Download Conservation of phylogeographic lineages under climate change

Global Ecology and Biogeography, (Global Ecol. Biogeogr.) (2013) 22, 93–104
bs_bs_banner
RESEARCH
PA P E R
Conservation of phylogeographic
lineages under climate change
geb_774
93..104
Manuela D’Amen*, Niklaus E. Zimmermann and Peter B. Pearman
Landscape Dynamics Unit, Swiss Federal
Research Institute, WSL, Zuercherstrasse 111,
CH-8903 Birmensdorf, Switzerland
ABSTRACT
Aim We address the unexplored question of whether the lack of information on
intra-specific diversity inherent in species-level niche modelling might bias evaluation of the conservation requirements of species and phylogeographic lineages
under changing climates. We test for directional biases that might arise due to these
methodological differences in ways of assessing risks from climate change.
Location The African continent.
Methods We identified from peer-reviewed studies that used both nuclear and
plastid markers the distribution of deep phylogeographic divisions within nine
species of African mammals and their phylogeographic lineages. We fitted ecological niche models to describe currently suitable, occupied climates and to project the
shift of suitable climate to two future time slices. We applied gap analysis to reveal
potential changes in the protection of phylogeographic diversity owing to climatic
shifts.
Results We found that, within species, most phylogeographic lineages differ in the
climates they experience and have substantial geographic separation. Models that
do not distinguish these subspecific units often fail to identify potential risks of
climate change to lineages. Modelled potential effects of climate change on the
geographic extent of suitable climate vary in both direction and magnitude. Predictions of the persistence of suitable climate in current protected areas for the
resident lineages differ on average by factor of 2 between species and lineage
models.
Main conclusions Our study develops an original synthetic approach by combining niche modelling, projected climate change, phylogeographic information and
gap analysis. We clearly identify the potential benefits of using the new approach to
evaluate risks to the conservation of intra-specific genetic diversity that are posed by
climate change. Our results suggest that prudent conservation strategies need to
incorporate potential differences in climate tolerance among lineages when planning
conservation measures for species confronted with environmental change.
*Correspondence: Manuela D’Amen, Landscape
Dynamics Unit, WSL, Zuercherstrasse 111,
CH-8903 Birmensdorf, Switzerland.
E-mail: [email protected]
Keywords
Africa, biodiversity, gap analysis, genetic diversity, global warming, mammals,
phylogeography, protected areas, species distribution models.
INTRODUCTION
The conservation of species is a fundamental goal of national
and international programmes that aim to stop the ongoing loss
of biodiversity (Ennos et al., 2005). This is because species constitute a recognizable currency for developing and evaluating
conservation policy and management activities (Mace, 2004).
© 2012 Blackwell Publishing Ltd
Nonetheless, the frequent discovery of phylogeographic lineages
within species motivates and justifies a reconsideration of strategies for conserving intra-specific diversity (Crandall et al.,
2000). Such conservation efforts are crucial for the preservation
of the historical components of diversity and to maintain the
potential for species to mount adaptive responses to environmental change. The existence of phylogeographic lineages in
DOI: 10.1111/j.1466-8238.2012.00774.x
http://wileyonlinelibrary.com/journal/geb
93
M. D’Amen et al.
widely distributed species has additional implications for conservation planning because of changing global climate.
Future changes in climates are likely to affect species ranges,
the distribution of biodiversity and the evolutionary history it
holds (Thuiller et al., 2011). Empirical evidence demonstrates
that ongoing range shifts in response to changing climate are
occurring globally (e.g. Parmesan & Yohe, 2003). These shifts
can be successfully predicted (Araújo et al., 2005) by ecological
niche models (ENMs), which constitute an important tool for
projecting potential shifts in the distributions of suitable conditions for species (Guisan & Zimmermann, 2000), under the
assumption that the species’ ecological niche does not evolve to
meet changing conditions (Peterson, 2011). These statistical distribution models generally focus on species and incorporate no
information on the spatial distribution of intra-specific genetic
variability.
Climate change and its effects on the geographic distribution of suitable climate could challenge the future effectiveness
of national parks and other biological reserves as conservation
tools for preserving biodiversity (Hannah et al., 2002). While
environmental tolerances or local adaptation might allow some
species to persist in parts of their current ranges, other species,
even if currently protected, will survive only by colonizing
newly suitable areas (Araújo et al., 2004). Thus, research needs
to evaluate the long-term effectiveness of current protected
areas and to establish priorities for providing connectivity of
reserve networks. Some studies suggest a reduction of the
long-term efficacy of current reserves under climate change
(e.g. D’Amen et al., 2011), but these have focused only on the
species level and have ignored intra-specific lineage structure.
Despite the importance of preserving genetic diversity, no
studies have explored how species-level niche modelling could
bias evaluation of the conservation requirements of species
and intra-specific phylogeographic lineages under conditions
of changing climate.
The incorporation of the structure of phylogeographic lineages when modelling the impacts of climate change might
improve our understanding of the potential risks faced by
species exposed to climate change. The omission of phylogeographic structure from niche modelling exercises can lead to
some lineages having little representation in the resulting species
models (Pearman et al., 2010). This can, on the one hand, lead to
underestimation of the climate tolerances of species and overestimation of the effects of climate change on the potential
future distribution of species. On the other hand, climate change
could threaten particular lineages, but these threats could go
undetected. The integration of intra-specific variability into the
projection of the potential effects of climate change on species
could better identify future conservation needs that are specific
to particular lineages in geographically restricted portions of a
species’ range.
This study arises from the focus in Pearman et al. (2010) on
incorporating phylogeographic information in predictive
models for species, and aims to assess risks to lineages that are
posed by climate change and the resulting implications for conservation planning. Here, we address the unexplored question:
94
do continental-scale assessments of the future effectiveness of
protected areas, based on the projection of ENMs that are built
with species-level occurrence data, mask risks to particular phylogeographic lineages? We focus on several mammal species in
Africa, selected because of the availability of robust phylogeographic information. We fit ENMs for species and their component phylogeographic lineages to project the shift of suitable
climate at two future points in time. In particular, we are interested in the geographic distribution of the climate that is currently occupied by lineages and how this distribution will
change in the future. We then apply gap analysis to document
potential changes in the protection of phylogeographic diversity
in the current reserve network owing to climatic shifts (other
potential impacts on future species ranges, such as land-use/
land-cover change, are beyond the scope of this paper). We
additionally focus on two related issues regarding the estimation
of range change and reserve representation: (1) the evaluation at
the species level of the implications of modelling approaches
that either include or ignore lineage structure; (2) the effect on
assessment at the lineage level of projections derived by ENMs
that are built with species-level data versus the use of independent niche models for each lineage. To our knowledge no previous study has addressed this fundamental issue for preserving
genetic diversity.
M AT E R I A L S A N D M E T H O D S
Species data
We address potential effects of climate change on several species
of large mammal inhabiting the African continent. We surveyed
published data from multi-genome phylogeographic analyses
because the identification of phylogeographic lineages with
mitochondrial DNA sequences alone can cause substantial bias
when inferring the evolutionary history of species (Ballard &
Whitlock, 2004). We excluded from consideration those studies
that did not report clear definition of intra-specific divisions,
and those that did not consider substantial parts of species
ranges in Africa. As a result, we selected 10 phylogeographic
studies (Table 1). We used species and infra-specific nomenclature according to the International Union for Conservation of
Nature (IUCN). When infra-specific nomenclature did not correspond to documented phylogeographic units or was nonexistent, we named lineages based on their geographic
distributions.
We derived species ranges from the digital distribution maps
of the IUCN Red List of threatened species (IUCN, 2010). These
data were available as shapefile polygons and contained the most
up to date (October 2010) known range of species. For conservation reasons, the IUCN provided only relevant national
boundaries to describe range limits for the black rhinoceros.
This lack of detail led us to exclude this species from further
analyses. We used the IUCN polygons to select occupied cells in
a grid of 1° resolution and encompassing the whole African
continent (2642 cells). For each species, we considered as
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
African mammals and global warming
occupied all the cells that intersected the relevant polygons. The
use of relatively large grid cells probably reduced errors of false
presence in the data set. We defined distribution boundaries for
the phylogeographic lineages according to the information contained in the IUCN database, when available, otherwise we got
this information from relevant publications for the target
species (Table 1, see Fig. S1 in Supporting Information). We
excluded from the analysis three lineages that occupied ranges of
Table 1 Mammals’ species selected on
the basis of the existing literature on
phylogeographic structure.
ID
< 20 1° cells because ENMs that are built with small data sets can
be unstable to the omission or inclusion of single occurrences
(Stockwell & Peterson, 2002).
Climatic predictors
We obtained current bioclimatic layers at a resolution of 30″
from the WorldClim database, which contains 19 bioclimatic
Species
Intra-specific lineages
IUCN status
Reference
1
Acinonyx jubatus
Aepyceros melampus
3
Diceros bicornis†
4
Equus zebra
5
Giraffa camelopardalis
6
Hippotragus equinus
7
Kobus ellipsiprymnus
8
Kobus kob
9
Loxodonta africana
VU
CR
–
–
–
LC
–
–
–
CR
CR
VU
CR
CR
VU
VU
VU
LC
EN
–
–
–
–
–
EN
LC
–
–
LC
LC
NT
LC
–
–
VU
–
–
LC
–
–
–
–
Charruau et al. (2011)
2
–
A. j. venaticus
A. j. soemmerringii
A. j. raineyi
A. j. jubatus
–
E Africa unit
S Africa unit
SW Africa unit*
–
D. b. longipes
D. b. bicornis
D. b. michaeli
D. b. minor
–
E. z. hartmannae
E. z. zebra
–
G. c. rothschildi
G. c. reticulata
G. c. tippelskirchi
G. c. giraffa
G. c. angolensis
G. c. antiquorum
G. c. peralta*
–
W Africa unit
SE Africa unit
–
K. e. ellipsiprymnus
K. e. defassa
–
E Africa unit
WC Africa unit
–
L. a. cyclotis
L. a. africana
–
P. a. africanus
P. a. sundevallii
P. a. massaicus
P. a. aeliani*
10
Phacochoerus africanus
Lorenzen et al. (2006)
Harley et al. (2005)
Moodley & Harley (2005)
Brown et al. (2007)
Alpers et al. (2004)
Lorenzen et al. (2006)
Lorenzen et al. (2007)
Rohland et al. (2010)
Muwanika et al. (2003)
VU, Vulnerable; CR, Critically Endangered; LC, Least Concern; EN, Endangered; NT, Near
Threatened.
*Lineages not considered in the analyses due to their restricted distribution (i.e. fewer than 20
occurrence cells).
†Species not included in the analyses because of the unavailability of distribution data.
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
95
M. D’Amen et al.
variables that are derived from averaged monthly precipitation
and temperature for the period 1950–2000 (Hijmans et al.,
2005). Potential values for these variables at two future time
points (2050 and 2080) were derived from climatic grids of 10′
resolution from the HadCM3 global circulation model for the
SRES A1b emission scenario (Ramirez & Jarvis, 2008). Because
of the demonstrative purpose of this work, we use only one of
the several available climate models as a case study. In order to
match the resolution of the species and lineage distribution
grids, we rescaled all climatic layers to 1° resolution by averaging
the climate values within each map cell. We avoided the effect of
inter-correlation among independent variables by calculating
the variance inflation factor (VIF) for the 19 bioclimatic grids.
We removed variables until all VIFs were below 5 (Belsley et al.,
1980) and, as a result, retained six predictors (Table S1).
Ecological niche models
We modelled the climatic niche using the generalized boosted
model algorithm (GBM; Friedman et al., 2000), a topperforming algorithm for predicting species distributions (Elith
et al., 2006). For each species and lineage, we fitted GBMs with
the ‘gbm’ library in R 2.12.1 (Ridgeway, 2006; Elith et al., 2008; R
Development Core Team, 2010). Model fit was evaluated using
10-fold cross-validated area under the curve (AUC) scores
(Fielding & Bell, 1997). We fitted climatic models representing
the realized niche of each lineage (ENM-L, ecological niche
models for lineage), using as presences the occupied 1° cells that
intersected lineage boundaries and all other cells as (assumed)
true absences. We built models for species by applying three
approaches: The ecological niche models for species (ENM-S)
consisted of simply training a GBM model on all cells occupied
by the species (i.e. the occurrences pooled across all relevant
lineages). A composite model (ENM-C) was developed from the
ENM-L predictions to summarize the niche across all the lineages. To build the ENM-C, we first standardized the distribution of average probability of occurrence for each lineage along
the environmental gradients considered in the modelling procedure to make the models comparable. Then, we calculated the
mean probability of occurrence of at least one of the related
lineages for the map cells using the multiplicative probability
method described in Pearman et al. (2010). We constructed
binary presence–absence maps by implementing a threshold for
species presence that maximized the true skill statistic (TSS;
Allouche et al., 2006). Finally, we built an aggregated model
(ENM-A) that combined binary projections of the ENM-Ls of
each species, so that we assigned species presence to each cell
that was predicted suitable for at least one lineage (Fig. 1).
We predicted changes in the number of climatically suitable
cells for species and lineages using each modelling approach. We
used these projections to indicate how species and lineage ranges
could shift in response to climate change under an assumption
of unlimited dispersal. In addition, we calculated the future
persistence of climatically suitable conditions within the current
geographic ranges of species according to the three approaches
to constructing species models. We also estimated the propor96
tion of the current geographic range of lineages that could
remain climatically suitable in the future according to projections derived from ENM-Ls (no dispersal for lineages) and from
the three species models (i.e. prediction of suitable climate for
the species as a whole in the areas currently occupied by the
lineages).
Reserves data set and gap analysis
We obtained data on the location of existing protected areas in
Africa from the World Database on Protected Areas 2010 (IUCN
and UNEP-WCMC, 2010), which includes data on reserve
extent, location and classification of management designation.
We calculated the proportion of area protected in each 1° cell.
On the basis of this information, we performed a gap analysis by
calculating the percentage of current, modelled distribution that
intersects the current reserve system for each species and lineage
(Jennings, 2000). We then evaluated for 2050 and 2080 the
potential change in the proportion of area protected within the
areas of modelled suitable climate. We did this for each species
and lineage according to projections derived from different
models and dispersal assumptions. First, we projected potential
future ranges from ENM-Ls, ENM-Ss, ENM-Cs and ENM-As
under an unlimited dispersal scenario. Second, we assessed the
degree of persistence of climatic suitability for species and lineages within the currently protected portions of their distributions. For species, we based this estimation on projections of the
persistence of suitable conditions, as derived from each of the
three species models. For lineages, we used both ENM-L projections (i.e. to indicate the proportion of protected area that is
projected to remain climatically suitable specifically for the considered lineage) and species models. By basing the analysis of
lineage risks to climate change on the species models, we
obtained an estimation of currently suitable, protected area that
remains so in the future if one does not distinguish for which
lineage the climate will actually remain suitable. This allowed us
to represent an assumption, which accompanies the use of ecological niche models, of unlimited dispersal of lineages among
the remaining areas of suitable climate for the species.
R E S U LT S
Projected species models
We observe that distribution models for both species and lineages display high values of cross-validated AUC, ranging
between 0.89 and 0.99 (Table S2). Under an unlimited dispersal
scenario, changes in the extent of suitable climatic conditions
(i.e. potential range size) as ascertained from species models
(ENM-S, ENM-C, ENM-A) are similar in both future time
slices, with the exception of Hippotragus equinus (Tables 2 & S3).
Generally, the inclusion of lineage information does not significantly affect predictions of change in range size at species level
(Wilcoxon matched pairs test, n = 9; Tables 2 & S3). For approximately half of the species, the species models agree in forecasting
a substantial reduction in the number of climatically suitable
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
African mammals and global warming
Species level
Lineages level
Species
distribution
Phylogeographic
analysis
Lineages
distribution
GBM calibration
ENM-S
ENMs-L
Current
climate
2050
climate
2080
climate
Projections
ENM-S
ENMs-L
ENM-C
ENM-A
Persistence of suitable climate in
Unlimited dispersal
Persistence of suitable climate in
Unlimited dispersal
Current range
Range change
Current range
Range change
Current protected
range
Protected
range change
Current protected
range
Protected
range change
(9 species x 3 models
x 3 time slices)
(9 species x 3 models
x 3 time slices)
(25 lineages x 4 models
x 3 time slices)
(25 lineages x 1 model
x 3 time slices)
Figure 1 Flow chart of the modelling design. ENM, ecological niche model: for lineage (-L), for species (-S), composite (-C), and
aggregate (-A).
cells, reaching up to -90% for Equus zebra (Tables 2 & S3). Only
small changes (< 10%) are projected for three species, but for
Kobus kob all species models project a substantial increase in
extent of area with suitable climate by 2080 (Table 2). In future
time slices ENM-Ss, ENM-Cs and ENM-As agree in projecting
the persistence of some areas of suitable climate within the
current range of every species (Table S4).
eages differ in the magnitude of this variation (Tables 2 & S3).
For about one-third of the study lineages, we find that ENM-Ls
project the persistence of suitable climatic conditions by 2080 in
< 20% of the currently occupied lineage range. In particular, we
find that ENM-Ls predict the retention of suitable climate in no
cells in the current known range of two giraffe lineages
(Table S5).
Projected lineage models
Model bias
The geographic overlap in the distribution of suitable climate
for conspecific lineages under the current conditions resulting
from ENM-L projections of most species is very small, and
mainly limited to distribution borders (Figs 2 & S2). The consistent differences among conspecific lineages in the distribution
of suitable climatic conditions may indicate a potential for differential responses of lineages to climate change. Indeed, in
future time slices, ENM-L projections of the distribution of
suitable climate show contrasting intra-specific responses for
more than half of the species (Tables 2 & S3). Even when projections of species and lineage models agree in the sign of the
trend in extent of suitable climate conditions, conspecific lin-
Consideration of ENM-Ls when projected to future climates
reveals a bias of species models in predicting lineage response to
climate change (i.e. a discrepancy in terms of over- or underestimation of lineage range change). For six out of nine species
(i.e. Acinonyx jubatus, Giraffa camelopardalis, Hippotragus
equinus, Kobus ellipsiprymnus, Loxodonta africana, Phacochoerus
africanus) ENM-Ls predict trends in the extent of future suitable
climate differing in sign and magnitude from those obtained
from the projection of the corresponding ENM-Ss (Figs 2 & S2,
Tables 2 & S3). Moreover, ENM-Ls predict a median percentage
area of persistent suitable conditions within the current geographic range of lineages that is smaller than predicted by the
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
97
M. D’Amen et al.
Table 2 Percentage change (D) in areal extent and in protected area with suitable climatic conditions by 2080 under the assumption of
unlimited dispersal.
Species models
D Areal extent
D Protected area
ENM-L
D Areal extent
D Protected area
Acinonyx jubatus
ENM-S
ENM-C
ENM-A
–
-51.48
-38.71
-33.09
–
-70.72
-73.50
-71.73
Aepyceros melampus
ENM-S
ENM-C
ENM-A
Equus zebra
ENM-S
ENM-C
ENM-A
Giraffa camelopardalis
ENM-S
ENM-C
ENM-A
–
-44.78
-47.95
-45.84
–
-93.75
-91.30
-90.18
–
-44.84
-52.38
-37.08
–
-61.83
-67.05
-62.08
–
-94.86
-94.57
-86.70
–
-67.13
-75.15
-66.14
–
A. j. venaticus
A. j. soemmerringii
A. j. raineyi
A. j. jubatus
–
E Africa unit
S Africa unit
–
-22.65
9.09
-62.10
-42.65
–
-36.16
-49.16
–
-59.09
-17.63
-80.53
-80.38
–
-44.82
-73.75
–
E. z. hartmannae
E. z. zebra
–
-88.33
-92.73
–
-84.49
-95.08
–
–
Hippotragus equinus
ENM-S
ENM-C
ENM-A
Kobus ellipsiprymnus
ENM-S
ENM-C
ENM-A
Kobus kob
ENM-S
ENM-C
ENM-A
Loxodonta africana
ENM-S
ENM-C
ENM-A
Phacochoerus africanus
ENM-S
ENM-C
ENM-A
–
-21.89
10.85
-0.84
–
-0.10
-6.41
-8.47
–
64.55
79.75
89.80
–
11.76
8.38
9.86
–
1.00
-2.60
-3.45
–
-38.12
-9.45
-18.60
–
-9.20
-15.94
-17.00
–
65.08
72.75
78.15
–
-8.22
-5.46
-4.35
–
-2.11
-8.28
-10.97
–
G. c. rothschildi
G. c. reticulata
G. c. tippelskirchi
G. c. giraffa
G. c. angolensis
G. c. antiquorum
–
W Africa unit
SE Africa unit
36.96
77.01
-4.08
-10.77
-96.47
-59.7
–
206.78
-43.14
20.16
-10.47
-89.85
-78.65
-99.72
-68.57
–
207.03
-58.51
–
K. e. defassa
K. e. ellipsiprymnus
–
-21.03
45.36
–
–
E Africa unit
WC Africa unit
–
125.86
87.17
–
–
L. a. cyclotis
L. a. africana
–
-7.71
6.72
–
-13.00
-8.76
–
P. a. africanus
P. a. sundevallii
P. a. massaicus
–
–
72.56
-58.92
-46.24
66.76
-68.23
-49.20
10.92
-21.04
83.95
75.98
ENM-L, ecological niche models for lineages; ENM-S, ecological niche models for species; ENM-C, composite model; ENM-A, aggregated model.
corresponding species models by almost a factor of 2 (Fig. 3a).
These differences between the predictions of species models and
ENM-Ls are statistically significant for both 2050 and 2080 after
Bonferroni correction (Table S6).
Future climate suitability in protected areas
Gap analyses of all species models predict that by the year 2050
all but two species (Kobus kob and L. africana) will experience a
decrease in extent of protected area with suitable climatic conditions compared with current levels (Table S3). By the year
98
2080, species models for four species predict a relatively large
reduction of protected climatically suitable area (> 60%;
Table 2). Only the kob could experience an increase in protected
area with suitable climate by 2080 (Table 2). Results from gap
analyses of species models indicate that persistence of suitable
climate in the current protected areas is < 50% for four species
by 2050 (Table S7). We find no evidence of statistical differences
for any species in gap analysis of projections from ENM-Ss,
ENM-Cs or ENM-As (Wilcoxon matched pairs test, n = 9).
Gap analysis results for lineages vary, depending on the modelling approach (Table 2, Fig. 3b). Gap analyses of projections
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
African mammals and global warming
Figure 2 Shift of suitable climatic conditions predicted by the ecological niche models for species (ENM-S) and for lineages (ENM-L) for
Phacocerus africanus. The black line indicates the current known distribution of P. africanus according to IUCN (2010) (Africa Albers equal
area conic projection).
Figure 3 Percentage of the current geographic ranges (a) and of the current protected range (b) of lineages that remain climatically
suitable by 2080 as predicted by different modelling approaches: ENM-L, ecological niche models for lineages; ENM-S, ecological niche
models for species; ENM-C, composite model; ENM-A, aggregated model.
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
99
M. D’Amen et al.
2050
2080
Persistence (%)
ENM-S
ENM-C
ENM-A
ENM-L
ENM-S
ENM-C
ENM-A
ENM-L
0–20
21–50
51–80
81–100
3
7
6
9
3
9
4
9
2
9
3
11
3
11
8
3
7
5
7
6
11
3
5
6
7
6
5
7
13
5
4
3
Table 3 Number of phylogeographic
lineages showing percentages of
protected area within current
distributions that retain suitable climate,
according to projections from four
different modelling approaches.
ENM-L, ecological niche models for lineages; ENM-S, ecological niche models for species; ENM-C,
composite model; ENM-A, aggregated model.
from ENM-Ls of four species (G. camelopardalis, H. equinus, K.
ellipsiprymnus and P. africanus) indicate contrasting trends
among conspecific lineages in the degree to which suitable
climate falls in any protected area in the future, within or outside
of current ranges (Tables 2 & S3). Further, we compare the
retention of suitable climatic conditions within the current protected distribution of lineages as projected by ENM-Ls and
species models (i.e. prediction of suitable climate for the species
as a whole in the areas currently occupied by the lineages). Our
results show significantly lower persistence of suitable climate in
the currently protected distribution of lineages according to
ENM-Ls compared with predictions from all species models
(Table S8, Fig. 3b). For example, ENM-Ls project that almost
twice as many lineages will have < 20% of current protected area
remaining climatically suitable as of 2080, compared with projections from ENM-Ss and ENM-As (Table 3).
DISCUSSION
The means to secure the conservation of phylogeographic lineages under changing climate and environment have been little
studied (Pfenninger et al., 2007). We detected considerable
among-lineage differences in the distributions of current suitable climate. Such variability leads to divergent projections of
the degree of change in future potential distributions. This calls
into question the validity of species models that do not account
for phylogeographic structure and lineage identity when such
structure exists. Our results suggest that no information relevant
to differential potential for persistence of lineages can be
obtained from existing approaches to modelling the climate
occupancy of species and projecting it to the future. In contrast,
the projection of single ‘lineage’ models of suitable climate provides an alternative prospective on the potential response of
species to climate change. This can, in turn, affect the evaluation
of the potential future amount of suitable area falling into
reserves, particularly when accounting for dispersal constraints
in the modelling implementation.
Phylogeographic structure and risks from
climate change
Genetic variation can be partitioned into two complementary
components. One is the selected (or functional) diversity arising
directly from adaptive evolution due to natural selection, while
100
the second is the neutral heritage of the population resulting
from the effects of neutral evolutionary forces such as genetic
drift, mutation or migration (Bonin et al., 2007). Phylogeographic studies are usually undertaken with neutral molecular
markers, which allow the detection of the effects of demographic history on patterns of lineage distribution and divergence. Nonetheless, in species with strong phylogeographic
structure, lineages can differ in the environmental space they
occupy (Murphy & Lovett-Doust, 2007; Pearman et al., 2010).
Considering our case study, among-lineage differences in behavioural, life-history and morphological traits exist for at least five
out of nine species (E. zebra, Nowak, 1999; Groves & Bell, 2004;
G. camelopardalis, Brown et al., 2007; K. ellipsiprymnus, East,
1998; K. kob, Lorenzen et al., 2007; L. africana, Groves & Grubb,
2000; Grubb et al., 2000). Most conspecific phylogeographic lineages in this study differ in the climates they experience and have
substantial geographic separation within a single geographic
and climate space (i.e. the geographic area and the climate that
the whole species occupies; Figs 2 & S2). This differentiation
leads to divergent projections of the degree of change in future
area of suitable climate for the lineages in six out of the nine
study species. These differences go undetected when considering
projections by species models. Together, these differences
suggest that our lineages diverge in characters additional to the
neutral genes used to identify phylogeographic structure and
emphasize the importance of accounting for inter-specific
lineage structure when planning conservation measures.
Overall, our results illustrate the importance of considering
multiple levels of biodiversity in the assessment of the potential
impacts of climate change on the distribution of suitable conditions for widely distributed species.
Understanding the current geographic distributions of suitable climate for phylogeographic lineages and the potential
alteration of these distributions due to changing climate could
be critical for management for the conservation of biodiversity.
Our findings suggest that species-level models are frequently
inadequate to provide such conservation information relevant
to lineages, and that bias was common in all three species model
approaches. The lack of differences among species models that
distinguish the contribution of lineages and those that do not is
probably due to the uniformity in prevalence among conspecific
lineages. Indeed the development of a composite niche model
has proven useful when some subspecific taxa are geographically
restricted, relative to other related taxa and, thus, contribute
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
African mammals and global warming
little to modelled distribution at the level of the species
(Pearman et al., 2010). In our case study, the majority of conspecific lineages have comparable geographic extents, thus they
contribute equally to the niche occupancy displayed by the
species, even under the classical species distribution model technique. All species-level modelling approaches underestimate the
risks of loss of suitable climate for some phylogeographic lineages in current protected areas. This probably occurs because
species models produce an estimate of the geographic area that
under future conditions remains climatically suitable for one or
more of the conspecific lineages, but not specifically for the
lineage that is originally present. Thus, projecting species
models to future climates could predict that a reserve remains
climatically suitable. However, whether this suitability refers to
the local populations or to populations from another portion of
the species range remains unclear. This uncertainty biases
assessment of the potential effects of climate change towards
underestimation of impacts on reserve adequacy.
The alternative approach of projecting single ‘lineage’ models
to changed climate conditions can influence the evaluation of
the future amount of suitable habitat area falling into reserves.
The weight that should be given to this perspective depends
on specific conservation goals and other biological considerations. For instance, the incorporation of phylogeographic
information could be particularly valuable when dispersal limitation is included in modelling that aims to assess potential,
climate-induced range shifts (Meier et al., 2011). The introduction of dispersal constraints could differentially affect the projection of models of species and lineages. Species models reflect
the assumption that dispersal constraints act only on displacements that occur outside the initial species boundaries. Under
this approach dispersal limitations would not be relevant to
members of subspecific units as they colonize areas within the
original species range. In contrast, dispersal constraints applied
to projections from lineage models reflect the assumption that
constraints apply beyond the original limits of the relevant
lineage itself. This probably produces more restrictive forecasts
of potential distributional shifts for the species as a whole
because lineage members are not allowed unlimited dispersal
within the part of the species range that remains climatically
suitable for the species. Such contrasting assumptions and projections can cause substantial differences in the assessment of
the geographic distribution of species in the future and, consequently, in setting conservation priorities. An approach that
specifically addresses climate suitability for lineages deserves
closer inspection and further development.
Data limitation and potential bias
The choice of African mammal species for this case study is
primarily due to the availability of robust phylogeographic
studies for multiple species for a single geographic area. Our
focus on tropical species, which might not experience extreme
climatic conditions (Araújo & Guisan, 2006), could put in question the ecological importance of the magnitude of modelled
climate change that we observed. On the other hand, Rapoport’s
rule (Stevens, 1989) suggests that some tropical species are likely
to be climate specialists and could be relatively sensitive to climatic variability and trends. Further, climate sensitivity in
African tropical forests and savannas might be much higher
than in more poleward-situated regions because of relatively
higher projected climate change in the tropics (Beaumont et al.,
2011). We also recognize that focusing on modelling climate
suitability can produce an overestimation of the projected suitable area for species and lineages because we do not take into
account other important drivers of mammal distributions, such
as land cover/use and inter-specific interactions. Moreover,
other approaches, such as an ensemble of different modelling
techniques and climatic scenarios (Araújo & New, 2006), can be
helpful in obtaining more reliable consensus projections of
future species distributions than our illustrative but comparatively simple one-algorithm approach. Finally, our estimates of
future protected area with suitable climate assume that protected areas do not increase or decrease in extent (or effectiveness). For these reasons, our results should not be used directly
for deriving practical conservation initiatives for African
mammals. The results nonetheless indicate the importance of
modelling potential intra-specific variation in response to
changing climatic and environmental factors.
Our choice of data on the current distribution of the
mammals in Africa could potentially influence our results. We
use the most up-to-date species distribution information. This
could potentially result in underestimation of the climate tolerances of species or lineages because, for some species, current
ranges are much reduced compared with their historical states
due to human pressures. Further, the extent of local adaptation
of lineages to the current climates they experience, and maladaptation to climates outside of the current distributions of these
species, is both unknown and in need of focused research.
Drastic reduction in population numbers for many of these
species (Novellie et al., 1992; Fennessy & Brown, 2010), and
historical bottlenecks in others (O’Brien et al., 1983; Whitehouse & Harley, 2001), could have caused substantial genetic
depletion, which might diminish the breadth of environmental
tolerance of lineages. Until sufficient knowledge is available to
suggest otherwise, we advise identification and use of the most
conservative criterion (i.e. to model both species and lineages on
the basis of the actual range) for conservation planning of key
taxa, given that the ultimate purpose of such modelling exercises
is to improve biodiversity conservation.
Climate change and evolutionary response
Modelling currently provides the most comprehensive and flexible approach to generating probabilistic projections of changes
in biodiversity under future climate scenarios (McMahon et al.,
2011). While many species may move to favourable climatic
habitats, others will overcome climate variation via plastic
changes, or through evolutionary adaptation. New studies can
incorporate evolution into distribution modelling (e.g. Kearney
et al., 2009), but such studies apply particularly to species with
short generation times and large population sizes (Lynch &
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
101
M. D’Amen et al.
Lande, 1993; Hoffmann & Sgrò, 2011; McMahon et al., 2011).
This raises the question of whether long-lived organisms, such
as large mammal species that are locally depleted in number due
to the high human pressure, can evolve to keep up with rapid
climate change. Currently there is no fast and inexpensive way to
determine whether a lineage will need to shift geographically to
track favourable conditions (due to niche conservatism) or has
the potential to mount an adaptive response without moving,
based upon standing genetic variation. A conservative approach
would be to secure the representation of lineages in conservation areas (at least within their current geographic distribution).
In this way, one might be more likely to capture the entire
distribution of phenotypic, behavioural and physiological variation with which some populations might be able to persist under
future climate change.
CONCLUSIONS
To become effective, programmes for biodiversity conservation
need to bring both spatial and temporal perspectives to the
forefront of their research agendas. Phylogeography, the study of
the geographic distribution of genealogical lineages, can contribute these essential components – time and space – to inform
conservation planning. We show that the combination of phylogeographic and modelling approaches can improve our ability
to forecast shifts in the potential ranges of lineages under
climate change. The approach we demonstrate contributes to a
more realistic assessment of the effects of climate change on
biodiversity through the inclusion of phylogeographic information. When these forecasts are used to inform gap analysis, we
obtain valuable information on risk to (and perhaps in some
cases, opportunity for) the future representation of phylogeographic lineages and, thus, to the conservation of the evolutionary potential of species. The early recognition of the elements of
biodiversity that are threatened by climate change is crucial for
planning effective countermeasures.
ACKNOWLEDGEMENTS
We thank four anonymous referees and P. Bombi for helpful
comments on previous versions of the manuscript. D. Salvi provided valuable advice regarding molecular phylogeographic
studies. P.B.P. acknowledges the support of SNF grants
CRSII3_125240 ‘SPEED’ and 31003A_122433 ‘ENNIS’. P.B.P.
and N.E.Z. acknowledge the support of EU grants GOCECT-2007-036866 ‘ECOCHANGE’ and ENVCT-2009-226544
‘MOTIVE’.
REFERENCES
Allouche, O., Tsoar, A. & Kadmon, R. (2006) Assessing the accuracy of species distribution models: prevalence, kappa and the
true skill statistic (TSS). Journal of Applied Ecology, 43, 1223–
1232.
Alpers, D.L., Van Vuuren, B.J., Arctander, P. & Robinson, T.J.
(2004) Population genetics of the roan antelope (Hippotragus
102
equinus) with suggestions for conservation. Molecular Ecology,
13, 1771–1784.
Araújo, M.B. & Guisan, A. (2006) Five (or so) challenges for
species distribution modelling. Journal of Biogeography, 33,
1677–1688.
Araújo, M.B. & New, M. (2006) Ensemble forecasting of species
distributions. Trends in Ecology and Evolution, 22, 42–47.
Araújo, M.B., Cabeza, M., Thuiller, W., Hannah, L. & Williams,
P.H. (2004) Would climate change drive species out of
reserves? An assessment of existing reserve-selection methods.
Global Change Biology, 10, 1618–1626.
Araújo, M.B., Pearson, R.G. & Thuiller, W. (2005) Validation of
species–climate impact models under climate change. Global
Change Biology, 11, 1504–1513.
Ballard, J.W.O. & Whitlock, M.C. (2004) The incomplete
natural history of mitochondria. Molecular Ecology, 13, 729–
744.
Beaumont, L.J., Pitman, A., Perkins, S., Zimmermann, N.E.,
Yoccoz, N.G. & Thuiller, W. (2011) Impacts of climate change
on the world’s most exceptional ecoregions. Proceedings of the
National Academy of Sciences USA, 108, 2306–2311.
Belsley, D.A., Kuh, E. & Welsch, R.E. (1980) Regression diagnostics: identifying influential data and sources of collinearity. John
Wiley and Sons, New York.
Bonin, A., Nicole, F., Pompanon, F., Miaud, C. & Taberlet, P.
(2007) Population adaptive index: a new method to help
measure intraspecific genetic diversity and prioritize
populations for conservation. Conservation Biology, 21,
697–708.
Brown, D.M., Brenneman, R.A., Koepfli, K.-P., Pollinger, J.P.,
Milá, B., Georgiadis, N.J., Louis, E.E., Grether, G.F., Jacobs,
D.K. & Wayne, R.K. (2007) Extensive population genetic
structure in the giraffe. BMC Biology, 5, 57. doi:10.1186/17417007-5-57.
Charruau, P., Fernandes, C., Orozco-Terwengel, P., Peters, J.,
Hunter, L., Ziaie, H., Jourabchian, A., Jowkar, H., Schaller, G.,
Ostrowski, S., Vercammen, P., Grange, T., Schlötterer, C.,
Kotze, A., Geigl, E.-M., Walzer, C. & Burger, P.A. (2011)
Phylogeography, genetic structure and population divergence
time of cheetahs in Africa and Asia: evidence for long-term
geographic isolates. Molecular Ecology, 20, 706–724.
Crandall, K.A., Bininda-Emonds, O.R.P., Mace, G.M. & Wayne,
R.K. (2000) Considering evolutionary processes in conservation biology. Trends in Ecology and Evolution, 15, 290–295.
D’Amen, M., Bombi, P., Pearman, P.B., Schmatz, D.R., Zimmermann, N.E. & Bologna, M.A. (2011) Will climate change
reduce the efficacy of protected areas for amphibian conservation in Italy? Biological Conservation, 144, 989–997.
East, R. (1998) Antelope specialist group report. IUCN, Gland,
Switzerland.
Elith, J., Graham, C.H., Anderson, R.P. et al. (2006) Novel
methods improve prediction of species’ distributions from
occurrence data. Ecography, 29, 129–151.
Elith, J., Leathwick, J.R. & Hastie, T. (2008) A working guide to
boosted regression trees. Journal of Animal Ecology, 77, 802–
813.
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
African mammals and global warming
Ennos, R.A., French, G.C. & Hollingsworth, P.M. (2005) Conserving taxonomic complexity. Trends in Ecology Evolution,
20, 164–168.
Fennessy, J. & Brown, D. (2010) Giraffa camelopardalis. IUCN
2010. IUCN Red List of threatened species. Version 2010.4.
Available at: http://www.iucnredlist.org (accessed May 2011).
Fielding, A.H. & Bell, J.F. (1997) A review of methods for the
assessment of prediction errors in conservation presence/
absence models. Environmental Conservation, 24, 38–49.
Friedman, J., Hastie, T. & Tibshirani, R. (2000) Additive logistic
regression: a statistical view of boosting. Annals of Statistics,
28, 337–374.
Groves, C.P. & Bell, C.H. (2004) New investigations on the taxonomy of the zebras genus Equus, subgenus Hippotigris.
Mammalian Biology, 69, 182–196.
Groves, C.P. & Grubb, P. (2000) Do Loxodonta cyclotis and L.
africana interbreed? Elephant, 2, 4–7.
Grubb, P., Groves, C.P., Dudley, J.P. & Shoshani, J. (2000) Living
African elephants belong to two species: Loxodonta africana
(Blumenbach, 1797) and Loxodonta cyclotis (Matschie, 1900).
Elephant, 2, 1–4.
Guisan, A. & Zimmermann, N.E. (2000) Predictive habitat distribution models in ecology. Ecological Modelling, 135, 147–
186.
Hannah, L., Midgley, G.F., Lovejoy, T.E., Bond, W.J., Bush, M.,
Lovett, J.C., Scott, D. & Woodward, F. (2002) Conservation of
biodiversity in a changing climate. Conservation Biology, 16,
264–268.
Harley, E.H., Baumgarten, I., Cunningham, J. & O’Ryan, C.
(2005) Genetic variation and population structure in remnant
populations of black rhinoceros, Diceros bicornis, in Africa.
Molecular Ecology, 14, 2981–2990.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. & Jarvis, A.
(2005) Very high resolution interpolated climate surfaces for
global land areas. International Journal of Climatology, 25,
1965–1978.
Hoffmann, A.A. & Sgrò, C.M. (2011) Climate change and evolutionary adaptation. Nature, 470, 479–485.
IUCN (2010) IUCN Red List of threatened species. Version
2010.4. Available at: http://www.iucnredlist.org (accessed
March 2011).
IUCN and UNEP-WCMC (2010) The world database on protected areas (WDPA): annual release. UNEP-WCMC, Cambridge, UK. Available at: http://www.wdpa.org (accessed July
2010).
Jennings, M.D. (2000) Gap analysis: concepts, methods, and
recent results. Landscape Ecology, 15, 5–20.
Kearney, M., Porter, W.P., Williams, C., Ritchie, S. & Hoffmann,
A.A. (2009) Integrating biophysical models and evolutionary
theory to predict climatic impacts on species’ ranges: the
dengue mosquito Aedes aegipti in Australia. Functional
Ecology, 23, 528–538.
Lorenzen, E.D., Simonsen, B.T., Kat, P.W., Arctander, P. &
Siegismund, H.R. (2006) Hybridization between subspecies of
waterbuck (Kobus ellipsiprymnus) in zones of overlap with
limited introgression. Molecular Ecology, 15, 3787–3799.
Lorenzen, E.D., De Neergaard, R., Arctander, P. & Siegismund,
H.R. (2007) Phylogeography, hybridization and Pleistocene
refugia of the kob antelope (Kobus kob). Molecular Ecology, 16,
3241–3252.
Lynch, M. & Lande, R. (1993) Evolution and extinction in
response to environmental change. Biotic interactions and
global change (ed. by P.M. Kareiva, J.G. Kingsolver and R.B.
Huey), pp. 234–250. Sinauer, Sunderland, MA.
Mace, G.M. (2004) The role of taxonomy in species conservation. Philosophical Transactions of the Royal Society B: Biological Sciences, 359, 711–719.
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 in Ecology and Evolution, 26, 249–259.
Meier, E.S., Lischke, H., Schmatz, D.R. & Zimmermann, N.E.
(2011) Climate, competition and connectivity affect future
migration and ranges of European trees. Global Ecology and
Biogeography, 21, 164–178.
Moodley, Y. & Harley, E.H. (2005) Population structuring in
mountain zebras (Equus zebra): the molecular consequences
of divergent demographic histories. Conservation Genetics, 6,
953–968.
Murphy, H.T. & Lovett-Doust, J. (2007) Accounting for regional
niche variation in habitat suitability models. Oikos, 116,
99–110.
Muwanika, V.B., Nyakaana, S., Siegismund, H.R. & Arctander, P.
(2003) Phylogeography and population structure of the
common warthog (Phacochoerus africanus) inferred from
variation in mitochondrial DNA sequences and microsatellite
loci. Heredity, 91, 361–372.
Novellie, P., Lloyd, P.H. & Joubert, E. (1992) Mountain zebras.
Zebras, asses and horses: an action plan for the conservation of
wild equids (ed. by P. Duncan), pp. 6–9. IUCN, Gland,
Switzerland.
Nowak, R. (1999) Perissodactyla; Equidae. Walker’s mammals of
the world, Vol. 2, 6th edn (ed. by R. Nowak), pp. 1020–1021.
The Johns Hopkins University Press, Baltimore, MD.
O’Brien, S.J., Wildt, D., Goldman, D., Merril, C.R. & Bush, M.
(1983) The cheetah is depauperate in genetic variation.
Science, 221, 459–462.
Parmesan, C. & Yohe, G. (2003) A globally coherent fingerprint
of climate change impacts across natural systems. Nature, 421,
37–42.
Pearman, P.B., D’Amen, M., Graham, C.H., Thuiller, W. & Zimmermann, N.E. (2010) Within-taxon niche structure: niche
conservatism, divergence and predicted effects of climate
change. Ecography, 33, 990–1003.
Peterson, A.T. (2011) Ecological niche conservatism: a timestructured review of evidence. Journal of Biogeography, 38,
817–827.
Pfenninger, M., Nowak, C. & Magnin, F. (2007) Intraspecific range dynamics and niche evolution in Candidula
land snail species. Biological Journal of the Linnean Society, 90,
303–317.
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd
103
M. D’Amen et al.
R Development Core Team (2010) R: a language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria.
Ramirez, J. & Jarvis, A. (2008) High resolution statistically downscaled future climate surfaces. International Centre for Tropical
Agriculture, CIAT, Cali, Colombia. Available at: http://
www.ccafs-climate.org/ (accessed March 2011).
Ridgeway, G. (2006) Generalized boosted models: a guide to the
gbm package. R vignette. [unpublished PDF document].
Available at: http://cran.r-project.org/web/packages/gbm/
vignettes/gbm.pdf (accessed March 2011).
Rohland, N., Reich, D., Mallick, S., Meyer, M., Green, R.E., Georgiadis, N.J., Roca, A.L. & Hofreiter, M. (2010) Genomic DNA
sequences from mastodon and woolly mammoth reveal deep
speciation of forest and savanna elephants. PLoS Biology, 8, 12,
e1000564.
Stevens, G.C. (1989) The latitudinal gradient in geographical
range: how so many species coexist in the tropics. The American Naturalist, 133, 240–256.
Stockwell, D.R.B. & Peterson, A.T. (2002) Controlling bias in
biodiversity data. Predicting species occurrences: issues of scale
and accuracy (ed. by J.M. Scott, P.J. Heglund and M.L. Morrison), pp. 537–546. Island Press, Washington, DC.
Thuiller, W., Lavergne, S., Roquet, C., Boulangeat, I., Lafourcade,
B. & Araujo, M.B. (2011) Consequences of climate change on
the tree of life in Europe. Nature, 470, 531–534.
Whitehouse, A.M. & Harley, E.H. (2001) Post-bottleneck
genetic diversity of elephant populations in South Africa,
revealed using microsatellite analysis. Molecular Ecology, 10,
2139–2149.
S U P P O RT I N G I N F O R M AT I O N
Additional Supporting Information may be found in the online
version of this article:
Figure S1 Geographic distribution of phylogeographic lineages
according to IUCN information and relevant publications for
the target species.
Figure S2 Geographic distribution of suitable climatic conditions predicted by lineage distribution models for the years
2000, 2050 and 2080.
Table S1 Bioclimatic grids utilized as predictors in generalized
boosted algorithm models.
Table S2 Number of 1° cells predicted to be climatically suitable
under the current climatic conditions for species and intraspecific lineages.
Table S3 Percentage change in areal extent and in protected area
with suitable climatic conditions by 2050 under the assumption
of unlimited dispersal.
Table S4 Percentage of persistence of suitable climatic conditions within the current range of every species.
Table S5 Degree of persistence of climatically suitable cells in
the known geographic ranges of intra-specific lineages.
Table S6 Results from the Wilcoxon matched pairs tests conducted to evaluate whether the persistence in the current ranges
of lineages differed among model predictions.
104
Table S7 Number of species with different percentages of protected surface that will remain suitable under future climate
(persistence), according to different modelling approaches.
Table S8 Results from the Wilcoxon matched pairs tests conducted to evaluate whether the potential occupancy in the lineages’ protected areas differed among model predictions under
future climate.
As a service to our authors and readers, this journal provides
supporting information supplied by the authors. Such materials
are peer-reviewed and may be re-organized for online delivery,
but are not copy-edited or typeset. Technical support issues
arising from supporting information (other than missing files)
should be addressed to the authors.
BIOSKETCHES
Manuela D’Amen has a range of interests in spatial
ecology and conservation biology. Her recent work has
ranged from forecasting the effects of climate change on
faunal diversity and conservation to the development of
the best methods for spatial analysis of biodiversity
data. She is also interested in combining phylogenetic
information and niche modelling to provide insight
into niche dynamics and to contribute to predicting
impacts of climate change.
Niklaus E. Zimmermann is an ecologist and head
of the research unit in landscape dynamics at the
Swiss Federal Research Institute WSL, Birmensdorf,
Switzerland. He is interested in macroecology,
ecological theory and niche evolution. His research has
recently focused on the impacts of climate and land-use
change on species, ecosystems and biodiversity at a
range of spatial scales, and on evolutionary aspects
related to climate patterns and climate change.
Peter B. Pearman is interested in how species are
distributed along environmental gradients, both
biotic and abiotic, which has led him to use both
manipulative and correlative approaches. He has
contributed to fieldwork on four continents. Many of
his earlier publications address aspects of patterns of
distribution of amphibians. He is currently interested in
questions surrounding the structure of the niche and its
evolution. He addresses this with ecological data and
data from both phylogeographic and phylogenetic
analyses.
Author contributions: M.D. and P.B.P. conceived the
questions and analysis; M.D. assembled data, conducted
analyses and modelling, and assembled all tables and
graphics; M.D. and P.B.P. wrote the first draft; M.D.,
P.B.P. and N.Z. contributed to subsequent revision.
Editor: José Alexandre F. Diniz-Filho
Global Ecology and Biogeography, 22, 93–104, © 2012 Blackwell Publishing Ltd