?Correspondingauthor.Tel.:+49-471-4831-1652;fax:+49-471-4831-1425.
E-mailaddress:cpusch@awi-bremerhaven.de(C.Pusch).
ARTICLEINPRESS1686
C.Puschetal./Deep-SeaResearchI51(2004)1685–1708
stockbybothspecieswasestimatedtoamountto11.1–26.7%intheSouthShetlandIslandsregion.Thisestimateemphasizestheimportantroleofmesopelagic?shintheAntarcticecosystemasaprevalentconsumerofkrill.r2004ElsevierLtd.Allrightsreserved.
Keywords:Marine?sh;Myctophidae;Mesopelagiczone;Communitystructure;Stomachcontent;Predation;Dailyration
1.Introduction
One-quarterofallknown?shspeciesintheSouthernOceanliveinthemesopelagicandbathypelagiczones(Kock,1992).Myctophidsarethedominant?shfamilyinthesezones,asfarasdiversity,biomassandabundanceareconcerned.Thirty-threemyctophidspeciesareknownfromtheSouthernOcean,ofwhich11haveacircum-polardistribution.Althoughtheirgeographicaldistributionandtaxonomyhavebeendescribed(Hulley,1981;McGinnis,1982),comparativelyfewstudieshaveexaminedtheverticaldistributionofoceanicmicronektonbyintensivedepthstrati-?edsampling(TorresandSomero,1988;Lancraftetal.,1989;Piatkowskietal.,1994;Duhamel,1998).AllthesestudiesindicatethatthecommonAntarcticmyctophidsaredielverticalmigrators.KingGeorgeIslandislocatedinthesouthernpartoftheDrakePassageandisstronglyin?uencedbytheAntarcticCircumpolarCurrent(ACC).TheACCisthemajoroceanographicfeatureoftheSouthernOcean;itisanextensiveeastward?owingcircumpolarcurrent(Hofmannetal.,1996).TheupperwatersoftheACCinthestudyareacomprisedAntarcticSurfaceWater(ASW)andtheassociatedCircumpolarDeepWater(CDW),which?owsfromtheBellingshau-senSeaintosouthernDrakePassage(SteinandHeywood,1994).ThestudysiteontheslopeofKingGeorgeIslandischaracterizedbyashelf-breakfrontresultinginenhancedproductionandahigherkrillabundancecomparedtooceanicwaters.ForthisreasontheareanorthofKingGeorgeIslandisoneofthemostimportantkrill?shingregionsoftheSouthernOcean(Ichiietal.,1996).
Myctophidsplayasigni?cantroleasconsumersofzooplanktoninthefoodweboftheSouthernOcean(Lancraftetal.,1989).Asintheotherworldoceans,theyoccupythethirdlevelandareconsumersofthesecondorder.Theyareanimportantfoodsourceforthepredatorsofhighertrophiclevelslikebenthopelagic?sh(Bulmanetal.,2002),seabirds(Guinetetal.,1996),furseals(Chereletal.,1997)andsquid(Rodhouseetal.,1992;Phillipsetal.,2001).Anyestimationofenergytransportwithinthepelagicsystemmustincludeanalysesoftheindividualdietcompositionofthemesopelagic?shandtheirratesoffoodconsumption.
Althoughthedietcompositionofthemostabundantmyctophidspeciesisdocumented(Ro-wedder,1979a;AscenioandMoreno,1984;Williams,1985;KozlovandTarverdiyeva,1989;Lancraftetal.,1991;Hoddelletal.,2000)onlyafewstudieshaveinvestigatedfeedingchronology(Rowedder,1979a)andestimateddailyrations(Gerasimova,1990;Pakhomovetal.,1996).
Krill(Euphausiasuperba)playsanimportantroleasthekeypreyitemofanumberoftoppredators,especiallyintheAtlanticSectoroftheSouthernOcean(Barlowetal.,2002).Becauseoftheirhighbiomass(thetotalstockoftheSouthernOceanwasestimatedbyLancraftetal.(1989)toaccountfor133–191milliontonnes),mesopelagic?shcouldbeoneofthemostimportantpredatorsofoceaniczooplankton(Lancraftetal.,1989;Pakhomovetal.,1996).Numerousstudieshaveshownthatmyctophidsplayasigni?cantroleintheconsumptionofjuvenileandadultkrill(Rembiszewskietal.,1978;Rowedder,1979a;Williams,1985).Thisconclusionhasmorerecentlybeenchallengedbyasuggestionthatasubstantialconsumptionofkrilloccursonlyduringcertainperiodsandwithinspeci?cregions(Pakhomovetal.,1996).
Finally,itshouldbenotedthatapreliminaryanalysisofcommunitystructureofthemesopela-gic?shassemblagefromcruiseANTXIV/2ofRV
ARTICLEINPRESSC.Puschetal./Deep-SeaResearchI51(2004)1685–1708
1687
‘‘Polarstern’’hasalreadybeenpresentedbyHulleyetal.(1998),andtheresultsusedforacladisticanalysisofthemyctophidtribeElectronini(Hul-ley,1998).Inthepresentstudy,however,commu-nitystructureofthemesopelagiccommunityovertheslopeofKingGeorgeIslandisanalysedingreaterdetail.Thecommunitypatternandtheverticaldistributionarerelatedtothefeedingecologyofthemostabundantmesopelagic?shspecies.Fourmyctophidsandonegempylidareanalysedwithrespecttodiet,feedingchronologyanddailyrations.Inconclusion,anestimationismadeofthepredationimpactofthemesopelagic?shcommunityonthekrillstockintheregionofKingGeorgeIsland.
2.Materialandmethods2.1.Sampling
DatawerecollectedduringthecruiseANTXIV/2ofRV‘‘Polarstern’’inNovember/Decem-ber1996.ThestudyareawaslocatedovertheslopenorthwestofKingGeorgeIsland(SouthShetlandIslands),insouthernDrakePassage(Fig.1).
Thesamplingprogramconsistedof16hauls,withtheobjectivetosamplethreedifferentdepthhorizons:200–300m,400–500mandnear-bottom(2–28mabovesea?oor)atsoundingsof400,600and800m(Table1)(Kocketal.,1998).Mesope-lagic?shweresampledwithapelagictrawlPT-1088withanestimatedmouthopeningof200m2(width20mandaheightof10–12m).Themeshsizewas12mminthecodend.Itwasexpectedthatjuvenilemyctophids(SLo30mm)wouldnotbesampledadequatelywiththisnetcon?guration(Gartneretal.,1988).AnSCANMARdepthsensorcontrolledthesampledepthandnetopen-ingduringtrawling.Towingtimevariedbetween30and60min;trawlspeedsrangedfrom3.5to4.0knots.Shipspeedwasincreasedduringnetdeploymentanddecreasedduringretrieval.Thisprocedureminimizestheeffectsofnetcontamina-tionby?shresidentinwaterlayersabovethe?shingdepth.Station73wasexcludedfromcommunityanalysisasthenetsnaggedonthebottomduringtrawling(Hulleyetal.,1998).All?sheswereidenti?edtospeciesaccordingtothemostrecentkeys(GonandHeemstra,1990).Fishfromtheentiresampleorasub-sampleofeachspeciesfromeachstationwerecountedandweighed,andstandardlengths(SL)weretaken
?S61.5517057556564605359635452695868731500 m1000 m500 m200 m62.0orgeIs.KingGe60.059.559.058.558.057.5°WFig.1.Samplinglocalitiesin1996.Lineindicateshydrographicsectionthroughthestudyarea.
ARTICLEINPRESS1688
C.Puschetal./Deep-SeaResearchI51(2004)1685–1708
Table1
Stationdatafor16PT-1088trawlsamplesST51525354555758596063646568697073
aSamplingLocalityLongitudeLatitude581430W591170W591190W591310W591340W591160W591130W591320W591440W591320W591350W591510W591090W591170W591020W581440W
611360S611450S611450S611480S611470S611430S611470S611480S611480S611480S611470S611490S611440S611460S611380S611360S
Date199630November1December1December1December1December3December3December3December4December4December5December5December5December6December6December6December
Time(Local)16:49–17:4908:00–08:3010:57–11:2714:50–15:2017:35–18:0508:19–08:5020:26–21:2623:20–00:2002:15–03:1820:30–21:1101:06–01:4104:26–04:5621:10–21:4101:01–01:3204:50–05:2021:07–21:10
DayTimeaDayDayDayDayDayDayDayNightNightDayNightDayDayNightDayDay
Global
Radiation(wm2)197.875.0153.0230.3145.8179.712.40.00.098.50.080.646.00.023.020.0
SampleDepth(m)400–450415–515273–302283–295400–450397–465283–325431–495380–440520–580610–640750–800340–360560–597790–825550–575
BottomDepth(m)786–875660–730635–738475–652680–995730–790380–387850–1058681–690555–608731–7921287–1468360–367584–599810–833570–580
Ship
Speed(km)4.04.04.04.04.04.03.53.73.54.04.04.03.53.73.73.5
De?nedbysunset22:06andsunrise03:23.
tothenearestmillimetrewithslidingcallipers.Hydrographicdatawerecollectedbyconductivity,temperatureanddepthcasts(CTD,22stations).TheCTDwasdeployedateachstationinadvanceofthetrawltows.2.2.Dietanalysis
Dietanalysiswasperformedonthe?vemostabundantmesopelagic?shspecies.Amaximumof20individualsofthesespecieswasselectedfromeachsample.Insamplescontaining420speci-mensindividualswerechosenhaphazardly.Fishwereweighedwet,measured(SL,mmbelow)andthewholestomachremoved.Thedryweightof?shspecimenswasdeterminedbyoven-dryingspeci-mensat801Cuntilconstantweightwasreached.Preyorganismswereidenti?edtothelowestpossibletaxonandmeasuredunderabinocularmicroscope.Dryweightofthepreyorganismswasreconstructedbylength–weightregressionstakenfromtheliterature(Mizdalski,1988;Groegeretal.,2000).
Threeindices,thefrequencyofoccurrence(Fi)ofeachpreyiteminnon-emptystomachs,thepercentageofeachfooditembynumber(Ni)tothetotalnumber,andthepercentagebydryweight
ley,1979).(DWi)werecalculated(GeorgeandHad
Bythefollowingequation,allthreeindiceswerecombinedtodescribethepreyutilizationbythe‘RelativeImportanceIndex’(RI)foreachpreycategoryi(GeorgeandHadley,1979;Hyslop,1980):
eFitNitDWiT?100RIi?Ps;
eFitNitDWiTi?1
e1T
wheresisthenumberofpreycategories.
FeedingchronologywasanalysedbytheSto-machContentIndex(SCI):SCIe%T?
dryweightofstomachcontent
?100:
bodydryweight
e2T
Inaddition,thestageofdigestionofeachpreyitemwasdeterminedbythemodi?edmethodofPearcyetal.(1979):Stage1=undigestedprey,Stage2=slightlydigestedwithsomeappendagesdamaged,butbodyshapestillpreserved,Stage3:bodyshapeofpreydeformed.Theratioofdigestionstageswascalculatedforeachtimeintervalbasedonthesecriteria.
Dailyration(meandailyfoodconsumption,Cw)ofthefourmyctophidswasinvestigatedbythe
ARTICLEINPRESSC.Puschetal./Deep-SeaResearchI51(2004)1685–1708
1689
methodofEggers(1977):Cw?I?24=T;
e3T
whereIisthedailyaverageSCI(%)andTthegutpassagetime(h).
Inthisstudy,nogutpassagetimedatawerecollected.Therefore,datafromtheliteraturewereused.TwoestimatesofegestiontimesforAntarc-ticmyctophidsareavailable:the?rstestimatebyRowedder(1979a)forElectronaantarcticaandthesecondbyGerasimova(1990)forE.carlsbergi.Thesestudiesestimatedtheegestiontimetobe8and8.5h,respectively.Inourstudy,weusedthe8.5hsuggestedbyGerasimova.Forthecalcula-tionofdailyrationsomeauthors(Pakhomovetal.,1996)haverecommendedthesubstitutionof24?10inEq.(3)forspeciesthathaveanactivefeedingperiodof10h.Otherauthorsarguethatthissubstitutionintroducesasigni?cantconserva-tivebias(Williamsetal.,2001).Wecalculatedtwoalternativedailyrationsusingboth10and24.2.3.Dataanalysis
Densityandbiomassdataformesopelagic?shwerecalculatedasindividualsper?lteredwatervolume.The?lteredvolumewascalculatedbymultiplyingthetrawleddistanceofthevesselwiththeestimatedmouthopening(200m2)ofthePT-1088.
CommunitystructurewasinvestigatedwiththePrimer-E5Softwarepackage(ClarkeandWar-wick,2001).Toreducetheweightingofdominantspecies,thedensitiesweresquare-roottransformedpriortothecomputationofthetriangularsimilaritymatricesbasedonBray-Curtissimila-rities(Fieldetal.,1982).Theresultsofthelatterwereclassi?edbyhierarchicalagglomerativeclus-teranalysisusingthegroupaveragelinkingmethod,andordinatedbyanon-metric,multi-dimensionalscalingtechnique(MDS).
TheBIOENVsub-routinewasusedtorelatethecommunitypatternsofthemesopelagic?shassemblagetosixenvironmentalvariables:lightintensity(W/m2)(indicatingtimeofday),mini-mumandmaximumvaluesofsampledepth,bottomdepth,temperatureandsalinityatthesampleddepthhorizon.Valuesofaveragelightintensitywerelogtransformedtovalidatetheuse
ofnormalizedEuclideandistanceforthecalcula-tionoftheabioticsimilaritymatrix.WeappliedtheSpearmanRankcorrelationtorelatethebioticmatrixbasedonmesopelagic?shabundanceswiththeabioticsimilaritymatrix(ClarkeandAins-worth,1993).
TwodifferentwatermasseswerediscernibleovertheslopeofKingGeorgeIsland,separatedatadepthof450m(seeSection3).Mesopelagic?shsamplesaboveandbelowthisdepthwerecom-paredwithdifferentsub-routinesofthePrimercomputerprogram.
One-wayanalysisofsimilarity,ANOSIM(ClarkeandWarwick,1994),wasemployedtotestthehypothesisofnodifferencesinmesopelagic?shassemblageaboveandbelow450m.Thissubrou-tinecomparestheaverageranksimilaritieswithintheprede?nedgroupsofsampleswiththeaveragesimilaritybetweengroups.Valuescloseto1indicateastrongseparationbetweengroups,whileavalueof0indicatesnodifferencesbetweengroups.
Thesimilaritypercentageroutine(SIMPER)(ClarkeandWarwick,1994),whichwasappliedtosquare-roottransformedmesopelagic?shabun-dances,identi?edthecontributionfromindividualspeciestothedissimilaritiesbetween(thedeepandtheshallow)samplegroups.
Variousunivariateindiceswerecalculatedinordertocharacterizethespeciesassemblagesofthedeepandshallowgroupofsamples:speciesnumber,Shannon’sdiversityindex(H0)(ShannonandWeaver,1949)andPielou’sEvennessIndex(J)(Pielou,1975).Astheseindicesareknowntobein?uencedbysamplesize,wealsocalculatedthetaxonomicdiversityDandtaxonomicdistinctnessD?
;whichconsiderthetaxonomicrelatednessofspecies(WarwickandClarke,1995).TaxonomicdiversityDisempiricallyrelatedtoH0butcontains,inaddition,informationonthetaxo-nomicseparationofthespeciesinasample,i.e.besidesthedistributionofindividualsamongspeciesitalsotakesintoaccountthedistributionofspeciesinthetaxonomicsystembyweightingtheco-occurrencesofspeciesaccordingtothedegreeofseparationinthehierarchicalclassi?ca-tion(1=differentspecies,2=differentgenera,3=differentfamilies,4=differentorders).D?is
陆坡水域鱼类群落结构与摄食生态
