K26-05060-journal_pone_0128713 ( .PDF )

PAGE 1

RESEARCHARTICLERange-WideGeneticAnalysisofLittleBrown Bat( Myotislucifugus )Populations: EstimatingtheRiskofSpreadofWhite-Nose SyndromeMaartenJ.Vonhof1,2* ,AmyL.Russell3,CassandraM.Miller-Butterworth41 DepartmentofBiologicalSciences,WesternMichiganUniversity,Kalamazoo,Michigan,UnitedStatesof America, 2 EnvironmentalandSustainabilityStudiesProgram,WesternMichiganUniversity,Kalamazoo, Michigan,UnitedStatesofAmerica, 3 DepartmentofBiology,GrandValleyStateUniversity,Allendale, Michigan,UnitedStatesofAmerica, 4 PennStateBeaver,Monaca,Pennsylvania,UnitedStatesofAmerica maarten.vonhof@wmich.eduAbstractThelittlebrownbat( Myotislucifugus )isoneofthemostwidespreadbatspeciesinNorth Americaandisexperiencingseverepopulationdeclinesbecauseofanemergingfungaldisease,white-nosesyndrome(WNS).Tomanageandconservethisspecieseffectivelyitis importanttounderstandpatternsofgeneflowandpopulationconnectivitytoidentifypossiblebarrierstodiseasetransmission.However,littleisknownaboutthepopulationgenetic structureoflittlebrownbats,andtodate,nostudieshaveinvestigatedpopulationstructure acrosstheirentirerange.WeexaminedmitochondrialDNAandnuclearmicrosatellitesin 637littlebrownbats(includingallcurrentlyrecognizedsubspecificlineages)from29locationsacrossNorthAmerica,toassesslevelsofgeneticvariationandpopulationdifferentiationacrosstherangeofthespecies,includingareasaffectedbyWNSandthosecurrently unaffected.Weidentifiedconsiderablespatialvariationinpatternsoffemaledispersaland significantgeneticvariationbetweenpopulationsineasternversuswesternportionsofthe range.Overalllevelsofnucleargeneticdifferentiationwerelow,andthereisnoevidencefor anymajorbarrierstogeneflowacrosstheirrange.However,patternsofmtDNAdifferentiationarehighlyvariable,withhigh STvaluesbetweenmostsamplepairs(including betweenallwesternsamples,betweenwesternandeasternsamples,andbetweensome easternsamples),whilelowmitochondrialdifferentiationwasobservedwithintwogroupsof samplesfoundincentralandeasternregionsofNorthAmerica.Furthermore,theAlaskan populationwashighlydifferentiatedfromallothers,andwesternpopulationswerecharacterizedbyisolationbydistancewhileeasternpopulationswerenot.Thesedataraisethe possibilitythatthecurrentpatternsofspreadofWNSobservedineasternNorthAmerica maynotapplytotheentirerangeandthattheremaybebroad-scalespatialvariationinthe riskofWNStransmissionandoccurrenceifthediseasecontinuestospreadwest. PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 1/23 a11111 OPENACCESS Citation: VonhofMJ,RussellAL,Miller-Butterworth CM(2015)Range-WideGeneticAnalysisofLittle BrownBat( Myotislucifugus )Populations:Estimating theRiskofSpreadofWhite-NoseSyndrome.PLoS ONE10(7):e0128713.doi:10.1371/journal. pone.0128713 AcademicEditor: FilipposA.Aravanopoulos, AristotleUniversityofThessaloniki,GREECE Received: November19,2014 Accepted: April29,2015 Published: July8,2015 Copyright: 2015Vonhofetal.Thisisanopen accessarticledistributedunderthetermsofthe CreativeCommonsAttributionLicense ,whichpermits unrestricteduse,distribution,andreproductioninany medium,providedtheoriginalauthorandsourceare credited. DataAvailabilityStatement: Allsequencedatahas beensubmittedtoGenbank(accessionnumbers KM363877-KM363977andKM382071-382147).All otherrelevantdataareinthepaperandits SupportingInformationfiles. Funding: Fundingforthisstudywasprovidedby WesternMichiganUniversity,thePennsylvaniaGame CommissionStateWildlifeGrantProgram (CooperativeAgreementNo.231100),andthe PennsylvaniaStateUniversityResearch DevelopmentGrants.

PAGE 2

IntroductionUnderstandinghowhostmovementpatternsinfluencethetransmissionofpathogensiscritical tothedevelopmentofeffectivepreventionandcontrolstrategies,andtotheconservationand managementofhostpopulationsduringandafterdiseaseoutbreaks.However,formanyhost species,dataonindividualmovementsandcontactratesaredifficultorimpossibletocollect becauseofcrypticbehavior,thegeographicscaleofmovements,ormethodologicalconsiderationsthatlimitourabilitytofollowindividualsthroughtimeandspace.Evidencefromempiricalstudiesemployingpopulationandlandscapegeneticapproacheshasdemonstratedthat landscapefeatures,suchasmountainsandriversthatlimithostgeneflow,oftenrepresentbarrierstodiseasetransmission[ 1 – 6 ],althoughalternativemechanismsofpathogendispersal, includinghumansandotherhighlymobileintermediatehosts,mayoverridetheinfluenceof primaryhostpopulationgeneticstructure[ 1 ].Nevertheless,wheretheyexist,suchbarriersto hostgeneflowcanhaveadramaticimpactoninitialdiseaseestablishment,therateanddirectionofdiseasespread,spatialpatternsofhostresistance,anddynamicsandgeneticstructureof pathogenpopulations[ 1 – 6 ].Assumingthatratesofcontactamongindividualsleadingtogene flowareindicativeofcontactsthatcouldresultindiseasetransmission,geneticmethodsprovideausefulalternativetotraditionaldemographicapproachesasameansofexamininghost movementsandtheirimpactondiseasetransmission[ 1 ]. White-nosesyndrome(WNS)isanemergingfungaldiseasecausinghighlevelsofmortality inhibernatingNorthAmericanbats[ 7 – 9 ].Thecausativeagent, Pseudogymnoascusdestructans (hereafter Pd ),isacold-lovingfungusthataffectsbatsduringhibernationandsubsequent arousal. Pd causescharacteristiccup-likeerosionsoftheepidermisofthewingsandmuzzle, andmayinvadesebaceousglandsandhairfollicles[ 10 ].Mortalityofbatslikelyoccursthrough thelossofphysiologicalhomeostasis[ 11 ],possiblyassociatedwithdehydrationandelectrolyte depletion[ 12 13 ],leadingtomorefrequentarousalbehaviorandprematurelossoffatreserves [ 14 15 ].SinceitwasfirstdiscoveredinNewYorkStateduringthewinterof2006 – 07,WNShas sincespreadto27additionalstatesandfiveCanadianprovinces,andisknowntoaffectatleast sevenspeciesofhibernatingbats[ 16 ].Mortalityratesvaryconsiderablyamongspeciesbutcan beveryhigh( > 90%forlittlebrownbats, Myotislucifugus ,andnorthernlong-earedbats, M septentrionalis [ 9 ]),andcumulativemortalityofallaffectedbatspecieshasbeenestimatedat 5.7to6.7millionindividualsasofJanuary2012[ 17 ].Therapidemergence,andthegeographic andtaxonomicspreadofthediseasehaveraisedseriousconcernsaboutthelong-termsurvival ofhibernatingbatspeciesineasternNorthAmerica,andhavehighlightedourlackofknowledgeofthefactorsthatmayinfluenceWNStransmissionandspreadtocurrentlyunaffected regions. LittlebrownbatswereamongthefirstspeciestobediagnosedwithWNS[ 7 ],andpopulationmodelsindicatethatifmortalityratesstayconstant,thisspeciescouldbeextirpatedfrom thenortheasternUnitedStateswithin16years[ 18 ].Hibernatingpopulationsofallsizeshave beenaffectedbyWNS,buttheprobabilityofinfectionincreaseswithincreasingcolonysize [ 19 20 ],althoughmortalitywithinpopulationsisdensity-independentandcharacterizedby frequency-dependenttransmission[ 21 ].Thus,thereisahighprobabilitythatlittlebrownbat populationsincurrentlyaffectedregionswillbehighlyreducedandpossiblybeextirpatedin comingdecades.Additionally,WNSmayposeathreattotheentirespeciesifthediseasecontinuestospreadacrossthespecies ’ range.Further,becauselittlebrownbatsareoneoftwo affectedbatspecieswhosegeographicrangesspantemperateNorthAmerica,theymaydrive transmissionof Pd toanovelsuiteofwesternNorthAmericanhibernatingbatspeciesthat mightotherwiseremaingeographicallyisolatedfromthedisease.Therefore,itiscrucialthat Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 2/23 CompetingInterests: Theauthorshavedeclared thatnocompetinginterestsexist.

PAGE 3

weunderstandtheprobabilityof Pd transmissionacrosstherangeoflittlebrownbats,and whethertherearebarrierstogeneflowthatcouldrestrictthegeographicspreadofWNS. Hereweapplygeneticapproachestounderstandlevelsofgeneflowandpopulationconnectivityinthelittlebrownbat.Thissmall(6 – 10g)insectivorousbatspeciesisamongthemost widespread( Fig1 )andwell-studiedinNorthAmerica[ 22 23 ].Duringthesummer,reproductivefemalesformmaternitycoloniesinbuildings,trees,orcreviceswhereparturitionandpostnatalcaretakeplace,whilemalesandnon-reproductivefemalestypicallyroostsolitarily[ 22 ]. Inwinter,bothsexescongregateinhibernacula,andmatingtakesplaceduringthepre-hibernationswarmingperiod,orduringhibernationitself[ 24 25 ].Thesizeofhibernatingpopulationsmayvaryconsiderably,ontheorderof10 ’ sto100,000 ’ s,butmostofthelargerknown hibernaculaoccurinkarstregionsofeasternNorthAmerica,andverylittleisknownaboutthe distributionorsizeofhibernaculainwesternNorthAmerica.Becauseindividualsfrommany breedinggroupscometogetheratswarmingorhibernationsiteswithmalesthatmayormay nothaveoriginatedfromthesamebreedinggroup[ 26 27 ],thesesiteshavebeensuggestedto represent ‘ hotspots ’ ofgeneflowfortemperatebats[ 28 – 30 ].Thus,patternsofgeneflowwill representtheinterplayofmovementsofindividualsbetweensummerand/orwinterpopulations,andlevelsandspatialpatternsofconnectivityamongsummerandwinterpopulations thatdeterminethecompositionofmatingaggregations. Therearecurrentlyfiverecognizedsubspeciesoflittlebrownbats( M l alascensis M l carissima M l lucifugus M l pernox ,and M l relictus [ 22 31 ];see Fig1 )basedonmorphology, buttheextenttowhichthesesubspeciesdivergegeneticallyisunclear.Coalescentanalysesof nuclearDNA(nucDNA)andmitochondrialDNA(mtDNA)suggestthatsomesubspeciesmay representindependentevolutionarylineages,butthat M lucifugus maybeparaphyleticwith respecttothewesternlong-earedbat, M evotis [ 32 ].Inaddition,twodistinctmtDNAlineages (correspondingto M. l carissima and M l lucifugus )co-occurinsouthernAlbertaandnorthcentralMontana,butthesetwogroupsofbatsarenotdifferentiatedbasedonnuclearmicrosatelliteDNAormorphology,suggestingthatthesubspeciesinquestionmayinterbreedand exchangegenes[ 33 ].Asinglemitochondriallineagecorrespondingto M l lucifugus was observedintheMinnesotapopulations,andtherewasastrongsignalofpopulationexpansion datingto18kya[ 34 ].EnvironmentalnichemodelingbasedonconditionsduringtheLastGlacialMaximum(LGM)indicatedthepresenceofasinglelargerefugiumextendingacrossthe southeasternandsouth-centralUnitedStates,andmorefragmentedrefugiainthesouthern portionofthemountainouswesternUnitedStates[ 34 ],suggestingapossiblemechanismfor lineagedifferentiationwithinthisspecieswhereseparationintodisjunctglacialrefugiawasfollowedbysubsequentpost-glacialrangeexpansionandsecondarycontact. Fewstudieshaveexaminedgeneticvariationinlittlebrownbats,andtherehasbeenno comprehensiverange-widepopulationgeneticanalysisofthisspecies.Fine-scalegeneticstudiesinMinnesotadescribedhighlevelsofmtDNAstructureandweakbutsignificantnucDNA structureamongmaternitycolonies[ 35 ],apatternconsistentwithmanyothertemperatebat speciescharacterizedbyfemalephilopatrytosummerbreedinghabitatandextensivegeneflow viamatingatswarmingsitesduringtheautumn[ 36 – 41 ].InsouthesternCanada,highlevelsof geneticconnectivitywereidentifiedamongswarmingsites,howeverminimalstructuringat bothmtDNAandnucDNAmarkerssuggesteddispersalmayoccurinbothsexes,althoughit maybemale-biased[ 42 ].InPennsylvania,nosignificantnuclearstructurewasidentified amonghibernatingpopulations,butthesepopulationswerestructuredmatrilineally.This mtDNAstructurewascorrelatedwithlocaltopography,whichmayhavedelayedthespreadof WNStowesternpartsofthestate[ 6 ]. TherapidspreadofWNSthrougheasternNorthAmericanpopulationsoflittlebrownbats (andotheraffectedspecies)suggeststhatfewbarrierstotransmissionexistwithinthecurrent Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 3/23

PAGE 4

Fig1.MapofCanadaandtheUnitedStatesshowingthedistributionofdescribed Myotislucifugus subspecies(modifiedfrom[ 17 ])andsamplinglocations. a)showssamplinglocationswithpiecharts indicatingfrequenciesofmtDNAsubspecificclades(subspecificdesignationsareindicatedinthelegendand colorsfollowthoseusedin Fig2 )ineachpopulation,whileb)showsgroupingsofpopulations(orangeand bluedots)withinwhichpairwise STvaluesbasedonmtDNAhaplotypefrequencieswerelowversus populationsthatweresignificantlydifferentiatedfromallotherpopulations(purpledots;highpairwise STvalueswithallothersampledpopulations).OnesampledpopulationinMichigan(shownwithablackdotina) Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 4/23

PAGE 5

rangeofthedisease.HereweutilizemtDNAsequenceandnucDNAmicrosatellitevariation fromalargesampleoflittlebrownbatscollectedacrosstherangeofthespeciestoaddressthe followingobjectives:1)assesslevelsofgeneticvariationinlittlebrownbatpopulations,includingareasaffectedbyWNSandthosecurrentlyunaffected;2)quantifygeneticdifferentiation amongpopulationssampledacrosstherangeofthespecies,includingpopulationsineastern NorthAmericawithinthecurrentrangeofWNS,aswellasadditionalpopulationssituated botheastandwestofthetransitionbetweentheGreatPlainsandRockyMountains;and3) assessthecurrentgeographicdistributionofandlevelsofgeneticdifferentiationamongcurrently-recognizedsubspecificlineages. Therearefewphysiographicbarriersthatwouldlimitmovementofhighlyvagileorganisms eastoftheRockyMountains.PhylogeographicstudiesofwidespreadbatsandbirdsinNorth wasnotincludedinmtDNAanalyses.Populationabbreviationsaredetailedin Table1 ,andcolorsinpie chartsina)correspondtocladesshownin Fig2 .Datasourcesforthemapinclude:nationalatlas.gov, iucnredlist.org,andESRIData&Maps2006throughArcGIS( S1File ). doi:10.1371/journal.pone.0128713.g001 Fig2.PhylogenetictreeshowingrelationshipsbetweenNearctic Myotis (sensu[ 58 ]),andthepresenceofthreedistinctcladeswithin M lucifugus basedonmaximumlikelihoodanalysisofpartialCOIsequencesinPhyML3.0. AmemberoftheNeotropical Myotis clade( M austroriparius )was includedastheoutgroup.Leavesarecollapsedtohighlightwell-supportedclades,andtheverticaldimensionofthetrianglesisproportionaltoth enumberof samplesincluded.SH-likebranchsupportvaluesareprovidedforallmajorclades.Cladescontaining M lucifugus aredesignatedbythespecific abbreviationfollowedbythesubspeciesname(e.g., M l lucifugus referstothenominalsubspecies).Notethatoneclade(including M l carissima )also containsmembersofotherspecies(including M evotis M keenii ,and M thysanodes )aspreviouslydescribed[ 32 ]. doi:10.1371/journal.pone.0128713.g002 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 5/23

PAGE 6

Americatypicallyreportlittledifferentiationamongpopulationswithineasternandcentral portionsofNorthAmerica,significantdifferentiationamongeasternandwesternpopulations, andhigherlevelsofdifferentiationamongpopulationswithinthemountainouswest[ 41 43 – 46 ].WepredictthattheRockyMountainswillrepresentabarriertogeneflow,andthatwe willthereforeobservegeneticdifferentiationbetweensamplesiteseastversuswestoftheGreat Plains-RockyMountainstransition.Further,becauseofhighertopographicvariability,wepredicthigherlevelsofgeneticdifferentiationamongsamplesitesinthemountainouswestthan ineasternNorthAmerica.Ifsubspecificlineagesrepresentreproductively-isolatedunitsthat aroseduringtheLGM,thenwepredictthatpatternsofdifferentiationatbothnucDNAand mtDNAmarkerswillmatchthedescribedgeographicdistributionofsubspecies.Ourstudy providesvaluabledataonpopulationconnectivityandhenceopportunitiesforWNStransmissionacrosstherangeoflittlebrownbatsthatmaybeusedtoinformthemanagementandconservationofaffectedspecies.Methods SamplecollectionTissuesampleswereobtainedduringthesummer(betweenMayandAugust)from637individualsat29locationsacrosstherangeoflittlebrownbats( Table1 S1Table ,and Fig1 ).Two 3mmbiopsypunches,onefromeachwing,weretakenfromeachbatandstoredin5MNaCl with20%DMSO[ 47 ].Thebatswerereleasedaftersampling.Themajorityofpopulationsampleswerecollectedatmaternitycolonies(N=16)orsingleorseveralclosely-spaced( < 10km) nettingsites(N=12).However,theIdahosampleconstitutedbatscollectedin8different countiesinthesoutheasternportionofthestate.Whensamplescamefrommorethanonecapturelocation,centroidswerecalculatedandusedasapproximatesamplelocations.EthicsstatementOneauthor(MJV)collectedsamplesforthisstudyfromonepopulationinMichigan.ThesampleswerecollectedunderpermissiongrantedbytheStateofMichiganDepartmentofNatural Resources(PermitSC1257),andthemethodswereapprovedbytheWesternMichiganUniversityInstitutionalAnimalCareandUseCommitteeProtocolNumber08-05-03.Allother sampleswerecollectedbyuniversityandgovernmentresearchersperformingotherresearch whowererequiredtohaveappropriatepermitsandothernecessarypermissiontoundertake theirwork.MitochondrialDNAsequencingandanalysisTotalgenomicDNAwasisolatedusingDNeasyTissueKits(Qiagen,ValenciaCA).Weamplifiedandsequenceda636bpfragmentofthemitochondrialcytochromecoxidasesubunitI (COI)geneusingprimersHCO2198andLCO1490[ 48 ]orprimersVF1andVR1[ 49 ].Bats fromallsamplesitesexceptMichiganweresequenced,foratotalof617individuals.PCRs wereconductedin25 lvolumescontaining0.4 Mofeachprimerand20 – 50ngofDNAtemplate,usingIllustraPuReTaqReady-To-GoPCRbeads(GEHealthcareLifeSciences,PittsburghPA).Whenreconstitutedto25 lwithwater,thesebeadscontained2.5unitsPuReTaq DNApolymerase,200 MeachdNTPin10mMTris-HCl(pH9),50mMKCl,1.5mM MgCl2,andanunspecifiedconcentrationofbovineserumalbumin(BSA).Nootheradditives wereaddedtothesolution.Cyclingconditionsconsistedofonecycleof5minat94C,30 cyclesof30secat94C,45secat68Cand1minat72C,andafinalcycleof2minat72C. PCRproductswerepurifiedbydigestionwithexonucleaseIandshrimpalkalinephosphatase Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 6/23

PAGE 7

Table1.SampledlittlebrownbatpopulationsanddiversitystatisticsformitochondrialCOIsequences( NSEQ,numberofindividualssequenced; NHAP,numberofhaplotypes; NHAP-UN,numberofhaplotypesuniquetoasite; h ,haplotypediversity; ,nucleotidediversity),andmicrosatellite genotypes( NGEN,numberofindividualsgenotyped; HO,observedheterozygosity; HE,expectedheterozygosity; NA,meannumberofallelesper locus; AR ,allelicrichness; AR-P ,privateallelicrichness; FIS,inbreedingcoefficient). Provinceor State CountyAbbrev.WNS-Status NSEQNHAPNHAP-UNh NGENHOHENAARAR-PFISAlaskaKenai Peninsula AKNeg18550.650.0017180.8640.8148.07.670.01-0.061 CaliforniaMariposaCA-MaNeg10210.200.00160 CaliforniaMonoCA-MoNeg15210.480.00080 CaliforniaSiskiyouCA-SiNeg20760.840.0026200.9220.92013.212.200.12-0.003 CaliforniaShastaCA-ShNeg11210.180.00140 WashingtonSkagitWANeg17860.640.0175170.8690.91512.211.800.170.050 British ColumbiaSouth BC-SNeg30760.610.0157300.8890.90314.611.800.120.015 IdahoMultipleIDNeg26640.520.0038230.8600.90413.611.890.260.049 WyomingCarbonWYNeg30970.790.0148300.8700.90415.012.090.020.037 AlbertaSouth AB-SNeg28960.660.0103290.9200.91816.112.920.12-0.002 AlbertaNorth AB-NNeg28950.790.0045290.9000.91615.912.760.260.017 British ColumbiaNorth BC-NNeg15850.700.0040150.9190.91513.313.330.14-0.003 ManitobaMBNeg11410.750.00440 OntarioON-1Neg15770.890.00660 MinnesotaStLouisMNNeg20730.730.0046200.9170.89413.812.290.29-0.025 WisconsinMarquetteWI-MaNeg20940.870.0057200.8890.89513.612.040.390.007 WisconsinSaukWI-SaNeg20720.800.0045160.8680.90512.011.700.430.041 MichiganCassMINeg0200.8280.88113.011.640.650.061 KentuckyRowanKYPos2913110.840.0043330.8860.90015.712.190.330.016 OhioFair eldOHPos311150.880.0059300.8630.88114.311.270.260.021 TennesseeBlountTNPos32410.460.0008300.8520.90514.811.910.230.059 West Virginia RaleighWVPos20730.740.00330 OntarioON-2Pos301240.800.00310 Pennsylvania BlairPAPos211240.880.0035210.8280.89413.611.900.160.018 MarylandWashingtonMDPos3412100.910.0041300.8860.89616.012.150.220.004 NewYorkJeffersonNYPos201380.930.0036180.8630.90013.012.080.180.026 NewJerseyMorrisNJ-MoPos301240.900.0041310.8520.89415.011.930.290.045 NewJerseySalemNJ-SaPos20720.780.00240 QuebecQBPos16740.780.0019300.8480.89014.311.430.10.047 Overall6177.84.50.720.00515100.8760.89713.911.900.230.020 Populationsareorderedwesttoeast.WNS-StatusdenotessamplinglocalitieswithinstatesorprovincesthatwerepositiveornegativeforWNSasof 2012. doi:10.1371/journal.pone.0128713.t001 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 7/23

PAGE 8

(EXOSAP),andweresequencedinbothdirections,usingtheamplificationprimers,attheUniversityofArizonaGeneticsCoreFacility.SequenceswereeditedusingCodonCodeAligner3.0 (GeneCodesCorp.)andalignedusingthedefaultsettingsinMAFFT[ 50 ].MicrosatellitegenotypingandanalysisWegenotypedindividualsatelevenhighlyvariablemicrosatellitelociusingprimerspreviously developedforothervespertilionidbats(IBatCA5,CA11,CA43,CA47,andM23[ 51 ];MS3D02 andMS3F05[ 52 ];E24andG9[ 53 ];Cora_F11_C04[ 54 ];Coto_G02F_H10R[ 55 ]).Wedidnot genotypepopulationsampleswith 15individuals(CA-Ma,CA-Mo,CA-Sh,MB-1,MB-2; Table1 ),withtheexceptionofBC-Nthathadbeengenotypedforanotherstudy.BasedonpreliminaryresultsoflowlevelsofdifferentiationineasternNorthAmericaandtoreducecosts, wedidnotgenotypeindividualsfromthreeadditionalsiteseastoftheRockyMountains(ON2,WV,andNJ-Sa).Thetotalnumberofindividualsgenotypedwas510.Amplificationswere carriedoutinfourmultiplexreactionsandtwosingle-locusamplifications(see S2Table ),and weresubsequentlypooledintothreedifferentloadsforfragmentanalysisonanABI3130 sequencer.Thebasiccyclingconditionsconsistedof1minat94C,threecyclesof30secat 94C,20secatTa(54or60C),and5secat72C,33cyclesof15secat94C,20secatTa,and 10secat72C,followedbyafinalextensionat72Cfor30min.Someamplificationsrequired additionalcyclesortheremovalofthefinalextensionstep( S2Table ).Fragmentswereanalyzed andscoredusingGeneMarkersoftware(SoftGeneticsLLC,StateCollege,PA).MitochondrialDNAanalysisTodescribeoveralllevelsofmtDNAdiversitywithinpopulations,wecalculatedhaplotype( h ) andnucleotide( )diversitiesinDNASPv.5.10.1[ 56 ],anddeterminedthenumberofprivate haplotypesineachsiteaftercollapsingsequencesfromtheentiredatasettouniquehaplotypes usingFABOX[ 57 ].Weusedseveralpopulationgeneticapproachestoestablishwhethercurrentpatternsofvariationareindicativeofthepresenceofdistinctgeneticclusters.Wecalculatedpairwise FSTvaluesbetweensitesandtestedforsignificancewith10,000permutationsin Arlequinv.3.11[ 58 ]toidentifypairsofsitesthatweregeneticallydistinct.Wealsoperformed ananalysisofmolecularvariance(AMOVA[ 59 ])todescribetherelativeamountofgenetic variationwithinandamongsites.Basedoninitialpairwise FSTresults,wethenperformed nestedAMOVAstoidentifynaturalgroupsofsites.Siteswereinitiallygroupedtogetherifthey hadlowpairwise FSTvalues,andtheanalysiswasrerun.Anyambiguoussites(sitesthathad low FSTvalueswithsitesinmorethanonegroup)weresequentiallymovedbetweengroups andtheanalysiswasrerun.Alllogicalcombinationsweretestedtoidentifythegroupingthat minimizedamong-site/within-groupvariationandmaximizedbetween-groupvariation. Totestthesignificanceofdefinedsubspecificlineageswithin M lucifugus usingournationwidedataset,weusedamaximumlikelihoodphylogeneticapproachimplementedinPHYMLV.3.0[ 60 ].WesequencedCOIforotherNorthAmerican Myotis spp.( M californicus M ciliolabrum M evotis M keenii M leibii M sodalis M thysanodes ,and M volans ;cf[ 61 ]),anda memberoftheNeotropical Myotis clade( M austroriparius ),whichwasusedastheoutgroup (see S3Table forlistofspecimens).Weusedthebestfitmodelofsequenceevolution(HKY +G)asdeterminedusingMegav.5.0[ 62 ],withthegammadistributionofvariabilityofrates amongsitescalculatedempiricallyfromthedata,SPRmovestoexploretreespace,andSH-Like proceduretoassessbranchsupports[ 60 ].Theproportionofeachsampledpopulationfalling withineachsubspecificcladewasthencalculatedandplottedonamapproducedinARC-GIS v.10.1tovisualizethegeographicdistributionoftheclades. Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 8/23

PAGE 9

MicrosatelliteDNAanalysisDeviationsfromHardy-Weinbergequilibrium(HWE)wereestimatedforeachlocus,andloci wereconfirmedtobeinlinkageequilibriumusingFSTATv.2.9.3[ 63 ].Totestfordifferences inlevelsofgeneticdiversityamongsitesandregions,severalindicesofnucleargeneticdiversity wereestimated,includingnumberofallelesperlocus,allelicrichness,andtheinbreedingcoefficient( FIS)usingFSTAT,privateallelicrichnessusingHP-RAREv.1.0[ 64 ],andobservedand expectedheterozygosityusingGENODIVE[ 65 ].Wethentestedfordifferencesamongsites (orgroupsofsites)inallelicrichnessand FISinFSTAT,andexpectedheterozygosityinGENODIVE,using10,000permutations.Testswereperformedamongclustersofsitesidentified usingclusteringtechniques(seebelow),andamongsitesfallingwithinstatesorprovincesthat wereWNS-positive(KY,OH,TN,WV,ON-2,PA,MD,NY,NJ-Mo,NJ-Sa,QB; Table1 )or WNS-negative(allothersites)asof2012,althoughitshouldbenotedthattissueswerecollectedpriortotheemergenceofWNSattheselocalitiesinallcases. Weappliedthreedifferentapproachestodeterminethemostlikelynumberofdistinct geneticclustersindependentoforiginalsamplinglocations,asdifferentclusteringalgorithms canproducedifferentsolutionsandconcordanceamongmultipletechniquesissuggestiveof thepresenceofastronggeneticsignal[ 66 ].First,weutilizedthemodel-basedBayesianclusteringapproachinSTRUCTUREv.2.3.3[ 67 68 ]withpopulationmembershipasaprior[ 69 ].To determinetheoptimalnumberofclusters( K ),weran10runsper K ,for K =1 – 10,witha 100,000MCMCiterationburn-infollowedby400,000iterationsusingtheadmixturemodel withcorrelatedallelefrequencies.Themostlikelynumberofclusterswasdeterminedusingthe Evannoetal.[ 70 ]methodimplementedintheprogramSTRUCTUREHARVESTER[ 71 ].The Evannoetal.[ 70 ]methodisnotinformativeforthehighestandlowest K ;therefore,ifthehighestloglikelihoodvaluewasobservedfor K =1or10acrossallreplicates,weacceptedthatas the K withthehighestprobability.Forthebestvalueof K weusedCLUMPP[ 72 ]toharmonize individualassignmentstoclusters. Second,weappliedthe K meansclusteringapproachinGENODIVEv.2.0asoutlinedby Meirmans[ 65 ].ThisapproachisbasedonanAMOVAframeworkandusesasimulated annealingalgorithmtominimizetheamong-populations/within-groupssumofsquares throughmaximizationof FCT,thevarianceamongclustersrelativetothetotalvariance.We determinedthemostlikelynumberofclustersusingthePseudoF summarystatistic,which performsbetterthanthealternativeBayesianInformationCriterion(BIC)whenmigration ratesarehighandmatingisrandom[ 65 ]. ThethirdapproachwasthatofDuchesneandTurgeon[ 73 ]implementedinthesoftware FLOCK.Samplesarerandomlypartitionedinto K clusters( 2),allelefrequenciesareestimatedforeachofthe K clusters,andeachgenotypeisthenreallocatedtotheclusterwiththe highestlikelihoodscore.Repeatedreallocationbasedonlikelihoodscores(20iterationsper run)resultedingeneticallyhomogeneousclusterswithinarun.Fiftyrunswerecarriedoutfor each K ,andattheendofeachrunthesoftwarecalculatedtheloglikelihooddifference(LLOD) scoreforeachgenotype(thedifferencebetweentheloglikelihoodofthemostlikelyclusterfor thegenotypeandthatofitssecondmostlikelycluster)andthemeanLLODoverallgenotypes. Strongconsistencyamongruns(resultingin ‘ plateaus ’ ofidenticalmeanLLODscores)isused toindicatethemostlikelynumberofclusters[ 73 ].Basedonthisanalysis,were-rerantheiterativereallocationprocedureforthemostlikely K andplottedthemeanLLODscoreagainstgeographicallyorderedsitesasameansofidentifyingadmixturelevelsbetweengeneticclusters. Forallthreemethodsofgeneticclusteridentification,wetestedwhetherclusterassignment wasvalidbyquantifyingthenumberofindividualswithineachsamplethatwereallocatedto eachcluster,andthenbuildingan r c contingencytablewhere r isthenumberofgenetic Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 9/23

PAGE 10

clustersand c isthenumberofsamplesites.Wethentestedforrandomallocationtothe geneticclustersacrossempiricalsamplesusingalikelihood-ratiotestwithWilliams ’ correction withthenullhypothesisthatclusterassignmentswererandomacrosssampledsites.Arejection ofthenullhypothesisindicatedthatthatclustercompositionwasunlikelytoberandomacross thesamples,andthatclusterassignmentswerethereforevalid[ 74 ].Inaddition,giventhat mostclusteringtechniquesassumethatgenotypicproportionswithineachclusterareinHWE andatlinkageequilibrium,wetestedidentifiedclustersforcompliancewiththeseassumptions assuggestedbyGuillotetal.[ 66 ].TotestwhetherclusterassignmentwasindependentofsubspecificmtDNAclademembershipwecouldnotsimplytestforanassociation,ascluster assignmentwasconfoundedbyspatialvariationinthedistributionofmtDNAclades.Therefore,wecompiledclusterandclademembershipforindividualsineachofthefoursitesthat containedmembersofmorethanonemtDNAclade(see Results Fig1a ),andperformedalikelihood-ratiotesttodeterminewhetherclusterassignmentswereindependentofsubspecific clademembershipwithinheterogeneouspopulations. Thelevelofgeneticdifferentiationamongpre-definedsitesandanalternativegrouping basedonsubspecificclademembership,whereindividualswereclassifiedasbelongingtothe M l alascensis M l carissima ,or M l lucifugus cladesbasedonthemtDNAphylogenetic analysis,wasdeterminedbycalculatingpairwisedistancemeasures,including FST[ 75 ],and twomeasuresindependentoftheamountofwithin-populationdiversity:Jost ’ sD[ 76 ],and G ”ST[ 77 ].Differencesinthemagnitudeofpairwisedistancemeasuresamonggroupsofsites wastestedusing10,000permutationsinGENODIVE( G ”STandJost ’ sD)andFSTAT( FST).As withmtDNA,weperformedanAMOVAonmicrosatellitegenotypesusingARLEQUIN,and subsequentlyperformednestedAMOVAsbygroupingsiteswithlow FSTvaluestoidentifythe groupingthatmaximizedamong-groupvariationandminimizedamong-site/within-group variation.IsolationbydistanceThereisconsiderableevidencetosuggestthat,regardlessofthealgorithmemployed,clustering methodsareconfoundedbythepresenceofisolationbydistance(IBD),suchthatconsistent clinalgeneticvariationmaybemisinterpretedbyclusteringalgorithmsasthepresenceofdistinctclusterseventhoughthereisnotruebarriertogeneflow[ 66 78 – 80 ].Wethereforetested forIBDinmitochondrialandnuclearDNAbothglobally(includingallsampledlocations)and withinidentifiedclusters(formicrosatellitedataonly).WeconductedaManteltestcomparing standardizedgeneticdistance[ FST/(1FST)]andthenaturallogofgeographicdistance[ 81 ] usingtheIBDWebService[ 82 ].Tocalculatebetween-sitegeographicdistances,polylineswere constructedfromX,YcoordinatesinArcGIS10.1.Thegeodesicdistanceofthesepolylineswas calculatedusingthe “ Shape.length@meters ” command.Forthemicrosatellites,wefollowedthe recommendationsofGuillotetal.[ 66 ]:weplottedgeneticdistance( FST)againstgeographic distancewhiledifferentiatingbetweendatapointsforsitepairsthatbelongedtothesame geneticclusteranddatapointsforsitepairsbelongingtodifferentclusters.Ifclustersarereal, thenforanygivengeographicdistance,geneticdistancebetweensitepairsindifferentclusters shouldconsistentlybegreaterthandistancebetweensitepairsfallingwithinthesamecluster (e.g.[ 83 84 ]).Inaddition,assuggestedbyMeirmans[ 80 ],weperformedapartialManteltest toinvestigatetheassociationbetweenthematrixofgeneticdistances( FSTand G ”ST)anda matrixofclustermembershipforthemicrosatellitedata,withthematrixofgeographicaldistancesasacovariate.Ifthereisnotarelationshipbetweengeneticdistanceandclustermembershipaftercontrollingforgeographicdistance,thenclustermembershipislikelyconfounded byIBD.Giventwomainclusters(see Results ),wecodedclustermembershipintwoways:a) Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 10/23

PAGE 11

bothmembersofasamplingsitepairbelongingtothesamecluster(1)ornot(2);orb)both membersofasamplingsitepairbelongingtoawesterncluster(1),bothbelongingtoaneastern cluster(2),orsamplesitesbelongindifferentclusters(3).PartialManteltestswerecarriedout inPASSaGEsoftwareandsignificancewasassessedwith10,000permutations[ 85 ].Results GeneticdiversityWeobserved148uniquehaplotypescharacterizedby104segregatingsitesamongthe617individualssequenced.Thenumberofhaplotypespersiterangedfrom2 – 13(mean:7.8),andthe numberofhaplotypesuniquetoasiterangedfrom1 – 11(mean:4.5; Table1 ).Therewasnosignificantdifferenceinnucleotidediversity( )betweensiteseastversuswestoftheGreatPlainsRockyMountainstransition(Mann-WhitneyU-Test;MeanEast:0.00,West:0.01, P =0.829), orbetweensitesinstatesthatwerepositiveornegativeforWNSasofthe2012 – 2013winter season(Mann-WhitneyU-Test;WNS-Neg:0.01,WNS-Pos:0.00, P =0.132).However,sites westoftheGreatPlains-RockyMountainstransitionhadsignificantlylowerhaplotypediversitythanthoseeastoftheboundary(Mann-WhitneyU-Test;East:0.80,West:0.56, P =0.002), andWNS-freesitesasof2012alsohadsignificantlylowerhaplotypediversitythanWNSaffectedsites(Mann-WhitneyU-Test;WNS-Neg:0.65,WNS-Pos:0.81, P =0.015),although thislatterresultislikelyconfoundedbythehighproportionofwesternsitesintheWNS-Neg group. Althoughweoriginallytyped11microsatelliteloci,twoloci(E24andCOTO_G02_H10) hadhighnullallelefrequenciesandweredroppedfromfurtheranalyses.Theremainingnine lociallmetHWEexpectationsandwereunlinked.Meanobservedandexpectedheterozygositieswerehigh(0.876and0.897,respectively),aswasthemeannumberofallelesperlocus (13.9)andallelicrichness(11.9),althoughprivateallelicrichnesswaslow(0.23; Table1 ;see S4 Table fordiversitystatisticsforeachlocus).Comparingsample-levelmeasuresofgeneticdiversityamongidentifiedclustersandamongsamplesinWNS-positiveandWNS-negativestates orprovincesrevealednosignificantdifferencesinobservedorexpectedheterozygosity,allelic richness,or FIS( P > 0.05inallcasesbasedonpermutationtests).Differentiationof M lucifugus subspeciesPhylogeneticanalysisofthemtDNACOIsequencesconfirmedtheexistenceofthreeclades withinlittlebrownbatsroughlycorrespondingtopreviouslydefinedsubspecies( Fig2 ).FollowingCarstensandDewey[ 32 ],werefertothesecladesas M l lucifugus M l carissima ,and M l alascensis (hereafterlucifugus,carissima,andalascensis,respectively).Thecarissimaclade hadthreeotherspecies( M evotis M keenii M thysanodes )nestedwithinit,aspreviously described[ 32 ].Wehavefocusedon M lucifugussensustricto here.Addressingthetaxonomic relationshipsamong M lucifugus M evotis M keenii ,and M thysanodes isbeyondthescope ofthispaper;thereforeweignoredthepresenceoftheseadditionaltaxaintherestofouranalyses.Forthemostpart,cladesweregeographicallyrestrictedandfollowedpreviouslydefined geographicranges( Fig1a ).AlmostallofthesampledsiteseastoftheGreatPlains-Rocky Mountainstransition(withtheexceptionoftheAlberta-Southsite)werecomposedentirelyof individualsclassifiedaslucifugus.Thecarissimacladewasrestrictedtomountainousbutnot coastalregionsofthewestandaportionofthewesternplains(insouthernAlberta).ThealascensiscladewasfoundalongthePacificCoastandcoastalmountainranges( Fig1a ).However, individualsinthelucifuguscladewerealsosampledatlocationswestoftheGreatPlains-Rocky Mountainstransition,includingtheBritishColumbia-South(BC-S),Wyoming(WY),and Washington(WA)samples.TheAlberta-South(AB-S)sampleeastoftheGreatPlains-Rocky Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 11/23

PAGE 12

Mountainstransitionalsocontainedbothlucifugusandcarissimahaplotypes(seealso[ 33 ]). NoalascensishaplotypeswerefoundattheBC-Ssite,althoughrangemapsplacethissitein thatsubspecies ’ range;furthermore,theAlaska(AK)sampleclusteredwiththealascensis clade,eventhoughitfellwithinthedescribedrangeofthelucifugusclade.SpatialpatternsofpopulationgeneticstructureMitochondrialDNA. AMOVAanalysisconsideringallsamplesasasinglegrouprevealed highlevelsofdifferentiation( FST=0.721).Weiterativelygroupedsiteswithlowpairwise FSTvaluestodeterminethebestarrangementsofsitesthatmaximized ‘ among-group ’ andminimized ‘ among-site/within-group ’ variationintheAMOVAframework.Mostsampleswere highlydivergentfromallothers(76%ofpairwisecomparisonshad FST> 0.2,and62% were > 0.5; S5Table and Fig1b ),butweidentifiedtwogroupsofsites(oneinthecentral UnitedStatesandCanadaeastoftheRockyMountains,andoneineasternNorthAmerica) withinwhichdivergencewaslow( FST=-0.019 – 0.130; Fig1b ).Aftergroupingthesesites togetherinanAMOVAanalysis,among-groupvariation( FCT)accountedfor73.5%ofvariationinhaplotypefrequencies,andamong-site/within-groupvariationaccountedfor1.4%. MicrosatelliteDNA. Allclusteringmethodsemployed[Bayesianclustering(STRUCTURE),repeatedreallocation(FLOCK),and K meansclustering(GENODIVE)]identified K =2asthemostlikelynumberofgeneticclusters,roughlycorrespondingtoclusterseastversuswestoftheGreatPlains-RockyMountainstransition,andnotcorrespondingtosubspecies affiliation(Figs 3 and 4 ).However,foursitesinthetransition,namelyBritishColumbia-North (BC-N),Alberta-North(AB-N),AB-S,andWYhadintermediateQ(STRUCTURE)orLLOD (FLOCK)valuesindicativeofadmixture,andtherewasacleargradationinthisregionbetween geneticclusters(Figs 3 and 4 ).Onaverage,theBC-NandAB-Nsiteshadhigherproportional membershipwiththeeasterncluster,andAB-SandWYsiteswiththewesterncluster,and thereforetheyweregroupedaccordinglyinallsubsequentanalyses.Totestthevalidityofthe mostlikelynumberofclustersidentifiedusingeachofthethreealgorithms,weperformedcontingencytableanalyseswiththenullhypothesisthatclusterassignmentswererandomwith respecttosamplingsites.Inallcaseswerejectedthenullhypothesis(STRUCTURE: G =521.5; FLOCK: G =422.6;GENODIVE: G =77.1;df=20and P < 0.001inallcases)andconcluded thatclusterswerevalid.However,allclustersidentifiedbythethreealgorithmsfailedtomeet HWEexpectations( P < 0.05inallcases).Acrossthefoursites(WA,BC-S,AB-S,andWY) containingbothcarissimaandlucifugusmtDNAhaplotypes(settingasidethe3alacensis-type individualsinWA),clustermembershipwasindependentofsubspeciesaffiliationfortheclustersdefinedbyFLOCK( 2=1.65, P =0.1990)andGENODIVE( 2=0.22, P =0.6390),but wasnotforSTRUCTURE( 2=5.65, P =0.0175). AMOVAanalysisofmicrosatellitegenotypesindicatedweakbutsignificantpopulation structure(global FST=0.0161, P < 0.001;proportionofvariationwithinsites=0.984).The groupingofsitesthatmaximizedamong-groupvariationandminimizedamong-site/withingroupvariationincludedagroupcontainingAKonly,awesterngroupofsamples(CaliforniaSiskiyou(CA-Si),WA,BC-S,Idaho(ID),andWY),andaneasterngroupcontainingallother samples(variationamonggroups=2.72%, P < 0.001;variationamongsiteswithin groups=0.41%, P < 0.001).Generally, FSTvaluesbetweenAlaskaandallothersiteswerehigh andsignificant(0.049 – 0.089; S6Table ). FSTvaluesbetweensitesinthewesternandeastern groupsrangedfrom0.004 – 0.045,valuesbetweensiteswithinthewesterngrouprangedfrom 0.002 – 0.018,andvaluesbetweensiteswithintheeasterngrouprangedfrom-0.005 – 0.013( S6 Table ).AnAMOVAgroupingindividualsbasedontheirsubspecificmembershipsimilarly indicatedweakbutsignificantpopulationstructure(global FST=0.022, P < 0.001,proportion Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 12/23

PAGE 13

ofvariationwithinsubspecies=0.978).Themaximumpairwise FSTwasbetweenalascensis andlucifugus(0.0270, P < 0.001),theminimumwasbetweenalascensisandcarissima(0.0141, P < 0.001),anddifferentiationbetweencarissimaandlucifuguswasintermediate(0.0200, P < 0.001). Geneticdistancemeasureswerehigheramongsiteswithinthewesterngroupthanamong siteswithintheeasterngroup,andpermutationtestsapproachedsignificanceatthe P < 0.05 level( FST:West:0.023,East:0.002, P =0.0610; G ”ST:West:0.220,East:0.016, P =0.057;Jost ’ s D:West:0.202,East:0.014, P =0.057).Theseresultswerelargelydrivenbytheinclusionofthe AKsiteinthewesterngroup,andifthissitewasremoved,geneticdistancemeasureswerestill consistentlyhigherwithinthewestbutnotsignificantlydifferentbetweengroups( FST:West: 0.007,East:0.002, P =0.361; G ”ST:West:0.088,East:0.016, P =0.273;Jost ’ sD:West:0.081, East:0.014, P =0.274). Isolationbydistance. TotestforisolationbydistanceweperformedManteltestsonthe logarithmofgeographicdistanceandstandardizedgeneticdistance[ FST/(1FST)].Therewere clearsignalsofIBDforbothmitochondrial( r =0.346, P < 0.0001; Fig5a )andmicrosatellite DNA( r =0.537, P < 0.0001; Fig5b )acrosstherangeoflittlebrownbats.Withinidentified clusters,therewasnosignalofIBDintheeast( r =-0.307, P =0.9925; Fig5c ),buttherewas withinthewest( r =0.913, P =0.0411; Fig5d ).TotestifIBDinthewestwasdisproportionately drivenbytheAlaskapopulation,were-rantheanalysiswiththatsiteremoved,andthesignal ofIDBremained( r =0.880, P =0.0069; S1Fig ). Toassessthevalidityofclustersgiventhepatternofisolationbydistance,weplottedgeographicandgeneticdistance( FST)basedonmicrosatellitesaccordingtoclustermembership (pointsidentifiedseparatelyforcomparisonswithinthesameclustervs.indifferentclusters; Fig6 ).Therewasnoclearseparationbetweensitepairsinthesamevs.differentclusters,and lowgeneticdistancesforagivengeographicdistancewereobservedregularlyforsitepairsin differentclusters.However,partialMantelteststoexaminetheassociationbetweenthematrix Fig3.Proportionalmembership(Q)of M lucifugus togeneticclustersfor K =2estimatedusingSTRUCTUREwithsamplinglocationasprior information. Eachbarisasingleindividual,sampledpopulationsaredelineatedbyblacklinesandareorderedbygeographicalsamplinglocationfromwest toeast.Colorsdistinguishgeneticclusters(blueforproportionalmembershipinthewesterncluster,orangeforproportionalmembershipintheea stern cluster). doi:10.1371/journal.pone.0128713.g003 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 13/23

PAGE 14

ofgeneticdistancesandamatrixofclustermembershipforthemicrosatellitedata,withthe matrixofgeographicaldistancesasacovariate,weresignificant( P < 0.05inallcases)regardlessofcodingscheme(see Methods )orgeneticdistancemeasure( FSTor G ”ST)used.DiscussionTheemergenceandspreadofWNShasdecimatedbatpopulationsinaffectedareasandraised thespecterofextinctionfor Myotislucifugus M septentrionalis ,andotherhighly-affectedbat species.Therefore,understandingpopulationconnectivityandpossiblebarrierstodisease transmissionisvitaltotheongoingmanagementandconservationofaffectedspeciesandthe developmentofmitigationstrategiestolimitdiseasespreadandassociatedmortality.Here, usingacombineddatasetofmtDNAsequencesandnuclearmicrosatellitegenotypesforlittle brownbatsfromacrosstheirrange,wedemonstrateconsiderablespatialvariationinpatterns offemaledispersalandsignificantgeneticvariationbetweensitesineasternversuswestern portionsoftherangeoflittlebrownbats.Whethertheobservedvariationisrepresentativeof discretegeneticclustersratherthanisolationbydistanceisdebatable(seebelow),butoverall,it isclearthatlevelsofnucleargeneticdifferentiationarelow,andthereisnoevidenceforany majorbarrierstonucleargeneflowacrosstherangeoflittlebrownbats.However,somekey Fig4.Meanlog-likelihooddifference(LLOD)betweentwogeneticclustersobtainedbyFLOCKalongaseriesofgeographicallyorderedsitesfrom westtoeast. Forpresentation,populationsintheGreatPlains-RockyMountainstransitionzone(BC-NAB-N,AB-S,andWY)areorderedbyLLODto demonstratethetransitionamongclusters. doi:10.1371/journal.pone.0128713.g004 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 14/23

PAGE 15

spatialpatternsemergefromouranalyses,namely(1)patternsofmtDNAdifferentiationare highlyvariable,withhigh FSTvaluesbetweenmostsamplepairs(includingbetweenallwestern samples,betweenwesternandeasternsamples,andbetweensomeeasternsamples),whilelow mitochondrialdifferentiationwasobservedwithintwogroupsofsamplesfoundincentraland easternregionsofNorthAmerica(showninAMOVAandpairwise FSTanalyses; Fig1b );(2) thesitefromAlaskaishighlydifferentiatedfromallothersinourstudy(showninAMOVA andpairwise FSTanalyses);and(3)westernsitesarecharacterizedbysignificantisolationby distancebasedonmicrosatellites,whilethoseintheeastarenot.Thesedataraisethepossibility thatthecurrentpatternsofspreadof Pd observedineasternNorthAmericamaynotapplyto theentirerangeofthelittlebrownbat,andthattheremaybebroad-scalespatialvariationin theriskofWNStransmissionandoccurrenceifthediseasecontinuestospreadwest. Thepresenceofisolationbydistanceisamajorconfoundingfactorwhenexamininglevels ofgeneticstructureamongpopulationsofwidespreadspeciessuchaslittlebrownbats,because clinalvariationmaybeinterpretedasthepresenceofdiscreteclustersevenintheabsenceof barrierstogeneflow[ 66 78 80 ].Allclusteringmethodsweemployedonourmicrosatellitedata identifiedthepresenceoftwogeneticclusters,roughlydividingsiteseastvs.westoftheGreat Plains-RockyMountainstransition.However,wealsoobservedastrongpatternofisolationby distance,indicatingthattheseobservedclustersmaybeanartifactofdispersallimitationand clinalvariationacrosstheverylargerangeofthisspecies.Additionalanalysestotestthe Fig5. Standardizedgeneticdistance[ FST/(1FST)]plottedagainstthelogarithmofgeographicdistanceincludingallsampledpopulationsformtDNA(a),and microsatellites(b),andfortheeastern(c)andwestern(d)populationclustersbasedonmicrosatellites. doi:10.1371/journal.pone.0128713.g005 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 15/23

PAGE 16

validityofclustersprovidedmixedresults.Ontheonehand,geneticdistancesbetweeneastern andwesternsiteswererelativelylow(thebestgroupingintheAMOVAexplainedonly2.7%of thevariationinmicrosatelliteallelefrequencies),aManteltestshowednoclearseparation betweensitepairsinthesameversusdifferentclusters( Fig6 ),andidentifiedclustersfailedto meetHWexpectations,suggestingthatclustersdidnotrepresentgeneticallypanmicticpopulations(cf.[ 66 ]).Ontheotherhand,partialManteltestscontrollingforgeographicdistance revealedasignificantcorrelationbetweenclustermembershipandgeneticdistance,suggesting thatwesternandeasternclustersweredifferentiateddespitethesignalofisolationbydistance (cf.[ 80 ]). Whatisclearfromthesedataisthatthereissignificantgeneticvariationamongsamples fromeasttowestacrosstherangeoflittlebrownbats(althoughwhethertheyrepresentdiscrete geneticclustersratherthanisolationbydistanceisdebatable),butlevelsofnucDNAgenetic differentiationwererelativelylow,andtherewasnounambiguousevidenceforanymajorbarrierstogeneflowthatmightseverelyrestrictthespreadofWNS.Geneflowintemperatebats ismediatedthroughthepermanentdispersalofindividualsandtheexchangeofgenesatmatingcongregationsduringswarmingandhibernation.Thelackofisolationbydistanceandlow levelsofnucDNAdifferentiationamongsitesineasternNorthAmericaisconcordantwiththe continuousspreadofWNSfromitsorigininNewYork,andindicatesthatgeneflowviamatinghasoccurredoverwidegeographicareas.Furthermore,thediseasehaspassed,oriscurrentlypassing,throughregionsinwhichtherearelowlevelsofmtDNAdifferentiationamong Fig6.Pairwisegenetic( FST)basedonmicrosatellitesandgeographicdistancevalueshighlightingpopulationpairswithinthesamecluster(blue dots)andindifferentclusters(orangedots). doi:10.1371/journal.pone.0128713.g006 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 16/23

PAGE 17

sites(onegroupintheeasternUnitedStatesandonegroupinthecentralUnitedStatesand Canadianprovinces; Fig1b ).Mosttemperatebatsarecharacterizedbyrelativelyhighlevelsof femalephilopatryandmale-biaseddispersal,resultinginsignificantmatrilinealgeneticstructuringofpopulations(e.g.[ 35 86 – 88 ]).However,ourdatasuggestthattheexchangeoffemales amongpopulationsacrosslargeportionsoftherangeoflittlebrownbatsisanon-trivialsource ofgeneflowthatmaybecontributingtothespreadofWNS,andareconsistentwithsimilar inferencesoffemaledispersalamongpopulationsoversmallerspatialscalesinlittlebrownbats [ 6 33 35 42 ]andotherbatspecies[ 38 89 ]basedonmtDNA.Inaddition,thesedataareconsistentwithbandingdatashowingextensivemovementsbyindividuallittlebrownbatsofboth sexesoverhundredsofkilometersbetweensummerroosts,swarmingsites,andhibernacula withinandbetweenyearsincentralCanada[ 27 42 ].Bandingstudiesoflittlebrownbatsin otherpartsoftheirrangewouldhelptoresolvewhethertheobserveddifferencesarebetter explainedbyhistoricaldemographyorcurrentgeneflow.ThehighlevelsofobservedmtDNA differentiationinotherportionsoftherange,particularlyinthewest,suggestimportantspatial variationinfemaledispersalpatterns,andhighlighttheneedtoconsiderpermanentmovementsofbothmalesandfemalesandincorporateregionalvariationindispersalratesanddistancesinmodelsofWNStransmissiondynamics. GiventhelackofmajorphysiographicbarrierseastoftheRockyMountainsandthehigh levelsofgeneflowweinferred,itislikelythatWNSwillcontinuetoexpanditsrangeacross easternNorthAmerica.CurrentmodelsofdiseasespreadindicatethatWNSexhibitscharacteristicsofanexpandingepizooticwhereinrelativelydistantsiteshavelowerinfectionrisk,but overtimeinfectionratesincreaseandtheeffectofdistancediminishesasthedisease ‘ fillsin ’ behindtheexpansionfront[ 19 20 ].However,thesemodelsarebasedondatafromeasternportionsoftherangewheregeneflowishigh.Evenwithinthisregiontherateofspreadmaybe restrictedbyconsistentspatialvariationinabove-groundorcavemicroclimatesthatmaylimit thesurvivalandgrowthof Pd (cf.[ 20 21 ]),orbyspatialvariationintopographyorlanduse thatlimitsmovementsanddispersalbybats[ 6 ].InPennsylvania,forexample,topographical featuressuchastheAppalachianhighplateauandtheAlleghenyFrontescarpmentmayhave influencedseasonalmigrationpatternsoffemalebats,therebylimitingmatrilinealgeneflow anddiseasetransmissionratesamongpopulations[ 6 ].Indeed,twogeneticallydistinctpopulationsofwinteringcolonieswereobservedoneithersideoftheAlleghenyFront[ 6 ];hibernating coloniesoflittlebrownbatslocatedonthewesternAppalachianhighplateauwereinfected withWNS1 – 2yearslaterthancoloniesinthecentralmountainousandeasternlowland regionsofthestate[ 16 ].Thus,topographicorclimaticvariationmayslowthespreadofWNS throughsomeareasbylimitingpopulationconnectivityofthehostorthesurvivalandgrowth of Pd ,andmayexplainsomeoftheobservedspatialvariationtodateintherateanddirection ofWNSspreadthrougheasternNorthAmerica. Ourgeneticdataindicatinglowerlevelsofpopulationconnectivityinthewestsuggestthat ifWNSreacheswesternpopulations,therateofdiseasespreadmaydecline.ThehighmtDNA FSTvaluesamongwesternpopulations( S5Table )indicatethatfemalemovementsarehighly restrictedrelativetoeasternpopulations.OveralllevelsofnucDNAgeneflowamongwestern siteswerereducedrelativetotheeast,andwesternsiteswerecharacterizedbyisolationbydistancebasedonmicrosatelliteswhileeasternsiteswerenot.Theseresultsmay,inpart,berelated tothegreatertopographicalandecologicalheterogeneityinthewest,whichincludesmultiple mountainranges,plateaus,basins,andcoastallowlands,andwhichhasbeenimplicatedin recurrentphylogeographicpatternsinawidevarietyofothertaxa[ 90 ].Hibernationbehavior ispoorlycharacterizedforwesternNorthAmericanlittlebrownbatpopulations.Allknown largehibernatingpopulations( > 10,000individuals)aredescribedfromeasternNorthAmerica,andidentifiedhibernaculainthemountainouswesttypicallyhavelowercensussizesthan Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 17/23

PAGE 18

manyhibernaculaintheeast.Thehighphysiographicvariationinthemountainouswestmay limitpopulationconnectivityandthescaleofbatmovements,andthehighdensityofmines andcavesinmanyregionsinthewestmayresultinsmallerandmorediffusehibernatingcoloniesrelativetoeasternNorthAmerica.Comparativedataonconnectivitybetweensummer andwintersites(asin[ 27 ])areurgentlyrequiredtoquantifyspatialandtemporalpatternsof movementoflittlebrownbatsinthewesternportionoftheirrangeandtopredictpotential ratesofWNStransmission.Further,themostdistantpopulationwesampled(inAlaska)was byfarthemostdivergentfromallotherpopulations,andwerequiremuchmoredensesamplinginthewesternportionoftherangeoflittlebrownbatstodetermineifanyotherpopulationsareequallyormoreisolatedandhencemayhavereducedcontactrateswithother regionalpopulations. Thespatialvariationinpopulationconnectivityweobservedwaslargelyindependentof subspecificaffiliation.Phylogeneticanalysisrevealedthepresenceofthreedivergentlineages basedonmtDNA(correspondingtopreviouslydefinedsubspecies M l alascensis M l carissima and M l lucifugus ;asin[ 32 ]),withthenotablefindingofmultiplelineagesatthesame samplinglocationsinsouthernBritishColumbia,southernAlberta(asin[ 33 ]),andWyoming. AlthoughCarstensandDewey[ 32 ]providedsomesupportfrommtDNAandnuclearintrons fordiscreteevolutionarylineageswithin M lucifugus ,wefoundthatclustermembershipbased onmicrosatelliteswasindependentofsubspecificaffiliation,andweestimatedlowlevelsof nucDNAdifferentiationamongsubspecies( FST=0.022inAMOVAanalysis).Ourobserved discrepancybetweenmtDNAandnucDNAsignalsmaybedueinparttohomoplasy,particularlyatrapidly-evolvingmicrosatelliteloci.Alternatively,incompletelineagesortingmightbe producingdiscordantpatternsamongloci,particularlyforaspecieswitharelativelylarge effectivepopulationsize( Ne 400,000basedonCarstensandDewey ’ s[ 32 ]estimateof for M lucifugus )andrelativelyrecentdivergence(divergencefromwestern Myotis sp.approximately1 – 1.5Mya[ 32 ]).Athirdalternativeisthatourfindingoflowerdifferentiationat nucDNAcomparedtomtDNAisthatpatternsatmtDNAmayreflectgeneticdifferentiation thatevolvedandwerereinforcedaspopulationsuseddistinctglacialrefuges(asin[ 34 ]),while patternsatnucDNAreflectsecondarycontactparticularlymediatedviamalegeneflow. Ourdatashowextensivespatialvariationinlevelsofconnectivityamonglittlebrownbat populationsandprovidevaluableinformationforunderstandingpastandfuturepatternsof WNSspread.However,theuseofgeneticmethodstoinferpatternsoftransmissionassumes thatpatternsofgeneflowareindicativeofthemovementofinfectiousindividuals,andwe mustrecognizethattheriskofdiseasetransmissionmaybehigherthangeneticdatamayindicatebecausetheremaybemorecontactsamonginfectedandsusceptibleindividuals,including amongmembersofmultiplespecies,thanjustthosethatleadtogeneflow.Urgentresearchis requiredtodeterminehowandwhenindividualbatsmaybeexposedto Pd spores,andhow contactsofvaryingdurationsandseasonaltimingsinfluencetheriskofWNStransmission. Ultimatelyweneedtolearnwhetherbriefcontactsduringmatingcanresultintransferof sporesleadingtoinfectionorwhetherpermanentdispersalsaredrivingtransmission.Theusefulnessofourgeneticdataonlittlebrownbatsalsorestsontheassumptionthatintraspecific transmissiondynamicsoutweightheimpactofcross-speciestransmission,giventhatmultiple sympatricbatspeciesareaffectedbyWNS.Thisassumptionmaybejustified,aspost-WNS populationdeclinesofaffectedbatspeciesarenotinfluencedbypopulationsizesofother affected,cohabitingbatspecies[ 21 ],butresearchisrequiredtoassesshowoftencross-species transmissionmaytakeplaceandhowtherateofintroductionofinfectivepropagulestoenvironmentalreservoirsisinfluencedbymultiplespeciescohabitingthesamehibernaculum. Inconclusion,thisstudyidentifiedhighlevelsofgeneticvariationamongpopulationsoflittlebrownbatsacrosstheirrange,andmitochondrialDNAsequencesrevealedconsiderable Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 18/23

PAGE 19

spatialvariationinpatternsoffemaledispersal.Overalllevelsofnucleargeneticdifferentiation among M lucifugus populationsarelow,andwedidnotidentifyanymajorbarrierstogene flowacrosstheirrange.However,levelsofgeneticdifferentiationatbothmtDNAandmicrosatellitesaresignificantlyhigheramongpopulationstothewestoftheGreatPlains-Rocky Mountainstransition,suggestingthatthecurrentpatternofspreadofWNSandriskoftransmissionof Pd observedineasternNorthAmericamaynotapplytotheentirerangeofthelittle brownbat.SupportingInformationS1Fig.Standardizedgeneticdistance[ FST/(1 – FST)]formicrosatellitesplottedagainstthe logarithmofgeographicdistanceforthewesternpopulationclusterwiththeAlaskapopulationremoved. (DOCX) S1File.Datasourcesandpermissionsusedtoconstruct Fig1 (PDF) S1Table.Listof Myotislucifugus specimensincludedinanalyses. (DOCX) S2Table.Locusinformationfor11microsatellitesusedtoamplify M lucifugus (DOCX) S3Table.Listofspecimens(allinthegenus Myotis ),yearofcollection,samplinglocalities, andGenbankaccessionnumbersforCOIsequencesusedinphylogeneticanalysis. (DOCX) S4Table.Diversityofmicrosatelliteloci,includingobserved( HO)andexpectedheterozygosity( HE),numberofalleles( NA),allelicrichness( AR )andtheinbreedingcoefficient ( FIS). (DOCX) S5Table.Pairwise STvaluesamongpopulationsbasedonmtDNACOIsequences. (DOCX) S6Table.Pairwise FST(lowerdiagonal)andJost ’ s D (upperdiagonal)basedonnucDNA microsatellites. (DOCX)AcknowledgmentsWearegratefultonumerousfieldresearchersforprovidingsamples:E.Adams,E.Britzke,C. Butchkoski,R.Christophersen,M.Delorme,C.Dobony,E.Gates,K.Getz,J.Gruver,M.Gumbert,D.Hobson,J.Johnson,H.Kaarakka,C.Lausen,S.Loeb,F.Martinez,H.Niederriter,G. Nordquist,D.Pierson,W.Rainey,D.Redell,D.Solick,C.Stihler,K.Stroebel,M.Valent,J.Van deVenter,andC.Willis.Wealsothankthefollowingmuseumsforprovidingtissuesamples: AngeloStateNaturalHistoryCollection,CarnegieMuseumofNaturalHistory,Centrode InvestigacionesBiolgicasdelNoroesteS.C.,MuseumofVertebrateZoologyatBerkeley, NationalMuseumofNaturalHistory,andtheUniversityofAlaskaMuseumoftheNorth.We thankE.ClareforgenerouslysharingunpublishedmtDNAsequences.WealsothankJ.Rosensternforassistancewithlabwork,J.Glatzforhelpwithproducingmaps,andS.Gillforprovidingvaluablecommentsonearlierversionsofthismanuscript. Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 19/23

PAGE 20

AuthorContributionsConceivedanddesignedtheexperiments:MJVALRCMM.Performedtheexperiments:MJV ALRCMM.Analyzedthedata:MJVALRCMM.Contributedreagents/materials/analysis tools:MJVALRCMM.Wrotethepaper:MJVALRCMM.References1. BiekR,RealLA(2010)Thelandscapegeneticsofinfectiousdiseaseemergenceandspread.MolEcol 19:3515 – 3531.doi: 10.1111/j.1365-294X.2010.04679.x PMID: 20618897 2. RealLA,BiekR(2007)Spatialdynamicsandgeneticsofinfectiousdiseasesonheterogeneouslandscapes.JRSocInterface4:935 – 948.PMID: 17490941 3. CullinghamCI,MerrillEH,PybusMJ,BollingerTK,WilsonGA,ColtmanDW(2010)Broadandfinescalegeneticanalysisofwhite-taileddeerpopulations:estimatingtherelativeriskofchronicwasting diseasespread.EvolAppl4:116 – 131.doi: 10.1111/j.1752-4571.2010.00142.x PMID: 25567957 4. VollmerSA,BormaneA,DinnisRE,SeeligF,DobsonADM,AanensenDM,etal.(2011)Hostmigration impactsonthephylogeographyofLymeBorreliosisspirochaetespeciesinEurope.EnvironMicrobiol 13:184 – 192.doi: 10.1111/j.1462-2920.2010.02319.x PMID: 20722696 5. GrayRR,SalemiM(2012)Integrativemolecularphylogeographyinthecontextofinfectiousdiseases onthehuman-animalinterface.Parasitology139:1939 – 1951.doi: 10.1017/S0031182012001102 PMID: 22931895 6. Miller-ButterworthCM,VonhofMJ,RosensternJ,TurnerG,RussellA(2014)Geneticstructureoflittle brownbats( Myotislucifugus )correspondswithspreadofwhite-nosesyndromeamonghibernacula.J Hered105:354 – 364.doi: 10.1093/jhered/esu012 PMID: 24591103 7. BlehertDS,HicksAC,BehrM,MeteyerCU,Berlowski-ZierBM,BucklesEL,etal.(2009)Batwhitenosesyndrome:Anemergingfungalpathogen?Science323:227.doi: 10.1126/science.1163874 PMID: 18974316 8. LorchJM,MeteyerCU,BehrMJ,BoylesJG,CryanPM,HicksAC,etal.(2011)Experimentalinfection ofbatswith Geomycesdestructans causeswhite-nosesyndrome.Nature480:376 – 378.doi: 10.1038/ nature10590 PMID: 22031324 9. TurnerGG,ReederDM,ColemanJTH(2011)Afive-yearassessmentofmortalityandgeographic spreadofWhite-noseSyndromeinNorthAmericanbatsandalooktothefuture.BatResNews52: 13 – 27. 10. MeteyerCU,BucklesEL,BlehertDS,HicksAC,GreenDE,Shearn-BochslerV,etal.(2009)Histopathologiccriteriatoconfirmwhite-nosesyndromeinbats.JVetDiagnInvest21:411 – 414.PMID: 19564488 11. CryanPM,MeteyerCU,BoylesJG,BlehertDS(2010)Wingpathologyofwhite-nosesyndromeinbats suggestslife-threateningdisruptionofphysiology.BMCBiol8:135.doi: 10.1186/1741-7007-8-135 PMID: 21070683 12. WillisCKR,MenziesAK,BoylesJG,WojciechowskiMS(2011)Evaporativewaterlossisaplausible explanationformortalityofbatsfromwhite-nosesyndrome.IntegrCompBiol51:364 – 373.doi: 10. 1093/icb/icr076 PMID: 21742778 13. CryanPM,MeteyerCU,BlehertDS,LorchJM,ReederDM,TurnerGG,etal.(2013)Electrolytedepletioninwhite-nosesyndromebats.JWildlDis49:398 – 402.doi: 10.7589/2012-04-121 PMID:23568916 14. ReederDM,FrankCL,TurnerGG,MeteyerCU,KurtaA,BritzkeER,etal.(2012)Frequentarousal fromhibernationlinkedtoseverityofinfectionandmortalityinbatswithwhite-nosesyndrome.PLoS One7:e38920.doi: 10.1371/journal.pone.0038920 PMID: 22745688 15. VerantML,MeteyerCU,SpeakmanJR,CryanPM,LorchJM,BlehertDS(2014)White-nosesyndrome initiatesacascadeofphysiologicdisturbancesinthehibernatingbathost.BMCPhysiology14:e10. 16. U.S.FishandWildlifeService(2014)CurrentWNSOccurrenceMap.MapbyC.Butchkoski.Available: http://whitenosesyndrome.org/about/where-is-it-now 17. U.S.FishandWildlifeService(2012)NorthAmericanbatdeathtollexceeds5.5millionfromwhite-nose syndrome.U.S.FishandWildlifeServiceNewsRelease. 18. FrickWF,PollockJF,HicksAC,LangwigKE,ReynoldsDS,TurnerGG,etal.(2010)AnemergingdiseasecausesregionalpopulationcollapseofacommonNorthAmericanbatspecies.Science329: 679 – 682.doi: 10.1126/science.1188594 PMID: 20689016 19. WilderAP,FrickWF,LangwigKE,KunzTH(2011)Riskfactorsassociatedwithmortalityfromwhitenosesyndromeamonghibernatingbatcolonies.BiologyLett7:950 – 953. Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 20/23

PAGE 21

20. ThogmartinWE,KingRA,SzymanskiJA,PruittL(2012)Space-timemodelsforapanzooticinbats, withafocusontheendangeredIndianabat.JWildlDis48:876 – 887.doi: 10.7589/2011-06-176 PMID: 23060489 21. LangwigKE,FrickWF,BriedJT,HicksAC,KunzTH,KilpatrickAM(2012)Sociality,density-dependenceandmicroclimatesdeterminethepersistenceofpopulationssufferingfromanovelfungaldisease,white-nosesyndrome.EcolLett15:1050 – 1057.doi: 10.1111/j.1461-0248.2012.01829.x PMID: 22747672 22. FentonMB,BarclayRMR(1980)Myotislucifugus.MammSpecies142:1 – 8. 23. ReidFA(2006)MammalsofNorthAmerica.Boston:HoughtonMifflinCompany. 24. DavisWH,HitchcockHB(1965)Biologyandmigrationofthebat, Myotislucifugus ,inNewEngland.J Mammal46:296 – 313. 25. ThomasDW,FentonMB,BarclayRMR(1979)Socialbehaviorofthelittlebrownbat, Myotislucifugus I.Matingbehavior.BehavEcolSociobiol6:129 – 136. 26. RiversNM,ButlinRK,AltringhamJD(2006)AutumnswarmingbehaviourofNatterer'sbatsintheUK: Populationsize,catchmentareaanddispersal.BiolConserv127:215 – 226. 27. NorquayKJO,Martinez-Nu ezF,DuboisJE,MonsonKM,WillisCKR(2013)Long-distancemovementsoflittlebrownbats( Myotislucifugus ).JMammal94:506 – 515. 28. KerthG,KieferA,TrappmannC,WeishaarM(2003)Highgenediversityatswarmingsitessuggesthot spotsforgeneflowintheendangeredBechstein'sbat.ConservGenet4:491 – 499. 29. VeithM,BeerN,KieferA,JohannesenJ,SeitzA(2004)Theroleofswarmingsitesformaintaining geneflowinthebrownlong-earedbat( Plecotusauritus ).Heredity93:342 – 349.PMID: 15241447 30. FurmankiewiczJ,AltringhamJ(2007)Geneticstructureinaswarmingbrownlong-earedbat( Plecotus auritus )population:evidenceformatingatswarmingsites.ConservGenet8:913 – 923. 31. WilsonDE,ReederDM(2005)Mammalspeciesoftheworld:Ataxonomicandgeographicreference, thirdedition.Baltimore:JohnsHopkinsUniversityPress. 32. CarstensBC,DeweyTA(2010)Speciesdelimitationusingacombinedcoalescentandinformation-theoreticapproach:AnexamplefromNorthAmerican Myotis bats.SystematicBiol59:400 – 414.doi: 10. 1093/sysbio/syq024 PMID: 20547777 33. LausenCL,DelisleI,BarclayRMR,StrobeckC(2008)BeyondmtDNA:nucleargeneflowsuggeststaxonomicoversplittinginthelittlebrownbat( Myotislucifugus ).CanJZool86:700 – 713. 34. DixonMD(2011)Post-Pleistocenerangeexpansionoftherecentlyimperiledeasternlittlebrownbat ( Myotislucifuguslucifugus )fromasinglesouthernrefugium.EcolEvol1:191 – 200.doi: 10.1002/ece3. 20 PMID:22393495 35. DixonMD(2011)PopulationgeneticstructureandnatalphilopatryinthewidespreadNorthAmerican bat Myotislucifugus .JMammal92:1343 – 1351. 36. BilginR,KaratasA,CoramanE,MoralesJC(2008)Themitochondrialandnucleargeneticstructureof Myotiscapaccinii (Chiroptera:Vespertilionidae)intheEurasiantransition,anditstaxonomicimplications.ZoolScr37:253 – 262. 37. KerthG,PetrovB,ContiA,AnastasovD,WeishaarM,GazaryanS,etal.(2008)Communallybreeding Bechstein'sbatshaveastablesocialsystemthatisindependentfromthepostglacialhistoryandlocationofthepopulations.MolEcol17:2368 – 2381.doi: 10.1111/j.1365-294X.2008.03768.x PMID: 18429964 38. VonhofMJ,StrobeckC,FentonMB(2008)Geneticvariationandpopulationstructureinbigbrownbats ( Eptesicusfuscus ):isfemaledispersalimportant?JMammal89:1411 – 1420. 39. BryjaJ,KanuchP,FornuskovaA,BartonickaT,RehakZ(2009)Lowpopulationgeneticstructuringof twocrypticbatspeciessuggeststheirmigratorybehaviourincontinentalEurope.BiolJLinnSoc96: 103 – 114. 40. LackJB,WilkinsonJE,vandenBusscheRA(2010)Range-widepopulationgeneticstructureofthepallidbat( Antrozouspallidus ) — incongruentresultsfromnuclearandmitochondrialDNA.ActaChiropt12: 401 – 413. 41. TurmelleAS,KunzTH,SorensonMD(2011)Ataleoftwogenomes:contrastingpatternsofphylogeographicstructureinawidelydistributedbat.MolEcol20:357 – 375.doi: 10.1111/j.1365-294X.2010. 04947.x PMID: 21143331 42. BurnsLE,FrasierTR,BrodersHG(2014)Geneticconnectivityamongswarmingsitesinthewiderangingandrecentlydeclininglittlebrownbat( Myotislucifugus ).EcolEvol4:4130 – 4149.doi: 10.1002/ ece3.1266 PMID: 25505539 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 21/23

PAGE 22

43. GibbsHL,DawsonRJG,HobsonKA(2000)LimiteddifferentiationinmicrosatelliteDNAvariation amongnorthernpopulationsoftheyellowwarbler:evidenceformale-biasedgeneflow?MolEcol9: 2137 – 2147.PMID: 11123625 44. KimuraM,CleggSM,LovetteIJ,HolderKR,GirmanDJ,MilB,etal.(2002)Phylogeographical approachestoassessingdemographicconnectivitybetweenbreedingandoverwinteringregionsina Nearctic-Neotropicalwarbler( Wilsoniapusilla ).MolEcol11:1605 – 1616.PMID: 12207712 45. JonesKL,KrapuGL,BrandtDA,AshleyMV(2005)Populationgeneticstructureinmigratorysandhill cranesandtheroleofPleistoceneglaciations.MolEcol14:2645 – 2657.PMID: 16029467 46. IrwinDE,IrwinJH,SmithTB(2011)GeneticvariationandseasonalmigratoryconnectivityinWilson's warblers( Wilsoniapusilla ):species-leveldifferencesinnuclearDNAbetweenwesternandeastern populations.MolEcol20:3102 – 3115.doi: 10.1111/j.1365-294X.2011.05159.x PMID: 21689190 47. WorthingtonWilmerJ,BarrattE(1996)Anon-lethalmethodoftissuesamplingforgeneticstudiesof chiropterans.BatResNews37:1 – 3. 48. HebertPDN,CywinskaA,BallSL,DeWaardJR(2003)BiologicalidentificationsthroughDNAbarcodes.ProcRSocLondSerB-BiolSci270:313 – 321. 49. IvanovaNV,DewaardJR,HebertPDN(2006)Aninexpensive,automation-friendlyprotocolforrecoveringhigh-qualityDNA.MolEcolNotes6:998 – 1002. 50. KatohK,StandleyDM(2013)MAFFTMultipleSequenceAlignmentSoftwareVersion7:Improvements inperformanceandusability.MolBiolEvol30:772 – 780.doi: 10.1093/molbev/mst010 PMID: 23329690 51. Oyler-McCanceSJ,FikeJA(2011)Characterizationofsmallmicrosatellitelociisolatedinendangered Indianabat( Myotissodalis )foruseinnon-invasivesampling.ConservGenetResour3:243 – 245. 52. TrujilloRG,AmelonSK(2009)Developmentofmicrosatellitemarkersin Myotissodalis andcross-speciesamplificationin M grisescens M leibii M lucifugus ,and M septentrionalis .ConservGenet10: 1965 – 1968. 53. CastellaV,ReudiM,ExcoffierL,IbanezC,ArlettazR,HausserJ(2000)IstheGibraltarStraitabarrier togeneflowforthebat Myotismyotis (Chiroptera:Vespertilionidae)?MolEcol9:1761 – 1772.PMID: 11091312 54. PiaggioAJ,FigueroaJA,PerkinsSL(2009)Developmentandcharacterizationof15polymorphic microsatellitelociisolatedfromRafinesque'sbig-earedbat, Corynorhinusrafinesquii .MolEcolResour 9:1191 –1193.doi: 10.1111/j.1755-0998.2009.02625.x PMID: 21564872 55. PiaggioAJ,MillerKEG,MatocqMD,PerkinsSL(2009)Eightpolymorphicmicrosatellitelocideveloped andcharacterizedfromTownsend'sbig-earedbat, Corynorhinustownsendii .MolEcolResour9:258 – 260.doi: 10.1111/j.1755-0998.2008.02243.x PMID: 21564620 56. LibradoP,RozasJ(2009)DnaSPv5:AsoftwareforcomprehensiveanalysisofDNApolymorphism data.Bioinformatics25:1451 – 1452.doi: 10.1093/bioinformatics/btp187 PMID: 19346325 57. VillesenP(2007)FaBox:anonlinetoolboxforfastasequences.MolEcolNotes7:965 – 968. 58. ExcoffierL,LavalG,SchneiderS(2005)Arlequinver.3.0:Anintegratedsoftwarepackageforpopulationgeneticsdataanalysis.EvolBioinformOnline1:47 – 50. 59. ExcoffierL,SmousePE,QuattroJM(1992)AnalysisofmolecularvarianceinferredfrommetricdistancesamongDNAhaplotypes:applicationtohumanmitochondrialDNArestrictiondata.Genetics 131:479 – 491.PMID: 1644282 60. GuindonS,DufayardJ-F,LefortV,AnisimovaM,HordijkW,GascuelO(2010)Newalgorithmsand methodstoestimatemaximum-likelihoodphylogenies:AssessingtheperformanceofPhyML3.0.Syst Biol59:307 – 321.doi: 10.1093/sysbio/syq010 PMID: 20525638 61. StadelmannB,LinL-K,KunzTH,RuediM(2007)MolecularphylogenyofNewWorld Myotis (Chiroptera,Vespertilionidae)inferredfrommitochondrialandnuclearDNAgenes.MolPhylogentEvol43: 32 – 48. 62. TamuraK,PetersonD,PetersonN,StecherG,NeiM,KumarS(2011)MEGA5:Molecularevolutionary geneticsanalysisusingmaximumlikelihood,evolutionarydistance,andmaximumparsimonymethods. MolBiolEvol28:2731 – 2739.doi: 10.1093/molbev/msr121 PMID: 21546353 63. GoudetJ(1995)FSTAT(Version1.2):AcomputerprogramtocalculateF-statistics.JHered86:485 – 486. 64. KalinowskiST(2005)HP-RARE1.0:acomputerprogramforperformingrarefactiononmeasuresof allelicrichness.MolEcolNotes5:187 – 189. 65. MeirmansPG(2012)AMOVA-basedclusteringofpopulationgeneticdata.JHered103:744 – 750.doi: 10.1093/jhered/ess047 PMID: 22896561 66. GuillotG,LebloisR,CoulonA,FrantzAC(2009)Statisticalmethodsinspatialgenetics.MolEcol18: 4734 – 4756.doi: 10.1111/j.1365-294X.2009.04410.x PMID: 19878454 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 22/23

PAGE 23

67. PritchardJK,StephensM,DonnellyP(2000)Inferenceofpopulationstructureusingmultilocusgenotypedata.Genetics155:945 – 959.PMID: 10835412 68. FalushD,StephensM,PritchardJK(2003)Inferenceofpopulationstructureusingmultilocusgenotype data:linkedlociandcorrelatedallelefrequencies.Genetics164:1567 – 1587.PMID: 12930761 69. HubiszMJ,FalushD,StephensM,PritchardJK(2009)Inferringweakpopulationstructurewiththe assistanceofsamplegroupinformation.MolEcolResour9:1322 – 1332.doi: 10.1111/j.1755-0998. 2009.02591.x PMID: 21564903 70. EvannoG,RegnautS,GoudetJ(2005)DetectingthenumberofclustersofindividualsusingthesoftwareSTRUCTURE:asimulationstudy.MolEcol14:2611 – 2620.PMID: 15969739 71. EarlDA,vonHoldtBM(2012)STRUCTUREHARVESTER:awebsiteandprogramforvisualizing STRUCTUREoutputandimplementingtheEvannomethod.ConservGenetResour4:359 – 361. 72. JakobssonM,RosenbergNA(2007)CLUMPP:aclustermatchingandpermutationprogramfordealing withlabelswitchingandmultimodalityinanalysisofpopulationstructure.Bioinformatics23:1801 – 1806.PMID: 17485429 73. DuchesneP,TurgeonJ(2012)FLOCKprovidesreliablesolutionstothe"numberofpopulations"problem.JHered103:734 – 743.doi: 10.1093/jhered/ess038 PMID: 22615162 74. DuchesneP,TurgeonJ(2009)FLOCK:amethodforquickmappingofadmixturewithoutsourcesamples.MolEcolResour9:1333 – 1344.doi: 10.1111/j.1755-0998.2009.02571.x PMID: 21564904 75. WeirBS,CockerhamCC(1984)Estimating F -statisticsfortheanalysisofpopulationstructure.Evolution38:1358 – 1370. 76. JostL(2008)GSTanditsrelativesdonotmeasuredifferentiation.MolEcol17:4015 – 4026.PMID: 19238703 77. MeirmansPG,HedrickPW(2011)Assessingpopulationstructure: FSTandrelatedmeasures.MolEcol Resour11:5 – 18.doi: 10.1111/j.1755-0998.2010.02927.x PMID: 21429096 78. FrantzAC,CellinaS,KrierA,SchleyL,BurkeT(2009)UsingspatialBayesianmethodstodetermine thegeneticstructureofacontinuouslydistributedpopulation:clustersorisolationbydistance?JAppl Ecol46:493 – 505. 79. SchwartzM,McKelveyK(2009)Whysamplingschemematters:theeffectofsamplingschemeon landscapegeneticresults.ConservGenet10:441 – 452. 80. MeirmansPG(2012)Thetroublewithisolationbydistance.MolEcol21:2839 – 2846.doi: 10.1111/j. 1365-294X.2012.05578.x PMID: 22574758 81. RoussetF(1997)Geneticdifferentiationandestimationofgeneflowfrom F -statisticsunderisolationby distance.Genetics145:1219 – 1228.PMID: 9093870 82. JensenJL,BohonakAJ,KelleyST(2005)Isolationbydistance,webservice.BMCGenet6:13.PMID: 15760479 83. McRaeB,BeierP,HuynhL,DeWaldL,KeimP(2005)Habitatbarrierslimitgeneflowandilluminate historicaleventsinawiderangingcarnivore,theAmericanpuma.MolEcol14:1965 – 1977.PMID: 15910319 84. FontaineMC,BairdSJE,PiryS,RayN,TolleyKA,DukeS,etal.(2007)Riseofoceanographicbarriers incontinuouspopulationsofacetacean:thegeneticstructureofharborporpoisesinoldworldwaters. BMCBiology5:30.PMID: 17651495 85. RosenbergMS,AndersonCD(2011)PASSaGE:PatternAnalysis,SpatialStatisticsandGeographic Exegesis.Version2.MethodEcolEvol2:229 – 232. 86. CastellaV,RuediM,ExcoffierL(2001)Contrastedpatternsofmitochondrialandnuclearstructure amongnurserycoloniesofthebat Myotismyotis .JEvolutionBiol14:708 – 720. 87. KerthG,MayerF,PetitE(2002)Extremesex-biaseddispersalinthecommunallybreeding,nonmigratoryBechstein'sbat( Myotisbechsteinii ).MolEcol11:1491 – 1498.PMID: 12144668 88. MoussyC,HoskenDJ,MathewsF,SmithGC,AegerterJN,BearhopS(2013)Migrationanddispersal patternsofbatsandtheirinfluenceongeneticstructure.MammalRev43:183 – 195. 89. BogdanowiczW,Lesi skiG,Sadkowska-TodysM,GajewskaM,RutkowskiR(2013)Population geneticsandbatrabies:acasestudyof Eptesicusserotinus inPoland.ActaChiropt15:35 – 56. 90. ShaferABA,CullinghamCI,CoteSD,ColtmanDW(2010)Ofglaciersandrefugia:adecadeofstudy shedsnewlightonthephylogeographyofnorthwesternNorthAmerica.MolEcol19:4589 – 4621.doi: 10.1111/j.1365-294X.2010.04828.x PMID: 20849561 Range-WidePopulationGeneticAnalysisofLittleBrownBats PLOSONE|DOI:10.1371/journal.pone.0128713July8,2015 23/23