Localextinctionandcolonisationinnativeandexoticfish inrelationtochangesinlanduse

Dorothe´eKoppA,JordiFiguerolaB,ArthurCompinA,Fre´de´ricSantoulA

andRe´gisCe´re´ghinoA,C

AEcoLab,LaboratoireEcologie FonctionnelleetEnvironnement,UMR5245,Universite´de

Toulouse, 118routedeNarbonne,31062ToulouseCedex9,France.

BEstacionBiologicadeDon˜ana,AvenuenidadeMariaLuisas/n,PabellondelPeru,

41013Sevilla,Spain.

CCorrespondingauthor.Email:

Abstract. Distributionpatternsofmanynativeandexoticfishspeciesarewelldocumented,yetlittleisknownaboutthe temporaldynamicsofnativeandexoticdiversityinrelationtochangesinlanduse.Wehypothesisedthatcolonisationrates wouldbehigherforexoticfishspeciesandthatextinctionrateswouldbehigherfornativespeciesinlargestreamsystems. Wealsopredictedthatcold-water specieswouldbemoreimpactedthanthermallytolerantspecies.Totestthese hypotheses,weusedgeneralisedlinearmixedmodelstocomparechangesinnativeandexoticfishspeciesrichnessover

10yearsinaFrenchdrainagebasinsubjectedtolandscapealterations.Exoticfishweremoresusceptibletolocalextinction thanthenativeones.Extinctionwasgreateramongcold-tolerantspeciesandathigherelevations.Colonisationbyexotic specieswashigheratlowerelevations.Althoughadecadeofexpandingurbanisation affectedfishcolonisation, agriculturallandsexperiencedhigherextinctionrates.Inthecontextofglobalchangesinlanduseandpopulation pressure,ourstudysuggeststhatthetemporaldynamicsoffishdiversityaredrivenbylandscapealterationsaswellasby thethermaltoleranceofspecies.

Additionalkeywords:agriculture,freshwaterfishes,introductions,thermalpreferences,urbanisation.

Introduction

Worldwide, manycountrieshaveestablishednational Biodiversity ActionPlans(BAPs,e.g.USA,Australia,New Zealand,Tanzania).Thesepolicychangeswereimplementedin responsetoagrowingconcernthatdevelopmentoverprevious decadeswashavingasignificant impactuponbiodiversity, despiteconsiderable benefitstothecitizensofnations.Under thisframework, assessingtheeffectsoflandscapealterations (e.g.urbanisation,agriculture)oncommunitydynamicsisnowa priorityforbiologicalconservation(McKinney2002).Muchof thecurrentresearchhasreporteddeclinesinthediversityof targetorganisms(e.g.variousinvertebrates, fish,mammalsor plants)inrelationtolandscapealterations,fromlocaltoglobal scales.Apartfromtheirdirecteffectsupondiversity and ecologicalprocesses,changesinlanduseorhabitatdestruction canalso promotebiologicalinvasionsby exoticspecies.These biologicalinvasionsconstituteasignificant component of globalenvironmentalchangesandmayaffecthumanhealthand wealth,ecosystem functions andnativebiologicaldiversity (Vitouseketal.1996).Freshwatersareparticularlysubjectedto habitatdegradation,hydrologicalterations,pollutionandthe spreadofinvasivespecies(Dudgeonet al.2006).Forexample,

recentestimatesindicatethat25%oftheworld’sfreshwaterfish arethreatenedwithextinction(Vie´etal.2009).

Many species present a metapopulation structure, with colonisation/extinctionoccurringatlocalscales(Case1991).

Theequilibriumbetweenbothprocessesdetermines trendsin population ranges.However,researchhasfocussedoncharac- terisingcolonisationandwavesofestablishment(e.g.Leprieur etal.2008;Gozlanetal.2010)andlittleisknownaboutthe

temporalstabilityofbothexoticandnativepopulations in invadedareas(Strayeretal.2006).Inthiscontextofspecies lossandbiologicalinvasions,theresponseoflocalcommunities

toenvironmentalchanges canbeperceivedintermsofcoloni- sation bysomesetsofspecies andextinctionofotherspecies overtime.Localcolonisationisdefinedasaspeciesbeing

detectedinalocationfromwhichitwaspreviously absent, whereaslocalextinctionisdefinedastheabsenceofaspeciesin alocationwhereitwaspreviouslypresent(Nicholsetal.1998). Assumingthat exotic species coloniseecosystemsas agri-

culturalorurbanlandcoverincreases(Pooletal.2010),one maypredictthatexoticandnativespecies differintheir colonisation/extinction responses to environmental changes

(Koppetal.2009).

Spatialpatterns ofnativeandexoticspeciesrichnesshave been extensively studied along gradients of disturbance (e.g.Koppetal.2009).However, littleisknownaboutthe temporaldynamicsofnativeandnon-nativediversityinrelation tochanging landscapes.Althoughhabitatdegradation,climate changeorbiologicalinvasionscancauseimmediateextinctions, thereisoftenaconsiderablelagbetweenenvironmentalchanges andthesubsequentextinctions(Oldenetal.2010).Moststudies of thedistributionsof exoticspecieshavebeenbasedon short- termsurveysandhave,thus,yieldedsnapshots thatlacka temporalcontext(Strayer etal.2006).Inlightofrecenturban andagriculturaldevelopment,short-termassessmentsarelikely tobeinadequatetodescribecolonisation/extinctionpatternsand therolesthatexoticspeciesplayovertimeintheareasthey invade.Usinglong-termdataonspeciesoccurrenceorcompar- ingpastvspresenthabitatcharacteristicsmaybeusedtoaddress theseaims(Oldenetal.2010).

Inthisstudy,wefocussedontheAdour-Garonnestream system(France).Thedrainagebasinofthissystemissubjected

tourbanisationandextensiveagriculture.Riverfisharethemost

frequentlyintroducedfreshwater organismsworldwide (Leprieuretal.2008);therefore, theyarerelevantmodel organismstostudythecontributionofnativeandexoticspecies tobiodiversity patternsinaspatialandtemporalcontext.Our specificaimwastoassess theextentofchanges innativeand exoticfishspeciesrichnessoverthepastdecade.Assumingthat increasinganthropogenicimpactenhancesthechancesofsuc- cessfulinvasion byexoticfishesandthatexoticandnative speciesdifferintheirresponsestoenvironmental changes (watertemperature,urbanisation,agriculture),wehypothesised highercolonisationratesforexoticspeciesandhigherextinction ratesfornativeones.Inthecontext ofglobalwarming,wealso predictedthatcold-water specieswouldbemoreproneto extinctionthanthermallytolerantspecies.Totestthesehypoth- eses,weanalysedthepatternsofcolonisationandextinctionby nativeandexoticfishat100sitesfrom1994–97to2004–07,in relationtochanges inlanduse,thelocationofsiteswithinthe streamsystemandthethermalpreferencesofspecies(fromcold towarmwaters).

Materialsandmethods

Studyareaanddatacollection

TheAdour-Garonnestreamsystem(south-westernFrance)hasa

116000km2 drainagebasin.Weselected100samplingsites rangingfrom6to1800mabovesealevel(a.s.l.,seeFig.S1,

availableasSupplementaryMaterialtothispaper),representing

riversfromhighmountain,plainandcoastalareas.Samples werecollectedbyelectrofishing duringlow-flowperiods.To avoidpseudo-absencedata,eachsitewassampledthreetimes between1994and1997andthenthreetimesbetween2004and

2007. The sampling reaches (,100m long) were similar betweentheperiods.Aspecieswiththreeabsencesduringa givenperiodwasconsideredasabsent.Thebiologicalvariables assignedtoeachsitewerethecolonisationbynewfishspeciesor theextinctionofspecies from oneperiod totheother.Asfish temperaturepreferences haveastronginfluenceuponspecies’ potentialrangewithintherivercontinuum(Buissonetal.2008), fishweredividedintowarm-,cool-andcold-watertypeswith

regardtotheirpreferredthermalconditions(Magnusonetal.

1979).Wethenincludedthethermaltolerances offish:cold-, cool-orwarm-tolerant(accordingtoKottelatandFreyhof2007) asexplanatoryvariablesintheanalysis.

For each site, a geographic information system (GIS, MapinfoProfessional7.8,Troy,NY)wasusedtodelineatea

geographicalbufferzonerepresentinga1000-mradiuscentred

onthesite.Thissizefallswithin thatofthe‘reach buffer’ defined byAllan(2004)asabufferof100toseveral hundred metresinwidthoneachbankandsomehundredsofmetrestoa kilometreinlength).Inpractice, thewidthofthebufferzoneis oftenadaptedtothelandscape characteristicsofthestudied areasandrangesfrom30to100moneachsideoftheriver (discussedbyCompinandCe´re´ghino 2007).Sampling sites werethencharacterisedusingelevationabovesealevel(m)and threeland-cover variablesintended toaccountforanthropo- genicpressure.Thethreeland-covervariables,describedforthe twoperiods1994–97and2004–07,werepercentageareawithin abufferzonecoveredbyforest(areasoccupied byforestand woodlandswith nativeor exoticconiferousor deciduoustrees; scrubandherbaceousvegetationassociations), urbanareas (industrial, commercialandtransportunits;artificialandnon- agriculturalvegetatedareas) and agriculturalareas (arable lands,permanent cropsandpasture).Digitalland-cover information wasobtainedfromtheCORINEland-cover data- baseforEurope(EuropeanEnvironment Agency, eea.europa.eu/,accessed June2010).Thisdatabasewasgener- atedfromorthorectifiedsatelliteimagesand providesthematic GISmaplayersincludingupto44land-cover classeswitha mappingscale of1:100000.Thesefourvariableswerechosen because theycharacterisethelocationofsamplingsiteswithin thestreamsystemandwithintheregionallandscapemosaicand theyareeasytodescribeusingaGIS.

Dataanalyses

Weanalysedtherelationshipofdifferentecologicalfactorsand thepatternsofcolonisationandextinctionin100differentsites infourdifferentriverbasins.Wefittedageneralised linear mixedmodel(GLMM)tothepresence–absencedatacollected foreachspeciesineachdifferentsitein1994–97and2004–07. Bothbasinandsitewithinabasinwereincludedasrandom factorstocontrolforlocalnon-independence ofthedata.We modelledthenon-independenceofdatacomingfromthesame speciesbyassumingacommonpositivecorrelationbetween datafromthesamespeciescomingfromdifferentlocalitiesand azerocorrelationwiththedataofotherfishspecies (see Blackburn andDuncan2001forasimilarapproach).Wecon- structedtwodifferentmodels:oneforlocalcolonisations and anotherforlocalextinctions. Forthemodeloflocalcolonisa- tions,weconsidered onlythedataofnegativecensusforeach speciesxlocalityin1994–97andusedthepresence/absence datafor2004–07astheresponsevariablemodelling theprob- abilityofobtainingapositivecountin2004–07. Forthe extinctionsmodel,weusedthe dataofpositivecensusforeach speciesxlocalityin 1994–97andagain usedthepresence/ absencedataastheresponsevariable,modellingtheprobability of anegativecensusin2004–07.

AllcalculationsweredonewiththeGLIMMIXprocedurein

SAS9.2.(SASInstitute,Cary,NC,USA)usingabinomial

Table1. Modelsanalysingthepatternsof fishextinctionforcold-tolerantspecies(cold),cool-tolerantspecies(cool), thermally-tolerantspecies(warm) andfornativevs.exotic species

OnlyvariableswithP,0.05areinterpretedasstatisticallysignificant.Estimatescorrespondtoslopevaluesassociatedtoeach factorandareonlyreportedforsignificanteffectsretained inthefinalmodel

Extinction / F-value / d.f. / P-value / Estimate±s.e.
Elevation / 6.79 / 1,932 / 0.009 / 0.5448±0.2090
Cold / 4.08 / 1,32 / 0.05 / Non-cold,—0.7309±0.6320;cold,0
Cool / 0.07 / 1,31 / 0.79 / –
Warm / 0.13 / 1,31 / 0.72 / –
Nativevsexotic / 18.51 / 1,32 / 0.0001 / Exotics,1.2880±0.2994;native, 0
Urbanisation / 0.28 / 1,931 / 0.60 / –
Agriculture / 7.53 / 1,932 / 0.006 / 0.0103±0.0037
Changeurbanlandcover / 0.54 / 1,931 / 0.46 / –
Changeagriculturallandcover / 0.14 / 1,931 / 0.70 / –

Table2. Modelsanalysingthepatternsoffishcolonisation(forlegend seeTable1)

OnlyvariableswithP,0.05areinterpretedasstatisticallysignificant.Estimatescorrespondtoslopevaluesassociatedtoeach factorandareonlyreportedforsignificanteffectsretained inthefinalmodel

Colonisation / F-value / d.f. / P-value / Estimate±s.e.
Elevation / 76.95 / 1,2692 / ,0.0001 / —1.2457±0.1420
Cold / 10.07 / 1,34 / 0.003 / Non-cold,1.2534±0.3894;cold,0
Cool / 0.78 / 1,33 / 0.38 / –
Warm / 5.02 / 1,34 / 0.02 / Non-warm,0.6912±0.3085;warm,0
Nativevsexotic / 0.67 / 1,33 / 0.42 / –
Urbanisation / 0.69 / 1,2691 / 0.41 / –
Agriculture / 1.03 / 1,2691 / 0.31 / –
Changeurbanlandcover / 21.80 / 1,2692 / ,0.0001 / 0.0796±0.0170
Changeagriculturallandcover / 1.11 / 1,2692 / 0.29 / –

distributed errorandalogitlinkfunction. Wefollowed a backwardsselectionprocedure.Theleastsignificantvariable wasexcludedfromthemodelwhichwasthenrecalculateduntil onlyexplanatoryvariablesincreasingfitwithaP,0.05were retained.Environmentalvariableswerelog-transformedtofita normaldistribution(elevation)orrankedwhennotnormalised byusualtransformations(%urban,%agriculturalarea,changes incoversofbothlanduses).Aslandusewasclassifiedinthree exclusivecategories(i.e.%urban,agriculturalandforestareas), astrongcolinearityexistedbetweenthethreevariables.Forthis reason,onlytwoof thecategories(% urbanand% agricultural areas)wereincludedinfurtheranalyses,butstatisticallysignifi- cantrelationshipsofthe samesignforbothvariablesmustalso beinterpretedasanegativerelationship forthethirdone.The categoryexcludedfromtheanalyseswaschosenat randombut qualitativelyidenticalresultsemergedwhenexcludinganother ofthehabitatchangecategories. Changesinlandcoverwere calculatedasthedifferenceintheproportionofhabitatbetween

2004–07and1994–97.

Results

Thirty-seven fishspecieswerefoundintheAdour-Garonne basin,amongwhich16wereexotics(according toKeithand Allardi 2001; forcompletelistseeTableS1,availableasSup- plementaryMaterialtothispaper).Ourmodelsindicatethat elevationandland-usevariablescouldexplainlocalextinction andcolonisationpatternsinriverfishinthisbasin.

Fishextinction

Theextinctionmodelrevealedthatlocalextinctionwas signif- icantly higher among cold-tolerant species (F1,32¼4.08; P¼0.05;Table1),sothattheextinctionratewashigherat higherelevation(F1,932 ¼6.79;P¼0.009).Exoticspecieshada highertendencytolocalextinctioncompared withthenative ones(F1,32¼18.51;P¼0.0001;Table1).Theintensification ofagriculture(upto46%atsome ofthestudiedsites)fostered local extinction of fish species (F1,932 ¼7.53; P¼0.006; Table1).

Fishcolonisation

Thecolonisation modelrevealedthatcolonisation hasbeen greateratlowelevations(F1,2792¼76.95;P,0.0001;Table2). Colonisationappearedtobelowerforbothcold-andthermally- tolerantspecies,eithernativeorexotic(F1,34¼10.07;P¼0.003 andF1,34¼5.02;P¼0.02respectively).Finally,theexpansion ofurbanlandsbetween 1994–97and2004–07(upto17%) positively affected colonisation (F1,2792¼21.8; P,0.0001; Table2).

Discussion

IntheAdour-Garonne streamsystem,fishcolonisationwas higheratlowelevations.Accordingtothe‘bioticresistance hypothesis’,thehigherfishrichnessatdownstreamsites(Kopp etal. 2009)shouldpreventtheestablishmentofnon-native

speciesthroughcompetitiveexclusion(Leprieuretal.2008; Oldenetal.2010).However,becausecolonisationwashigherin theseareas,wecaninferthathigherspeciesrichness didnot limitthecapacityofexoticspeciestoestablish.Instead,the hypothesis of‘bioticacceptance’canbeapplicableinoursys- tem.Thishypothesispredictsthatthefactorsthatsupportgreater nativerichnessinaregion(e.g.abundantresourcesandhabitat heterogeneity) alsopromotetheestablishment ofnon-native species.Therefore,biologicalinteractions mayhaveplayeda negligibleroleintheobservedpatterns,contrarytoenviron- mentalfactors.

Cold-tolerant species,whichexperiencedthegreatest extinctionrates,aretypicallyconfinedtomountainstreamsin

theAdour-Garonnebasin.AccordingtoBalcombeetal.(2011),

thesecold-watertolerantspeciesareparticularlysensitiveto increasedstreamtemperature,especiallywhentheyarelocated attheirtemperaturelimits(Morrongielloetal.2011)because temperaturehasadirectphysiological effectonthesefish.As fisharepoikilotherms, watertemperaturestronglycontrols populationdynamicsthroughmetabolism,growthandfecundity (Lobon-Cerviaetal.1996),thus,formingakeyphysicochemi- calhabitatfilter(Poff1997)thatdetermines species’potential refugesitesalongtherivercontinuum (Murawski 1993). Mountainstreamsmayalsorepresentharshenvironmentswhich arephysicallystressful(higherrivercompetenceanderosive forcesgeneratedthroughthecombination ofslopewithother variablessuchaswaterdepthandcurrentvelocityandsnowmelt floods).Thesenaturalenvironmentalfactors,combinedwith lowconnectivityamongsuitable sitesincoldheadwater streams,mayberesponsibleforlowlevelsofcolonisation at highaltitudesites(GidoandBrown1999).

Atthesametime,theintensificationofagriculture(including pasturelandsinmountainousareas)mayhavefosteredlocal extinctionoffishspecies. Theagriculturallandscape ofsouth- westernFranceandwesternEuropeingeneral,hasradically changedoverthepastdecades(Meeusetal.1990)asaresultof land-use conversion,intensificationofproductionsystems and abandonment oftraditionalpractices.Theprincipalpressures causingbiodiversitylossinthecountrysidearehabitatfragmen- tation,degradation anddestructionduetoland-usechange. Importantsourcesofagriculture-derived pollutionincludethe inflowofnutrients,pesticidesandheavymetalsfromdiffuseand pointsources(Allan2004).Fragmentationimpactstheconnec- tivitybetweensuitablehabitats(Fahrig2003)andultimately affectscommunitydynamics(Denoe¨landFicetola2008).

Expansion ofurbanlandspositivelyaffectedcolonisation. Althoughourmodeldoesnotallowustodistinguish whether expansionofurbanlandsfavoursnativeorexoticfishcolonisa- tion,closescrutinyof thedatasetrevealedthatcolonizerswere mainlyexoticspeciessuch asPseudorasboraparva(topmouth gudgeon)andSilurusglanis(Europeancatfish).Arecentstudy intheUnitedStatesrevealedthatnon-nativespeciesdominate watershedssupportinghigh densitiesofdams, roads and urban andagriculturallandscapes, whereaswatersheds characterised byupstreamlandprotectionsupportfishcommunities witha strongcomplementofnativespecies(Pooletal. 2010).Accord- ingtoLeprieuretal.(2008),the‘humanactivityhypothesis’ (i.e.bydisturbingnaturallandscapesandincreasingthe impor- tationofnon-nativespecies,humanactivitiesfacilitategreater

levelsofestablishment)bestexplainsthepatternsofnon-native fishrichnessinriverbasins. Inourstudyarea,landscapes characterisedby increasingproportionsof urbanlanduse were successfully colonisedbyexoticspecies,thus,supporting the ideathatassociation withhumansincreasethechancesof successfuldispersalandestablishmentinnon-nativespecies.

Inconclusion, thecolonisation andextinctionpatterns highlightedinthisstudyonlypartiallymatchedourapriori

expectations.Overadecade,exoticfishweremoresusceptible tolocalextinctionthanthenativeones.The‘threetens’rule(one imported speciesin10appearsinthewild,onein10ofthese becomeestablishedandonein10ofestablishednon-indigenous

speciesbecomesapest)whichwasinitiallyformulated for exoticplants(Williamson andFitter1996)probablydoesnot applytoexoticfish,becausetheseanimalsareintentionally

introduced byhumansinpotentiallysuitablehabitats.There- fore,itislikelythatextinctioninexoticfish wasnotrelatedto unsuccessfulsettlement,butrathertoexternalfactorssuchas

watertemperatureor flowvariability(Costelloeetal.2010).

Regardlessoftheirbiogeographicstatus(eithernativeor exotic),cold-tolerantfishweremorepronetoextinctionthanthe cool-orthermallytolerantones.Thismaybeduetotherapidity of climatechangeswhicharepredictedtoexceedtheabilityof manyspeciestoadapttonewenvironmental conditions (Morrongielloetal.2011;Oldenetal.2011).Therefore,our resultssupportthehypothesesthat:(i)allfishcommunitiesare susceptible toinvasionregardlessofnativespeciesrichness (MoyleandLight1996; GidoandBrown1999);and(ii)water temperature,oritschanges,isamajordriveroffishextinctionin human-impactedlandscapes.Inlightoffutureclimatescenarios, ourstudyfurthersuggeststhat mountainrangesare theareasat greatestextinctionriskforpoikilotherms andourabilityto detectresponses ofnativeandexoticspeciestolandscape alterationsusingacombination ofsimpleenvironmental vari- ablesexemplifiesacost-effectivetechniqueforassessingareas atgreaterinvasionriskinlargestreamsystems.

Acknowledgements

WewishtothanktheFrenchOfficeNational del’EauetdesMilieux Aquatiques(ONEMA)andmorespecificallyDrN.Poulet,forprovidingus withthefishdata.Threeanonymous reviewersandDrA.Boultonmade usefulcommentsonanearlierversionof thispaper.

References

Allan,J.D.(2004).Landscapesandriverscapes:theinfluenceoflanduseon streamecosystems.AnnualReviewofEcologyEvolutionandSystemat- ics 35, 257–284. doi:10.1146/ANNUREV.ECOLSYS.35.120202.

110122

Balcombe, S.R.,Sheldon,F.,Capon,S.J.,Bond,N.R.,Hadwen, W.L., Marsh,N.,andBernays,S.J.(2011). Climate-changethreatstonative fishindegradedriversandfloodplainsoftheMurray-DarlingBasin, Australia.MarineandFreshwater Research62, 1099–1114. doi:10.1071/MF11059

Blackburn,T.M.,andDuncan,F.P.(2001).Determinantsofestablishment successin19introducedbirds.Nature414,195–197.doi:10.1038/

35102557

Buisson,L.,Blanc,L.,andGrenouillet,G.(2008).Modelling streamfish speciesdistributioninarivernetwork:therelativeeffectsoftemperature versusphysicalfactors.EcologyFreshwaterFish17, 244–257. doi:10.1111/J.1600-0633.2007.00276.X

Case,T.J.(1991).Invasionresistance,speciesbuild-upandcommunity collapseinmetapopulation models withinterspeciescompetition.Bio- logicalJournaloftheLinnean Society.Linnean SocietyofLondon42,

239–266.doi:10.1111/J.1095-8312.1991.TB00562.X

Compin,A.,andCe´re´ghino,R.(2007).Spatialpatternsofmacroinvertebrate functionalfeedinggroupsinstreamsinrelationtophysicalvariablesand land-coverinsouthwesternFrance.LandscapeEcology22,1215–1225. doi:10.1007/S10980-007-9101-Y

Costelloe,J.F.,Reid,J.R.W.,Pritchard,J.C.,Puckridge,J.T.,Bailey,V.E., andHudson,P.J.(2010).Arealienfishdisadvantaged byextremely variablefollowregimesinarid-zonerivers?MarineandFreshwater Research61,857–863.doi:10.1071/MF09090

Denoe¨l,M.,andFicetola,G.F.(2008).Conservationofnewtguildsinan agriculturallandscape ofBelgium:theimportanceofaquaticand terrestrialhabitats.Aquatic Conservation:MarineandFreshwater Ecosystems18,714–728.doi:10.1002/AQC.853

Dudgeon,D.,Arthington,A.H.,Gessner,M.O.,Kawabata,Z.I.,Knowler, D.J.,Leveque,C.,Naiman,R.J.,Prieur-Richard, A.H.,Soto,D., Stiassny,M.L.J.,andSullivan,C.A.(2006).Freshwaterbiodiversity: importance,threats,statusandconservation challenges.Biological ReviewsoftheCambridgePhilosophicalSociety81, 163–182.

Fahrig,L.(2003).Effectsofhabitatfragmentationon biodiversity.Annual

ReviewofEcologyEvolutionandSystematics34,487–515.doi:10.1146/ ANNUREV.ECOLSYS.34.011802.132419

Gido,K.B.,andBrown,J.H.(1999).InvasionofNorthAmericandrainages byalienfishspecies.Freshwater Biology42,387–399.doi:10.1046/ J.1365-2427.1999.444490.X

Gozlan,R.E.,Britton, J.R.,Cowx,I.,andCopp,G.H.(2010). Current knowledgeonnon-nativefreshwaterfishintroductions.JournalofFish Biology76,751–786.doi:10.1111/J.1095-8649.2010.02566.X

Keith,P.,andAllardi,J.(2001).‘AtlasdesPoisonsd’EauDoucedeFrance.’ (PatrimoinesNaturels:Paris.)

Kopp, D., Syva¨ranta, J., Figuerola, J., Compin, A., Santoul, F., and

Ce´re´ghino,R.(2009).Environmentaleffectsrelatedtothelocalabsence ofexoticfish.BiologicalConservation142,3207–3212.doi:10.1016/ J.BIOCON.2009.07.030

Kottelat,M.,andFreyhof,J.(2007).‘HandbookofEuropeanFreshwater

Fishes.’(PublicationsKottela: Berlin.)

Leprieur,F.,Beauchard,O.,Blanchet,S.,Oberdorff,T.,andBrosse,S. (2008).Fishinvasions intheworld’s riversystems:whennatural processesareblurredbyhumanactivities. PLoSBiology6,404–410.

Lobon-Cervia,J.,Dgebuadze,Y.,Utrilla,C.G.,Rincon,P.A.,andGranado- Lorencio,C.(1996).ThereproductivetacticofdaceincentralSiberia: evidencefortemperatureregulationofthespatio-temporalvariabilityof itslifehistory.Journal ofFishBiology48,1074–1087.doi:10.1111/ J.1095-8649.1996.TB01805.X

Magnuson,J.J.,Crowder,L.B.,andMedvick,P.A.(1979).Temperatureas an ecological resource.AmericanZoologist19,331–343.

McKinney,M.L.(2002).Urbanization,biodiversity,andconservation.

Bioscience52, 883–890.doi:10.1641/0006-3568(2002)052[0883: UBAC]2.0.CO;2

Meeus,J.H.A., Wijermans,M.P.,andVroom,M.L. (1990).Agricultural

landscapesinEuropeandtheirtransformation.LandscapeandUrban

Planning18,289–352.doi:10.1016/0169-2046(90)90016-U Morrongiello,J.R.,Beatty,S.J.,Bennett,J.C.,Crook,D.A.,Ikedife,D.N.

E.N.,Kennard,M.J.,Kerezsy,A.,Lintermans,M.,McNeil,D.G., Pusey,B.J.,andRayner,T.(2011).Climatechangeanditsimplications forAustralia’sfreshwaterfish.MarineandFreshwaterResearch62,

1082–1098.doi:10.1071/MF10308

Moyle,P.B.,andLight,T.(1996).Biologicalinvasionsoffreshwater:

empirical rules and assembly theory. Biological Conservation 78,

149–161.doi:10.1016/0006-3207(96)00024-9

Murawski, S.A.(1993).Climatechangeandmarinefishdistributions: forecastinghistoricalanalogy.TransactionsoftheAmericanFisheries Society122, 647–658.doi:10.1577/1548-8659(1993)122,0647: CCAMFD.2.3.CO;2

Nichols,J.D.,Boulinier, T.,Hines,J.E.,Pollock,K.H.,andSauer,J.R. (1998). Estimatingratesoflocalspeciesextinction,colonization,and turnoverinanimalcommunities.EcologicalApplications8,1213–1225. doi:10.1890/1051-0761(1998)008[1213:EROLSE]2.0.CO;2

Olden,J.D.,Kennard,M.J.,Leprieur,F.,Tedesco,P.A.,Winemiller,K.O., andGarc´ıa-Berthou,E.(2010).Conservationbiogeographyoffresh- waterfishes:recentprogressandfuturechallenges.Diversity & Distributions16,496–513.doi:10.1111/J.1472-4642.2010.00655.X

Olden,J.D.,Kennard,M.J.,Lawler,J.J.,andPoff,N.L.(2011).Challenges andopportunitiesinimplementingmanagedrelocationforconservation offreshwaterspecies.ConservationBiology 25,40–47.doi:10.1111/ J.1523-1739.2010.01557.X

Poff,N.L.(1997).Landscapefiltersandspeciestraits:towardsmechanistic understandingandpredictioninstreamecology.JournaloftheNorth AmericanBenthologicalSociety16, 391–409.doi:10.2307/1468026

Pool,T.K.,Olden,J.D.,Whittier,J.B.,andPaukert,C.P.(2010).

Environmentaldriversoffishfunctionaldiversityandcomposition in theLowerColoradoRiverBasin.CanadianJournalofFisheriesand AquaticSciences67,1791–1807.doi:10.1139/F10-095

Strayer,D.L.,Eviner,V.T.,Jeschke,J.M.,andPace,M.L.(2006).

Understandingthelong-termeffectsofspeciesinvasions.Trendsin

EcologyEvolution21,645–651.doi:10.1016/J.TREE.2006.07.007

Vie´,J.-C.,Hilton-Taylor, C.,andStuart,S.N.(2009).‘Wildlifeina ChangingWorld–anAnalysisofthe2008IUCNRedListofThreatened Species.’(IUCN:Gland,Switzerland.)

Vitousek,P.M.,D’Antonio,C.M.,Loope,L.L.,andWestbrooks,R.(1996).

Biologicalinvasionsasglobalenvironmentalchange.AmericanScien- tist84,468–478.

Williamson,M.,andFitter,A.(1996).Thevaryingsuccessofinvaders.

Ecology77,1661–1666.doi:10.2307/2265769