This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Strong population structure deduced from genetics, otolith chemistry and parasite abundances explains vulnerability to localized fishery collapse in a large Sciaenid fish, Protonibea diacanthus
Laura Taillebois1,2 | Diane P. Barton1,3 | David A. Crook1 | Thor Saunders3 | Jonathan Taylor3 | Mark Hearnden3 | Richard J. Saunders4,5 | Stephen J. Newman6 | Michael J. Travers6 | David J. Welch7 | Alan Greig8 | Christine Dudgeon9 | Safia Maher9 | Jennifer R. Ovenden9
CorrespondenceLaura Taillebois, Research Institute for the Environment and Livelihoods, Charles Darwin University,Darwin,NT,Australia.Email: [email protected]
AbstractAspressureoncoastalmarineresourcesisincreasingglobally,theneedtoquantita-tivelyassessvulnerablefishstocks iscrucial inordertoavoidtheecologicalconse-quencesof stockdepletions.SpeciesofSciaenidae (croakers,drums) are importantcomponentsoftropicalandtemperatefisheriesandareespeciallyvulnerabletoex-ploitation.Theblack-spottedcroaker,Protonibea diacanthus,istheonlylargesciaenidincoastalwatersofnorthernAustraliawhereitistargetedbycommercial,recreationalandindigenousfishersdueto itsfoodvalueandpredictableaggregatingbehaviour.Localizeddeclinesintheabundanceofthisspecieshavebeenobserved,highlightingtheurgentrequirementbymanagersforinformationonfine-andbroad-scalepopula-tionconnectivity.ThisstudyexaminedthepopulationstructureofP. diacanthus across north-westernAustraliausingthreecomplementarymethods:geneticvariationinmi-crosatellitemarkers,otolithelementalcompositionandparasiteassemblagecomposi-tion. The genetic analyses demonstrated that there were at least five geneticallydistinctpopulationsacrossthestudyregion,withgeneflowmostlikelyrestrictedbyinshorebiogeographicbarrierssuchastheDampierPeninsula.Theotolithchemistryand parasite analyses also revealed strong spatial variation among locationswithin
Nearshore coastal ecosystems provide an important source of foodfor humanpopulations, supporting 90%of the globalwild fish har-vest whilst accounting for only 7% of the oceanworldwide (Paulyetal.,2002).Asglobaldemandforseafoodincreases,manyexploitedspecies in these regionshavedeclinedand largenumbersof fisher-iesarecurrently fishedatunsustainable levels (Jacksonetal.,2001;Smithetal.,2010).Sustainablemanagementofexploitedfishpopu-lations requires detailed knowledge of population structure, naturalabundanceandthedegreeofecologically relevantexchangeamongharvestedstocks(Paulyetal.,2002).Theconceptof“stocks”aspop-ulationunitshaslongbeenanchoredinfisheriesscienceandcanbedefinedasgroupsoffishwithinaspeciesthatareself-recruiting;sharesimilar growth rates and rates of natural and non-naturalmortality;andmayreactmoreorlessindependentlytoharvesting(Cadrin,Kerr,&Mariani,2013).
Methods fordelineating stockshaveadvancedconsiderably inrecent years and include genetic techniques, acoustic telemetry,tagging,otolithchemistry,demographicanalysis,otolithshapeandmeristicdata (Hawkinsetal., 2016).Genetic approaches speak toboth inter-and intragenerational timescalesas theytrack the life-history stages of the fish (i.e., from fertilized eggs to adults) thatare interchanged between locations andwhich subsequently par-ticipate insuccessfulspawning.Otolithmulti-elementalsignaturesandparasiteassemblagesareecologicalmarkersrecordingtheen-vironmentthejuvenileandadultfishinhabitandreflectprocessesoccurringwithintheindividuals’lifetime.Theintegrationofmultipletechniquesthatoperateoverdifferent temporalandspatialscalesmakes it possible to overcome many of the limitations of singletechniqueapproachesandgreatlystrengthensthe inferenceavail-able from stock structure studies (Abaunza etal., 2008; Begg &Waldman,1999;Lleonart&Maynou,2003;Waldman,1999;Welchetal.,2009,2015).Asanexample, Izzoetal. (2017)usedan inte-gratedapproachtorevealfourstocksofthesardineSardinops sagax in Australian waters when it was previously considered a semi-continuousmeta-population.
Fishes of the family Sciaenidae, commonly known as croakers,are important components of commercial, recreational and indige-nousfisheriesintropicalandtemperateregionsworldwide(Lenanton&Potter, 1987).They are targeted for their flesh and, increasingly,for their swim bladders, which are sold fresh or dried in South- East Asia(Ghoshetal.,2009;Sadovy&Cheung,2003;Tuuli,2010).ManyspeciesofSciaenidaehavedeclined in recentdecadesduetoover-exploitation,andseveralarenowconsideredthreatened.Theblack-spotted croaker Protonibea diacanthus (Lacapede, 1802) is a largespecies (>1.5mmax. total length; up to 42kgmass) that iswidelydistributed throughout coastalwaters and estuaries of the tropicalIndo-WestPacific (Sasaki,2001). InAustralia, it isdistributedalongthe northern coast from Shark Bay, Western Australia to HerveyBay,Queensland (Bray,2011).The speciesgrows rapidly andhas amaximum recorded age of 13years (Phelan, 2008). Spawning ofP. diacanthusinAustraliaoccursfromAugusttoDecember,andtheyappeartoproducepelagiceggsandhaveapelagiclarvalphase(Froese&Pauly,2016;Leis&Carson-Ewart,2000).Whilst inshoreseasonalbreeding aggregations ofP. diacanthus are thought to occur (Welchetal.,2014),directevidenceofthebehaviourofthespeciesislimited(Phelan,2008)andlittleisknownaboutthesizeandlocationsoftheaggregations.Commercialandrecreationalsectorsbothcontributetototal landings (667 tonnes in2005),butcatcheshave increased re-centlydueprimarilytogrowthinthecommercialsector(from43to250tonsperannumbetween1995and2005;Coleman,2004;Phelan&Elphick,2006). Intensiverecreational fishingontheeastcoastofQueensland(Bowtell,1995,1998)hasalsobeenattributedwithre-ducing fishabundanceand size to thepointwhere thecatchesarealmostexclusivelylimitedtoimmaturefish.Thereisalsoevidenceofover-exploitationofP. diacanthusoutsideofAustralianwaters:over-fishinginIndiainthe1980sledtolocalextirpation(James,1994)andafisherythatoncethrivedinHongKongnolongerexists(Sadovy&Cheung,2003).
ofthespeciesacrossnorth-westernAustraliawiththeexpectationofspatially distinct stocks,whichwould be consistentwith the obser-vationoflocalizeddepletion.Moreover,weaimedtotesttheexpec-tation that genetics provides informationon a broader spatial scalethanotolithchemistryandparasitedata in thiscoastallydistributedspecies.Todosoandtodemonstratethepowerofintegratingacrosstechniques,wecontrastedpopulationstructureresultsfromthethreemethodsandfocusedononeexamplewheretheassignmentofindi-vidualstoadjacentpopulationswasmethod-specific.Finally,weuseourresultstoaddresstheappropriatespatialscaleforstockassess-mentandsuggest fisheriesmanagementarrangementsforP. diacan-thusacrossnorth-westernAustralia.
2 | MATERIAL AND METHODS
2.1 | Samples
Atotalof298fishwerecollectedfrom11samplinglocationsacrossnorth-western Australia from Roebuck Bay (Western Australia) totheVanderlinIslandsintheGulfofCarpentaria(NorthernTerritory)(Figure1,Table1).Thedeclineof thespecies (Phelan,2002)meantthat samples could not be obtained from north-eastern Australia.The lack of knowledge on the location and size of the aggregationgroundsinP. diacanthusdidnotallowtheexclusivecollectionofma-ture fishonspawninggrounds,andthus, fishofdifferentsizewereincludedintheanalysis.SampleswerecollectedwiththeassistanceofIndigenousmarinerangers,fishingtouroperatorsandrecreationaland
commercialfishers.Thetotallength(TL),standardlength(SL)andsexofeachspecimenwererecorded.Forgeneticanalyses,muscletissueorfinclipsweresampledandimmediatelyplacedintovialscontain-ingmoleculargrade95%ethanolor20%dimethylsulfoxide(DMSO)solutionin5mol/LNaClandstoredat4°Cinthefieldand−20°Cinthe laboratory. The sagittal otolithswere dissected from each fish,cleanedandrinsedthoroughly,dried,andstoredinpaperenvelopespriortopreparationfortraceelementanalysis.Gillsandinternalbodyorganswereremoved,placedintoindividualbagsandfrozenuntilex-amination for parasites. The same individualswere used for otolithmicrochemistry,geneticandparasiteanalyses.
2.2 | Data analysis
Populationstructurewastestedinanhierarchicalfashion,withgenet-icsbeingdeployedat thebroadscale toguidepopulationstructureandecologicalmarkersusedatthefinerscale.Thenatureofotolithmicrochemistry reflects fine-scale spatial variation between envi-ronmentsandisnotusefulforbroad-scalevariation.Indeed, lackofsignificance (i.e.,otolithsimilarity)onabroadspatial scaledoesnotindicateecologicalpopulationhomogeneity.Fine-scalegeneticanal-yseswereused to integrateacross the threemethods for apairoflocationsinNWAustraliaforwhichonepresentedastrikinggeneticheterogeneity.
F IGURE 1 Locationofthe11samplingsitesacrossnorthernAustraliashowingthetwojurisdictions(WesternAustraliaandNorthernTerritory).ThedetailsforeachofthesamplinglocationsareprovidedinTable1
The potential for null alleles, large allele dropout and stutteringto interferewith scoring accuracywas checked for eachmicrosatel-lite locus in each sample usingMICROCHECKERversion 2.2.3 (VanOosterhout, Hutchinson,Wills, & Shipley, 2004). Summary statisticsformicrosatellite loci, includingthenumberofalleles,allelicrichness,expected and observed heterozygosity and fixation indexes, wereobtained for each sampling location using GENALEX version 6.502(Peakall&Smouse,2006).Testsofconformanceofgenotypicpropor-tions to Hardy-Weinberg equilibrium expectations were carried outfor each locus at each sample location, and testsof genotypic equi-libriumbetweenpairsofmicrosatellites (linkagedisequilibrium)werecarriedoutforeachsamplelocation,usinganexactprobabilitytestasimplementedinGENEPOPversion4.5(Rousset,2008).TheexacttestwasestimatedusingaMarkovChainthatemployed1,000dememo-rizations,500batchesand1,000 iterationsperbatch.WeaccountedforthelinkagedisequilibriummultipletestingbyapplyingaBonferronicorrection to the p-value.Atestwithap-value<.0045(0.05/numberoflociconsideredineachmultipletest)wasrequiredtobeconsideredsignificant.
Inordertoassesswhetherlocationscouldbetreatedasindepen-dentgeneticallycohesivegroupsonabroadspatialscale,andtotestfortheextentofadmixturewithinthegroups,twocontrastingclus-tering approacheswere used. Population assignment and clusteringwas performedusing theBayesianmodel-based clustering programSTRUCTUREversion2.3.4(Pritchard,Wen,&Falush,2003).Forcom-parativepurposesonthebroadspatial scale,adiscriminantanalysisofprincipalcomponents(DAPC,Jombart,Devillard,&Balloux,2010)available in the adegenetpackage(Jombart,2008)forRversion3.3.1
(RCoreTeam,2017)wasalsoused.Anhierarchicalapproachwassub-sequently taken toexplorepotential substructurewithin thegroupsrevealed by the STRUCTURE broad- scale analysis. The hierarchical analyseswere also performed using STRUCTURE.Methodology forSTRUCTUREandDAPCisdetailedinAppendixA1inAppendixS1.
To test alternate post hocobservationsofgeneticstructurerevealedby our different approaches, we assessed the per cent of geneticvarianceexplainedbythegroupingsdeducedfromSTRUCTUREandDAPC results using an analysis of molecular variance (AMOVA) asimplemented in ARLEQUIN. The significance of any differentiationwasdeterminedbypermutationof22,000replicates.
PairwiseFSTvalueswere calculated (i) to further investigate thepopulationstructurewithinthegroupsrevealedwiththeclusteringap-proachesand(ii)totestwhetherP. diacanthusisconstrainedbyapat-ternofisolationbydistance(IBD).Fixationindices(FST)betweenpairsofsamplelocalitieswereestimatedasimplementedinARLEQUINver-sion3.5.2.2(Excoffier&Lischer,2010).WeperformedaManteltestof FST/(1−FST) (G)versusgeographic (D)distances among locationsusingARLEQUIN.
Toinvestigatethefine-scalepopulationstructureandprovidein-formationonthedispersalof fish, themicrochemistryoftheotolithwasexaminedintwosectionsoftheotoliths:thenearcoreandthemargin. The near core area is located just outside the first opaquezone ~500 μmfromthecoreoftheotolithandrepresentsthe juve-nilephase;andtheotolithmargin(edge)wassampledadjacenttothesulcus acusticustorepresentgrowthintheperiodpriortocapture.Atotalof11traceelements(7Li, 25Mg,23Al,49Ti, 53Cr, 55Mn,60Ni, 63Cu, 66Zn, 88Sr, 138Ba)andtheinternalstandard(43Ca)wereanalysedfromthetwoablationzonesineachotolith.Detailsonthegeneralmethod-ologyfortheotolithpreparation,thelaserablationanddatatransfor-mationarepresentedinAppendixA2inAppendixS1.Forallelements,theratioofelementisotopeintensitytoCaintensitywasusedtoes-timate the element:43Ca ratio. These ratios were converted to molar
Parasiteswerealsousedasbiologicaltagstoinvestigatethefine-scalepopulationstructureofP. diacanthus.Directlifecycleparasitesgenerallyhaveshort-livedlarvalstagesthatareunabletocoverlargedistancesandshouldreflectlocalityoforigin.Indirectlifecyclepar-asites,ontheotherhand,cantravelthroughmultiplehostsandthushaveadistributionpatternthatisultimatelyatthemercyofanum-berofdifferentfactors.Actual lifecyclesofthemajorityofaquaticparasites remainunknown,andwecanonlygeneralizeas tohostswithinthecyclebasedoncurrentknowledge.Initially,analyseswereperformedforexternalparasites(=direct)andlarvalstages(=indirectorpermanentparasites)separately.Bothanalysesshowedsignificantseparation among sites. We subsequently combined the data fordirectand indirectparasites for the finalanalyses,as thisprovidedmore robust and biologically reliable and interpretable results (seebelow).
The otolith and parasite datawere broken down into three geo-graphic regions for statistical analyses (i) because these ecologicalmarkersarethoughttooperateoversmallerspatialandtemporalscalescomparedtothegenetics;hence,theyrequireasmallerscaleofanal-ysisand(ii)totestofthecurrentjurisdictionalmanagementboundarybetweenWesternAustraliaandtheNorthernTerritoryandtheregionalmanagementunitsacrosstheNorthernTerritory.Eachregionincludedthegeographicallyclosestsamplelocationfromtheadjacentregiontoallowinter-regioncomparisons.TheWesternregionincludesalloftheWesternAustralianlocations(RB,CSandWy)aswellasWa,whichistheclosestNorthernTerritorysamplelocation(Figure1).TheDarwinregionincludesalltheNorthernTerritorylocationsfromWatoMI(Wa,PI,OD,BIandMI).TheArnhem/GulfregionincludesalltheArnhemLandandGulfofCarpentariapopulationsoftheNorthernTerritory(Ma,ASandVI)aswellasMI,whichistheclosestNorthernTerritorysamplelocation.
All data analysis was performed using R (R Core Team, 2017).Pairwisedistributionsofallvariablesusedineachanalysisforthepar-asiteandotolithdatasetswereinspectedforextremeoutliers,unsatis-factorycovariatedistributions,andthepresenceofcollinearity.Wherenecessaryabundancedatawerelog(Y+1)transformed.Homogeneity
Fishareknowntoaccumulatelonger-livedparasitespecieswithage(Rohde,1982),whichwillaffectdifferencesinparasitefaunaifsamples contain individuals of different ages. All parasite speciesthatwereselectedonprevalencewereexaminedfor theircorrela-tionwith host length (as an indicator of age) using linearmodels.Wherelengthwassignificantlycorrelatedwithabundance,numbersgreaterthanzerowereadjustedtothemeanhostsize(totallength)usingthemethodsdescribedinMoore,Buckworth,Moss,andLester(2003).
Spatial variation in parasite assemblages and the otolith nearcoreandmarginchemistryamongregionsand locationswithin re-gionswere investigated separately using single-factormultivariateanalysisofvariance (MANOVA). Lineardiscriminant function anal-ysis (LDFA)was conducted toprovide a statistical andvisual indi-cationof thesimilaritiesofeither theparasitecommunitiesor themulti-elementalotolithmicrochemicalsignaturesamongsamplesattheregionalspatialscaleusingtheMASSpackageinR(Venables&Ripley, 2002). LDFA classification success rates and an associatedproportional chance criterion (the expected proportion of correctclassification by chance alone) were calculated (Poulin & Kamiya,2015)foreachlocationwithintheregions.Toshowtheseparationachieved for the groups in the analysis, the first two discriminantfunctionsforeachindividualwereplottedandthe95%confidenceellipses around the centroid means of the first two discriminantfunctionsforeachgroupinthesampleusingtheellipsepackageinR(Murdoch&Chow,2013).
Totesttheeffectofdistancebetween locationsontheparasiteassemblages, and the otolith elemental concentrations in the nearcoreandthemargin,weperformedaManteltestoftheJaccarddis-similarityusingthemeandifferencesinnumberandconcentrations,respectively, between each pair of locationsversus geographic dis-tancesamonglocationsusingtheveganpackageinR(Oksanenetal.,2017).
The individualassignmentoffishwastestedthroughaBayesianapproach across all fish for otoliths andparasites.Classes andpos-terior probabilities were calculated by jackknife cross-validation toassign individuals tomost probable locationswithin each region. Ina couple of locationswhere the genetic structure gave evidenceofgeneticheterogeneitywithinlocations,STRUCTUREwasusedtocal-culateposteriorprobabilitiesandassignindividualstomostprobablelocations. Individual mis-assignments were compared across tech-niquestotestwhethereachtechniqueprovideddifferentinformationornot(i.e.,seewhetherthesamefishweremis-assignedtothesamelocationsbetweentechniques).
3 | RESULTS
Genotypesfrom11microsatellitelociwereobtainedfor284individu-als of P. diacanthus.Thelevelofmissingdatawaslow(2.15%)andas
6 | TAILLEBOIS ET AL.
suchwas retainedand identifiedasmissingdata in the subsequentanalyses (missingdatawerecalculatedas thenumberofgenotypesthatcouldnotbescoredoverthetotalnumberofgenotypespresentintheidealdataset,11×284).Thenumberofallelesperlocusrangedfrom5 (Prd046) to21 (Prd012) (AppendixB3 inAppendixS2). Thedatawere freeof scoringerrorsasevaluatedbyMICROCHECKER.Genotypeswererescoredandcorrecteduntiltherewasnoevidenceof null alleles at these or other locations and loci. Deviations from Hardy-Weinbergproportions(HWP)weredetected14times(p- value <.05 for Prd044atRB,Prd023atRB,WaandBI,Prd042 at OD, Prd012 atWaandMa,Prd018atCSandMa,Prd020atPI,ODandAS,Prd036 atRBandPrd024atPI)outofthe121testsperformedforeachlocusagainst each location. Therewasno significant test for linkagedis-equilibriumbetweenpairsoflociacrossalllocationsafterBonferronicorrection(p-value<.0045).Theanalysisproceededwithoutremov-inglociorindividualsfromthedataset.Heterozygositywasmoderatetohigh forall loci acrossall locations (0.659±0.179,AppendixB3inAppendixS2)andgenerallysimilartoexpectationsformarinefish(approximately0.7,DeWoody&Avise,2000).
Four distinct populationswere identified usingBayesianmodel-basedclusteringmethods.Theoptimumnumberofgenetic clustersreturnedby∆K in STRUCTURE was 5 without LOCPRIORand3withLOCPRIOR. The LOCPRIORparameterenhancedthepopulationstruc-turepattern(AppendixB4inAppendixS2);hence,thefollowingre-sults describe the observed pattern using location information as aprior.ForK=3,fourgeographicallydistinctpopulationsorgroupsofpopulationswererevealed.ThewesternpopulationsRBandCSweredistinctfromeachother,andagroupofcentralnorthernpopulationsWy-Wa-PI-OD-BI-MI-Maappeared tobe similar.TherewasalsoaneasterngroupwithlocationsASandVIappearingsimilartoeachother.DespitethesimilaritybetweenAS-VIandCSgenotypeswhenK=3,theywereassumed tobedifferentpopulationsorgroupofpopula-tions because of their geographic distance (Figure2a). Independentinvestigations of hierarchical structure within the northern groupWy-Wa-PI-OD-BI-MI-Ma and within the Gulf locations AS-VI didnot reveal further substructurewithin those twogroups (Figure2a).Overall,thegroupsfavouredbySTRUCTUREwerethefollowingfour:RB,CS,Wy-Wa-PI-OD-BI-MI-Ma,AS-VI.
In successive K-means clustering using find.clusters in R, the Bayesian InformationCriterion (BIC) showedan initial sharpdeclinein values until K=5afterwhichBICvalues increased (Figure2b).Tomatch the number of genetic clusters revealed by find.clusters, the individualswerepartitioned intofivegroupsof locationsthatwouldtake into account the genetic structure previously highlightedwithSTRUCTURE.TheGulfofCarpentarialocationsASandVIwereheresplit basedon their geographic distance.The five groups used as apriorintheDAPCwereasfollows:RB,CS,Wy-Wa-PI-OD-BI-MI-Ma,AS andVI.Theoptimal numberofPCswith thehighestmean suc-cessfulassignmentandthelowestmeansquarederrorwas30.Whenpartitioningtheindividualsintofivegroups,DAPCrevealedoverlap-pingdistributionsofgeographicgroupsontheordinationplotthatin-dicatedalowdegreeofgeneticdifferentiationbetweenthesegroupsorlocations.TheDAPCseparatedRBfromotherlocationsalongthe
AMOVAwasusedtocomparethedifferentscenariossupportedbySTRUCTUREandDAPC.The4-and5-groupscenariohighlightedbySTRUCTUREandDAPC,respectively,explained2.15%(p-value=.000)and 2.28% (p-value=.000) of the total genetic variation. As bothscenariosexplainedverysimilaramonggroupgeneticvariation,theyarebothlikelytorepresentbiologicalpopulationstructure.However,the 4-group scenario revealed by STRUCTUREwill be used in thediscussion,asitistheresultofunsupervisedanalyses,whichwasnotthecaseforDAPC.
ThepairwiseFST values confirmed some of the results found with the clustering approaches.A pattern of genetic differentiationwithlowbutsignificantpopulation-pairwiseFSTvalues(range0.009–0.054,Table2)wasobserved,andtheoverallFSTvaluewas0.014.Thepat-ternrevealedthattwolocationsinthewesternpartofthesamplingarea (RBandCS)and two locations in theeasternpartof thesam-plingarea(ASandVI)weregeneticallydistinctfromalmostallotherlocationssurveyedinthestudy.Bycontrast,thesevenlocations(Wy,Wa,PI,OD,BIMI,Ma)inbetweentheeasternandwesternlocationsformed an undifferentiated groupwhere only three of 42 pairwiseFST valueswere significant and noneof themwere significant afterBonferronicorrection.
Toevaluatethespatialprocessesthatdrivepopulationstructure,anisolation-by-distancemodelwasexamined.TheMantelcorrelationcoefficient r between thegenetic andgeographicdistancematriceswas equal to0.655 (coefficient of determination r2 =.429) andwasstatistically significant (p-value=.001). There was a linear relation-shipbetweenFST/(1−FST)andgeographicdistances(Figure3).Thus,fishfromnearbylocationstendedtobegeneticallymoresimilarthanexpected by chance and genetic differences increased linearlywithdistance. There was no correlation between the mean elemental otolith composition and the geographic distance between locations(nearcorer=−.225,p-value=.85;marginr=.449,p-value=.05).Thecorrelationcoefficientbetweentheparasitesabundanceandthegeo-graphicdistancebetweenlocationswaslowandmarginallysignificant(r=.434,p-value=.04).
Ecologicalmarkers(i.e.,otolithmicrochemistryandparasiteabun-dances) provided further information about population structure inP. diacanthus and information on the individuals’ life history.Of the11 elements measured, concentrations of 7Li, 25Mg,55Mn,60Ni, 63Cu, 66Zn, 88Sr and 138BainthenearcoresandmarginsofP. diacanthus oto-lithsweremeasuredabovetheLODinatotalof286fish(AppendixB1 inAppendixS2).A totalof291P. diacanthuswereexamined forthe parasitology component of this study. Overall, all fishwere in-fectedwithatleastoneparasiteindividualandatotalof44differentparasites specieswere identified (AppendixB2 inAppendix S2).Ofthese,11wereexcludedfromtheanalysesbasedonprevalenceandafurtherfourgroups(e.g.,Anisakidnematodelarvae,AncyrocephalineandDiplectanidmonogeneans)wereexcludedbasedon issueswithidentificationand/orenumeration.Themeanparasitespeciesrichness
Themulti-elementalsignaturesof thenearcoreandmarginsandtheparasiteabundancesdifferedsignificantlyamongthethreebroad-scale regions (Western, Darwin, Arnhem/Gulf) and among locationswithin each of the three regions (MANOVA, all testsp- value <.001, Table3,Figure4).AverageclassificationsuccessintheLDFAamongalllocationswas54%fortheotolithmargin,31%fortheotolithnearcoreand67%fortheparasites(Table4)andwasmuchhigherthanexpectedbychance (9%for theotolithand10%for theparasites), suggestinga nonrandomdistributionof theotolithmicrochemistry andparasiteabundancevariabilityacrossthesampledrange.Classificationsuccess
forbothmargin(71%fortheWesternregion,57%fortheDarwinre-gionand74%fortheArnhem/Gulfregion)andnearcore(57%fortheWesternregion,44%fortheDarwinregionand38%fortheArnhem/Gulfregion)washigherwhenthedataforeachregionwereanalysedseparately,butthe%classificationsuccessrelativetothe%expectedby chance was higher when all sites were analysed simultaneously(Table4, Figure4). Similar to the otolith chemistry analysis, separa-tionof theanalyses into regions increased reclassificationsuccess inparasite abundances,with81% for theWestern region,56% for theDarwinregionand89%fortheArnhem/Gulfregion(Table4,Figure4).Forbothotolithsandparasites,theclassificationsuccesswasgreaterfortheArnhemGulfandWesternregionsthanfortheDarwinregion.
F IGURE 2 SummaryofBayesianmodel-basedanddiscriminantanalysisofprincipalcomponents(DAPC)approachesforpopulationclusteringofmicrosatellitedatafromProtonibea diacanthus.(a)PlotoftheBayesianmodel-basedpopulationclusteringusingSTRUCTUREforK=3geneticclustersacrosstheentiredataset,forK=2withinthenortherngroup(Wy-Wa-PI-OD-BI-MI-Ma)andthegulflocations(AS-VI).Eachverticallinerepresentsanindividualandtheposteriorprobabilityproportionsofitsgenotypeassignedtothedifferentgeneticclusters.Individualsareplottedfromwesttoeastalongthesamplinggradient,andpopulationsareabbreviatedfollowingTable1.RedarrowspointattheindividualsfromRBthatareassignedtoCS.Theadmixturemodelandpopulationinformationwereusedintheanalysis.(b)ChangesinmeanBayesianInformationCriterion(BIC)valuesinsuccessiveK-meansclustering.(c)OrdinationplotoftheDAPCforthefivepredefinedclusters(RB,CS,Wy-Wa-PI-OD-BI-MI-Ma=NorthernGroup,VIandAS).Individualgenotypesappearasdots,andlocationsaredepictedbycoloursand95%inertiaellipses.Thebottom-leftinsetshowsthenumberofprincipalcomponentretainedandthecumulativevariancetheyexplain;andthebottom-rightinsetshowstheeigenvaluesforthefourprincipalcomponentsinrelativemagnitudewithx-axisandy-axisconstitutingthefirsttwoprinciplecomponents,respectively
location immediatelyprior to collection (i.e., samples collected fromthesamelocationinthesameyear).Incontrast,thenearcore(juve-nilephase)samplesincorporatedgrowthfromdifferentyearsbecausefish ineach locationwereofvariableagesandsourcesmayvarybyageandcohort.Individualmis-assignmentswerecomparedbetweennear core andmargin, and therewas no evidenceofmatchingpat-ternofmis-assignment(AppendixB5inAppendixS2).Fishthatweremis-assignedusing thenear corewerenotnecessarilymis-assignedusing the margin; and if they were, the assigned location was notnecessarily thesame.ClassificationsuccesswasgenerallyhigherfortheparasitesthantheotolithexceptBIandWywherebothmarginsandnearcoreshadhigherclassification rates thanparasites;andPIandASwhere only themargins had higher classification rates thanthe parasites. Individual mis-assignments were compared betweenparasitesandmargin,andtherewasnoevidenceofmatchingpatternofmis-assignment(AppendixB5inAppendixS2).Fishthatweremis-assignedusingtheparasiteswerenotnecessarilymis-assignedusingthemargin,andiftheywere,theassignedlocationwasnotnecessarilythe same.
F IGURE 3 Isolation-by-distanceanalysisgeneratedfrom1,000Mantelrandomizations.GeneticdistanceFST/(1−FST)againstgeographicdistance(km)andcorrespondingvalues
locationswithindividualspresentingmatchinggenotypeproportionstoindividualsfromCSlocation(Figure2a).FurtherfullBayesianassign-menttestsbetweenRBandCSrevealedthatindividualshavingoutliergeneticsignaturedidnotnecessarilyhavedistinctotolithorparasitecharacteristics.Ofthethree individualsthatweremis-assignedwiththe genetics, fish3167 and3540were alsomis-assignedusing thenearcoreandtheparasites,respectively(Table5).Assumingparasitesand otolith margins provide information about postlarval develop-ment;andnearcoreandgeneticprovideinformationabouttheloca-tionofearlystagedevelopment,wetestedthehypothesesthatotolithnearcorewouldcorrespond to thegenetics; and thatparasitesand
otolithmarginwouldprovide similar assignments too.Off theeightfishthatweremis-assignedusingthenearcore,onlyonewassimilarlymis-assignedforthegenetics (fish3167).Ofthefourfishthatweremis-assignedfor themarginandthetwofor theparasites,onlyonewasmis-assignedforbothparasitesandmargin(fish3518).
4 | DISCUSSION
In this study, the combined application of genetic markers, otolithchemistry and parasites across the same specimens provided three
F IGURE 4 Plotsofthefirsttwodiscriminantfunctionscoresshowingspatialvariationintheparasiteassemblageandthemulti-elementalotolithnearcoreandmarginsignaturesofProtonibea diacanthuscollectedfrom11locationswithinthreeregions:Westernregion,DarwinregionandArnhem/Gulfregion.Ellipsesare95%confidenceintervalsaroundthegroupcentroidforeachlocationwithineachregion,anddatapointsrepresentindividualfish
10 | TAILLEBOIS ET AL.
independentlinesofevidenceandconsiderablepowertodetectthehighlycomplexpopulationstructureofP. diacanthus across northern Australia. Relatively few studies to date have used this integratedapproachtoinvestigatepopulationstructureandconnectivityorforstock identification (Abuanza etal., 2008; Welch etal., 2015; Izzoetal.,2017).Thesestudiesalsoconcludedthattheintegrateduseofmultipletechniquesconsiderablyimprovedconfidenceinstockiden-tificationusingindexofstockdifferences(ISD)thresholdsdefinition(Izzoetal.,2017;Welchetal.,2015)andwereparticularlyeffectivefor specieswith broadly dispersing larvae or adults that exhibit lit-tleornogeneticstructureacrosstheirrange(Waldman,1999).Asacaseinpoint,ourstudyrevealedtheexistenceofmajorgeneticbreaksalongthenorth-westernAustraliancoastlineandthusthepresenceofseveraldiscretebreedingpopulationsofP. diacanthus.Priortosettle-ment,geneticresultssuggestthatcurrentsandgeographicbarrierstodispersalmaylimitexchangeamongcollectionlocations.Theecologi-calmarkers(otolithsandparasites)stronglysuggestthatP. diacanthus isgenerallysedentaryfollowinglarvalsettlement.
4.1 | Site fidelity and dispersal
Whilsttheresultsofthisstudysuggesthighlevelsofsitefidelityafterlarval settlement for P. diacanthusbasedonthefine-scalepopulationstructure,itshouldbenotedthattherewasoverlapinthemultivariateanalysesofparasiteassemblagecompositionsandotolithelementalsignaturesbetweensome locations thatpotentially reflectdispersalbyaproportionof juvenileor adult fish. Similar to the IBDpatternobservedinthegeneticanalyses,thecorrelationbetweengeographicdistance andparasite assemblagediversity showedadecline in thesimilarity of parasite assemblageswith increasing distance. The ge-neticIBDpatternindicatesthatreproductionismorelikelytooccurbetween individuals in the same location. The positive correlationbetweendistanceandparasiteassemblagedissimilaritycouldbe in-terpreted in a similar manner, suggesting that parasites are sharedbythedispersaloffishwithinandamongneighbouringpopulations,whichmayberesponsiblefortherelativesimilarityofneighbouringlocations. Alternatively, itmay be that similarity in parasite assem-blagesamongneighbouringsitesisdrivenby(i)dispersalofparasitesindependentofP. diacanthusmovement(e.g.,viaotherhosts);(ii)thespatial distributions of the endemic areas of the different parasitespecies(MacKenzie&Abaunza,1998);or(iii)habitatpreferencesoftheparasite species,withneighbouring locationshavingmore simi-larhabitatconditionsthandistant locations (MacKenzie&Abaunza,1998).Underthesethreescenarios,dispersaloffishamonglocationswouldnotbenecessarytoexplaintherelationshipbetweenparasiteassemblagesimilarityandgeographicdistance.
In the case of otolith chemistry, overlap inmulti-elemental sig-natures among locations is likely to be driven by local geochemicalcharacteristics, including the underlying seabed and nearby rivercatchmentgeology,aswellasphysicochemicalfactorssuchaswatertemperatureandsalinityandtheireffectsonfishphysiologyandlocalecologicalprocesses(Campana&Thorrold,2001;Elsdon&Gillanders,2002,2004).Whilstsuchfactorscouldpotentiallyvarysystematicallyovergeographicgradients,wefoundnoevidenceofarelationshipbe-tween multi-elemental signatures and geographic distance. Indeed,severalverydistantsites(e.g.,CSandMI)showedhighlevelsofover-lapinotolithchemicalsignatures,butitisextremelyunlikelythattherewasanyexchangeof individualfishbetweenthesedistant locationsgiven theweightofevidenceacross theother techniques (parasitesandgenetics)oflocationfidelityandlimiteddispersal.Rather,themis-classificationofindividualstodistantlocationsappearstobeduetothe high number of locations relative to the limited number of ele-mentsthatcontributedtolocationseparationinthediscriminantfunc-tionanalysis.Similaroverlapandmisclassificationinotolithchemicalsignatures among locations have been reported previously both infreshwater (e.g.,Macdonald&Crook,2014)andmarine (e.g.,Miller,Banks,Gomez-Uchida,&Shanks,2005)studies.
Aspreviouslymentioned,thegenerallylowerclassificationsuccessinthenearcorecomparedtothemargincan,at least inpart,beat-tributedtothefactthatthenearcoresamples incorporatedgrowthfromdifferentyears (i.e.,differentagecohorts)withineach location(seealsoGillanders,2002). It isalsopossiblethatmovementbyfish
afterthejuvenilephaseaccountsforsomethereducedclassificationsuccess inthenearcoreanalyses.Yetdespitethepotential forvari-ability among fish of different ageswithin a location, our analysesconsistently revealed highly significant spatial differences in otolithchemistry signatures among locations. This finding agreeswith ourconclusionoflocation-levelpopulationstructurebasedonparasiteas-semblagesand,further,suggeststhatatleastsomespatialstructuringismaintainedfromtheearlystagesofthelifehistoryofP. diacanthus. Eventhoughageswereavailableforoursamples,thelownumberofsamplesperlocationcombinedwiththewiderangeofagesresultedinlownumberspercohort.Hence,cohortscouldnotbecomparedinthisstudy,andfurtherresearchisrequiredtodeterminetheextenttowhichmis-classificationoffishcanbeattributedtotemporalvariationinotolithchemicalsignaturesversusotherpotentialcauses(e.g.,immi-grationfromotherlocations).
4.2 | Population structure and connectivity
Atbroaderspatialscales,thegeneticanalysesrevealedtheexistenceofmajordiscontinuitiesalongthenorth-westernAustraliancoastline,resulting inseveralgeneticallydiscretepopulationsofP. diacanthus. This finding is based onmixed age samples, andwewould predictstrongerspatialstructuring ifmature fishwerecollectedonspawn-ing grounds.Major genetic breaks distinguished populations at RB,CS,ahomogenousgroupoflocationsintheNorthernTerritory(fromWytoMa), theoffshoreregionof theASandtheVI intheGulfofCarpentaria.ThedeepestgeneticstructureobservedwasbetweenRBinWesternAustraliaandtherestofthesamplingrange.ThedistancebetweenRBandthenearestlocationCS(~400km)maypartlyexplainthispronouncedgeneticbreak.However,similarorgreaterdistancesbetweenotherlocations(e.g.,425kmbetweenMaandMIor630kmbetweentheASandVI)didnotresultinsuchgeneticdiscontinuity.
Thegeomorphologyandlocalhydrodynamicregimeofthenorth-ern coastline of Western Australia may influence the population
geneticstructureofP. diacanthusacrossitsrange.Thetidalregime,thecompositionofthesubstrateandtheextentoftheriverinenetworkinfluence the size of themangrove communities in this area.Thus,the Kimberley and Canning bioregions separated by the DampierPeninsula (Figure1) (Semeniuk,1996;Thackway&Cresswell,1998)differsignificantlyinmangrovecharacteristics.ThetipoftheDampierPeninsulahasbeenidentifiedasanimportantbiogeographicbreakformarine species (Hutchins,2001;Travers,Potter,Clarke,Newman,&Hutchins,2010;Wilson,2013)andcoincidewithapronouncedshiftintheunderlyinggeology(e.g.,fromsedimentarysandstoneinthenorthtounconsolidatedsandandsiltinthesouth)andassociateddominantbenthichabitats (e.g., fromcoralreefs inthenorthtosoftsubstratecommunitiesinthesouth).Thisbarrieralsocoincideswithaprominentincrease in tidal currents and associated water turbidity to the north. The Kimberley-Canning border has the largest tropical tidal range(~12m)andsomeofthefastesttidalcurrentsintheworld(2.5m/s),andexperiencesmassivevolumesoffreshwaterinahighlyturbidplumefromtheFitzroyRiverduringthemonsoonalwetseason(Wolanski&Spagnol,2003).ThediscontinuitybetweentheKimberleyandCanninggeomorphology(macrotidal,jagged“ria”coastlinewithextensiveman-groveforestsvssmoothercoastlinewithnorivers)createscontrastedmarinecoastalhabitats,inhibitinglarvaldispersalandreinforcingspa-tialpopulationstructure.Likewise,quantitativeanalysesoffishassem-blageshaverevealedtwomainbiogeographicregions,theKimberleyand Canning coasts (Hutchins, 2001; Travers, Newman, & Potter,2006;Travers,Potter,Clarke,&Newman,2012;Traversetal.,2010;Wilson,2013),matchingthegeneticgroupsreportedhere.OurstudyisthefirsttoidentifyacleargeneticbreakbetweentheKimberleyandtheCanningbioregionsincoastalfish.Acrossthesamebroaderarea,studiesonthreadfinspecies(E. tetradactylum and Polydactylus macro-chir)alsoreportedastrongeast-westphylogeographicbreakbetweenRBandpopulationsofVanDiemenGulforintheGulfofCarpentaria(Horne, Momigliano, Welch, Newman, & van Herwerden, 2011;Horne,Momigliano,Welch,Newman,&VanHerwerden,2012),but
lackedintermediatesamplingbetweentheselocationsthatmayhaveidentified a Kimberley-Canning break. Similarly, a genetic study ongreymackerel (Scomberomorus semifasciatus) highlighted the distinc-tivenessbetweenWesternAustralia(PilbararegiontothesouthoftheCanning) and theNorthernTerritory, but again lacked intermediatesampling(Brodericketal.,2011).
The samples from ofWy at the mouth of the Cambridge Gulfwithin thebroaderJosephBonaparteGulf (JBG)weredistinct fromCS in theKimberley region.Thehydrodynamics of theJBG, in par-ticular the effect of the seasonal current flowand thehighbottomstress,arelikelytobefactorslinkedtotheisolationofWyfromtherest of theKimberley area (Condie, 2011).The geographic distancebetweenCSandWymaybeatleastinpartresponsibleforthisgeneticgap,recognizingthatintermediatesamplingbetweentheselocationsmayhave identifiedadditionalpopulationsubdivision.Wy ispartofageneticallyhomogenousgroup thatspans thewesternsideof theNorthernTerritory.Anexplanation for the isolationof theNorthernTerritorygroupistheyear-roundpersistentArnhemCurrent(Condie,2011)thateffectivelyisolateswatersacrosstheNorthernTerritory.
TheVIaresituatedatthewesternbaseoftheGulfofCarpentaria,anotherareadominatedbyspecificcirculationflows.TheGulfisashal-lowsemi-enclosedseawithverylongflushingtimesandlongresidencetimesforparticles (Condie,2011).Thewatercirculation intheGulf istriggeredbyaseasonalchangeinthewinddirection.TransportinandoutoftheGulfofCarpentariaisinfluencedbyseasonalwindvariationsaswellasthewestwardArnhemCurrentthatoffsetstransportintotheGulfduringthesummermonsoonseason(Condie,2011).Moreover,thesouthernpartoftheGulfofCarpentariaisevenmoreisolatedfromtheArafuraandTimorseasthanthenorthernsectionaseachpartoftheGulfexhibitsitsownseasonalcurrentpattern.Thecurrentflow,throughthetransportationofpresettlementlifestagessuchaseggsandlarvae,couldaccountfortheseparationoftheVIfromtheremainingNorthernTerritory locations. Conversely, in the study of S. semifasciatus, individ-ualsfromtheVIweregroupedwithlocationsontheNorthernTerritorycoastandnotwiththeothereasternGulfofCarpentarialocationsintheQueenslandjurisdiction(Brodericketal.,2011).
4.3 | Integration across methods: example of RB and CS locations
Despitetheoverallevidenceofastronggeneticbreak,Bayesiananal-ysisat the individual levelshowedthatRBconsistedofanadmixedpopulation,withsomefishhavinggenotypessimilartoCSindividuals(Table5).Ocean currents cangreatly influence recruitmentpatternsoforganisms through thedispersalofpropagules (e.g., Feutryetal.,
2013). Modelling of water currents along the northern Australiancoasthasshownthat the“IndonesianThroughFlow” influences themajorseasonalsouthwardsurfacecurrentalongthenorthernWesternAustraliacoast(knownasthe“HollowayCurrent”)(D’Adamo,Fandry,Buchan, & Domingues, 2009). This narrow boundary current, thatmainlyinfluencesthecontinentalshelfintheCanningandKimberleybioregionsandpredominantly flowssouthwestthroughouttheyear,albeitwithseasonalvariations (Bahmanpour,Pattiaratchi,Wijeratne,Steinberg,&D’Adamo,2016), likelyfavoursthesouthwarddispersaloflarvaefromCStoRB(Traversetal.,2010).However,seasonalvari-ationisalsoobserved,withanorthwardflowofcurrentduringsum-mer(January-March)thatcontributestowaterbuild-upintheGulfofCarpentaria.Duringthesummerspawningmonths,P. diacanthus lar-vae from CS would mainly drift northward but could still drift south-ward to RB.The seasonality and direction of theHolloway currentflowmayexplainthegeneticheterogeneityofRBsamples.AlthoughmostoftheindividualsfromRBareclosetomaturityandpossiblytheoldestfishamongoursamples(meantotallength1,018mm),Wyalsohaslargerfish(meantotallength1,061mm)anddonotpresentsuchageneticdistinctiveness.ThesegregationofgenotypesintheRBpopu-lationdidnotseemtodependontheyearofsamplingnortobelinkedto the fish ageor cohort, although the limited temporalwindowofoursamplingregimedidnotallowstatisticaltestingofthishypothesis.
Usingallthreetechniquestodetectmigrantsthroughmis-assignmentstestsforcontemporaryconnectivitybetweenlocations.Neithertheoto-lith or parasite data showed evidence ofmatchingmis-assignment togenotypesattheindividuallevelbetweenRBandCSlocations.Ifthemi-grantsarrivedasjuveniles,theywouldhavethegeneticsignatureofthesourcepopulation,butastheymature inthenewlocation,theywouldacquire the local parasite andotolith signature. If themigrants arrivedasadults,theywouldhavethegeneticandnearcoresignaturesofthesourcepopulation,andpossiblyintermediateparasitefaunacomposition.Alternatively, the genetically distinct immigrants could be the descen-dantsoftrueimmigrants—hence,theywouldbegeneticallydistinct,buthavethelocalotolith(nearcoreandmargin)andparasitesignatures.Theabsenceofamatchingpatternbetweengeneticsandecologicalmethodsinindividualmis-assignmentrevealsthatthegeneticallymis-assignedin-dividualsfromRBmay(i)notbecontemporaryimmigrantsbutratherageneticsignalofancestralimmigrantsor(ii)becontemporaryimmigrantsfromsitesnotsampled in thestudy.Fine-scalesamplingandtemporalseriescollectionwouldberequiredtofurtherinvestigatetheprocessesthatdrivethegeneticcompositionoffishintheRBpopulation.
4.4 | Management implications
ThehighlycomplexpopulationstructureofP. diacanthus shown by this studyacrossnorthernAustraliaemphasizesthelikelyvulnerabilityofthespecies tooverfishing, and importantly, guides theappropriate spatialscaleforstockassessmentandfisheriesmanagement.Inparticular,thepopulation genetics andotolith andparasites data suggest that stockassessmentsandfisheriesmanagementonaspatialscaleofaslittleas100s of kmmay be justified. Given genetic breaks among broad re-gions,itisunlikelythatimmigrationofrecruitsfromotherregionswould
14 | TAILLEBOIS ET AL.
contributetopopulationrecoveryintheshorttermshouldoneormoreregionssufferseverepopulationdeclines.Atafinerscale,theparasiteandotolithchemistryanalysessuggestthatmovementofrecruitsamongsites(e.g.,reefcomplexes)isgenerallylimitedandthatheavilydepletedsitesmaybe highly reliant on self-recruitment for populationmainte-nance and, therefore,may be slow to recover once depleted.WithintheWesternAustralia fisheries management jurisdiction, RB, CS andWyweregeneticallydistinctfromeachother, indicatingthat location-specificstockassessmentsandlocation-specificmanagementarrange-ments for P. diacanthusneedtobeconsideredbyfisheriesmanagers,oralternativelythatthelocation-specific“stocks”identifiedareconsideredwithinassessment,monitoringandmanagementframeworks.
Protonibea diacanthus is targetedbycommercial, recreationalandindigenous fisheries, which makes population modelling challengingoverlimitedspatialscalesof100sofkms.Thedifficultywithmodellingoverthisspatialscaleisthelogisticalchallengesassociatedwithcollect-ingmatching stock-specificdata suchas reproduction and spawningdynamics,growthrates,exploitationratesandnaturalmortality.Thislogisticalchallengecomeswithasignificantcostburden.Thesefactorsindicate that fine-scalemanagement is problematical once the addi-tionalcostofcompliance,regulationandenforcementisadded.Withinthatframework,considerationneedstobegiventoarangeofotheroptionsthatlimitfishingeffort,includinglimitedentry,limitingfishingeffort, consideration of targeted fisheries spatial closures, rotationalharvestsoracombinationoftheseapproacheswithothertraditionalfisheriesmanagementarrangements(e.g.,co-operativemanagementoflocalfisherieswithIndigenouspeople).Thisapproachhasbeenusedinthe Northern Territory where a series of reef- scale closure areas have beenimplementedwhichhassubsequentlyshowntoaidtherecoveryof P. diacanthus and Lutjanus johnii (NTG2016).Consideration is alsobeinggiventodevelopingfine-scalearealcatchlimitsforP. diacanthus basedoninformationfromthisstudy(ThorSaunderspers.com)demon-stratingthatmanagementatthescaleofpopulationsisachievable.
5 | CONCLUSIONS
SeverallinesofevidencesupportamodeloflowdispersalinP. diacanthus. Significant spatial geneticpopulation structurewas supportedby finerspatial-scale parasite assemblage and otolith chemistry data. There isevidenceofasmallamountofmigration—mostlikelyinthepelagiclarvalphase—fromadjacentpopulationscorrespondingtoprevailingcurrents,whichwasinferredfromintegrationacrossdatafromthreesources.
OurfindingthatP. diacanthus ishighlystructuredacrossnorthernAustraliafromRBtotheVIhasdemonstratedthatthecombinationofecologicalandgeneticmarkersprovidesamoreholisticviewofpop-ulationstructurewhendeployedtogetheronthesamesetofindivid-uals.Geneticsprovides theoverall framework forapplyingecologicalmethodsandprovidesalso considerable informationaboutexchangebetween populations. Sensitive ecological methods, such as otolithchemistryandparasites,provideresolutionofthefine-scalespatialsep-arationwithinandbetweencollectionlocations,butareunabletopro-videbroaderscale informationover intergenerational timescales.Our
ThefindingthatP. diacanthusacrossnorth-westernAustraliacon-sistsofasetofspatiallystructuredpopulationswithlowconnectivitysetsaframeworktodevelopapproachesfortheassessmentandspa-tialmanagementofthespecies.ThecollapseoftheNorthQueenslandstock(Phelan,Gribble,&Garrett,2008)andstocksinIndiaandChina(James,1994;Sadovy&Cheung,2003)demonstratesthatP. diacan-thusispronetolocaldepletionandourfindingssuggestthatthisvul-nerability is likelydue to its lowdispersalandhigh location-specificfidelity.Stockassessmentmodelsforthisspeciescannotassumefreeandfrequentmovementofindividualsamongthegeneticpopulationsstockboundaries.Becausepopulationrecoverywillmainlydependonlimitedinterchangeofearlylife-stagessuchaseggsorlarvaebetweenneighbouringpopulations(stocks),itmaytakemultiplegenerationstoreplenishdepletedstocksandthereispotentialthatsomeheavilyde-pletedstocksmaynotrecoveratall.
ACKNOWLEDGEMENTS
This research was supported by the Fisheries Research andDevelopment Corporation (Project 2013/017); LT received addi-tional funding support from the North Australia Marine ResearchAlliance (NAMRA—AIMS/ANU/CDU/NT Government) PostdoctoralFellowship.TheauthorsthankstafffromtheDepartmentofFisheries(Government of Western Australia) and the Northern TerritoryDepartment of Primary Industries as well as the Indigenous ma-rine rangers from Borroloola, Maningrida, Wadeye, Larrakia andDambimangari for the collection of specimens. We would like tothankJeffPrunster andMarkDouglas for kindlyproviding samplesfrom Roebuck Bay and Wyndham. We thank Samuel Williams attheMolecular Fisheries Laboratory and Sean Corley at theAnimalGeneticsLaboratoryfortheirassistanceinthelaboratory.Wethankthefollowingpeoplefortheiridentificationandassistancewithmanyof the parasites included within this study: Frank Moravec, GeoffBoxshall,IanBeveridge,TerryMiller,JoseLluque,JessMorgan,DelaneKritsky,LesleySmalesandShokoofehShamsi.
DATA ARCHIVING STATEMENT
Data files with individual information (genotypes, parasites assem-blageandabundanceandotolithsmicrochemistry)areavailablefromDRYAD(https://doi.org/10.5061/dryad.9rs5d).
ETHICS APPROVAL
This work was conducted under Charles Darwin University AnimalEthicspermitA13014.VoucherspecimensofparasiteswillbedepositedwithinthecollectionsoftheMuseumandArtGalleryoftheNorthernTerritory(MAGNT),theAustralianHelminthologicalCollection(AHC)andMarine Invertebrates section of the South AustralianMuseum
Abaunza,P.,Murta,A.G.,Campbell,N.,Cimmaruta,R.,Comesaña,A.S.,Dahle,G.,&MacKenzie,K. (2008). Stock identity of horsemackerel(Trachurus trachurus)intheNortheastAtlanticandMediterraneanSea:Integratingtheresultsfromdifferentstockidentificationapproaches.Fisheries Research, 89, 196–209.
Bahmanpour,M.H.,Pattiaratchi,C.,Wijeratne,E.M.S.,Steinberg,C.,&D’Adamo,N.(2016).Multi-yearobservationofHollowaycurrentalongtheshelfedgeofnorthwesternAustralia.Journal of Coastal Research, 75, 517–521.
Begg,G.A.,&Waldman,J.R. (1999).Anholistic approach to fish stockidentification. Fisheries Research, 43,35–44.
Bowtell,B.(1995July).Heedjewfishwarnings.Fish and Boat.Bowtell,B.(1998July).Hugeschoolsofblackjewabout.Fish and Boat.Bray, D. J. (2011). Black Jewfish, Protonibea diacanthus. Fishes of
Australia. Retrieved from http://www.fishesofaustralia.net.au/Home/species/671
Broderick,D.,Ovenden,J.R.,Buckworth,R.C.,Newman,S.J.,Lester,R.J.G.,&Welch,D.J.(2011).Geneticpopulationstructureofgreymack-erel Scomberomorus semifasciatusinnorthernAustralia.Journal of Fish Biology, 79,633–661.
Bush,A.O.,Aho,J.M.,&Kennedy,C.R. (1990). Ecologicalversusphy-logenetic determinants of helminth parasite community richness.Evolutionary Ecology, 4, 1–20.
Bush,A.O.,Lafferty,K.D.,Lotz,J.M.,&Shostak,A.W.(1997).Parasitologymeetsecologyonitsownterms:Margolisetal.revisited.The Journal of Parasitology, 83,575–583.
Cadrin,S.X.,Kerr,L.A.,&Mariani,S.(2013).Stockidentificationmethods:Anoverview.InS.X.Cadrin,L.A.Kerr,&S.Mariani(Eds.),Stock identi-fication methods: Applications in fisheries science(pp.1–6).Amsterdam:Elsevier.
Campana, S. E., &Thorrold, S. R. (2001). Otoliths, increments, and ele-ments: Keys to a comprehensive understanding of fish populations?Canadian Journal of Fisheries and Aquatic Sciences, 58,30–38.
Coleman, A. P. M. (2004). The national recreational fishing survey: TheNorthern Territory. Fishery report. Department of Business, IndustryandResourcesDevelopment,NorthernTerritory.
Condie, S. A. (2011). Modeling seasonal circulation, upwelling and tidalmixing in theArafuraandTimorSeas.Continental Shelf Research, 31, 1427–1436.
D’Adamo,N., Fandry, C., Buchan, S., &Domingues, C. (2009).Northernsources of the Leeuwin current and the “Holloway Current” on theNorth West Shelf. Journal of the Royal Society of Western Australia, 92, 53–66.
DeWoody, J.A.,&Avise, J. C. (2000).Microsatellitevariation inmarine,freshwater and anadromous fishes compared with other animals.Journal of Fish Biology, 56,461–473.
Elsdon,T. S., &Gillanders, B.M. (2002). Interactive effects of tempera-tureandsalinityonotolithchemistry:Challengesfordeterminingen-vironmental histories of fish. Canadian Journal of Fisheries and Aquatic Sciences, 59, 1796–1808.
Elsdon,T.S.,&Gillanders,B.M.(2004).Fishotolithchemistryinfluencedbyexposure tomultipleenvironmentalvariables.Journal of Experimental Marine Biology and Ecology, 313, 269–284.
Ghosh,S.,Mohanraj,G.,Asokan,P.K.,Dhokia,H.K.,Zala,M.S.,&Bhint,H. M. (2009). Flourishing trade of air bladders at Okha, Gujarat.Marine fisheries information service technical and extension series No. 201.CentralMarineFisheriesResearchInstitute.
Gillanders,B.M.(2002).Temporalandspatialvariabilityinelementalcom-position of otoliths: Implications for determining stock identity andconnectivityofpopulations.Canadian Journal of Fisheries and Aquatic Sciences, 59, 669–679.
Hawkins, S. J., Bohn, K., Sims, D. W., Ribeiro, P., Faria, J., Presa, P., &Burrows,M. T. (2016). Fisheries stocks from an ecological perspec-tive:Disentangling ecological connectivity from genetic interchange.Fisheries Research, 179,333–341.
Horne,J.B.,Momigliano,P.,Welch,D.J.,Newman,S.J.,&vanHerwerden,L. (2011). Limited ecological population connectivity suggestslow demands on self-recruitment in a tropical inshore marine fish(Eleutheronema tetradactylum: Polynemidae). Molecular Ecology, 20, 2291–2306.
Horne,J.B.,Momigliano,P.,Welch,D.J.,Newman,S.J.,&VanHerwerden,L.(2012).Searchingforcommonthreadsinthreadfins:PhylogeographyofAustralian polynemids in space and time.Marine Ecology Progress Series, 449,263–276.
Hutchins, J. B. (2001). Biodiversity of shallow reef fish assemblages inWesternAustralia using a rapid censusing technique.Records of the Western Australian Museum, 20, 247–270.
Izzo,C.,Ward,T.M., Ivey,A.R.,Suthers, I.M.,Stewart,J.,Sexton,S.C.,& Gillanders, B. M. (2017). Integrated approach to determiningstock structure: Implications for fisheries management of sardine,Sardinops sagax, in Australian waters. Reviews in Fish Biology and Fisheries, 27, 267–284.
James,P. S.B.R. (1994). Endangered,vulnerable and raremarine fishesandanimals.InP.Dehadrai,P.Das,&S.R.Verma(Eds.),Threaten fishes of India. Proceeding of the National Seminar of Endangered Fishes of India (pp.271–295).Muzzaffarnagar:NatureConservator.
Leis,J.M.,&Carson-Ewart,B.M.(2000).The larvae of Indo-Pacific coastal fishes: An identification guide to marine fish larvae, 2nd ed.. Leiden, Boston:Brill.
Lenanton,R.C.J.,&Potter,I.C.(1987).Contributionofestuariestocom-mercial fisheries in temperateWesternAustralia and the conceptofestuarinedependence.Estuaries, 10,28–35.
Lleonart, J., & Maynou, F. (2003). Fish stock assessments in theMediterranean:Stateoftheart.Scientia Marina, 67,37–49.
Macdonald,J.I.,&Crook,D.A.(2014).Nurserysourcesandcohortstrengthofyoungoftheyearcommoncarp(Cyprinus carpio)underdifferingflowregimes inaregulatedfloodplainriver.Ecology of Freshwater Fish, 23, 269–282.
MacKenzie,K.,&Abaunza,P.(1998).Parasitesasbiologicaltagsforstockdiscrimination of marine fish: A guide to procedures and methods.Fisheries Research, 38, 45–56.
Miller,J.A.,Banks,M.A.,Gomez-Uchida,D.,&Shanks,A.L.(2005).Acom-parison of population structure in black rockfish (Sebastes melanops)as determined with otolith microchemistry and microsatellite DNA.Canadian Journal of Fisheries and Aquatic Sciences, 62, 2189–2198.
Moore,B.R.,Buckworth,R.C.,Moss,H.,&Lester,R.J.G.(2003).Stockdis-criminationandmovementsofnarrow-barredSpanishmackerelacrossnorthernAustraliaasindicatedbyparasites.Journal of Fish Biology, 63, 765–779.
Murdoch,D.,&Chow,E.D.(2013).ellipse: Functions for drawing ellipses and ellipse-like confidence regions. R package version 0.3-8.
Oksanen,J.,Blanchet,G.,Friendly,M.,Kindt,R.,Legendre,P.,McGlinn,D.,&Wagner,H.(2017).Vegan: Community ecology package. R package ver-sion 2.4-3.
Pauly,D.,Christensen,V.,Guénette,S.,Pitcher,T.J.,Sumaila,U.R.,Walters,C. J., & Zeller, D. (2002). Towards sustainability in world fisheries.Nature, 418, 689–695.
Phelan,M. J. (2002). Fishery biology andmanagement of black jewfishProtonibeadiacanthus(Sciaenidae)aggregationsnearInjinoocommu-nity,farnorthernCapeYork.FRDC Report No. 98/135.
Phelan, M. J. (2008). Assessment of the implications of target fishingon black Jewfish (Protonibea diacanthus) aggregations in NorthernTerritory. FRDC Report No. 91. Department of Primary IndustryFisheriesandMines,NorthernTerritory.
Phelan,M.J.,&Elphick,D.(2006).Coastal line fishery status reports 2005. NorthernTerritory:DepartmentofPrimaryIndustryFisheriesandMines.
Poulin,R.,&Kamiya,T.(2015).Parasitesasbiologicaltagsoffishstocks:A meta-analysis of their discriminatory power. Parasitology, 142, 145–155.
Pritchard,J.K.,Wen,W.,&Falush,D.(2003).Documentation for STRUCTURE software: Version 2.3.
RCoreTeam(2017).R: A language and environment for statistical comput-ing.Vienna,Austria:RFoundationforStatisticalComputing.Retrievedfromhttp://www.R-project.org/
Rohde,K.(1982).Ecology of marine parasites.Brisbane,Australia:UniversityofQueenslandPress.
Sadovy,Y.,&Cheung,W.L.(2003).Nearextinctionofahighlyfecundfish:Theonethatnearlygotaway.Fish and Fisheries, 4, 86–99.
Sasaki,K. (2001).Sciaenidae.Croakers (drums).FAO species identification guide for fishery purposes. The living marine resources of the Western Central Pacific, 5,2791–3380.
Semeniuk, V. (1996). Coastal forms and Quaternary processes alongthe arid Pilbara coast of northwestern Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 123, 49–84.
daSilva,A.R.(2016).biotools: Tools for biometry and applied statistics in ag-ricultural science. R package version 3.0.Retrievedfromhttps://CRAN.R-project.org/package=biotools
Taillebois, L.,Dudgeon,C.,Maher, S., Crook,D.A., Saunders,T., Barton,D. P., … Ovenden, J. R. (2016). Characterization, development andmultiplexingofmicrosatellitemarkersinthreecommerciallyexploitedreeffishandtheirapplicationforstockidentification.PeerJ,https://doi.org/10.7717/peerj.2418.
Thackway,R.,&Cresswell, I. (1998). Interim marine and coastal regionali-sation for Australia: An ecosystem-based classification for marine and
coastal environments. Canberra: Commonwealth Department of theEnvironment.
Travers,M.J.,Newman,S.J.,&Potter, I.C. (2006). Influenceof latitude,waterdepth,dayv.nightandwetv.dryperiodsonthespeciescompo-sitionofreeffishcommunitiesintropicalWesternAustralia.Journal of Fish Biology, 69, 987–1017.
Travers,M.J.,Potter, I.C.,Clarke,K.R.,Newman,S.J.,&Hutchins,J.B.(2010).Theinshorefishfaunasoversoftsubstratesandreefsonthetropicalwest coast ofAustralia differ and changewith latitude andbioregion.Journal of Biogeography, 37, 148–169.
Tuuli,C.D.(2010).The croaker fishery and dried swimbladder trade in Hong Kong, and the reproductive biology of the greyfin croaker, Pennahia anea.TheUniversityofHongKong.Retrievedfromhttp://hdl.handle.net/10722/132328
Van Oosterhout, C., Hutchinson, W. F., Wills, D. P. M., & Shipley, P.(2004).MICRO-CHECKER: Software for identifying and correctinggenotypingerrors inmicrosatellitedata.Molecular Ecology Notes, 4, 535–538.
Venables,W.N., &Ripley, B.D. (2002).Modern applied statistics with S. Springer-Verlag.Retrievedfromhttp://www.r-project.org
Waldman, J. R. (1999). The importance of comparative studies in stockanalysis. Fisheries Research, 43,237–246.
Welch,D.J.,Buckworth,R.C.,Ovenden,J.R.,Newman,S.J.,Broderick,D.,Lester,R.J.G,&Gribble,N.A.(2009).Determinationofmanagementunits for grey mackerel fisheries in northernAustralia. FRDC Report No. 2005/010. Fishing and fisheries research centre, James CookUniversity,Townsville,Australia.
Welch, D. J., Newman, S. J., Buckworth, R. C., Ovenden, J. R.,Broderick, D., Lester, R. J. G., … Stapley, J, (2015). Integrating dif-ferent approaches in the definition of biological stocks: A north-ern Australian multi-jurisdictional fisheries example usinggrey mackerel, Scomberomorus semifasciatus. Marine Policy, 55, 73–80.
Wilson, B. (2013). The biogeography of the Australian North West Shelf: Environmental change and life’s response. Newnes: Elsevier.
Wolanski, E., & Spagnol, S. (2003). Dynamics of the turbiditymaximuminKingSound,tropicalWesternAustralia.Estuarine, Coastal and Shelf Science, 56, 877–890.
SUPPORTING INFORMATION
Additional Supporting Information may be found online in thesupportinginformationtabforthisarticle.
How to cite this article:TailleboisL,BartonDP,CrookDA,etal.Strongpopulationstructurededucedfromgenetics,otolithchemistryandparasiteabundancesexplainsvulnerabilitytolocalizedfisherycollapseinalargeSciaenidfish, Protonibea diacanthus. Evol Appl. 2017;00:1–16. https://doi.org/10.1111/eva.12499