Present and Potential Future Distribution of Common Vampire Bats in the Americas and the Associated Risk to Cattle


previous item | next item

Citation
Present and Potential Future Distribution of Common Vampire Bats in the Americas and the Associated Risk to Cattle

Material Information

Title:
Present and Potential Future Distribution of Common Vampire Bats in the Americas and the Associated Risk to Cattle
Series Title:
Plos One
Creator:
Lee, Dana N.
PapeÅŸ, Monica
Van den Bessche, Ronald A.
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Cattle ( local )
Livestock ( local )
Rabies ( local )
Bats ( local )
Climate Change ( local )
Mexico ( local )
Brazil ( local )
Ecological Niches ( local )
Genre:
serial ( sobekcm )

Notes

Abstract:
Success of the cattle industry in Latin America is impeded by the common vampire bat, Desmodus rotundus, through decreases in milk production and mass gain and increased risk of secondary infection and rabies. We used ecological niche modeling to predict the current potential distribution of D. rotundus and the future distribution of the species for the years 2030, 2050, and 2080 based on the A2, A1B, and B1 climate scenarios from the Intergovernmental Panel on Climate Change. We then combined the present day potential distribution with cattle density estimates to identify areas where cattle are at higher risk for the negative impacts due to D. rotundus. We evaluated our risk prediction by plotting 17 documented outbreaks of cattle rabies. Our results indicated highly suitable habitat for D. rotundus occurs throughout most of Mexico and Central America as well as portions of Venezuela, Guyana, the Brazilian highlands, western Ecuador, northern Argentina, and east of the Andes in Peru, Bolivia, and Paraguay. With future climate projections suitable habitat for D. rotundus is predicted in these same areas and additional areas in French Guyana, Suriname, Venezuela and Columbia; however D. rotundus are not likely to expand into the U.S. because of inadequate ‘temperature seasonality.’ Areas with large portions of cattle at risk include Mexico, Central America, Paraguay, and Brazil. Twelve of 17 documented cattle rabies outbreaks were represented in regions predicted at risk. Our present day and future predictions can help authorities focus rabies prevention efforts and inform cattle ranchers which areas are at an increased risk of cattle rabies because it has suitable habitat for D. rotundus.
Original Version:
Plos One, Vol. 7, no. 8 (2012-08-10).

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
This item is licensed with the Creative Commons Attribution License. This license lets others distribute, remix, tweak, and build upon this work, even commercially, as long as they credit the author for the original creation.
Resource Identifier:
K26-05086 ( USFLDC: LOCAL DOI )
k26.5086 ( USFLDC: LOCAL Handle )

USFLDC Membership

Aggregations:
University of South Florida
Karst Information Portal

Postcard Information

Format:
serial

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

PresentandPotentialFutureDistributionofCommon VampireBatsintheAmericasandtheAssociatedRiskto CattleDanaN.Lee * ,MonicaPapes ¸,RonaldA.VanDenBusscheDepartmentofZoology,OklahomaStateUniversity,Stillwater,Oklahoma,UnitedStatesofAmericaAbstractSuccessofthecattleindustryinLatinAmericaisimpededbythecommonvampirebat, Desmodusrotundus ,through decreasesinmilkproductionandmassgainandincreasedriskofsecondaryinfectionandrabies.Weusedecologicalniche modelingtopredictthecurrentpotentialdistributionof D.rotundus andthefuturedistributionofthespeciesfortheyears 2030,2050,and2080basedontheA2,A1B,andB1climatescenariosfromtheIntergovernmentalPanelonClimateChange. Wethencombinedthepresentdaypotentialdistributionwithcattledensityestimatestoidentifyareaswherecattleareat higherriskforthenegativeimpactsdueto D.rotundus .Weevaluatedourriskpredictionbyplotting17documented outbreaksofcattlerabies.Ourresultsindicatedhighlysuitablehabitatfor D.rotundus occursthroughoutmostofMexico andCentralAmericaaswellasportionsofVenezuela,Guyana,theBrazilianhighlands,westernEcuador,northernArgentina, andeastoftheAndesinPeru,Bolivia,andParaguay.Withfutureclimateprojectionssuitablehabitatfor D.rotundus is predictedinthesesameareasandadditionalareasinFrenchGuyana,Suriname,VenezuelaandColumbia;however D. rotundus arenotlikelytoexpandintotheU.S.becauseofinadequate‘temperatureseasonality.’Areaswithlargeportionsof cattleatriskincludeMexico,CentralAmerica,Paraguay,andBrazil.Twelveof17documentedcattlerabiesoutbreakswere representedinregionspredictedatrisk.Ourpresentdayandfuturepredictionscanhelpauthoritiesfocusrabiesprevention effortsandinformcattlerancherswhichareasareatanincreasedriskofcattlerabiesbecauseithassuitablehabitatfor D. rotundus .Citation: LeeDN,Papes ¸M,VanDenBusscheRA(2012)PresentandPotentialFutureDistributionofCommonVampireBatsintheAmericasandtheAssociated RisktoCattle.PLoSONE7(8):e42466.doi:10.1371/journal.pone.0042466 Editor: R.MarkBrigham,UniversityofRegina,Canada Received April9,2012; Accepted July9,2012; Published August10,2012 Copyright: 2012Leeetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsunrestricted use,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: Theauthorshavenosupportorfundingtoreport. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. *E-mail:dana.lee10@okstate.eduIntroductionSincetheintroductionofdomesticlivestockintotheNew World,vampirebat-transmittedrabieshasbeentheprimary diseaseprobleminlivestock[1],and Desmodusrotundus ,the commonvampirebat,hasservedasamajorconstrainttothe successofthecattleindustry[2],[3]. D.rotundus canfeedfromthe bloodofanymammal,butreadilyfeedsoncattle[4],[5], primarilybecausecattleareamorepredictablepreysourcethan wildlife[3]. D.rotundus havebeenreportedtoroostnearaherdand feedrepeatedly[1].Inareaswithhighbatdensity,asingle individualhasreceived12bitesinonenightandhaduptofour batsfeedingatatime[6].Cattleattempttoshakethebatoff,but thisisonlyatemporaryreprieve. Nightlyattacksby D.rotundus cannegativelyimpactthehealthof cattlebycausingadecreaseinmassgain,decreasedmilk production,increasedsecondarybacterialinfections,andincreasedriskofrabiesorotherdiseases[3],[7],[8].Inaddition totheinitialvolumeofbloodloss,theanticoagulantsecretedinthe salivaof D.rotundus causesbloodtoseepfromthewoundforhours aftertheinitialbite[9].SchmidtandBadger[10]reportedthat cattleownersestimatedfrequentbitingcouldreducetheamount ofmilkproducedbyasinglecow260Lperyearanddecrease meatproductionofanindividual39.7kgperyear.Thompsonet al.[11]foundcattlefromtypicaltropicalregionsthatwereinpoor conditionhadasignificantincreaseinmilkproductionwhenthey wereinjectedwithananticoagulantandthusmitigatedthe negativeeffectsof D.rotundus .Theyconcludedthatcattleinthese areasexperienceothersourcesofstresssuchasextremeclimate, inadequatediet,andotherparasites,thereforeprotectionfrom D. rotundus iscritical.However,anempiricalstudyinColumbiadid notfindacorrelationbetweenthenumberofvampirebatbites andmilkproduction[12].Thereisstillnotaconsensusonthe effectsofbloodlossoncattle. Nightlyparasitismpotentiallyaffectsmeatandmilkproduction, buttheprimarylimitingfactorforlivestockproductionthroughout LatinAmericaisvampirebat-transmittedrabies[3],[2].In1968, over500,000cattlediedfrombat–transmittedrabiesinLatin America[13].Withtheinitiationofbatcontrolmethodsand vaccinesforcattle,thesenumbersdeclinedto9,904reportedcases in1983[14],1,831in1993[15],6,088in2000[16],3,327in 2002[8],and1,580in2006[15],[17].Whilethenumbersof reportedrabiesfatalitieshavedecreased,theseareonlyconservativeestimates.Thescarcityofdiagnosticlabsimpedestestingof mostcattlefounddeadinthefield,suggestingtheactualrateof mortalityduetorabiesishigher[14].Milkandmeatfroman animalinfectedwithrabiesmaystillcontainthevirus,but fortunately,pasteurizationandcookingmeattopropertemperaPLOSONE|www.plosone.org1August2012|Volume7|Issue8|e42466

PAGE 2

tureskillsthevirus[18–20].Todatetherehasbeenno documentationofahumanrabiescaseresultingfromlivestock intheU.S.[21]. Itishardtoestimatetheimpactof D.rotundus onthecattle industryduetoalackofaccuratereporting,particularlyinrural areas[2],[8],[22]butAchaandAlba[14]estimated D.rotundus wereresponsibleforlossesgreaterthan$40millionUSduring 1983andagainin1984.Theselosses,coupledwithcostsofvarious preventivemeasures,canbeasignificanteconomicproblemfor the18countriesaffectedbybovinerabiesinLatinAmerica[22]. Duetothelargeexpenseofcontrollingthespreadofbovinerabies andmitigatingtheproductionlossescausedby D.rotundus ,the mosteffectivecourseofactionwouldbeforcountriestofocus effortsonareaswithinLatinAmericawherelargenumbersof cattleand D.rotundus co-occur.However,itisdifficulttodetect suchregionallocationsbecausethepotentialareaforoverlapistoo great[16].Aneffectivewaytopredictdistributionsisthrough modelingspecies’ecologicalniches[23].Thismethoddetects associationsbetweenenvironmentalvariables[intheformof GeographicInformationSystems(GIS)layers]andlocalitiesof knownoccurrencesofspeciestogenerateaprobabilityofthe speciespresenceineachpixelofthestudyarea.Thesepredictions canthenbeplottedonadigitalmapusingGISsoftware.One specificuseofnichemodelingistoidentifypotentialareasfor diseasetransmissionbyhighlightingareasenvironmentallysuitable forboththehostandvectorspecies[24],[25].Thusfar,ecological nichemodelinghasbeenusedtopredictpossibleareasatriskfor outbreaksofanthrax[26],denguefever[27],chagasdisease[28], chytridmycosis[29],plague[30],andhemorrhagicfevercausedby filoviruses[31]. Giventheestimateof70millioncattleatriskinareaswhere rabieshasbeenreportedinthepast10years[14],[32],webelieve ecologicalnichemodelingcouldbeabeneficialtooltopredict areaswherecattlecouldpotentiallyhaveagreaterriskofrabies andothernegativeeffectsof D.rotundus .Cattlerabiesoccurrences appeartobelinkedtoseasonalclimatevariationandanincrease inbatpopulationsize[33],thereforewegeneratedanenvironmentalsuitabilitymapfor D.rotundus andusedapublisheddataset ofpredictedcattledensitytoindicateareasthatmayhaveahigher relativeriskofcommonvampirebatpredationorsuitable conditionsforcattlerabiesoutbreaks.Ascattledensityhasalready beenshowntobeanimportantfactortoexplainthespatial clusteringpatternof D.rotundus [34],wehypothesizeareaswitha highdensityofcattleandsuitableenvironmentalconditionsfor D. rotundus couldsufferthegreatesteffectsofbothnightlyparasitism andriskforrabies.Wealsoinvestigatedifthedistributionof D. rotundus wouldchangeandpossiblyextendintocurrently unsuitableareas,includingtheUnitedStates,withfutureclimate predictions.Climatechangehasalreadybeenpredictedtoimpact thedistributionofEuropeanbats[35]and D.rotundus inMexico [36].Thechangeinamountofsuitablehabitatmayintroducebat predationoncattlenotcurrentlyaffected;howeverourresultsdo notaccountforfuturecattledistributions.Methods Desmodusrotundus predictedpotentialdistributionTogeneratethepresentdaypotentialdistributionmapfor D. rotundus ,thestudyareawasdelimitedusingtheknownspecies’ distributionfromMexicosouththroughCentralAmericato Uruguay,Argentina,andChile,specificallyfrom28 u Nto33 u S [37].Therearenoknownoccurrencesof D.rotundus onBaja peninsula(Mexico)orintheCaribbeanislands,exceptTrinidad, Tobago,andMargaritaIsland,sotheseareaswereexcludedfrom thepresentdayprediction.Museumrecordsof D.rotundus (9,741) weredownloadedfromtheGlobalBiodiversityInformation Facility[38](http://www.gbif.org/).Thisorganizationservesas adataportaltoallowfreeaccessofinformationaboutnatural historymuseumholdings.Occurrencedatafor D.rotundus collected before1940wereremovedbecauseGISenvironmentaldataare notavailableforthattimeframe.Recordslackinglatitudeand longitudecoordinatesweregeoreferencedinGEOlocatev.3.22 [39](http://www.museum.tulane.edu/geolocate/).Thiswebapplicationusestextualdescriptionsofspecimencollectinglocalities toassignlatitudeandlongitudecoordinatestospecimens. Dependinguponthedetailforthecollectinglocality,the georeferenceswereassignedlow,medium,orhighconfidence basedonthegeographicextentoftheerrorassociatedwiththe georeference.Recordswithmediumorhighconfidencescores wereincludedintheoccurrencedataset.Allpointswereplottedin ArcMap10[40]toconfirmthatgeoreferencedlocalitiescorrespondwithoriginaldescriptions.Finally,duplicaterecordswere removed,leaving984spatiallyuniqueoccurrencepointsfor D. rotundus . GISclimaticlayersrepresentingminimumandmaximum temperature,andprecipitation,averagedoverthelastfivedecades (1950–2000,hereafter‘‘present’’),wereobtainedfromthedata portaloftheResearchProgramonClimateChange,Agriculture andFoodSecurityoftheConsultativeGrouponInternational AgriculturalResearch[41](http://www.ccasfs-climate.org/).To predictthedistributionof D.rotundus infutureclimates,we downloadedfromthesamesourceclimatemodeldata(temperatureandprecipitation)for2021–2040(hereafter2030),2041– 2060(hereafter2050),and2071–2090(hereafter2080),downscaledfromMIROC3.2GeneralCirculationModel(GCM),one oftheGCMsusedintheFourthAssessmentReportofthe IntergovernmentalPanelonClimateChange(IPCC)[42].We usedtheA2,A1B,andB1emissionscenariosincludedintheIPCC SpecialReportonEmissionScenarios.IntheA2scenario,the focusisonregionaleconomicdevelopmentandslowchange towardscleanertechnology.Itisalsocharacterizedbyanincrease ofCO2concentrationto1250ppmandtemperatureby3.4 u Cin 2100.TheA1Bscenariorepresentscurrenttrendsinwhichhuman energyusecontinuestoincrease(notrelyingononeparticular energysource),butCO2emissionsarestabilizedtosomedegreeby technologicaladvancesandpublicawareness.AnestimatedCO2concentrationof850ppmandtemperatureincreaseof2.8 u Cis used.IntheB1scenario,thehumanpopulationpeaksandstartsto declinearound2050.Thereisaswitchtousingcleaner technology,CO2concentrationsincreaseto600ppm,and temperaturerisesby1.8 u C[43].Presentandfuturetemperature andprecipitationvariableswereusedtocalculate19‘‘bioclimatic variables’’forpresent,2030,2050,and2080periodsrepresenting quarterlyandmonthlyclimateseasonalityandextremes[44]. BioclimaticvariablesweregeneratedinESRIArcInfousing availableAMLcode(http://www.worldclim.org/bioclim).All environmentalvariablelayershada1km2resolutionandwere maskedtotheextentofthestudyareainArcMap10[40]. Environmentallayersand D.rotundus occurrenceswereusedin Maxentv.3.3.3k[45],[46]toruntheecologicalnichemodels becausethemaximumentropyalgorithmrequirespresence–only dataandhasbeenshowntoproducereliableresults[47],[48]. Maxentcontraststheenvironmentalconditionsassociatedwith presencespointswithrandombackgroundpointsthatsample availableenvironmentalspacewherethespeciescouldpotentially occur.Additionally,Maxentuses‘‘features’’,functionsderived fromtheenvironmentalvariables,asparameterstokeepthemodel fromoverfittingthedata[46].WeusedtheautofeaturesoptionNicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org2August2012|Volume7|Issue8|e42466

PAGE 3

whichchoosesthefeaturesappropriateforthenumberof occurrencesinthedataset.Giventherelativelylargeoccurrence datasetavailable,weoptedfortherandomseedandtest percentageoptionsinMaxenttorandomlysplittheoccurrence points(984)intotrainingandtestingdatasetseachwithhalfofthe datapoints(492).Jackknifingwasappliedforallenvironmental layerstodetermineindividualpercentageofcontributiontothe modeloverallaccuracygain.Wethenrankedthevariablesby percentcontribution.Wechosetousethefivevariablesthat contributedmorethan5%torefineourpredictionsinafinal model.Theecologicalnichemodelgeneratedusingthesubsetof environmentalvariableswasprojectedonthe2030,2050,and 2080environmentaldatasetsforeachofthethreeclimate scenarios,resultinginninepredictions.Predictingspecies’ distributionsusingprojectionsofecologicalnichemodelson futureclimatedatasetsisanappropriatemethodforgaining insightstopossiblechangesinspeciesdistributions[49],[50],and usedtodocumentbothrangeexpansions[51–53]andreductions [54],[55]. Todirectlycomparepresentdaypotentialdistributionof D. rotundus toitspotentialdistributionin2030,2050,and2080,we identifiedpixelsthatwereunsuitableunderpresentconditionsbut becamesuitableinthefuturepredictions.Weconvertedthe continuousprobabilityofpresencevaluestoabinaryoutputby applyinga10%omissionerrorthresholdtotheMaxentoutputs. Thismethodassignspixelswithaprobabilityofpresencevalueless thanthelowestvaluecorrespondingto10%ofthetrainingpoints avalueofzero(absent),andpixelswithaprobabilityofpresence abovethisvaluearegivenavalueofone(present).Thisconversion ismoresensitiveto‘‘outliers’’(locationswherethespecieswas collecteddespitealowpredictedprobabilityofsuitability)and constrainsthepixelsinitiallypredictedaspresent[56].Wewere alsointerestedinidentifyingtheenvironmentalvariablethatmost influencesthedifferencesbetweenthepresentdayandeachfuture prediction.Maxentv.3.3.3k[45],[46]canaddressthisquestionby measuringthesimilaritybetweenpresentandfutureclimatesfor eachenvironmentalvariable.Thevariablewiththelargest dissimilarityvalueforeachpixelisthenplottedongeographic space[57].Finally,thepresentnichemodelwasevaluatedfor accuracyusingtheareaunderthecurve(AUC)ofthereceiver operatorcharacteristicwhichplotstheproportionofpresences predictedabsent(omissionerror)againsttheproportionofarea predictedpresent.AnAUCvalueof1indicatesaperfect predictionand0.5isapredictionnobetterthanrandom[58]. However,theusefulnessROCAUCtoevaluatingmodelaccuracy isincreasinglyquestioned[59–61].Aclearerbutperhaps oversimplifiedassessmentisprovidedbytheomissionerroralone.CattleatriskpredictionProjectedcattledensitydatafor2005wereobtainedfromthe FoodandAgricultureOrganizationoftheUnitedNations(FAO) [62].TheAnimalProductionandHealthDivisionoftheFAO maintainsapublicdatabasecontaininggeoreferenceddataon livestocknumbersbutthesenumbersareatdifferentspatialscales fordifferentregions.TheFAOusedthisdatabasewithvegetative, geological,environmental,demographic,andclimaticvariablesto interpolateandextrapolatethedensityofcattlefortheworldat 1km2resolution.Pixelsindeserts,highmountains,closedcanopy forests,andhighlyurbanizedareaswerecodedasunsuitable habitatforcattle.Theresultingpredictioncanbedownloaded fromtheFAOwebsite(http://www.fao.org/AG/AGAInfo/ resources/en/glw/GLW_dens.html)andusedasalayerfor additionalprocessinginArcMap10.Tohighlighttheareaswith suitablehabitatforvampirebatsandincludeameasureofcattle density,thecontinuousMaxentoutputforthepresentday D. rotundus distributionwasconvertedtoabinaryoutputusinga10% omissionerrorastheminimumthresholdvaluefollowingthe methodsexplainedearlier.Thecattledensitylayerwasthen maskedtoonlyshowpixelscorrespondingtopredictedpresences of D.rotundus .Finally,locationsfor17casesofcattlerabies outbreaksreportedinthescientificliteratureorfoundthrough ProMED-mail(http://www.promedmail.org)[63],whichisa databasecontainingrecentalertsoninfectiousdiseases,were plottedonthecattleriskpredictionmap.Weassignedgeographic coordinatestoeachrecordusingGeoNetNamesServiceonline gazetteer(http://geonames.nga.mil/ggmagaz/)andestimatedthe georeferencinguncertaintybasedonthegeographicextentofthe localitydescriptionoftheseoutbreaks.Weusedthegeoreferencing uncertaintymeasuretomapazoneofuncertainty(GISbuffer) aroundeachoutbreak.Insomecases,thegeoreferencing uncertaintywasonly2–5kmsoweappliedaminimumzoneof uncertaintyof10kmtoallrecordsbecauseLord[33]reportedthe majorityofbovinerabiesoutbreaksreach5–10kmwide.The numberofpixelsintheuncertaintyzonepredictedatriskbyour modelwerethencalculated.Itisimportanttonotesomerecords onProMED-maildonotreporthowthecattleacquiredrabies,but weassumedvampirebatswerethevector.Results Desmodusrotundus predicteddistributionFiveoftheclimatevariables(precipitationseasonality,temperatureseasonality,precipitationofthewettestmonth,precipitation ofthedriestmonth,andmeantemperatureofthecoldestmonth) contributedmosttothemodel(Table1).Thesewerethevariables chosentoincludeinthefinalmodelusedtopredictthepresentday andfuturedistributionsof D.rotundus .AtrainingAUCof0.826 andatestingAUCof0.805indicatedthepresentmodel performedwellusingonlythetopfiveenvironmentalvariables. OurpresentmodelpredictedmostofMexicoandCentral Americatohavesuitableenvironmentalconditionsfor D.rotundus (Fig.1).Otherregionsofhighsuitabilityincludeportionsof Venezuela,Guyana,theBrazilianhighlands,westernEcuador, andeastoftheAndesinPeru,Bolivia,Paraguay,andnorthern Argentina. Generallyregionsofsuitabilityinthepresentdaymodels, Mexico,CentralAmerica,Venezuela,Guyana,westernEcuador andPeru,andBolivia,alsohadhighsuitabilitywhenthemodel wasprojectedtofutureclimatesfor2030,2050,and2080(Fig.2). Differencesamongtheclimatescenariosweremostobviousinthe amountofsuitablehabitatfor D.rotundus inBrazil.Therewerealso areasthatwouldbecomesuitablefor D.rotundus inthefuture climates.TheseincludedFrenchGuyana,Suriname,andadditionalportionsofVenezuelaandColumbia(Fig.2).The CaribbeanregionandFloridahadsuitablehabitatfor D.rotundus underfutureclimatesbuttheseregionswerenotincludedinthe presentdaymodelbecausetherearenomuseumrecordsfrom theseareas.NochangesoccurredinCentralAmericaunderfuture scenarios,andnosuitableregionswerepredictedwithanyofthe futureclimatescenariosintheU.S.,exceptforsouthernFlorida (Fig.2).Thisoutcomeisincontradictionwithpreviously hypothesizedwiderangeexpansionintotheU.S.[36],which canbeexplainedbyregionaldifferencesin‘temperature seasonality’ofthepresentdayandfutureclimates(Fig.3).The lackofexpansionintonewareasinSouthAmericacouldbe explainedbydifferencesin‘meantemperatureofcoldestmonth’ (Fig.3).NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org3August2012|Volume7|Issue8|e42466

PAGE 4

CattleatriskpredictionWhenthe10%omissionerrorthresholdwasappliedtothe presentday D.rotundus distribution,51.0%ofthestudyareawas classifiedassuitablehabitatforvampirebats.Thisareaincluded mostofCentralAmerica,whichalsohasahighdensityofcattle perkm2,withtheexceptionoftheYucatanPeninsula(Fig.4).The YucatanPeninsulawassuitableforbatsbutwasclassifiedas unsuitableforcattleintheFAOAnimalProductionandHealth Divisioncattledataset.TheBrazilianhighlandsalsocontainland suitablefor D.rotundus andlargenumbersofcattle.Whilethe easternslopeoftheAndeswaspredictedtobesuitablefor D. rotundus ,mostofthismountainousregionisnotsuitableforcattle ranching.The17documentedcasesofcattlerabiesweregenerally inareaswherecattlewerepredictedatrisk(Fig.4).Twelveofthe 17caseshadpixelspredictedatriskwithinthegeoreferencing uncertaintyzone(Table2).Uponcloserinvestigationthefive outbreakswithnopixelspredictedatriskwereinveryclose proximitytopixelsinlocationsatrisk(1–12km). Figure1.Currentpotentialsuitablehabitatforthecommonvampirebat,Desmodusrotundus.basedonfourdifferent environmentaldatasets. (A)present,(B)2030,(C)2050,(D)2080.Blackdotsindicatespatiallyuniqueknownoccurrencesfor D.rotundus whichwereusedinmodelconstruction. doi:10.1371/journal.pone.0042466.g001 NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org4August2012|Volume7|Issue8|e42466

PAGE 5

Table1. Jackkniferesultsindicatingvariablepercentcontributionstothemodel.EnvironmentalVariableContributiontofirstmodelContributiontofinalmodel Precipitationseasonality28.342.4 Temperatureseasonality23.424.7 Precipitationofwettestmonth11.617.5 Precipitationofdriestmonth7.24.8 Meantemperatureofcoldestmonth5.310.5 Precipitationofcoldestquarter4.2 Meantemperatureofcoldestquarter4 Meantemperatureofwettestquarter3.2 Annualprecipitation2.5 Precipitationofdriestquarter2.3 Meantemperatureofdriestquarter1.8 Maxtemperatureofwarmestmonth1.5 Temperatureannualrange1.1 Meandiurnalrange1.0 Meantemperatureofwarmestquarter0.8 Isothermality0.7 Precipitationofwettestquarter0.6 Annualmeantemperature0.4 Precipitationofwarmestquarter0.1 doi:10.1371/journal.pone.0042466.t001 Figure2.Futurepotentialsuitablehabitatforthecommon vampirebat,Desmodusrotundus,basedonthreeclimate scenariosandtimeframes. (A)2030scenarioA2,(B)2030scenario A1B,(C)2030scenarioB1,(D)2050scenarioA2,(E)2050scenarioA1B, (F)2050scenarioB1,(G)2080scenarioA2,(H)2080scenarioA1B,(I) 2080scenarioB1. doi:10.1371/journal.pone.0042466.g002 Figure3.Dissimilaritymapsindicatingwhichenvironmental factorwasmostdissimilarbetweenpresentdaypredictions andninefuturepredictions. (A)2030scenarioA2,(B)2030scenario A1B,(C)2030scenarioB1,(D)2050scenarioA2,(E)2050scenarioA1B, (F)2050scenarioB1,(G)2080scenarioA2,(H)2080scenarioA1B,(I) 2080scenarioB1.Coloredpixelsrepresentingdissimilaritybetween presentdayandfuturepredictions:Blueformeantemperatureof coldestmonth,purplefortemperatureseasonality,greenforprecipitationseasonality,yellowforprecipitationofdriestmonth,andpinkfor precipitationofwettestmonth. doi:10.1371/journal.pone.0042466.g003 NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org5August2012|Volume7|Issue8|e42466

PAGE 6

DiscussionOurmodelindicatesseveralenvironmentalcharacteristicsthat explainthedistributionof D.rotundus throughoutMexico,Central andSouthAmerica.Temperatureandprecipitationvariablesare consistentwithknownecologicalrequirementsof D.rotundus . ‘Meantemperatureofthecoldestmonth’and‘temperature seasonality’(differencebetweensummerandwinter)areamong someofthemostimportantpredictorsofhabitatsuitability.This agreeswithpreviousresearch,whichsuggeststhedistributionof thisspeciesismostlimitedbythecoldesttemperatureinwinter. D. rotundus cannotsurviveinareasthathavetemperaturesbelow15 u C [13]becausethermoregulationinthesecoldtemperaturesrequires moreenergythananindividualcanconsumeonanightlybasis [64]. D.rotundus alsopreferslocationswithlessthan45%humidity [65],whichcanexplainwhy‘precipitationofthewettestmonth’ Figure4.Cattledensitiesperkm2showninpixelswithpredictedsuitablehabitatforthecommonvampirebat,Desmodusrotundus. CattledensityincreaseswithshadesofredandgraypixelsindicateareaspredictedtobeunsuitableforcattlebytheFoodandAgriculture OrganizationoftheUnitedNations[55].Greendotsindicatedocumentedcattlerabiesoutbreaksandblackcirclesrepresentuncertaintyzonefor eachrecord. doi:10.1371/journal.pone.0042466.g004 NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org6August2012|Volume7|Issue8|e42466

PAGE 7

and‘precipitationseasonality’and‘precipitationofthedriest month’arefoundtocontributegreatlytothemodel. Usedcollectively,thesecharacteristicsdepicttheknown distributionof D.rotundus wellandcanbeappliedtopredictions of D.rotundus ’sdistributioninfutureclimates.Twoofthese environmentalvariables,‘meantemperatureofthecoldestmonth’ and‘temperatureseasonality’canbeinterpretedaslimitingfactors of D.rotundus expansioninSouthAmericaandtheU.S., respectively.Interestingly,differentenvironmentalcharacteristics appearaslimitingfactorsinthetwocontinents,butnot surprisinglyasmultiplevariablesarerequiredtoproperlydescribe aspecies’fundamentalniche. Whenthepredicteddistributionfor D.rotundus iscombinedwith cattledensitydata,areasinMexico,andCentralandSouth Americathathavecattlewithahigherrelativeriskofharmful effectsfromvampirebatparasitismarehighlighted.Mostof Mexico,CentralAmerica,Paraguay,andtheBrazilianhighlands arehighlysuitableforboth D.rotundus andcattle.Cattleinthis regionarelikelytobesympatricwith D.rotundus ,sufferfrom commonvampirebatbites,andhaveagreaterriskofcontracting rabies.Resultsfromourcattleatriskpredictionarenotsurprising consideringMexicoandBrazilarebothroutinelylistedinthetop threecountrieswiththemostreportedcasesofcattlerabies[14], [15],[17].UnfortunatelyforthecattleindustryinLatinAmerica, morelandbecomessuitablefor D.rotundus ifclimatechange followsanyscenarioweused.Itisalsoimportanttonotethereis suitablehabitatfor D.rotundus intheCaribbean.Withthe exceptionofTrinidad,Tobago,andMargaritaIsland,thereare currentlynovampirebatsinthisregionbutourresultspredictthey couldbesuccessfulinvadersifcattlearealsopresent.Finally,our modelpredictsthemajorityofcattleintheU.S.aresafefromthe negativeimpactsof D.rotundus ,despiteglobalwarmingtrends. WhilewerecognizebreedingdistributionsofNorthAmerican birdshavealreadymovednorthward[66],ourresultssuggest D. rotundus willbelimitedby‘temperatureseasonality’andnot expandintotheU.S.throughMexico.Thisresultcontradicts anotherreport[36]whichsuggeststheGulfcoastofTexasand Louisianamaybecomecapableofinvasion.Theconlfictinresults isbasedonadifferenceinenvironmentalvariablesconsidered.It seemsMistryandMoreno-Valdez[36]madeinitialconclusionson therangeexpansionofvampirebatsafterexaminingonlyasingle temperatureincrease,whilethisstudyusesclimatescenariosand severalclimatevariablessummarizingannualandseasonal temperatureandprecipitationtrends.Theagreementbetween thisandthepreviousreportisthatsouthernBajaCalifornia, Florida,andthecoastsofMexicocouldbecomesuitableforthe vampirebatwithfutureclimates. Asexpected,the17documentedoutbreaksofcattlerabies occurredwithinornearbyareasatriskforharmfuleffectsof D. rotundus ,suggestingsuccessfulutilityofourprediction.Inaddition totheusefulnessofourriskprediction,thereareotherpatternsof rabiestransmissionthatcouldbeusedincombinationwithour resultstohelpauthoritiesfocuspreventionefforts.Epidemiological characteristicsofvampirebattransmittedrabiesincattlehave beenassociatedwithtopographicalandgeographicalfeatures[67]. Migrationpatternsofoutbreaksusuallyfollowriversbecausethere areampletreesforthebatstoroost[33],[67].Whenstrainsof cattlerabiesisolatedfromBrazilwereexamined,groupingsof differentphylogeneticstrainscouldbeexplainedbyelevation boundaries[67].AlsoinBrazil,regressionanalysisindicated clusterpatternsofvampireattacksoncattlecouldbeexplainedby ‘distancetoforest’,‘proportionofsugarcane’,and‘cattledensity’ [34].InVenezuela,Mexico,andArgentina,thenumberof outbreakswascorrelatedwithprecipitationandtheseasonalityof vampirereproduction[33].Currentlymostcountrieshave scatteredeffortsthatcanonlyrespondtoareaswherecattlehave died[33],butconsideringthesefactorsalongwithourecological nichemodelpredictionsiscriticaltomitigatingthespreadofcattle rabies. Avarietyofmethodsareemployedtoreducetheharmfuleffects from D.rotundus ,includingdestroyingroostswithfireordynamite [65]orcementingthemclosed;howeverthesemethodsalsoaffect Table2. Cattlerabiesoutbreaksusedtoevaluatecattleatriskprediction.OutbreakCitationSizeofUncertaintyZone%‘‘atrisk’’pixels Guasipati,Venezuela[28]10km100% Olmedo,Manab ´ ,Ecuador[75]10km100% Floresto ´ polis,Parana ´ ,Brazil[75]42km77% BelaVistadoPara ´ so,Parana ´ ,Brazil[75]43km66% Aldama,Tamaulipas,Mexico[56]10km50% Parana ´ ,Brazil[75]656km46% Guarayos,SantaCruz,Bolivia[75]305km32% Salta,Argentina[28]680km26% Maltrata,Veracruz,Mexico[75]10km20% RioGrandedoSul,Brazil[75]723km11% Oventeni,Atalaya,UcayaliRegion,Peru[75]320km10% IslaApipe ´ ,Argentina[28]27km3% LosChiles,Alajuela,CostaRica[75]10km0% RanchoSantaGertrudis,Tamaulipas,Mexico[75]10km0% SantoTome ´ ,Corrientes,Argentina[75]10km0% Saposoa,SanMartin,Peru[75]10km0% SevillaDonBosco,Morona-Santiago,Ecuador[75]10km0% doi:10.1371/journal.pone.0042466.t002 NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org7August2012|Volume7|Issue8|e42466

PAGE 8

anyspeciesthatcohabitatswith D.rotundus ,suchasthethreatened Dekeyser’snectarbat, Lonchophylladekeyseri [68].Alternatively,an anticoagulantpoison,diphacinone,isusedtokillthebats.When injectedintocattle,feeding D.rotundus willreceivealethaldose [10],[65],[69],[70].Thistreatmentissafeonadultcattlebutis notrecommendedonsucklingcalves[7].Additionally,the chemicalmustberoutinelyinjected[71].Diphacinonecanalso bemixedwithvaselineandplacedonacapturedbat.Afterthebat returnstotheroosts,this‘‘vampiricidegel’’istransferredtocolony mates.Asthebatsthengroomthemselvestheyingestthechemical [70]. Asrabiesisthemostimportantthreattothecattle,non–lethal methodsofbatcontrolandcattleprotectionincludevaccinationof eitherspecies.Arabiesvaccineforcattlewascreatedintheearly 1970’s,butitisnotwidelyused[13],[72].Manyranchersdonot vaccinatetheircattleunlesstherewasarecentrabiesoutbreak, evenwhen D.rotundus areknowntobeinthearea[1],[73].Even thoughvaccinationofallcattleispossible,thecostofroutinely vaccinationscanbeprohibitiveforsmalleroperations[10].An oralvaccinecanbemixedwithvaselineandappliedtothebatin thesamemannerasthe‘‘vampiricidegel’’.Thevaccineisalso transferredtootherbatsintheroost.Asbatsgroomthemselves, theybegintodevelopimmunitytorabiesafteringestion[74].The costofthistreatmentmethodwasanalyzedwithestimatesfrom Massadetal.[3]andfoundtobecheaperthanboththe ‘‘vampiricidegel’’andcattlevaccines[74].Regardlessof managementstrategy,ourpredictionshelphighlightareasthat shouldreceivepriority. Ourresultsprovideacurrentpotentialdistributionof D.rotundus andcanbeusedtoindicateareaswherecattlemaybeatan increasedriskofbeingnegativelyaffectedbythesebats. Additionally,ourdatacanbecomparedwiththepublishedcattle densitydatasettolocateareaswithsuitablehabitatforcattlebut not D.rotundus .Eventhoughitcanbehardtodelineatetheseareas atsuchalargegeographicscale,mapsofsmallerregionscaneasily begeneratedwithafinerscale.Wewereabletopredictpotential changeindistributionofthecommonvampirebatunderthreeof themultipleclimatechangescenariosproposed.Itisimportantto keepinmindthesearesimplypredictionsandarenotindicativeof acertainfuture.Ourresultswillneedtobere-assessedperiodically whenupdatedandmorerefinedfutureclimatedataareavailable. Changesinlandcoverusecouldalsoaffectourfuturepredictions. Previousecologicalchangesfromanaturaltoamoreruraland agriculturallandscapehavefavored D.rotundus expansion[33], [68],andthistrendwilllikelycontinueinthefuture.The populationinLatinAmericaisprojectedtoincreaseto665million by2020andthedemandforlivestockproductionwillintensify [75].Tomeetthesedemands,itisimperativethecattleindustry minimizethenegativeimpactsfrom D.rotundus .Itisdifficultto protectallcattlefromthenegativeeffectsof D.rotundus ,butwe believeourriskmaphassignificantimplicationsfordetermining areasthatwouldbenefitmostfromrabiesimmunity.AcknowledgmentsWewouldliketothankJeremyWilkinsonforassistancewithArcMap10 andMaxent.WealsothankRichardDolmanandanonymousreviewers forhelpfulcommentsonearlierversionsofthemanuscript.AuthorContributionsConceivedanddesignedtheexperiments:DNLMPRAVDB.Performed theexperiments:DNLMP.Analyzedthedata:DNLMP.Contributed reagents/materials/analysistools:MP.Wrotethepaper:DNL.Edited manuscript:MPRAVDB.References1.TurnerDC(1995)Thevampirebat:afieldstudyinbehaviorandecology. Baltimore:JohnsHopkinsUniversityPress.54p. 2.BelottoA,LeanesLF,SchneiderMC,TamayoH,CorreaE(2005)Overviewof rabiesintheAmericas.VirusRes111:5–12. 3.MassadE,CoutinhoFAB,BurattiniMN,SallumPC,LopezLF(2001)Amixed ectoparasite–microparasitemodelforbat–transmittedrabies.TheorPopulBiol 60:265–279. 4.DelpietroHA,MarchevskyN,SimonettiE(1992)Relativepopulationdensities andpredationofthecommonvampirebat( Desmodusrotundus )innaturaland cattle–raisingareasinnorth–eastArgentina.PrevVetMed14:13–20. 5.VoigtCC,KelmDH(2006)Hostpreferenceofthecommonvampirebat ( Desmodusrotundus ;Chiroptera)assessedbystableisotopes.JMammal87:1–6. 6.GreenhallAM,SchmidtU,Lopez-FormentW(1971)Attackingbehaviorofthe vampirebat, Desmodusrotundus ,underfieldconditionsinMexico.Biotropica3: 136–141. 7.EliasDJ,ThompsonRD,SavariePJ(1978)Effectsoftheanticoagulant diphenadioneonsucklingcalves.BullEnvironContamToxicol20:71–78. 8.KuzminIV,RupprechtCE(2007)Batrabies.In:JacksonAC,WunnerWH, editors.Rabies.SanDiego:AcademicPress.pp.259–307. 9.GreenhallAM(1972)Thebitingandfeedinghabitsofthevampirebat, Desmodus rotundus .JZool168:451–461. 10.ScmhidtKM,BadgerDD(1979)Somesocialandeconomicaspectsin controllingvampirebats.ProcOklaAcadSci59:112–114. 11.ThompsonRD,EliasDJ,MitchellGC(1977)Effectsofvampirebatcontrolon bovinemilkproduction.JWildlManage41:736–739. 12.CoenCE(2002)Economicimpactofvampirebatactivityonmilkproductionby dairycows.In:Comparativenutritionalecologyoftwogeneraofvampirebats: Desmodusrotundus and Diaemusyoungi .Ph.D.dissertation,CornellUniversity, Ithica,NewYork.pp.11–47. 13.Arellano-SotaC(1988b)Vampirebat–transmittedrabiesincattle.RevInfect Dis10:Suppl4,S707–S709. 14.AchaPN,AlbaAM(1988)Economiclossesdueto Desmodusrotundus .In: GreenhallAM,SchmidtU,editors.Naturalhistoryofvampirebats.Florida: CRCPressInc.pp.207–214. 15.PanAmericanHealthOrganization,VeterinaryPublicHealthProgram(1993) EpidemiologicalsurveillanceofrabiesintheAmericas.28p. 16.MayenF(2003)HaematophagousbatsinBrazil,theirroleinrabies transmission,impactonpublichealth,livestockindustryandalternativestoan indiscriminatereductionofbatpopulation.JVetMedSerB50:469–472. 17.PanAmericanHealthOrganization,VeterinaryPublicHealthProgram(2007) EpidemiologicalsurveillanceofrabiesintheAmericas.SeptemberMonthly report.6p. 18.SikesRK(1970)Guidelinesforthecontrolofrabies.AmJPublicHealthNations Health60:1133–1138. 19.LeedomJM(2006)Milkofnonhumanoriginandinfectiousdiseasesinhumans. ClinInfectDis43:610–615. 20.NelL,LeRouxK,AtlasR(2009)MeetingtherabiescontrolchallengeinSouth Africa.Microbe4:61–65. 21.GibbonsRV(2002)Cryptogenicrabies,bats,andthequestionofaerosol transmission.AnnEmergMed39:528–536. 22.MeltzerMI,RupprechtCE(1998)Areviewoftheeconomicsoftheprevention andcontrolofrabies.Pharmacoeconomics4:481–498. 23.PetersonAT(2001)Predictingspecies’geographicaldistributionsbasedon ecologicalnichemodeling.Condor103:599–605. 24.PetersonAT(2006)Ecologicnichemodelingandspatialpatternsofdisease transmission.EmergingInfectDis12:1822–1826. 25.PetersonAT(2007)Ecologicalnichemodellingandunderstandingthe geographyofdiseasetransmission.VetItal43:393–400. 26.BlackburnJK,McNysetKM,CurtisA,Hugh-JonesME(2007)Modelingthe geographicdistributionof Bacillusanthracis ,thecausativeagentofanthrax disease,forthecontiguousUnitedStatesusingpredictiveecologicniche modeling.AmJTropMedHyg77:1103–1110. 27.Machado-MachadoEA(2012)Empiricalmappingofsuitabilitytodenguefever inMexicousingspeciesdistributionmodeling.ApplGeogr33:82–93. 28.SarkarS,StrutzSE,FrankDM,RivaldiCL,SisselB,etal.(2010)Chagas diseaseriskinTexas.PLoSNeglectedTropDisdoi:10.1371/journal.pntd.0000836 29.MurrayKA,RetallickRWR,PuschendorfR,SkerrattLF,RosauerD,etal. (2011)Assessingspatialpatternsofdiseaserisktobiodiversity:implicationsfor themanagementoftheamphibianpathogen, Batrachochytriumdendrobatidis .JAppl Ecol48:163–173. 30.NeerinckxSB,PetersonAT,GulinckH,DeckersJ,LeirsH(2008)Geographic distributionandecologicalnicheofplagueinsub–SaharanAfrica.IntJHealth Geographicsdoi:10.1186/1476–072X–7–54 31.PetersonAT,BauerJT,MillsJN(2004)Ecologicandgeographicdistributionof FilovirusDisease.EmergingInfectDis10:40–47.NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org8August2012|Volume7|Issue8|e42466

PAGE 9

32.KuzminIV,BozickB,GuagliardoSA,KunkelR,ShakJR,etal.(2011)Bats, emerginginfectiousdiseases,andtherabiesparadigmrevisited.EmergHealth ThreatsJdoi:10.3402/ehtj.v4i0.7159 33.LordRD(1980)Anecologicalstrategyforcontrollingbovinerabiesthrough eliminationofvampirebats.In:Proceedingsofthe9thVertebratePest Conference.VertebrPestConfProcPaper22:170–175. 34.GomesMN,MonteiroAMV,LewisN,GoncalvesCA,FilhoVDN(2010) LandscaperiskfactorsforattacksofvampirebatsoncattleinSaoPaulo,Brazil. PrevVetMed93:139–146. 35.RebeloH,TarrosoP,JonesG(2010)Predictedimpactofclimatechangeon Europeanbatsinrelationtotheirbiogeographicpatterns.GlobalChangeBiol 16:561–576. 36.MistryS,Moreno-ValdezA(2008)ClimateChangeandBats.BATSMagazine 26:8–11. 37.GreenhallAM,JoermannG,SchmidtW(1983) Desmodusrotundus .Mamm Species202:1–6. 38.GlobalBiodiversityInformationFacility.Available:http://www.gbif.org/. Accessed2011Sept29. 39.RiosNE,BartHL(2010)GEOLocate(Version3.22)computersoftware.Tulane UniversityMuseumofNaturalHistory,BelleChasse,LA.Available:http:// www.museum.tulane.edu/geolocate/.Accessed2011October24. 40.ESRI(2011)ArcGISDesktop:Release10.EnvironmentalSystemsResearch Institute,Redlands,CA. 41.JonesPG,ThorntonPK,HeinkeJ(2009)Generatingcharacteristicdaily weatherdatausingdownscaledclimatemodeldatafromtheIPCC’sFourth Assessment.Available:http://www.ccafs-climate.org/documentation/.Accessed2011December15. 42.SolomonS,QinD,ManningM,ChenZ,MarquisM,etal.(2007)Climate change2007:Thephysicalsciencebasis.ContributionofWorkingGroupIto theFourthAssessmentReportoftheIntergovernmentalPanelonClimate Change.Cambridge:CambridgeUniversity.996p. 43.IPCC(2007)ClimateChange2007:SynthesisReport.ContributionofWorking GroupsI,II,andIIItotheFourthAssessmentReportoftheIntergovernmental PanelonClimateChange.CoreWritingTeam,PachauriRK,ReisingerA, editors.Sitzerland:IPCC.104p. 44.HijimansRJ,CameronSE,ParraJL,JonesPG,JarvisA(2005)Veryhigh resolutioninterpolatedclimatesurfacesforgloballandareas.IntJClimatol25: 1965–1978.Available:http://worldclim.org/.Accessed2011Sept29. 45.PhillipsSJ,AndersonRP,SchapireRE(2006)Maximumentropymodelingof speciesgeographicdistributions.EcolModel190:231–259. 46.PhillipsSJ,Dud ´ kM(2008)ModelingofspeciesdistributionswithMaxent:new extensionsandacomprehensiveevaluation.Ecography31:161–175. 47.ElithJ,GrahamCH,AndersonRP,Dud ´ kM,FerrierS,etal.(2006)Novel methodsimprovepredictionofspecies’distributionsfromoccurrencedata. Ecography29:129–151. 48.ElithJ,PhillipsSJ,HastieT,Dud ´ kM,CheeYE,etal.(2011)Astatistical explanationofMaxEntforecologists.DiversDistrib17:43–57. 49.PearsonRG,DawsonTP(2003)Predictingtheimpactsofclimatechangeonthe distributionofspecies:arebioclimateenvelopemodelsuseful?Globalecol Biogeogr12:361–371. 50.Arau ´joMB,PearsonRG,ThuillerW,ErhardM(2005)Validationofspeciesclimateimpactmodelsunderclimatechange.GlobalChangeBiol11:1504– 1513. 51.PetersonAT,ShawJ(2003)Lutzomyiavectorsforcutaneousleishmaniasisin SouthernBrazil:ecologicalnichemodels,predictedgeographicdistributions, andclimatechangeeffects.IntJParasitol33:919–931. 52.BradleyBA,WilcoveDS,OppenheimerM(2010)Climatechangeincreasesrisk ofplantinvasionintheEasternUnitedStates.BiolInvasions12:1855–1872. 53.SaupeEE,PapesM,SeldenPA,VetterRS(2011)Trackingamedically importantspider:climatechange,ecologicalnichemodeling,andtheBrown Recluse( Loxoscelesreclusa ).PLoSONE6,e17731.doi:10.1371/journal.pone.0017731 54.PetersonAT,Sa ´nchez-CorderoV,Sobero ´nJ.BartleyJ,BuddemeierRW,etal. (2001)EffectsofglobalclimatechangeongeographicdistributionsofMexican Cracidae.EcolModel144:21–30. 55.LoarieSR,CarterBE,HayhoeK,McMahonS,MoeR,etal.(2008)Climate changeandthefutureofCalifornia’sendemicflora.PLoSONE3,e2502. doi:10.1371/journal.pone.0002502 56.PearsonRG(2007)Species’DistributionModelingforConservationEducators andPractitioners.Synthesis.AmericanMuseumofNaturalHistory.Available: http://ncep.amnh.org. 57.ElithJ,KearneyM,PhillipsS(2010)Theartofmodelingrange-shiftingspecies. Methodsinecolevol1:330–342. 58.FieldingAH,BellJF(1997)Areviewofmethodsfortheassessmentofprediction errorsinconservationpresence/absencemodels.EnvironConserv24:38–49. 59.LoboJM,Jime ´nez-ValverdeA,RealR(2008)AUC:amisleadingmeasureofthe performanceofpredictivedistributionmodels.GlobalEcologyandBiogeography17:145–151. 60.PetersonAT,Papes ¸M,Sobero ´n(2008)Rethinkingreceiveroperatingcharacteristicanalysisapplicationsinecologicalnichemodeling.EcolModel213:63– 72. 61.Jime ´nez-ValverdeA(2012)Insightsintotheareaunderthereceiveroperating characteristiccureve(AUC)asadiscriminationmeasureinspeciesdistribution modelling.GlobalEcologyandBiogeography21:498–507. 62.WintGRW,RobinsonTP(2007)Griddedlivestockoftheworld,Foodand AgricultureOrganizationoftheUnitedNations,Rome:131p. 63.ProMED-mail.Rabiesvirus,cows–Mexico,CentralAmerica,SouthAmerica. Available:http://www.promedmail.org.Accessed2012July21. 64.McNabBK(1973)Energeticsandthedistributionofvampires.JMammal54: 131–144. 65.Flores-CrespoR,Arellano-SotaC,(1991)Biologyandcontrolofthevampirebat.In:BaerGMeditor.Thenaturalhistoryofrabies.Florida:CRCPressInc. pp.461–476. 66.HitchAT,LebergPL(2007)BreedingdistributionsofNorthAmericanbirds speciesmovingnorthasaresultofclimatechange.ConservBiol21:534–539. 67.KobayashiY,SatoG,MochizukiM,HiranoS,ItouT,etal.(2008)Molecular andgeographicanalysesofvampirebat–transmittedcattlerabiesincentral Brazil.BMCVetResdoi:10.1186/1746–6148–4–44 68.AguiarLMS,BritoD,MachadoRB(2010)Docurrentvampirebat( Desmodus rotundus )populationcontrolpracticesposeathreattoDekeyser’snectarbat’s ( Lonchophylladekeyseri )longtermpersistenceintheCerrado?ActaChiropterol12: 275–282. 69.ThompsonRD,MitchellGC,BurnsRJ(1972)Vampirebatcontrolbysystemic treatmentoflivestockwithananticoagulant.Science177:806–808. 70.GonzalezR,MitchellGC(1976)VampirebatcontrolprogramsinLatin America.In:Proceedingsofthe7thVertebratePestConference.VertebrPest ConfProcPaper22:254–257. 71.Arellano-SotaC(1988a)Biology,ecology,andcontrolofthevampirebat.Rev InfectDis10:Suppl4,S615–S619. 72.SternerRT,SunB(2004)Relativefactorcostsofwildliferabiesimpactsinthe U.S.In:Proceedings21stVertebratePestConference.VertebrPestConfProc Paper390:185–189. 73.Mart ´ nez–BurnesJ,Lo ´pezA,Medell ´ nJ,HainesD,LozaE,etal.(1997)An outbreakofvampirebat–transmittedrabiesincattleinnortheasternMexico. CanVetJ38:175–177. 74.AlmeidaMF,MartorelliLFA,AiresCC,SallumPC,MassadE(2005)Indirect oralimmunizationofcaptivevampires, Desmodusrotundus .VirusRes111:77–82. 75.vanderZijppJ(2003)FutureoflivestockproductioninLatinAmericaandcrosscontinentaldevelopments.ArchivosLatinoamericanosdeProductio ´nAnimal. 11:50–56.NicheModelingofVampireBatsandCattle PLOSONE|www.plosone.org9August2012|Volume7|Issue8|e42466


printinsert_linkshareget_appmore_horiz

Download Options

close


  • info Info

    There are only PDFs associated with this resource.

  • link PDF(s)



Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.