Landscape fragmentation affects responses of aviancommunities to climate changeMARTA A JARZYNA1 W ILL IAM F PORTER 1 BR IAN A MAURER 1 2 B EN JAMIN
ZUCKERBERG3 and ANDREW O FINLEY4
1Department of Fisheries and Wildlife Michigan State University East Lansing MI 48824 USA 2Center for Statistical Training
and Consulting Michigan State University East Lansing MI 48824 USA 3Department of Forest and Wildlife Ecology
University of Wisconsin-Madison Madison WI 53706 USA 4Departments of Forestry and Geography Michigan State
University East Lansing MI 48824 USA
Abstract
Forecasting the consequences of climate change is contingent upon our understanding of the relationship between
biodiversity patterns and climatic variability While the impacts of climate change on individual species have been
well-documented there is a paucity of studies on climate-mediated changes in community dynamics Our objectives
were to investigate the relationship between temporal turnover in avian biodiversity and changes in climatic condi-
tions and to assess the role of landscape fragmentation in affecting this relationship We hypothesized that commu-
nity turnover would be highest in regions experiencing the most pronounced changes in climate and that these
patterns would be reduced in human-dominated landscapes To test this hypothesis we quantified temporal turn-
over in avian communities over a 20-year period using data from the New York State Breeding Atlases collected dur-
ing 1980ndash1985 and 2000ndash2005 We applied Bayesian spatially varying intercept models to evaluate the relationship
between temporal turnover and temporal trends in climatic conditions and landscape fragmentation We found that
models including interaction terms between climate change and landscape fragmentation were superior to models
without the interaction terms suggesting that the relationship between avian community turnover and changes in cli-
matic conditions was affected by the level of landscape fragmentation Specifically we found weaker associations
between temporal turnover and climatic change in regions with prevalent habitat fragmentation We suggest that
avian communities in fragmented landscapes are more robust to climate change than communities found in contigu-
ous habitats because they are comprised of species with wider thermal niches and thus are less susceptible to shifts in
climatic variability We conclude that highly fragmented regions are likely to undergo less pronounced changes in
composition and structure of faunal communities as a result of climate change whereas those changes are likely to be
greater in contiguous and unfragmented habitats
Keywords biodiversity bird communities climate change global change habitat fragmentation temporal turnover
Received 8 September 2014 and accepted 11 January 2015
Introduction
The ability to predict the ecological consequences of cli-
mate change and develop sound conservation strategies
depends on our understanding of the relationship
between changes in biodiversity patterns and climatic
variability To date the majority of research regarding
the implications of climate change to biodiversity
has evaluated responses of individual species (eg
Parmesan amp Yohe 2003 La Sorte amp Thompson 2007
Marini et al 2009 Zuckerberg et al 2009) The variabil-
ity of individual species responses is predicted to lead
to disruptions of communities and ecosystems (Brotons
amp Jiguet 2010) but the complex nature of ecological
interactions makes it difficult to extrapolate from the
scales of individuals to the community or ecosystem
level (Walther et al 2002) To fully understand conse-
quences of climate change however it is imperative to
develop a more comprehensive understanding on
whether climate change can lead to broadscale changes
in community structure and composition
Measuring community turnover has potential to offer
a useful and simple way of evaluating the implications
of climate change to ecological communities (eg
Tuomisto 2010ab Barton et al 2013) Community
turnover can be quantified in multiple ways but gener-
ally is an aggregate of gains and losses of species
comprising a community (eg Tuomisto 2010b) Tradi-
tionally turnover has been used to describe biodiver-
sity patterns across space rather than time (spatial
turnover) but it can also be viewed in terms of a
Correspondence Marta A Jarzyna Department of Ecology and
Evolutionary Biology Yale University 165 Prospect Street
New Haven CT 06520-8106 USA tel +1 203 432 1597
fax +1 203 432 5176
e-mail martajarzynayaleedu
2942 copy 2015 John Wiley amp Sons Ltd
Global Change Biology (2015) 21 2942ndash2953 doi 101111gcb12885
temporal change in community composition that has
occurred at the same site across a particular time period
(temporal turnover) Given that spatial turnover in
species composition is thought to be at least partially
related to spatial variation in environmental variables
(Gaston et al 2007 Buckley amp Jetz 2008 Gutierrez-
Canovas et al 2013 Siefert et al 2013) it is likely that
temporal turnover reflects temporal changes in climate
and could be a useful indicator of climate change
impacts on biodiversity
Even though the impacts of climate change on biodi-
versity are likely to be felt across a wide range of ecosys-
tems different communities are likely to respond to
climate change in disparate ways because a wide range
of environmental factors might influence speciesrsquo vulner-
abilities to climate change For example simulation stud-
ies suggest that the interaction between climate and land
cover will influence the way in which biodiversity
responds to climate change (Travis 2003 Jeltsch et al
2011) Specifically habitat fragmentation is thought to
affect the rate of distributional shifts associated with cli-
mate change allowing species in less fragmented habi-
tats to disperse faster and farther (Warren et al 2001
Opdam amp Wascher 2004) Lack of suitable habitat
between speciesrsquo current distribution and the future suit-
able climatic conditions might prevent a species with
some limiting intrinsic characteristics such as sedentary
life style or short dispersal distances from migrating to
these new climatically suitable locations If species in
contiguous (ie unfragmented) habitats shift their distri-
butions faster as a result of climate change than species
in fragmented habitats then changes in communities
should also follow the same pattern Thus stronger asso-
ciations of temporal turnover with changing climatic
conditions in contiguous habitats might be expected
In addition to more pronounced distributional shifts
other mechanisms might also lead or contribute to
stronger associations of temporal turnover with chang-
ing climate in unfragmented landscapes First commu-
nities in contiguous habitats tend to be comprised
largely of habitat specialists (Estavillo et al 2013)
which generally have narrower habitat and climatic
niches (Barnagaud et al 2012) and thus might be more
vulnerable to changing climatic conditions Conse-
quently the same amount of change in climatic condi-
tions would be expected to exert a stronger effect on
communities in contiguous landscapes Secondly frag-
mented habitats tend to support smaller populations
than contiguous unfragmented regions Small local
population sizes generally increase the effect of demo-
graphic or environmental stochasticity (Gaublomme
et al 2014) resulting in higher turnover in community
composition compared with that of regions supporting
large populations (Banks-Leite et al 2012)
While recent research suggests that climate change is
one of the factors causing temporal shifts in community
composition (Devictor et al 2008 Kampichler et al
2012 Prince amp Zuckerberg 2015) the influence of land-
cover and habitat fragmentation is still unclear For
example Devictor et al (2008) found that composition
of avian communities in France changed to include
higher proportion of high-temperature dwelling species
and suggested (although not explicitly tested) that this
response of communities to climate warming might dif-
fer between different types of land-cover and levels of
landscape fragmentation Kampichler et al (2012) sug-
gested a connection between land-cover and climate
change and bird community change by showing a
directional change in the proportion of high-tempera-
ture dwelling species as well as in the level of habitat
specialization of species comprising the community In
eastern North America Prince amp Zuckerberg (2015)
found that the composition of winter birds has become
increasingly dominated by warm-adapted species and
this community reshuffling was greatest in regions
experiencing long-term warming patterns However
despite these initial investigations the relationship
between temporal turnover in biodiversity and the
interaction of climatic changes and landscape fragmen-
tation to the best of our knowledge has not yet been
explicitly evaluated We intend to fill this research gap
by investigating the drivers of temporal turnover in
avian biodiversity across New York using Breeding
Bird Atlas data (Andrle amp Carroll 1988 McGowan amp
Corwin 2008) Our primary hypothesis was that tem-
poral turnover in avian assemblages is related to cli-
mate change but the strength of this relationship varies
with the degree of habitat fragmentation Specifically
we predict high rates of temporal turnover in regions
experiencing greater climate change and that these cli-
mate-mediated turnover rates would be higher in land-
scapes characterized by more contiguous habitat
Contiguous habitat is likely to facilitate higher rates of
turnover via three ecological mechanisms (i) allowing
species to disperse faster in response to climate change
(ii) supporting communities comprised of species more
vulnerable to changing climatic conditions and (iii)
increasing the ratio of deterministic to stochastic
processes
Materials and methods
Site description
The study area is the State of New York USA New York cov-
ers 128 401 km2 including 4240 km2 of inland water The cli-
mate of New York is affected by the statersquos broad elevation
gradient as well as by the proximity to lakes Erie and Ontario
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2943
and the Atlantic Ocean The Adirondack and Catskill Moun-
tains in the east are among the highest regions of the state
with elevation varying between 600 and 1500 m The elevation
of the southwestern New York ranges from approximately
300ndash900 m while northwestern New York as well as Hudson
River valley and New York City are low-lying with elevation
ranging from approximately 0ndash200 m
In general New York State is covered by approximately
512 forest (deciduous evergreen and mixed forests) 134
pasture and hay 82 cultivated crops 29 scrub and shrub
10 grassland and herbaceous 80 wetland cover 87
developed land and 02 barren land (Homer et al 2004)
New York offers a broad gradient of landscape fragmentation
The regions of the Adirondack and Catskill Mountains are the
least fragmented with expansive regions of forest land The
land cover of the Hudson River valley is fragmented by
residential and commercial development whereas the
landscapes of western New York are characterized mostly by
agriculturendashforest mosaic
Breeding Bird Atlas
We used the New York State Breeding Bird Atlas (BBA) to
characterize changes in avian communities through time BBA
is a statewide survey that documented the distribution of
breeding birds in New York To date BBA has been conducted
in two time periods 1980ndash1985 (hereafter BBA1980 Andrle amp
Carroll 1988) and 2000ndash2005 (hereafter BBA2000 McGowan
amp Corwin 2008) For both BBAs a grid system was used to
define the basic unit for reporting data The BBA reporting
unit (a block) was a square measuring 5 9 5 km a total of
5335 blocks covered the entirety of New York State This data-
set represents one of the largest and finest resolution atlases in
the world (Gibbons et al 2007)
A total of 242 species were recorded for BBA1980 and 248
species were recorded for BBA2000 (Andrle amp Carroll 1988
McGowan amp Corwin 2008 see Table S1 in Appendix S1 for
the list of all species used in this analysis) Observations were
made by skilled birders who spent at least 8 h in each block
visited all cover types in each block and included at least one
nighttime visit to document nocturnal species Observer effort
was recorded for each BBA and reported as a number of per-
son hours (measured as the sum of the number of hours spent
in each block x the number of people surveying each block
McGowan amp Zuckerberg 2008) The BBA represents a detec-
tionnondetection dataset nondetection indicates that species
could not be found given search criteria (McGowan amp Corwin
2008)
Habitat fragmentation
Habitat fragmentation was identified using the National Land
Cover Data (NLCD) derived from the Landsat Thematic Map-
per satellite data (Homer et al 2004) Even though the NLCD
are available for four time periods (ie 1992 2001 2006 and
2011) there is no land-cover data readily available for the time
period of BBA1980 Thus to assess whether temporal turnover
is related to landscape fragmentation and whether landscape
fragmentation affects responses of communities to climate
change we used NLCD 2001 To ensure that land-cover or
land-use change could not have significantly affected changes
in avian community composition over the time period of
1980ndash2005 we used the NLCD 19922001 Retrofit Land Cover
Change Product (Fry et al 2009) We quantified the amount of
land-cover change that occurred between 1992 and 2001 and
found that on average only 17 of the BBA blockrsquos area chan-
ged over this time period (median = 13 min = 0
max = 216) and approximately 97 of all blocks underwent
land-cover change of less than 5 (see Fig S1 in Appendix
S1) We thus felt confident that land-cover change had minor
if any effect on changes in avian community composition
Prior to landscape analysis we consolidated the following
land-cover classes open water and perennial ice and snow
classes into one class of waterice open space developed low
intensity developed medium intensity developed high inten-
sity developed into one class of developed land cultivated
crops and pasturehay into one class of agriculture and
deciduous forest evergreen forest mixed forest and forested
wetland into one class of forest Consolidation of these land-
cover types improved accuracy of classification and simplified
the environment for purposes of our evaluation Upon consoli-
dation we examined the following land-cover classes water
ice developed barren land forest scrubshrub grassland
herbaceous agriculture and wetlands
We used FRAGSTATS 41 (McGarigal et al 2012) and the
Geospatial Modelling Environment (GME httpwwwspa-
tialecologycomgme) to quantify landscape fragmentation
Because we focused our analysis on a diverse suite of species
with varying habitat requirements we chose a landscape-scale
variable to capture broadscale variation in habitat fragmenta-
tion We chose edge density (ED) as a measure of landscape
fragmentation because an increase in habitat edge is a primary
outcome of habitat fragmentation (Hargis et al 1998) ED was
also reported as an effective tool for evaluating landscape
fragmentation and performed better than other landscape
fragmentation indices (Hargis et al 1998) However in situa-
tions when landscape consists entirely of one cover type the
ED would be 0 regardless of the type of land cover present
Therefore to differentiate between landscapes that consisted
entirely of natural land cover (eg forest) and those mostly
developed (eg cities) we also calculated the proportion of
developed land (DEVEL) metric and included it as a variable
in our models (see Fig S2 in Appendix S1 for maps of all
explanatory variables)
Climatic trends
The climate data were derived from the PRISM (Parameter-ele-
vation Regressions on Independent Slopes Model) climate
mapping system (Daly amp Gibson 2002) PRISM consists of inter-
polated monthly maximum and minimum temperatures and
precipitation at a 25-arcmin resolution from 1891 to 2010 for
the entire contiguous United States We calculated the magni-
tude of the 25-year (1980ndash2005) trend in average monthly max-
imum and minimum temperatures and in total monthly
precipitation using ordinary least squares regression The
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2944 M A JARZYNA et al
slope of the OLS regression indicated the magnitude of the
trend and reflected the amount of change in climatic variables
that occurred between 1980 and 2005 We then interpolated
the trend magnitudes for each BBA block and averaged the
monthly values to reflect average breeding season (May
through September) trend magnitude in maximum and mini-
mum temperatures (TMAXTrend and TMINTrend respec-
tively) and in total precipitation (PRECIPTrend) TMAXTrend
and TMINTrend were expressed in degC25 years PRECIP-
Trend was expressed in mm25 years An average increase in
maximum temperatures of the breeding season was 063degC(median = 065 degC) over the 1980ndash2005 time period while an
average increase in minimum temperatures was 141 degC(median = 141 degC) over the same time period On average
precipitation increased by 10 mm (median = 10 mm) over the
1980ndash2005 time period (see Figs S2 and S3 in Appendix S1 for
maps and histograms of all explanatory variables)
Other model covariates
We included elevation in all the models to account for topo-
graphical variation of New York State We used digital eleva-
tion models for New York State to quantify mean elevation in
each block (ELEV) Additionally increasing survey effort often
results in a higher number of recorded species (Tobler et al
2008) Drastically different survey effort between BBA1980
and BBA2000 could increase the values of temporal turnover
which could be a result of different number of recorded spe-
cies rather than actual changes in species identities through
time To account for potential survey effort bias for each BBA
block we calculated the difference in the number of person
hours between BBA2000 and BBA1980 standardized by the
number of person hours in BBA1980 (EFF = (EFF2000 ndashEFF1980)EFF1980) and included it as a covariate in our
models We did not detect significant correlation among any
of our explanatory variables (Fig S3 in Appendix S1)
Statistical analysis
We quantified temporal turnover (TURN) as an aggregate of
species losses (ie species recorded at a particular site during
BBA1980 but no longer detected during BBA2000) and species
gains (ie species not detected at a particular site during
BBA1980 but recorded during BBA2000) within a particular
site Specifically temporal turnover was calculated as follows
TURN =Ethorn C
Ethorn Cthorn Peth1THORN
where E is the number of species lost in a particular BBA block
between BBA1980 and BBA2000 C is the number of species
gained in a particular BBA block between BBA1980 and
BBA2000 and P is number of species present in a particular
BBA block during both BBAs (ie persistence) TURN metric
is the complement of Jaccardrsquos similarity index (Jaccard 1912)
In addition to temporal turnover we quantified its compo-
nents proportion of species lost (ie species recorded at a par-
ticular site during BBA1980 but no longer detected during
BBA2000 hereafter called extinction EXT) and proportion of
species gained (ie species not detected at a particular site
during BBA1980 but recorded during BBA2000 hereafter
called colonization COL) in each block between BBA1980 and
BBA2000 EXT and COL were calculated as follows
EXT =E
Ethorn Peth2THORN
COL =C
Cthorn Peth3THORN
For TURN we used the sum of E C and P as the denomi-
nator of Eqn (1) For EXT we used the sum of E and P as the
denominator because only species present in a block in
BBA1980 could have gone extinct For colonization we used
the sum of C and P as the denominator We recognize that
species already present in a block in BBA1980 (ie those
included in P metric) could not have colonized that block and
that species other than those included in C and P metrics
could have potentially colonized a site However we were
simply interested in the proportion of species that colonized
the site relative to all the species found within the block in the
second time period Furthermore using the sum of C and P as
the denominator of Eqn (3) allows COL metric to be equiva-
lent to EXT The values of TURN EXT and COL are bounded
by 0 and 1 values approaching 0 indicate low temporal turn-
over extinction or colonization in a BBA block between
BBA1980 and BBA2000 while values approaching 1 indicate
high temporal turnover extinction or colonization in a BBA
block between BBA1980 and BBA2000
To ensure that observed temporal turnover extinction and
colonization are at least partially a result of deterministic fac-
tors rather than stochastic processes we used North American
Breeding Bird Survey (httpswwwpwrcusgsgovbbs) as
independent dataset to verify whether change in community
composition increases with temporal lag We calculated tem-
poral turnover extinction and colonization that occurred
between 1980 and any subsequent year until 2005 and found
that each of the three community change metrics increased
with increasing temporal lag (see Fig S4 in Appendix S1)
Increase in community turnover with temporal lag implies
that shifts in community composition were unlikely to be a
result of stochastic processes alone but rather were partly dri-
ven by deterministic factors
We evaluated five competing statistical models (Table 1)
Model 1 included all climatic variables (TMAXTrend TMIN-
Trend and PRECIPTrend) as well as elevation (ELEV) and
survey effort (EFF) Model 2 included all climatic variables
ED ELEV and EFF Model 3 included all climatic variables
ED percent developed land (DEVEL) ELEV and EFF Model
4 included all climatic variables ED ELEV EFF and the inter-
action terms between the climatic variables and ED Model 5
included all climatic variables ED DEVEL ELEV ED and
the interaction terms between all climatic variables and ED
(Table 1) To ease the comparison and interpretation of the
coefficient estimates we standardized all the explanatory vari-
ables to z-scores with a mean = 0 and standard deviation = 1
We used a Bayesian model estimation procedure Each
response variable (temporal turnover TURN extinction EXT
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2945
or colonization COL) at each location (s) was modeled as a
binomial random variable Let y(s) be a response variable
then y(s) ~ B(p(s)N(s)) where N(s) is the total number of spe-
cies found at least once in a block from BBA1980 to BBA2000
(for models of TURN) the total number of species found in
BBA1980 (for models of EXT) or the total number of species
found in BBA2000 (for models of COL) For each BBA block
let p(s) = y(s)N(s) Also let x(s) be a vector containing the
explanatory variables for each model In all models we
included spatially structured random effects by adding all the
variance associated with the spatial process to the model inter-
cept thus accounting for the potential spatial autocorrelation
A generalized linear model approach was used to relate
responses to explanatory variables Using the logit link g(s) = log[p(s)1-p(s)] the spatially varying intercept model
was
gethsTHORN frac14 wethsTHORN thorn bxethsTHORN eth4THORNwhere b is a vector of regression coefficients and w(s) is a ran-
dom effect representing spatially varying adjustments to the
intercept We assume each element in w(s) arises from a spa-
tial Gaussian process (GP) (see eg Banerjee et al 2004 or
Cressie amp Wikle 2011 for more details) Specifically w(s) ~ GP
(0 C(s s0)) where s and s0 are any two locations within the
study area the spatial covariance C(s s0) = rq(s s0 ) with
variance parameter r2 correlation function q( ) and spatial
decay parameter The exponential spatial correlation was
assumed for q()Prior distributions on the parameters complete the hierar-
chical model specification (eg Gelman et al 2004) The glo-
bal regression coefficients brsquos followed a normal distribution
priors with mean and variance parameters equal to 0 and 100
respectively (ie N(0 100)) while the variance component r2
was assigned inverse gamma distribution prior with shape
and scale parameters of 2 and 1 respectively (ie IG(2 1))
The spatial decay parameter followed an informative uni-
form prior with support that ranged from 1 km to the maxi-
mum distance between any two grid cells (ie approximately
350 km) We ran three MCMC chains for 20 000 iterations
each Convergence was diagnosed using the GelmanndashRubin
diagnostic (Gelman amp Rubin 1992) As appropriate for a
Bayesian framework candidate model fit to the observed data
was assessed using the deviance information criterion (DIC
Spiegelhalter et al 2002) Lower values of DIC indicate
improved fit Model explanatory power was evaluated using
pseudo-R2 Pseudo-R2 was calculated as 1 ndash SSresSStot where
SSres is the sum of squared differences between the observed
and the predicted values of the dependent variable and SStot is
the sum of squared differences between the observed and the
mean value of the dependent variable Larger values of R2
suggest better fit Model accuracy was compared using root
mean square error (RMSE) that was calculated as the square
root of the mean squared deviations between fitted and
observed outcomes Lower values of RMSE indicate improved
accuracy
Prior to statistical analysis we removed all blocks that did
not have continuous land-cover coverage blocks with more
than 50 open water coverage and those that did not have
survey effort data We withheld 10 of the data for theTable
1Fivecompetingmodelsoftemporalturnoverextinctionan
dcolonizationincluded
thefollowingexplanatory
variablesmag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseason(TMAXTrend)mag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseason
(TMIN
Trend)
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitation
ofthebreed
ingseason(PRECIPTrend)
edgeden
sity
(ED)
percentdev
eloped
land
(DEVEL)interactionbetweenthemag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMAXTrendED)inter-
actionbetweenthemag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMIN
TrendED)interactionbetween
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitationofthebreed
ingseasonan
ded
geden
sity
(PRECIPTrendED)elev
ation(ELEV)an
dsu
rvey
effort(EFF)
Model
Explanatory
variables
TMAXTrend
TMIN
Trend
PRECIPTrend
ED
DEVEL
TMAXTrendED
TMIN
TrendED
PRECIPTrendED
ELEV
EFF
Model
1x
xx
xx
Model
2x
xx
xx
x
Model
3x
xx
xx
xx
Model
4x
xx
xx
xx
xx
Model
5x
xx
xx
xx
xx
x
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2946 M A JARZYNA et al
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
temporal change in community composition that has
occurred at the same site across a particular time period
(temporal turnover) Given that spatial turnover in
species composition is thought to be at least partially
related to spatial variation in environmental variables
(Gaston et al 2007 Buckley amp Jetz 2008 Gutierrez-
Canovas et al 2013 Siefert et al 2013) it is likely that
temporal turnover reflects temporal changes in climate
and could be a useful indicator of climate change
impacts on biodiversity
Even though the impacts of climate change on biodi-
versity are likely to be felt across a wide range of ecosys-
tems different communities are likely to respond to
climate change in disparate ways because a wide range
of environmental factors might influence speciesrsquo vulner-
abilities to climate change For example simulation stud-
ies suggest that the interaction between climate and land
cover will influence the way in which biodiversity
responds to climate change (Travis 2003 Jeltsch et al
2011) Specifically habitat fragmentation is thought to
affect the rate of distributional shifts associated with cli-
mate change allowing species in less fragmented habi-
tats to disperse faster and farther (Warren et al 2001
Opdam amp Wascher 2004) Lack of suitable habitat
between speciesrsquo current distribution and the future suit-
able climatic conditions might prevent a species with
some limiting intrinsic characteristics such as sedentary
life style or short dispersal distances from migrating to
these new climatically suitable locations If species in
contiguous (ie unfragmented) habitats shift their distri-
butions faster as a result of climate change than species
in fragmented habitats then changes in communities
should also follow the same pattern Thus stronger asso-
ciations of temporal turnover with changing climatic
conditions in contiguous habitats might be expected
In addition to more pronounced distributional shifts
other mechanisms might also lead or contribute to
stronger associations of temporal turnover with chang-
ing climate in unfragmented landscapes First commu-
nities in contiguous habitats tend to be comprised
largely of habitat specialists (Estavillo et al 2013)
which generally have narrower habitat and climatic
niches (Barnagaud et al 2012) and thus might be more
vulnerable to changing climatic conditions Conse-
quently the same amount of change in climatic condi-
tions would be expected to exert a stronger effect on
communities in contiguous landscapes Secondly frag-
mented habitats tend to support smaller populations
than contiguous unfragmented regions Small local
population sizes generally increase the effect of demo-
graphic or environmental stochasticity (Gaublomme
et al 2014) resulting in higher turnover in community
composition compared with that of regions supporting
large populations (Banks-Leite et al 2012)
While recent research suggests that climate change is
one of the factors causing temporal shifts in community
composition (Devictor et al 2008 Kampichler et al
2012 Prince amp Zuckerberg 2015) the influence of land-
cover and habitat fragmentation is still unclear For
example Devictor et al (2008) found that composition
of avian communities in France changed to include
higher proportion of high-temperature dwelling species
and suggested (although not explicitly tested) that this
response of communities to climate warming might dif-
fer between different types of land-cover and levels of
landscape fragmentation Kampichler et al (2012) sug-
gested a connection between land-cover and climate
change and bird community change by showing a
directional change in the proportion of high-tempera-
ture dwelling species as well as in the level of habitat
specialization of species comprising the community In
eastern North America Prince amp Zuckerberg (2015)
found that the composition of winter birds has become
increasingly dominated by warm-adapted species and
this community reshuffling was greatest in regions
experiencing long-term warming patterns However
despite these initial investigations the relationship
between temporal turnover in biodiversity and the
interaction of climatic changes and landscape fragmen-
tation to the best of our knowledge has not yet been
explicitly evaluated We intend to fill this research gap
by investigating the drivers of temporal turnover in
avian biodiversity across New York using Breeding
Bird Atlas data (Andrle amp Carroll 1988 McGowan amp
Corwin 2008) Our primary hypothesis was that tem-
poral turnover in avian assemblages is related to cli-
mate change but the strength of this relationship varies
with the degree of habitat fragmentation Specifically
we predict high rates of temporal turnover in regions
experiencing greater climate change and that these cli-
mate-mediated turnover rates would be higher in land-
scapes characterized by more contiguous habitat
Contiguous habitat is likely to facilitate higher rates of
turnover via three ecological mechanisms (i) allowing
species to disperse faster in response to climate change
(ii) supporting communities comprised of species more
vulnerable to changing climatic conditions and (iii)
increasing the ratio of deterministic to stochastic
processes
Materials and methods
Site description
The study area is the State of New York USA New York cov-
ers 128 401 km2 including 4240 km2 of inland water The cli-
mate of New York is affected by the statersquos broad elevation
gradient as well as by the proximity to lakes Erie and Ontario
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2943
and the Atlantic Ocean The Adirondack and Catskill Moun-
tains in the east are among the highest regions of the state
with elevation varying between 600 and 1500 m The elevation
of the southwestern New York ranges from approximately
300ndash900 m while northwestern New York as well as Hudson
River valley and New York City are low-lying with elevation
ranging from approximately 0ndash200 m
In general New York State is covered by approximately
512 forest (deciduous evergreen and mixed forests) 134
pasture and hay 82 cultivated crops 29 scrub and shrub
10 grassland and herbaceous 80 wetland cover 87
developed land and 02 barren land (Homer et al 2004)
New York offers a broad gradient of landscape fragmentation
The regions of the Adirondack and Catskill Mountains are the
least fragmented with expansive regions of forest land The
land cover of the Hudson River valley is fragmented by
residential and commercial development whereas the
landscapes of western New York are characterized mostly by
agriculturendashforest mosaic
Breeding Bird Atlas
We used the New York State Breeding Bird Atlas (BBA) to
characterize changes in avian communities through time BBA
is a statewide survey that documented the distribution of
breeding birds in New York To date BBA has been conducted
in two time periods 1980ndash1985 (hereafter BBA1980 Andrle amp
Carroll 1988) and 2000ndash2005 (hereafter BBA2000 McGowan
amp Corwin 2008) For both BBAs a grid system was used to
define the basic unit for reporting data The BBA reporting
unit (a block) was a square measuring 5 9 5 km a total of
5335 blocks covered the entirety of New York State This data-
set represents one of the largest and finest resolution atlases in
the world (Gibbons et al 2007)
A total of 242 species were recorded for BBA1980 and 248
species were recorded for BBA2000 (Andrle amp Carroll 1988
McGowan amp Corwin 2008 see Table S1 in Appendix S1 for
the list of all species used in this analysis) Observations were
made by skilled birders who spent at least 8 h in each block
visited all cover types in each block and included at least one
nighttime visit to document nocturnal species Observer effort
was recorded for each BBA and reported as a number of per-
son hours (measured as the sum of the number of hours spent
in each block x the number of people surveying each block
McGowan amp Zuckerberg 2008) The BBA represents a detec-
tionnondetection dataset nondetection indicates that species
could not be found given search criteria (McGowan amp Corwin
2008)
Habitat fragmentation
Habitat fragmentation was identified using the National Land
Cover Data (NLCD) derived from the Landsat Thematic Map-
per satellite data (Homer et al 2004) Even though the NLCD
are available for four time periods (ie 1992 2001 2006 and
2011) there is no land-cover data readily available for the time
period of BBA1980 Thus to assess whether temporal turnover
is related to landscape fragmentation and whether landscape
fragmentation affects responses of communities to climate
change we used NLCD 2001 To ensure that land-cover or
land-use change could not have significantly affected changes
in avian community composition over the time period of
1980ndash2005 we used the NLCD 19922001 Retrofit Land Cover
Change Product (Fry et al 2009) We quantified the amount of
land-cover change that occurred between 1992 and 2001 and
found that on average only 17 of the BBA blockrsquos area chan-
ged over this time period (median = 13 min = 0
max = 216) and approximately 97 of all blocks underwent
land-cover change of less than 5 (see Fig S1 in Appendix
S1) We thus felt confident that land-cover change had minor
if any effect on changes in avian community composition
Prior to landscape analysis we consolidated the following
land-cover classes open water and perennial ice and snow
classes into one class of waterice open space developed low
intensity developed medium intensity developed high inten-
sity developed into one class of developed land cultivated
crops and pasturehay into one class of agriculture and
deciduous forest evergreen forest mixed forest and forested
wetland into one class of forest Consolidation of these land-
cover types improved accuracy of classification and simplified
the environment for purposes of our evaluation Upon consoli-
dation we examined the following land-cover classes water
ice developed barren land forest scrubshrub grassland
herbaceous agriculture and wetlands
We used FRAGSTATS 41 (McGarigal et al 2012) and the
Geospatial Modelling Environment (GME httpwwwspa-
tialecologycomgme) to quantify landscape fragmentation
Because we focused our analysis on a diverse suite of species
with varying habitat requirements we chose a landscape-scale
variable to capture broadscale variation in habitat fragmenta-
tion We chose edge density (ED) as a measure of landscape
fragmentation because an increase in habitat edge is a primary
outcome of habitat fragmentation (Hargis et al 1998) ED was
also reported as an effective tool for evaluating landscape
fragmentation and performed better than other landscape
fragmentation indices (Hargis et al 1998) However in situa-
tions when landscape consists entirely of one cover type the
ED would be 0 regardless of the type of land cover present
Therefore to differentiate between landscapes that consisted
entirely of natural land cover (eg forest) and those mostly
developed (eg cities) we also calculated the proportion of
developed land (DEVEL) metric and included it as a variable
in our models (see Fig S2 in Appendix S1 for maps of all
explanatory variables)
Climatic trends
The climate data were derived from the PRISM (Parameter-ele-
vation Regressions on Independent Slopes Model) climate
mapping system (Daly amp Gibson 2002) PRISM consists of inter-
polated monthly maximum and minimum temperatures and
precipitation at a 25-arcmin resolution from 1891 to 2010 for
the entire contiguous United States We calculated the magni-
tude of the 25-year (1980ndash2005) trend in average monthly max-
imum and minimum temperatures and in total monthly
precipitation using ordinary least squares regression The
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2944 M A JARZYNA et al
slope of the OLS regression indicated the magnitude of the
trend and reflected the amount of change in climatic variables
that occurred between 1980 and 2005 We then interpolated
the trend magnitudes for each BBA block and averaged the
monthly values to reflect average breeding season (May
through September) trend magnitude in maximum and mini-
mum temperatures (TMAXTrend and TMINTrend respec-
tively) and in total precipitation (PRECIPTrend) TMAXTrend
and TMINTrend were expressed in degC25 years PRECIP-
Trend was expressed in mm25 years An average increase in
maximum temperatures of the breeding season was 063degC(median = 065 degC) over the 1980ndash2005 time period while an
average increase in minimum temperatures was 141 degC(median = 141 degC) over the same time period On average
precipitation increased by 10 mm (median = 10 mm) over the
1980ndash2005 time period (see Figs S2 and S3 in Appendix S1 for
maps and histograms of all explanatory variables)
Other model covariates
We included elevation in all the models to account for topo-
graphical variation of New York State We used digital eleva-
tion models for New York State to quantify mean elevation in
each block (ELEV) Additionally increasing survey effort often
results in a higher number of recorded species (Tobler et al
2008) Drastically different survey effort between BBA1980
and BBA2000 could increase the values of temporal turnover
which could be a result of different number of recorded spe-
cies rather than actual changes in species identities through
time To account for potential survey effort bias for each BBA
block we calculated the difference in the number of person
hours between BBA2000 and BBA1980 standardized by the
number of person hours in BBA1980 (EFF = (EFF2000 ndashEFF1980)EFF1980) and included it as a covariate in our
models We did not detect significant correlation among any
of our explanatory variables (Fig S3 in Appendix S1)
Statistical analysis
We quantified temporal turnover (TURN) as an aggregate of
species losses (ie species recorded at a particular site during
BBA1980 but no longer detected during BBA2000) and species
gains (ie species not detected at a particular site during
BBA1980 but recorded during BBA2000) within a particular
site Specifically temporal turnover was calculated as follows
TURN =Ethorn C
Ethorn Cthorn Peth1THORN
where E is the number of species lost in a particular BBA block
between BBA1980 and BBA2000 C is the number of species
gained in a particular BBA block between BBA1980 and
BBA2000 and P is number of species present in a particular
BBA block during both BBAs (ie persistence) TURN metric
is the complement of Jaccardrsquos similarity index (Jaccard 1912)
In addition to temporal turnover we quantified its compo-
nents proportion of species lost (ie species recorded at a par-
ticular site during BBA1980 but no longer detected during
BBA2000 hereafter called extinction EXT) and proportion of
species gained (ie species not detected at a particular site
during BBA1980 but recorded during BBA2000 hereafter
called colonization COL) in each block between BBA1980 and
BBA2000 EXT and COL were calculated as follows
EXT =E
Ethorn Peth2THORN
COL =C
Cthorn Peth3THORN
For TURN we used the sum of E C and P as the denomi-
nator of Eqn (1) For EXT we used the sum of E and P as the
denominator because only species present in a block in
BBA1980 could have gone extinct For colonization we used
the sum of C and P as the denominator We recognize that
species already present in a block in BBA1980 (ie those
included in P metric) could not have colonized that block and
that species other than those included in C and P metrics
could have potentially colonized a site However we were
simply interested in the proportion of species that colonized
the site relative to all the species found within the block in the
second time period Furthermore using the sum of C and P as
the denominator of Eqn (3) allows COL metric to be equiva-
lent to EXT The values of TURN EXT and COL are bounded
by 0 and 1 values approaching 0 indicate low temporal turn-
over extinction or colonization in a BBA block between
BBA1980 and BBA2000 while values approaching 1 indicate
high temporal turnover extinction or colonization in a BBA
block between BBA1980 and BBA2000
To ensure that observed temporal turnover extinction and
colonization are at least partially a result of deterministic fac-
tors rather than stochastic processes we used North American
Breeding Bird Survey (httpswwwpwrcusgsgovbbs) as
independent dataset to verify whether change in community
composition increases with temporal lag We calculated tem-
poral turnover extinction and colonization that occurred
between 1980 and any subsequent year until 2005 and found
that each of the three community change metrics increased
with increasing temporal lag (see Fig S4 in Appendix S1)
Increase in community turnover with temporal lag implies
that shifts in community composition were unlikely to be a
result of stochastic processes alone but rather were partly dri-
ven by deterministic factors
We evaluated five competing statistical models (Table 1)
Model 1 included all climatic variables (TMAXTrend TMIN-
Trend and PRECIPTrend) as well as elevation (ELEV) and
survey effort (EFF) Model 2 included all climatic variables
ED ELEV and EFF Model 3 included all climatic variables
ED percent developed land (DEVEL) ELEV and EFF Model
4 included all climatic variables ED ELEV EFF and the inter-
action terms between the climatic variables and ED Model 5
included all climatic variables ED DEVEL ELEV ED and
the interaction terms between all climatic variables and ED
(Table 1) To ease the comparison and interpretation of the
coefficient estimates we standardized all the explanatory vari-
ables to z-scores with a mean = 0 and standard deviation = 1
We used a Bayesian model estimation procedure Each
response variable (temporal turnover TURN extinction EXT
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2945
or colonization COL) at each location (s) was modeled as a
binomial random variable Let y(s) be a response variable
then y(s) ~ B(p(s)N(s)) where N(s) is the total number of spe-
cies found at least once in a block from BBA1980 to BBA2000
(for models of TURN) the total number of species found in
BBA1980 (for models of EXT) or the total number of species
found in BBA2000 (for models of COL) For each BBA block
let p(s) = y(s)N(s) Also let x(s) be a vector containing the
explanatory variables for each model In all models we
included spatially structured random effects by adding all the
variance associated with the spatial process to the model inter-
cept thus accounting for the potential spatial autocorrelation
A generalized linear model approach was used to relate
responses to explanatory variables Using the logit link g(s) = log[p(s)1-p(s)] the spatially varying intercept model
was
gethsTHORN frac14 wethsTHORN thorn bxethsTHORN eth4THORNwhere b is a vector of regression coefficients and w(s) is a ran-
dom effect representing spatially varying adjustments to the
intercept We assume each element in w(s) arises from a spa-
tial Gaussian process (GP) (see eg Banerjee et al 2004 or
Cressie amp Wikle 2011 for more details) Specifically w(s) ~ GP
(0 C(s s0)) where s and s0 are any two locations within the
study area the spatial covariance C(s s0) = rq(s s0 ) with
variance parameter r2 correlation function q( ) and spatial
decay parameter The exponential spatial correlation was
assumed for q()Prior distributions on the parameters complete the hierar-
chical model specification (eg Gelman et al 2004) The glo-
bal regression coefficients brsquos followed a normal distribution
priors with mean and variance parameters equal to 0 and 100
respectively (ie N(0 100)) while the variance component r2
was assigned inverse gamma distribution prior with shape
and scale parameters of 2 and 1 respectively (ie IG(2 1))
The spatial decay parameter followed an informative uni-
form prior with support that ranged from 1 km to the maxi-
mum distance between any two grid cells (ie approximately
350 km) We ran three MCMC chains for 20 000 iterations
each Convergence was diagnosed using the GelmanndashRubin
diagnostic (Gelman amp Rubin 1992) As appropriate for a
Bayesian framework candidate model fit to the observed data
was assessed using the deviance information criterion (DIC
Spiegelhalter et al 2002) Lower values of DIC indicate
improved fit Model explanatory power was evaluated using
pseudo-R2 Pseudo-R2 was calculated as 1 ndash SSresSStot where
SSres is the sum of squared differences between the observed
and the predicted values of the dependent variable and SStot is
the sum of squared differences between the observed and the
mean value of the dependent variable Larger values of R2
suggest better fit Model accuracy was compared using root
mean square error (RMSE) that was calculated as the square
root of the mean squared deviations between fitted and
observed outcomes Lower values of RMSE indicate improved
accuracy
Prior to statistical analysis we removed all blocks that did
not have continuous land-cover coverage blocks with more
than 50 open water coverage and those that did not have
survey effort data We withheld 10 of the data for theTable
1Fivecompetingmodelsoftemporalturnoverextinctionan
dcolonizationincluded
thefollowingexplanatory
variablesmag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseason(TMAXTrend)mag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseason
(TMIN
Trend)
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitation
ofthebreed
ingseason(PRECIPTrend)
edgeden
sity
(ED)
percentdev
eloped
land
(DEVEL)interactionbetweenthemag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMAXTrendED)inter-
actionbetweenthemag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMIN
TrendED)interactionbetween
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitationofthebreed
ingseasonan
ded
geden
sity
(PRECIPTrendED)elev
ation(ELEV)an
dsu
rvey
effort(EFF)
Model
Explanatory
variables
TMAXTrend
TMIN
Trend
PRECIPTrend
ED
DEVEL
TMAXTrendED
TMIN
TrendED
PRECIPTrendED
ELEV
EFF
Model
1x
xx
xx
Model
2x
xx
xx
x
Model
3x
xx
xx
xx
Model
4x
xx
xx
xx
xx
Model
5x
xx
xx
xx
xx
x
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2946 M A JARZYNA et al
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
and the Atlantic Ocean The Adirondack and Catskill Moun-
tains in the east are among the highest regions of the state
with elevation varying between 600 and 1500 m The elevation
of the southwestern New York ranges from approximately
300ndash900 m while northwestern New York as well as Hudson
River valley and New York City are low-lying with elevation
ranging from approximately 0ndash200 m
In general New York State is covered by approximately
512 forest (deciduous evergreen and mixed forests) 134
pasture and hay 82 cultivated crops 29 scrub and shrub
10 grassland and herbaceous 80 wetland cover 87
developed land and 02 barren land (Homer et al 2004)
New York offers a broad gradient of landscape fragmentation
The regions of the Adirondack and Catskill Mountains are the
least fragmented with expansive regions of forest land The
land cover of the Hudson River valley is fragmented by
residential and commercial development whereas the
landscapes of western New York are characterized mostly by
agriculturendashforest mosaic
Breeding Bird Atlas
We used the New York State Breeding Bird Atlas (BBA) to
characterize changes in avian communities through time BBA
is a statewide survey that documented the distribution of
breeding birds in New York To date BBA has been conducted
in two time periods 1980ndash1985 (hereafter BBA1980 Andrle amp
Carroll 1988) and 2000ndash2005 (hereafter BBA2000 McGowan
amp Corwin 2008) For both BBAs a grid system was used to
define the basic unit for reporting data The BBA reporting
unit (a block) was a square measuring 5 9 5 km a total of
5335 blocks covered the entirety of New York State This data-
set represents one of the largest and finest resolution atlases in
the world (Gibbons et al 2007)
A total of 242 species were recorded for BBA1980 and 248
species were recorded for BBA2000 (Andrle amp Carroll 1988
McGowan amp Corwin 2008 see Table S1 in Appendix S1 for
the list of all species used in this analysis) Observations were
made by skilled birders who spent at least 8 h in each block
visited all cover types in each block and included at least one
nighttime visit to document nocturnal species Observer effort
was recorded for each BBA and reported as a number of per-
son hours (measured as the sum of the number of hours spent
in each block x the number of people surveying each block
McGowan amp Zuckerberg 2008) The BBA represents a detec-
tionnondetection dataset nondetection indicates that species
could not be found given search criteria (McGowan amp Corwin
2008)
Habitat fragmentation
Habitat fragmentation was identified using the National Land
Cover Data (NLCD) derived from the Landsat Thematic Map-
per satellite data (Homer et al 2004) Even though the NLCD
are available for four time periods (ie 1992 2001 2006 and
2011) there is no land-cover data readily available for the time
period of BBA1980 Thus to assess whether temporal turnover
is related to landscape fragmentation and whether landscape
fragmentation affects responses of communities to climate
change we used NLCD 2001 To ensure that land-cover or
land-use change could not have significantly affected changes
in avian community composition over the time period of
1980ndash2005 we used the NLCD 19922001 Retrofit Land Cover
Change Product (Fry et al 2009) We quantified the amount of
land-cover change that occurred between 1992 and 2001 and
found that on average only 17 of the BBA blockrsquos area chan-
ged over this time period (median = 13 min = 0
max = 216) and approximately 97 of all blocks underwent
land-cover change of less than 5 (see Fig S1 in Appendix
S1) We thus felt confident that land-cover change had minor
if any effect on changes in avian community composition
Prior to landscape analysis we consolidated the following
land-cover classes open water and perennial ice and snow
classes into one class of waterice open space developed low
intensity developed medium intensity developed high inten-
sity developed into one class of developed land cultivated
crops and pasturehay into one class of agriculture and
deciduous forest evergreen forest mixed forest and forested
wetland into one class of forest Consolidation of these land-
cover types improved accuracy of classification and simplified
the environment for purposes of our evaluation Upon consoli-
dation we examined the following land-cover classes water
ice developed barren land forest scrubshrub grassland
herbaceous agriculture and wetlands
We used FRAGSTATS 41 (McGarigal et al 2012) and the
Geospatial Modelling Environment (GME httpwwwspa-
tialecologycomgme) to quantify landscape fragmentation
Because we focused our analysis on a diverse suite of species
with varying habitat requirements we chose a landscape-scale
variable to capture broadscale variation in habitat fragmenta-
tion We chose edge density (ED) as a measure of landscape
fragmentation because an increase in habitat edge is a primary
outcome of habitat fragmentation (Hargis et al 1998) ED was
also reported as an effective tool for evaluating landscape
fragmentation and performed better than other landscape
fragmentation indices (Hargis et al 1998) However in situa-
tions when landscape consists entirely of one cover type the
ED would be 0 regardless of the type of land cover present
Therefore to differentiate between landscapes that consisted
entirely of natural land cover (eg forest) and those mostly
developed (eg cities) we also calculated the proportion of
developed land (DEVEL) metric and included it as a variable
in our models (see Fig S2 in Appendix S1 for maps of all
explanatory variables)
Climatic trends
The climate data were derived from the PRISM (Parameter-ele-
vation Regressions on Independent Slopes Model) climate
mapping system (Daly amp Gibson 2002) PRISM consists of inter-
polated monthly maximum and minimum temperatures and
precipitation at a 25-arcmin resolution from 1891 to 2010 for
the entire contiguous United States We calculated the magni-
tude of the 25-year (1980ndash2005) trend in average monthly max-
imum and minimum temperatures and in total monthly
precipitation using ordinary least squares regression The
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2944 M A JARZYNA et al
slope of the OLS regression indicated the magnitude of the
trend and reflected the amount of change in climatic variables
that occurred between 1980 and 2005 We then interpolated
the trend magnitudes for each BBA block and averaged the
monthly values to reflect average breeding season (May
through September) trend magnitude in maximum and mini-
mum temperatures (TMAXTrend and TMINTrend respec-
tively) and in total precipitation (PRECIPTrend) TMAXTrend
and TMINTrend were expressed in degC25 years PRECIP-
Trend was expressed in mm25 years An average increase in
maximum temperatures of the breeding season was 063degC(median = 065 degC) over the 1980ndash2005 time period while an
average increase in minimum temperatures was 141 degC(median = 141 degC) over the same time period On average
precipitation increased by 10 mm (median = 10 mm) over the
1980ndash2005 time period (see Figs S2 and S3 in Appendix S1 for
maps and histograms of all explanatory variables)
Other model covariates
We included elevation in all the models to account for topo-
graphical variation of New York State We used digital eleva-
tion models for New York State to quantify mean elevation in
each block (ELEV) Additionally increasing survey effort often
results in a higher number of recorded species (Tobler et al
2008) Drastically different survey effort between BBA1980
and BBA2000 could increase the values of temporal turnover
which could be a result of different number of recorded spe-
cies rather than actual changes in species identities through
time To account for potential survey effort bias for each BBA
block we calculated the difference in the number of person
hours between BBA2000 and BBA1980 standardized by the
number of person hours in BBA1980 (EFF = (EFF2000 ndashEFF1980)EFF1980) and included it as a covariate in our
models We did not detect significant correlation among any
of our explanatory variables (Fig S3 in Appendix S1)
Statistical analysis
We quantified temporal turnover (TURN) as an aggregate of
species losses (ie species recorded at a particular site during
BBA1980 but no longer detected during BBA2000) and species
gains (ie species not detected at a particular site during
BBA1980 but recorded during BBA2000) within a particular
site Specifically temporal turnover was calculated as follows
TURN =Ethorn C
Ethorn Cthorn Peth1THORN
where E is the number of species lost in a particular BBA block
between BBA1980 and BBA2000 C is the number of species
gained in a particular BBA block between BBA1980 and
BBA2000 and P is number of species present in a particular
BBA block during both BBAs (ie persistence) TURN metric
is the complement of Jaccardrsquos similarity index (Jaccard 1912)
In addition to temporal turnover we quantified its compo-
nents proportion of species lost (ie species recorded at a par-
ticular site during BBA1980 but no longer detected during
BBA2000 hereafter called extinction EXT) and proportion of
species gained (ie species not detected at a particular site
during BBA1980 but recorded during BBA2000 hereafter
called colonization COL) in each block between BBA1980 and
BBA2000 EXT and COL were calculated as follows
EXT =E
Ethorn Peth2THORN
COL =C
Cthorn Peth3THORN
For TURN we used the sum of E C and P as the denomi-
nator of Eqn (1) For EXT we used the sum of E and P as the
denominator because only species present in a block in
BBA1980 could have gone extinct For colonization we used
the sum of C and P as the denominator We recognize that
species already present in a block in BBA1980 (ie those
included in P metric) could not have colonized that block and
that species other than those included in C and P metrics
could have potentially colonized a site However we were
simply interested in the proportion of species that colonized
the site relative to all the species found within the block in the
second time period Furthermore using the sum of C and P as
the denominator of Eqn (3) allows COL metric to be equiva-
lent to EXT The values of TURN EXT and COL are bounded
by 0 and 1 values approaching 0 indicate low temporal turn-
over extinction or colonization in a BBA block between
BBA1980 and BBA2000 while values approaching 1 indicate
high temporal turnover extinction or colonization in a BBA
block between BBA1980 and BBA2000
To ensure that observed temporal turnover extinction and
colonization are at least partially a result of deterministic fac-
tors rather than stochastic processes we used North American
Breeding Bird Survey (httpswwwpwrcusgsgovbbs) as
independent dataset to verify whether change in community
composition increases with temporal lag We calculated tem-
poral turnover extinction and colonization that occurred
between 1980 and any subsequent year until 2005 and found
that each of the three community change metrics increased
with increasing temporal lag (see Fig S4 in Appendix S1)
Increase in community turnover with temporal lag implies
that shifts in community composition were unlikely to be a
result of stochastic processes alone but rather were partly dri-
ven by deterministic factors
We evaluated five competing statistical models (Table 1)
Model 1 included all climatic variables (TMAXTrend TMIN-
Trend and PRECIPTrend) as well as elevation (ELEV) and
survey effort (EFF) Model 2 included all climatic variables
ED ELEV and EFF Model 3 included all climatic variables
ED percent developed land (DEVEL) ELEV and EFF Model
4 included all climatic variables ED ELEV EFF and the inter-
action terms between the climatic variables and ED Model 5
included all climatic variables ED DEVEL ELEV ED and
the interaction terms between all climatic variables and ED
(Table 1) To ease the comparison and interpretation of the
coefficient estimates we standardized all the explanatory vari-
ables to z-scores with a mean = 0 and standard deviation = 1
We used a Bayesian model estimation procedure Each
response variable (temporal turnover TURN extinction EXT
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2945
or colonization COL) at each location (s) was modeled as a
binomial random variable Let y(s) be a response variable
then y(s) ~ B(p(s)N(s)) where N(s) is the total number of spe-
cies found at least once in a block from BBA1980 to BBA2000
(for models of TURN) the total number of species found in
BBA1980 (for models of EXT) or the total number of species
found in BBA2000 (for models of COL) For each BBA block
let p(s) = y(s)N(s) Also let x(s) be a vector containing the
explanatory variables for each model In all models we
included spatially structured random effects by adding all the
variance associated with the spatial process to the model inter-
cept thus accounting for the potential spatial autocorrelation
A generalized linear model approach was used to relate
responses to explanatory variables Using the logit link g(s) = log[p(s)1-p(s)] the spatially varying intercept model
was
gethsTHORN frac14 wethsTHORN thorn bxethsTHORN eth4THORNwhere b is a vector of regression coefficients and w(s) is a ran-
dom effect representing spatially varying adjustments to the
intercept We assume each element in w(s) arises from a spa-
tial Gaussian process (GP) (see eg Banerjee et al 2004 or
Cressie amp Wikle 2011 for more details) Specifically w(s) ~ GP
(0 C(s s0)) where s and s0 are any two locations within the
study area the spatial covariance C(s s0) = rq(s s0 ) with
variance parameter r2 correlation function q( ) and spatial
decay parameter The exponential spatial correlation was
assumed for q()Prior distributions on the parameters complete the hierar-
chical model specification (eg Gelman et al 2004) The glo-
bal regression coefficients brsquos followed a normal distribution
priors with mean and variance parameters equal to 0 and 100
respectively (ie N(0 100)) while the variance component r2
was assigned inverse gamma distribution prior with shape
and scale parameters of 2 and 1 respectively (ie IG(2 1))
The spatial decay parameter followed an informative uni-
form prior with support that ranged from 1 km to the maxi-
mum distance between any two grid cells (ie approximately
350 km) We ran three MCMC chains for 20 000 iterations
each Convergence was diagnosed using the GelmanndashRubin
diagnostic (Gelman amp Rubin 1992) As appropriate for a
Bayesian framework candidate model fit to the observed data
was assessed using the deviance information criterion (DIC
Spiegelhalter et al 2002) Lower values of DIC indicate
improved fit Model explanatory power was evaluated using
pseudo-R2 Pseudo-R2 was calculated as 1 ndash SSresSStot where
SSres is the sum of squared differences between the observed
and the predicted values of the dependent variable and SStot is
the sum of squared differences between the observed and the
mean value of the dependent variable Larger values of R2
suggest better fit Model accuracy was compared using root
mean square error (RMSE) that was calculated as the square
root of the mean squared deviations between fitted and
observed outcomes Lower values of RMSE indicate improved
accuracy
Prior to statistical analysis we removed all blocks that did
not have continuous land-cover coverage blocks with more
than 50 open water coverage and those that did not have
survey effort data We withheld 10 of the data for theTable
1Fivecompetingmodelsoftemporalturnoverextinctionan
dcolonizationincluded
thefollowingexplanatory
variablesmag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseason(TMAXTrend)mag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseason
(TMIN
Trend)
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitation
ofthebreed
ingseason(PRECIPTrend)
edgeden
sity
(ED)
percentdev
eloped
land
(DEVEL)interactionbetweenthemag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMAXTrendED)inter-
actionbetweenthemag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMIN
TrendED)interactionbetween
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitationofthebreed
ingseasonan
ded
geden
sity
(PRECIPTrendED)elev
ation(ELEV)an
dsu
rvey
effort(EFF)
Model
Explanatory
variables
TMAXTrend
TMIN
Trend
PRECIPTrend
ED
DEVEL
TMAXTrendED
TMIN
TrendED
PRECIPTrendED
ELEV
EFF
Model
1x
xx
xx
Model
2x
xx
xx
x
Model
3x
xx
xx
xx
Model
4x
xx
xx
xx
xx
Model
5x
xx
xx
xx
xx
x
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2946 M A JARZYNA et al
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
slope of the OLS regression indicated the magnitude of the
trend and reflected the amount of change in climatic variables
that occurred between 1980 and 2005 We then interpolated
the trend magnitudes for each BBA block and averaged the
monthly values to reflect average breeding season (May
through September) trend magnitude in maximum and mini-
mum temperatures (TMAXTrend and TMINTrend respec-
tively) and in total precipitation (PRECIPTrend) TMAXTrend
and TMINTrend were expressed in degC25 years PRECIP-
Trend was expressed in mm25 years An average increase in
maximum temperatures of the breeding season was 063degC(median = 065 degC) over the 1980ndash2005 time period while an
average increase in minimum temperatures was 141 degC(median = 141 degC) over the same time period On average
precipitation increased by 10 mm (median = 10 mm) over the
1980ndash2005 time period (see Figs S2 and S3 in Appendix S1 for
maps and histograms of all explanatory variables)
Other model covariates
We included elevation in all the models to account for topo-
graphical variation of New York State We used digital eleva-
tion models for New York State to quantify mean elevation in
each block (ELEV) Additionally increasing survey effort often
results in a higher number of recorded species (Tobler et al
2008) Drastically different survey effort between BBA1980
and BBA2000 could increase the values of temporal turnover
which could be a result of different number of recorded spe-
cies rather than actual changes in species identities through
time To account for potential survey effort bias for each BBA
block we calculated the difference in the number of person
hours between BBA2000 and BBA1980 standardized by the
number of person hours in BBA1980 (EFF = (EFF2000 ndashEFF1980)EFF1980) and included it as a covariate in our
models We did not detect significant correlation among any
of our explanatory variables (Fig S3 in Appendix S1)
Statistical analysis
We quantified temporal turnover (TURN) as an aggregate of
species losses (ie species recorded at a particular site during
BBA1980 but no longer detected during BBA2000) and species
gains (ie species not detected at a particular site during
BBA1980 but recorded during BBA2000) within a particular
site Specifically temporal turnover was calculated as follows
TURN =Ethorn C
Ethorn Cthorn Peth1THORN
where E is the number of species lost in a particular BBA block
between BBA1980 and BBA2000 C is the number of species
gained in a particular BBA block between BBA1980 and
BBA2000 and P is number of species present in a particular
BBA block during both BBAs (ie persistence) TURN metric
is the complement of Jaccardrsquos similarity index (Jaccard 1912)
In addition to temporal turnover we quantified its compo-
nents proportion of species lost (ie species recorded at a par-
ticular site during BBA1980 but no longer detected during
BBA2000 hereafter called extinction EXT) and proportion of
species gained (ie species not detected at a particular site
during BBA1980 but recorded during BBA2000 hereafter
called colonization COL) in each block between BBA1980 and
BBA2000 EXT and COL were calculated as follows
EXT =E
Ethorn Peth2THORN
COL =C
Cthorn Peth3THORN
For TURN we used the sum of E C and P as the denomi-
nator of Eqn (1) For EXT we used the sum of E and P as the
denominator because only species present in a block in
BBA1980 could have gone extinct For colonization we used
the sum of C and P as the denominator We recognize that
species already present in a block in BBA1980 (ie those
included in P metric) could not have colonized that block and
that species other than those included in C and P metrics
could have potentially colonized a site However we were
simply interested in the proportion of species that colonized
the site relative to all the species found within the block in the
second time period Furthermore using the sum of C and P as
the denominator of Eqn (3) allows COL metric to be equiva-
lent to EXT The values of TURN EXT and COL are bounded
by 0 and 1 values approaching 0 indicate low temporal turn-
over extinction or colonization in a BBA block between
BBA1980 and BBA2000 while values approaching 1 indicate
high temporal turnover extinction or colonization in a BBA
block between BBA1980 and BBA2000
To ensure that observed temporal turnover extinction and
colonization are at least partially a result of deterministic fac-
tors rather than stochastic processes we used North American
Breeding Bird Survey (httpswwwpwrcusgsgovbbs) as
independent dataset to verify whether change in community
composition increases with temporal lag We calculated tem-
poral turnover extinction and colonization that occurred
between 1980 and any subsequent year until 2005 and found
that each of the three community change metrics increased
with increasing temporal lag (see Fig S4 in Appendix S1)
Increase in community turnover with temporal lag implies
that shifts in community composition were unlikely to be a
result of stochastic processes alone but rather were partly dri-
ven by deterministic factors
We evaluated five competing statistical models (Table 1)
Model 1 included all climatic variables (TMAXTrend TMIN-
Trend and PRECIPTrend) as well as elevation (ELEV) and
survey effort (EFF) Model 2 included all climatic variables
ED ELEV and EFF Model 3 included all climatic variables
ED percent developed land (DEVEL) ELEV and EFF Model
4 included all climatic variables ED ELEV EFF and the inter-
action terms between the climatic variables and ED Model 5
included all climatic variables ED DEVEL ELEV ED and
the interaction terms between all climatic variables and ED
(Table 1) To ease the comparison and interpretation of the
coefficient estimates we standardized all the explanatory vari-
ables to z-scores with a mean = 0 and standard deviation = 1
We used a Bayesian model estimation procedure Each
response variable (temporal turnover TURN extinction EXT
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2945
or colonization COL) at each location (s) was modeled as a
binomial random variable Let y(s) be a response variable
then y(s) ~ B(p(s)N(s)) where N(s) is the total number of spe-
cies found at least once in a block from BBA1980 to BBA2000
(for models of TURN) the total number of species found in
BBA1980 (for models of EXT) or the total number of species
found in BBA2000 (for models of COL) For each BBA block
let p(s) = y(s)N(s) Also let x(s) be a vector containing the
explanatory variables for each model In all models we
included spatially structured random effects by adding all the
variance associated with the spatial process to the model inter-
cept thus accounting for the potential spatial autocorrelation
A generalized linear model approach was used to relate
responses to explanatory variables Using the logit link g(s) = log[p(s)1-p(s)] the spatially varying intercept model
was
gethsTHORN frac14 wethsTHORN thorn bxethsTHORN eth4THORNwhere b is a vector of regression coefficients and w(s) is a ran-
dom effect representing spatially varying adjustments to the
intercept We assume each element in w(s) arises from a spa-
tial Gaussian process (GP) (see eg Banerjee et al 2004 or
Cressie amp Wikle 2011 for more details) Specifically w(s) ~ GP
(0 C(s s0)) where s and s0 are any two locations within the
study area the spatial covariance C(s s0) = rq(s s0 ) with
variance parameter r2 correlation function q( ) and spatial
decay parameter The exponential spatial correlation was
assumed for q()Prior distributions on the parameters complete the hierar-
chical model specification (eg Gelman et al 2004) The glo-
bal regression coefficients brsquos followed a normal distribution
priors with mean and variance parameters equal to 0 and 100
respectively (ie N(0 100)) while the variance component r2
was assigned inverse gamma distribution prior with shape
and scale parameters of 2 and 1 respectively (ie IG(2 1))
The spatial decay parameter followed an informative uni-
form prior with support that ranged from 1 km to the maxi-
mum distance between any two grid cells (ie approximately
350 km) We ran three MCMC chains for 20 000 iterations
each Convergence was diagnosed using the GelmanndashRubin
diagnostic (Gelman amp Rubin 1992) As appropriate for a
Bayesian framework candidate model fit to the observed data
was assessed using the deviance information criterion (DIC
Spiegelhalter et al 2002) Lower values of DIC indicate
improved fit Model explanatory power was evaluated using
pseudo-R2 Pseudo-R2 was calculated as 1 ndash SSresSStot where
SSres is the sum of squared differences between the observed
and the predicted values of the dependent variable and SStot is
the sum of squared differences between the observed and the
mean value of the dependent variable Larger values of R2
suggest better fit Model accuracy was compared using root
mean square error (RMSE) that was calculated as the square
root of the mean squared deviations between fitted and
observed outcomes Lower values of RMSE indicate improved
accuracy
Prior to statistical analysis we removed all blocks that did
not have continuous land-cover coverage blocks with more
than 50 open water coverage and those that did not have
survey effort data We withheld 10 of the data for theTable
1Fivecompetingmodelsoftemporalturnoverextinctionan
dcolonizationincluded
thefollowingexplanatory
variablesmag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseason(TMAXTrend)mag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseason
(TMIN
Trend)
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitation
ofthebreed
ingseason(PRECIPTrend)
edgeden
sity
(ED)
percentdev
eloped
land
(DEVEL)interactionbetweenthemag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMAXTrendED)inter-
actionbetweenthemag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMIN
TrendED)interactionbetween
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitationofthebreed
ingseasonan
ded
geden
sity
(PRECIPTrendED)elev
ation(ELEV)an
dsu
rvey
effort(EFF)
Model
Explanatory
variables
TMAXTrend
TMIN
Trend
PRECIPTrend
ED
DEVEL
TMAXTrendED
TMIN
TrendED
PRECIPTrendED
ELEV
EFF
Model
1x
xx
xx
Model
2x
xx
xx
x
Model
3x
xx
xx
xx
Model
4x
xx
xx
xx
xx
Model
5x
xx
xx
xx
xx
x
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2946 M A JARZYNA et al
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
or colonization COL) at each location (s) was modeled as a
binomial random variable Let y(s) be a response variable
then y(s) ~ B(p(s)N(s)) where N(s) is the total number of spe-
cies found at least once in a block from BBA1980 to BBA2000
(for models of TURN) the total number of species found in
BBA1980 (for models of EXT) or the total number of species
found in BBA2000 (for models of COL) For each BBA block
let p(s) = y(s)N(s) Also let x(s) be a vector containing the
explanatory variables for each model In all models we
included spatially structured random effects by adding all the
variance associated with the spatial process to the model inter-
cept thus accounting for the potential spatial autocorrelation
A generalized linear model approach was used to relate
responses to explanatory variables Using the logit link g(s) = log[p(s)1-p(s)] the spatially varying intercept model
was
gethsTHORN frac14 wethsTHORN thorn bxethsTHORN eth4THORNwhere b is a vector of regression coefficients and w(s) is a ran-
dom effect representing spatially varying adjustments to the
intercept We assume each element in w(s) arises from a spa-
tial Gaussian process (GP) (see eg Banerjee et al 2004 or
Cressie amp Wikle 2011 for more details) Specifically w(s) ~ GP
(0 C(s s0)) where s and s0 are any two locations within the
study area the spatial covariance C(s s0) = rq(s s0 ) with
variance parameter r2 correlation function q( ) and spatial
decay parameter The exponential spatial correlation was
assumed for q()Prior distributions on the parameters complete the hierar-
chical model specification (eg Gelman et al 2004) The glo-
bal regression coefficients brsquos followed a normal distribution
priors with mean and variance parameters equal to 0 and 100
respectively (ie N(0 100)) while the variance component r2
was assigned inverse gamma distribution prior with shape
and scale parameters of 2 and 1 respectively (ie IG(2 1))
The spatial decay parameter followed an informative uni-
form prior with support that ranged from 1 km to the maxi-
mum distance between any two grid cells (ie approximately
350 km) We ran three MCMC chains for 20 000 iterations
each Convergence was diagnosed using the GelmanndashRubin
diagnostic (Gelman amp Rubin 1992) As appropriate for a
Bayesian framework candidate model fit to the observed data
was assessed using the deviance information criterion (DIC
Spiegelhalter et al 2002) Lower values of DIC indicate
improved fit Model explanatory power was evaluated using
pseudo-R2 Pseudo-R2 was calculated as 1 ndash SSresSStot where
SSres is the sum of squared differences between the observed
and the predicted values of the dependent variable and SStot is
the sum of squared differences between the observed and the
mean value of the dependent variable Larger values of R2
suggest better fit Model accuracy was compared using root
mean square error (RMSE) that was calculated as the square
root of the mean squared deviations between fitted and
observed outcomes Lower values of RMSE indicate improved
accuracy
Prior to statistical analysis we removed all blocks that did
not have continuous land-cover coverage blocks with more
than 50 open water coverage and those that did not have
survey effort data We withheld 10 of the data for theTable
1Fivecompetingmodelsoftemporalturnoverextinctionan
dcolonizationincluded
thefollowingexplanatory
variablesmag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseason(TMAXTrend)mag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseason
(TMIN
Trend)
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitation
ofthebreed
ingseason(PRECIPTrend)
edgeden
sity
(ED)
percentdev
eloped
land
(DEVEL)interactionbetweenthemag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
emax
imum
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMAXTrendED)inter-
actionbetweenthemag
nitudeofthe25
-year(198
0ndash200
5)tren
din
averag
eminim
um
temperature
ofthebreed
ingseasonan
ded
geden
sity
(TMIN
TrendED)interactionbetween
mag
nitudeofthe25
-year(198
0ndash20
05)tren
din
averag
etotalprecipitationofthebreed
ingseasonan
ded
geden
sity
(PRECIPTrendED)elev
ation(ELEV)an
dsu
rvey
effort(EFF)
Model
Explanatory
variables
TMAXTrend
TMIN
Trend
PRECIPTrend
ED
DEVEL
TMAXTrendED
TMIN
TrendED
PRECIPTrendED
ELEV
EFF
Model
1x
xx
xx
Model
2x
xx
xx
x
Model
3x
xx
xx
xx
Model
4x
xx
xx
xx
xx
Model
5x
xx
xx
xx
xx
x
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2946 M A JARZYNA et al
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
validation of modelrsquos predictive performance Ultimately we
used a total of N = 4271 blocks to fit the models and 473
blocks to validate the models Holdout set predictions were
compared to the observed values using root mean square
error (RMSEpred) where lower values indicate improved per-
formance
The models were run in R 2151 statistical package (R
Development Core Team 2013 httpwwwr-projectorg)
using package spBayes (Finley et al 2007) In a Bayesian
framework parameter estimates have valid posterior distribu-
tions and inferences are made using mean or median as well
as the credible intervals of these posterior distributions Credi-
ble intervals overlapping zero imply that the parameter in
question is not different from zero Summaries of parameter
estimates were generated using the R coda package (Plummer
et al 2006)
Results
Temporal turnover
The mean value of TURN was 0384 (variance 00097)
This indicates that on average approximately 62 of
the species were common to both BBAs and 38 of spe-
cies had either colonized or become extinct within the
average block (Fig 1) The northeastern part of the
state which is the location of the Adirondack Moun-
tains experienced the highest rates of temporal turn-
over across BBA1980 and BBA2000 (Fig 1) Central and
northwestern New York underwent lowest temporal
turnover (Fig 1)
Model selection indicated that the variation in tempo-
ral turnover in avian communities was best accounted
for by the most complex model (Model 5 Table 2) Val-
ues of pseudo-R2 implied that Model 5 also explained a
slightly larger proportion of variation than other
models (Table 2) Similarly values of RMSE suggested
that the most complex model had slightly higher accu-
racy than the remaining models (Table 2) Selection of
the most complex model points to the importance of
the interactions between landscape fragmentation and
change in climatic conditions in explaining temporal
turnover in species composition between BBA1980 and
BBA2000 The values of RMSEpred indicated that pre-
dictive abilities of the most complicated model were
better than predictive abilities of the remaining models
(Table 2) However despite Model 5 having the best
predicting abilities it tended to overpredict low values
of TURN while high values tended to be underpredict-
ed (see Fig S5 in Appendix S1 for model validation plot
for the best model)
Parameter estimates from the top model indicated
that temporal turnover was positively related to
TMAXTrend suggesting that higher temporal turnover
was observed in regions where maximum temperatures
increased the most (Fig 2 see Table S2 in Appendix S1
for parameter estimates resulting from all temporal
turnover models) Temporal turnover was also posi-
tively associated with TMINTrend but the effect of
trends in minimum temperatures on TURN interacted
with ED (Fig 3a) In regions of low landscape fragmen-
tation large increases in minimum temperatures were
associated with high temporal turnover but the effect
of TMINTrend reversed and significantly diminished
in locations with high landscape fragmentation There
was a negative association of temporal turnover and
ED but the presence of multiple ED interaction terms
in the model makes it difficult to interpret the main
effects of ED on temporal turnover We found a posi-
tive relationship between temporal turnover and
DEVEL (Fig 2) indicating that highly urbanized
(a) (b) (c)
Fig 1 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) observed between 1980ndash
1985 and 2000ndash2005 in avian communities across New York Temporal turnover was calculated using New York State Breeding Bird
Atlas (BBA) as a proportion of species gained or lost between 1980ndash1985 (ie BBA1980) and 2000ndash2005 (ie BBA2000) in a particular
site (ie a Breeding Bird Atlas block) relative to all species recorded during at least one time period extinction was calculated as a pro-
portion of species lost between BBA1980 and BBA2000 in a particular site relative to all species present in a block during BBA1980 colo-
nization was calculated as a proportion of species gained between BBA1980 and BBA2000 in a particular site relative to all species
present in a block during BBA2000 High values of all three measures indicate high temporal turnover extinction and colonization
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2947
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
regions generally experienced increased temporal turn-
over The relationship between TURN and ELEV was
positive (Fig 2) suggesting that high-elevation regions
experienced higher temporal turnover
Extinction
The mean value of EXT for all species was 0210 (vari-
ance 00139) which indicates that on average approxi-
mately 21 of species of the original assemblage had
become extinct within the average block (Fig 1) The
northeastern part of the state (ie the Adirondack
Mountains) and regions surrounding the Finger Lakes
displayed higher than average rates extinction (Fig 1)
Model selection indicated that the most complex
model (Model 5) provided the best fit to the variation
in the proportion of extinction events (Table 2) Values
of pseudo-R2 implied that explanatory abilities of
Model 5 were slightly better than those of other models
(Table 2) Similarly values of RMSE suggested that
Model 5 had slightly higher accuracy than the remain-
ing models (Table 2) Selection of the most complex
model indicates that interactions between land-cover
fragmentation and change in climatic conditions con-
tribute to explaining extinction rates observed in avian
communities The values of RMSEpred indicated that
predictive abilities of Model 5 were slightly lower than
those of Model 3 but better than predictive abilities of
the remaining models (Table 2) Model 5 in general un-
derpredicted low values of extinction and overpredict-
ed high values of extinction (see Fig S5 in Appendix
S1)
Parameter estimates of the best model indicated that
extinction in avian communities was positively related
to TMINTrend although the effect of changes in mini-
mum temperatures on extinction rates interacted with
ED (Fig 2 see Table S3 in Appendix S1 for parameter
estimates resulting from all extinction models) In
regions of low landscape fragmentation large increases
in minimum temperatures were associated with high
proportion of extinction events but the effect of TMIN-
Trend diminished or reversed in locations with high
landscape fragmentation (Fig 3b) Extinction was posi-
tively associated with PRECIPTrend (Fig 2) but this
positive relationship between EXT and PRECIPTrend
was only detected in regions with high landscape frag-
mentation (Fig 3c) There was a negative association of
extinction and ED but the presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED on extinction in avian
communities There was a positive relationship
between extinction and DEVEL (Fig 2) indicating
increased proportion of extinction events in highly
urbanized regions We found a positive relationship
Table 2 Comparison of five competing models of temporal turnover proportion of extinction events and proportion of coloniza-
tion events using deviance information criterion (DIC) pseudo-R2 and root mean square error (RMSE) pD indicates the effective
number of parameters Predictive performance of the models was evaluated using RMSEpred Pseudo-R2 was calculated as 1 ndash SSres
SStot where SSres is the sum of squared differences between the observed and the predicted values of the dependent variable and
SStot is the sum of squared differences between the observed and the mean value of the dependent variable Larger values of
pseudo-R2 suggest better fit RMSE was calculated as the square root of the mean squared deviations between fitted and observed
outcomes RMSEpred was calculated as the square root of the mean squared deviations between the observed values from the hold-
out set and the predicted values resulting from all the models Lower values of RMSE and RMSEpred indicate improved accuracy
and better predictive power respectively For description of the models see Table 1
Measure of community change Model DDIC pD Pseudo-R2 RMSE RMSEpred
Temporal turnover Model 5 00 1093 0299 00825 00803
Model 3 195 1061 0296 00826 00805
Model 4 353 1075 0297 00826 00806
Model 2 591 1050 0294 00828 00808
Model 1 946 1023 0291 00830 00813
Extinction Model 5 00 1136 0316 00976 00971
Model 4 380 1135 0314 00978 00975
Model 3 398 1114 0313 00977 00970
Model 2 768 1099 0312 00978 00973
Model 1 1514 1107 0308 00981 00978
Colonization Model 5 00 1133 0205 01008 01006
Model 3 05 1098 0204 01008 01007
Model 4 91 1109 0204 01008 01007
Model 2 127 1089 0204 01009 01007
Model 1 231 1086 0203 01009 01008
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2948 M A JARZYNA et al
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
between EXT and ELEV indicating that high-elevation
regions experienced higher levels of temporal turnover
in avian communities
Colonization
The mean value of COL was 025 (variance 00128)
indicating that on average 25 of all species found in a
block during BBA2000 were the ones that had colonized
the site since BBA1980 Colonization was in general
higher in eastern and western parts of the state while
central and southern New York underwent lower colo-
nization rates (Fig 1) However the spatial pattern of
proportion of colonization events was less clumped
than the one displayed by temporal turnover or
extinction
Model selection procedure selected more than one
model as the top model accounting for the variation
in the proportion of colonization events The top two
models included the most complex model (Model 5)
and the model with main effects of all explanatory
variables but without the interaction terms (Model 3
Table 2) Values of pseudo-R2 implied that Model 5
also explained a slightly larger proportion of variation
than other models (Table 2) The values of RMSE sug-
gested that the most complex model also had slightly
higher accuracy than other models (Table 2) How-
ever despite the fact that the most complex model
was selected among the top models none of the inter-
action terms was associated with COL The most com-
plex model also showed the poorest predictive
abilities as indicated by the values of RMSEpred
(Table 2) Model validation also indicated that it un-
derpredicted the low values of colonization and over-
predicted high values of colonization (see Fig S5 in
Appendix S1)
Colonization was positively related to TMAXTrend
and negatively related to PRECIPTrend (Fig 2 see
Table S4 in Appendix S1 for parameter estimates result-
ing from all colonization models) Although ED was
negatively related to COL presence of multiple ED
interaction terms in the model makes it difficult to
interpret the main effects of ED We found a positive
relationship between colonization and DEVEL (Fig 2)
which indicates that highly urbanized regions
experienced highest colonization rates
Discussion
Our findings supported our hypothesis that changes in
avian community composition are related to changes in
climate but the strength of the relationship between
temporal turnover and climate change is affected by
the magnitude of landscape fragmentation Associa-
tions were generally weaker between community
change and climatic change in regions with widespread
habitat fragmentation but these relationships varied
based on temporal turnover extinction and coloniza-
tion dynamics We found that temporal turnover and
extinction were generally associated with increasing
rates of minimum temperature but these climate-medi-
ated patterns interacted strongly with fragmentation
and were diminished in fragmented areas Colonization
events were more likely in regions experiencing
increases in maximum temperature and decreases in
precipitation but unlike extinction patterns these rela-
tionships were not affected by fragmentation Although
we found that overall rates of community turnover
were higher in regions experiencing warming the
climatic drivers of temporal turnover extinction and
Fig 2 Coefficient estimates (ie median values of the posterior
distribution and associated 95 credible intervals) of the best
model for temporal turnover proportion of extinction events
and proportion of colonization events Abbreviations of the
explanatory variables are as follows magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season (TMAXTrend) magnitude of the 25-year (1980ndash
2005) trend in average minimum temperature of the breeding
season (TMINTrend) magnitude of the 25-year (1980ndash2005)
trend in average total precipitation of the breeding season (PRE-
CIPTrend) edge density (ED) percent developed land
(DEVEL) interaction between the magnitude of the 25-year
(1980ndash2005) trend in average maximum temperature of the
breeding season and edge density (TMAXTrendED) interac-
tion between the magnitude of the 25-year (1980ndash2005) trend in
average minimum temperature of the breeding season and edge
density (TMINTrendED) interaction between magnitude of the
25-year (1980ndash2005) trend in average total precipitation of the
breeding season and edge density (PRECIPTrendED) elevation
(ELEV) and survey effort (EFF)
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2949
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
colonization events varied geographically across the
study region
There are at least three plausible not mutually exclu-
sive ecological mechanisms that might be driving the
pattern of associations between temporal community
change climatic change and landscape fragmentation
First communities in contiguous landscapes might be
able to respond more quickly to changes in tempera-
tures and precipitation perhaps due to fewer barriers
to dispersal Such an interpretation would align with
research suggesting that unfragmented habitats allow
higher rates of species distributional shifts associated
with climate change (Opdam amp Washer 2004) If that
were the case communities in fragmented habitats
would lag behind changing climatic conditions and
therefore be less capable of adapting to shifting climate
space This lag would initially be reflected in communi-
ties being compositionally similar across time and thus
showing no temporal turnover However through
time it is likely that species will eventually disappear
from locations in which they can no longer find suitable
climatic conditions Such localized species extinctions
will ultimately result in higher community temporal
turnover and higher extinction rates perhaps on a par
with those currently observed in unfragmented
landscapes
While birds are highly mobile organisms not all
birds display similar dispersal abilities (Gaston amp
Blackburn 2002) There are significant differences in
dispersal distances between migratory and resident
bird species with migratory birds capable of dispersing
farther (Sutherland et al 2000) The differences in dis-
persal abilities might suggest that distributional shifts
of migratory birds might not be restricted by landscape
fragmentation to the same extent as those of resident
birds although research suggests otherwise For exam-
ple Villard et al (1995) showed that the area of land-
scape patches affects the spring resettlement by
Neotropical migratory birds suggesting that landscape
matrix might be one of the factors influencing resettle-
ment of returning migrants However mechanisms
through which migratory birds reoccupy their breeding
ranges are still poorly known and further research is
needed to explore whether landscape fragmentation
might be impeding distributional shifts of migratory
birds to the same extent as shifts of resident birds
How plausible is the unimpeded movement explana-
tion in New York State Using the same dataset
Zuckerberg et al (2009) documented shifts in breeding
ranges of 129 species of birds They found that the birds
exhibiting shifts in their range boundaries all had
different breeding habitat associations suggesting that
land cover was not a factor in the observed range
expansion Specifically they found no evidence for
more pronounced range shifts of forest breeding birds
in unfragmented habitats such as those of the Adiron-
dack and Catskill Mountains Given their findings it
seems unlikely that contiguous habitats facilitating the
climate change-driven range shifts may be at the root of
the higher rates of temporal turnover extinction and
colonization we are observing
A second plausible mechanism is that avian commu-
nities in fragmented landscapes are more robust to
changes in temperatures and precipitation resulting
from climate change than communities found in contig-
uous habitats If that were the case we would expect
little or no temporal change in community composition
and weaker associations with changing climatic
(a) (b) (c)
Fig 3 Effects size plots resulting from the best models for temporal turnover (a) and proportion of extinction events (b c) for the inter-
action of the magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) and
edge density (ED) and the interaction of the magnitude of the 25-year (1980ndash2005) trend in total precipitation of the breeding season
(PRECIPTrend) and edge density (ED) The interaction terms shown here are the ones whose credible intervals did not overlap zero
Yellow (temporal turnover) and blue (extinction) points are the observed data points and are plotted to indicate the levels of confidence
for the predicted surface
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2950 M A JARZYNA et al
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
conditions in fragmented landscapes Indeed heteroge-
neous and fragmented habitats are often composed of a
large proportion of habitat generalists (Tscharntke
et al 2012 Estavillo et al 2013) Habitat generalists
tend to have wider thermal breadths than habitat spe-
cialists (Barnagaud et al 2012) which potentially
allows them to tolerate greater changes in climatic con-
ditions On the other hand narrower thermal niches of
habitat specialists (Barnagaud et al 2012) might make
them vulnerable to even small increases in tempera-
tures Thus in this scenario we would expect commu-
nities in contiguous or less fragmented habitats (ie
comprised mainly of habitat specialists) to continue to
undergo relatively large compositional changes as a
result of climate change while those found in frag-
mented regions (ie comprised largely of generalists)
to show relatively small changes until perhaps some
thermal threshold is reached
Preliminary analysis of temporal turnover and pro-
portion of extinction and colonization events in com-
munities of birds belonging to different habitat
associations revealed that temporal community change
in generalist birds was lower than that of birds with
other more specialized habitat associations (eg forest
grassland wetland specialists Jarzyna MA data not
shown) This lower temporal change in communities of
generalist birds perhaps suggests that generalist birds
are indeed more robust to climatic changes lending
support to our second interpretation of the pattern of
stronger associations between community change and
climatic change in regions with unfragmented land-
scapes Furthermore Kampichler et al (2012) found
that highly contiguous regions such as forests and
heathlands in the Netherlands underwent large tempo-
ral change in community composition but the level of
community specialization remained constant and rela-
tively high Thus we suggest that relative contributions
of habitat generalists and specialists in ecological com-
munities might be at least partially contributing to the
divergent patterns of the associations between temporal
community change and climate change
Lastly the relative importance of stochastic and
deterministic processes in fragmented and contiguous
landscapes might be yet another factor contributing to
the observed pattern of relationships between temporal
turnover and changes in climatic conditions Turnover
in community composition in highly fragmented land-
scapes is generally pervasive (Krauss et al 2003
Borgella amp Gavin 2005 Banks-Leite et al 2012) This
high compositional turnover is expected to result from
small local population sizes which increase the effect
of demographic or environmental stochasticity (Gau-
blomme et al 2014) On the other hand stochastic
events tend to be less important in contiguous
landscapes because independent individual events tend
to average out in populations of larger sizes (Lande
1993 Goulart et al 2013) Thus in unfragmented habi-
tats deterministic processes such as changes in climatic
conditions are expected to be a major contributor to
turnover in community composition We were able to
attribute a larger portion of variation in temporal turn-
over to changing climatic conditions in regions of con-
tiguous land cover than in fragmented landscapes
While determining which of the last two alternative
explanations ndash relative contributions of habitat general-
ists and specialists or increased importance of deter-
ministic processes in contiguous habitats ndash is the main
mechanism driving the divergent community changendashclimate change associations is beyond the scope of this
manuscript we suggest that both processes are likely to
have played important roles
We found that high-elevation regions experienced
more pronounced changes in community composition
Elevation had the strongest influence on the values of
overall temporal turnover while proportion of extinc-
tion events was slightly less strongly affected by this
variable Indeed a positive effect of elevation on turn-
over in community composition is well supported by
scientific literature For example Stegen et al (2013)
found a positive relationship between habitat heteroge-
neity ndash measured as elevation range ndash and spatial and
temporal turnover in communities of breeding birds
across the continental United States while Coyle et al
(2013) found a positive effect of elevation on species
richness of North American birds In New York high-
elevation regions tend to also be more contiguous By
accounting for the influence of elevation on community
turnover we were able to disentangle the effects of
landscape fragmentation and elevation on the response
of avian communities to climate change
We recognize that rare species might be occasionally
missed during surveying potentially leading to percep-
tion that temporal turnover is low Additionally non-
random variation in the proportion of rare or poorly
detectable species across sites could potentially lead to
apparent patterns of high temporal turnover While
such bias resulting from presence of rare species within
communities has been well-documented in the litera-
ture (Vellend 2001 Bacaro amp Ricotta 2007) so far few
solutions have been offered
Our findings provide new insights into the drivers of
temporal change in avian communities and have signif-
icant implications for conservation theory and practice
Current conservation strategies commonly focus on
conserving large regions with undisturbed and contigu-
ous habitats that generally support high species diver-
sity often comprising of species with narrow habitat
niches (ie habitat specialists) Our research suggests
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2951
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
that while ecological communities in these protected
areas will be less subject to stochastic variation they
will be more sensitive to changing environmental con-
ditions and will likely undergo the most drastic compo-
sitional changes as a result of climate change It is
therefore necessary that current conservation strategies
are put in the context of climate change
Acknowledgements
We would like to thank thousands of volunteers who partici-pated in both New York State Breeding Bird Atlases We alsothank Kimberley Corwin and Kevin McGowan for supplyingatlas databases and Colin M Beier Daniel Bishop and JohnWiley for supplying climate data We would like to thankAllen Hurlbert and an anonymous reviewer for helpful com-ments on the earlier versions of the manuscript The manu-script benefited from discussions with members of the Booneand Crockett Quantitative Wildlife Center at Michigan StateUniversity This study received financial support from NASAGrant NNXO9AK16G and the Boone and Crockett ClubAndrew O Finley was supported by the National ScienceFoundation grants DMS-1106609 EF-1137309 EF-1241868 andEF-125322
References
Andrle RF Carroll JR (1988) The Atlas of Breeding Birds in New York State Cornell Uni-
versity Press Ithaka NY
Bacaro G Ricotta C (2007) A spatially explicit measure of beta diversity Community
Ecology 8 41ndash46
Banerjee S Carlin BP Gelfand AE (2004) Hierarchical Modeling and Analysis for Spatial
Data Chapman amp HallCRC Boca Raron FL
Banks-Leite C Ewers RM Metzger JP (2012) Unraveling the drivers of community
dissimilarity and species extinction in fragmented landscapes Ecology 93 2560ndash
2569
Barnagaud J-Y Devictor V Jiguet F Barbet-Massin M Le Viol I Archaux F (2012)
Relating habitat and climatic niches in birds PLoS ONE 7 e32819
Barton PS Cunningham SA Manning AD Gibb H Lindenmayer DB Didham RK
(2013) The spatial scaling of beta diversity Global Ecology and Biogeography 22
639ndash647
Borgella R Jr Gavin TA (2005) Avian community dynamics in a fragmented tropical
landscape Ecological Applications 15 1062ndash1073
Brotons L Jiguet F (2010) Bird communities and climate change In Effects of Climate
Change on Birds (eds Moller AP Fielder W Berthold P) pp 275ndash294 Oxford Uni-
versity Press Oxford
Buckley LB Jetz W (2008) Linking global turnover of species and environments Pro-
ceedings of the National Academy of Sciences of the United States of America 105
17836ndash17841
Coyle JR Hurlbert AH White EP (2013) Opposing mechanisms drive richness pat-
terns of core and transient bird species The American Naturalist 181 E83ndashE90
Cressie N Wikle CK (2011) Statistics for Spatio-Temporal Data John Wiley and Sons
Hoboken NJ
Daly C Gibson W (2002) Parameter-estimation on Independent Slopes Model (PRISM)
The PRISM Climate Group Corvallis OR
Devictor V Julliard R Couvet D Jiguet F (2008) Birds are tracking climate warming
but not fast enough Proceedings of the Royal Society B 275 2743ndash2748
Estavillo C Pardini R da Rocha PLB (2013) Forest loss and the biodiversity threshold
an evaluation considering species habitat requirements and the use of matrix habi-
tat PLoS ONE 8 e823369
Finley AO Banerjee S Carlin BP (2007) spBayes an R package for univariate and
multivariate hierarchical point-referenced spatial models Journal of Statistical Soft-
ware 19 1ndash24
Fry JA Coan MJ Homer CG Meyer DK Wickham JD (2009) Completion of the
National Land Cover Database (NLCD) 1992ndash2001 Land Cover Change Retrofit
product pp 18 US Geological Survey Open-File Report 2008ndash1379
Gaston KJ Blackburn TM (2002) Large-scale dynamics in colonization and extinction
for breeding birds in Britain Journal of Animal Ecology 71 390ndash399
Gaston KJ Davies RG Orme CDL et al (2007) Spatial turnover in the global avifauna
Proceedings of the Royal Society B-Biological Sciences 274 1567ndash1574
Gaublomme E Eggermont H Hendrickx F (2014) Local extinction processes rather
than edge effects affect ground beetle assemblages from fragmented and urban-
ised old beech forests Insect Conservation and Diversity 7 82ndash90
Gelman A Rubin DB (1992) Inference from iterative simulation using multiple
sequences Statistical Science 7 457ndash511
Gelman A Carlin JB Stern HS Rubin DB (2004) Bayesian Data Analysis 2nd edn
Chapman amp HallCRC Boca Raton FL
Gibbons DW Donald PF Bauer H-G Fornasari L Dawson IK (2007) Mapping avian
distributions the evolution of bird atlases Bird Study 54 324ndash334
Goulart FF Salles P Machado RB (2013) How may agricultural matrix intensification
affect understory birds in an Atlantic Forest landscape A qualitative model on
stochasticity and immigration Ecological Informatics 18 93ndash106
Gutierrez-Canovas C Millan A Velasco J Vaughan IP Ormerod SJ (2013) Contrast-
ing effects of natural and anthropogenic stressors on beta diversity in river organ-
isms Global Ecology and Biogeography 22 796ndash805
Hargis CD Bissonette JA David JL (1998) The behavior of landscape metrics com-
monly used in the study of habitat fragmentation Landscape Ecology 13 167ndash186
Homer C Huang C Yang L Wylie B Coan M (2004) Development of a 2001 national
landcover database for the United States Photogrammetric Engineering and Remote
Sensing 70 829ndash840
Jaccard P (1912) The distribution of the flora in the alpine zone New Phytologist 11
37ndash50
Jeltsch F Moloney KA Schwager M Keuroorner K Blaum N (2011) Consequences of cor-
relations between habitat modifications and negative impact of climate change for
regional species survival Agriculture Ecosystems and Environment 145 49ndash58
Kampichler C Turnhout CAM Devictor V van der Jeugd HP (2012) Large-scale
changes in community composition determining land use and climate change sig-
nals PLoS ONE 7 e35272
Krauss J Steffen-Dewenter I Tscharntke T (2003) Local species immigration extinc-
tion and turnover of butterflies in relation to habitat area and habitat isolation
Oecologia 442 591ndash602
La Sorte FA Thompson FR (2007) Poleward shifts in winter ranges of North Ameri-
can birds Ecology 88 1803ndash1812
Lande R (1993) Risks of population extinction from demographic and environmental
stochasticity and random catastrophes The American Naturalist 142 911ndash927
Marini MA Barbet-Massin M Lopes LE Jiguet F (2009) Predicted climate-driven bird
distribution changes and forecasted conservation conflicts in a neotropical
savanna Conservation Biology 23 1558ndash1567
McGarigal K Cushman SA Ene E (2012) FRAGSTATS v4 Spatial Pattern Analysis
Program for Categorical and Continuous Maps Computer software program pro-
duced by the authors at the University of Massachusetts Amherst MA
McGowan M Corwin K (2008) The Second Atlas of Breeding Birds in New York State
Cornell University Press Ithaka NY
McGowan K Zuckerberg B (2008) Summary of results In The Second Atlas of Breeding
Birds in New York State (eds McGowan K Corwin K) pp 15ndash42 Cornell University
Press Ithaka NY
Opdam P Wascher D (2004) Climate change meets habitat fragmentation linking
landscape and biogeographical scale levels in research and conservation Biological
Conservation 117 285ndash297
Parmesan C Yohe G (2003) A globally coherent fingerprint of climate change impacts
across natural systems Nature 421 37ndash42
Plummer M Best N Cowles K Vines K (2006) CODA convergence diagnosis and
output analysis for MCMC R News 6 7ndash11
Prince K Zuckerberg B (2015) Climate change in our backyards the reshuffling
of North Americarsquos winter bird communities Global Change Biology 21 572ndash
585
Siefert A Ravenscroft C Weiser MD Swenson NG (2013) Functional beta-diversity
patterns reveal deterministic community assembly processes in eastern North
American trees Global Ecology and Biogeography 22 682ndash691
Spiegelhalter DJ Best NG Carlin B van der Linde A (2002) Bayesian measures of
model complexity and fit (with discussion) Journal of the Royal Statistical Society
Series B 64 583ndash639
Stegen JC Freestone AL Crist TO et al (2013) Stochastic and deterministic drivers of
spatial and temporal turnover in breeding bird communities Global Ecology and
Biogeography 22 202ndash212
Sutherland GD Harestad AS Price K Lertzman KP (2000) Scaling of natal dispersal
distances in terrestrial birds and mammals Conservation Ecology 4 Art 16
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
2952 M A JARZYNA et al
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953
Tobler MW Carrillo-Percastegui SE Leite Pitman R Mares R Powell G (2008) An
evaluation of camera traps for inventorying large- and medium-sized terrestrial
rainforest mammals Animal Conservation 11 169ndash178
Travis JMJ (2003) Climate change and habitat destruction a deadly anthropogenic
cocktail Proceedings of the Royal Society B-Biological Sciences 270 467ndash473
Tscharntke T Tylianakis JM Rand TA et al (2012) Landscape moderation of
biodiversity patterns and processes ndash eight hypotheses Biological Reviews 87 661ndash
685
Tuomisto H (2010a) A diversity of beta diversities straightening up a concept gone
awry Part 2 Quantifying beta diversity and related phenomena Ecography 33
23ndash45
Tuomisto H (2010b) A diversity of beta diversities straightening up a concept gone
awry Part 1 Defining beta diversity as a function of alpha and gamma diversity
Ecography 33 2ndash22
Vellend M (2001) Do commonly used indices of b-diversity measure species turn-
over Journal of Vegetation Science 12 545ndash552
Villard M-A Merriam G Maurer BA (1995) Dynamics and subdivided populations
of Neotropical migratory birds in a fragmented temperate forest Ecology 76 27ndash
40
Walther GR Post E Convey P et al (2002) Ecological responses to recent climate
change Nature 416 389ndash395
Warren MS Hill JK Thomas JA et al (2001) Rapid responses of British butterflies to
opposing forces of climate and habitat change Nature 414 65ndash69
Zuckerberg B Woods AM Porter WF (2009) Poleward shifts in breeding bird distri-
butions in New York State Global Change Biology 15 1866ndash1883
Supporting Information
Additional Supporting Information may be found in the online version of this article
Appendix S1
Table S1 List of all bird species (N = 256) from the New York State Breeding Bird Atlas (BBA) used in the analysisTable S2 Parameter estimates (brsquos) of each covariate resulting from all tested models of temporal turnover (starting with the bestmodel as indicated by DIC)Table S3 Parameter estimates (brsquos) of each covariate resulting from all tested models of extinction (starting with the best model asindicated by DIC)Table S4 Coefficient estimates (brsquos) of each covariate resulting from all tested models of colonization (starting with the best modelas indicated by DIC)Figure S1 Percent of the New York State Breeding Bird Atlas blocks that underwent land cover or land use change between 1992and 2001 The NLCD 19922001 Retrofit Land Cover Change Product was used to quantify the amount of land cover changeFigure S2 Spatial distribution of all explanatory variables magnitude of the 25-year (1980ndash2005) trend in average maximum tem-peratures (a) magnitude of the 25-year (1980ndash2005) trend in average minimum temperatures (b) magnitude of the 25-year (1980ndash2005) trend in total precipitation (c) edge density (d) percent developed land (e) and elevation (f)Figure S3 Scatterplot and histograms of all explanatory variables Abbreviations of the explanatory variables are as follows magni-tude of the 25-year (1980ndash2005) trend in average maximum temperature of the breeding season (TMAXTrend) magnitude of the 25-year (1980ndash2005) trend in average minimum temperature of the breeding season (TMINTrend) magnitude of the 25-year (1980ndash2005) trend in average total precipitation of the breeding season (PRECIPTrend) edge density (ED) percent developed land(DEVEL) elevation (ELEV)Figure S4 Temporal turnover (a) proportion of extinction events (b) and proportion of colonization events (c) as a function of tem-poral lag Data from 128 Breeding Bird Survey routes in New York State surveyed between 1980 and 2005Figure S5 Model validation for the best model for temporal turnover proportion of extinction events and proportion of coloniza-tion events observed between BBA1980 and BBA2000 using plots of predicted vs observed values
copy 2015 John Wiley amp Sons Ltd Global Change Biology 21 2942ndash2953
RESPONSES OF COMMUNITIES TO CLIMATE CHANGE 2953