Unit III Geo

RESEARCH ARTICLE

Climate-Related Local Extinctions Are Already

Widespread among Plant and Animal Species

John J. Wiens*

Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona, United States of

America

* [email protected]

Abstract

Current climate change may be a major threat to global biodiversity, but the extent of spe-

cies loss will depend on the details of how species respond to changing climates. For exam-

ple, if most species can undergo rapid change in their climatic niches, then extinctions may

be limited. Numerous studies have now documented shifts in the geographic ranges of spe-

cies that were inferred to be related to climate change, especially shifts towards higher

mean elevations and latitudes. Many of these studies contain valuable data on extinctions of

local populations that have not yet been thoroughly explored. Specifically, overall range

shifts can include range contractions at the “warm edges” of species’ ranges (i.e., lower lati-

tudes and elevations), contractions which occur through local extinctions. Here, data on cli-

mate-related range shifts were used to test the frequency of local extinctions related to

recent climate change. The results show that climate-related local extinctions have already

occurred in hundreds of species, including 47% of the 976 species surveyed. This frequency

of local extinctions was broadly similar across climatic zones, clades, and habitats but was

significantly higher in tropical species than in temperate species (55% versus 39%), in ani-

mals than in plants (50% versus 39%), and in freshwater habitats relative to terrestrial and

marine habitats (74% versus 46% versus 51%). Overall, these results suggest that local

extinctions related to climate change are already widespread, even though levels of climate

change so far are modest relative to those predicted in the next 100 years. These extinctions

will presumably become much more prevalent as global warming increases further by

roughly 2-fold to 5-fold over the coming decades.

Author Summary

Climate change is an important threat to the world’s plant and animal species, including

species on which humans depend. However, predicting how species will respond to future

climate change is very difficult. In this study, I analyze the extinctions caused by the cli-

mate change that has already occurred. Numerous studies find that species are shifting

their geographic ranges in response to climate change, typically moving to higher eleva-

tions and latitudes. These studies also contain valuable data on local extinctions, as they

document the loss of populations at the “warm edge” of species’ ranges (lower elevations

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 1 / 18

a11111

OPEN ACCESS

Citation: Wiens JJ (2016) Climate-Related Local

Extinctions Are Already Widespread among Plant

and Animal Species. PLoS Biol 14(12): e2001104.

doi:10.1371/journal.pbio.2001104

Academic Editor: Anthony Barnosky, University of

California-Berkeley, United States of America

Received: September 14, 2016

Accepted: November 3, 2016

Published: December 8, 2016

Copyright: © 2016 John J. Wiens. This is an open access article distributed under the terms of the

Creative Commons Attribution License, which

permits unrestricted use, distribution, and

reproduction in any medium, provided the original

author and source are credited.

Data Availability Statement: All relevant data are

within the paper and its Supporting Information

files.

Funding: The author(s) received no specific

funding for this work.

Competing Interests: The authors have declared

that no competing interests exist.

Abbreviations: GLM, general linear model; GLMM,

general linear mixed model.

and latitudes). Here, I use these data to show that recent local extinctions related to climate

change have already occurred in hundreds of species around the world. Specifically, among

976 species surveyed, local extinctions occurred in 47%. These extinctions are common

across climatic zones, habitats, and groups of organisms but are especially common in trop-

ical regions (which contain most of Earth’s species), in animals (relative to plants), and in

freshwater habitats. In summary, this study reveals local extinctions in hundreds of species

related to the limited global warming that has already occurred. These extinctions will

almost certainly increase as global climate continues to warm in the coming decades.

Introduction

Anthropogenic climate change may be a major driver of biodiversity loss in the next 100 years,

but the possible impacts of climate change on species survival remain highly uncertain [1–3].

Global mean annual temperatures increased by ~0.85˚C between 1880 and 2012 and are likely

to rise by an additional 1˚C to 4˚C by 2100 [4]. Modeling studies have predicted that various

levels of species loss will result from this future climate change, ranging from 0% to >50% of

all species currently known [3]. This uncertainty has many sources (e.g., different climate

models and different hypotheses about species dispersal). One of the most important sources

of uncertainty hinges on the details of how species respond to climate change. For example, if

species can evolve rapidly enough in response to changing climate, then species extinctions

due to climate change might actually be limited [5,6].

Species can potentially respond to climate change in several ways. The most important case

to consider may be that when the species’ present-day (realized) climatic niche no longer

occurs within the species’ current geographic range (because of the potential for global extinc-

tion of the species under these conditions). In this case, the possible responses of the species

include the following: (i) undergoing niche shifts, such that the species’ realized niche changes

to incorporate these new climatic conditions (e.g., through plastic changes and/or by evolu-

tionary adaptation to the modified abiotic and/or biotic conditions), (ii) dispersing to track

the original climatic conditions over space (e.g., moving to higher latitudes or elevations), and

(iii) going extinct [5–8]. While each of these responses has been shown in some cases (at least

in local populations), the relative frequency of each is still unclear [7,8]. However, changes in

species’ geographic ranges have been especially well documented [9–11].

These data on geographic range shifts contain important but underutilized information on

how species respond to climate change. Range shifts observed under climate change typically

involve an overall shift towards higher latitudes and higher elevations [9–11]. These shifts can

be composed of one (or both) of two types of changes (Fig 1): (i) range expansions at the cool

edge of the species range (higher latitudes and elevations) and (ii) range contractions at the

warm edge (lower latitudes and elevations). The presence of warm-edge contractions is criti-

cally important. A warm-edge contraction occurs when populations from one or more locali-

ties at the lowest latitudes or elevations of a species’ regional distribution disappear (i.e., are

inferred to no longer occur at those localities), leading to an overall shift in the species range

towards higher latitudes or elevations. These contractions indicate that species are failing to

shift their niches sufficiently to tolerate these new conditions and that these populations are

instead going extinct (referred to as “local extinction” hereafter). This must be true regardless

of the specific mechanism of local extinction (e.g., elevated death rates, increased emigration,

or declining recruitment). The many papers that have assessed range shifts and that have

included surveys of warm-edge populations can therefore provide a wealth of data about

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 2 / 18

which species have (and have not) undergone local extinctions potentially related to climate

change. These data are particularly useful because published papers on range shifts need not

be strongly biased towards documenting warm-edge contractions, given that many studies

that included data on warm edges also surveyed the cool edge. Thus, even though studies that

failed to find any range shifts might go unpublished (a potential source of bias), studies that

documented an overall range shift need not show a warm-edge contraction.

Here, I analyze the extensive data on range shifts to examine the prevalence of local extinc-

tions related to modern climate change. I also provide a synthesis of inferred local extinction

across habitats, climatic zones, and taxonomic groups. I systematically searched the literature

for studies that examined shifts in species’ ranges at their warm edges, shifts that were consid-

ered (in the original studies) to be related to current climate change. Hundreds of examples of

local extinctions were found across diverse climatic zones, habitats, and taxonomic groups.

Not all species exhibiting range shifts showed warm-edge contractions, but ~50% of the species

surveyed had local extinctions inferred to be related to climate change. These results suggest

that even the relatively small changes in climate that have already occurred are sufficient to

cause widespread local extinctions and that many species may be unable respond to climate

change fast enough to avoid extinction as global climate warms even further.

Results

The Web of Science was searched repeatedly between December 2014 and March 2016 using

keywords related to climate change, range shifts, and local extinctions (see Materials and

Fig 1. Hypothetical example illustrating the two components of a geographic range shift associated with climate change. The large open circle

indicates the species’ overall geographic range. Small dark blue circles indicate populations before climate change. After climate change, the overall

geographic range is shifted northward (large open circle), both through the range expansion (new populations; small light blue circles) added at the

northern, “cold” edge of the species range and range contraction (local extinction of original populations; small red circles) at the southern, “warm” edge

of the species range. Similar patterns occur for range shifts along an elevational gradient. Modified from Cahill et al. [12].

doi:10.1371/journal.pbio.2001104.g001

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 3 / 18

Methods). All studies that monitored the warm edge of at least one species’ range and that tied

their results to climate change with explicit statistical analyses were included. Importantly,

studies can document overall range shifts but need not find that the warm-edge populations

that they examined had local extinctions.

A total of 27 studies (Table 1; [13–39]) met all the necessary criteria to address potential cli-

mate-associated warm-edge range shifts (see Materials and Methods). The sampled species

were broadly distributed across clades (e.g., animals = 716; plants = 260) and regions (e.g.,

Asia = 332; Europe = 268; Madagascar = 30; Oceania = 58; North America = 233; South Amer-

ica = 55). Among the 976 unique species surveyed, 460 species had warm-edge contractions,

and 516 did not (S1 Appendix). Therefore, local extinctions related to climate change are

already very common (47.1% of species examined), even given the relatively modest rise in

global temperatures that has occurred so far (less than 1˚C increase in global mean annual

temperature; [4]).

These 976 species spanned many clades, habitats, and regions (Table 1; S1 Appendix).

Comparison between those species that showed warm-edge contractions and those that did

not provides potential insights into which species may be most sensitive to climate change, in

terms of the climatic zones and habitats that they occur in and the clades that they belong to.

Furthermore, there is no evidence that there were more species with local extinctions in studies

that ended more recently, were of longer duration, or began earlier (based on midpoints for

ranges of values; Table 1). Specifically, regression analyses of the proportion of species with

local extinctions against (i) the study end date, (ii) the duration of the study, and (iii) the study

start date all yielded nonsignificant results (end date: r2 = 0.001, p = 0.8910; duration: r2 = 0.045, p = 0.2896; start date r2 = 0.047, p = 0.2788; after removing nine studies with four or fewer species: end date: r2 = 0.146, p = 0.1181; duration: r2 = 0.132, p = 0.1376, but unexpect- edly trending towards fewer extinctions in studies with longer durations; start date r2 = 0.177, p = 0.0821, with more extinctions in studies beginning more recently, not earlier). Therefore, the frequency of local extinctions was initially compared across species in different studies,

regardless of differences in the duration, beginning, or end date of the study in which they

were surveyed.

Overall, the frequency of local extinctions was similar (close to 50%) across most climatic

zones, habitats, gradients, and clades. Nevertheless, there were some significant differences.

First, local extinctions were significantly more common in species from tropical and subtropi-

cal regions (combined and referred to as tropical hereafter for brevity) than in those from tem-

perate regions (p < 0.0001; Chi-squared test, testing the assumption of equal frequencies of local extinction among species between regions; subsequent p-values are also from Chi- squared tests). Specifically, 54.6% of the 504 included tropical species had local extinctions,

whereas only 39.2% of the 472 temperate species did (Fig 2A). The pattern was even stronger

when only considering terrestrial species on elevational gradients (54.6% of 504 tropical spe-

cies versus 28.2% of 301 temperate species), which applied to all plants and most animals. In

part, this pattern of more frequent tropical extinction arose from a much lower frequency of

extinctions for temperate plants (59.4% of 155 tropical species versus 8.6% of 105 temperate

species; p < 0.0001). The very low frequency of temperate extinctions in plants was based on a single study from very high latitudes [19]. Nevertheless, there were also significantly more

local extinctions in tropical animals (52.4% of 349 tropical species versus 38.8% of 196 temper-

ate species; p = 0.0022), if one compares terrestrial species on elevational gradients. This restriction also made them more comparable to the sampled plants (all from terrestrial, eleva-

tional gradients) and still encompassed most sampled animal species (76.1%; 545 of 716 spe-

cies). Across all animals, the difference was not significant (p = 0.2309), possibly because of the influence of temperate marine and freshwater species (see below). Among the most well-

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 4 / 18

Table 1. Summary information on the 27 range-shift studies used to document local extinctions related to climate change. Studies are listed alpha-

betically by first author. The major taxonomic group surveyed is given (Taxon, all groups are animals except for “Plant”), along with the total number of species

surveyed (Total Species), the percentage of those species with one or more local extinctions (% Local Extinction), the general habitat type (Habitat; including

terrestrial, freshwater, and marine), the climatic region (tropical-subtropical versus temperate), the geographic region where the study was conducted (note

that North America here extends to Central America), the type of range shift (latitudinal, elevational), the dates of the initial survey and the resurvey, and the

duration in between (for surveys and/or resurveys spanning multiple years, the midpoint of each was used to calculate the duration).

Reference Taxon Total

Species

% Local

Extinction

Habitat Climatic

Region

Geographic

Region

Range

Shift

Initial

Survey

Resurvey

Date

Duration

Angelo and

Daehler [13]

Plant 4 50 Terrestrial Tropical Oceania

(Hawaii)

Elevational 1966–

1967

2008 41.5

Beever et al.

[14]

Mammal 1 100 Terrestrial Temperate North America Elevational 1898–

1956

2003–2006 77.5

Brusca et al.

[15]

Plant 27 56 Terrestrial Tropical North America Elevational 1963 2011 48

Chen et al. [16] Insect 208 56 Terrestrial Tropical Asia Elevational 1965 2007 42

Comte and

Grenouillet [17]

Fish 31 74 Fresh. Temperate Europe Elevational 1980–

1992

2003–2009 20

Dieker et al.

[18]

Insect 2 50 Terrestrial Temperate Europe Elevational 1958–

1986

2008–2009 36.5

Felde et al. [19] Plant 105 9 Terrestrial Temperate Europe Elevational 1900 2008 108

Forero-Medina

et al. [20]

Bird 55 29 Terrestrial Tropical South America Elevational 1969 2010 41

Franco et al.

[21]

Insect 3 100 Terrestrial Temperate Europe Latitudinal 1970–

1999

2004–2005 20

Freeman and

Freeman [22]

Bird 54 74 Terrestrial Tropical Oceania (New

Guinea)

Elevational 1965 2012 47

Hiddick et al.

[23]

Marine

invertebrates

65 55 Marine Temperate Europe Latitudinal 1986 2000 14

Hitch and

Leberg [24]

Bird 1 100 Terrestrial Temperate North America Latitudinal 1967–

1971

1998–2002 31

Menendez

et al. [25]

Insect 39 54 Terrestrial Temperate Europe Elevational 1981–

1993

2006–2007 24

Moritz et al.

[26]

Mammal 27 41 Terrestrial Temperate North America Elevational 1914–

1920

2003–2006 87.5

Myers et al.

[27]

Mammal 8 12 Terrestrial Temperate North America Latitudinal 1883–

1980

1981–2006 62

Nye et al. [28] Fish 28 50 Marine Temperate North America Latitudinal 1968 2008 40

Perry et al. [29] Fish 10 40 Marine Temperate North America Latitudinal 1997 2001 24

Ploquin et al.

[30]

Insect 16 69 Terrestrial Temperate Europe Elevational 1988–

1989

2007–2009 19.5

Pomara et al.

[31]

Squamate 1 100 Terrestrial Temperate North America Elevational 1965 2008 43

Raxworthy

et al. [32]

Amphibian-

Squamate

30 37 Terrestrial Tropical Madagascar Elevational 1993 2003 10

Rowe et al. [33] Mammal 4 25 Terrestrial Temperate North America Elevational 1927–

1929

2006–2008 79

Rubal et al. [34] Mollusca 7 29 Marine Temperate Europe Latitudinal 1917,

1940

2011 94

Sheldon [35] Insect 1 0 Terrestrial Temperate North America Elevational 1977–

1978

2006 28.5

Telwala et al.

[36]

Plant 124 60 Terrestrial Tropical Asia Elevational 1849–

1850

2007–2010 159

Tingley et al.

[37]

Bird 92 25 Terrestrial Temperate North America Elevational 1900–

1930

1980–2006 78

Warren and

Chick [38]

Insect 2 0 Terrestrial Tropical North America Elevational 1973–

1974

2012 38.5

(Continued )

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 5 / 18

sampled groups of animals, tropical extinction was significantly more common in birds

(51.4% of 109 tropical species versus 37.1% of 124 temperate species; p = 0.0284), but not in insects (local extinctions in 55.2% of 210 tropical species versus 59.0% of 61 temperate species;

p = 0.6007). For other animal groups, the species sampled here were either predominantly tem- perate (mammals, fish, and marine invertebrates) or tropical (squamate reptiles and amphibi-

ans), and so did not allow for similar within-clade comparisons.

Overall, the frequency of climate-related local extinctions (Fig 2B) was similar in terrestrial

(45.6% of 835 species) and marine environments (50.9% of 110; p = 0.2964). In contrast, the frequency in freshwater species was substantially higher (74.2% of 31; p = 0.0053 across all three habitats). However, the estimate for freshwater species was based on a single study of

European fishes [17]. Comparing fish only (all temperate) also supported a significantly higher

frequency of extinction in freshwater environments relative to marine environments

(p = 0.0240; local extinctions in 47.4% of 38 marine species versus 74.2% of 31 freshwater spe- cies). All marine species included here were temperate animals, but there was no significant

difference in extinction frequencies between marine and terrestrial environments when only

temperate animals were compared (p = 0.1676; marine: 50.9% of 110 species, terrestrial: 42.9% of 226 species). Terrestrial and freshwater species remained significantly different in this more

restricted comparison (p = 0.0011). The frequency of local extinctions (Fig 2C) was somewhat lower for species surveyed along

elevational gradients relative to those on latitudinal gradients (elevational: 45.8% of 836 spe-

cies; latitudinal: 55.0% of 140 species; p = 0.0439). Most (78.6%) species measured along latitu- dinal gradients were marine (and all marine studies focused on latitudinal gradients), and all

were temperate. Again, most species included here were based on studies of elevational gradi-

ents in terrestrial environments.

Local extinctions were also broadly similar in frequency across taxonomic groups (Fig 3).

Nevertheless, local extinctions were significantly more common (p = 0.0018) in animals (50.1% of 716) than plants (38.8% of 260). This difference was reduced when comparing only

animals and plants on terrestrial, elevational gradients (47.3% of 556 animal species versus

38.8% of 260 plant species; p = 0.0236). Among these latter species, the plant–animal difference was nonsignificant for tropical species (and was actually reversed: local extinctions in 52.4% of

349 tropical animal species versus 59.4% of 155 tropical plants; p = 0.1500) but was strong for temperate species (38.6% of 207 temperate animal species versus 8.6% of 105 temperate plants;

p < 0.0001). The frequencies of local extinctions across different animal groups (Fig 3) were broadly

similar to the overall value for animals (50.1%), but with higher values in insects (56.1% of 271

species; based on six studies; Table 1) and fish (59.4% of 69 species; three studies) relative to

mammals (35.0% of 40 species; four studies), birds (43.8% of 233 species; five studies), amphib-

ians (36.8% of 19 species; one study), and squamate reptiles (lizards and snakes; 41.7% of 12

species; two studies). Local extinctions were also broadly similar in frequency in various

groups of marine invertebrates, including crustaceans (46.7% of 15; one study), annelids

(64.5% of 31; one study), and molluscs (45.4% of 22; two studies). The frequency in echino-

derms was lower (25.0%; one study) but was based on a very small sample size (4 species).

Table 1. (Continued )

Reference Taxon Total

Species

% Local

Extinction

Habitat Climatic

Region

Geographic

Region

Range

Shift

Initial

Survey

Resurvey

Date

Duration

Zuckerberg

et al. [39]

Bird 31 71 Terrestrial Temperate North America Both 1980–

1985

2000–2005 20

doi:10.1371/journal.pbio.2001104.t001

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 6 / 18

Results were generally similar using both general linear models (GLMs; see below) and gen-

eral linear mixed models (GLMMs; see next paragraph). GLM results are given in full in S2

Appendix and are summarized here. Simultaneously including all 976 species and most vari-

ables (habitat [terrestrial versus freshwater versus marine], climatic regions [tropical versus

temperate], taxonomic group [plants versus animals], survey type [latitudinal versus eleva-

tional], and study dates [start date, end date, and duration in between]) showed that most

Fig 2. The frequency of local extinctions related to climate change across different climatic regions,

habitats, and gradients. (A) Species are categorized as temperate or tropical (based on the location of the

study), and the percentage of species with one or more local extinctions is shown, along with the sample sizes

of species in each region. (B) Species are categorized as terrestrial, freshwater, or marine, and the frequency

of species with local extinctions is shown (along with total species per habitat). (C) Species are categorized

based on whether they were surveyed along elevational or latitudinal transects. Vertical lines indicate 95%

confidence intervals on the estimated frequency of species with local extinctions.

doi:10.1371/journal.pbio.2001104.g002

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 7 / 18

variables had significant effects on the frequency of extinction, except for the study dates.

There were strong effects of habitat and climate (p < 0.00001) but weaker effects of taxonomic group (p = 0.0246). Results were similar when excluding study dates and taxonomic group. Including geographic regions showed that most regions had no significant effect (except for

Madagascar and South America). Given that Madagascar and South America were represented

by one study each, these region effects were not considered further. Furthermore, the effects of

climatic region, habitat, taxonomic group, and survey type remained significant when geo-

graphic regions were included. Comparing species only on terrestrial elevational gradients

(805 species in total) further confirmed the significant effects of climate and taxonomic group.

Similarly, considering plants only (260 species) also confirmed the significant effects of cli-

matic region. Considering only terrestrial animals on elevational gradients (545 species)

showed a significant effect of climate (p = 0.0023) after removing study dates, which had no significant effect. Considering birds alone (233 species) and including climatic region, survey

type, and study dates showed that climatic region, survey type, start date, and end date had sig-

nificant effects. For insects (271 species), when including climatic region, study dates, and sur-

vey type, no variables were significant. For fish (69 species), a model including habitat

(freshwater versus marine), study dates, and survey type showed that no variables were signifi-

cant. However, habitat was significant if other variables were removed. Similarly, for temperate

animals (367 species), a model including habitat, survey type, and study dates showed that

only habitat and survey type were significant. Comparison of plants and animals on terrestrial

elevational gradients (including study dates) showed that extinction is significantly different

between temperate plants and animals (more common in animals), but not between tropical

ones. Across animals, the effects of taxonomic group were limited and depended on the other

variables included. If only taxonomic groups and study dates were included, then annelids,

fish, and insects showed significantly more extinction (p = 0.03–0.05). Including habitat and survey type (and removing study dates) showed stronger effects in fish and annelids (as well as

in crustaceans and molluscs), but not in insects.

Fig 3. The frequency of local extinctions related to climate change across different taxonomic groups. The percentage of species with

one or more local extinctions in each taxonomic group is shown, along with the total sample size of species surveyed in that group. For ease of

presentation, four different groups of marine invertebrates (annelids, crustaceans, molluscs, and echinoderms) are shown together. Frequencies

for these four groups were averaged to obtain a single value, and sample sizes of species across groups were summed. Squamate reptiles

include lizards and snakes. Vertical lines indicate 95% confidence intervals on the estimated frequency of species with local extinctions.

doi:10.1371/journal.pbio.2001104.g003

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 8 / 18

Results were also broadly similar using GLMMs, with study identity included as a random

effect. Results are summarized below and given in full in S3 Appendix. The impacts of study

dates were somewhat counterintuitive (and rarely significant), and analyses including them

sometimes failed. When most variables were included (habitat, climatic region, taxonomic

group [plant versus animal], survey type, and study dates), all variables were significant except

for study dates and taxonomic group, with strong effects of habitat, climatic region, and survey

type. When study dates were removed, only habitat and survey type were significant. When

geographic regions were included (and study dates excluded), only South America had a sig-

nificant effect, and habitat, taxonomic group, and climatic region were significant or margin-

ally significant. Comparing tropical and temperate species on terrestrial, elevational gradients

showed significant effects of climatic region (p = 0.0017) and taxonomic group (p = 0.0119), but not of study dates. When study dates were removed, no variables were significant. Plants

alone showed a significant effect of climatic region (p < 0.0001), but analyses failed if study dates were included. Animals on terrestrial, elevational gradients showed no significant effect

of climatic region (again, study dates had to be excluded). Considering birds alone showed no

significant effect of climate but a significant effect of survey type (excluding study dates).

Insects showed no significant effects of climate or survey type, regardless of whether study

dates were included. Analyses of fish failed unless study dates and survey type were excluded,

but habitat alone (marine versus freshwater) had a significant effect (p = 0.0265). Analyses of temperate animals (367 species) including habitat, survey type, and study dates showed only

habitat type as significant (p = 0.0307), but excluding study dates showed significant effects of habitat and survey type. Comparing only temperate plants and animals showed a significant

effect of taxonomic group, when study dates were included (p = 0.0116) or excluded (p = 0.0005; study dates had no significant effect). In contrast, there was no significant effect of taxonomic group when comparing tropical plants and animals (504 species total; excluding

study dates). Analyses of animals alone showed no significant effect of taxonomic group.

In summary, several patterns emerged as significant across all (or most) analyses. First,

there were significant effects of climatic region overall, with extinction more common in tropi-

cal regions. This was present in plants across all analyses and generally present in animals. Ani-

mals showed significantly more extinction than plants overall and when comparing temperate,

but not tropical, species. There were significant effects of habitat on animals overall (higher

extinction in freshwater), even when considering fish alone. Finally, GLM analyses showed

some effects of taxonomic groups across animals (with higher extinction in fish and annelids)

and possibly in insects, molluscs, and crustaceans. The GLMM analyses did not show these

group effects, possibly because many animal groups are included based on a single study.

Discussion

The results of this study show that local extinctions (inferred to be related to climate change) are

already widespread and have occurred in hundreds of species. Roughly half of the 976 species

that were surveyed for range shifts showed evidence of local extinctions (47%). This proportion

was surprisingly similar across diverse climatic regions, habitats, and taxonomic groups. The

results here suggest that even the modest changes in climate that have occurred so far are enough

to drive local populations in many species to extinction. The results here also suggest that local

populations in many species cannot shift their climatic niches rapidly enough to prevent extinc-

tion. This pattern of widespread local extinction seems likely to become even more prevalent as

the global climate warms further (by roughly 2 to 5-fold [4]) in the next several decades.

The results here showed generally similar patterns of local extinction across climatic zones,

habitats, and clades. Nevertheless, most analyses showed that local extinctions were

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 9 / 18

significantly more common in tropical species (Fig 2A), in freshwater species (Fig 2B), and in

animals. A greater impact of climate change on tropical species has been predicted by several

authors (e.g., [40–42]). This prediction is related to the narrower climatic niche widths for

temperature-related variables in tropical species that are associated with reduced temperature

seasonality in the tropics (e.g., [43,44]) and lower rates of temperature-related climatic niche

change in tropical species (e.g., [42]). The results here provide support for this prediction

based on documented local extinctions that have already occurred: species in tropical regions

had local extinctions more frequently than those in temperate regions (54.6% versus 39.2%),

especially when species were compared on terrestrial, elevational gradients (54.6% versus

28.2%). This pattern was strongest in plants and when animals were compared on terrestrial

elevational gradients. Overall, these results further support the idea that the negative impacts

of climate change on biodiversity are more frequent (per species) in tropical regions [40–42],

where biodiversity is highest.

Climate-related local extinctions were also similar in frequency in marine and terrestrial spe-

cies (Fig 2B) but were more common in freshwater species (although freshwater habitats were

represented by a single study). Freshwater species may be especially susceptible to changes in

precipitation patterns (e.g., drought), which can substantially alter or eliminate their habitats

(e.g., [45]), quickly resulting in local extinction. In contrast, marine species may experience less

impact from changes in precipitation. Furthermore, they may be buffered from temperature

changes because they can potentially adjust the temperatures that they experience by movement

within the water column (more so than is possible for most freshwater species; [46,47]).

The frequency of local extinctions was also broadly similar across diverse taxonomic groups

(~35%–60%; Fig 3), including plants, insects, fish, amphibians, squamate reptiles, endothermic

vertebrates (birds and mammals), and many marine invertebrates (annelids, crustaceans, and

molluscs). However, local extinctions were significantly more common in animals than plants

(and animals are far more species-rich than plants). They were also relatively common in

insects (the most species-rich group of animals) and fish (the most species-rich group of verte-

brates). Local extinctions were not particularly common in amphibians (36.7%) or squamate

reptiles (41.7%), although both groups were included here based primarily on one study [32].

Nevertheless, both groups appear to have been strongly impacted by climate change overall.

For example, many amphibian species have undergone sharp declines and global extinctions,

many of which are thought to be caused by an interaction between climate change and an

infectious disease (chytrid fungus; [48]). However, these chytrid studies were not included

here because they were not focused on surveying warm-edge populations over time. Similarly,

local extinctions related to climate change have been documented in many lizard species [49].

Again, these were not included here because they were not based on a systematic survey of

warm-edge populations. Nevertheless, if the species studied by Sinervo et al. [49] were

included here, the frequency of local extinctions in squamates would go from 41.7% (of 12 spe-

cies) to 77.4% (of 124 species), but with the caveat that their study focused on documenting

local extinctions and so might overestimate this frequency. It should also be noted that the

well-publicized declines in amphibian populations globally are not necessarily inconsistent

with the frequency of local extinction observed here. For example, a global assessment of

amphibian populations [50] noted declines in 43% of amphibian species (compare to the 47%

of all species here with local extinctions and the 37% for amphibians), but these declines also

included those unrelated to climate change (e.g., habitat destruction and overexploitation).

Thus, the frequency of climate-related declines here is not necessarily an underestimation rela-

tive to the declines documented by the global amphibian assessment [50].

A major conclusion of this study is that populations of many species are already unable to

undergo niche shifts that are fast enough to prevent local extinction from climate change. The

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 10 / 18

rate is emphasized here because even if the absolute amount of niche change needed to avoid

extinction might be attainable, it might require more time to achieve than is allowed by the

rapid pace of anthropogenic climate change. Given this result, and that climate is predicted to

change even further in the near future, the persistence of many species might depend largely

on their ability to successfully shift their geographic ranges to higher latitudes or elevations

and remain within their original climatic niche. Indeed, the summary here shows numerous

instances of cool-edge expansions (in 367 of 904 species, with cool edges that were stable in

371 others and contracted in 166 others).

Unfortunately, these movements may be impeded for many species by one or more factors.

First, human impacts may prevent species from successfully dispersing (including agriculture,

roads, and urbanization), or these human impacts may simply leave them no habitat to dis-

perse to (e.g., [51,52]). Second, many species are already confined to islands, peninsulas, and

mountaintops, where dispersal to higher latitudes or elevations may not be possible (e.g., [53]).

Third, even if dispersal is unimpeded by human or natural barriers, it may simply occur too

slowly to allow species to remain within their climatic niche (e.g., [54,55]).

The combination of these potential limits to dispersal and the widespread local extinctions

documented here is troubling. However, the results here do not rule out the possibility that

rapid niche shifts will occur in some populations of many species in the future, preventing

global extinctions. Indeed, roughly half of the species surveyed showed no local extinctions,

and most species had some populations that persisted locally (but again, this is under the lim-

ited climate change that has already occurred). The future persistence of species will depend

on many factors [6,8], including rates and patterns of climate change at each location, dis-

persal, niche shifts, local climatic microrefugia [56], and the contribution of population-level

niche width to species-level niche width (e.g., whether species are broadly tolerant or locally

specialized to different climatic conditions across their ranges [44]). Most importantly, I sug-

gest that the patterns of present-day local extinctions obtained from range-shift studies should

be part of the evidence used to predict species persistence in the future.

There are several potential sources of bias that may have influenced some aspects of these

results but should not overturn the major conclusions. First, “local extinction” means that

individuals of a given species are entirely absent from a location that they previously occupied.

However, it can be difficult to distinguish between extinction and a substantial decline in

abundance that causes the species to go undetected at a given location (e.g., [57]), and studies

did not necessarily provide statistical evidence for the absence of a species at a site. Here, the

estimates of previous researchers were used, and it was assumed that they adequately docu-

mented local absences (otherwise, their estimates of range shifts would also be erroneous).

Furthermore, strong declines that make a species undetectable at a given site might soon lead

to local extinction. Second, there may be a bias in terms of unpublished results. Specifically,

some researchers who monitored the warm edge of a population but failed to find any changes

associated with climate change may not have published their negative results. Such a reporting

bias would lead to overestimating the proportion of species experiencing local extinction in

this study. Nevertheless, local extinctions were still documented in hundreds of species across

regions and clades, even if there are hundreds of additional species in which these local extinc-

tions did not occur. Additionally, numerous species (n = 171) showed evidence of a cool-edge expansion without a corresponding contraction in the warm edge. Thus, a species can undergo

a range shift but without local extinction, which should limit this source of publication bias.

Third, it was assumed that previous researchers correctly associated the patterns that they

observed with climate change. In theory, other factors such as overharvesting or habitat

destruction may have contributed to the observed local extinctions in some cases (e.g., [21]).

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 11 / 18

Again, the analyses here primarily assume that the main conclusions of these previous studies

were not erroneous.

Finally, despite the widespread pattern of warm-edge contractions and local extinctions,

521 species showed no local extinctions at the warm edge, indicating that they have success-

fully persisted in the face of the climate change that has occurred so far. However, even these

species might still go globally extinct when global climate changes further. Additionally, con-

trary to the overall trend, 54 species were documented here as having expansions at both their

warm edge and their cool edge (6.0% of 904 species with data on both cool and warm edges).

One scenario by which this may occur is if cool-edge limits are set by colder temperatures

(allowing expansion as global climate warms) and warm-edge limits are set by low precipita-

tion (allowing warm-edge expansion), given that precipitation may increase in some areas

because of climate change [4]. Indeed, some studies have found evidence for warm-edge

expansions through this mechanism [58]. It is also important to note that local extinctions

related to climate change need not be confined to the warm edge of the species range and so

might actually be underestimated here. For example, there could be climate-related local

extinctions far from the warm edge that are associated with certain microclimates (e.g., equato-

rially facing slopes at the cool edge of a species range; [59]).

In summary, the results here show that widespread local extinctions (seemingly related to

climate change) have already occurred in hundreds of species, with broadly similar patterns of

extinction across diverse clades, habitats, and climatic regions. Importantly, levels of climate

change so far are limited relative to those generally predicted for the next 100 years [4]. The results

here suggest that many species are unable to shift their niches rapidly enough to prevent local

extinction. This inference of climate change outpacing niche change supports predictions from

other sources, including transplant experiments in plants [60], phylogenetic analyses of rates of

niche change in plants and animals [42,61,62], and projections based on selection, heritability,

and temperature tolerances in lizards [49]. Local extinctions from climate change might also

impact species that many human populations depend on for food, such as grasses (e.g., wheat,

rice, and corn [62]). More generally, this study demonstrates that analyses of range shifts can pro-

vide extensive data on local extinctions related to climate change that have already occurred.

These local extinctions offer a potentially important but underutilized source of information for

the challenging task of predicting patterns of species survival and extinction in the future.

Materials and Methods

Selection of Studies

Web of Science searches were initially conducted from December 2014 to April 2015 using the

Boolean search terms Topic = (global warming OR climate change) AND Topic = (local

extinction OR range contraction OR range shift). A second Web of Science search was con-

ducted between April 2015 and May 2015 to identify additional studies potentially missed by

the first set of keywords, using the search terms TS = (global warm� OR climate change) AND

TS = (extinction� OR contraction� OR range shift�), excluding results from TS = (global

warming OR climate change) AND TS = (local extinction OR range contraction OR range

shift). Each set of Web of Science results was sorted by relevance and then binned into subsets

of 50. Searching was ceased when less than 1 in 50 studies per subset was relevant (see below

for criteria). Finally, a third Web of Science search was performed on 1 March 2016 to find

more recently published studies. This third search used the keywords TS = (global warm� OR

climate change) AND TS = (extinction� OR contraction� OR range shift�). A total of 1,530

results were found in this third search. Results were sorted by relevance, and the first 300

(~20%) were examined. The last 40 of these 300 included no relevant studies.

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 12 / 18

Some additional studies were also found that were listed as references in the papers identi-

fied by these initial Web of Science searches. The reference list was also checked against a

recent review study [11], which also conducted thorough searches of the literature on climate-

related range shifts. Three studies were added from that survey which were not initially

included here. Finally, several relevant studies were also found in the survey of Gibson-Rene-

mer et al. [63], which had similar rules for inclusion of studies. Although those authors did not

conduct a systematic search of the literature (as done here), they nevertheless included five

studies not found in the searches described above. These were also added here.

In theory, the fact that “extinction” and “contraction” were included as keywords might

have biased the results to include more papers documenting local extinctions and range con-

tractions than would be obtained from a search of range-shift studies that excluded these as key-

words (possibly leading to overestimation of the frequency of local extinctions). However, this

seems unlikely in practice. First, these were included as “or” keywords, along with “range shifts.”

Examining the keywords and titles of the 27 selected papers showed that most were focused on

overall range shifts, with no mention of local extinction (extinction or extirpation are men-

tioned in the titles of only 4 of 27 studies and as keywords in only 4 of the 21 studies with key-

words; “contraction” is mentioned in only 1). Furthermore, the fact that the survey results here

were checked against another recent review on range shifts [11], and that three missing studies

were added, also makes this potential bias seem unlikely. In other words, if many range-shift

studies were missed because of this bias, they should have been added at that point.

Overall, these searches were extensive but may not be truly exhaustive. Regardless, many

studies were found that documented local extinctions, and finding more studies that did so

would not overturn this main conclusion.

Studies were included that monitored one or more populations at the warm edge of a spe-

cies’ range (the edge that is lower in elevation or closer to the equator) over a relatively long

time span. Studies were only included that spanned an interval of at least 10 years. The mean

study duration was ~50 years (range = 14 to 159; Table 1). Studies were included that related

their findings on range shifts to climate change through an explicit statistical analysis (but not-

ing that these inferences could still be incorrect, for example, if other factors instead of climate

change caused local extinctions of a particular species). The included studies all documented

populations along elevational or latitudinal transects at two or more discrete time points.

Some recent studies have inferred climate-related range shifts based on overall trends in lat-

itudinal and elevational distributions across a large number of localities over time, rather than

systematically resurveying specific localities at different time points (e.g., [64]). These studies

are valuable for documenting range shifts in general but were excluded here, since they do not

unambiguously represent local extinctions (because the overall patterns described might be

driven solely by range expansions instead).

Categorizing Species

Studies that documented warm-edge range contractions (and that were linked to climate

change by the authors of the original studies) were considered evidence of climate-associated

local extinction, regardless of changes at the cool edge. Studies differed in whether they

reported changes at the population level (e.g., [28,37]) or species level (e.g., [33]). The analysis

here was conducted at the species level. Therefore, if populations of the same species differed

in the pattern of their range shifts, the species was categorized as showing evidence of local

extinction if at least one population did so.

Most species were included in only one study. However, the plant species Anthoxanthum odoratum was included by both Angelo and Daehler [13] (in Hawaii) and Felde et al. [19] (in

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 13 / 18

Europe). However, since this species is not native to Hawaii, it was excluded from the dataset

of Angelo and Daehler [13], along with all other nonnative species in that study.

For each study, it was noted whether the range shifts were elevational or latitudinal, as well

as the general habitat of the organisms (i.e., terrestrial, freshwater, or marine), the higher taxa

to which they belonged, the specific geographic location of the study, and whether the species

occurred in a tropical or subtropical region (arbitrarily defined as within 35˚ of the equator) or

in a temperate region (>35˚). Species were assigned to these climatic regions based solely on

the location where they were surveyed, rather than on their overall geographic range. Species

were also assigned to taxonomic categories, including plants, insects, fish, amphibians, birds,

mammals, and squamate reptiles (i.e., lizards and snakes), as well as marine annelids, crusta-

ceans, echinoderms, and molluscs. The beginning and end dates of the study were also noted

(e.g., the date of the initial survey and the subsequent resurvey) and were used to estimate the

duration of the study. Some studies provided a range of dates for the start and/or end date. In

these cases, the midpoint of each range of dates was used to estimate the start, end, and dura-

tion (Table 1). Data for all species are provided in S1 Appendix.

The studies included (Table 1) spanned many geographic regions (e.g., North America,

South America, Europe, Asia, and Oceania). Many studies were conducted in North America

(n = 13; here extending to Central America) and Europe (n = 8), but the actual number of spe- cies sampled was more broadly distributed among regions (e.g., Asia = 332; Europe = 268;

Madagascar = 30; Oceania = 58; North America = 233; and South America = 55). Africa and

Australia were not represented, although nearby Madagascar and New Guinea were. The num-

bers of temperate and tropical species included were nearly equal. Further, there was no clear

hypothesis for why particular continents alone should be an important factor influencing the

frequency of local extinctions (e.g., separate from temperate versus tropical effects).

Statistical Analyses

Chi-squared analyses were initially used to compare the proportion of climate-associated local

extinctions across some categories (i.e., tropical versus temperate; freshwater versus marine

versus terrestrial; and latitudinal versus elevational gradients), testing the null hypothesis that

frequencies of local extinction were equal between these categories. A series of analyses were

conducted to assess whether frequencies of local extinction were higher in tropical regions rel-

ative to temperate regions, after accounting for the potential influence of different habitats,

gradients, and clades (see Results). Similar analyses were conducted to assess the impacts of

different habitats and clades (i.e., plants versus animals). However, potential analyses were

restricted by the available data. For example, it was not possible to compare the effect of tropi-

cal versus temperate climates on marine or freshwater organisms, since only temperate marine

and freshwater species were included here. For this reason, different sets of analyses were con-

ducted for each question.

These analyses were then repeated using GLMs and GLMMs, both in R. These analyses

were implemented treating the presence of warm-edge local extinction in a species as the bino-

mial, dependent variable. GLMM analyses were conducted using the R package lme4 [65]. GLMM analyses treated the study (from which the species data were obtained) as the random

variable and the other variables as the fixed variables. GLM and GLMM analyses initially

included all species and all or most variables and were then restricted to smaller sets of species

(and variables) to test additional hypotheses and reduce potentially confounding effects (as in

the Chi-squared analyses).

Phylogenetic information was not incorporated here, since phylogenies and comparable

branch lengths spanning all the included species were not available (especially species-level

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 14 / 18

phylogenies for fish, insects, plants, and marine invertebrates). Nevertheless, some analyses

were conducted to assess patterns within and between clades (see Results).

Supporting Information

S1 Appendix. Data for the 976 species used in this study.

(XLSX)

S2 Appendix. Results of GLM analyses, showing variable coefficients.

(DOC)

S3 Appendix. Results of GLMM analyses, showing variable coefficients.

(DOC)

Acknowledgments

I am very grateful to D. Davison for her extensive work on the initial literature searches for

this study. I thank E. Beever, J. Lenoir, and A. Phillimore for many helpful comments that

greatly improved the manuscript.

Author Contributions

Conceptualization: John J. Wiens.

Investigation: John J. Wiens.

Writing – original draft: John J. Wiens.

Writing – review & editing: John J. Wiens.

References 1. Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ,Collingham YC, et al. Extinction risk

from climate change. Nature. 2004; 427:145–148. doi: 10.1038/nature02121 PMID: 14712274

2. Bellard C, Bertelsmeier C, Leadley P, Thuiller W, Courchamp F Impacts of climate change on the future

of biodiversity. Ecol Lett. 2012; 15:365–77. doi: 10.1111/j.1461-0248.2011.01736.x PMID: 22257223

3. Urban MC Accelerating extinction risk from climate change. Science. 2015; 348: 571–573. doi: 10.

1126/science.aaa4984 PMID: 25931559

4. Stocker TF et al. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to

the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. http://www.

climatechange2013.org/report/full-report

5. Holt RD The microevolutionary consequences of climate change. Trends Ecol Evol. 1990; 5:311–315.

doi: 10.1016/0169-5347(90)90088-U PMID: 21232381

6. Williams SE, Shoo LP, Isaac JL, Hoffmann AA, Langham G. Towards an integrated framework for

assessing the vulnerability of species to climate change. PLoS Biol. 2008; 6:2621–2626. doi: 10.1371/

journal.pbio.0060325 PMID: 19108608

7. Gienapp P, Teplitsky C, Alho JS, Mills JA, Merilä J. Climate change and evolution: disentangling envi-

ronmental and genetic responses. Mol Ecol. 2008; 17:167–178. doi: 10.1111/j.1365-294X.2007.03413.

x PMID: 18173499

8. Moritz C, Agudo R The future of species under climate change: resilience or decline? Science. 2013;

341:504–508. doi: 10.1126/science.1237190 PMID: 23908228

9. Parmesan C, Yohe G A globally coherent fingerprint of climate change impacts across natural systems.

Nature. 2003; 421:37–342. doi: 10.1038/nature01286 PMID: 12511946

10. Chen I-C, Hill JK, Ohlemüller R, Roy DB, Thomas CD Rapid range shifts of species associated with

high levels of climate warming. Science. 2011; 333: 1024–1026. doi: 10.1126/science.1206432 PMID:

21852500

11. Lenoir J, Svenning J-C Climate-related range shifts–a global multidimensional synthesis and new

research directions. Ecography. 2015; 38:15–28.

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 15 / 18

12. Cahill AE, Aiello-Lammens ME, Fisher-Reid MC, Hua X, Karanewsky CJ, Ryu HY, et al. How does cli-

mate change cause extinction? Proc R Soc Lond B. 2013; 280:2012890.

13. Angelo CL, Daehler CC. Upward expansion of fire-adapted grasses along a warming tropical elevation

gradient. Ecography. 2013; 36:551–559.

14. Beever EA, Ray C, Wilkening JL, Brussard PF, Mote PW Contemporary climate change alters the pace

and drivers of extinction. Global Change Biol. 2011; 17:2054–2070.

15. Brusca RC, Wiens JF, Meyer WM, Eble J, Franklin K, Overpeck JT, et al. Dramatic response to climate

change in the Southwest: Robert Whittaker’s 1963 Arizona Mountain plant transect revisited. Ecol Evol.

2013; 3:3307–3319. doi: 10.1002/ece3.720 PMID: 24223270

16. Chen IC, Hill JK, Shiu HJ, Holloway JD, Benedick S, Chey VK, et al. 2011 Asymmetric boundary shifts

of tropical montane Lepidoptera over four decades of climate warming. Global Ecol. Biogeogr. 2011;

20:34–45.

17. Comte L, Grenouillet G. Do stream fish track climate change? Assessing distribution shifts in recent

decades. Ecography. 2013; 36:1236–1246.

18. Dieker P, Drees C, Assmann T Two high-mountain burnet moth species (Lepidoptera, Zygaenidae)

react differently to the global change drivers climate and land-use. Biol Conserv. 2011; 144:2810–2818.

19. Felde VA, Kapfer J, Grytnes J-A Upward shift in elevational plant species ranges in Sikkilsdalen, central

Norway. Ecography. 2012; 35:922–932.

20. Forero-Medina G, Terborgh J, Socolar SJ, Pimm SL Elevational ranges of birds on a tropical montane

gradient lag behind warming temperatures. PLoS ONE. 2011; 6:e28535. doi: 10.1371/journal.pone.

0028535 PMID: 22163309

21. Franco AMA, Hill JK, Kitschke C, Collingham YC, Roy DB, Fox R, et al. Impacts of climate warming and

habitat loss on extinctions at species’ low-latitude range boundaries. Global Change Biol. 2006;

12:1545–1553.

22. Freeman BG, Freeman AMC Rapid upslope shifts in New Guinean birds illustrate strong distributional

responses of tropical montane species to global warming. Proc Natl Acad Sci USA. 2014; 111:4490–

4494. doi: 10.1073/pnas.1318190111 PMID: 24550460

23. Hiddick JG, Burrows MT, Garcia Molinos J Temperature tracking by North Sea benthic invertebrates in

response to climate change. Global Change Biol. 2015; 21:117–129,

24. Hitch AT, Leberg PL Breeding distributions of North American bird species moving north as a result of

climate change. Conserv Biol 2007; 21:534–539. doi: 10.1111/j.1523-1739.2006.00609.x PMID:

17391203

25. Menendez R, Gonzalez-Megias A, Jay-Robert P, Marquez-Ferrando R Climate change and elevational

range shifts: evidence from dung beetles in two European mountain ranges. Glob Ecol Biogeogr. 2014;

23:646–657.

26. Moritz C, Patton JL, Conroy CJ, Parra JL,White GC, Beissinger SR Impact of a century of climate

change on small-mammal communities in Yosemite National Park, USA. Science. 2008; 322:261–264.

doi: 10.1126/science.1163428 PMID: 18845755

27. Myers P, Lundrigan BL, Hoffman SMG, Haraminac AP, Seto SH Climate-induced changes in the small

mammal communities of the Northern Great Lakes Region. Global Change Biol. 2009; 15:1434–1454.

28. Nye JA, Link JS, Hare JA, Overholtz WJ Changing spatial distribution of fish stocks in relation to climate

and population size on the Northeast United States continental shelf. Mar Ecol Prog Ser. 2009;

393:111–129.

29. Perry AL, Low PJ, Ellis JR, Reynolds JD Climate change and distribution shifts in marine fishes. Sci-

ence. 2005; 308:1912–1915. doi: 10.1126/science.1111322 PMID: 15890845

30. Ploquin EF, Herrera JM, Obeso JR Bumblebee community homogenization after uphill shifts in mon-

tane areas of northern Spain. Oecologia. 2013; 173:1649–60. doi: 10.1007/s00442-013-2731-7 PMID:

23852029

31. Pomara LY, LeDee O, Martin KJ, Zuckerberg B Demographic consequences of climate change and

land cover help explain a history of extirpations and range contraction in a declining snake species.

Global Change Biol. 2014; 20:2087–2099. doi: 10.1111/gcb.12510 PMID: 24357530

32. Raxworthy CJ, Pearson RG, Rabibisoa N, Rakotondrazafy AM, Ramanamanjato JB, Raselimanana

AP, et al. Extinction vulnerability of tropical montane endemism from warming and upslope displace-

ment: a preliminary appraisal for the highest massif in Madagascar. Global Change Biol. 2008;

14:1703–1720.

33. Rowe RJ, Finarelli JA, Rickart EA Range dynamics of small mammals along an elevational gradient

over an 80-year interval. Global Change Biol. 2010; 16:2930–2943

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 16 / 18

34. Rubal M, Veiga P, Cacabelos E, Moreira J, Sousa-Pintos I Increasing sea surface temperature and

range shifts of intertidal gastropods along the Iberian Peninsula. J Sea Res. 2013; 77:1–10.

35. Sheldon AL Possible climate-induced shift of stoneflies in a southern Appalachian catchment. Freshw

Sci. 2012; 31:765–774.

36. Telwala Y, Brook BW, Manish K, Pandit MK Climate-induced elevational range shifts and increase in

plant species richness in a Himalayan biodiversity epicentre. PLoS One. 2013; 8:e57103. doi: 10.1371/

journal.pone.0057103 PMID: 23437322

37. Tingley MW, Koo MS, Moritz C, Rush AC, Beissinger SR The push and pull of climate change causes

heterogeneous shifts in avian elevational ranges. Global Change Biol. 2012; 18:3279–3290.

38. Warren RJ, Chick L Upward ant distribution shift corresponds with minimum, not maximum, tempera-

ture tolerance. Global Change Biol. 2013; 19:2082–2088

39. Zuckerberg B, Woods AM, Porter WF Poleward shifts in breeding bird distributions in New York State.

Global Change Biol. 2009; 15:1866–1883.

40. Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak DC, et al. Impacts of climate

warming on terrestrial ectotherms across latitude. Proc Natl Acad Sci USA. 2008; 105:6668–6672. doi:

10.1073/pnas.0709472105 PMID: 18458348

41. Huey RB, Deutsch CA, Tewksbury JJ, Vitt LJ, Hertz PE, Álvarez Pérez HJ, et al. Why tropical forest liz-

ards are vulnerable to climate warming. Proc R Soc Lond B. 2009; 276:1939–1948.

42. Jezkova T, Wiens JJ Rates of change in climatic niches in plant and animal populations are much

slower than projected climate change. Proc R Soc Lond B. 2016;20162104.

43. Janzen DH Why mountain passes are higher in the tropics. Am Nat. 1967; 101:233–249.

44. Quintero I, Wiens JJ What determines the climatic niche width of species? The role of spatial and tem-

poral climatic variation in three vertebrate clades. Global Ecol Biogeogr. 2013; 22:422–432

45. Koehn JD, Hobday AJ, Pratchett MS, Gillanders BM Climate change and Australian marine and fresh-

water environments, fishes and fisheries: synthesis and options for adaptation. Mar Freshw Res. 2011;

62:1148–1164.

46. Walther GN, Post E, Convey P, Menzel A, Parmesan C, Beebee TC, et al. Ecological responses to

recent climate change. Nature. 2002; 416:389–395. doi: 10.1038/416389a PMID: 11919621

47. Dulvy NK, Rogers SI, Jennings S, Stelzenmuller V, Dye SR, Skjoldal HR Climate change and deepen-

ing of the North Sea fish assemblage: a biotic indicator of warming seas. J Appl Ecol. 2008; 45:1029–

1039.

48. Rohr JR, Raffel TR Linking global climate and temperature variability to widespread amphibian declines

putatively caused by disease. Proc. Natl Acad Sci USA. 2010; 107:8269–8274. doi: 10.1073/pnas.

0912883107 PMID: 20404180

49. Sinervo B, Méndez-de-la-Cruz F, Miles DB, Heulin B, Bastiaans E, Villagrán-Santa Cruz M, et al. Ero-

sion of lizard diversity by climate change and altered thermal niches. Science. 2010; 328:894–899. doi:

10.1126/science.1184695 PMID: 20466932

50. Stuart S, Chanson JS, Cox NA, Young BE, Rodrigues ASL, Fischman DL, et al. Status and trends of

amphibian declines and extinctions worldwide. Science. 2004; 306:1783–1786. doi: 10.1126/science.

1103538 PMID: 15486254

51. Brook BW, Sodhi NS, Bradshaw CJA Synergies among extinction drivers under global change. Trends

Ecol Evol. 2008; 23:453–460. doi: 10.1016/j.tree.2008.03.011 PMID: 18582986

52. Klausmeyer KR, Shaw MR Climate change, habitat loss, protected areas and the climate adaptation

potential of species in Mediterranean ecosystems worldwide. PLoS ONE. 2009; 4:e6392. doi: 10.1371/

journal.pone.0006392 PMID: 19641600

53. Colwell RK, Brehm G, Cardelús CL, Gilman AC, Longino JT Global warming, elevational range shifts,

and lowland biotic attrition in the wet tropics. Science. 2008; 322:258–261. doi: 10.1126/science.

1162547 PMID: 18845754

54. Loarie SR, Duffy PB, Hamilton H, Asner GP, Field CB, Ackerly DD. The velocity of climate change.

Nature. 2009; 462:1052–1055. doi: 10.1038/nature08649 PMID: 20033047

55. Schloss CA, Nunez TA, Lawler JJ Dispersal will limit ability of mammals to track climate change in the

Western Hemisphere. Proc Natl Acad Sci USA. 2012; 109:8606–8611. doi: 10.1073/pnas.1116791109

PMID: 22586104

56. Dobrowski SZ A climatic basis for microrefugia: the influence of terrain on climate. Global Change Biol.

2010; 17:1022–1035.

57. Tingley MW, Beissinger SR Detecting range shifts from historical species occurrences: new perspec-

tives on old data. Trends Ecol Evol. 2009; 24:625–633. doi: 10.1016/j.tree.2009.05.009 PMID:

19683829

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 17 / 18

58. Crimmins SM, Dobrowski SZ, Greenberg JA, Abatzoglou JT, Mynsberge AR Changes in climatic water

balance drive downhill shifts in plant species’ optimum elevations. Science. 2011; 331:324–327. doi:

10.1126/science.1199040 PMID: 21252344

59. Lenoir J, Svenning J-C Latitudinal and elevational range shifts under contemporary climate change. In

Levin SA, editor. Encyclopedia of biodiversity, 2nd edition. Elsevier, 2013. pp 599–611.

60. Etterson JR, Shaw RG Constraint to adaptive evolution in response to global warming. Science. 2001;

294:151–154. doi: 10.1126/science.1063656 PMID: 11588260

61. Quintero I, Wiens JJ Rates of projected climate change dramatically exceed past rates of climatic niche

evolution among vertebrate species. Ecol Lett. 2013; 16:1095–1103. doi: 10.1111/ele.12144 PMID:

23800223

62. Cang FA, Wilson AA, Wiens JJ. Climate change is projected to outpace rates of niche change in

grasses. Biol Lett. 2016; 12:20160368. doi: 10.1098/rsbl.2016.0368 PMID: 27677813

63. Gibson-Reinemer DK, Sheldon KS, Rahel FJ Climate change creates rapid species turnover in mon-

tane communities. Ecol Evol. 2015; 5:2340–2347. doi: 10.1002/ece3.1518 PMID: 26120424

64. Kerr JT, Pindar A, Galpern P, Packer L, Potts SG, Roberts SM, et al. Climate change impacts on bum-

blebees converge across continents. Science. 2015; 349:177–180. doi: 10.1126/science.aaa7031

PMID: 26160945

65. Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J. Stat. Softw.

2015; 67:1–48.

Climate Change and Extinction

PLOS Biology | DOI:10.1371/journal.pbio.2001104 December 8, 2016 18 / 18