ellsworth, D.S., West, J.B., Tilman, D., Knops, J.M.H.,
Naeem, S., Trost, J., 2006. Nitrogen limitation constrains sustainability of
ecosystem response to CO2. Nature 440, 922e925.
Rozendaal, D.M.A., Zuidema, P.A., 2011. Dendroecology in the tropics: a review.
Trees-Struct. Funct. 25, 3e16.
Sampson, R.N., Bystriakova, N., Brown, S., Gonzalez, P., Irland, L.C., Kauppi, P.,
Sedjo, R., Thompson, I.D., 2005. Timber, fuel and fiber. In: Hassan, R., Scholes, R.,
Ash, N. (Eds.), Ecosystems and Human Well-being: Current State and Trends.
Island Press, Washington, D.C, pp. 243e269.
Sedjo, R., Sohngen, B., 2012. Carbon sequestration in forests and soils. Annu. Rev.
Resour. Econ. 4, 127e144.
Seppa€, H., Alenius, T., Muukkonen, P., Giesecke, T., Miller, P.A., Ojala, A.E.K., 2009.
Calibrated pollen accumulation rates as a basis for quantitative tree biomass
reconstructions. Holocene 19, 209e220.
Simard, I., Morin, H., Lavoie, C., 2006. A millennial-scale reconstruction of spruce
budworm abundance in Saguenay, Quebec, Canada. Holocene 16, 31e37.
Simard, I., Morin, H., Potelle, B., 2002. A new paleoecological approach to recon-
struct long-term history of spruce budworm outbreaks. Can. J. For. Res. Rev. Can.
Rech. For. 32, 428e438.
Smith, D., Nayyar, K., Schreve, D., Thomas, R., Whitehouse, N., 2014. Can dung
beetles from the palaeoecological and archaeological record indicate herd
concentration and the identity of herbivores? Quat. Int. 341, 119e130.
Sugita, S., 2007a. Theory of quantitative reconstruction of vegetation I: pollen from
large sites REVEALS regional vegetation composition. Holocene 17, 229e241.
Sugita, S., 2007b. Theory of quantitative reconstruction of vegetation II: all you need
is LOVE. Holocene 17, 243e257.
Sutton, M.A., 2011. The European Nitrogen Assessment: Sources, Effects and Policy
Perspectives. Cambridge Unversity Press, Cambridge.
Swetnam, T.W., Thompson, M.A., Sutherland, E.K., 1985. Spruce Budworms Hand-
book: Using Dendrochronology to Measure Radial Growth of Defoliated Trees.
US Department of Agriculture. Agricultural Handbook No. 639, 39 pages.
Swetnam, T.W., Betancourt, J.L., 1998. Mesoscale disturbance and ecological
response to decadal climatic variability in the American Southwest. J. Clim. 11,
3128e3147.
Swetnam, T.W., Lynch, A.M., 1993. Multicentury, regional-scale patterns of western
spruce budworm outbreaks. Ecol. Monogr. 63, 399e424.
Therrell, M.D., Stahle, D.W., Mukelabai, M.M., Shugart, H.H., 2007. Age, and radial
growth dynamics of Pterocarpus angolensis in southern Africa. For. Ecol. Manag.
244, 24e31.
Turner, M.G., Hargrove, W.W., Gardner, R.H., Roe, W.H., 1994. Effects of fire on
landscape heterogeneity in Yellowstone National Park, Wyoming. J. Veg. Sci. 5,
731e742.
Tylianakis, J.M., Didham, R.K., Bascompte, J., Wardle, D.A., 2008. Global change and
species interactions in terrestrial ecosystems. Ecol. Lett. 11, 1351e1363.
UK National Ecosystem Assessment, World Conservation Monitoring Centre, 2011.
UK National Ecosystem Assessment: Technical Report. United Nations Envi-
ronment Programme World Conservation Monitoring Centre, Cambridge.
Vorosmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water resources:
vulnerability from climate change and population growth. Science 289,
284e288.
Walker, L.R., Wardle, D.A., 2014. Plant succession as an integrator of contrasting
ecological time scales. Trends Ecol. Evol. 29, 504e510.
Walters, B.B., Ronnback, P., Kovacs, J.M., Crona, B., Hussain, S.A., Badola, R., Primavera, J.H.,
Barbier, E., Dahdouh-Guebas, F., 2008. Ethnobiology, socio-economics and man-
agement of mangrove forests: a review. Aquat. Bot. 89, 220e236.
Wardle, D.A., Bardgett, R.D., Klironomos, J.N., Setala, H., van der Putten, W.H.,
Wall, D.H., 2004. Ecological linkages between aboveground and belowground
biota. Science 304, 1629e1633.
Welsh, K.E., Dearing, J.A., Chiverrell, R.C., Coulthard, T.J., 2009. Testing a cellular
modelling approach to simulating late-Holocene sediment and water transfer
from catchment to lake in the French Alps since 1826. Holocene 19, 785e798.
Westerling, A.L., Hidalgo, H.G., Cayan, D.R., Swetnam, T.W., 2006. Warming and
earlier spring increase western U.S. forest wildfire activity. Science 313,
940e943.
Whitehouse, N.J., Kirleis, W., 2014. The world reshaped: practices and impacts of
early agrarian societies. J. Archaeol. Sci. 51, 1e11.
Willis, K.J., Birks, H.J.B., 2006. What is natural? The need for a long-term perspective
in biodiversity conservation. Science 314, 1261e1265.
Willis, K.J., Bailey, R.M., Bhagwat, S.A., Birks, H.J.B., 2010. Biodiversity baselines,
thresholds and resilience: testing predictions and assumptions using palae-
oecological data. Trends Ecol. Evol. 25, 583e591.
Willis, K.J., Braun, M., Sumegi, P., Toth, A., 1997. Does soil change cause vegetation
change or vice versa? A temporal perspective from Hungary. Ecology 78, 740e750.
Wolfe, A.P., Hobbs, W.O., Birks, H.H., Briner, J.P., Holmgren, S.U., Ingo


lfsson, O.,
Kaushal, S.S., Miller, G.H., Pagani, M., Saros, J.E., Vinebrooke, R.D., 2013. Strati-
graphic expressions of the HoloceneeAnthropocene transition revealed in
sediments from remote lakes. Earth-Sci. Rev. 116, 17e34.
Zaehle, S., 2013. Terrestrial nitrogenecarbon cycle interactions at the global scale.
Philos. Trans. R. Soc. B Biol. Sci. 368, 1621.