Environmental DNA. 2019;00:1–10.  | 1wileyonlinelibrary.com/journal/edn3

Received:24February2019  |  Revised:15August2019  |  Accepted:12September2019DOI: 10.1002/edn3.39

M E T H O D S A N D S O F T W A R E

Development of highly sensitive environmental DNA methods for the detection of Bull Sharks, Carcharhinus leucas (Müller and Henle, 1839), using Droplet Digital™ PCR

Katherine E. Schweiss1  | Ryan N. Lehman1 | J. Marcus Drymon2 | Nicole M. Phillips1

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycited.©2019TheAuthors.Environmental DNApublishedbyJohnWiley&SonsLtd

1SchoolofBiological,Environmental,andEarthSciences,TheUniversityofSouthernMississippi,Hattiesburg,Mississippi2CoastalResearchandExtensionCenter,MississippiStateUniversity,Biloxi,Mississippi

CorrespondenceKatherineE.Schweiss,SchoolofBiological,Environmental,andEarthSciences,TheUniversityofSouthernMississippi,Hattiesburg,Mississippi39406.Email:katieeschweiss@gmail.com

Funding informationUniversityofSouthernMississippi;InstitutionalDevelopmentAwardNationalInstitutesofGeneralMedicalSciencesoftheNIH,Grant/AwardNumber:P20‐GM103476

AbstractBackground: Asapexandmesopredators,elasmobranchsplayacrucialroleinmain‐tainingecosystemfunctionandbalanceinmarinesystems.Elasmobranchpopulationsworldwideareindeclineasaresultofexploitationviadirectandindirectfisheriesmortalitiesandhabitatdegradation;however,alackofinformationondistribution,abundance,andpopulationbiologyformostspecieshinderstheireffectivemanage‐ment.EnvironmentalDNAanalysishasemergedasacost‐effectiveandnon‐invasivetechniquetofillsomeofthesedatagaps,butoftenrequiresthedevelopmentofspe‐cies‐specificmethodologies.Aims: Here,weestablishedeDNAmethodologyappropriatefortargetedspeciesde‐tectionsofBullSharks,Carcharhinus leucas,inestuarinewatersinthenorthernGulfofMexico.Materials and Methods: WecompareddifferentQIAGEN®DNeasy® extractionkitprotocolsanddevelopedaspecies‐specificDropletDigital™PCR(ddPCR)assaybydesigningprimersandaninternalprobetoamplifya237basepairportionoftheND2geneinthemitochondrialgenomeofC. leucas.Tovalidatethedevelopedmethods,watersampleswerecollectedfromknownC. leucashabitatandfromanexsituclosedenvironmentcontainingasingleC. leucasindividual.TheeffectivenessoftheassayinanopenenvironmentwasthenassessedbyplacingoneC. leucasintoaflow‐throughmesocosmsystemandwatersampleswerecollectedevery30minfor3hr.Results: ThedevelopedC. leucas‐specificassayhastheabilitytodetecttargetDNAconcentrationsinareactionaslowas0.6copies/μl.DdPCRreactionsperformedonwatersamplesfromknownhabitatand30minafterasharkwasaddedtotheclosedenvironmentcontained1.62copies/μland166.6copies/μloftargetC. leucas eDNA,respectively.Carcharhinus leucas eDNAwas detected in the flow‐through systemwithin30min,butconcentrationsremainedlowandvariablethroughoutthedura‐tionoftheexperiment.

K E Y W O R D S

elasmobranch,GulfofMexico,habitatuse,mitochondrialgenome,threatenedspecies,watersample

2  |     SCHWEISS Et al.

1  | INTRODUC TION

Elasmobranchs (sharks, skates, and rays) play a crucial role inma‐rineecosystemsasapexandmesopredators,influencingpreyabun‐dance, behavior, and trophic interactions across multiple trophiclevels inmarinefoodwebs (Ferretti,Worm,Britten,Heithaus,andLotze2010;Ritchieetal.2012).Healthyelasmobranchpopulationshelptomaintainecosystemfunction,increasebiodiversity,andbuf‐feragainstinvasivespeciesandtransmissionofdiseases(Heithaus,Frid,Wirsing,andWorm2008;Ritchieetal.2012).However,manyelasmobranchpopulationsare indeclineasaresultofexploitationviadirectandindirectfisheriesmortalitiesandhabitatdegradation(Dulvyetal.2014).Thelifehistorystrategiesofmanyelasmobranchsare characterized by late maturity, longevity, and low fecundity,making the recovery of exploited populations a biologically slowprocess (Garcia et al., 2008;Hoenig andGruber1990).Accordingto the InternationalUnion forConservationofNature (IUCN)RedListofThreatenedSpecies,one‐quarterofelasmobranchspeciesareestimatedtobethreatenedwithextinctionandalmostone‐halfarecategorizedasDataDeficient,meaningthereareinsufficientdatatoproperlyassesstheirconservationstatus(Dulvyetal.2014).Robustdata on species distribution, abundance, biology, and populationbiologyarenecessarytoenactappropriateconservationstrategiesforthemaintenanceofhealthyelasmobranchpopulations;unfortu‐nately,suchdataareoften incompleteor lackingformanyspecies(Dulvyetal.2014).

Analysis of environmental DNA (eDNA) has recently emergedasanalternative,powerfulapproachtofilldatagapsonthedistri‐bution, habitat use, abundance, andpopulationbiologyof aquaticspecies (Ficetola,Miaud,Pompanon,andTaberlet2008), includingelasmobranchs(Sigsgaardetal.2016).AllorganismsleavetracesofDNA in the environment through sheddingof cellular debris, skincells, blood, and biologicalwaste, all ofwhich can be collected inwatersamples (Rees,Maddison,Middleditch,Patmore,andGough2014); however, differences in how organisms shed DNA (i.e.,mucus, scales, feces) suggest that eDNA accumulationmay differacross species (Le Port, Bakker, Cooper, Huerlimann, andMariani2018), requiring taxon‐specific research. In targeted species de‐tections,watersamplesaretypically filtered,DNAextractionsareperformedontheresultingparticulatematerial,andextractedDNAsamples are analyzed using a quantitative real‐time polymerasechain reaction (qRT‐PCR) platform with species‐specific primers,developed to amplify a smallDNA fragment in the target species(Foote et al. 2012; Taberlet, Coissac, Hajibabaei, and Rieseberg2012). The collection ofwater samples is a cost‐effective and ef‐ficientmethodofsurveyingelasmobranchpopulationswhencom‐paredtotraditionalsurveymethodsinvolvingsettingnetsorlines,whichcanhavehighincidenceofbycatchandinflictvaryingdegreesofstress tobothtargetandnontargetspecies (Larsonetal.2017;Lewison,Crowder,Read,andFreeman2004).Post‐releaserecoveryandsurvivaltendstovarywidelyacrossspecies,withsomespeciesbeingparticularlysensitivetonetcaptureandhandling(Stobutzki,Millter,Heales, andBrewer2002).With awell‐designed sampling

scheme,eDNAmethodologiesofferincreasedsensitivityfordetect‐ingthepresenceofrarespecieswhilenegatingtheneedtocapture,handle,orevenobservethetargetspecies(Portetal.2016;Reesetal.2014).Inelasmobranchs,eDNAmethodshavebeenusedintar‐geted speciesdetections for theCriticallyEndangeredLargetoothSawfish,Pristis pristis (Simpfendorfer et al. 2016), theEndangeredMaugeanSkate,Zearaja maugeana(Weltzetal.2017),theVulnerableChileanDevilRay,Mobula tarapacana (Garganetal.2017),andtheVulnerable Great White Shark, Carcharodon carcharias (Lafferty,Benesh,Mahon, Jerde, and Lowe 2018). Furthermore, eDNA hasbeen used to assess population characteristics in the EndangeredWhaleshark,Rhincodon typus(Sigsgaardetal.2016)andtoestimatesharkdiversityintropicalhabitatsusingmetabarcoding(Bakkeretal.2017;Boussarieetal.2018).

Bull Sharks,Carcharhinus leucas (Müller and Henle, 1839), arefoundintemperate,subtropical,andtropical latitudesgloballyandaredistinctiveasoneofonlyafewsharksthatcanusefreshwaterforextendedperiodsof time (Thorson1962;Thorson1971;Thorson,Cowan,andWatson1973).Asupper trophic levelpredators, theyplayacrucialroleinmaintainingecosystemhealthacrossbothma‐rineandfreshwaterhabitats(Every,Pethybridge,Fulton,Kyne,andCrook 2017; Polovina, Abecassis, Howell, andWoodworth 2009;Ritchieetal.2012).Usingacoustic telemetrydata toexamine thehabitatuseofC. leucasinnorthernGulfofMexicowaters,Drymonet al. (2014) foundC. leucasmaypreferentially selecthigher‐qual‐ity,less‐urbanizedrivers,althoughaspatiallylimitedacousticarrayhinderedafullevaluationofthispattern.TargetedeDNAsurveysofC. leucascouldprovideacost‐effective,sensitivemethodtoexam‐inethispatternmorewidely,astherecouldbesubstantialecologicalimplicationsofsuchhabitatpreference.Here,weestablishaneDNAmethodologyappropriatefortargetedspeciesdetectionsofC. leu‐casinestuarinewatersinthenorthernGulfofMexico.Specifically,we compare total eDNA yields for different QIAGEN® DNeasy® DNAextractionkitprotocolsanddevelopaspecies‐specificC. leu‐caseDNAassayusingarelativelynovel,Bio‐Rad®DropletDigital™PCR (ddPCR), platform to detect low quantities of target DNA.Finally,weapplythesemethodsto investigatethedetectabilityofC. leucaseDNAinknownhabitatinthenorthernGulfofMexicoandinexsituclosedandflow‐throughenvironmentscontainingasingleC. leucas individual.

2  | MATERIAL S AND METHODS

2.1 | Laboratory controls

Strictlaboratorycontrolswereimplementedthroughoutthisstudytoreducetheriskofcross‐contaminationandcontaminationbyex‐ogenous DNA (see Deiner,Walser, Mächler, and Altermatt 2015;Goldberg et al. 2016). Water processing, DNA extractions, andPCR amplifications were conducted in physically separated labo‐ratory spaces to prevent cross‐contamination between stages.Negative controls were incorporated into every stage of sampleprocessing,andPCRwasperformedonthemtocheckforpotential

     |  3SCHWEISS Et al.

contamination. Filter negatives contained target‐free, autoclaveddeionized water, DNA extraction negatives contained no filteredparticulatematerial,andPCRamplificationnegativescontainednoDNA;allnegativecontrolsproducednegativeresults,indicatingnocontaminationhadoccurred.TheddPCRassay conditionsused tocarryoutthesenegativecontroltestsaredescribedbelow.

2.2 | Water sample collection and filtration

Water samples throughout this studywerecollected justbelow thesurfaceofthewater in1Lhigh‐densitypolyethyleneNalgene®bot‐tlesprecleanedina10%bleachsolutionandsanitizedunderultravio‐let(UV)lightfor20min.Newgloveswereusedtocollecteachwatersampleandsampleswerestoredoniceinacooleruntilfiltrationusingavacuumpumpcouldtakeplace,whichoccurredwithin24hrofcol‐lection(seePilliodetal.2013),exceptwhereotherwisenoted.Watersampleswerefilteredinadedicated,precleanedlaboratoryspacethathad never had C. leucastissueortotalgenomicDNA(gDNA)present.Each1Lwatersamplewasinvertedatleastthreetimestoensureho‐mogenizationofparticulatematterandwasthenvacuum‐filteredusing47‐mm‐diameter, 0.8‐μm nylon filters, which were replaced whencloggingoccurredevery~350ml (e.g., threefiltersper1L)andpre‐servedin95%ethanolatroomtemperature,unlessnotedotherwise(seeAppendixS1).Duringallwaterfiltration,filterswerehandledwith

designatedsterileforcepsforeachsampleandgloveswerechangedinbetweensamplestoavoidcross‐contamination.

2.3 | DNA extraction methods

Due to thewidevarietyofDNAextractionmethodsused ineDNAliterature(Renshaw,Olds,Jerde,McVeigh,andLodge2015),wecom‐paredeDNAextractionkitstoestablishanappropriatemethodforthenylonfiltersusedtofilterwatersamplesinthisstudy.TheQIAGEN® DNeasy®Blood&TissueKitisafrequentchoiceforDNAextractionsfromfiltersineDNAstudies,butwithnumerousvariations(seeReesetal.2014).TheperformanceofthiskitusingtheGoldbergetal.(2011)variationincorporatingQIAshredder™spincolumnswascomparedtothatofanextractionkitdesignedspecificallyforwatersamples,theQIAGEN®DNeasy®PowerWater®Kit.TheGoldbergetal.(2011)pro‐tocolincorporatingQIAshredder™spincolumnswasselectedbecauseinpreliminarytrials,ityieldedhigherrelativequantitiesofDNAcom‐paredtosomeothervariations(AppendixS2).Additionally,fourvaria‐tionsofphysicaldisruptionmethodstodislodgetheparticulatematterfrom the filters prior to digestionwere testedwith each extractionmethod:(a)nophysicaldisruption,(b)beadbeating,(c)filterscraping,and(d)freezingfilterswithliquidnitrogenandcrushingthemusinganautoclavedmortarandpestle.TheQIAGEN®DNeasy®PowerWater® Kitcontainedbeadbeatingaspartofthestandardmanufacturer'spro‐tocol,sothisstepwaseliminatedforthenophysicaldisruptionvaria‐tiontodetermineifthisstepwasacriticalfactorinDNAyields.Three×1Lwatersamplereplicateswereused ineachextraction/physicaldisruption treatment, collected fromMobileBay,Alabamausing thewatercollectionandfiltrationprotocolsdescribed.Toeliminatethefil‐terpreservationstep,thefiltersforeach1Lsamplewereimmediatelyplaced into theappropriate lysisbuffers (seeHinloetal.2017).TheDNAextractsforeach1LwatersamplewerecombinedandtheDNAqualitieswereassessedusing2%agarosegelandtherelativequanti‐tiesweremeasuredusingThermoFisherScientificNanoDrop™spec‐trophotometertechnology,witheachextractmeasuredfourtimes.

2.4 | Development of a species‐specific assay

Todevelopaspecies‐specificassay,primersandan internalprobeweremanuallydesignedinconservedregionsofthemitochondrial(mtDNA)NADHdehydrogenase2(ND2)genewithinC. leucas,butvariableregionsacross23geneticallysimilar,exclusionelasmobranchspecies, using sequences available fromGenBank and aligned viaCodonCodeAlignerv.7.0(seeAppendixS3).Forward(BULLND2F6:5′‐TCCGGGTTTATACCCAAATG‐3′) and reverse (BULLND2R5: 5′‐GAAGGAGGATGGATAAGATTG‐3′) primers were designed firstto PCR‐amplify a 237base pair portion of themtDNAND2genein C. leucas. The primers were first tested using gDNA extractedfromfiveC. leucasindividualsfromnorthernGulfofMexicowatersusingconventionalPCR.EachPCRconsistedof10mMTAQbuffer,1.5mMMgCl2,0.3μMofeachprimer,0.1mMdNTPs,1UofTaqpol‐ymerase,~25ng/μlofeachDNAextract,andPCR‐gradewaterforafinalreactionvolumeof25μl.PCRcyclingconditionsbeganwith

TA B L E 1  EighteengeneticallysimilarexclusionelasmobranchspeciesfoundintheGulfofMexico

Common name Species name

NurseShark Ginglymostoma cirratum

ShortfinMako Isurus oxyrinchus

DuskySmoothhound Mustelus canis

TigerShark Galeocerdo cuvier

GreatHammerhead Sphyrna mokarran

ScallopedHammerhead Sphyrna lewini

Bonnethead Sphyrna tiburo

AtlanticSharpnoseShark Rhizoprionodon terraenovae

LemonShark Negaprion brevirostris

FinetoothShark Carcharhinus isodon

BlacknoseShark Carcharhinus acronotus

SandbarShark Carcharhinus plumbeus

SpinnerShark Carcharhinus brevipinna

DuskyShark Carcharhinus obscurus

SilkyShark Carcharhinus falciformis

BlacktipShark Carcharhinus limbatus

CownoseRay Rhinoptera bonasus

AtlanticStingray Hypanus sabina

Note: These18geneticallysimilarexclusionspecies,andCarcharhinus leucas,weretestedforspecies‐specificityofthedevelopedprimersandinternalprobeontheBio‐Rad®QX200™DropletDigital™PCRplat‐form.AlltissuesampleswerecollectedfromtheGulfofMexico.

4  |     SCHWEISS Et al.

initialdenaturationat94°Cfor5min,followedby35cyclesof94°Cfor30s,59°Cfor30s,and72°Cfor30s,finalextensionat72°Cfor7min,andafinalholdat4°C.Primerswerealsotestedagainstoneindividualofeachof18othergeneticallysimilar,localexclusionspecies,collectedfromtheGulfofMexico(Table1)toassessspeci‐ficity.TheprimersamplifiedDNAinthetargetspecies,C. leucas,butalsoamplifiedDNAfromsomeofthenontargetspeciestested.Toincreasethespeciesspecificityoftheassay,aninternalPrimeTime® double‐quenched ZEN™/IOWA Black™ FQ probe labeled with 6‐FAM(BULL_IBFQ:5’‐CAACACTAACTATAAGTCCTAACCCAATC‐3’)wasdesignedtoamplifythetargetgeneinonlyC. leucas.

DdPCR reaction mixtures and cycling conditions were opti‐mized forC. leucas by systematically adjusting the concentrationsofprimers(300–1,000nM)andinternalprobe(100–250nM),cyclenumber (30–40 cycles), ramp rate (0.5–2.0°C/s), annealing tem‐perature (54–66°C), elongation time (1–2min), and the amountofgDNA(0.2–25.0ng/μl).TheoptimizedddPCRmixturecontained1XBio‐Rad®ddPCRsupermixforprobes(nodeoxyuridinetriphosphate(dUTP)),750nMofeachprimer,and250nMofprobe,and1.1μlofextractedDNA,adjustedtoafinalvolumeof22μlwithPCR‐gradewater.DdPCRdropletsweregeneratedforeach22μlreactionusingthe Bio‐Rad® QX200™ AutoDG™ Droplet Digital™ PCR System(Instrumentno.773BR1456)andthermalcyclingconditionswereasfollows,usingaramprateof1°C/s:initialdenaturationat95°Cfor10min,followedby35cyclesof94°Cfor30sand56°Cfor2min,followedbyenzymedeactivationat98°Cfor10min,andafinalholdat4°C.Toensuretheoptimizedassaywasspecies‐specificforC. leu‐casusing theddPCRplatform, theprimersandprobewere testedusingtheseddPCRreactionandcyclingconditions,inreplicatesofthree,with0.2ng/μlofgDNAextractedfromfiveC. leucas individ‐ualsandoneindividualofeachof18othergeneticallysimilar,localexclusionspecies,collectedfromtheGulfofMexico(Table1).

All ddPCR data were analyzed with the Bio‐Rad® QX200™Droplet Reader and QuantaSoft™ software using the Rare EventDetection (RED) analysis, a manual detection threshold of 3,000amplitude(Figure1),andalimitofdetection(LoD)ofthedevelopedassay.TheLoDisconsideredthelowestconcentrationofC. leucas DNAthatcanreliablybedetectedusingtheoptimizedassaycondi‐tions.The lowerLoDwasdeterminedbyconductingddPCRswithgDNA from twoC. leucas individualsusing a sixfold seriesof10Xdilutions(e.g.,1:10to1:1,000,000),fromastartingconcentrationof25.0ng/μl.MeansandstandarderrorsofdetectedDNAconcentra‐tion(copies/μl)werecalculatedforeachindividual,acrossthethreeddPCRreplicatesforeachdilution.

2.5 | Collection of positive water samples

Carcharhinus leucaseDNAsampleswereobtainedviathecollec‐tion ofwater samples from knownC. leucas habitat and ex situexperiments. These experimentswere conducted in accordancewiththelawsofthestateofAlabamaandundertheIACUCproto‐cols(IACUCProtocolNumber974304).Allmeasuresweretakentoreducethepainorstresstheanimalunderwentduringtesting;

therefore, thewaterused in theex situexperimentswere fromnaturalsharkhabitat.WaterwascollectedfromthecoastalwatersofMobileBay,Alabama,knownC. leucashabitat,inApril2017andplaced into aprecleaned, circular fiberglass, closed‐system tank(~120cmwideandheldavolumeof~711L),andsix×1LwatersampleswereimmediatelycollectedfromthistanktodeterminewhethertargeteDNAwaspresent in theambientwater.Abub‐blerwasadded to the tank tokeep the systemoxygenatedandone wild‐caught juvenile maleC. leucas, ~930mm total length,wasaddedtothetank.ToacquireaconfirmedpositiveC. leucas eDNAsample,after30min, six×1Lwatersampleswereagaincollected from the tank. Thesewater sampleswere used in as‐pectsofmethoddevelopment (seeAppendixS1)andtovalidatethedevelopedgeneticassay.

TotesttheeffectivenessofthedevelopedC. leucasassayinanopensystemwithasingletargetspeciespresent,aflow‐throughmesocosm (~365 cmwide containing a volume of ~14,500 L) atDauphin Island Sea Lab, Alabamawasmaintained inApril 2017.The flow rate of the mesocosm was designed to mimic flow ina coastal system at ~30 cm3/hr,with complete system turnoverat approximately 2 hr. One wild‐caught juvenile male C. leucas,~930mmtotal length,was introduced to this systemand five×

F I G U R E 1  RawoutputoftheoptimizedDropletDigital™PCR(ddPCR)forthedesignedCarcharhinus leucasspecificassayshowingoneddPCRreplicateforoneindividual(0.2ng/μlofgenomicDNA)andonereplicatefortheddPCRnegativefromtheBio‐Rad® QX200™DropletReader.Eachdropletineachwellwasclassifiedaseitherpositive(bluedroplets)ornegative(graydroplets)fortargetDNA,basedonamanualdetectionthresholdsetto3,000amplitude(thehorizontalpinkline)usingtheQuantaSoft™RareEventDetectionanalysis.Eventnumberreferstothenumberofdropleteventsgeneratedforagivenwellorsample;Ch1amplitudemeasurementreferstotheleveloffluorescenceemittedbyadropletevent;andeachcolumnisasinglewell

     |  5SCHWEISS Et al.

1Lwatersampleswerecollectedimmediately(time0.0),spanningthediameterofthemesocosm;thissamplingregimewasrepeatedevery0.5hrfor3hr,allowingforcompleteturnoverofthesystem.Watersampleswerestoredina−20°Cfreezerfor1month,duetolaboratoryequipmentconstraints, similar toBakkeret al. (2017)andGarganetal. (2017),andwerethawedat roomtemperaturepriortofiltration.

Water samples from these experiments were vacuum‐filteredusing 47‐mm‐diameter nylon 0.8‐μm filters (three per 1 L), whichwere preserved in 95% ethanol at room temperature (AppendixS1) andDNAextractions followed theGoldberget al. (2011)pro‐tocolincorporatingtheQIAshredder™spincolumns(AppendixS2).DdPCR amplificationswere carried out in replicates of five, usingtheoptimizedC. leucasassaypreviouslydescribedinthisstudy.AllddPCRreactionsweresetupusingaerosolbarrierfilterpipettetipsanddesignatedpipettes,separatefromthoseusedinsettingupPCRreactions,wereusedtoaddeDNAextractstothereactions.DdPCRresultswereanalyzedusingtheBio‐Rad®QX200™DropletReaderand QuantaSoft™ RED analysis, a manual detection threshold of3,000amplitude,andtheLoD.

3  | RESULTS

3.1 | Optimal eDNA methods

TheGoldbergetal. (2011)protocolusingtheQIAGEN®DNeasy® Blood&TissueKitandQIAshredder™spincolumnsyieldedhigherrelative quantities of total eDNA from filters compared to theQIAGEN®DNeasy®PowerWater®Kitprotocol,acrossallvariationsinphysicaldisruptionmethods(Figure2).TheDNAyieldsfromthefour physical disruption methods used with the Goldberg et al.(2011)protocolweresimilar:NophysicaldisruptionyieldedatotalDNA average of 61.19 ng/μl (SE = 1.65), bead beating the filtersyielded56.83ng/μl (SE = 6.75), filter scraping yielded56.78ng/μl (SE=1.77),and freezing filterswith liquidnitrogenandcrush‐ingyielded64.93ng/μl(SE=2.36)(Figure2).SincethetotalDNAyieldsweresimilaracrossthesemethodsandbecausetheadditionofaphysicaldisruptionstep is time‐consumingandallowsforanadditionalopportunity forcontaminationbyexogenousDNA,wedeterminedtheoptimalDNAextractionmethodforourpurposestobetheGoldbergetal. (2011)protocolwithnophysicaldisrup‐tionmethod.

Thecombinationofprimersandprobedesignedinthisstudyweredemonstratedtobespecies‐specificforC. leucasinourstudyareabysuccessfullyamplifyingtargetDNAinallddPCRreplicatesfor the fiveC. leucas individuals andnot amplifyingDNA in anyoftheddPCRreplicatesofthe18localexclusionspeciesorPCRnegative controls. The LoD, as determined using the Bio‐Rad® QX200™ Droplet Reader and QuantaSoft™, was the 1:10,000dilution, corresponding to 2.5 pg of targetDNA in the reaction(Figure3).Therewereseveralpositivedropletspresentabovethemanual threshold in the1:10,000dilutions,andthestandarder‐rorsdidnot includezerooroverlapwiththoseofthe1:100,000

dilutions. In contrast, there were no positive droplets in the1:100,000dilutionsandthestandarderrorsoverlappedwithzero,suggestingC. leucas DNA could not be reliably detected at thisdilution (Figure3).Usingthenumberofcopiesof targetDNA/μl in the 1:10,000 dilutions and applying the lower standard errorastherelaxeddetectionthresholdforeachofthetwoindividuals(seeBakeretal.2018),theaverageLoDthresholdwasdeterminedtobe0.6copies/μlinareaction.

3.2 | Analysis of water samples

UsingthedevelopedddPCRassayandtheQuantaSoft™REDanaly‐siswithamanualdetectionthresholdof3,000amplitude,anaverageof1.62copies/μl(SE=0.12)ofC. leucasDNAwasdetectableintheddPCRreactionsfromwatersamplescollectedfromknownhabitat,Mobile Bay,without visually confirming the presence ofC. leucas (Figure4).IntheexsitupositiveeDNAexperiment,30minafteraC. leucaswasaddedtotheclosedtankcontainingthiswater, largeamountsoftargeteDNAwerepresent,withanaverageconcentra‐tionof166.6copies/μl(SE=3.01)intheddPCRreactions(Figure4).Intheflow‐throughmesocosmexperiment,whenapplyingalowerLoDof0.6copies/μltothedataanalysis,targetC. leucasDNAwasnotdetectable inanyof theddPCRreplicatesat time0.0butwasdetectableinallddPCRreplicates0.5hrafterthesharkwasadded(Figure 5). Average target eDNA concentration peaked by 1.0 hr,withanaverageof5.8copies/μl (SE=0.27)acrossallddPCRrep‐licates,andthendeclinedoverthenexthour (Figure5).By2.0hr,theaverageconcentrationofC. leucaseDNAdippedbelowtheLoD,withpositivedetectionsinonlytwoofthefiveddPCRreplicatesforthissample(Figure5).Therewasasecond,smallerspikeinC. leucas eDNAby2.5hr,thatagaindecreased,buttheaverageconcentration

F I G U R E 2  ConcentrationsofDNAextractsfromwatersamplesusingtheQIAGEN®DNeasy®Blood&TissueKitwiththeGoldbergetal.(2011)protocolandtheQIAGEN®DNeasy® PowerWater®Kit,incombinationwithadditionalphysicaldisruptionmethods.SEbarswereusedtoshowtheerrorinmeanDNAconcentrationsbetweencategories,usingfourThermoFisherScientificNanoDrop™spectrophotometerreadingspersample.TheDNAextractsforeach1Lwatersamplewerecombinedandeachcategorycontainedthree×1Lwatersamplereplicates

6  |     SCHWEISS Et al.

oftargetDNAremaineddetectableat3.0hr,althoughonlytwoofthefiveddPCRreplicatesforthissamplehadconcentrationsabovetheLoD(Figure5).

4  | DISCUSSION

TheuseofeDNAasatooltostudythedistributionandecologyofmarinespecieshasincreasedsubstantiallyinrecentyears(Bakkeretal.2017;Footeetal.2012;Laffertyetal.2018;Portetal.2016).However,carefulconsiderationandoptimizationofthemethodsemployedinsuchstudiesarenecessary,ultimatelyallowingforanappropriateinterpretationoftheresults.Here,wefoundfilteringwaterwithnylon0.8‐μmfilters,preservingthefiltersin95%etha‐nol(AppendixS1),andthenperformingDNAextractionsusingtheGoldbergetal.(2011)protocolwiththeQIAGEN®DNeasy®Blood&TissueKitandQIAshredder™spincolumns tobeanappropri‐atemethodofisolatingtotaleDNAfromwatercollectedfromthenorthern Gulf of Mexico. Although the number of replicates intheexperimentwassmall,theGoldbergetal.(2011)protocolwasfoundtooutperformthePowerWater®kitacrossallfourphysicaldisruptionmethods,despitethelatterbeingspecificallydesignedandmarketed for eDNAextractions fromwater samples, andatahighercost.ThetotalDNAyieldsusedtoevaluate theperfor‐mancesof theseextractionmethods areunlikely tobe accurateinanabsolutesenseduetothe inabilityofNanoDrop™spectro‐photometertechnologytodecipherDNAfromotherpossiblebio‐logicalmacromolecules,buttherelativedifferencesbetweenDNAyieldsweresubstantial.Thecombinationofprimersand internalprobefor themtDNAND2genedesigned inthisstudyareopti‐mizedforC. leucasintheestuariesinthenorthernGulfofMexico;however,whethertheyareappropriate(e.g.,species‐specific)foruse in other geographic regions, such as northern Australia, orin fullymarinewaters,wheretheremaybeadditionalspeciesofcloselyrelatedcarcharhinidspresent,requiresfurthertesting.TheLoDdeterminedinthisstudyshowsthesensitivityanddetectioncapabilityofthedevelopedassayandwasdemonstratedtobesuf‐ficientforC. leucaseDNAdetectioninMobileBayandinexsitupositive samples.However, the LoDmay require further refine‐mentthroughadditionaldilutionseriesbetweenthe1:10,000and1:100,000 dilutions before being used in data analysis for largenumbers of field samples. Furthermore, due to potential differ‐encesacrossddPCRmachines,werecommendtheLoDtobere‐fined independently for eachmachine, using the LoD here as astartingreferencepointforthisassay.

The ability of ddPCR to detect low concentrations of targetDNA,forexample,2.5pgofC. leucasDNAinthisstudy,meansthisplatformmaybeless likelytoproducefalsenegativeswhenusedalongside an appropriate sampling regime and water processingmethods(e.g.,spatialanddepthcoverage,volumecollected,filterporesize).FalsenegativescanoccurwhentargetDNAiscapturedinwatersamplesbutisnotdetectedduetolimitationsofthege‐netic assays employed (Darling andMahon 2011; Ficetola et al.2015;Goldbergetal.2016;Lahoz‐Monfort,Guillera‐Arroita,andTingley2016).Todate,themajorityofstudiesthatuseeDNAintar‐getedspeciesdetectionshaveusedqRT‐PCR,butthedetectionca‐pabilitiesofthisplatformmaybelimited,whencomparedtothoseofddPCR(Doi,Takahara,etal.2015;Doi,Uchii,etal.,2015).The

FIGURE 3 Limitofdetection(LoD)testsusinga6‐fold10Xdilutionseries(1:10–1:1,000,000)oftotalgenomicDNA(gDNA)fromtwoCarcharhinus leucasindividualsfromthenorthernGulfofMexico.(a)ThemeanDNAconcentrations(copynumber/μl)andstandarderrorbarswerecalculatedfromthreeDropletDigital™PCR(ddPCR)replicatesforeachoftwoindividuals,usingamanualdetectionthresholdof3,000amplitudeandtheRareEventDetectionanalysissettingontheBio‐Rad®QX200™DropletReaderandQuantaSoft™software.The1:10and1:1,000,000werenotgraphedduetooversaturationofthePCRproduct,andthelackofDNAcopiespresentshowingnopositivedropletdetections,respectively.TheLoD(0.6copies/μl)isrepresentedbyadottedline.(b)RawdropletoutputofddPCRserialdilutionproductsfromoneddPCRreplicateofoneC. leucasindividualdetectedbytheBio‐Rad®QX200™DropletReaderandQuantaSoft™software.Eachdropletineachwellwasclassifiedaseitherpositive(bluedroplets)ornegative(graydroplets)fortargetDNA.EachwellisseparatedbyyellowbarsandcorrespondstothesamedilutionconcentrationsgraphedinFigure3a,labeledwitheachdilutionseriesitrepresents

(a)

(b)

     |  7SCHWEISS Et al.

differenceindetectionabilitiesbetweenthetwoPCRplatformsislikelyduetofundamentaldifferencesinhowtheyquantifytargetDNA.DdPCRquantifiesthestartingDNAcopynumberpresentinasampleusingend‐pointPCRwithoutreferencetoastandard(abso‐lutequantification)(Whaleetal.2012),makingitamoresensitiveand precise assay, ideal for eDNA applications targeting a singletargetspecies.Additionally,theREDanalysissettingusingtheBio‐Rad®QuantaSoft™softwareisdesignedtoidentifylowcopynum‐bersoftargetDNAinabackgroundlargelycomposedofnontargetDNAcopies (Bio‐Rad®DropletDigital™PCRApplicationsGuide).GiventheabilityofddPCRtodetectsuchlowquantitiesofDNA,itmayreplaceqRT‐PCRineDNAresearch(Doi,Uchii,etal.,2015;Nathan,Simmons,Wegleitner, Jerde,andMahon2014)assessingthedistribution,habitatuse,andabundanceofspeciesfoundinlowabundanceand/orareofconservationconcern(Bakeretal.2018;

Hunteretal.2018;Tréguieretal.2014), includingelasmobranchs(Bohmannetal.2014;Laffertyetal.2018).However,wecautionthattheabilitytodetectsuchlowquantitiesofDNAalsoincreasesthe potential for false positives (Goldberg et al. 2016; Huggett,Cowen,andFoy2015).AlleDNAstudies,butespeciallythoseusingddPCR,requirestrictfieldandlaboratorycontrolsandproceduresbeinplacetoreducethepotentialforfalsepositives,typicallytheresultofcontaminationbyexogenousDNAorcross‐contaminationof samples (seeFicetola,Taberlet, andCoissac2016). InadditiontothecontaminationcontrolsdescribedbyGoldbergetal.(2016),Deineretal.(2015),andPortetal.(2016),whenusingddPCR,wealsosuggest: (a)usingtwocleaningmethodsfordecontaminationof all field and water filtration equipment (e.g., a bleach wash,plus autoclaving, and/orUV light exposure), (b) thatwater filtra‐tion isconducted ina laboratoryspacethathasneverhadtissueorgDNAfromthetargetspeciespresent, (c) thatglovesandanytoolsarechangedinbetweensamplesduringwaterfiltration(seePilliod,Goldberg,Arkle,andWaits2013),(d)thatnegativesbein‐corporated into field collection,water filtration,DNAextraction,andPCR,witheachnegativerunthroughtoPCR(seeBakkeretal.2017;Jerde,Mahon,Chadderton,andLodge2011),(e)thatadesig‐natedpipette,separatefromthatusedtosetupreactions,beusedtoaddDNAextractstoddPCRreactions,and(f)thatmultiplerep‐licatesforeachsamplearerunduringddPCR(seeReesetal.2014).Strictfieldandlaboratorycontrolswillensuretheauthenticityandreliabilityof eDNA results,which is increasingly critical in eDNAresearchusinghighlysensitivetechnologies,suchasddPCR,espe‐ciallywhentheresultsofsuchstudieswillbeusedtoinformcon‐servationandmanagement initiatives (Hunteretal.2017;Hunteretal.2018).

F I G U R E 4  RawDropletDigital™PCR(ddPCR)outputfromtheambientwatersampleinMobileBay,theCarcharhinus leucaseDNApositivewatersampletakenfromaclosedsystem30minafteraddingtheshark,andeachnegativecontrolfromtheBio‐Rad® QX200™DropletReader.Eachdropletineachwellwasclassifiedaseitherpositive(bluedroplets)ornegative(graydroplets)fortargetDNAbasedonamanualdetectionthresholdsetto3,000amplitude(thehorizontalpinkline)usingtheQuantaSoft™RareEventDetectionanalysis.Eventnumberreferstothenumberofdropleteventsgeneratedforagivenwellorsample;Ch1amplitudemeasurementreferstotheleveloffluorescenceemittedbyadropletevent;andeachcolumnisasinglewell.Columns,orwells,areseparatedbyyellowbars;ColumnD01correspondstooneddPCRreplicatefromtheambientMobileBaywatersampleandF01correspondstooneddPCRreplicatefromtheC. leucaseDNApositivewatersample.ColumnsB11,A12,andB12correspondtooneddPCRreplicatefromeachnegativecontrolincorporatedandshowsnocontaminationoccurredduringanystageofthisexperiment

F I G U R E 5   Carcharhinus leucasmeaneDNAconcentrations(unitofmeasure)inaflow‐throughmesocosmdetectedusingtheBio‐Rad®QX200™DropletReaderandQuantaSoft™usingamanualdetectionthresholdof3,000amplitudewiththeRareEventDetectionanalysissetting.EachtimepointsamplewasruninDropletDigital™PCR(ddPCR)replicatesoffive,andstandarderrorbarswereusedtoshowthevariationinconcentrationestimatesacrossthefiveddPCRreplicatesforeachsample.Thelowerlimitofdetection,foundtobeatleast0.6copies/μlinthisstudy,isindicatedbyadottedline

8  |     SCHWEISS Et al.

Fundamental researchon the accumulation, persistence, anddegradationofelasmobrancheDNA isnecessary to improve theinterpretation of results in eDNA field research. Here, we haveshown that after adding a shark into closed and flow‐throughsystems,targeteDNAwasdetectablewithin30min.Intheflow‐throughsystem,theinitialspikeintargeteDNAthatoccurredbe‐tween 0.5 and 1.0 hr could be due to initial stress experiencedby the shark after being added to the mesocosm, causing it toexpelmoreDNA (e.g., Barnes et al. 2014). The overall decreaseintargeteDNAbetween1.0and2.0hrmaybetheresultoftheshark acclimating to theenvironment and releasing lessDNAorturnoverofwaterinthemesocosmifthesharkisreleasingDNAintothesysteminpulsesratherthancontinuously;however,thishasnotbeenexplicitlyexploredinelasmobranchs.TheinabilitytodetectC. leucasDNAinsomeoftheddPCRreplicatesat2.0and3.0hr,despitetheconfirmedpresenceofasharkandtheuseofahighlysensitiveddPCRassay,suggeststheremayhavebeenverylittleC. leucasDNApresentat those times,whichcouldoccur ifDNAwasshed inpulses,andthenflowedoutof themesocosm.However, thispatterncouldalsobe indicativeofsamplingerror,where C. leucasDNAwaspresent,butnotcaptured,highlightingtheneed forcarefulconsiderationofsampling regimeaswellasthe interpretation of the results of eDNA studies. Becauseme‐socosmwatersampleswere frozenaftercollection, itcannotbecompletely ruledout that theeDNAdegradedprior to filtration(Hinloetal.2017;Takahara,Minamoto,andDoi2015);however,theconcentrationsofthetotaleDNAextractsfromthesesampleswerenotunusuallylowcomparedtotheothereDNAextractsana‐lyzedforthisstudy.Furthermore,othereDNAstudieshavefrozenwatersamplespriortofiltrationwithoutapparentnegativeeffects(Bakkeretal.2017;Garganetal.2017)makingitunlikelytobethesoleexplanationfor theobservedpatternsofC. leucasDNAde‐tectedinthisexperiment.Ideally,theseexperimentsshouldhavebeenreplicatedand includedasecondtankwithoutasharkasanegative control, with water samples filtered immediately aftercollection;however,duetolimitedfacilitiesandtheconstraintsofusingliveanimals,theseimprovementstothestudydesignwerenotfeasible.Regardless,thisisthefirstelasmobrancheDNAstudythathasplacedasingletargetanimal intoclosedandthenopen,flow‐throughsystemstoquantifytargeteDNAfromasinglean‐imalovertime,creatingabaselinefor futureexsituresearch. Incomparison,othereDNAstudiesofelasmobranchshaveacquiredpositiveeDNAsamplesbycollectingwatersamplesfromaquariawith the target species present (e.g., Simpfendorfer et al. 2016)orcollectingwatersamplesfromknownhabitats,butwithoutvi‐sually confirming the presence of the target species (e.g.,Weltzetal.2017).FuturestudiesshouldassessDNAaccumulationoverdifferent timescales thanpresentedhere,aswellashowalteredflow rates, water conditions (pH, temperature), weather condi‐tions (photoperiod, cloud cover), and number and size of targetspeciesimpacttheaccumulationandpersistenceofelasmobrancheDNAinmarinesystems.

ACKNOWLEDG MENTS

ThisresearchwasfundedbyTheUniversityofSouthernMississippiand was supported by the Mississippi INBRE, funded underInstitutionalDevelopmentAward(IDeA)numberP20‐GM103476from theNational Institutes ofGeneralMedical Sciences of theNationalInstitutesofHealth.WewouldliketoshowourgratitudetoDeanGrubbs,JillHendon,andTobyDaly‐Engelforprovidinguswith elasmobranch tissue samples. Thank you to Emily Seubert,GrantLockridge,andMattJargowskyforcollectionofBullSharkspecimens andmesocosmmaintenance and to Ruth Carmichaelforaccesstofreezerspace.ThankyoutoJoshuaSpeed,LondonWilliams, and Michael Garrett for kindly accommodating us atUniversityofMississippiMedicalCenterandprovidingaccesstoallDropletDigital™PCRequipment.WewouldalsoliketothankGlenmoreShearer,LoganBlancett,andJonathanLindnerforad‐vice,assistance,andaccesstoequipmentusedinthisstudy.HelpprovidedbyElannaThompson,MadisonVitale,andArielWilliamsin collection, filtering, and measurement of water samples wasgreatlyappreciated.

CONFLIC T OF INTERE S T

Theauthorsdeclarethattheyhavenoconflictofinterest.

AUTHOR CONTRIBUTIONS

Allauthorscontributedtotheconceptionanddesignofthestudy,theacquisitionand the interpretationof thedata, andwriting themanuscript.

ORCID

Katherine E. Schweiss https://orcid.org/0000‐0002‐4795‐1933

Nicole M. Phillips https://orcid.org/0000‐0002‐4138‐4966

DATA AVAIL ABILIT Y S TATEMENT

Thedatasetsgeneratedduringand/oranalyzedduringthecurrentstudyareavailablefromN.M.Phillipsonreasonablerequest.

R E FE R E N C E S

Baker,C. S., Steel,D.,Nieukirk, S.,&Klinck,H. (2018). EnvironmentalDNA(eDNA)fromthewakeofthewhales:DropletdigitalPCRfordetection and species identification. Frontiers in Marine Science,5(133),https://doi.org/10.3389/fmars.2018.00133

Bakker, J.,Wangensteen,O.S.,Chapman,D.D.,Boussarie,G.,Buddo,D., Guttridge, T. L., … Mariani, S. (2017). Environmental DNA re‐vealstropicalsharkdiversityincontrastinglevelsofanthropogenicimpact. Scientific Reports, 7(1), 16886. https://doi.org/10.1038/s41598‐017‐17150‐2

Barnes,M.A.,Turner,C.R.,Jerde,C.L.,Renshaw,M.A.,Chadderton,W.L.,&Lodge,D.M.(2014).EnvironmentalconditionsinfluenceeDNA

     |  9SCHWEISS Et al.

persistence inaquaticsystems.Environmental Science & Technology,48(3),1819–1827.https://doi.org/10.1021/es404734p

Bio‐Rad®.Droplet Digital™ PCR Applications Guide.Retrievedfromhttp://www.bio‐rad.com/webroot/web/pdf/lsr/literature/Bulletin_6407.pdf.

Bohmann,K.,Evans,A.,Gilbert,M.T.P.,Carvalho,G.R.,Creer,S.,Knapp,M.,…DeBruyn,M.(2014).EnvironmentalDNAforwildlifebiologyand biodiversitymonitoring. Trends in Ecology and Evolution,29(6),358–367.https://doi.org/10.1016/j.tree.2014.04.003

Boussarie, G., Bakker, J., Wangensteen, O. S., Mariani, S., Bonnin, L.,Juhel,J.B.,…Mouillot,D.(2018).EnvironmentalDNAilluminatesthedarkdiversityof sharks.Science Advances,4(5), eaap9661.https://doi.org/10.1126/sciadv.aap9661

Darling,J.A.,&Mahon,A.R. (2011).Frommolecules tomanagement:Adopting DNA‐based methods for monitoring biological invasionsin aquatic environments.Environmental Research,111(7), 978–988.https://doi.org/10.1016/j.envres.2011.02.001

Deiner,K.,Walser, J.C.,Mächler,E.,&Altermatt,F. (2015).Choiceofcapture and extraction methods affect detection of freshwaterbiodiversity fromenvironmentalDNA.Biological Conservation,183,53–63.https://doi.org/10.1016/j.biocon.2014.11.018

Doi, H., Takahara, T., Minamoto, T., Matsuhashi, S., Uchii, K., &Yamanaka, H. (2015). Droplet digital polymerase chain reaction(PCR) outperforms real‐time PCR in the detection of environ‐mentalDNA froman invasive fish species.Environmental Science & Technology, 49(9), 5601–5608. https://doi.org/10.1021/acs.est.5b00253

Doi, H., Uchii, K., Takahara, T., Matsuhashi, S., Yamanaka, H., &Minamoto, T. (2015). Use of droplet digital PCR for estima‐tion of fish abundance and biomass in environmental DNA sur‐veys. PloS One, 10(3), e0122763. https://doi.org/10.1371/journal.pone.0122763

Drymon,J.M.,Ajemian,M.J.,&Powers,S.P. (2014).Distributionanddynamichabitatuseof youngbull sharksCarcharhinus leucas in ahighly stratified northern Gulf of Mexico estuary. PloS One, 9(5),e97124.https://doi.org/10.1371/journal.pone.0097124

Dulvy,N.K.,Fowler,S. L.,Musick, J.A.,Cavanagh,R.D.,Kyne,P.M.,Harrison, L. R.,…White,W. T. (2014). Extinction risk and conser‐vationof theworld’ssharksandrays.eLife,3,e00590.https://doi.org/10.7554/eLife.00590

Every,S.L.,Pethybridge,H.R.,Fulton,C.J.,Kyne,P.M.,&Crook,D.A.(2017).Nichemetricssuggesteuryhalineandcoastalelasmobranchsprovide trophicconnectionsamongmarineand freshwaterbiomesinnorthernAustralia.Marine Ecology Progress Series,565,181–196.https://doi.org/10.3354/meps11995

Ferretti, F., Worm, B., Britten, G. L., Heithaus, M. R., & Lotze, H.K. (2010). Patterns and ecosystem consequences of shark de‐clines in the ocean. Ecology Letters, 13(8), 1055–1071. https://doi.org/10.1111/j.1461‐0248.2010.01489.x

Ficetola,G.F.,Miaud,C.,Pompanon,F.,&Taberlet,P. (2008).Speciesdetection using environmental DNA from water samples. Biology Letters,4(4),423–425.https://doi.org/10.1098/rsbl.2008.0118

Ficetola, G. F., Pansu, J., Bonin, A., Coissac, E., Giguet‐Covex, C., DeBarba,M.,…Taberlet,P. (2015).Replication levels, falsepresencesand the estimation of presence/absence from eDNAmetabarcod‐ing data.Molecular Ecology Resources, 15(3), 543–556. https://doi.org/10.1111/1755‐0998.12338

Ficetola,G.F.,Taberlet,P.,&Coissac,E.(2016).HowtolimitfalsepositivesinenvironmentalDNAmetabarcoding?Molecular Ecology Resources,16(3),604–607.https://doi.org/10.1111/1755‐0998.12508

Foote,A.D., Thomsen, P. F., Sveegaard, S.,Wahlberg,M.,Kielgast, J.,Kyhn, L. A., … Gilbert,M. T. P. (2012). Investigating the potentialuseofenvironmentalDNA(eDNA)forgeneticmonitoringofmarinemammals. PloS One, 7(8), e41781. https://doi.org/10.1371/journal.pone.0041781

Garcia,V.B.,Lucifora,L.O.,&Myers,R.A. (2008).The importanceofhabitatandlifehistorytoextinctionriskinsharks,skates,raysandchimaeras. Proceedings of the Royal Society B: Biological Sciences,275(1630),83–89.https://doi.org/10.1098/rspb.2007.1295

Gargan,L.M.,Morato,T.,Pham,C.K.,Finarelli,J.A.,Carlsson,J.E.,&Carlsson,J.(2017).DevelopmentofasensitivedetectionmethodtosurveypelagicbiodiversityusingeDNAandquantitativePCR:Acasestudyofdevilrayatseamounts.Marine Biology,164(5),112.https://doi.org/10.1007/s00227‐017‐3141‐x

Goldberg,C.S.,Pilliod,D.S.,Arkle,R.S.,&Waits,L.P.(2011).Moleculardetection of vertebrates in stream water: A demonstration usingRockyMountaintailedfrogsandIdahogiantsalamanders.PloS One,6(7),e22746.https://doi.org/10.1371/journal.pone.0022746

Goldberg,C.S.,Turner,C.R.,Deiner,K.,Klymus,K.E.,Thomsen,P.F.,Murphy,M.A.,…Taberlet,P.(2016).CriticalconsiderationsfortheapplicationofenvironmentalDNAmethods todetectaquatic spe‐cies.Methods in Ecology and Evolution,7(11),1299–1307.https://doi.org/10.1111/2041‐210X.12595

Heithaus,M.R.,Frid,A.,Wirsing,A. J.,&Worm,B. (2008).Predictingecological consequences of marine top predator declines. Trends in Ecology and Evolution,23(4), 202–210. https://doi.org/10.1016/j.tree.2008.01.003

Hinlo, R., Gleeson, D., Lintermans, M., & Furlan, E. (2017). Methodsto maximise recovery of environmental DNA from water sam‐ples. PloS One, 12(6), e0179251. https://doi.org/10.1371/journal.pone.0179251

Hoenig, J. M., & Gruber, S. H. (1990). Life‐history patterns inElasmobranch: Implications for fisheries Management. NOAA Technical Report NMFS,90,1–15.

Huggett, J. F.,Cowen, S.,&Foy,C.A. (2015).Considerations fordigi‐talPCRasanaccuratemoleculardiagnostictool.Clinical Chemistry,61(1),79–88.https://doi.org/10.1373/clinchem.2014.221366

Hunter,M.E.,Dorazio,R.M.,Butterfield,J.S.,Meigs‐Friend,G.,Nico,L.G.,&Ferrante, J.A. (2017).Detection limitsofquantitativeanddigitalPCRassaysandtheirinfluenceinpresence–absencesurveysofenvironmentalDNA.Molecular Ecology Resources,17(2):221–229.https://doi.org/10.1111/1755‐0998.1261

Hunter,M.E.,Meigs‐Friend,G.,Ferrante,J.A.,Kamla,A.T.,Dorazio,R.M.,Diagne,L.K.,…Reid,J.P.(2018).SurveysofenvironmentalDNA(eDNA):AnewapproachtoestimateoccurrenceinVulnerableman‐ateepopulations.Endangered Species Research,35,101–111.https://doi.org/10.3354/esr00880

Jerde,C. L.,Mahon,A. R., Chadderton,W. L., & Lodge,D.M. (2011).“Sight‐unseen” detection of rare aquatic species using environ‐mental DNA. Conservation Letters, 4(2), 150–157. https://doi.org/10.1111/j.1755‐263X.2010.00158.x

Lafferty, K. D., Benesh, K. C.,Mahon, A. R., Jerde, C. L., & Lowe, C.G. (2018). Detecting southern California’s white sharks with en‐vironmental DNA. Frontiers in Marine Science, 5(355), https://doi.org/10.3389/fmars.2018.00355

Lahoz‐Monfort,J.J.,Guillera‐Arroita,G.,&Tingley,R.(2016).Statisticalapproaches to account for false‐positive errors in environmentalDNAsamples.Molecular Ecology Resources,16(3),673–685.https://doi.org/10.1111/1755‐0998.12486

Larson,E.R.,Renshaw,M.A.,Gantz,C.A.,Umek,J.,Chandra,S.,Lodge,D.M.,&Egan,S.P.(2017).EnvironmentalDNA(eDNA)detectstheinvasive crayfishesOrconectes rusticus andPacifastacus leniuscu‐lusinlargelakesofNorthAmerica.Hydrobiologia,800(1),173–185. https://doi.org/10.1007/S10750‐017‐3210‐7

LePort,A.,Bakker,J.,Cooper,M.,Huerlimann,R.,&Mariani,S.(2018).EnvironmentalDNA(eDNA):Avaluabletoolforecologicalinterfer‐enceandmanagementofsharksandtheirrelatives.InL.Carrier,M.Heithaus,&C.Simpfendorfer(Eds.),Shark research: Emerging technol‐ogies and applications for the field and laboratory(pp.255–283).BocaRaton,FL:CRCPress.

10  |     SCHWEISS Et al.

Lewison, R. L., Crowder, L. B., Read, A. J., & Freeman, S. A. (2004).Understanding impacts of fisheries bycatch on marine mega‐fauna.Trends in Ecology and Evolution,19(11),598–604.https://doi.org/10.1016/j.tree.2004.09.004

Nathan,L.M.,Simmons,M.,Wegleitner,B.J.,Jerde,C.L.,&Mahon,A.R.(2014).QuantifyingenvironmentalDNAsignalsforaquaticinvasivespecies acrossmultiple detection platforms. Environmental Science and Technology, 48(21), 12800–12806. https://doi.org/10.1021/es5034052

Pilliod,D.S.,Goldberg,C.S.,Arkle,R.S.,&Waits,L.P.(2013).Estimatingoccupancyandabundanceofstreamamphibiansusingenvironmen‐talDNA from filteredwater samples.Canadian Journal of Fisheries and Aquatic Sciences, 70(8), 1123–1130. https://doi.org/10.1139/cjfas‐2013‐0047

Polovina, J. J., Abecassis,M., Howell, E. A., &Woodworth, P. (2009).Increasesintherelativeabundanceofmid‐trophiclevelfishescon‐current with declines in apex predators in the subtropical NorthPacific,1996–2006.Fishery Bulletin,107(4),523–531.

Port,J.A.,O’Donnell,J.L.,Romero‐Maraccini,O.C.,Leary,P.R.,Litvin,S.Y.,Nickols,K.J.,…Kelly,R.P.(2016).Assessingvertebratebiodiver‐sityinakelpforestecosystemusingenvironmentalDNA.Molecular Ecology,25(2),527–541.https://doi.org/10.1111/mec.13481

Rees, H. C.,Maddison, B. C.,Middleditch, D. J., Patmore, J. R.M., &Gough,K.C. (2014).Review:Thedetectionofaquaticanimal spe‐ciesusingenvironmentalDNA–areviewofeDNAasasurveytoolinecology.Journal of Applied Ecology,51(5),1450–1459.https://doi.org/10.1111/1365‐2664.12306

Renshaw,M.A.,Olds,B.P.,Jerde,C.L.,McVeigh,M.M.,&Lodge,D.M.(2015).Theroomtemperaturepreservationoffilteredenvironmen‐talDNAsamplesandassimilationintoaphenol–chloroform–isoamylalcoholDNAextraction.Molecular Ecology Resources,15(1),168–176.https://doi.org/10.1111/1755‐0998.12281

Ritchie, E. G., Elmhagen, B., Glen, A. S., Letnic, M., Ludwig, G., &McDonald, R. A. (2012). Ecosystem restoration with teeth: Whatrole forpredators?Trends in Ecology and Evolution,27(5), 265–271. https://doi.org/10.1016/j.tree.2012.01.001

Sigsgaard,E.E.,Nielson,I.B.,Bach,S.S.,Lorenzen,E.D.,Robinson,D.P.,Knudsen,S.W.,…Thomsen,P.F.(2016).Populationcharacteris‐ticsofa largewhalesharkaggregation inferred fromseawateren‐vironmentalDNA.Nature Ecology and Evolution,1(1),4.https://doi.org/10.1038/s41559‐016‐0004

Simpfendorfer,C.A.,Kyne,P.M.,Noble,T.H.,Goldsbury,J.,Basiita,R.K., Lindsay, R., … Jerry, D. R. (2016). Environmental DNA detectsCritically Endangered largetooth sawfish in the wild. Endangered Species Research,30,109–116.https://doi.org/10.3354/esr00731

Stobutzki, I. C., Millter, M. J., Heales, D. S., & Brewer, D. T. (2002).Sustainability of elasmobranchs caught as bycatch in a tropicalprawn(shrimp)trawlfishery.Fishery Bulletin,100(4),800–821.

Taberlet, P., Coissac, E., Hajibabaei, M., & Rieseberg, L. H. (2012).Environmental DNA.Molecular Ecology, 21(8), 1789–1793. https://doi.org/10.1111/j.1365‐294X.2012.05542.x

Takahara,T.,Minamoto,T.,&Doi,H.(2015).Effectsofsampleprocess‐ingonthedetectionrateofenvironmentalDNAfromtheCommonCarp(Cyprinuscarpio).Biological Conservation,183,64–69.https://doi.org/10.1016/j.biocon.2014.11.014

Thorson,T.B.(1962).PartitioningofbodyfluidsintheLakeNicaraguasharkandthreemarinesharks.Science,138(3541),688–690.https://doi.org/10.1126/science.138.3541.688

Thorson,T.B.(1971).Movementofbullsharks,Carcharhinusleucas,be‐tweenCaribbeanSeaandLakeNicaraguademonstratedbytagging.Copeia,1971(2),336–338.https://doi.org/10.2307/1442846

Thorson,T.B.,Cowan,C.M.,&Watson,D.E.(1973).BodyfluidsolutesofjuvenilesandadultsoftheeuryhalinebullsharkCarcharhinusleu‐casfromfreshwaterandsalineenvironments.Physiological Zoology,46(1),29–42.https://doi.org/10.1086/physzool.46.1.30152514

Tréguier,A.,Paillisson,J.M.,Dejean,T.,Valentini,A.,Schlaepfer,M.A.,&Roussel,J.M.(2014).EnvironmentalDNAsurveillanceforinverte‐bratespecies:AdvantagesandtechnicallimitationstodetectinvasivecrayfishProcambarusclarkiiinfreshwaterponds.Journal of Applied Ecology,51(4),871–879.https://doi.org/10.1111/1365‐2664.12262

Weltz, K., Lyle, J. M., Ovenden, J., Morgan, J. A., Moreno, D. A., &Semmens, J.M. (2017).ApplicationsofenvironmentalDNAtode‐tectanendangeredmarineskatespeciesinthewild.PloS One,12(6),e0178724.https://doi.org/10.1371/journal.pone.0178124

Whale,A.S.,Huggett,J.F.,Cowen,S.,Speirs,V.,Shaw,J.,Ellison,S.,…Scott,D.J.(2012).ComparisonofmicrofluidicdigitalPCRandconven‐tional quantitativePCR formeasure copynumbervariation.Nucleic Acids Research,40(11),e82.https://doi.org/10.1093/nar/gks203

SUPPORTING INFORMATION

Additional supporting information may be found online in theSupportingInformationsection.

How to cite this article:SchweissKE,LehmanRN,DrymonJM,PhillipsNM.DevelopmentofhighlysensitiveenvironmentalDNAmethodsforthedetectionofBullSharks,Carcharhinus leucas(MüllerandHenle,1839),usingDropletDigital™PCR.Environmental DNA. 2019;00:1–10. https://doi.org/10.1002/edn3.39