PRIMARY RESEARCH PAPER

Projected compositional shifts and loss of ecosystem servicesin freshwater fish communities under climate changescenarios

Shekhar R. Biswas . Richard J. Vogt . Sapna Sharma

Received: 3 November 2016 / Revised: 17 April 2017 / Accepted: 18 April 2017

� Springer International Publishing Switzerland 2017

Abstract What are the projected impacts of climate

change on community composition and consequen-

tially on the distribution of functional traits? Answers

to these questions are somewhat unclear but critical for

designing ecological management strategies. Here we

forecast potential impacts of climate change on

freshwater lake fish communities of Ontario, Canada,

by contrasting species composition, species richness,

functional diversity and functional composition for

present versus projected communities under ‘‘best-

case’’ and ‘‘business-as-usual’’ climate change sce-

narios. Results indicate that the composition of

projected communities differs from present, and

includes a shift from cold- and cool-water species to

warm-water species. Species richness in projected

communities is estimated to increase by 60–81%, but

functional diversity is estimated to decline. These

projected communities are estimated to have on

average 22% shorter mean body length, 38% lighter

body weight and 36% less fecundity than present.

Also, the present configuration of sport and commer-

cially important fishes are projected to decline in their

distribution, potentially impacting ecosystem services

associated with commercial and recreational fisheries.

Together, climate change may initiate a compositional

shift that may result in an important shift in commu-

nity functional structure, which is likely to affect

important aquatic ecosystem services.

Keywords Compositional shift � Ecosystemservices � Species distribution models � Trait analysis

Introduction

Broad-scale shifts in species distributions in response

to climate change are well documented (Parmesan &

Yohe, 2003; Chen et al., 2011) and are typically a

result of species having to respond to local environ-

ments according to their thermal habitat requirements

(Tonn & Magnuson, 1982). Of course, changes in

species distributions will have an impact on local

community dynamics, composition and diversity

Communicated by Handling editor: Begona Santos

Electronic supplementary material The online version ofthis article (doi:10.1007/s10750-017-3208-1) contains supple-mentary material, which is available to authorized users.

S. R. Biswas � S. Sharma

Department of Biology, York University, 4700 Keele

Street, Toronto, ON M3J 1P3, Canada

S. R. Biswas (&)

Faculty of Natural Resources Management, Lakehead

University, 955 Oliver Road, Thunder Bay, ON P7B 5E1,

Canada

e-mail: shekhar.forest@gmail.com

R. J. Vogt

Department of Biological Sciences, University of Quebec

at Montreal, CP 8888, Succ. Centre Ville, Montreal,

QC H3C 3P8, Canada

123

Hydrobiologia

DOI 10.1007/s10750-017-3208-1

(Walther, 2010; Simpson et al., 2011) via ecological

interactions that result as species begin to invade novel

habitats. Since each species can play relatively unique

ecological functions and deliver somewhat unique

ecological services (Dıaz et al., 2007), any such

change in local species composition or diversity

should impact ecosystem functions and services

(Cardinale et al., 2012; Hooper et al., 2012), or the

resilience of ecological communities (Oliver et al.,

2015a, b). These community-scale consequences of

climate change are still somewhat poorly understood

but important for designing ecological management

strategies for the continued provision of important

ecosystem services.

One way to anticipate the impact of climate change

on community-scale ecosystem functions and services

is to project future communities under climate change

scenarios (Guisan & Thuiller, 2005; Botkin et al.,

2007), followed by comparative analyses of commu-

nity trait structure for present versus projected com-

munities. Here functional traits allow quantification of

species characteristics (i.e. morphological, structural

and behavioural.) that influence species performance

or fitness (Violle et al., 2007; Nock et al., 2016). By

emphasizing the characteristics that define how organ-

isms interact with one another and their changing

environments, a functional trait approach can provide

insights into the ecological consequences of regional

shifts in species distributions (Hooper et al., 2005). In

fact, as the pattern of trait distribution of a lake

changes, inferences can be made regarding systematic

changes in ecosystem functioning or provision of

ecosystem services (Civantos et al., 2012; Thuiller

et al., 2014; Barbet-Massin & Jetz, 2015; Gauzere

et al., 2015; Mokany et al., 2015). For example, a

large-scale shift in species’ size or body mass distri-

butions will have ramifications for overall ecosystem

productivity (Millennium-Ecosystem-Assessment,

2005). Further, future trait distributions can be fore-

casted once present day relationships between traits

and local environments are understood (Mokany et al.,

2016).

Here we focus on modelling freshwater fish species

and trait distributions, which define how co-occurring

species interact with one another and their habitats,

across a landscape of lakes and for various regional

climate warming scenarios. Changes in freshwater fish

species distributions in response to warming climates

and changing availability of suitable thermal habitat

have been well documented in lakes (Shuter et al.,

1980; Magnuson et al., 1990, 1997; Alofs et al., 2014).

Typically, coldwater fishes are defined as fishes with

an optimal thermal range between 8 and 12�C, cool-water fishes as fishes with an optimal thermal range

between 16 and 20�C, and warm-water fishes as fishes

with an optimal thermal range between 22 and 26�C(Christie & Regier, 1988; Magnuson et al., 1990).

Under warming climates, warm-water fishes, such as

centrarchids (e.g. smallmouth bass and rock bass),

have invaded lakes at the northern extent of their range

(Alofs et al., 2014), and coldwater fishes, such as

cisco, have become extirpated from lakes near the

southern extents of their range (Sharma et al., 2011).

Species range shifts like these are predicted to become

exacerbated under scenarios of climate change with

forecasts of range expansion of predatory warm-water

fishes, northern shifts of cool-water fishes and extir-

pation of coldwater fishes from inland lakes (Chu

et al., 2005; Sharma et al., 2007; Van Zuiden et al.,

2016). Ramifications for the remaining fishes in the

community are still poorly understood.

We therefore aim to understand how changing

climates may impact freshwater fish community

composition and functional trait distributions across

a landscape of lakes to assess the impacts of climate

change on (i) species composition, (ii) species rich-

ness, (iii) functional diversity and (iv) functional

composition. We compared present day communities

with those forecast under business-as-usual and best-

case climate change scenarios in 2070 using a

700-lake dataset from Ontario, Canada. This region

is compositionally dynamic, as it includes both the

present northern range limit of warm-water fishes and

the southern extent of the range of cold-water species

(Chu et al., 2005; Sharma et al., 2007; Van Zuiden

et al., 2016). By 2070, as warm-water habitats become

more prevalent, we expect that species richness will

increase in many lakes as new species begin to invade

what were once predominantly colder habitats. Cor-

respondingly, lake functional diversity will increase if

warm-water invaders introduce new functional char-

acteristics to community trait distributions in lakes

altered by a warming climate. Further, we explore the

ramifications of changes to fish community structure

and trait distributions for provision of ecosystem

services that may result from climate warming.

Hydrobiologia

123

Methods

Study area and data

The study area is inland lakes in Ontario, Canada

(Fig. 1). These lakes support a wide range of com-

mercial and recreation fisheries including walleye

(Sander vitreus), whitefish (Coregonus clupeiformis),

perch (Perca flavescens), pike (Esox Lucius) and lake

trout (Salvelinus namaycush). The Ontario Ministry of

Natural Resources and Forestry Broad-scale Monitor-

ing (OMNRF-BsM) Program aims to collect data on

*700 Ontario lakes every five years to inform

landscape level management; and to-date has col-

lected detailed information on lake fish communities

and lake characteristics from 2008 to 2012 for over

700 lakes (Sandstrom et al., 2010). Data that were

collected included latitude, longitude, mean lake

depth, lake surface area, secchi depth, pH and

presence/absence of 129 fish species. Fish data were

collected between June–September, using both large-

and small-mesh gillnets (Sandstrom et al., 2010).

We obtained the most recent (2008–2012) geo-

referenced fish community dataset from OMNRF-

BsM program for 722 inland lakes. Our final lake

dataset consisted of 645 lakes, following the removal

of lakes that were species poor (species richness was

\5% of total richness in the dataset). We also removed

rare species (occurrence in less than 5% of lakes)

resulting in a regional species richness of 40 species

(i.e. a total of 40 species were included in the final

dataset). Rare species and sites can have a dispropor-

tionate influence on statistical analysis (Jackson &

Harvey, 1989).

For the same set of lakes, we obtained historical

climate data (monthly mean temperature and precip-

itation) represented as climate averages from 1950 to

2000 fromWorldClim database (Hijmans et al., 2005),

Fig. 1 Map of Ontario highlighting study lakes and lake-wise present mean annual air temperature (�C) averaged over the period of

1950–2000. Climatic data were obtained from the WorldClim database, with a spatial resolution of 30 arc-seconds (*1 km)

Hydrobiologia

123

with a resolution of 30 arc-seconds (*1 km). When

deriving climate estimates and estimating future

climate change projections, the IPCC recommends

the use of climate normals from 1950 to 2000 to reduce

the influence of inter-annual variability in weather on

climate forecasts (IPCC, 2013). By using the averaged

climate over decades, we are focusing on the effects of

climate variability on fish populations, rather than

weather. The timeline difference for climate data

(1950–2000) and fish data (2008–2012) should not be

a problem because species distributional shift is a

gradual process; this climatic baseline is widely used

in the species distribution modelling; and presently,

this is the most updated climate normal (at a fine

spatial resolution of 1 km2) we obtained. The most

conservative greenhouse gas emissions or ‘‘best-case’’

scenario (Representative Concentration Pathway,

RCP = 2.6) and ‘‘business-as-usual’’ greenhouse gas

emissions scenarios (RCP = 8.5) were obtained from

the Canadian Centre for Climate Modelling and

Analysis (CCMA) CMIP5 (IPPC Fifth Assessment)

Global Climate Model (GCM) for 2070. Here the year

2070 represents climate projections averaged for

2060–2080. The mean annual air temperatures for

our study lakes are predicted to increase by

1.81 ± 1.35�C for RCP 2.6 and 3.63 ± 1.31�C for

RCP 8.5 (Fig. S1 in Electronic Supplementary

Materials).

Projecting species composition and generating

future communities

Generalized linear models were developed for all 40

fish species to understand the environmental factors

structuring the occurrence of each species in the

community (Table 1). Eighty percent of the dataset

was used to train the models (N = 516 lakes), and the

remaining 20% of the data were used to validate the

models (N = 129 lakes). That is, we first randomly

split the whole dataset into training and validation sets;

and then, the same training dataset was used for

modelling the occurrence of each species and the same

validation dataset was used for validating the occur-

rence of each species. While the response variable was

the presence/absence of each species, explanatory

variables included lake surface area, mean depth,

secchi depth, seasonal mean air temperatures, seasonal

precipitation and mean annual air temperatures and

precipitation. These environmental and climatic

variables are known to be relevant for structuring

Ontario fish communities (Jackson & Harvey, 1989).

Environmental variables were tested for normality and

multicollinearity (see Electronic Supplementary

Materials, Fig. S2). Forward selection with a dual

criterion (a = 0.05 and R2adj) was used to identify

significant environmental predictor variables for each

species (Blanchet et al., 2008).

Receiver Operating Characteristics (ROC) curves

were used to identify thresholds that maximize the

sensitivity (ability to correctly predict species pres-

ence) and specificity (ability to correctly predict

species absence) for each species distribution model.

This procedure is recommended when species pres-

ences and absences are not equal within the data

(Sharma & Jackson Sharma & Jackson, 2008). We

retained the best model for each species based on the

lowest Akaike Information Criterion (AIC) value.

Following the development and validation of species

distribution models and appropriate ROC thresholds

for each species, we projected the generalized linear

model in 2070 using the most conservative scenario

(RCP = 2.6) and business-as-usual scenario

(RCP = 8.5) of the CCMAGCM. Finally, by stacking

the predicted occurrence of all 40 species in each lake,

lake-wise future communities were generated (Guisan

& Zimmermann, 2000; Dubuis et al., 2011; Guisan &

Rahbek, 2011). One of the major advantages of this

stacking approach to construct lake-wise future com-

munities is that they yield lake-specific species

composition (Zurell et al., 2016), which is essential

for the follow up functional trait analyses involving

both quantitative and categorical traits. However, our

bio-climate based projections essentially reflect the

aspect of habitat suitability (Araujo & Peterson, 2012)

and did not consider other constrains that could also

limit species occurrence at a particular lake, such as

biotic interactions, dispersal or historical factors

(Guisan & Rahbek, 2011; Zurell et al., 2016). We

therefore evaluated our projection efficiency (model

uncertainty) and found satisfactory at the level of

individual species (sensitivity range: 0.50–0.86; speci-

ficity range: 0.51–0.94; see Table 1) but over predic-

tion for species richness (correlation between

projected richness and actual richness, r = 0.61,

n = 129 validation lakes), which is a well acknowl-

edged phenomenon in the stacking approach of

community generation (Zurell et al., 2016). As such,

our projected communities should be considered a

Hydrobiologia

123

Table

1Species-wiselogisticmodelsforpredictingfuture

distributionsoffishes

ininlandlakes

ofOntario

Species

Localnam

eModel

usedto

predictfuture

presence

Optimal

threshold

Sensitivity

Specificity

Noof

occurrence

Salvelinusfontinalis(M

itchill,1814)

Brooktrout

1=ð1

þe�

ð11:87�0:64tm

eanjja�2:23logðsaÞþ3:41logðsdÞÞÞ

0.20

0.60

0.92

64

Salvelinusnamaycush

(Walbaum,1792)

Laketrout

1=ð1

þe�

ð�2:27�0:42tm

eanjjaþ7:04logðm

dÞþ

5:20logðsdÞÞÞ

0.47

0.82

0.83

286

Coregonusclupeiform

is(M

itchill,1818)

Lakewhitefish

1=ð1

þe�

ð0:61þ1:92logðsaÞ�0:38tm

eanjjaþ0:70logðm

dÞÞÞ

0.56

0.72

0.72

360

CoregonusartediLesueur,1818

Cisco

1=ð1

þe�

ð�4:44þ1:73logðsaÞþ2:29logðm

d�

2:53logðsdÞ�

0:17tm

eanatÞ Þ

0.57

0.82

0.68

402

Osm

erusmordax(M

itchill,1814)

Rainbow

smelt

1=ð1

þe�

ð�19:71þ2:39logðm

dÞþ

0:08prec s

onþ0:93logðsaÞþ0:37tm

eanjjaÞ Þ

0.11

0.55

0.82

62

EsoxluciusLinnaeus,1758

Northernpike

1=ð1

þe�

ð�0:80þ1:81logðsaÞ�0:24tm

eanat�4:19logðsdÞÞÞ

0.71

0.80

0.63

488

EsoxmasquinongyMitchill,1824

Muskellunge

1=ð1

þe�

ð�22:18þ0:96tm

eanjjaþ1:18logðsaÞ�2:25logðm

dÞþ

2:87logðsdÞÞÞ

0.22

0.67

0.93

42

Catostomuscatostomus(Forster,1773)

Longnose

sucker

1=ð1

þe�

ð8:91�0:76tm

eanjjaþ2:42logðm

d�

1:53logðsdÞÞÞ

0.17

0.50

0.85

77

Catostomuscommersonii(Lacepede,1803)

Whitesucker

1=ð1

þe�

ð2:55�0:54tm

eansonþ1:34logðsaÞÞ Þ

0.91

0.88

0.57

617

Moxostomamacrolepidotum

(Lesueur,

1817)

Shorthead

redhorse

1=ð1

þe�

ð�13:64þ1:43logðsaÞþ0:10prec jja�1:59logðm

dÞÞÞ

0.20

0.64

0.88

83

ChrosomuseosCope,

1861

Northernredbelly

dace

1=ð1

þe�

ð�3:25�2:03logðsaÞþ0:07prec s

onÞ Þ

0.16

0.50

0.91

33

Couesiusplumbeus(A

gassiz,

1850)

Lakechub

1=ð1

þe�

ð5:76þ3:00logðsdÞ�

0:57tm

eanjjaÞ Þ

0.14

0.67

0.69

87

Notemigonuscrysoleucas(M

itchill,1814)

Golden

shiner

1=ð1

þe�

ð�12:52þ0:84tm

eanjja�0:81logðm

d�

0:75logðsaÞÞ Þ

0.23

0.61

0.72

146

Notropisatherinoides

Rafinesque,

1818

Emeraldshiner

1=ð1

þe�

ð�2:96þ1:15logðsaÞ�1:28logðsdÞ�

0:22tm

eansonÞÞÞ

0.29

0.66

0.79

147

Luxiluscornutus(M

itchill,1817)

Commonshiner

1=ð1

þe�

ð�7:63þ0:07prec s

onþ1:20logðm

dÞÞÞ

0.17

0.70

0.64

146

Notropisheterodon(Cope,

1865)

Blackchin

shiner

1=ð1

þe�

ð�20:94þ1:07tm

eanjjaÞ Þ

0.05

0.75

0.56

45

Notropisheterolepis

Eigenmann&

Eigenmann,1893

Blacknose

shiner

1=ð1

þe�

ð�2:16�0:15tm

eanatþ1:86logðsdÞÞÞ

0.17

0.67

0.51

114

Notropishudsonius(Clinton,1824)

Spottailshiner

1=ð1

þe�

ð�1:81þ1:72logðsaÞ�1:54logðm

d�

0:38tm

eansonÞ Þ

0.47

0.69

0.69

320

Notropisvolucellus(Cope,

1865)

Mim

icshiner

1=ð1

þe�

ð�22:05þ0:13prec jjaþ2:39logðsdÞþ

0:43tm

eanjjaÞ Þ

0.10

0.75

0.79

41

Pimephalesnotatus(Rafinesque,

1820)

Bluntnose

minnow

1=ð1

þe�

ð�17:71þ0:79tm

eanjjaþ1:30logðm

dÞþ

2:58logðsdÞÞÞ

0.24

0.68

0.79

102

PimephalespromelasRafinesque,

1820

Fatheadminnow

1=ð1

þe�

ð7:61�1:79logðsaÞ�0:49tm

eanjjaþ3:71logðsdÞÞÞ

0.07

0.71

0.81

37

Sem

otilusatromaculatus(M

itchill,1818)

Creek

chub

1=ð1

þe�

ð�3:90�2:65logðsaÞþ0:06prec s

onþ2:75logðm

dÞÞÞ

0.05

0.75

0.80

37

Margariscusmargarita

(Cope,

1867)

Pearldace

1=ð1

þe�

ð�1:29þ3:02logðsdÞ�

0:76logðsaÞ�0:23tm

eansonÞ Þ

0.07

0.70

0.69

42

Ameiurusnebulosus(Lesueur,1819)

Brownbullhead

1=ð1

þe�

ð�16:15þ0:88tm

eanjjaÞÞÞ

0.28

0.72

0.85

136

Lota

lota

(Linnaeus,1758)

Burbot

1=ð1

þeð

�ð�

4:67þ2:01logðsaÞþ3:76logðm

d�

0:29tm

eanjjaÞ Þ

0.47

0.70

0.78

266

Pungitiuspungitius(Linnaeus,1758)

Ninespine

stickleback

1=ð1

þe�

ð�8:05þ1:46logðsaÞþ3:11logðsdÞ�

0:22tm

eanatÞ Þ

0.11

0.71

0.81

55

Hydrobiologia

123

Table

1continued

Species

Localnam

eModel

usedto

predictfuture

presence

Optimal

threshold

Sensitivity

Specificity

Noof

occurrence

Percopsisomiscomaycus(W

albaum,1792)

Troutperch

1=ð1

þe�

ð�2:61þ1:93logðsaÞ�3:19logðsdÞ�

0:32tm

eansonÞ Þ

0.48

0.79

0.74

282

Ambloplitesrupestris

(Rafinesque,

1817)

Rock

bass

1=ð1

þe�

ð�17:93þ0:92tm

eanjjaþ0:91logðsaÞÞ Þ

0.54

0.70

0.81

294

Lepomisgibbosus(Linnaeus,1758)

Pumpkinseed

1=ð1

þe�

ð�20:65þ1:19tm

eanjjaÞÞÞ

0.32

0.73

0.67

196

LepomismacrochirusRafinesque,

1819

Bluegill

1=ð1

+e�

ð�24:26þ1:29tm

eanjja�2:54logðm

dÞþ

3:67logðsdÞÞÞ

0.23

0.62

0.89

59

MicropterusdolomieuLacepede,

1802

Smallm

outh

bass

1=ð1

+e�

ð�19:74þ0:98tm

eanjjaþ0:82logðsaÞþ1:55logðm

dÞÞÞ

0.54

0.79

0.76

329

Micropterussalmoides

(Lacepede,

1802)

Largem

outh

bass

1=ð1

þe�

ð�5:36þ1:21tm

eanmam�3:55logðm

dÞþ

3:32logðsdÞþ

0:60logðsaÞÞ Þ

0.30

0.86

0.94

101

Pomoxisnigromaculatus(Lesueur,1829)

Black

crappie

1=ð1

þe�

ð�22:58þ1:27tm

eanjja�1:75logðm

dÞÞÞ

0.20

0.75

0.86

65

Perca

flavescens(M

itchill,1814)

Yellow

perch

1=ð1

+e�

ð�11:97þ2:41logðsaÞþ0:61tm

eanjja�3:88logðsdÞÞÞ

0.85

0.84

0.67

567

Sander

canadensis(G

riffith

&Smith,1834)

Sauger

1=ð1

þe�

ð�6:47�5:74logðsdÞþ

2:02logðsaÞ�0:32tm

eanatÞÞÞ

0.07

0.63

0.88

38

Sander

vitreus(M

itchill,1818)

Walleye

1=ð1

þe�

ð�4:21þ1:99logðsaÞ�1:93logðm

dÞþ

0:18tm

eanjja�2:70logðsdÞÞÞ

0.64

0.82

0.57

469

Etheostomanigrum

Rafinesque,

1820

Johnnydarter

1=ð1

þe�

ð�8:45�0:33tm

eanatþ4:18logðsdÞþ

1:23logðsaÞÞ Þ

0.08

0.75

0.87

33

Percinacaprodes

(Rafinesque,

1818)

Logperch

1=ð1

þe�

ð�8:09þ1:18logðsaÞþ1:54logðsdÞþ

0:17tm

eanjjaÞÞÞ

0.31

0.60

0.70

171

CottusbairdiiGirard,1850

Mottledsculpin

1=ð1

þe�

ð�5:86þ1:55logðm

dÞþ

0:61logðsaÞÞ Þ

0.09

0.86

0.78

45

CottuscognatusRichardson,1836

Slimysculpin

1=ð1

þe�

ð�6:75þ2:33logðm

dÞþ

0:03prec s

onÞ Þ

0.06

0.86

0.59

49

sdsecchidepth,sa

surfacearea,mdmeandepth,tmeanjjameansummer

temperature

(June–July–August),tmeansonmeanfalltemperature

(September–October–Novem

ber),

tmeandjfmeanwintertemperature

(Decem

ber–January–February),tmeanmammeanspringtemperature

(March–April–May)

Hydrobiologia

123

generous estimate and should be interpreted with

caution.

Trait data and community-weighted traits

Rao’s quadratic entropy, a measure of functional

diversity, was computed by using nine traits relevant

to trophic and thermal niche, yield, demography,

habitat use, commercial importance and adaptation

potential (Table 2). Trait data were compiled from

locally available Ontario freshwater fishes life history

database (http://www.ontariofishes.ca/) and global

fish database (http://www.fishbase.ca/). For categori-

cal traits, we used the verbatim categories found in the

database.

Functional diversity of all traits and community-

weighted mean value or dominant states of individual

traits were computed using function ‘‘dbFD’’, which

implements a flexible distance-based framework to

compute multidimensional functional diversity

indices, in the R library ‘‘FD’’ (Laliberte & Legendre,

2010); we used Gower’s distance. Traditionally,

correlated traits are avoided in computing functional

diversity. Among the selected nine traits in our study,

only species length and weight traits were correlated to

each other. However, we kept both length and weight

in our analyses because species length and weight are

important proxies for body size and biomass, respec-

tively. Moreover, we were mainly interested in

understanding changes in the values of individual

traits. As implemented in the FD package, for numeric

traits, lake-wise community-weighted mean repre-

sents the average trait value across all species in a lake,

and for categorical traits, community-weighted mean

represents the most frequently (i.e. mode) represented

trait state in a lake.

Data analyses

We used permutational multivariate analysis of vari-

ance test using distance matrices to test whether fish

species composition differs between present and

future climate change scenarios (Anderson, 2001).

We used a Bray-Curtis dissimilarity matrix to sum-

marize species composition and used 999 permuta-

tions to determine statistical significance. We also

used non-metric multidimensional scaling to visualize

the compositional trends for different time periods and

climate change scenarios.

We then tested whether species richness, functional

diversity and the mean values of community-weighted

quantitative traits (functional composition; i.e.

Table 2 List of species life history traits used to quantify functional diversity and community-weighted trait means in inland lake

fish communities of Ontario

Trait Ecological significance Coding Trait states/units

Trophic breadth Trophic niche Numeric Number of prey phyla consumed from diet studies

Environment Habitat use Category Benthic, Benthopelagic and Pelagic

Thermal guild Thermal niche Category Cold-water, Cool-water and Warm-water speciesa

Spawning season Demography Category Fall, Spring, Summer, Spring-summer, Winter

Average fecundity Demography Numeric The average number of mature eggs produced

by a female fish per year

Maximum length Body size/yield Numeric Cm

Maximum weight Body size/yield Numeric Kg

Economic importance Ecological services Category Bait; Commercial; Panb; Sportc; Commercial and sport;

Forage and bait; Forage, bait, commercial and sport;

Forage and commercial; Forage, commercial and sport

Disturbance tolerance Extinction risk Ordinal Intolerant (0), Moderate (1), Tolerant (2)

a Coldwater fishes: optimal thermal range between 8 and 12�C; cool-water fishes: optimal thermal range between 16 and 20�C;warm-water fishes: optimal thermal range between 22 and 26�C (Christie & Regier, 1988; Magnuson et al. 1990)b Pan: ‘‘Small fish species easily captured by angling, often found in large numbers, which is harvested for fun or food. Harvest limits

are generous or unlimited’’.c Sport: ‘‘Species that are harvested for personal use, recreation or challenge. Harvest of these species is usually regulated’’.

Hydrobiologia

123

maximum body length, maximum body weight, max-

imum fecundity, species’ disturbance tolerance and

trophic niche) differ between present and two fore-

casted communities, using generalized least square

regression with an induced restrictive correlation

structure (Zuur et al., 2009). The restrictive correlation

structure (i.e. compound symmetric structure)

accounts for the correlation between observations

within the same lakes for different time periods, i.e.

present versus under the scenarios of climate change

(Zuur et al., 2009). The generic model was imple-

mented in R as, gls (response * predictor, metho-

d = ’’REML’’, correlation = corCompSymm(-

form = *1|lake_id), data) using the package

‘‘nlme’’. Chi square tests were used to test if the

community-weighted categorical traits (i.e. thermal

guild, habitat use and economic importance) differ

between present and two forecasted communities

under climate scenarios.

Results

Species composition, richness and functional

diversity

Permutational multivariate analyses of variance indi-

cates that the composition of projected fish commu-

nities of Ontario under climate change scenarios is

different from present (P = 0.01, R2 = 0.29). The

difference is also evident in ordination space (Elec-

tronic Supplementary Materials, Fig. S3), in terms of

separation of lakes between present versus projected

communities (P\ 0.01; R2 = 0.31). The present

species richness of 10.75 ± 3.68 per lake is projected

to increase (P\ 0.01) on average by 60% (6.4 new

species per lake) based on the best-case climate

change scenario and 81% (8.7 new species per lake)

based on the business-as-usual climate change sce-

nario (Fig. 2a; Electronic Supplementary Materials,

Table S1). Under the most conservative scenario of

climate change, 80% of lakes are projected to have

increased species richness, whereas in a business-as-

usual scenario, 95% of lakes are projected to have

increased species richness (Fig. 2b–c). By contrast,

functional diversity is projected to decline (P\ 0.01)

between present and climate change scenarios

(Fig. 2d; Electronic Supplementary Materials,

Table S1).

Functional composition

Community-weighted mean trait values differed

between present and projected communities for traits

related to thermal guild, economic importance, max-

imum length, maximum weight, average fecundity,

trophic niche and disturbance tolerance (Fig. 3; Elec-

tronic Supplementary Materials, Tables S2–4). We did

not detect significant responses from traits related to

spawning season and habitat use (Electronic Supple-

mentary Materials, Fig. S4).

Presently, with respect to the thermal guild func-

tional trait, the proportion of lakes dominated bywarm-

water, cold-water and cool-water species are 4.8, 13.18

and 82%, respectively. According to our projection,

lakes dominated by the trait for warm-water species is

likely to increase by 10–40%,whereas lakes dominated

by the trait for cool-water species is likely to decrease

by up to 27% under scenarios of climate change. Lakes

dominated by the trait for cold-water fish is projected to

decrease by 80%, and will comprise only 0.2–1.8% of

total lakes (Fig. 3a; Electronic Supplementary Mate-

rials, Table S3). The trait for ‘‘commercial and sport’’

fish, including walleye, whitefish, yellow perch,

northern pike and lake trout, is projected to decline

from being found in 27% of lakes to 0.5% under

climate change scenarios (Fig. 3e; Electronic Supple-

mentaryMaterials, Table S4). ‘‘Forage and bait’’ fishes

are projected to predominate the fish community

representing approximately 93% of species composi-

tion based on species occurrence (Fig. 3e).

Presently, the means (±1 SD) of community-

weighted maximum body length, maximum body

weight and average fecundity are 68.29 ± 13.22 cm,

9.09 ± 3.24 kg and 100344.92 ± 45760.57 eggs/per

year. The projected communities under climate change

scenarios are estimated to have, on average, 21–24%

shorter community-weighted mean body length (P\0.01), 38–39% lighter mean body weight (P\ 0.01)

and 36–37% less mean fecundity (P\ 0.01) (Fig. 3b–

d; Electronic SupplementaryMaterials, Table S2). Such

reduction in community-weighted mean body size

suggests reduced yield under climate change scenarios.

The projected communities under the best-case climate

change scenario are estimated to have on average lower

mean trophic breadth (i.e. the average number of phyla

consumed by a species) than the present communities,

indicating the potential decline of predatory species in

projected communities (Fig. 3e; Electronic

Hydrobiologia

123

Supplementary Materials, Table S2). Disturbance tol-

erance in future communities is projected to increase

only under the business as usual climate change scenario

(RCP8.5), but themean trait value for all three scenarios

remains \1.0, suggesting that communities will still

have ‘‘moderate tolerance’’ in both present and fore-

casted scenarios (Fig. 3f).

Discussion

This study reinforces the idea that climate change is

likely to initiate a compositional shift that may favour

warm-water species at the expense of presently

abundant cool- and cold-water species (Chu et al.,

2005; Sharma et al., 2007; Van Zuiden et al., 2016;

Fig. 2 Species richness and functional diversity for present

versus projected fish communities of year-2070. a lake species

richness, b–c distributions of change in lake species richness fora given RCP, d lake functional diversity, e–f distributions of

changes in lake functional diversity for a given RCP. The values

of change in species richness or functional diversity (b–c, e–f) are derived by subtracting lake-specific present species

richness or functional diversity from future species richness or

functional diversity. In the boxplots (a, d), each box contains

middle half of the raw data for a given variable and for a given

climate scenario, horizontal line within each box represents the

median value, and whiskers, as a measure of spread, represent

the inter quartile ranges. Box marked with same letter did not

differ significantly (a = 0.05) among each other, as identified

by Tukey’s post hoc test

Hydrobiologia

123

Fig. 3 Community-

weighted trait states or

values for lake fish

communities. a mosaic plot

showing proportion of lakes

dominated by cold-water

(species that prefer\19�Cduring summer months),

cool-water (19–25�C) andwarm-water ([25�C) fish.b–e community-weighted

maximum length, weight,

fecundity and trophic

breadth. f economic

importance. g disturbance

tolerance. Boxes within the

same plot (b–g) marked with

same letter did not differ

significantly at a = 0.05,

identified by Tukey’s post

hoc test

Hydrobiologia

123

Van Zuiden & Sharma, 2016). This forecasted shift in

community composition is likely to result in an

important shift in community functional structure, as

supported by the changes in community-weighted

means for traits depicting thermal guild, maximum

length, maximum weight, average fecundity, trophic

breadth and economic importance (Fig. 3). Because

these traits are closely related to fisheries productivity

and profitability (Sumaila et al., 2011), the projected

shifts in functional composition can be useful in

developing strategies to buffer fisheries against the

impacts of climate change, and to ensure the continued

provision of aquatic ecosystem services into the

future. However, although our study offers the much

needed initial insights into the compositional shift and

potential changes in community trait structure, our

results should be interpreted with caution, given firstly

that we did not consider the potential negative effects

of inter-specific interactions (Alofs & Jackson,

2015a, b) or potential dispersal constraints (Melles

et al., 2015) in generating future communities.

Secondly, our study attributes climate change-associ-

ated changes in ecosystems services based on changes

in species occurrence data rather than changes in

abundances of the different species within the com-

munities. The latter is a more likely outcome of

climate change impacts given that refugia and fisheries

regulations could also limit the impacts of climate

change on inland lakes. Further studies incorporating

processes, such as inter-specific interactions, dispersal

and species abundance information, in climate change

community composition models would be

worthwhile.

Species richness and functional diversity

We project a 60–81% increase in species richness in

future freshwater fish communities as warm-water

habitat availability increases under warming scenarios

(Hawkins et al., 2003; Menendez et al., 2006).

Increases in species richness in warming aquatic

habitats (Sagarin et al., 1999) is in part attributable to

the gradual nature of the change in thermal habitat

(Daufresne & Boet, 2007). While coldwater fishes are

likely be lost under the most extreme predicted future

changes in temperature (Fig. 3a), gradual warming is

likely to ensure a mix of cool- and warm-water species

until such time as a future equilibrium state might

allow for competitive exclusion of fishes adapted to

cooler habitats. In fact, multiple generations will likely

be necessary before species richness might be

expected to decline (Wilson, 1990). Although larger

lakes in Ontario typically have greater species richness

than small lakes, the projected increase in species

richness under climate change scenarios was consis-

tent across all sizes of lakes (Fig. S5).

But projected increases in lake-wise species rich-

ness are not met with concomitant increases in

functional diversity. This discrepancy occurred

because several traits (2 out of 9) did not change

under climate change scenarios, and the remaining

traits shifted to smaller, lighter, smaller trophic

breadth (i.e. prey species) and less fecund species,

with some traits (describing cold-water, and large-

bodied, economically important fishes) being lost

entirely in forecasted communities (Mouillot et al.,

2014).

Species and functional composition

Present configuration of community composition is

forecasted to change to favour warm-water, smaller-

bodied fishes directly in response to increasing lake

water temperatures (Magnuson et al., 1990; Jaeger

et al., 2014; Melles et al., 2015; O’reilly et al., 2015).

Under warming scenarios, thermal habitat availability

for warm-water fishes is usually projected to increase,

which should result in an increase of warm-water

fishes at the expense of those adapted to cold waters

(Hondzo & Stefan, 1991; Chu et al., 2005; Van Zuiden

& Sharma, 2016). Accordingly, we found that the

proportion of lakes in which warm-water species will

be found may increase by 10% and 40% under best-

case and business as usual scenarios, respectively. The

proportion of lakes in which cool-water species are

found is likely to decrease by up to 27%, and cold-

water species will be lost entirely in the business as

usual warming scenario.

These climate-mediated projected shifts in com-

munity composition are expected to result in a

northward range expansion of both cool-water and

warm-water fishes (Chu et al., 2005; Alofs et al., 2014;

Van Zuiden et al., 2016), which may have strong

impacts on community-wide species interactions. For

example, the invasion of a warm-water fish like

smallmouth bass is projected to result in the loss of

25,000 forage fish populations in Ontario by 2050

(Jackson & Mandrak, 2002), the loss of up to 20,000

Hydrobiologia

123

lake trout populations across Canada by 2100 (Sharma

et al., 2009), and a three-fold reduction in walleye

abundance across Ontario lakes by 2070 (Van Zuiden

& Sharma, 2016) through predator–prey and compet-

itive interactions.

Our models suggest that the present configuration

of several traits describing important metrics of

fisheries yield may change as lakes warm in this

region. Specifically, we project that, on average,

future inland freshwater fish communities in Ontario’s

lakes will be 21–24% shorter, 38–39% lighter and

42–45% less fecund in 2070. These results are

consistent with a meta-analysis suggesting that cli-

mate change benefits smaller-bodied fishes (Daufresne

et al., 2009). Daufresne et al. (2009) presented three

ecological theories at community, population and

individual scales to infer mechanisms that explain why

smaller-bodied fishes are favoured in warming cli-

mates. For example, Bergmann’s rule operates at the

community scale, and suggests that warmer regions

favour smaller-bodied species (Bergmann, 1847). At

the population scale, James’ rule posits that smaller-

bodied individuals will be favoured within populations

found in warmer regions (James, 1970). Finally, at the

individual scale, the temperature-size rule suggests

that body size will decrease with increasing temper-

ature (Atkinson, 1994). Each of these ecological

theories suggest that body size distributions should

decrease with warming climates. However, though

Ontario fishes are typically known to attain greater

sizes in the north than south, we did not notice any

latitudinal bias regarding community-weighted body

size distributions under climate change scenarios

(Fig. S6).

We project that ‘‘business-as-usual’’ and ‘‘best-

case’’ climate change scenarioswill modify the present

configuration of cold-water commercial and sport

fisheries in Ontario’s inland lakes. Affected species

will include lake trout, northern pike, walleye, white-

fish and yellow perch, which are presently found in

27% of lakes. We forecast that the trait describing the

distribution of these species (commercial and sport

fish) will decline at the landscape scale, and will only

be found in 0.5% of lakes under projected climate

change. In contrast, the trait describing ‘‘forage and

bait’’ species will ultimately represent approximately

93% of species composition based on occurrence.

However, one must take into account that some of the

lakes that support commercial fisheries in Ontario, e.g.

Lake Nipigon, Rainy Lake and Lake of the Woods are

not sampled as part of the Broad-scale Monitoring

program (and the data used in this study), and that some

of the deep lakes that support many commercial

fisheries, e.g. Lake Nipigon may still offer coldwater

refugiawith climate change.Nevertheless, the increase

in dominance of forage and bait species combined with

community-wide shrinking body size may impact fish

yield and the provision of regional ecosystem services

under climate change scenarios (Sheridan & Bickford,

2011; Koenigstein et al., 2016). Alternatively, the

increasing dominance of small-bodied pan fish under

climate change scenarios may actually create alternate

fishing opportunities in the future.

One vital ecosystem service provided by fisheries is

food provision, which will be affected with this

forecasted reduction in species body size (Sumaila

et al., 2011). In 2014, approximately 11,684 tonnes of

freshwater fish was commercially harvested from

inland lakes in Ontario, and this catch was worth

*$33,434,000 CDN (http://www.dfo-mpo.gc.

ca). Reductions in the species body size and fecun-

dity may have important implications for fish yield,

and this could be a costly consequence of climate

change. In addition, sport or recreation fisheries also

represent important ecosystem services that will be

disrupted in Ontario under future scenarios of climate

change, and people who rely on the present configu-

ration of the cold-water commercial and sport fishery

are likely to experience consequential socio-economic

consequences of a warming climate (Nelson et al.,

2013). We however suggest that integrating consid-

erations of species diversity and functional composi-

tion are useful means of assessing how aquatic

communities are expected to change with a changing

climate, and can be helpful in crafting informed con-

servation and management strategies under circum-

stances of global environmental change (Villeger

et al., 2010; Cadotte et al., 2011).

Acknowledgements We thank Ontario Ministry of Natural

Resources for the fish data; Thomas Van Zuiden and Miranda

Chen for the climate data; Saiful Khan for the map of the study

area; and John Magnuson, Begona Santos and two anonymous

reviewers for valuable comments on an earlier version of this

manuscript. Funding for this research was provided by Natural

Sciences and Engineering Research Council Canada Discovery

Grant to SS and York University.

Hydrobiologia

123

References

Alofs, K. M. & D. A. Jackson, 2015a. The abiotic and biotic

factors limiting establishment of predatory fishes at their

expanding northern range boundaries in Ontario, Canada.

Global Change Biology 21: 2227–2237.

Alofs, K. M. & D. A. Jackson, 2015b. The vulnerability of

species to range expansions by predators can be predicted

using historical species associations and body size. Pro-

ceedings of the Royal Society B: Biological Sciences 282:

20151211.

Alofs, K. M., D. A. Jackson & N. P. Lester, 2014. Ontario

freshwater fishes demonstrate differing range-boundary

shifts in a warming climate. Diversity and Distributions 20:

123–136.

Anderson, M. J., 2001. A new method for non-parametric

multivariate analysis of variance. Austral Ecology 26:

32–46.

Araujo, M. B. & A. T. Peterson, 2012. Uses and misuses of

bioclimatic envelope modeling. Ecology 93: 1527–1539.

Atkinson, D., 1994. Temperature and organism size – a bio-

logical law for ectotherms? Advances in Ecological

Research 3: 1–58.

Barbet-Massin, M. &W. Jetz, 2015. The effect of range changes

on the functional turnover, structure and diversity of bird

assemblages under future climate scenarios. Global

Change Biology 21: 2917–2928.

Bergmann, C., 1847. Uber die verh altnisse der warme¨ oko-

nomie der thiere zuihrer grosse. Gott. Stud. 1: 595–708.

Blanchet, F. G., P. Legendre & D. Borcard, 2008. Forward

selection of explanatory variables. Ecology 89:

2623–2632.

Botkin, D. B., H. Saxe, M. B. Araujo, R. Betts, R. H. W.

Bradshaw, T. Cedhagen, P. Chesson, T. P. Dawson, J.

R. Etterson, D. P. Faith, S. Ferrier, A. Guisan, A. S. Hansen,

D. W. Hilbert, C. Loehle, C. Margules, M. New, M. J. So-

bel & D. R. B. Stockwell, 2007. Forecasting the effects of

global warming on biodiversity. BioScience 57: 227–236.

Cadotte, M. W., K. Carscadden & N. Mirotchnick, 2011.

Beyond species: functional diversity and the maintenance

of ecological processes and services. Journal of Applied

Ecology 48: 1079–1087.

Cardinale, B. J., J. E. Duffy, A. Gonzalez, D. U. Hooper, C.

Perrings, P. Venail, A. Narwani, G. M. Mace, D. Tilman,

D. A. Wardle, A. P. Kinzig, G. C. Daily, M. Loreau, J.

B. Grace, A. Larigauderie, D. S. Srivastava & S. Naeem,

2012. Biodiversity loss and its impact on humanity. Nature

486: 59–67.

Chen, I.-C., J. K. Hill, R. Ohlemuller, D. B. Roy & C.

D. Thomas, 2011. Rapid range shifts of species associated

with high levels of climate warming. Science 333:

1024–1026.

Christie, G. C. & H. A. Regier, 1988. Measures of optimal

thermal habitat and their relationship to yields for four

commercial fish species. Canadian Journal of Fisheries and

Aquatic Sciences 45: 301–314.

Chu, C., N. E. Mandrak & C. K. Minns, 2005. Potential impacts

of climate change on the distributions of several common

and rare freshwater fishes in Canada. Diversity and

Distributions 11: 299–310.

Civantos, E., W. Thuiller, L. Maiorano, A. Guisan & M.

B. Araujo, 2012. Potential impacts of climate change on

ecosystem services in Europe: the case of pest control by

vertebrates. BioScience 62: 658–666.

Daufresne, M. & P. Boet, 2007. Climate change impacts on

structure and diversity of fish communities in rivers. Global

Change Biology 13: 2467–2478.

Daufresne, M., K. Lengfellner & U. Sommer, 2009. Global

warming benefits the small in aquatic ecosystems. Pro-

ceedings of the National academy of Sciences of the United

States of America 106: 12788–12793.

Dıaz, S., S. Lavorel, F. De Bello, F. Quetier, K. Grigulis & T.

M. Robson, 2007. Incorporating plant functional diversity

effects in ecosystem service assessments. Proceedings of

the National Academy of Sciences 104: 20684–20689.

Dubuis, A., J. Pottier, V. Rion, L. Pellissier, J.-P. Theurillat &A.

Guisan, 2011. Predicting spatial patterns of plant species

richness: a comparison of direct macroecological and

species stacking modelling approaches. Diversity and

Distributions 17: 1122–1131.

Gauzere, P., F. Jiguet & V. Devictor, 2015. Rapid adjustment of

bird community compositions to local climatic variations

and its functional consequences. Global Change Biology

21: 3367–3378.

Guisan, A. & C. Rahbek, 2011. SESAM – a new framework

integrating macroecological and species distribution

models for predicting spatio-temporal patterns of species

assemblages. Journal of Biogeography 38: 1433–1444.

Guisan, A. &W. Thuiller, 2005. Predicting species distribution:

offering more than simple habitat models. Ecology Letters

8: 993–1009.

Guisan, A. & N. E. Zimmermann, 2000. Predictive habitat

distribution models in ecology. Ecological Modelling 135:

147–186.

Hawkins, B. A., R. Field, H. V. Cornell, D. J. Currie, J.-F.

Guegan, D. M. Kaufman, J. T. Kerr, G. G. Mittelbach, T.

Oberdorff, E. M. O’brien, E. E. Porter & J. R. G. Turner,

2003. Energy, water, and broad-scale geographic patetrns

in species richness. Ecology 84: 3105–3117.

Hijmans, R. J., S. E. Cameron, J. L. Parra, P. G. Jones & A.

Jarvis, 2005. Very high resolution interpolated climate

surfaces for global land areas. International Journal of

Climatology 25: 1965–1978.

Hondzo, M. & H. G. Stefan, 1991. Three case studies of lake

temperature and stratification response to warmer climate.

Water Resources Research 27: 1837–1846.

Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti,

S. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S.

Naeem, B. Schmid, H. Setala, A. J. Symstad, J. Vander-

meer & D. A. Wardle, 2005. Effects of biodiversity on

ecosystem functioning: a consensus of current knowledge.

Ecological Monographs 75: 3–35.

Hooper, D. U., E. C. Adair, B. J. Cardinale, J. E. K. Byrnes, B.

A. Hungate, K. L. Matulich, A. Gonzalez, J. E. Duffy, L.

Gamfeldt & M. I. O’Connor, 2012. A global synthesis

reveals biodiversity loss as a major driver of ecosystem

change. Nature 486: 105–108.

IPCC, 2013. Climate Change 2013: The Physical Science Basis.

Contribution of Working Group I to the Fifth Assessment

Report of the Intergovernmental Panel on Climate Change,

Hydrobiologia

123

Cambridge University Press, Cambridge, United Kingdom

and New York, NY, USA.

Jackson, D. A. & H. H. Harvey, 1989. Biogeographic associa-

tions in fish assemblages: local vs regional processes.

Ecology 70: 1472–1484.

Jackson, D. A. & N. E. Mandrak (eds), 2002. Changing Fish

Biodiversity: Predicting the Loss of Cyprinid Biodiversity

Due to Global Climate Change. American Fisheries Soci-

ety, Bethesda, MD.

Jaeger, K. L., J. D. Olden&N. A. Pelland, 2014. Climate change

poised to threaten hydrologic connectivity and endemic

fishes in dryland streams. Proceedings of the National

Academy of Sciences 111: 13894–13899.

James, F. C., 1970. Geographic size variation in birds and its

relationship to climate. Ecology 51: 365–390.

Koenigstein, S., F. C. Mark, S. Goßling-Reisemann, H. Reuter

& H.-O. Poertner, 2016. Modelling climate change impacts

on marine fish populations: process-based integration of

ocean warming, acidification and other environmental

drivers. Fish and Fisheries 17: 972–1004.

Laliberte, E. & P. Legendre, 2010. A distance-based framework

for measuring functional diversity from multiple traits.

Ecology 91: 299–305.

Magnuson, J. J., J. D. Meisner & D. K. Hill, 1990. Potential

changes in the thermal habitat of great lakes fish after

global climate warming. Transactions of the American

Fisheries Society 119: 254–264.

Magnuson, J. J., T. K. Kratz, T. F. Allen, D. E. Armstrong, B.

J. Benson, C. J. Bowser, D. W. Bolgrien, S. R. Carpenter,

T. M. Frost, S. T. Gower, T. M. Lillesand, J. A. Pike & M.

G. Turner, 1997. Regionalization of long-term ecological

research (LTER) on north temperate lakes. International

Association of Theoretical and Applied Limnology 26(Pt 2

26): 522–528.

Melles, S. J., C. Chu, K. M. Alofs & D. A. Jackson, 2015.

Potential spread of Great Lakes fishes given climate change

and proposed dams: an approach using circuit theory to

evaluate invasion risk. Landscape Ecology 30: 919–935.

Menendez, R., A. G. Megıas, J. K. Hill, B. Braschler, S.

G. Willis, Y. Collingham, R. Fox, D. B. Roy & C.

D. Thomas, 2006. Species richness changes lag behind

climate change. Proceedings of the Royal Society of

London B: Biological Sciences 273: 1465–1470.

Millennium-Ecosystem-Assessment, 2005. Ecosystems and

Human Well-being: Policy Responses. Island Press,

Washington, DC.

Mokany, K., J. J. Thomson, A. J. J. Lynch, G. J. Jordan & S.

Ferrier, 2015. Linking changes in community composition

and function under climate change. Ecological Applica-

tions 25: 2132–2141.

Mokany, K., S. Ferrier, S. R. Connolly, P. K. Dunstan, E.

A. Fulton, M. B. Harfoot, T. D. Harwood, A. J. Richardson,

S. H. Roxburgh, J. P. W. Scharlemann, D. P. Tittensor, D.

A.Westcott & B. A.Wintle, 2016. Integrating modelling of

biodiversity composition and ecosystem function. Oikos

125: 10–19.

Mouillot, D., S. Villeger, V. Parravicini, M. Kulbicki, J.

E. Arias-Gonzalez, M. Bender, P. Chabanet, S. R. Floeter,

A. Friedlander, L. Vigliola & D. R. Bellwood, 2014.

Functional over-redundancy and high functional

vulnerability in global fish faunas on tropical reefs. Pro-

ceedings of the National Academy of Sciences 111:

13757–13762.

Nelson, E. J., P. Kareiva, M. Ruckelshaus, K. Arkema, G.

Geller, E. Girvetz, D. Goodrich, V. Matzek, M. Pinsky, W.

Reid, M. Saunders, D. Semmens&H. Tallis, 2013. Climate

change’s impact on key ecosystem services and the human

well-being they support in the US. Frontiers in Ecology and

the Environment 11: 483–493.

Nock, C. A., R. J. Vogt & B. E. Beisner, 2016. Functional Traits.

In eLS. Wiley.

Oliver, T.H.,M. S.Heard, N. J. B. Isaac, D.B. Roy,D. Procter, F.

Eigenbrod, R. Freckleton, A. Hector, C. D. L. Orme, O.

L. Petchey, V. Proenca, D. Raffaelli, K. B. Suttle, G.

M. Mace, B. Martın-Lopez, B. A. Woodcock & J. M. Bul-

lock, 2015a. Biodiversity and resilience of ecosystem

functions. Trends in Ecology & Evolution 30: 673–684.

Oliver, T. H., N. J. B. Isaac, T. A. August, B. A. Woodcock, D.

B. Roy & J. M. Bullock, 2015b. Declining resilience of

ecosystem functions under biodiversity loss. Nat Com-

munications 6: 10122.

O’reilly, C. M., S. Sharma, D. K. Gray, S.E. Hampton, J.

S. Read, R. J. Rowley, P. Schneider, J. D. Lenters, P.

B. Mcintyre, B. M. Kraemer, G. A. Weyhenmeyer, D.

Straile, B. Dong, R. Adrian, M. G. Allan, O. Anneville, L.

Arvola, J. Austin, J. L. Bailey, J. S. Baron, J. D. Brookes, E.

De Eyto, M.T. Dokulil, D. P. Hamilton, K. Havens, A.

L. Hetherington, S. N. Higgins, S. Hook, L. R. Izmest’eva,

K. D. Joehnk, K. Kangur, P. Kasprzak, M. Kumagai, E.

Kuusisto, G. Leshkevich, D. M. Livingstone, S. Macintyre,

L. May, J. M. Melack, D. C. Mueller-Navarra, M. Nau-

menko, P. Noges, T. Noges, R. P. North, P.-D. Plisnier, A.

Rigosi, A. Rimmer, M. Rogora, L. G. Rudstam, J.

A. Rusak, N. Salmaso, N. R. Samal, D.E. Schindler, S.

G. Schladow, M. Schmid, S. R. Schmidt, E. Silow, M.

E. Soylu, K. Teubner, P. Verburg, A. Voutilainen, A.

Watkinson, C. E.Williamson&G. Zhang, 2015. Rapid and

highly variable warming of lake surface waters around the

globe. Geophysical Research Letters 42: 10773–10781.

Parmesan, C. & G. Yohe, 2003. A globally coherent fingerprint

of climate change impacts across natural systems. Nature

421: 37–42.

Sagarin, R. D., J. P. Barry, S. E. Gilman & C. H. Baxter, 1999.

Climate-related change in an intertidal community over

short and long time scales. Ecological Monographs 69:

465–490.

Sandstrom, S., M. Rawson & N. Lester, 2010. Manual of

Instructions for Broad-scale Fish Community Monitoring;

using LargeMesh Gillnets and Small Mesh Gillnets. 34. In:

O.M.O.N. Resources (ed.), Peterborough, ON.

Sharma, S. & D. A. Jackson, 2008. Predicting smallmouth bass

(Micropterus dolomieu) occurrence across North America

under climate change: a comparison of statistical approa-

ches. Canadian Journal of Fisheries and Aquatic Sciences

65: 471–481.

Sharma, S., D. A. Jackson, C. K. Minns & B. J. Shuter, 2007.Will northern fish populations be in hot water because of

climate change? Global Change Biology 13: 2052–2064.

Sharma, S., D. A. Jackson & C. K. Minns, 2009. Quantifying the

potential effects of climate change and the invasion of

Hydrobiologia

123

smallmouth bass on native lake trout populations across

Canadian lakes. Ecography 32: 517–525.

Sharma, S., M. J. Vander Zanden, J. J. Magnuson & J. Lyons,

2011. Comparing Climate Change and Species Invasions

as Drivers of Coldwater Fish Population Extirpations. Plos

One 6: e22906.

Sheridan, J. A. & D. Bickford, 2011. Shrinking body size as an

ecological response to climate change. Nature Climate

Change 1: 401–406.

Shuter, B. J., J. A. Maclean, F. E. J. Fry & H. A. Regier, 1980.

Stochastic simulation of temperature effects on first-year

survival of smallmouth bass. Transactions of the American

Fisheries Society 109: 1–34.

Simpson, S. D., S. Jennings, M. P. Johnson, J. L. Blanchard, P.-J.

Schon, D. W. Sims & M. J. Genner, 2011. Continental

shelf-wide response of a fish assemblage to rapid warming

of the sea. Current Biology 21: 1565–1570.

Sumaila, U. R., W. W. L. Cheung, V. W. Y. Lam, D. Pauly & S.

Herrick, 2011. Climate change impacts on the biophysics

and economics of world fisheries. Nature Climate Change

1: 449–456.

Thuiller, W., S. Pironon, A. Psomas, M. Barbet-Massin, F.

Jiguet, S. Lavergne, P. B. Pearman, J. Renaud, L. Zupan &

N. E. Zimmermann, 2014. The European functional tree of

bird life in the face of global change. Nature Communi-

cations 5: 3118.

Tonn, W. M. & J. J. Magnuson, 1982. Patterns in the species

composition and richness of fish assemblages in Northern

Wisconsin lakes. Ecology 63: 1149–1166.

Van Zuiden, T. M., M. M. Chen, S. Stefanoff, L. Lopez & S.

Sharma, 2016. Projected impacts of climate change on

three freshwater fishes and potential novel competitive

interactions. Diversity and Distributions 22: 603–614.

Van Zuiden, T. M. & S. Sharma, 2016. Examining the effects of

climate change and species invasions on Ontario walleye

populations: can walleye beat the heat? Diversity and

Distributions 22: 1069–1079.

Villeger, S., J. R. Miranda, D. F. Hernandez & D. Mouillot,

2010. Contrasting changes in taxonomic vs. functional

diversity of tropical fish communities after habitat degra-

dation. Ecological Applications 20: 1512–1522.

Violle, C., M.-L. Navas, D. Vile, E. Kazakou, C. Fortunel, I.

Hummel & E. Garnier, 2007. Let the concept of trait be

functional! Oikos 116: 882–892.

Walther, G. R., 2010. Community and ecosystem responses to

recent climate change. Philosophical Transactions of the

Royal Society B-Biological Sciences 365: 2019–2024.

Wilson, J. B., 1990. Mechanisms of species coexistence: twe-

leve explanations for Hutchinson’s paradox of the plank-

ton: evidence from New Zealand plant communities. New

Zealand Journal of Ecology 13: 17–42.

Zurell, D., N. E. Zimmermann, T. Sattler, M. P. Nobis & B.

Schroder, 2016. Effects of functional traits on the predic-

tion accuracy of species richness models. Diversity and

Distributions 22: 905–917.

Zuur, A. F., E. N. Ieno, N. J. Walker, A. A. Saveliev & G.

M. Smith, 2009. Mixed Effects Models and Extensions in

Ecology with R. Springer.

Hydrobiologia

123