ContentslistsavailableatFisheriesResearchjournalhomepagewwwelseviercom - PDF document

1 ContentslistsavailableatFisheriesResearc
ContentslistsavailableatFisheriesResearchjournalhomepage:www.elsevier.com/locate/ ShortcommunicationCatchascatchcan:Targetedandindiscriminatesmall-scaleshingofseahorsesinVietnamAllisonP.Stocks,SarahJ.Foster,NguyenK.Bat,AmandaC.J.Vincent Correspondingauthor.E-mailaddress:(A.P.Stocks). understandingofwildseahorsepopulationsorsheries,startingwithanunderstandingofthepressuresfacingseahorsepopulations(andVincent,2016InVietnam,seahorsesarecaughtinvastquantitiesforglobalexportanddomesticconsumption,butthisextractionisnotregulated(etal.,2006).AccordingtoCITESreports,thenumberofseahorsesex-portedfromVietnamperyearwasbetween20,000and90,000fortheyears2004through2011(UNEP-WCMC,2012).Thisislikelyanun-historicdatafrompre-CITEStradesurveysinVietnamsuggested2.2millionseahorseswerecaughtannuallyintrawlbycatchfromjustveofVietnams29coastalprovinces(Gilesetal.,2006),andCITESdataareknowntosuerfromerrorsand/oromissions(e.g.Fosteretal.,2016).Notonlyareseahorsesexportedinvastquantities,buttheyarealsoconsumedwithinVietnamastonicstopromotekidneyhealthandincreasesexualpotency(Gilesetal.,2006).SurveydataofsheriesinVietnamareoutofdate,andmayhaveunder-estimatedseahorsecatchvolumesbyfocusingonjustonetypeofgearthatcaughtseahorsesincidentally.Thereisevidencethattheexploitationofseahorsesiscausingde-clinesinwildpopulationsinVietnam(LongandVanHoang,1998;Gilesetal.,2006),ofupto95%(Stocks,2015).Despitethesedeclines,Vietnamhasnoocialmanagementmeasuresinplacetoregulateseahorseexploitation.TheconsequenceofVietnamslackofdataandofcapacitywasadecisionbyCITEStosuspendalltradeofonespeciesHippocampuskuda)fromVietnamasofMarch2013(CITES,2013Thisistherstexportbaneverimposedforanyfullymarinespeciesunderamultilateralenvironmentalagreement.IthighlightsanurgentneedtocollectsheriesandbiologicaldatainVietnamsincemaystillbetradeddomestically,andallotherseahorsespeciesinVietnammaystillbetradeddomesticallyandinternationally.AregioninsouthernVietnamreportedasahotspotforseahorsesheries(UtandTam,2012)waschosenforanin-depthstudyofdif-shingstrategiesandhowtheyaectseahorses.Theobjectivesofthisstudywereto:1)quantifytherateofseahorseextractionbyvariousshinggearsstandardizedtoeort;2)identifytheseahorsespeciescaughtandtheirlifehistorystates;3)determinetheoverallseahorsecatch;4)evaluateanindicatorofsheriessustainability;and5)eval-uatetheutilityoftheCITESrecommended10cmsizelimitasonecomponentofanadaptivemanagementplanforseahorsesheriesinsouthernVietnam.2.Materialsandmethods2.1.StudysiteThelocationofthisstudywasthePhuQuocDistrict,Vietnam,intheGulfofThailand(from9.4510.30°Nto103.55104.05°E,).InalargestudyofVietnameseseahorsebycatchandtrade,KienGiangprovinceinSouthWesternVietnamwasshowntohavethehighestseahorsecatchratesinthecountry(Gilesetal.,2006).Withinthisre-gion,anecdotalshingreportsidentiedPhuQuocasanareawhereseahorseswerecaughtinmultipleshinggears.Itisalsoanareaofconservationconcern,withshersnotingdeclinesofupto95%inseahorsecatchbetween2004and2014,despitereportedlyconsistentort(Stocks,2015).TheareacollectivelyknownasPhuQuoccontainsslargestisland,PhuQuocIsland,and21smallerisletsknownastheAnThoiIslands(landarea:593km).Fishinggroundsbetweentheislandandthemainlandaregreaterthan5000km.MaximumwaterdepthtotheeastofPhuQuocisapproximately10mandthesubstrateispredominantlysoft-bottom(sandormud)andseagrassOtero-Villanuevaetal.,2007).ThenorthandeastcoastsofPhuQuocIslandfaceinshoretomainlandCambodia(within10km)andVietnam(40km),whilethewestcoastandsouthernAnThoiislandsaremoreexposedtotheGulfofThailand.SurveyswerefocusedonportsalongtheeastcoastofPhuQuoc,shersoperatedinshallow(mostly10mdeep)groundsandcaughtseahorses.Fishinggroundsweregroupedintothreeregions:north,central,andsouth(Fig.1).Eachareacontainedonelarge,government-operatedport,aswellasmanysmallershingvil-lagesorbeacheswhereseahorseswerelanded.Thenorthernregion,closesttoCambodia,wascharacterizedbysandy,soft-bottomhabitatandoccasionalreefs.Thecentralregioncontainedsoft-bottomhabitatandpatchyseagrassbeds.Thesouthernregioncontainedthe13smallerAnThoiislands,surroundedbyreefsandsandybottom.Thenorthernandsouthernregionscontaindesignatedmarineprotectedareas(MPAs)(Fig.1).FishingactivityisprohibitedwithintheMPAcorezones,andshouldberegulatedinthebuerzones,butenforcementoftheMPAsisextremelylimitedandunregulatedshingstilloccurs.2.2.Fisheries-dependentsurveysFromApriltoJuly2014,catchesweresurveyedforspp.atninelandingsitesalongtheeastcoastoftheislands(Fig.1):sixweresmallshingvillagesorbeachesandthreewerelarge,govern-ment-operatedports.Thelocaleetusedavarietyofgeartypesincludingcompressor

2 divingequipment,trawlnets,crabnets,hooka
divingequipment,trawlnets,crabnets,hookandline,gillnets,andpurseseines.Whileallgeartypescaughtseahorsesatleastoccasionally,wefocusedoncompressordiversandbottomtraw-lers,geartypesthatregularlycatchandlandhighervolumesofsea-horses(Stocks,2015).Compressordiverswearmasksandweightbeltsandaresuppliedwithairpumpedfromtheoceansurfacethrougha300m-longplastichoseheldbythediversteeth,collectingorganismsbyhand.Thebottomtrawlersoperate612mboatswithupto60HPengines,anddragsinglenetsalongtheoceanoorthatarekeptopeneitherbyotterboards(ottertrawls)ora56mlongwoodenormetalbeam(beamtrawls).Theseshinggearswerethencategorizedbywhethertheytargetedseahorsesorcaughtthemincidentally(referredtohereafteras:targeteddivers,indiscriminatedivers,targettrawls,andindiscriminatetrawls).Siteswerevisitedinthemorningsandeveningsonaneardaily-basisinordertorecordcatchfromboatsthatshedatnightorduringtheday.Landingsweresampledfromatotalof305shingtrips(andabout100dierentboats);134tripsusedcompressordivinggearand171usedtrawlgear.Thespatialdistributionofsampledshingtripsacrossgeartypeswasasfollows(fortargeteddive,indiscriminatedive,targettrawl,andindiscriminatetrawl,respectively):northernregion:2,61,9,73;centralregion:13,54,20,37;southernregion:0,4,13,13.Samplingwasopportunisticanddidnotnecessarilyreectactualshingeortateachsite,butthiswasaccountedforinanalyses.Foreachshingtripdocumentedforthisstudy,wereaskedforinformationregardingshingeort(geartype,triplength,activeshingtime,distancefromshore,shingdepth,andfueluse),location,andbottomhabitat.Whenseahorseswerelandedthetotalnumberofindividualsand/ortotalmassofthecatchwasrecorded;onlymasswasrecordedforliveseahorses.Whenevershersallowed,thespecies,sex,height,mass,andreproductivestateofeachseahorseinthelandingswererecorded.Seahorseswereidentiedusingstandardseahorsetax-onomy(Lourieetal.,2004).SeahorseheightwasmeasuredasthelengthfromthetipofthecoronettothetipoftheoutstretchedtailLourieetal.,1999).Maleseahorseswereidentiedbythepresenceofabroodpouchor,forjuvenilemales,thepresenceofadarkenedovalzonewhereabroodpouchwasdeveloping;femaleswereidentiedbythelackofsuchfeaturesorbythepresenceofanovipositor().Maleswereconsideredmaturewherethebroodpouchwasdistendedorrecentlyemptied(Vincent,1994).Femaleswereassumedtomatureatthesamesizeasmales,asfemalematuritystatecanonlybedeterminedbydissectingovariesinfreshlydeadorpreservedspeci-mens(FosterandVincent,20042.3.Dataanalyses:sheriescatchandeAllcatchanalyseswerecarriedoutusingnumberofindividualA.P.Stocksetal. seahorses.Fornearlyallcatchlandings,boththetotalnumberandtotalmassoftheseahorsecatchwererecorded.Thesedatawereusedtocalculateaconversionfrommasstonumberofindividuals,usingagenerallinearmodel.Thisconversionwassubsequentlyusedtoconvertkilogramsofseahorsestonumberofindividualsforcatcheswhereonlymasswasrecorded.Landingsandtripdurationswereusedtocalculateseahorsecatchperuniteort(CPUE);measuresofeortwerecomparedusinggen-eralizedlinearmodelstostudytheirinuenceoncatchratesatsitesforeachgear.Analysisofvariance(ANOVA)testswereusedtoexaminemeanCPUEbysheryandshinglocation(datametassumptionsofANOVAtestsfornormalityoferrorsandequalvariance).WhereANOVAresultsweresignicant,weusedposthocgroupcomparisons(TukeyHonestSignicantDierences)todeterminewhichpairswerecantlydiTotalannualseahorsecatchwasestimatedbymultiplyingmeanCPUEofeachshery(pooledfortheentirestudyarea)bythenumberofboatsreportedlyoperatingeachgeartype(DECAFIREP,2015)andameasureofannualvesseleortbasedonweather,restdaysandgearrotation(150shingdaysperyearfordiversand200fortrawlers;Stocks,2015).Totalannualcatchwasalsocalculatedforeachlocation,usingmeanCPUEofeachsheryspecictothatlocation,extrapolatedbytheeetsizeoperatinginthatareaandthesamemeasuresofshingeort.AllstatisticalanalyseswerecarriedoutintheRstatisticalplatform(RDevelopmentCoreTeam;2.3.1.Catchcomposition,seahorsematurityandheightAgroupedchi-squaredtestwasusedtodeterminewhethertheproportionofeachspeciescaughtvariedbygeartypeandlocation.Theseproportionswerethenscaledtototalannuallandings.Mostseahorseswerecollecteddirectlyfromcatchlandings(85%,=1319).Supplementaryspecimenswerecollectedfrombuyersattheirmarketstalls,orfromseahorsecollectionsathome=227),butwereonlyusedtocalculatespeciesheightatmaturity(seebelow).Nearlyallseahorsesweremeasuredbyhand,butafewspecimenswerephotographedanddigitallymeasuredusingImageJ(U.S.NationalInstitutesofHealth,Bethesda,Maryland,USA).Alldigitallymeasuredheights(DMH)wereconvertedtohand-measuredheights(HMH)usingaconversionofHMH=2.1869(DMH)^).Aftermathematicalconversion,digitallymeasuredsampleheightsw

3 erenotsignicantlydierentfromhand-measure
erenotsignicantlydierentfromhand-measuredsampleheights(Mann-testforH.spinosissimus=0.23;H.trimaculatus=1846,=0.21).Therewasnosig-cantdierenceinmeanheightbetweendryandwetsamples-testforH.spinosissimus=17428,=0.61;=10014,=0.06)soallsampleswerepooledinsubsequentanalysis.Heightatphysicalmaturity(),thepointatwhich50%ofmaleseahorsesarereproductivelymature(butnotnecessarilyreproductivelyLawsonetal.,2014),wasdeterminedbyttingagenerallinearmodeltoabinaryclassicationofmalesasimmature(0)ormature(1)andcalculatingtheinectionpoint.Usingthismodeledforeachspecies,shwereassignedtojuvenileoradultinordertocalculatethesexratio(thenumberofmaturemalestothenumberofmaturefemalesofaspecies).Achi-squaretestwasusedtoidentifyiftheratiosweresignicantlydierentfromunityforeachTheindicatorofcatchsustainabilityusedinthisstudywastheproportionofthecatchthatwaslessthanHFroese,2004),whichwascalculatedforeachspeciesandeachshery.Finally,theHforeachspecieswascomparedto10cminordertodeterminethebiologicalprotectionconferredbytheCITESrecommended10cmMSL(CITES,2004),andtheproportionofseahorsecatchbydiversortrawlersbelow10cmwascalculatedtoassesshowthesizelimitwouldreducecatchesifenforced.3.Results3.1.DescriptionofAnestimated124bottomtrawlboatsand46compressordivingboatsoperatedfromtheislandsandregularlycaughtseahorses.Ofthistotal,sixdiveboatsand24trawlboats(amixofotterandbeamtrawls)soughtoutseahorsesspecically.Mostboats,however,shedindis-criminately(40diveboatsand100trawlboats),nottargetinganythingcally,butcatchingseahorsesalongwithavastnumberofotherspecies.Diversthattargetedseahorsesshedinareasthatweresandyoramixofsandanddebris.Trawlboatsthattargetedseahorsesoper-atedclosertoshoreinseagrassbedsandsandy-coarsesubstrate.Diversthatdidnottargetseahorsesshedinareasofsandorcoralreef,whiletrawlsthatshedindiscriminatelyoperatedinopen,sandyareas. Fig.1.MapofstudylocationinsouthernVietnamshowingPhuQuocIslandandthethirteenAnThoiIslands,aswellasmainlandVietnam(darkgrey)andCambodia(lightgrey).Starsindicatesitessampledsheriesdata(fromnorthtosouth):XaLuc,BaiNo,DaChong,BaiBon,HamNinhport,HamNinhvillage,AnThoi,HonThom,andHonRoi.TheareacontainstwoMPAs;solidlinesdenoteareasdesig-natedforfullprotection,dashedlinesdenotethedesignatedbuerzonearoundtheprotectedareas.Dot-dashlinesdenoteshingareas(north,central,andsouth).A.P.Stocksetal. 3.2.Catchperuniteort(CPUE)Fishingtripsthattargetedseahorseshadahigherlikelihoodofcatchingtheminanygiventripthandidtripsthatshedprimarilyforotherspecies.About93%oftargetdivingtripscaughtatleastoneseahorse(=14/15),whileonly36%ofindiscriminatedivetripscaughtatleastoneseahorse(=43/119).Similarly,95%oftargettrawltripscaughtatleastoneseahorse(=40/42),whileonlyabout50%oftheindiscriminatetrawltripscaughtatleastoneseahorse=61/123).Thecatchperuniteort(CPUE,seahorsesboat)wassig-cantlydierentamongtargetdivers,indiscriminatedivers,targettrawls,andindiscriminatetrawls(ANOVA,F=98.25,df=3,0.001).AposthocTukeytestshowedthattargetshingwascantlydierentfromnon-targetshing(0.001),butwithintargetandnon-targetshing(i.e.targetdiversandtargettrawls,andnon-targetdiversandnon-targettrawls)therewasnosignicantdif-ference(&#x-451;&#x.1 0;0.05).TargetdivershadthehighestCPUE,followedbytargettrawls(Table1).Incontrast,theCPUEfromindiscriminatetrawlswasmuchlower,whileindiscriminatedivershadthelowestCPUE(Table1).Seahorsecatchratesalsovariedregionally(F=34.18,df=2,0.001),witheachgroupbeingsignicantlydierentbasedonaposthocTukeytest(0.001).ThehighestCPUEfortargetdiverswasinthecentralregionwhileindiscriminatedivershadaslightlyhigherCPUEinthenorththanthecentralregion(Table2CPUEwashighestinthesouthforbothtargettrawlsandindiscriminatetrawls(Table23.3.ExtrapolationoftotalseahorsecatchperannumWeestimatedthatseahorselandingswerebetween269,000individualsyearwiththemajority(82%)comingfromthetargetandindiscriminatetrawlsheries(Table1).Thesenumbersrepresent12001800kgyearbasedonanestimated165liveseahorseskg.Estimatedtotalannualcatchesvariedsignibasedontheareainwhichboatsshed;thehighestvolumesofseahorseswerecaughtinthenorthandsouthbytargettrawls,followedbytargetdiversthatshedinthecentralregion(Table23.4.SeahorselandingsandsizeatmaturitybyspeciesThreespeciesofseahorseswereidentiedinlandings,whosepro-portionsincatchvariedbygearandlocation:thecommonseahorseHippocampuskudaBleeker1852,thehedgehogseahorseWeber1913,andthethree-spotseahorseLeach1814.Catchcompositionvariedbygeartype:targetdiverscaughtthehighestproportionofH.spinosissimus(n=371/996),whileindiscriminatetrawlsandtargettrawlscaughtthehighestpro-portionofH.trimaculatus(n=121/181)andH.ku

4 da(n=244/289)respectively(=494,d.f.=6,Hi
da(n=244/289)respectively(=494,d.f.=6,Hippocampusspino-dominateddivercatchesbynumberofindividuals(91.4%c.f.,H.trimaculatusand0.2%H.kudafortargetdivers;70.8%c.f.,H.trimaculatusand19.0%H.kudaforindiscriminatedivers).Ontheotherhand,trawlcatchesvariedintheirspeciescomposition:therelativeproportionofH.spinosissimus,H.trimaculatusH.kudanumberofindividualsfortargettrawlswas56.8,2.0,and41.2%,re-spectively,andforindiscriminatetrawlswas58.0,36.6,and5.4%,respectively.Whenscaledtoannualcatchnumberbyeetsize,targettrawlscaughtthegreatestnumberofH.spinosissimusH.kuda,whileindiscriminatetrawlscaughtthemostH.trimaculatusFig.2Catchcompositionalsovariedregionally(=184,d.f.=4,0.001).Catchinthenorthcomprised81.3%H.spinosissimus(n=295),10.2%H.trimaculatus(n=37),8.5%H.kuda(n=31);centralcatch:58.0%H.spinosissimus(n=467),10.0%H.trimaculatus(n=81),and32.0%H.kuda(n=258);southerncatch:78.7%(n=236),21.3%H.trimaculatus(n=64),andnoH.kuda(n=0).Whenscaledtoannualcatchnumberbyeetsize,thevastmajorityofH.kudawerecaughtinthecentralregion(70.2%,n=258),whilethemajorityofH.trimaculatusH.spinosissimuswerecaughtinthesouth(83.5%n=64,and56.7,n=236,respectively).Forallgeartypesandspecies,medianheightofcapturedseahorseswasremarkablysimilar(Fig.3).Therewasnodierenceamongmean Table1Meancatch-per-unit-eortofseahorses±standarderror(seahorsesboat)andestimatedtotallandingsperannumforthetrawlandcompressordivereetsofPhuQuocIsland,Vietnam.=numberofshingtripssampled.FisherynMeanCPUE(±S.E.)FleetSizeshingdaysperannumAnnuallandings(seahorsesyearMeanLower95%CIUpper95%CITargetdivers1531.8(±5.6)615028,62018,74238,498Indiscriminatedivers1171.3(±0.2)401507800544810,152Targettrawls4223.3(±2.9)24200111,84084,557139,123Indiscriminatetrawls1232.5(±0.8)10020050,00018,64081,360Total198,260127,386269,134*EstimatesofeetsizeandnumberofdaysshedarebasedoninterviewswithshersincommunitiesacrossPhuQuocIsland(Stocks,2015)andconsultationswithVietnamesesherieso Table2Meancatch-per-unit-eortofseahorses±standarderror(seahorsesboat)ofeachsherybyregionshedandestimatedtotallandingsperannum(seahorsesyear)forPhuQuocIsland,Vietnam.n=numberofshingtripssampled.NorthCentralSouthnMeanCPUE(±S.E.)FleetSizeAnnualnMeanCPUE(±S.E.)FleetSizeAnnualnMeanCPUE(±S.E.)FleetSizeAnnualTargetdivers214.5(±3.5)121751334.5(±6.1)525,8750NA0NA591.4(±0.3)153150541.2(±0.4)10180040.5(±0.5)151125Targettrawls918.1(±6.3)1243,4402014.7(±2.3)411,7601340.1(±5.4)864,160731.7(±0.6)3010,200373.2(±2.2)106400135.2(±1.9)6062,400A.P.Stocksetal. heightofH.kudaH.trimaculatuscaughtacrossthefour(F=2.0,df=3,=0.14;F=0.2,df=3,0.90,respectively).MeanheightsofH.spinosissimusdiddieramongthefourdisheries(F=9.75,df=3,0.001),withthesmallestseahorsescaughtbytargetdivers(posthocTukeytestSpeciesdieredminimallyinheightat50%malephysicalmaturitywas98.5mmforH.spinosissimus(95%C.I.=96.2100.8mm,=324),108.8mmforH.trimaculatus(95%C.I.=103.4114.2mm,=130)and106.1mmforH.kuda(95%C.I.=96.2116.1mm,=26)(Fig.3Innearlyallsheries,therewasanequalsexratioforH.kuda.MaletofemaleratiosofH.kudafortargetdiverswere:0:1,114:100=0.34),and15:15(=1);forindiscriminatediverswere9:11=0.65),32:22(=0.17),and7:2(=0.1);fortargettrawlswere64:55(=0.41),22:33(=0.14)and2:2(=1);andforindis-criminatetrawlswere9:8(=0.81),67:43(=0.02),and41:44=0.74).Onlycaughtincidentallyintrawlnetsshowedaskewedsexratio,withmorematuremalessampledthanmaturefemales(61%males,=5.24,df=1,=0.02).3.5.IndicatorsofcatchsustainabilityIndiscriminatediversandindiscriminatetrawlscaughtthehighestproportionsofimmatureseahorses,togethercatchinguptothreetimesmorejuvenilesthantargetshers.TheonlycasewhereindiscriminateshingresultedinalowerproportionofjuvenileswasforH.kudacaughtintrawlnets.TheproportionofjuvenileH.kuda,H.spinosissimusH.trimaculatuswas:0,33and12%,respectively,fortargetdivers;5,40and40%,respectively,forindiscriminatedivers;9,17and0%,respectively,fortargettrawls;and6,31and29%,respectively,forindiscriminatetrawls.TheproportionofcatchesofH.kuda,H.spinosissimusH.tri-lessthan10cmheight(therecommendedminimumsizelimit)was6,39and0%,respectively,fordivers,and10,32and11%,respectively,fortrawls.4.DiscussionThisistherstquantitativecomparisonoftargetedandincidentalseahorsecapturebytwoshingmethods:bottomtrawlingandcom-pressordiving.Wedemonstratethat,whilediverscaughtseahorsesat Fig.2.Estimatedannualcatchofseahorses(spp.)bytargetandindiscriminatediveandtrawlsheriesoperatingacrossthreegroundsonPhuQuocIsland,Vietnam(H.spinosissimusinblack,H.tri-ingrey,andH.kudainwhite). Fig.3.Comparisonofthesizeofseahorsescaughtbytargetandindiscriminatediveandtrawlsherieso

5 fPhuQuocIsland,Vietnam.MedianheightofH.k
fPhuQuocIsland,Vietnam.MedianheightofH.kudacaughtbyindiscriminatedivers(ID)=128mm,=15;targettrawls(TT)=118mm,=4;indiscriminatetrawls(IT)=115mm,=11;theoneH.kudameasuredfromtargetdivers(TD)was121mmheight.MedianheightofH.spinosissimuscaughtbyTD=102mm,=145;ID=108mm,=68;TT=119mm,=28;IT=104mm,=152.MedianheightofH.trimaculatuscaughtbyTD=120mm,=22;ID=114mm,=14;theoneH.trimaculatusmeasuredfromtargettrawlswas118mmheight;IT=120mm,=112.Widthofboxplotsisproportionaltosamplesize.Solidhorizontallinesindicateheightat50%maturity,withcondenceintervalsindicatedbydottedhorizontallines.A.P.Stocksetal. similarratestotrawls,bottomtrawlinghadamuchgreateroverallimpactonseahorsepopulationsbecauseofthelargescaleoftheThelargeoverallcatchfromPhuQuocIslandpointstowardahighvolumeofunmonitoredandunregulatedseahorseexploitationinVietnam.Fishingpressureintheareamaybeleadingtosize-speciimpactsandaectingthereproductivecapacityofsomeseahorsespe-cies.Regularmonitoringshouldbeexecutedinordertodeterminethelong-termimpactsofseahorseshingintheareaandaidthecontinuousadaptationofeectivemanagementforsustainablewildseahorsepo-pulations.Inthemeantime,precautionarymanagementeortsin-cludingminimumsizelimitsandimprovedMPAenforcementandcomplianceshouldbeimplementedtomitigateanyimpacts.Thisstudyshowshowasnapshotinvestigationcanbeusedtoinfersheriesimpactsintheabsenceoflong-termdata(Johannes,1998;Pillingetal.,2008),theusualscenarioforspeciesnotregardedasna-tionallyimportant.Inspeciesotherthanseahorses,suchsnapshotstu-diesofdata-poorsherieshavebeendeployedtoaidmanagement,compensatingforthedearthoftimeseriesinformation(Dowlingetal.,2008;MacCall,2009).Vietnamshowsapatternthatisconsistentwithseahorseresearchelsewhere,revealingthatrelativelylowseahorsecatchratesscaleuptolargecatchesafteraccountingforintensepressure(Lawsonetal.,2017).Thisstudyaccountedforthevastma-jorityofshingpressureonseahorsesinPhuQuoc,Vietnam,byen-compassingtheshingstrategiesthatcatchthelargestvolumesofseahorses(Stocks,2015).Nonetheless,manyothergeartypesusedintheareareportedlycatchseahorsesatleastoccasionally(Stocks,2015andsothetotalannualcatchislikelyhigherthanestimated.Moreover,thisstudyusedaconservativenumberofshingdayswhenconsideringpossibleseasonalvariationsincatchandchangesinsherbehaviour.Oursnapshotstudyprovidesafoundationforthelongtermmonitoringweproposebelow;longtermmonitoringwillallowrenementofcatchestimationsandidenticationofpatternsovertime.PhuQuocisoneofthefewdocumentedplaceswheretargetedshingoccurs;globally,about95%oftradedseahorsesarecaughtincidentally(Vincentetal.,2011).Moreover,shersonPhuQuocusebothdivinggearandtrawlnetstotargetseahorseswhereasallotherdocumentedtargetsheriesinvolvedcollectiononlybyhand(e.g.Brazil,India,IndonesiaandthePhilippines;MarichamyandLipton,1993;Rosaetal.,2011,2005;Vincentetal.,2007).TargetshersonPhuQuoccaughtseahorsesatratesthatwerehigherthanalldocu-mentedcasesofseahorseshingaroundtheworld(e.g.Gilesetal.,2006;Vincentetal.,2007;Perryetal.,2010).PaststudiesofseahorsesheriesandtradeexecutedinVietnamestimatedseahorseCPUEtobebetween0.12.5seahorsesboatbasedoninterviewandcatchlandingsdata(Gilesetal.,2006;Meeuwigetal.,2006).OtherestimatesthroughoutAsiainclude110seahorsesboatintheGulfofMannar,India(SalinandYohannan,2005),03seahorsesinMalaysia(Perryetal.,2010),2.93.4seahorsesinthePhilippines(Vincentetal.,2007),and0.6seahorsesboatinThailand(Perryetal.,2010).Intheseeries,seahorseswerecaughtasbycatch,althoughitispossiblethatboatstargetseahorsesintheseareas(asmentionedinPerryetal.,).FurtherinvestigationsofseahorseshinginVietnam(andelsewhere)mustexplicitlynotewhethertheseahorsesarebeingcaughtintentionallyorindiscriminatelyinordertoimprovecatchestimates.TheestimatedtotalannualseahorselandingsfromPhuQuocex-ceedsVietnamsannualdocumentedexports(UNEP-WCMC,2015),sotheexcessiseitherexportedillegallywithoutpermits,exportedwithpermitsbutgoesunreported,orconsumeddomestically.AccordingtocialCITESstatistics,whichtendtounderestimatetrueexports,thereportedseahorseexportsfromVietnamrangedfrom20to90,000in-dividualsperyearbetween2004and2011(Fosteretal.,2016).Thesecialexportsfortheentirecountryarelessthan45%ofourestimatedtotalannualseahorsecatchfromPhuQuocalone,andlessthan5%ofhistorictotalannualseahorsecatchestimatesinVietnam(Gilesetal.,).SeahorsesareconsumedinVietnamashealthtonicsbutthevolumeisunknownitseemsunlikely,however,thatdomesticcon-sumptionaccountsfortherestoftheunreportedcatch,sinceithasbeenobservedtoberelativelylow(Gilesetal.,2006).Previoustradesurveyssuggestthatillegal,unreportedandunregulated(IUU)exportsmaywelloccur(

6 Gilesetal.,2006),supportedbywelldocument
Gilesetal.,2006),supportedbywelldocumentedIUUtradeofawidevarietyofotherplantsandanimalsfromVietnam(2006;NgocandWyatt,2013PhuQuocsheriesmaybeplacingsize-specicpressureonseahorsepopulations,anditisunclearifcurrentselectivityisavoidinggrowthshing.Theseahorsesinourstudywere,onaverage,25%smallerbyweightthanseahorsesexaminedinasimilarsheriesstudyincen-tralVietnam,wherethecatchcompositionwasveryneartothatofPhuQuoc(comprisedmostlyofH.spinosissimusH.trimaculatusMeeuwigetal.,2006).Suchadierencemaysuggestadeclineinadultsizeoverthepasttenyears,ashasbeenreportedbyshersintheareaStocks,2015),butitcouldalsoreectgeographicdierencesinsea-horsepopulationsorseasonalinuencesonseahorsesize(andVincent,2005;Meeuwigetal.,2006Themajorityofseahorsessampledinthisstudywerelargerthanthemodeledspeciesheightat50%physicalmaturity,whichsuggestspo-pulationsarenotsubjecttogrowthovershing.Thatsaid,thisstudymayunderestimatetheproportionofindividualsbeingcaughtbeforetheycanreproduce;thesizeatwhichseahorsesstartreproducingislargerthanthesizeatwhichtheyarecapableofreproducing(al.,2014),andstudieshaveobservedthatfemalesmaturelaterthanmales(Thangarajetal.,2006H.kudaLongtermmonitoringofseahorsecatchandeortisneededtormwhetherPhuQuocseahorsesarevulnerabletocurrentlevelsofshingpressure.Manyofseahorseslifehistorytraitsmayincreasetheirsusceptibilitytooverexploitation,includinglengthyparentalcare,smallbroodsizes,lowmobility,smallhomerangesandsparsedis-tribution(asreviewedinFosterandVincent,2004).Furthermore,in-terviewswithshersonPhuQuocrevealeddeclinesofupto95%inseahorseCPUEbetween2004and2014,abovewhatcouldbeac-countedforwithincreasedshingeort(Stocks,2015).Asimplemonitoringprogramcouldprovideindicationsofwhetherseahorsesarebeingovershed(e.g.Fosteretal.,2014);eortsshouldfocusontheeetsoftargettrawlboatsanddiversthatcatchhighervolumesthantheindiscriminateeets.ThisessentialinformationonCPUEandinferredseahorseabundancewillaidVietnamscapacitytodevelopndings,adheretotheCITESrecommendations,anddevelopanationalmanagementplanforseahorses.Thehighvolumeofunmitigatedextractiondeterminedinthisstudy,combinedwithreportedcatchdeclines(Stocks,2015),suggeststhatPhuQuocsseahorsesherieswouldbenetfromprecautionaryman-agementtoolsinordertoensuretheirsustainability.Enforcingtheis-twomarineprotectedareasandencouragingsherstorespectthesereserveswouldbenetseahorsebenthichabitats,thoughthedocumentedbenetsofsmallMPAstoseahorsepopulationsarelimitedYasuéetal.,2012;Harastietal.,2014).Insteadlarge-scaleremovalofshinggearsmaybemoreeectiveatprotectingseahorsesYasuéetal.,2012).Seasonalshingrestrictionsexistinthearea,butareminimallyenforced.Minimumsizelimitsareoflittleusefortrawlsbecausethesegearscannoteasilyselectforsizeandsurvivalofdis-cardedseahorsesisassumedtobeverylow(FosterandVincent,2016Enforcingthe10cmsizelimitrecommendedbyCITES(CITES,2004could,however,beaneectivetoolformanagingdiverswhocanbemoreselectiveaboutwhatsizeseahorsestheyextract.The10cmMSLwouldconferreasonableprecautionaryprotectionofthethreespecies,sincetheymaturebeloworjustslightlyabovethisheight.ThisisacontributionfromProjectSeahorse.ThisstudywasmadepossiblethroughcollaborationswiththeResearchInstituteofMarineFisheries(RIMF),VietnamNationalUniversity,theVietnamCITESA.P.Stocksetal. ManagementAuthority,andthePhuQuocMPAsta.ManythankstoDoThanhAnandNguyenNhutThanhwhoprovidedassistancewithdatacollection,interpretation,andlogistics.ThankyoutoL.Aylesworth,C.Kinzett,K.Gillespie,J.Lawson,andothersfortheirsupportandreviews.WegreatlyappreciatethesupportforthisresearchprovidedbyMitacsandTheExplorersClub,andtheadditionalsupportprovidedbyGuylianChocolatesBelgiumandananonymousdonor.AppendixA.SupplementarydataSupplementarydataassociatedwiththisarticlecanbefound,intheonlineversion,atAthukorala,P.,2006.TradepolicyreformsandthestructureofprotectioninVietnam.WorldEcon.29,162Boisseau,J.,1967.LesregulationshormonalesdelincubationchezunVertébrémâle:recherchessurlareproductiondelCITES,2004.TradeinSeahorsesImplementationofDecision12.54,NoticationtothePartiesNo.2004/033CITES,2013.ImplementationofResolutionConf.12.8(Rev.CoP13),ReviewofcantTradeinSpecimensofAppendix-IISpecies,NoticationtothePartiesNo.DECAFIREP,2015.KienGiangData.HaNoiDowling,N.,Smith,D.,Knuckey,I.,Smith,A.,Domaschenz,P.,Patterson,H.M.,Whitelaw,W.,2008.Developingharveststrategiesforlow-valueanddata-poorsheries:casestudiesfromthreeAustraliansheries.Fish.Res.94,380Foster,S.J.,Vincent,A.C.J.,2004.Lifehistoryandecologyofseahorses:implicationsforconservationandmanagement.J.FishBiol.65,1Foster,S.J.,Vincent,A.C.J.,2016.MakingNon-De

7 trimentFindingsforSeahorsesFramework,Ver
trimentFindingsforSeahorsesFramework,Version4.(Vancouver,Canada)Foster,S.J.,Loh,T.-L.,Knapp,C.,2014.iSeahorseLandingsTrendsToolkitFoster,S.J.,Wiswedel,S.,Vincent,A.C.J.,2016.OpportunitiesandchallengesforanalysisofwildlifetradeusingCITESdataseahorsesasacasestudy.Aquat.Conserv.Mar.Freshw.Ecosyst.26,154Froese,R.,2004.Keepitsimple:threeindicatorstodealwithovershing.FishFish.5,Giles,B.G.,Ky,T.S.,Hoang,D.H.,Vincent,A.C.J.,2006.ThecatchandtradeofseahorsesinVietnam.Biodivers.Conserv.15,2497Harasti,D.,Martin-Smith,K.,Gladstone,W.,2014.Doesano-takemarineprotectedareatseahorses?PLoSOne9,e105462Johannes,R.,1998.Thecasefordata-lessmarineresourcemanagement:examplesfromtropicalnearshoresheries.TrendsEcol.Evol.13,243Lawson,J.M.,Foster,S.J.,Lim,a.C.O.,Chong,V.C.,Vincent,A.C.J.,2014.Novellife-historydataforthreatenedseahorsesprovideinsightintosheryeects.J.FishBiol.86,1Lawson,J.M.,Foster,S.J.,Vincent,A.C.J.,2017.Lowbycatchratesadduptobignumbersagenusofsmallshes.Fisheries42,19Long,N.,VanHoang,D.H.,1998.BiologicalparametersoftwoexploitedseahorsespeciesinaVietnameseshery.Mar.Biol.SouthChinaSea3,449Lourie,S.A.,Pritchard,J.C.,Casey,S.P.,Truong,S.K.,Hall,H.J.,Vincent,A.C.J.,1999.ThetaxonomyofVietnamsexploitedseahorses(familySyngnathidae).Biol.J.Linn.Soc.66,231Lourie,S.,Foster,S.,Cooper,E.,Vincent,A.,2004.AGuidetotheIdenticationofSeahorses.UniversityofBritishColumbiaandWorldWildlifeFund,Washington,DCMacCall,A.,2009.Depletion-correctedaveragecatch:asimpleformulaforestimatingsustainableyieldsindata-poorsituations.ICESJ.Mar.Sci.J.66,2267Marichamy,R.,Lipton,A.,1993.Largescaleexploitationofseahorse(HippocampuskudaalongthePalkBaycoastofTamilNadu.Mar.Fish.Inf.Serv.Tech.Ext.Ser.119,Martin-Smith,K.M.,Vincent,A.C.J.,2005.SeahorsedeclinesintheDerwentestuary,Tasmaniaintheabsenceofshingpressure.Biol.Conserv.123,533Meeuwig,J.J.,Hoang,D.H.,Ky,T.S.,Job,S.D.,Vincent,A.C.J.,2006.Quantifyingnon-targetseahorsesheriesincentralVietnam.Fish.Res.81,149Ngoc,A.,Wyatt,T.,2013.AgreencriminologicalexplorationofillegalwildlifetradeinVietnam.AsianJ.Criminol.8,129Otero-Villanueva,M.,del,M.,Ut,V.N.,2007.SeacucumbersheriesaroundPhuQuocArchipelago:across-borderissuebetweenSouthVietnamandCambodia.SPCBeche-de-merInf.Bull.25,32Perry,A.L.,Lunn,K.E.,Vincent,A.C.J.,2010.Fisheries,large-scaletrade,andconserva-tionofseahorsesinMalaysiaandThailand.Aquat.Conserv.Mar.Freshw.Ecosyst.20,Pilling,G.M.,Apostolaki,P.,Failler,P.,Floros,C.,Large,P.A.,Morales-Nin,B.,Reglero,P.,Stergiou,K.I.,Tsikliras,A.C.,2008.Assessmentandmanagementofdata-poorsheries.In:Payne,A.,Cotter,J.,Potter,T.(Eds.),AdvancesinFisheriesScience:50YearsonfromBevertonandHolt.BlackwellPublishing,pp.280Rosa,I.M.,Alves,R.R.,Bonifácio,K.M.,Mourão,J.S.,Osório,F.M.,Oliveira,T.P.,Nottingham,M.C.,2005.FishersknowledgeandseahorseconservationinBrazil.J.Ethnobiol.Ethnomed.1,12.Rosa,I.L.,Oliveira,T.P.R.,Osório,F.M.,Moraes,L.E.,Castro,A.L.C.,Barros,G.M.L.,Alves,R.R.N.,2011.FisheriesandtradeofseahorsesinBrazil:historicalperspective,currenttrends,andfuturedirections.Biodivers.Conserv.20,1951A.,Haywood,M.,2002.GuidanceForCITESScienticAuthorities:ChecklisttoAssistinMakingNon-DetrimentFindingsforAppendixIIExports.IUCNNo27Salin,K.R.,Yohannan,T.M.,2005.Fisheriesandtradeofseahorses,spp.,insouthernIndia.Fish.Manag.Ecol.12,269Stocks,A.P.,2015.DiversicationandDepletionofVietnameseSeahorseFisheries.TheUniversityofBritishColumbiaThangaraj,M.,Lipton,A.,Victor,A.,2006.OnsetofsexualmaturityincaptiverearedendangeredIndianseahorse,Hippocampuskuda.Curr.Sci.91,1714UNEP-WCMC,2015.CITESTradeDatabase.Availableat(lastAccessed15August2015).Ut,V.N.,Tam,T.C.,2012.Speciescompositionandshingstatusofseahorses(Hippocampusspp.)inPhuQuocIsland.In:ProceedingsofInternationalConferenceonBienDong.Vietnam.pp.1Vincent,A.C.J.,Meeuwig,J.J.,Pajaro,M.G.,Perante,N.C.,2007.Characterizingasmall-scale,data-poor,artisanalshery:seahorsesinthecentralPhilippines.Fish.Res.86,Vincent,A.C.J.,Foster,S.J.,Koldewey,H.J.,2011.Conservationandmanagementofseahorsesandother.J.FishBiol.78,1681Vincent,A.C.J.,SadovydeMitcheson,Y.J.,Fowler,S.L.,Lieberman,S.,2013.TheroleofCITESintheconservationofmarineshessubjecttointernationaltrade.FishFish.15,563Vincent,A.C.J.,1994.Seahorsesexhibitconventionalsexrolesinmatingcompetition,despitemalepregnancy.Behaviour128,135Yasué,M.,Nellas,A.,Vincent,A.C.J.,2012.Seahorseshelpeddrivecreationofmarineprotectedareas,sowhatdidtheseprotectedareasdofortheseahorses?Environ.Conserv.39,183ElectronicReferencesIUCN(2013).IUCNRedListofThreatenedSpeciesVersion2013.Availableat(lastAccessed15August2015).A.P.Stockseta