Citation
Environment and host species shape the skin microbiome of captive neotropical bats

Material Information

Title:
Environment and host species shape the skin microbiome of captive neotropical bats
Series Title:
PeerJ
Creator:
Lemieux-Labonté, Virginie
Tromas, Nicolas
Shapiro, B. Jesse
Lapointe, François-Joseph
Publication Date:
Language:
English
Physical Description:
1 online resource

Subjects

Subjects / Keywords:
Bats ( lcsh )
White-nose syndrome ( lcsh )
Microbiology ( lcsh )
Genre:
serial ( sobekcm )

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
K26-05071 ( USFLDC DOI )
K26.5071 ( USFLDC handle )

USFLDC Membership

Aggregations:
Karst Information Portal

Postcard Information

Format:
Serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

Submitted 2May2016 Accepted 10August2016 Published 20September2016 Correspondingauthor VirginieLemieuxLabont,virginie.lemieuxlabonte@umontreal.ca Academiceditor KimberlyBishop-Lilly AdditionalInformationand Declarationscanbefoundon page13 DOI 10.7717/peerj.2430 Copyright 2016Lemieux-Labontetal. Distributedunder CreativeCommonsCC-BY4.0 OPENACCESS Environmentandhostspeciesshapethe skinmicrobiomeofcaptiveneotropical bats VirginieLemieux-Labont,NicolasTromas,B.JesseShapiroand Fran cois-JosephLapointe DpartementdeSciencesBiologiques,UniversitdeMontral,Montral,Canada ABSTRACT Background .Awiderangeofmicroorganismsinhabitanimalskin.Thismicrobial communitymicrobiomeplaysanimportantroleinhostdefenseagainstpathogens anddisease.BatsChiroptera:Mammaliaareanecologicallyandevolutionarilydiversifiedgroupwitharelativelyunexploredskinmicrobiome.Thebatskinmicrobiome couldplayaroleindiseaseresistance,forexample,towhitenosesyndromeWNS, aninfectionwhichhasbeendevastatingNorthAmericanbatpopulations.However, fundamentalknowledgeofthebatskinmicrobiomeisneededbeforeunderstanding itsroleinhealthanddiseaseresistance.Captiveneotropicalfrugivorousbats Artibeus jamaicensis and Carolliaperspicillata provideasimplecontrolledsysteminwhichto characterizethefactorsshapingthebatmicrobiome.Here,weaimedtodeterminethe relativeimportanceofhabitatandhostspeciesonthebatskinmicrobiome. Methods .Weperformedhigh-throughput16SrRNAgenesequencingoftheskin microbiomeoftwodifferentbatspecieslivingincaptivityintwodifferenthabitats. Inthefirsthabitat, A.jamaicensis and C.perspicillata livedtogether,whilethesecond habitatcontainedonly A.jamaicensis. Results .Wefoundthatbothhabitatandhostspeciesshapethecompositionanddiversityoftheskinmicrobiome,withhabitathavingthestrongestinfluence.Cohabitating A.jamaicensis and C.perspicillata sharedmoresimilarskinmicrobiomesthanmembers ofthesamespecies A.jamaicensis acrosstwohabitats. Discussion .Theseresultssuggestthatincaptivity,theskinmicrobialcommunityis homogenisedbythesharedenvironmentsandindividualproximitiesofbatsliving togetherinthesamehabitat,attheexpenseoftheinnatehostspeciesfactors.The predominantinfluenceofhabitatsuggeststhatenvironmentalmicroorganismsor pathogensmightcolonizebatskin.Wealsoproposethatbatpopulationscoulddiffer inpathogensusceptibilitydependingontheirimmediateenvironmentandhabitat. Subjects Ecology,Microbiology,Zoology Keywords Batskinmicrobiome,Habitatmicrobiomeinteraction,Hostmicrobiomeinteraction INTRODUCTION Animalskinisanecosysteminhabitedbyhighlyvariableandcomplexcommunitiesof microorganisms Grice&Segre,2011 ,whichcanbedividedintoresidentandtransient floraacquiredfromtheenvironment Roth&James,1988 .Ahealthyskinmicrobiome contributestohostfitnessbyoccupyingpathogenadhesionsitesandproducingpathogen Howtocitethisarticle Lemieux-Labontetal.,Environmentandhostspeciesshapetheskinmicrobiomeofcaptiveneotropical bats. PeerJ4:e2430;DOI10.7717/peerj.2430

PAGE 2

inhibitors Roth&James,1988 ; Grice&Segre,2011 .Forexample,thesalamanderskin associatedbacteria Janthinobacteriumlividum confersresistancetothedevastatingfungal pathogen Batrachochytriumdendrobatidis Bruckeretal.,2008 ,possiblyexplainingwhy somesalamanderpopulationsdeclinewhileothersdonot.Competitiveinteractions betweenbeneficialandpathogenicskinmicrobesmaythereforeplayaroleinpreventing diseaseinwildanimals Belden&Harris,2007 .Despiteitsimportance,littleisknowabout thefactorsshapingtheskinmicrobiomeofwildanimalspecies. Exogenousenvironmentalandendogenoushost-specificfactorssuchpH,sebum production,temperatureandmoisture Grice&Segre,2011 areknowntoshapetheskin microbiome,buttherelativeinfluenceoftheseparametersdifferbetweenstudies.Inthegut microbiome,hosttaxaandphylogenyappearstohaveagreatereffectthantheenvironment ontheassemblageofbacterialcommunities Ochmanetal.,2010 ; Roeselersetal.,2011 ; Tzengetal.,2015 .Inprimates, Moelleretal. concludedthatsympatricmembersof differentspeciesi.e.,gorillasandchimpanzeessharingthesamehabitatharboramore similargutmicrobiomethanallopatricmembersofthesamespecies.Inneotropicalbats, gutmicrobiomeshavebeenproposedtobeinfluencedbyacomplexinteractionbetween exogenousandendogenousfactors Phillipsetal.,2012 ,withanimportantroleforhost taxaandphylogeny Phillipsetal.,2012 ; Carrillo-Araujoetal.,2015 Duetoitsdirectexposuretotheenvironment,theskinmicrobiomeissuspectedtobe muchmoredynamicthanthegutmicrobiome Romano-Bertrandetal.,2015 .Hence,the roleoftheenvironmentisexpectedtobestronginshapingtheskinmicrobiome.Studies oftheskinmicrobiomeofwildpopulationsofamphibianssuggestthathostspeciesdoes playamajorrolebecausetheskinmicrobiomesofcohabitatingspecieswerefoundto besignificantlydifferent McKenzieetal.,2012 ; Kuenemanetal.,2014 ; Walkeetal.,2014 However, Kuenemanetal. identifiedthehabitatasthesecondmostimportant parameterontheskinmicrobiomeofamphibianspecies.Indeed,theenvironmentshould actasabacterialreservoirfortheskinmicrobiomeandhostspeciesmaybeabletoselect particulartaxa Loudonetal.,2014 ; Walkeetal.,2014 .Consequently,itisthoughtthat hostandenvironmentalfactorsinteractcloselytoshapetheamphibianskinmicrobiome. Inspiteofrecentinvestigationsontheskinmicrobiomeofvariousanimalspecies,few studieshaveanalyzedtherelativeinfluenceofendogenoushostandexogenousenvironmentalfactorsontheskinmicrobiomeinnon-humanmammals.Inhumans,different variablesareexpectedtoshapetheskinmicrobiome,suchasbodysite,age,genderand habitat Fiereretal.,2008 ; Giacomoni,Mammone&Teri,2009 ; Yingetal.,2015 .Humans werefoundtosharemoresimilarmicrobiomeswiththeirdogsthanwithdogsfrom differenthouseholds Songetal.,2013 .Therefore,asharedhabitatmighthomogenize skinmicrobiomesacrossindividualsandevenacrossspecies. Incontrasttothegutmicrobiome,theskinmicrobialcommunityappearstobemuch moreinfluencedbyexposuretotheenvironment,includingenvironmentalmicrobes andabioticfactors Chengetal.,2015 .Yet,understandingthecomplexityoftheskin microbiomeclearlysuffersfromalackofstudiesacrossdifferentmammals.BatsChiroptera:Mammaliaarepartofoneofthemostecologicallyandevolutionarilydiversified mammalianorders Kunz&Fenton,2003 ; Voigt&Kingston,2016 ,providinganexcellent Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 2/19

PAGE 3

modeltostudyhowmicrobiomesvaryacrossrelatedhostspecies.Astheonlyflyingmammals,withacave-dwellinglifestyle,theskinmicrobiomeofbatsisprobablyuniqueamong allvertebrates.Additionally,theskinmicrobiomeprovidesapossibledefenseagainstwhite nosesyndromeWNS,askindiseasecausedbythefungus Pseudogymnoascusdestructans Pd Gargasetal.,2009 ; Lorchetal.,2011 ,formerlyknownas Geomycesdestructans ; Minnis&Lindner,2013 andresponsibleforthedeathofover6millionNorthAmerican bats USFish&WildlifeService,2012 .Notallbatspeciesareequallyaffectedbythedisease Turner,Reeder&Coleman,2011 ,suggestingthatbothhostgenetics,ecology,andmicrobiomesmightplayarolein Pd resistance.Inthelightofthisknowledge,basicinformation aboutfundamentalsourcesofvariationsinthebatskinmicrobiomeisbadlyneeded toharnessthepossibleimplicationsofthismicrobialcommunityinadiseasecontext. Jamaicanfruitbats Artibeusjamaicensis andSeba'sshort-tailledbats Carollia perspicillata provideconvenientanimalmodelstostudytheskinmicrobiomeof chiropterans.ThesespeciesofneotropicalbatsfamilyPhyllostomidaearewidely distributedinCentralandSouthAmerica.Theyshareagregariouslifestyle,witha polygynousharemsocialorganisationbasedonmaledefenseoftheroostingsites wherefemalesaggregate Porter,1978 ; Williams,1986 ; Ortega&Arita,1999 .Inthewild, bothspeciesroostinhollowtreesandcaves Morrison,1979 ; Williams,1986 ; Cloutier& Thomas,1992 ; Ortega&Arita,1999 ,where A.jamaicensis normallyaggregateinsmall groups<12individualsorverylargecolonies>500bats Arita&Vargas,1995 ,and C.perspicillata aggregateingroupsof10tomorethan100individuals Cloutier&Thomas, 1992 .Thesespeciesaregeneralistfrugivores Cloutier&Thomas,1992 ; Ortega&CastroArellano,2001 .Theyareeasilymaintainedincaptivity,wheretheycanevenbreed.The gutmicrobiomeofbothspecieshaverecentlybeencharacterized Carrillo-Araujoetal., 2015 ,whiletheskinmicrobialcommunitystillremainsunknownatthisdate. Here,westudiedhowhabitatfactorsincludingcolonyparametersanddietandhost speciescontributedtothestructureof A.jamaicensis and C.perspicillata skinmicrobiome understableenvironmentalconditionsi.e.,incaptivity.Althoughsuchtropicalspeciesare notaffectedorendangeredbythewhitenosesyndrome,theyprovideausefulmodeltostudy thefactorsthatshapetheskinmicrobiomeandmightprovideresistancetopathogens.The skinmicrobiomesofwildandcaptiveorganismsarecertainlydifferent Beckeretal.,2014 ; Loudonetal.,2014 ; Chengetal.,2015 ,butstudyingthemincaptivityispractical,allowing ustolimitenvironmentalfluctuationsthatmightobscuretheeffectsofhostspeciesin naturalsetting.Weusedhigh-throughput16Sampliconsequencingtoassessthetaxonomic compositionanddiversityoftheskinmicrobiomeofthesetwospeciesofbatssampledin twodifferentzooshabitats.Thisdesignallowedustocomparedbatgroupslivinginshared vs.separateenvironments.Theobjectivesweretoquantifythecontributionsofhabitat andhostspeciesinshapingthebatskinmicrobiome.Ourresultsshowasignificanteffect ofbothhabitatandhostspeciesontheskinmicrobiome,withhabitatplayingadominant role.Thisstudythusprovidesaninitialviewofwhatfactorsshapetheskinmicrobiomeof neotropicalbats.Thesefindingsprovidebasicknowledgeoftheskinmicrobiome,whichcan ultimatelybeappliedtothemanagementandconservationofthreatenedbatpopulations. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 3/19

PAGE 4

MATERIALSANDMETHODS Sampling Inthisstudy,42adultspecimensfromtwodifferentfrugivorousbatspeciesweresampled intwodifferentzoos.Namely,10 A.jamaicensis and12 C.perspicillata individualsall malesweresampledinDecember2014fromtheMontralBiodmeCanada,wherethey livetogetherinanartificialcavemaintainedatatemperatureof22 Cduringwinter,and 26 Cinsummer.Inaddition,20 A.jamaicensis individualsmalesand14femaleswere alsosampledinMarch2015fromtheGranbyZooCanada.Thesebatsalsoliveinan artificialcave,wherethetemperatureismaintainedat26 Callyearlong.Bothcoloniesof batswereestablishedin1992,withactualpopulationsizesof95individualsattheMontral Biodme A.jamaicensis and50 C.perspicillata ,and247individualsattheGranbyZoo. Skinmicrobiomesampleswereobtainedfromeachspecimenbyswabbingthebackand onewingfor30swithasterileWhatmanOmniswabFisherScientificsoakedinNaCl 0.15M.SwabstipswereejectedintoMobioPowersoilDNAisolationKittubesMoBio Laboratories,whichwerethenstoredat )]TJ/F61 10.9091 Tf 8.509 0 Td [(20 CuntilDNAextraction.Asanegative control,ahumidifiedsterileswabwasalsocollectedateachsamplingsite.Handlingof animalsattheGranbyZooaswellastheMontralBiodmewasapprovedbythelocal ethicscommitteesComitOprationsenConservationetRecherche,andBiodme's WelfareAnimalandEthicscommittee. DNAextraction,amplicationandsequencing BacterialgenomicDNAwasextractedfromeachswabusingtheMoBioPowersoilDNA isolationKitaccordingtothemanufacturer'sprotocols.Amplificationandsequencingwere thenperformedaspreviouslydescribed Preheimetal.,2013a .Librarieswereprepared usingatwo-stepPCR.ThefirstPCRamplifiesthehypervariableregionV4ofthe16S smallsubunitribosomalgenewithforwardprimerU515_f:ACACGACGCTCTTCCGAT CTYRYRGTGCCAGCMGCCGCGGTAAandreverseprimerE786_R:CGGCATTCCTG CTGAACCGCTCTTCCGATCTGGACTACHVGGGTWTCTAAT Caporasoetal.,2011 2 m lofextractedDNAequivalentDNAamountbysamplewasaddedtothePCR reactioncontaining14.25 m lofsterilewater,5 m lHFbuffer,0.5 m lDNTPs,0.25 m l PhusionHigh-FidelityDNAPolymeraseNewEnglandBiolabsinc.,and1.5 m lofforward andreverseprimers.AmplificationswereperformedwithaMastercycernexusGSX1 Eppendorfunderthefollowingconditions:initialdenaturationat98 Cfor30s;30 cyclesalternating98 Cfor25s,40sat54 C,35sat72 C,andfinalelongationstepfor oneminuteat72 C.Negativecontrolswereincludedintheamplificationsteptoaccount forpossiblecontamination.Eachsamplewasamplifiedinquadruplicateandpooledto limitpossiblePCRartefacts.AllPCRproductswerethenpurifiedwithZYMODNAClean &Concentrator TM -5ZYMORESEARCHfollowingthemanufacturer'sprotocol.The secondPCRstepconsistedofaddingprimerscontainingabarcodeindexandIllumina adaptersequencestoeachDNAamplicon.Todoso,4 m lofthefirststepamplification productwasaddedtoaPCRreactioncontaining10.25 m lofsterilewater,5 m lHFbuffer, 0.5 m lDNTPs,0.25 m lPhusionHigh-FidelityDNAPolymeraseand2.5 m lofforward primerPE-III-PCR-F:AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTAC Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 4/19

PAGE 5

ACGACGCTCTTCCGATCTandreverseprimerPE-III-PCR-001-096:CAAGCAGA AGACGGCATACGAGATNNNNNNNNNCGGTCTCGGCATTCCTGCTGAACCGCT CTTCCGATCTNindicatingtheuniquebarcode Preheimetal.,2013b .Indexingwas performedunderthefollowingthermalconditions:initialdenaturationat98 Cfor 30s,7cyclesalternating98 Cfor30s,30sat83 C,andfinally30sat72 C.This secondamplificationwasperformedintriplicate.Sampleswerepooledandpurifiedwith thePCRpurificationAgencourtAMPureXPBeckmanCoulter.Qubit2.0Fluorometer InvitrogenwasusedtomeasureDNAconcentrationineachsample.Indexedsamples werethenpooledtoobtainafinalconcentrationrangebetween10and20ng/ m l.DNA wasnextdilutedanddenaturedaccordingtothemanufacturer'sprotocolforpaired-end sequencingusingMiSeqReagentKitv2cycles2 250bponMiSeqIllumina. Dataanalysis 2,105,588sequenceswereamplifiedfrom41ofthe42skinsamples.Onespecimenof A.jamaicensis fromGranbyZoowasremovedfromthedatasetduetofailureofsequencing. Ameanof51,356sequenceswasobtainedpersample,withaminimumof2,247anda maximumof243,588sequences.RawsequencedataandmetadataareavailableonFigshare atDOI:10.6084/m9.figshare.3206668andDOI:10.6084/m9.figshare.3428159. Preclustering,qualityfiltering,primerremoval,mergingofrawsequences,and postclusteringdereplicatingwereperformedwiththeSmileTrainscriptshttps://github. com/almlab/SmileTrain/wiki/for16SdataprocessingusingUSEARCHv.7.0.1090 http://www.drive5.com/usearch/ Edgar,2010 .Distribution-basedclustering Preheimet al.,2013b usingthedbOTUcalleralgorithmhttps://github.com/spacocha/dbOTUcaller wasperformedtoclustersequencesintoOperationalTaxonomicUnitsOTUsby consideringthedistributionofDNAsequencesacrosssamplesandsequencedistances. ThecorrespondingOTUtableprovidingrelativeabundancesofbacterialtaxainthe differentsampleswasassignedwithQIIMEversion1.8.http://qiime.org/ Caporasoet al.,2010 usingGreenGenesdatabaserelease13_5http://greengenes.lbl.gov DeSantiset al.,2006 seeTableS1.Forcompositionalanalysis,thegenus Halomonas Shewanella and Lactobacillus wereidentifiedascontaminationbecauseoftheirhighproportioninnegative controls.Thesetaxawereconsequentlyfilteredoutfromallsamplespriortofurtheranalysis. TheLinearDiscriminantAnalysisLDAsizeEffectLEfSealgorithmhttps: //huttenhower.sph.harvard.edu/galaxy/ Segataetal.,2011 wasusedtoidentifytaxa andOTUscontributingthemosttodifferencesbetweenhabitatsandhostspecies.LEfSe detectssignificantdifferencesintaxaandOTUabundancewiththenon-parametricfactorial Kruskal-Wallissumranktest Kruskal&Wallis,1952 .Then,acanonicalmethodisapplied toestimatelinearcombinationsofOTUsthatprovidethebestdiscriminationamongbat speciesorhabitats. Toinvestigatethediversityoftheskinmicrobialcommunityalphadiversity,Shannon Hill,1973 andBalancedWeightedPhylogeneticDiversityBWPD Barker,2002 ; Vellend etal.,2011 ; McCoy&MatsenIV,2013 indiceswerecomputedfrommultiplerarefieddata sets.MultiplerarefactionconsistsofarepeatedsubsamplingoftheOTUtable.Thisprocedureisgenerallyusedtoensureamoreconsistentcomparisonbetweensamplesinwhichthe Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 5/19

PAGE 6

numberofsequencesdiffers.Fiftyiterationsofthedeepestsequencingdepthforeachsample wereusedinalphadiversitycalculations.TheShannonindex,whichincludesbothOTU richnessandevenness,wascomputedduetoitsreducedsensitivitytosampledepth differences Haegemanetal.,2013 ; Preheimetal.,2013a .BWPDisadiversitymeasurethat usesphylogeneticinformationtoevaluatediversityofmicrobialcommunitywherespecies delimitationisdifficult.ContrarytothePhylogeneticDiversityPDmeasure,BWPD accountsforabundanceandisrobusttosamplingdepthdifferencesbetweensamples McCoy&MatsenIV,2013 .Rversion3.1.3http://www.r-project.org/ RDevelopement CoreTeam,2015 wasusedforallstatisticalanalyses.Alphadiversityresultswere comparedbetweenhabitatsandspeciesusingnon-parametricWilcoxonSigned-Ranktest Wilcoxon,1945 .The p -valueforalltestswasadjustedwithHolm'ssequentialBonferroni Holm,1979 .Phylogeneticdiversityindiceswerecalculatedusingaphylogenetictree constructedwithFastTree2.1.8http://meta.microbesonline.org/fasttree/ Price,Dehal& Arkin,2010 Betadiversitywascalculatedbetweenbatmicrobiomesgroupedaccordingtohabitat andhostspecies.Phylogeny-basedweightedUniFracdistances Lozupone&Knight,2005 ; Lozuponeetal.,2007 andthesquarerootofJensenShannondivergenceJSD 1/2 Fuglede &Topse,2004 werecalculatedonunrarefieddataaspreviouslysuggested McMurdie& Holmes,2014 withthephyloseqpackagehttps://joey711.github.io/phyloseq/ McMurdie &Holmes,2013 .WeightedUniFrac,whichaccountsfordifferencesinabundance,is widelyusedtocomparedistancesbetweenmicrobialcommunities,althoughitissensitive todifferencesinsequencingdepthbetweensamples Lozuponeetal.,2011 .Toaddressthis problem,weusedrelativeOTUabundancestocalculateweightedUniFracdistances McMurdie&Holmes,2014 .JSDwasalsoselectedbecauseitisarobustmeasureofdivergence basedonthedistributionofrelativeabundancesbetweenmicrobialcommunities.Taking thesquarerootofJSDtransformsthismeasureintoaninterpretablemetric Preheimet al.,2013a .Allbetadiversityresultswerevisualizedwithnon-metricmultidimensional scalingNMDS Kruskal,1964 usingthephyloseq ordinate function. Totestforsignificantdifferencesamonggroupsofbats,weusedthepermutational multivariateanalysisofvariancePERMANOVA,ananalogofMANOVAforpartitioning distancematricesamongvarioussourcesofvariation Anderson,2001 .Thenullhypothesis ofthistestisthatthemetriccentroiddoesnotdifferbetweengroupsinourcase,hostspecies andhabitat Anderson&Walsh,2013 .PERMANOVAwascalculatedwiththe adonis functionintheveganpackagehttp://cran.r-project.org/package=vegan Oksanenetal., 2015 .Sincethistestissensitivetodatadispersionandmaythereforeconfusewithingroupvariationwithamong-groupvariation Anderson,2001 ,weperformedananalysis ofmultivariatehomogeneityPERMDISP Anderson,2006 withthe betadisper function totestifgroupsdifferedintheirdispersion.Thenullhypothesisofthistestisthatthe averagewithin-groupdispersionisthesameinallgroups Anderson&Walsh,2013 .In eachofthesetwotests,thenumberofpermutationswassetto9999.Forallanalyses,a p -valuethresholdof0.05wasconsideredsignificant. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 6/19

PAGE 7

Figure1Relativeabundancesofthesixdominantbacterialphylaintheskinmicrobiomeofcaptive neotropicalbats. ThecompletelistoftaxaisprovidedinTableS2. RESULTS Habitatandhostspeciesbothshapethecompositionofbatskin microbiomes Wefirstcharacterizedthetaxonomiccompositionofthebatskinmicrobiomebysequencing skinswabsfrom41captivebats.Weidentifiedfivedominantsharedphylaintheskin microbiomeofcaptivebatsFig.1:Actinobacteria%%,Proteobacteria% 36%,Firmicutes%%,Cyanobacteria%%,Bacteroidetes%%and Fusobacteria 1%.Attheorderlevel,LEfSeanalysisninetaxathatdifferedsignificantly byeitherhostspeciesorhabitatLDAscore 3.4, p < 0 : 05.Namely,fivetaxawere representativeof A.jamaicensis fromtheGranbyZoo,whereasoneandthreetaxawere respectivelyrepresentativeof A.jamaicensis and C.perspicillata fromtheBiodmeFig.2. Accordingtotheseresults, A.jamaicensis sampledfromtheGranbyZooappearstobethe mostdifferentgroupintermsofdifferentiallyabundanttaxa. Atfinertaxonomicresolution,aLEfSeanalysisattheOTUlevelalsorevealedthe importanceofhabitatinshapingtheskinmicrobiomecomposition.Weidentified924 OTUssignificantlyenrichedinaparticularhabitatFig.3Aalmosttwicethenumber ofOTUsenrichedinaparticularhostspeciesFig.3B.Theseresultssuggestthathabitat playsastrongerrolethanhostspeciesinshapingtheskinmicrobiome. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 7/19

PAGE 8

Figure2ResultsofLEfSeanalysisshowingthemaindifferencesamongbacterialordersintheskin microbiomeofcaptiveneotropicalbats. Significantresultsareidentifiedwithastar*.See`Methods' formoredetails.CpBiodme,Biodme C.perspicillata ;AjZoo,GranbyZoo A.jamaicensis ;AjBiodme, Biodme A.jamaicensis Figure3NumberofOTUsoperationaltaxonomicunitsenrichedindifferenthabitatsorhost species. ARepresentativeOTUsaccordingtohabitatBiodmeorGranbyZoo.BRepresentative OTUsaccordingtobatspecies A.jamaicencis or C.perspicillata .Theintersectionindicatesthenumbers ofOTUsthatdidnotdiffersignificantlybetweengroupsbyLEfSeanalysisMethods. Habitatisamajordeterminantofalphadiversity Wenextaskedwhetherthetotalamountofdiversityalphadiversityinthebatskin microbiomedifferedaccordingtohostspeciesorhabitatFig.4.BasedontheShannon indexofalphadiversity,wefoundthat A.jamaicensis fromBiodmeismostdiverse,and A.jamaicensis fromGranbyZooisleastdiverseFig.4A.Thus,thetwoBiodmespecies seemstoharboramorerichandevenskinmicrobiomecommunity.Shannondiversity betweenspecies A.jamaicensis and C.perspicillata isnotsignificantlydifferentFig.4B, Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 8/19

PAGE 9

Figure4Alphadiversitydifferssignificantlybyhabitat. AShannonindexcomparedacrossbat groups.BShannonindexcomparedacrossbatspecies.CShannonindexcomparedacrosshabitats. DBWPDindexphylogeneticmeasureacrossbatgroups.EBWPDindexacrossbatspecies. FBWPDindexacrosshabitats.Errorbarsrepresentstandarddeviations.AjBiodme,Biodme A.jamaicensis ;AjZoo,GranbyZoo A.jamaicensis ;CpBiodme,Biodme C.perspicillata .Non-parametric WilcoxonSigned-Ranktest p 0 : 05, p 0 : 01, p 0 : 001. WilcoxonSigned-Ranktest, V D 227, p D 0 : 134,whiletheBiodmebatsofbothspecies havesignificantlyhigherdiversitythanGranbyZoobatsFig.4C,WilcoxonSigned-Rank test, V D 400, p < 0 : 001. TheresultsbasedonShannondiversitywereconfirmedbyBWPD,anothermeasureof alphadiversity,whichaccountsforthephylogeneticrelatednessandrelativeabundanceof microbialtaxaFig.4D.AsobservedwithShannondiversity, A.jamaicensis fromBiodme hasthehighestalphadiversity,followedby C.perspicillata fromBiodmeand A.jamaicensis fromGranbyZooFig.4D.TheBWPDindexisnotsignificantlydifferentbetweenbats speciesFig.4E,WilcoxonSigned-Ranktest, V D 211, p D 0 : 300,whereassignificant differencesexistbetweenbatssampledfromtheBiodmeandGranbyZoohabitatsFig. 4F,WilcoxonSigned-Ranktest, V D 363, p < 0 : 001.ResultsoftheBWPDandShannon indexarethusconsistentandsuggesthabitatandpossiblyahabitat-speciesinteraction astheprincipalforcesshapingalphadiversityintheskinmicrobiomecommunity. Habitatandhostspeciesshapemicrobialcommunitybetadiversity Wenextusedbetadiversityanalysistoestimatetheeffectsofthehabitatandhostspeciesin shapingthecompositionoftheskinmicrobiome.Ourresultsshowthatsamplesareclusteredbybothhabitatandhostspecies,basedontwodifferentmetricsofbetadiversityFig. 5,andthatallsamplesareclearlydistinctfromnegativecontrolsFig.S1.UsingtheJSD 1/2 Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 9/19

PAGE 10

betadiversityindex,individualsofdifferentspeciesfromthesamehabitatBiodmeappear tobemorecloselyclusteredthanindividualsofthesamespecies A.jamaicensis from distincthabitatsFig.5A.TheweightedUniFracanalysisstilldiscriminatedthesamples byhabitatattheexpenseofhostspeciesFig.5B.OrdinationofonlyGranbyZoo,which includedbothmaleandfemale A.jamaicensis bats,doesnotshowanyclusteringaccording tosexdatanotshown.Howeverourlimitedsamplesizeonly6malespreventsusfrom drawingfirmconclusionontheinfluenceofsexonthebatskinmicrobiome.Sextherefore remainsapossibleconfoundingfactorofhabitat,becausetheBiodmecontainsallmale bats,wheretheGranbyzoowaspredominantlyfemale.Globally,betadiversityordinations thussuggestapredominantinfluenceofhabitatonskinmicrobiomebetadiversity. PERMANOVAanalysesalsostronglysupportedtheordinationresults.Habitat andhostspeciestogetherexplainedthemostvariationinJSD 1/2 .04%, adonis F D 16 : 212, p D 0 : 0001.Habitatappearedtobethemostimportantfactor,explaining 31.71%ofvariationinJSD 1/2 adonis F D 18 : 117, p D 0 : 0001.Infact,habitat explainedmorethantwicethevarianceexplainedbyhostspeciesfactors.95%, adonis F D 5 : 295, p D 0 : 0004.However,thePERMANOVAresultscouldbeaffected bynon-homogeneousdispersionofthedata.Indeed,the A.jamaicensis samples fromtheGranbyZooappearedtobelessdispersedinJSD 1/2 Fig.5A,andwe founddifferentlevelsofdispersionbyhabitat betadisper F D 33 : 4298, p D 0 : 0001 andbyhabitatandhostspeciescombined betadisper F D 7 : 0628, p D 0 : 0023, butnotforhostspeciesalonebetadisper, F D 3 : 5063, p D 0 : 0703.Nevertheless, theordinationclearlysupportsaclusteringpatternthatconfirmstheimportance ofhabitatandhostspeciesfactorscombined. RepeatingthesameanalysisonweightedUniFracdistancesyieldedsimilar PERMANOVAresultshabitatandhostspeciestogether:45.82%, adonis F D 16 : 066, p D 0 : 0001;habitat32.9%, adonis F D 19 : 1220, p D 0 : 0001;hostspecies:7.2%, adonis F D 3 : 0131, p D 0 : 0340.However,theUniFracdistancesdidnotvarysignificantlyindispersion byhabitat betadisper F D 1 : 6594, p D 0 : 2077orhabitatandhostspeciescombined betadisper F D 0 : 2244, p D 0 : 7989.Differentspeciesweredifferentlydispersed betadisper F D 8 : 0426, p D 0 : 0081.ResultsbasedonWeightedUniFracconfirmthathabitatandhost speciesareactingtogethertoshapetheskinmicrobiomeoftwospeciesofneotropicalbats livingindistincthabitats.Habitatappearstobethemajordriverofmicrobiomecommunity structure,withamoresubtlebutsignificantroleofhostspecies. DISCUSSION Theskinmicrobiomeisafirstlineofdefenseagainstpathogens Grice&Segre,2011 However,theskinmicrobiomeisstillpoorlyinvestigatedinmanyanimalseventhoughit couldplayaroleindiseaseoutcomes.Investigatingthefundamentalsourcesofvariationin theskinmicrobiomeisthereforeacriticalsteptowardunderstandingitsroleinhealthand diseases,andinhopeofeventuallydeployingmicrobiome-basedtherapiesagainstwildlife pathogens.Thisstudyexploredtherelativeinfluenceofendogenousandexogenousfactors i.e.,hostspeciesandhabitatinshapingtheskinmicrobiomeoftwospeciesoffrugivorous Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 10/19

PAGE 11

Figure5Microbiomesclustermainlybyhabitat,butalsobyhostspecies. ANon-metricmultidimentionalscalingofJSD 1/2 ofbatskinmicrobiomecomposition.Eachpointrepresentsasinglemicrobiome sample.2Dstress D 0.09.BNon-metricmultidimentionalscalingofweightedUniFracdistancesamong batskinmicrobiomes.2Dstress D 0.09. captivebats.However,theskinmicrobiomeinfluenceisexpectedtobedifferentinwild populationswithrespecttobatslivingincaptivity Beckeretal.,2014 ; Loudonetal.,2014 ; Chengetal.,2015 .Indeed,captivebatsarerestrictedtoalimitedarea,whereaswild individualsareusuallyexposedtosignificantenvironmentalvariationwhenforagingand roostinginnature.Theenvironmentalstabilityofthecavesinwhichcaptivebatsareliving couldinturnexplaintherelativeimportanceofthehabitatontheskinmicrobiomewith respecttohostspeciesinourresults. Thespeciesofbatsunderstudywerefoundtohaveskin-associatedmicrobial communitiessimilartothosecharacterizedinothermammalianorderssuchascarnivores Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 11/19

PAGE 12

dog,marsupialsTasmaniandevil,andprimateshuman Hoffmannetal.,2014 ; Chengetal.,2015 .Whereasthemostabundantbacterialphylai.e.,Actinobacteria, Proteobacteria,Firmicutes,Cyanobacteria,BacteroidetesandFusobacteriawere representedinallofthemammalsinvestigated,relativeabundancessometimesdiffered betweenspecies.Namely,Actinobacteriaisthemostabundantphyluminbatandhuman profilesalike, Grice&Segre and Ohetal. ,butonlythethirdmostabundant indogsandTasmaniandevils Hoffmannetal.,2014 ; Chengetal.,2015 .Suchdifferences couldbeexplainedbyhostspecies,samplingsiteontheskin,orhabitatvariation.Yet, captiveneotropicalbatswerealsofoundtoharborahigherproportionofCyanobacteria ontheirskin,aphylumalreadyidentifiedinthegutmicrobiomeofbats Phillipsetal., 2012 ,particularlyinwildpopulationsof C.perspicillata Carrillo-Araujoetal.,2015 .The presenceofthistaxonissuggestedtobeattributabletothecavehabitat,becausethistype ofmoistenvironmentisknowntobesuitablefortheestablishmentofCyanobacteriae.g., atthecaveentrancewherelightisavailable Albertano,2012 Wefoundapredominantinfluenceofhabitat,withaminorbutsignificantrole ofhostspeciesinshapingthemicrobiome.Specifically,thetwocohabitatingspecies, A.jamaicensis and C.perspicillata ,appearedtosharemoresimilarskinmicrobiomes,in termsofcompositionanddiversity,thanmemberofsamespecies A.jamaicensis from differenthabitats.Insuchgregariousanimals Porter,1978 ; Williams,1986 ; Ortega&Arita, 1999 ,individualsinhabitingasinglecavearepronetocontactwithoneanother,suchthat microbialtransferisfacilitatedincaptivity,bothdirectlyandindirectly.Inaddition,sharing thesamehabitatwithidenticalenvironmentalconditionslogicallyincursimilarconstraints ontheskinmicrobiomeofotherwisedifferenthostspecies.Contrarytopreviousstudies onamphibians,whichshowedsignificantdifferencesamongspeciescohabitinginthesame habitat McKenzieetal.,2012 ; Kuenemanetal.,2014 ; Walkeetal.,2014 ,habitatappearsto bethemainfactoractingontheskinmicrobiomeinbats,atleastincaptivity.Considering thathostspeciesandotherendogenicfactorsactincombinationwiththehabitattodrive theskinmicrobiomestructure,wesuggestthatbatpopulationscoulddifferindisease susceptibilitydependingontheirimmediateenvironment,aswellasthespeciesinvolved. Ourdefinitionofenvironmentalfactorshabitatsisbasedonacomparisonoftwo differentzoos,whichcoulddifferinotherconfoundingfactors.Someoftheseconfounding factorscanbeexcluded.Forexample,dietwasthesameinthetwozoos,andthebatcolonies wereestablishedatthesametimebothin1992.Therefore,dietandtimeincaptivitycan besafelyexcludedaspossibleconfoundersofenvironment.Otherfactorsdiddifferbetween zoos.Forexample,temperaturewas26 CallyearlongattheGranbyZooandrangedfrom 22intheBiodme.Therefore,weincludetemperatureaspartofwhatweconsidertobe ``environmentaleffects.''Thesexratioalsodifferedbetweenzoos,withonezooconsisting entirelyofmalesandtheotherprimarilyfemale.However,basedonbetadiversityanalyses, wefoundnoapparentdifferencesincommunitycompositionbetweensexes.Nevertheless, wecannotcompletelyexcludeenvironmentaleffectsbeingconfoundedwithsex.Sexis knowntoinfluencetheskinmicrobiomeinhumans Fiereretal.,2008 ; Yingetal.,2015 andithasbeenhypothesizedthatdifferencesinhormoneproductionandmetabolismmay affecttheskinmicrobiomecompositione.g.,pHandsebumproduction Giacomoni, Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 12/19

PAGE 13

Mammone&Teri,2009 .However,suchdifferencesarenotalwaysdetectable,especially forbodysitesthatdonotexhibitsexualdifferencese.g.,drysitesvs.moistsites Ohetal., 2012 .Inourstudy,batswereswabbedonthebackandthewingssiteswhichareunlikely toexperiencesexualdimorphism.Therefore,differencesbetweenzooslikelyrepresent environmentalfactorse.g.,differencesintemperature,humidityandenvironmental bacteria,althoughwecannotcompletelyexcludeaconfoundinginfluenceofsex. Thepredominanteffectoftheenvironmenthabitatinshapingtheskinmicrobiome ofbatsprovidesbothrisksandbenefitstothehostinthefaceofpathogens.Onthe negativeside,becausethebatskinmicrobiomevariesmoreaccordingtohabitat,itmight behighlysusceptibletoinvasionbypathogenssuchasthe Pd fungus.Onthepositiveside, theskinmicrobiomecouldbemanipulatedwithprobiotics.Thisobservationsuggests thatpreviouslyconsideredprobioticanti-fungalbacteriasuchas Pseudomonas Hoytet al.,2015 and Rhodococcusrhodochrous strainDAP96253 Cornelisonetal.,2014 could beintroduceddirectlyintobathabitatstoeasetheirimplantationinskincommunities. Probioticscouldthereforerepresentapromisingmanagementtoolagainstpathogenslike Pd .Ofcourse,suchmanagementtoolswouldhavetobevalidatedintherelevanthost speciesandenvironments. Ourinvestigationoftheskinmicrobiomeoftwoneotropicalspeciesofbatslivingin controlledhabitatsrevealedthecombinedinfluenceofendogenousandexogenousfactors. Theseresultsshowthatthecaptivebatskinmicrobiomeisshapedbothbyhabitatand hostspecies.Goingforward,itwillbeimportanttoextendourresultstoadditionalbat specieslivingincaptivityandtowildpopulationsofbats.Inbatsandothermammals,the skinmicrobiomehasthepotentialtobecomeanimportanttoolforpopulationhealth, conservationandmanagement. ACKNOWLEDGEMENTS TheauthorswouldliketothankPatrickPar,LouisLazure,AlexChandonnet,Nathalie MaroisfromGranbyZooandJacquesDancosse,Anne-MariePlante,EmikoWongfrom MontralBiodmefortheirassistanceinthesamplingprocedure,CatherineGirard,Ins LevadesandJulieMarleauforhelpwithexperiments,YvesTerratforhelpwithdata analysis,andthreeanonymousreviewersfortheircommentsonanearlierversionofthis manuscript. ADDITIONALINFORMATIONANDDECLARATIONS Funding ThisresearchwasfundedbyanNSERCdiscoverygrantsOGP0155251toFJLandBJSwas supportedbyaCanadaResearchChair.Thefundershadnoroleinstudydesign,data collectionandanalysis,decisiontopublish,orpreparationofthemanuscript. GrantDisclosures Thefollowinggrantinformationwasdisclosedbytheauthors: NSERCdiscoverygrants:OGP0155251. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 13/19

PAGE 14

CompetingInterests Theauthorsdeclaretherearenocompetinginterests. AuthorContributions VirginieLemieux-Labontconceivedanddesignedtheexperiments,performedthe experiments,analyzedthedata,contributedreagents/materials/analysistools,wrotethe paper,preparedfiguresand/ortables,revieweddraftsofthepaper. NicolasTromasconceivedanddesignedtheexperiments,analyzedthedata,contributed reagents/materials/analysistools,revieweddraftsofthepaper. B.JesseShapirocontributedreagents/materials/analysistools,revieweddraftsofthe paper. Fran cois-JosephLapointeconceivedanddesignedtheexperiments,contributed reagents/materials/analysistools,revieweddraftsofthepaper. AnimalEthics Thefollowinginformationwassuppliedrelatingtoethicalapprovalsi.e.,approvingbody andanyreferencenumbers: WelfareAnimalandEthic'scommitteeofBiodmedeMontralandGranbyZoo institutionapprovedthisbyletter. DataAvailability Thefollowinginformationwassuppliedregardingdataavailability: FigshareSequencingdataDOI:10.6084/m9.figshare.3428159;Figshare MetadataDOI:10.6084/m9.figshare.3206668. SupplementalInformation Supplementalinformationforthisarticlecanbefoundonlineathttp://dx.doi.org/10.7717/ peerj.2430#supplemental-information. REFERENCES AlbertanoP.2012. Cyanobacterialbiofilmsinmonumentsandcaves.In:WhittonBA, ed. EcologyofcyanobacteriaII .Heidelberg:Springer,317. AndersonMJ.2001. Anewmethodfornon-parametricmultivariateanalysisofvariance. AustralEcology 26 :32DOI10.1111/j.1442-9993.2001.01070.pp.x. AndersonMJ.2006. Distance-basedtestsforhomogeneityofmultivariatedispersions. Biometrics 62 :245DOI10.1111/j.1541-0420.2005.00440.x. AndersonMJ,WalshDCI.2013. PERMANOVA,ANOSIM,andtheManteltestinthe faceofheterogeneousdispersions:whatnullhypothesisareyoutesting? Ecological Monographs 83 :557DOI10.1890/12-2010.1. AritaHT,VargasJA.1995. Naturalhistory,interspecificassociation,andincidenceofthe cavebatsofYucatan,Mexico. TheSouthwesternNaturalist 40 :29. BarkerGM.2002. Phylogeneticdiversity:aquantitativeframeworkformeasurement ofpriorityandachievementinbiodiversityconservation. BiologicalJournalofthe LinneanSociety 76 :165DOI10.1111/j.1095-8312.2002.tb02081.x. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 14/19

PAGE 15

BeckerMH,Richards-ZawackiCL,GratwickeB,BeldenLK.2014. Theeffectofcaptivity onthecutaneousbacterialcommunityofthecriticallyendangeredPanamanian goldenfrog Atelopuszeteki BiologicalConservation 176 :199 DOI10.1016/j.biocon.2014.05.029. BeldenLK,HarrisRN.2007. Infectiousdiseasesinwildlife:thecommunityecology context. FrontiersinEcologyandtheEnvironment 5 :533DOI10.1890/060122. BruckerRM,HarrisRN,SchwantesCR,GallaherTN,FlahertyDC,LamBA,Minbiole KPC.2008. Amphibianchemicaldefense:antifungalmetabolitesofthemicrosymbiont Janthinobacteriumlividum onthesalamander Plethodoncinereus Journalof ChemicalEcology 34 :1422DOI10.1007/s10886-008-9555-7. CaporasoJG,KuczynskiJ,StombaughJ,BittingerK,BushmanFD,CostelloEK,Fierer N,GonzlezPeaA,GoodrichJK,GordonJI,HuttleyGA,KelleyST,KnightsD, KoenigJE,LeyRE,LozuponeCA,McdonaldD,MueggeBD,PirrungM,ReederJ, SevinskyJR,TurnbaughPJ,WaltersWA,WidmannJ,YatsunenkoT,ZaneveldJ, KnightR.2010. QIIMEallowsanalysisofhigh-throughputcommunitysequencing data. NatureMethods 7 :335DOI10.1038/nmeth.f.303. CaporasoJG,LauberCL,WaltersWA,Berg-LyonsD,LozuponeCA,Turnbaugh PJ,FiererN,KnightR.2011. Globalpatternsof16SrRNAdiversityatadepthof millionsofsequencespersample. ProceedingsoftheNationalAcademyofSciencesof theUnitedStatesofAmerica 108 :4516DOI10.1073/pnas.1000080107. Carrillo-AraujoM,Ta sN,Alcntara-HernndezRJ,GaonaO,SchondubeJE,Medelln RA,JanssonJK,FalcnLI.2015. Phyllostomidbatmicrobiomecomposition isassociatedtohostphylogenyandfeedingstrategies. FrontiersinMicrobiology 6 :Article447DOI10.3389/fmicb.2015.00447. ChengY,FoxS,PembertonD,HoggC,PapenfussAT,BelovK.2015. TheTasmanian devilmicrobiomeimplicationsforconservationandmanagement. Microbiome 3 :Article76DOI10.1186/s40168-015-0143-0. CloutierD,ThomasDW.1992. Carolliaperspicillata MammalianSpecies 417 :1 DOI10.2307/3504157. CornelisonCT,KeelMK,GabrielKT,BarlamentCK,TuckerTA,PierceGE,CrowSA. 2014. Apreliminaryreportonthecontact-independentantagonismof Pseudogymnoascusdestructans by Rhodococcusrhodochrous strainDAP96253. BMCMicrobiology 14 :246DOI10.1186/s12866-014-0246-y. DeSantisTZ,HugenholtzP,LarsenN,RojasM,BrodieEL,KellerK,HuberT,Dalevi D,HuP,AndersenGL.2006. Greengenes,achimera-checked16SrRNAgene databaseandworkbenchcompatiblewithARB. AppliedandEnvironmentalMicrobiology 72 :5069DOI10.1128/AEM.03006-05. EdgarRC.2010. SearchandclusteringordersofmagnitudefasterthanBLAST. Bioinformatics 26 :2460DOI10.1093/bioinformatics/btq461. FiererN,HamadyM,LauberCL,KnightR.2008. Theinfluenceofsex,handedness, andwashingonthediversityofhandsurfacebacteria. ProceedingsoftheNational AcademyofSciencesoftheUnitedStatesofAmerica 105 :17994 DOI10.1073/pnas.0807920105. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 15/19

PAGE 16

FugledeB,TopseF.2004. JensenShannondivergenceandHilbertspaceembedding. In: Proceedingsofthe2004IEEEinternationalsymposiumoninformationtheory ISIT ,31. GargasA,TrestMT,ChristensenM,VolkTJ,BlehertDS.2009. Geomycesdestructans sp.nov.associatedwithbatwhite-nosesyndrome. Mycotaxon 108 :147 DOI10.5248/108.147. GiacomoniP,MammoneT,TeriM.2009. Gender-linkeddifferencesinhumanskin. JournalofDermatologicalScience 55 :144DOI10.1016/j.jdermsci.2009.06.001. GriceEA,SegreJA.2011. Theskinmicrobiome. NatureReviewsMicrobiology 9 :244 DOI10.1038/nrmicro2537. HaegemanB,HamelinJ,MoriartyJ,NealP,DushoffJ,WeitzJS.2013. Robustestimationofmicrobialdiversityintheoryandinpractice. TheISMEJournal 7 :1092 DOI10.1038/ismej.2013.10. HillMO.1973. Diversityandevenness:aunifyingnotationanditsconsequences. Ecology 54 :427DOI10.2307/1934352. HoffmannAR,PattersonAP,DieselA,LawhonSD,LyHJ,StephensonCE,Mansell J,SteinerJM,DowdSE,OlivryT,SuchodolskiJS.2014. Theskinmicrobiomein healthyandallergicdogs. PLoSONE 9 :e83197DOI10.1371/journal.pone.0083197. HolmS.1979. Asimplesequentiallyrejectivemultipletestprocedure. Scandinavian JournalofStatistics 6 :65DOI10.2307/4615733. HoytJR,ChengTL,LangwigKE,HeeMM,FrickWF,KilpatrickAM.2015. Bacteria isolatedfrombatsinhibitthegrowthof Pseudogymnoascusdestructans ,thecausative agentofwhite-nosesyndrome. PLoSONE 10 :e0121329 DOI10.1371/journal.pone.0121329. KruskalJB.1964. Multidimensionalscalingbyoptimizinggoodnessoffittoanonmetric hypothesis. Psychometrika 29 :1DOI10.1007/BF02289565. KruskalWH,WallisWA.1952. Useofranksinone-criterionvarianceanalysis. Journalof theAmericanStatisticalAssociation 47 :583 DOI10.1080/01621459.1952.10483441. KuenemanJG,ParfreyLW,WoodhamsDC,ArcherHM,KnightR,McKenzieVJ.2014. Theamphibianskin-associatedmicrobiomeacrossspecies,spaceandlifehistory stages. MolecularEcology 23 :1238DOI10.1111/mec.12510. KunzTH,FentonMB.2003. Batecology .Chicago:UniversityofChicagoPress. LorchJM,MeteyerCU,BehrMJ,BoylesJG,CryanPM,HicksAC,BallmannAE, ColemanJ,RedellDN,ReederDM,BlehertDS.2011. Experimentalinfectionof batswith Geomycesdestructans causeswhite-nosesyndrome. Nature 480 :376 DOI10.1038/nature10590. LoudonAH,WoodhamsDC,ParfreyLW,ArcherH,KnightR,McKenzieV,Harris RN.2014. Microbialcommunitydynamicsandeffectofenvironmentalmicrobial reservoirsonred-backedsalamanders Plethodoncinereus TheISMEJournal 8 :830DOI10.1038/ismej.2013.200. LozuponeCA,HamadyM,KelleyST,KnightR.2007. Quantitativeandqualitative diversitymeasuresleadtodifferentinsightsintofactorsthatstructuremicrobial Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 16/19

PAGE 17

communities. AppliedandEnvironmentalMicrobiology 73 :1576 DOI10.1128/AEM.01996-06. LozuponeCA,KnightR.2005. UniFrac:anewphylogeneticmethodforcomparing microbialcommunities. AppliedandEnvironmentalMicrobiology 71 :8228 DOI10.1128/AEM.71.12.8228-8235.2005. LozuponeCA,LladserME,KnightsD,StombaughJ,KnightR.2011. UniFrac:an effectivedistancemetricformicrobialcommunitycomparison. TheISMEJournal 5 :169DOI10.1038/ismej.2010.133. McCoyCO,MatsenIVFA.2013. Abundance-weightedphylogeneticdiversitymeasures distinguishmicrobialcommunitystatesandarerobusttosamplingdepth. PeerJ 1 :e157DOI10.7717/peerj.157. McKenzieVJ,BowersRM,FiererN,KnightR,LauberCL.2012. Co-habitingamphibian speciesharboruniqueskinbacterialcommunitiesinwildpopulations. TheISME Journal 6 :588DOI10.1038/ismej.2011.129. McMurdiePJ,HolmesS.2013. Phyloseq:anRpackageforreproducibleinteractive analysisandgraphicsofmicrobiomecensusdata. PLoSONE 8 :e61217 DOI10.1371/journal.pone.0061217. McMurdiePJ,HolmesS.2014. Wastenot,wantnot:whyrarefyingmicrobiomedatais inadmissible. PLoSComputationalBiology 10 :e1003531 DOI10.1371/journal.pcbi.1003531. MinnisAM,LindnerDL.2013. .Phylogeneticevaluationof Geomyces andalliesreveals nocloserelativesof Pseudogymnoascusdestructans ,comb.nov.,inbathibernaculaof easternNorthAmerica. FungalBiology 117 :638 DOI10.1016/j.funbio.2013.07.001. MoellerAH,PeetersM,NdjangoJ-B,LiY,HahnBH,OchmanH.2013. Sympatric chimpanzeesandgorillasharborconvergentgutmicrobialcommunities. Genome Research 23 :1715DOI10.1101/gr.154773.113. MorrisonDW.1979. Apparentmaledefenseoftreehollowsinthefruitbat, Artibeus jamaicensis JournalofMammalogy 60 :11DOI10.2307/1379753. OchmanH,WorobeyM,KuoC-H,NdjangoJ-BN,PeetersM,HahnBH,Hugenholtz P.2010. Evolutionaryrelationshipsofwildhominidsrecapitulatedbygutmicrobial communities. PLoSBiology 8 :e1000546DOI10.1371/journal.pbio.1000546. OhJ,ConlanS,PolleyEC,SegreJA,KongHH.2012. Shiftsinhumanskinandnares microbiotaofhealthychildrenandadults. GenomeMedicine 4 :1 DOI10.1186/gm300. OksanenJ,BlanchetG,KindtR,LegendreP,MinchinPR,O'HaraRB,SimpsonGL, SolymosP,StevensM,WagnerH.2015. Vegan:communityecologypackage.R packageversion2.3-0. Availableathttp://CRAN.R-project.org/package=vegan OrtegaJ,AritaHT.1999. Structureandsocialdynamicsofharemgroupsin Artibeus jamaicensis Chiroptera:Phyllostomidae. JournalofMammalogy 80 :1173 DOI10.2307/1383168. OrtegaJ,Castro-ArellanoI.2001. Artibeusjamaicensis MammalianSpecies 662 :1 DOI10.1644/1545-1410<0001:AJ>2.0.CO;2. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 17/19

PAGE 18

PhillipsCD,PhelanG,DowdSE,McDonoughMM,FergusonAW,HansonJD,SilesL, Ordez-GarzaN,SanFranciscoM,BakerRJ.2012. Microbiomeanalysisamong batsdescribesinfluencesofhostphylogeny,lifehistory,physiologyandgeography. MolecularEcology 21 :2617DOI10.1111/j.1365-294X.2012.05568.x. PorterFL.1978. Roostingpatternsandsocialbehaviorincaptive Carolliaperspicillata JournalofMammalogy 59 :627DOI10.2307/1380245. PreheimSP,PerrottaAR,FriedmanJ,SmilieC,BritoI,SmithMB,AlmEJ.2013a. Computationalmethodsforhigh-throughputcomparativeanalysesofnatural microbialcommunities.In:EdwardFD,ed. Methodsinenzymology .SanDiego: ElsevierAcademicPress,353. PreheimSP,PerrottaAR,Martin-PlateroAM,GuptaA,AlmEJ.2013b. Distributionbasedclustering:usingecologytorefinetheoperationaltaxonomicunit. Appliedand EnvironmentalMicrobiology 79 :6593DOI10.1128/AEM.00342-13. PriceMN,DehalPS,ArkinAP.2010. FastTree2:approximatelymaximum-likelihood treesforlargealignments. PLoSONE 5 :e9490DOI10.1371/journal.pone.0009490. RDevelopementCoreTeam.2015. R:alanguageandenvironmentforstatistical computing.Vienna:RFoundationforStatisticalComputing. RoeselersG,MittgeEK,StephensWZ,ParichyDM,CavanaughCM,GuilleminK, RawlsJF.2011. Evidenceforacoregutmicrobiotainthezebrafish. TheISME Journal 5 :1595DOI10.1038/ismej.2011.38. Romano-BertrandS,Licznar-FajardoP,ParerS,Jumas-BilakE.2015. Impactde l'environnementsurlesmicrobiotes:focussurl'hospitalisationetlesmicrobiotes cutansetchirurgicaux. RevueFrancophonedesLaboratoires 2015 :75 DOI10.1016/S1773-035X-8. RothRR,JamesWD.1988. Microbialecologyoftheskin. AnnualReviewofMicrobiology 42 :441DOI10.1146/annurev.mi.42.100188.002301. SegataN,IzardJ,WaldronL,GeversD,MiropolskyL,GarrettWS,HuttenhowerC. 2011. Metagenomicbiomarkerdiscoveryandexplanation. GenomeBiology 12 :Article R60DOI10.1186/gb-2011-12-6-r60. SongSJ,LauberC,CostelloEK,LozuponeCA,HumphreyG,Berg-LyonsD,Caporaso JG,KnightsD,ClementeJC,NakielnyS,GordonJI,FiererN,KnightR.2013. Cohabitingfamilymemberssharemicrobiotawithoneanotherandwiththeirdogs. eLife 2 :e00458DOI10.7554/eLife.00458. TurnerGG,ReederDM,ColemanJTH.2011. Afive-yearassessmentofmortalityand geographicspreadofwhite-nosesyndromeinNorthAmericanbats,withalookat thefuture:updateofwhite-nosesyndromeinbats. BatResearchNews 52 :13. TzengT-D,PaoY-Y,ChenP-C,WengFC-H,JeanWD,WangD.2015. Effectsofhost phylogenyandhabitatsongutmicrobiomesoforientalriverprawn Macrobrachium nipponense PLoSONE 10 :e0132860DOI10.1371/journal.pone.0132860. USFish&WildlifeService.2012. NorthAmericanbatdeathtollexceeds5.5million fromwhite-nosesyndrome. NewsRelease Availableathttp://www.batcon.org/pdfs/ USFWS_WNS_Mortality_2012_NR_FINAL.pdf Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 18/19

PAGE 19

VellendM,CornwellWK,Magnuson-FordK,MooersA.2011. Measuringphylogeneticbiodiversity.In:MagurranAE,ed. Biologicaldiversity:frontiersinmeasurement andassessment .Oxford:OxfordUniversityPress,194. VoigtCC,KingstonT.2016. Batsintheanthropocene:conservationofbatsinachanging world .ChamHeidelberg:Springer. WalkeJB,BeckerMH,LoftusSC,HouseLL,CormierG,JensenRV,BeldenLK.2014. Amphibianskinmayselectforrareenvironmentalmicrobes. TheISMEJournal 8 :2207DOI10.1038/ismej.2014.77. WilcoxonF.1945. Individualcomparisonsbyrankingmethods. BiometricsBulletin 1 :80DOI10.2307/3001968. WilliamsCF.1986. Socialorganizationofthebat, Carolliaperspicillata Chiroptera: Phyllostomidae. Ethology 71 :265DOI10.1111/j.1439-0310.1986.tb00591.x. YingS,ZengDN,ChiL,TanY,GalzoteC,CardonaC,LaxS,GilbertJ,QuanZX.2015. Theinfluenceofageandgenderonskin-associatedmicrobialcommunitiesinurban andruralhumanpopulations. PLoSONE 10 :e0141842 DOI10.1371/journal.pone.0141842. Lemieux-Labontetal., PeerJ ,DOI10.7717/peerj.2430 19/19