Deep Sea Research II

Marine Research

Description

Niveau: Supérieur, Licence, Bac+1
Deep-Sea Research II 49 (2002) 3327–3349 Resource limitation of phytoplankton growth in the Crozet Basin, Subantarctic Southern Ocean P.N. Sedwicka,*, S. Blainb, B. Qu!eguinerc, F.B. Griffithsa,d, M. Fialae, E. Bucciarellib, M. Denisc aAntarctic CRC, GPO Box 252-80, Hobart, Tasmania 7001, Australia b Institut Universitaire Europ!een de la Mer, UMR-CNRS 6539, Place Nicolas Copernic, 29280 Plouzan!e, France cCentre d'Oceanologie de Marseille, Laboratoire d'Oc!eanographie et de Biog!eochimie, Campus de Luminy, Case 901, 13288 Marseille Cedex 09, France dCSIRO Division of Marine Research, Castray Esplanade, Hobart, Tasmania 7000, Australia eLaboratoire Arago, Universit!e Pierre et Marie Curie, UMR-CNRS 7621, 66651 Banyuls-sur-mer Cedex, France Accepted 20 July 2001 Abstract In January–February 1999, we performed shipboard iron- and macronutrient-addition experiments in the Crozet Basin, Indian sector of the Subantarctic Southern Ocean, to evaluate the sufficiency of ambient iron and macronutrient concentrations for algal growth. Experiments were conducted with near-surface seawater collected from three locations in a narrow latitudinal band characterized by relatively low algal biomass (o0.7 mg l1 chlorophyll a), low dissolved iron concentrations (o0.

• late summer


• seawater inside


• using seawater


• iron


• zone


• ocean


• subantarctic southern


• macronutrient-addition bottle


• crozet basin during

Sujets

• Huntsman
• Dufresne
• Hobart
• Research
• Southern
• Boys Life 5

Informations

• Publié par: profil-ontoe-2012
• Nombre de lectures: 11
• Langue: English

Extrait

Deep-SeaResearchII49(2002)3327–3349 ResourcelimitationofphytoplanktongrowthintheCrozet Basin,SubantarcticSouthernOcean a,bca,de P.N.Sedwick*,S.Blain,B.Que ! guiner,F.B.Griffiths,M.Fiala, cb E.Bucciarelli,M.Denis a AntarcticCRC,GPOBox252-80,Hobart,Tasmania7001,Australia b InstitutUniversitaireEurope ! endelaMer,UMR-CNRS6539,PlaceNicolasCopernic,29280Plouzane ! ,France c Centred’OceanologiedeMarseille,Laboratoired’Oce ! anographieetdeBioge ! ochimie,CampusdeLuminy,Case901, 13288MarseilleCedex09,France d CSIRODivisionofMarineResearch,CastrayEsplanade,Hobart,Tasmania7000,Australia e LaboratoireArago,Universite ! PierreetMarieCurie,UMR-CNRS7621,66651Banyuls-sur-merCedex,France Accepted20July2001 Abstract InJanuary–February1999,weperformedshipboardiron-andmacronutrient-additionexperimentsintheCrozet Basin,IndiansectoroftheSubantarcticSouthernOcean,toevaluatethesufficiencyofambientironandmacronutrient concentrationsforalgalgrowth.Experimentswereconductedwithnear-surfaceseawatercollectedfromthreelocations inanarrowlatitudinalbandcharacterizedbyrelativelylowalgalbiomass( o 0.7mgl
1 chlorophylla),lowdissolved ironconcentrations( o 0.33nM),andstrongmeridionalgradientsintemperature,salinityandmacronutrient concentrations:(1)thePolarFrontalZone(PFZ)near46 1 S,65 1 E( B 19mMnitrateand1.2mMsilicicacid);(2)the confluenceoftheSubantarcticandSubtropicalFronts(SAF/STF)near44 1 12 0 S,63 1 23 0 E( B 5.4mMnitrateand0.5mM silicicacid);and(3)thesouthernSubtropicalZone(STZ)near43 1 18 0 S,62 1 31 0 E( o 0.1mMnitrateand B 1.4mMsilicic acid).Ourexperimentalresultsrevealthreedistinctregimesofresourcelimitationofphytoplanktongrowth.Inthe PFZ,ironavailabilityexertedtheprimarylimitationonnitratedrawdownandbiomassaccumulation,thuscommunity growth,withsilicicacidavailabilityexertingasecondarylimitationondiatomgrowthandbiogenicsilicaproduction. WithintheSAF/STF,irondeficiencywasalsotheprimarylimitationonalgalcommunitygrowth;however,herewe observedevidenceofsecondarylimitationofnitratedrawdownandbiomassaccumulationbysilicicaciddeficiency,via controlofalgalcommunitystructure—suchthatironadditionpreferentiallystimulatedthegrowthofnon-diatom nanoplankton—suggestingthatthealgalcommunitywaspoisedclosetoco-limitationbyironandsilicicacid.As expected,ourexperimentalresultsindicatethatmacronutrients(nitrate/phosphate)weretheprimarylimitationon communitygrowthintheSTZwaters;however,ourresultsalsosuggestthatirondeficiencyimposedasignificant secondarylimitationoncommunitygrowth,particularlydiatomgrowth,suchthatthealgalcommunitywaspoised closetoco-limitationbymacronutrientsandiron.Weconcludethatthesesameregimesofresourcelimitationarelikely toregulatephytoplanktongrowthandexportproductionovermuchoftheopen-oceanSubantarcticregionduringthe midtolatesummer.r2002ElsevierScienceLtd.Allrightsreserved. *Correspondingauthor.Presentaddress:BermudaBiologicalStationforResearch,St.George’sGE01,Bermuda. E-mailaddress:[e-mail](P.N.Sedwick). 0967-0645/02/$-seefrontmatterr2002ElsevierScienceLtd.Allrightsreserved. PII:S0967-0645(02)00086-3

2338P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 1.Introduction TheSubantarcticregionofSouthernOcean, locatedbetweenthePolarandSubtropicalwater masses,accountsforasignificantfractionof marineprimaryproduction(Fieldetal.,1998), andisoneofthelargestoceanicsinksfor atmosphericCO 2 ,dueinparttobiologicaluptake andexportbyphytoplankton(Metzletal . ,1999). Fromsouthtonorth,thesurfacewaterswithinthis vastcircumpolarbeltincludetheAntarcticPolar Front,thePolarFrontalZone,theSubantarctic Front,theSubantarcticZone,andtheSubtropical Convergenceand/orSubtropicalFront(Nowlin andKlinck,1986;ParkandGambe ! roni,1997; Griffithsetal.,1999;RintoulandTrull,2001). Muchofthisregionischaracterizedbyhigh- nutrientlow-chlorophyll(HNLC)conditions— where‘‘nutrient’’referstonitrateandphos- phate—althoughlownitrateconcentrations ( o 1mM)areoftenobservedinsurfacewatersof theSubtropicalConvergence/SubtropicalFront (Clementsonetal.,1998;Boydetal.,1999; Griffithsetal.,1999).Silicicacid,whichisrequired forthegrowthofdiatoms,iscommonlydepleted fromlevelsof B 5mMormoreinspringtovery lowconcentrations( o 1mM)bymidtolate summer(Dugdaleetal.,1995;Nelsonetal., 2001;RintoulandTrull,2001),anddissolvediron concentrationsarelikelytobelow( o 0.5nM)in surfacewatersovermuchoftheregion(Sedwick etal.,1997,1999;Coaleetal.,1999;Measuresand Vink,2001).Thus,themajorcontrolsonalgal growthinopen-oceanwatersofthisregionmay includeironandsilicicacidavailability,inaddi- tiontolightavailabilityascontrolledbyvertical mixing,and,inthenorth,nitrateand/orphos- phateavailability(Martinetal.,1990a;Nelson andSmith,1991;Cullen,1991;Dugdaleetal., 1995;Banse;1996,Sedwicketal.,1997;Sundaand Huntsman,1997;Boydetal.,1999;Griffithsetal., 1999;Hutchinsetal.,2001).Atpresent,however, welackaclearunderstandingoftherelative importanceofthesefactorsinregulatingalgal growth,communitystructureandexportproduc- tionintheSubantarcticSouthernOcean. Numerousstudiesduringthepastdecadehave combinedtheresultsofwater-columnmeasure- ments,shipboardincubationexperimentsand, morerecently,mesoscaleperturbationexperi- ments,toinferthedominantcontrolsonalgal growthintheSouthernOcean(e.g.,Martinetal., 1990a;deBaaretal.,1990,1995;Lo . scheretal., 1997;vanLeeuweetal.,1997;Sedwickand DiTullio,1997;Takeda,1998;Boydetal.,2000; Smetacek,2001).However,onlyafewsuchstudies havebeenperformedintheSubantarcticregion, northoftheAntarcticPolarFront.Fieldstudies conductedinthePolarFrontalZonenear170 1 W, southeastofNewZealand,indicatethatphyto- planktoncommunitygrowthandbiomassare limitedbylowironconcentrationsovermuchof thegrowingseason(Coaleetal.,1999;Measures andVink,2001),andthatlowsilicicacid concentrations( o 5mM)maylimitthegrowthof diatoms,atleastduringthesummerandautumn (Francketal.,2000;Nelsonetal.,2001).In addition,experimentsperformedinthePolar FrontalZonearound140 1 E,southwestofTasma- nia,suggestthatlightavailabilitymaylimitalgal communitygrowthduringlatesummerand autumn(Boydetal.,2001).Furthernorth,in watersoftheSubantarcticZonesoutheastofNew Zealand,fielddatasuggestthatdiatomgrowthis limitedbyavailabilityofbothironandlight duringspring,andbyironand/orsilicicacid availabilityduringsummer,whereaswatersinand northoftheSubtropicalConvergenceareiron replete(Boydetal.,1999).Similarly,theresultsof fieldexperimentsconductedtothewestof TasmaniaindicatethatalgalgrowthinSubantarc- ticZonewatersislimitedbyironandsilicicacid deficiencyduringlatesummerandautumn, althoughtheseexperimentsalsoprovidedevidence ofironlimitationwithintheSubtropicalConver- gence(Sedwicketal.,1999;Hutchinsetal.,2001). Ageneralhypothesisthatisconsistentwiththe resultsofthesestudiesisthatphytoplankton communitygrowthinopen-oceanwatersofthe Subantarcticregionisprimarilyregulatedbythree resources—iron,silicicacid,andlight—theavail- abilityofwhichvaryseasonally.Growthlimita- tionduetoinsufficientlightislikelytobemost importantbetweenlateautumnandearlyspring, whenthesurfacemixed-layerdepthtypically exceeds100–200m(RintoulandTrull,2001).

P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 Althoughsurface-waterdissolvedironconcentra- tionsmaybehighestoverthisperiod(Measures andVink,2001),thelowirradiancealsomayserve toprecipitateorintensifygrowthlimitationby irondeficiency,giventheantagonisticrelationship betweentheironandlightrequirementsof phytoplankton(Raven,1990;SundaandHunts- man,1997),sothatalgalgrowthmaybephysio- logicallyco-limitedbyironandlight(Boydetal., 1999;Maldonaldoetal.,1999).Asthegrowing seasonprogresses,averagemixed-layerdepth decreases,andalgalbiomassincreases,thecon- centrationsofbothironandsilicicacidare depletedtoverylowvalues,whicharelikelyto limitthegrowthrateoflargerspecies,particularly diatoms.Atthistime,betweenlatespringand earlyautumn,eitherironorsilicicacidmayexert theproximatecontrolonalgalcommunitygrowth, andthephytoplanktoncommunitymaybepoised closetoco-limitationbytheseresources;infact, theavailabilityofbothresourcesmightsimulta- 2339 neouslylimitthegrowthofdifferentmembersof thealgalcommunity(Boydetal.,1999;Hutchins etal.,2001).However,itisimportanttoappreciate thattheabovehypothesisisbasedondata collectedduringarelativelysmallnumberoffield studiescarriedoutinonlytwonarrowsectorsof theSouthernOcean. Insummer1999,wewereprovidedwiththe opportunitytoexaminetheroleofresource availabilityinregulatingphytoplanktongrowth intheIndiansectoroftheSubantarcticSouthern Ocean,duringaresearchcruiseintheCrozet Basin,northwestofKerguelenIsland(Fig.1).This oceanographicsettingisquitedifferentfromthe SubantarcticregionnearNewZealandandTas- mania,inthattheSubantarcticandSubtropical Frontsconverge,constrainedbytheKerguelen PlateauinthesouthandtheAgulhasReturn CurrentFront(AgulhasFront)inthenorth(Park etal.,1993;ParkandGambe ! roni,1997).The transitionfromcold,fresh,nutrient-richpolar Fig.1.Antares-IVcruisetrackshowinglocationofCrozetBasin(basemapcourtesyofB.Delille).Inset:Detailofcruisetrackin CrozetBasin,showingapproximatepositionsofmajoroceanographicfeatures(Y.-H.Park,pers.comm.,1999)andStations1 0 (PFZ), 7(SAF/STF)and8(STZ).

0333P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 waterstowarm,salty,nutrient-poorsubtropical watersoccursoveranarrowlatitudinalrange (Fig.1),inaregionthatappearstobecharacter- izedbyrelativelylowdissolvedironconcentrations ( o 0.4nM)andlowalgalbiomass( o 0.7mgl
1 chlorophylla)duringthesummer.Herewereport theresultsfromaseriesofshipboardbioassay experiments,whichwereundertakentoinvestigate thesufficiencyofambientiron,silicicacid,and macronutrientconcentrationsforphytoplankton growthinthreedistinctlow-iron,low-chlorophyll regimes;namely,(1)thePolarFrontalZone(PFZ), withhighconcentrationsofnitrate( B 19mM)and lowconcentrationsofsilicicacid( B 1.2mM);(2) theconfluenceoftheSubantarcticandSubtropical Fronts(SAFandSTF),withmoderateconcentra- tionsofnitrate( B 5.4mM)andverylowconcen- trationsofsilicicacid( B 0.5mM);and(3)the southernedgeoftheSubtropicalZone(STZ),with verylowconcentrationsofnitrate( o 0.1 m M)and lowconcentrationsofsilicicacid( B 1.4mM).Our experimentalresultsdemonstratethreedistinct regimesofresourcelimitationwhichmayregulate algalcommunityproductionandspeciescomposi- tioninwatersoftheSubantarcticregionduring themidtolatesummer. 2.Methods The Antares -IVcampaignwasconducted aboardNOMarionDufresneintheCrozetBasin duringJanuaryandFebruary1999,aspartofthe France-JGOFSprogram.Amajoraimofthis expeditionwasanunderstandingofthefactors controllingphytoplanktonproductionandorganic matterexportinthissectoroftheSubantarctic SouthernOcean,aregionthatistypicallychar- acterizedbyrelativelylowalgalbiomass,except forareasimmediately‘‘downstream’’fromthe islandplatformsofCrozetandKerguelen(Moore etal.,1999).Ourstudyfocussedonanarea northwest(‘‘upstream’’)ofKerguelenIsland, boundedbylatitudes43–46 1 Sandlongitudes 61–65 1 E,wheretheconfluenceoftheSubantarctic andSubtropicalFrontswasalignedinasouthwest tonortheastdirection(Fig.1).Theapproximate positionsoftheSAFandSTF(Fig.1)aredefined byaxialtemperaturesandsalinitiesof6 1 Cand 34.3(SAF),and10 1 Cand34.8(STF),atawater depthof200m(Parketal.,1993;Park,pers. comm.,1999).Ourcruiseincludedhydrographic andbiogeochemicalmeasurementsatstations withinthePFZ,theSAF/STFconfluence,and thesouthernedgeoftheSTZ.Wemeasured dissolvedironinthewatercolumnandperformed iron-andmacronutrient-additionbottle-incuba- tionexperimentsusingseawaterandresident planktoncollectedfromtheuppermixedlayerat Stations1 0 ,7and8(Fig.1,Table1),whichare hereafterreferredtoasthePFZstation,theSAF/ STFstation,andtheSTZstation,respectively. Containersandapparatususedforsampling andanalysisofironinseawaterandforiron-and macronutrient-additionincubationexperiments wererigorouslycleanedusingthetrace-metal cleantechniquesdescribedbySedwicketal. (1997,2000).Water-columnsamplesforiron measurementswerecollectedincustom-built polycarbonatewatersamplerssuspendedfroma non-metallicline,asdescribedbySedwicketal. (1997).Uponrecovery,thewatersamplerswere immediatelytransferredintoadedicatedclean laboratoryvan,wherefilteredandunfiltered subsamplesweretakenunderClass-100filtered air(Sedwicketal.,1997).Subsampleswerefiltered throughacid-cleanedPoreticspolycarbonate membranes(0.4mm),usinganin-linesystem pressurizedwith0.2mm-filteredhigh-puritynitro- gengas.Thefilteredandunfilteredsubsamples werecollectedinacid-cleanedNalgenelow-density polyethylenebottlesandacidifiedwithSeastar ultrapureconcentratedhydrochloricacidasde- scribedbySedwicketal.(2000).DissolvedFe (measuredinthefiltered,acidifiedsubsamples)and total-dissolvableFe(TDFe,measuredinthe unfiltered,acidifiedsubsamples,andassumedto provideameasureofdissolved+particulateFe) weredeterminedattheAntarcticCRCinHobart byflow-injectionanalysiswithin-linepreconcen- trationonresin-immobilized8-hydroxyquinoline (Measuresetal.,1995).Thismethodhasa detectionlimitof B 0.02nMFeandoverall analyticaluncertaintiesof B 20%fordissolved Feand B 30%forTDFe,overtheconcentration rangesreportedinthisstudy(Sedwicketal.,2000).

P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 SAF/STF(7)44 1 12 0 S,63 1 23 0 E 6Feb.1999 1333 Table1 Summaryofambientconditionsandexperimentaltreatments StationPosition,startdateAmbientconditionsExperimentaltreatments PFZ(1 0 )46 1 S,65 1 E18.7mMnitrateControl(untreated) 14Jan.19991.18mMsilicicacid+Si(18.3mMSi) a 0.33nMdiss.iron+Fe(2nMFe) 0.21mgl
1 Chl a +Fe+Si(2nMFe,18.3mMSi) a 5.4mMnitrateControl(untreated) 0.5mMsilicicacid+Si(3.9mMSi) p 0.19nMdiss.iron b +Fe(2nMFe) 0.62mgl
1 Chla+Fe+Si(2nMFe,3.9mMSi) STZ(8)43 1 18 0 S,62 1 31 0 E o 0.1mMnitrateControl(untreated) 12Feb.19991.4mMsilicicacid+N+P(10mMN,0.63mMP) 0.09nMdiss.Fe+Fe(2nMFe) 0.36mgl
1 Chla+Fe+N+P(2nMFe,10mMN,0.63mMP) a Approximately0.8nMFeaddedtogetherwithcontaminatedSisolutioninStation1 0 experiment. b DissolvedFeconcentration(0.29nM)washigherthanTDFeconcentration(0.19nM)inthestartingseawateratthisstation. However,ironconcentrationsof0.19and0.29nMareanalyticallyindistinguishable,andthedissolvedFesampleisnotconsideredto havebeencontaminated. Unfilteredseawaterfortheshipboardincuba- tionexperimentswascollectedfromawaterdepth of B 20mthroughacid-cleanedpolyethylenetub- ingusinganOsmonicssolidTeflondiaphragm pump(Hutchinsetal.,1998;Sedwicketal.,1999). Thetubinginletwasloweredfromthewindward sideoftheshipusinganon-metallicline,and seawaterwaspumpedonboardwhileheadinginto thewindat B 1–2knots.Thepumpedseawater wasdischargedinsidethecleanlaboratoryvan, wheresamplecontainersusedintheexperiments wererinsedandfilled.Theseawaterwasnot screenedtoexcludelargerorganisms,althoughit shouldbenotedthatlargerheterotrophsarelikely tobeundersampledinthismethodofseawater collection.Thebottle-incubationexperimentswere similarindesigntothosedescribedbySedwick etal.(1999)andHutchinsetal.(2001).The pumpedseawaterwasgentlymixedinacid-cleaned 50-lNalgenepolyethylenecarboys,thenimmedi- atelytransferredintoacid-cleaned2.4-lNalgene polycarbonatebottles,usedfortheexperiments discussedhere,andacid-cleaned24-lNalgene polycarbonatecarboys,usedforconcurrent ‘‘dose-response’’iron-additionexperimentsthat examinedtheresponseofthealgalcommunityto variouslevelsofaddediron(seeBlainetal.,2002). Subsamplesfortheinitial(t¼0)analyseswere takendirectlyfromthe50-lmixingcarboys,except forthesubsamplesforironmeasurements,which weretakenfromapolycarbonatetrace-metal water-samplerthathadbeendirectlyfilledwith thepumpedseawater,andprocessedasdescribed forthewater-columnironsamples.The2.4-l experimentalcontainerswereimmediately amendedwithironand/ormacronutrientsas describedbelow,thencapped,sealedwithPVC tape,andsetincirculatingsurfaceseawaterinside polyethyleneincubatorslocatedondeck.These incubatorsshadedthebottlesto E 50%ofincident irradiance,whichwasintendedtosimulatethein situirradianceintheuppermixedlayer.Tempera- turesinthedeckincubatorsweremaintainedto within 7 4 1 Cofambientsea-surfacetemperature duringthecourseoftheexperiments. InthePFZstation(Station1 0 )experiment,the incubationbottleswereeither(1)enrichedwith silicicacid(+Si,18.3mM),(2)enrichedwithiron (+Fe,2nM),(3)enrichedwithironandsilicicacid (+Fe+Si,2nMand18.3mM,respectively),or(4) leftuntreatedascontrols.Ironwasaddedasa solutionofferricchloridein0.01Nhydrochloric acid,andsilicicacidwasaddedasasodium metasilicatesolutionwhichhadbeenpurifiedusing

2333P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 acolumnofresin-immobilized8-hydroxyquino- line.However,thispurificationschemeprovedto beinadequate,possiblyduetocomplexationof ironbysilicicacidinbasicsolution.Shipboard analysisofthepurifiedsodiummetasilicatesolu- tionrevealedthatitcontainedenoughironsuch that B 0.8nMironwasinadvertentlyaddedtothe incubationbottlestogetherwiththeaddedsilicic acid,thusinvalidatingthe+Sitreatmentsinthis experiment.Anew10mMsilicicacidstock solutionwaspreparedandpurifiedatseaforuse intheSAF/STFstation(Station7)experiment: reagentgradesodiummetasilicatewasdissolvedin deionizedwater,thenthissolutionwasfiltered througha0.2-mmpolycarbonatemembrane,di- lutedten-foldwithdeionizedwater,acidifiedtopH B 3.5withSeastarultrapurehydrochloricacid, thenpassedovercolumnsofresin-immobilized 8-hydroxyquinolineandWatersSep-PakC-18 packingmedia(thelattertoremove8-hydroxyqui- nolinewhichmayhavebledfromtheresin- immobilized8-hydroxyquinolinecolumnduring passageoftheacidicsolution;W.Landing,pers. comm.,1998).Thethus-purifiedsilicicacidsolution wasanalyzedatseaandfoundtocontain0.94mM Siand B 1nMFe,whichwasadequateforusein theSAF/STFstationexperiment(seebelow). IntheSAF/STFstation(Station7)experiment, theincubationbottleswereeither(1)enrichedwith silicicacid(+Si,3.9mM),(2)enrichedwithiron (+Fe,2nM),(3)enrichedwithironandsilicicacid (+Fe+Si,2nMand3.9mM,respectively),or(4) leftuntreatedascontrols.Inthiscase,the contaminantironaddedalongwiththesilicicacid solutionamountedto o 0.01nM,whichisas- sumednegligiblecomparedwiththeambientiron concentrationof B 0.2–0.3nMinseawatercol- lectedfortheexperiment.IntheSTZstation (Station8)experiment,theincubationbottleswere either(1)enrichedwithnitrateandphosphate (+N+P,10mMand0.63mM,respectively),(2) enrichedwithiron(+Fe,2nM),(3)enrichedwith iron,nitrateandphosphate(+Fe+N+P,2nM, 10mMand0.63mM,respectively),or(4)left untreatedascontrols.Nitrateandphosphatewere addedasaqueoussolutionsof20mMsodium nitrateand20mMsodiumdihydrogenphosphate, respectively,whichhadbeentwicepurifiedusing columnsofresin-immobilized8-hydroxyquinoline. Shipboardanalysisofthesesolutionsshowedthem tocontain0.65nMand6.0nMiron,respectively, suchthattheironaddedalongwiththenitrateand phosphatesolutionsamountedto o 0.01nM, whichisassumednegligiblerelativetotheambient ironconcentrationof B 0.1–0.2nMinseawater collectedfortheexperiment.Table1summarizes theambientconditionsandexperimentaltreat- mentsforthethreestationsinvestigated. Foreachexperimentaltreatment,duplicate2.4-l bottlesweresacrificed(i.e.,completelysub- sampled)atselectedtimepointsduringthecourse ofeachexperiment,withsubsamplestakenforthe followingmeasurements:dissolvednitrate+nitrite (N+N)andsilicicacid(Si),determinedin 0.45mm-filteredsolutionsbystandardmethods usingaTechniconsegmented-flowanalysissystem; chlorophylla(Chla),determinedin>0.2mm particulatematerialbyspectrofluorometry(Fiala etal.,1998);particulateorganiccarbon(POC)and particulateorganicnitrogen(PON),determinedin GF/F-filteredmaterialbyelementalanalyzer(Qu- e ! guineretal.,1997);particulatebiogenicsilica (BSi),determinedin>0.4mmparticulatematerial byspectrophotometryaftersodiumhydroxide digestion(Que ! guineretal.,1997);rateofprimary production,measuredbythe24-hincubation techniquedescribedbyQue ! guineretal.(1997); andrateofphotosynthesisvs.irradiance(Pvs.I), measuredbythesmall-bottleshort-termincuba- tiontechniquedescribedbyGriffithsetal.(1999). Theresultspresentedforthe24-hprimary productionmeasurements,P net ;provideanesti- mateofnetprimaryproductionoverthe24-h periodfollowingthetimeofsubsampling,inunits ofmgCm
3 d
1 .TheresultsshownfromPvs.I measurements,P max ;provideanestimateofgross light-saturatedrateofphotosynthesisclosetothe timeofsubsampling,inunitsofmgCm
3 h
1 , andarethusnotdirectlycomparablewiththe estimateddailynetproductionrates.Itshould benotedthatwehavechosennottonormalizethe valuesofP max tobiomassusingChla,because ironadditionisexpectedtosignificantlyalter thecellularChlacontentofoceanicphytoplank- toninsuchexperiments(GeiderandLaRoche, 1994).SamplesformeasurementsofdissolvedFe

P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 andtotal-dissolvableFeweretakenonlyfromthe startingseawater(t ¼ 0),duetothelimitedvolume oftheincubationbottles. Visualidentificationandenumerationofcells largerthan B 2mmwereperformedat400times magnificationusinganinvertedlightmicroscope withphasecontrast,afterovernightsettlingof B 20-mlsubsamplespreservedatseawithLugol’s iodinesolution.Wherepossible,morethan1000 cellswerecountedineachsample.Ateachofthe threestationswhereexperimentswereconducted,a significantcomponentofthenanoplanktoninthe startingseawaterandexperimentaltreatments wereroughlysphericalcells B 3–10mmindiameter, bothwithandwithoutflagella,whichwerenot definitivelyidentified,andareheretermed‘‘mon- ads’’assuggestedbyHasle(1978).Thesecellsare distinguishedfromthegenerallylarger,non- sphericalnanoflagellates,whichwehavetermed ‘‘flagellates’’.However,itshouldbenotedthat someofthemoredelicatephytoplanktonspecies, includingcoccolithophorids,whichmaycomprise animportantcomponentofthealgalcommunity northofthePolarFront(Findlay,1998),are poorlypreservedinLugol’siodinesolution (Throndsen,1978).Thus,thealgalcellsthatwe haveidentifiedbymicroscopyprovideuswithan incompletepictureofphytoplanktoncommunity composition,inthatpicoplanktonanddelicateor calcareousnanoplanktonwouldnothavebeen counted,sothattheseresultsmustbeconsideredas semi-quantitativeatbest.Italsoshouldbenoted thatwithourvisualcell-countingtechnique,wedid notattempttodistinguishbetweenautotrophsand heterotrophsinthedinoflagellate,flagellate,and monadgroupings.Foreachexperiment,selected samplespreservedwithLugol’siodinesolution werealsoexaminedusingshadow-casttransmis- sionelectronmicroscopy(TEM)attheAustralian AntarcticDivision,whichallowedagenus-level identificationofsomeofthemorenumerous nanoplanktoninthesamples. 3.Results Asinotherfieldstudiesthathaveusedship- boardbioassayexperiments(e.g.,Martinetal., 3333 1990a;deBaaretal.,1990;DiTullioetal.,1993; Fitzwateretal.,1996;Hutchinsetal.,2001),we inferrelativealgalcommunitygrowthratesinthe incubationbottlesfrom(1)accumulationofalgal biomass,asindicatedbydecreases(drawdown)in dissolvedN+NandSi,andincreasesinChla, POCandPON;and(2)relativeratesofprimary production,asindicatedbythe 14 C-basedesti- matesofdailynetprimaryproductionrate( P net ) andshort-termgrosslight-saturatedphotosynth- esisrate(P max ).Theseexperimentsareprincipally intendedtobediagnostic:wherebiomassaccumu- lationandratesofproduction/photosynthesisare significantlyenhancedinbottlestreatedwithiron- and/ormacronutrients,relativetocontroltreat- ments,weinferthatalgalcommunitygrowthrate islimitedbydeficienciesiniron-and/ormacro- nutrients,respectively.Thestatisticalsignificance ofdifferencesbetweenaveragevaluesofmeasured parameterswereassessedusingtheunpaired student’st-testatthe95%confidencelevel.In eachofourthreeexperiments,therewasrelatively littleornodrawdowninnitrateoraccumulation ofbiomass(Chla,POC,PON)inthecontrol treatments(seebelow).AsdiscussedbyFitzwater etal.(1996),thiswouldsuggestthattherewasno significantironcontaminationoftheseawater collectedandusedintheseexperiments. 3.1.PFZstation Fig.2ashowswater-columnprofilesoftempera- ture,salinity,density(s t ),fluorescence,N+N,Si, anddissolvedFeobtainedatthePFZstation (Station1 0 ).Heretheupper20mofthewater columnwasweaklystratified,probablyasaresult ofsurfacewarming,withawell-definedpycnocline betweendepthsof B 40and60mthatwasroughly coincidentwithamaximuminfluorescence.The relativelyhighconcentrationsofnitrate( B 20mM) andlowconcentrationsofsilicicacid(1–2mM)in theupperwatercolumnaretypicalofPFZwaters duringmidtolatesummer(RintoulandTrull, 2001;Nelsonetal.,2001).Withintheeuphotic zone,dissolvedironconcentrationswerehighest ( B 0.4nM)neartheseasurface(15mdepth)and atthebaseofthepycnocline(75mdepth),and lowest( B 0.2nM)nearthebaseoftheuppermixed

 ()  3334P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 (a)    Station 1' - PFZ 0 05  001  051  002  052  003  0246810 3333.53434.5352626.52727.500.51010203000.20.40.60.81 potential temp. (˚C) salinity
t fluorescence N+N (µM) dissolved Fe (nM) Si (µM) (b)   Station 7- SAF/STF 0 05  001  051  002  052  003  101112131433.53434.53535.5 25.52626.52700.5105101520 00.20.40.60.81 potential temp. (˚C) salinity
t fluorescence N+N (µM) dissolved Fe (nM) Si (µM) (c)    Station 8 - STZ 0 05  001  051  002  052  003  10121416182034.53535.53636.525.52626.52700.51051000.20.40.60.81 potential temp. (˚C) salinity
t fluorescence N+N (µM) dissolved Fe (nM) Si (µM) Fig.2.Verticalprofilesofpotentialtemperature,salinity,s t ;fluorescence,N+NandSi,anddissolvedFeat(a)PFZstation,(b)SAF/ STFstation,and(c)STZstation.Fluorescencedatawerecollectedbetween18:30and08:30localtime.

P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 layer(45mdepth).Total-dissolvableFeconcen- trations(datanotshown)werewithintheanaly- ticaluncertaintyofthedissolvedFe concentrationsthroughouttheupperwatercol- umn.Thedissolvedironconcentrationsarehigher thanvaluesof o 0.2–0.3nMFethathavebeen reportedforsurfacewatersinotheropen-ocean sitesneartheAntarcticPolarFront(e.g.,Martin etal.,1990b;deBaaretal.,1999;Measuresand Vink,2001),andmaythusreflectironinputsfrom sedimentsontheCrozetIslandplatform(located tothewestofStation1 0 ),asreporteddownstream fromtheKerguelenIslandplatform(Blainetal., 2001;Bucciarellietal.,2001),aswellasrecent inputsofmineralaerosol(presumablyfromsouth- ernAfrica)attheseasurface,asobservedinthe centralnorthPacific(Brulandetal.,1994). Atthisstationwesetouttoexaminethe potentialrolesofbothironandsilicicacidin limitingalgalgrowthrate,giventhatbothspecies werepresentatrelativelylowconcentrationsinthe uppermixedlayer( B 0.3nMand1.2mM,respec- tively).Thus,theincubationbottleswereamended withiron(+Fe),silicicacid(+Si),andiron+si- licicacid(+Fe+Si)(seeTable1).However,as detailedinSection2,thesilicicacidstocksolution usedinthisexperimentwassufficientlycontami- natedwithironsuchthatourexperimental treatmentsmustbeconsideredasonly+Feand +Fe+Si(i.e.the+Siand+Fe+Sitreatments eachcontainedsignificantadditionsofbothiron andsilicicacid).ResultsfromthePFZstation experimentareshowninFig.3.Littleorno significantchangesindissolvednutrients,algal biomassandalgalproductionrateswereobserved duringthefirst6daysoftheincubationexperi- ment,butafter9daystherehadbeensignificant increasesinN+Ndrawdown,Chla,POC,PON andP net inthe+Fe,+Siand+Fe+Sitreatments relativetothecontrolbottles.Significantdiffer- encesbetweenthe+Fe,+Siand+Fe+Si treatmentswereonlydiscernableintheBSiand P max data,withthegreatestincreasesinBSiand P max observedintheFe-contaminated+Sitreat- ments(althoughP max measurementswerenot replicatedforthesetreatments),followedbythe +Fe+Siand+Fetreatments.After9days incubation,theaveragemolarSi:Ndrawdown 3353 ratios(drawdowninSi/drawdowninN+N)were zerointhecontrolbottles(nomeasurableSi drawdown), o 0.1inthe+Fetreatments,andnear 0.2inthe+Siand+Fe+Sitreatments. Inthisexperiment,thenanoplanktonand microplanktoncountedbymicroscopywereiniti- allydominatedby B 3–10mmmonads(roughly sphericalcellswhichwerenotdefinitivelyidenti- fied),withlessernumbersofdinoflagellates, pennatediatoms(includingPseudo-nitzschia, Navicula,andFragilariopsisspp.),andother unidentifiablecells(Fig.4a).Themicroscope observationsindicatelittlechangeinalgal communitycompositioninthecontroltreatments after9daysincubation,whereasthelargeincreases inalgalbiomassinthe+Fe,+Siand+Fe+Si treatmentswereapparentlyduetoincreasesinthe numberofsmall,lightlysilicifiedpennatediatoms, mainlyofthegenusPseudo-nitzschia,and monads(Fig.4a).The+Fetreatmentscontained similarabundancesofthepennatediatoms andmonads,whereasthetreatmentscontaining bothFeandSiwerenumericallydominatedby pennatediatoms,whichweremostnumerous intheFe-contaminated+Sitreatments(Fig.4a). ObservationsmadeusingTEMsuggestthat themonadsthatgrewinthe+Fe,+Siand +Fe+Sitreatmentsweredominatedbyhapto- phytes,includingChrysochromulinaandPhaeocys- tisspecies(A.Davidson,pers.comm.2001). Inthisexperimentonly,subsampleswerealso analyzedbyflowcytometry,withresultssuggest- ingthatpicoeukaryoteswereatmostaroundhalf asabundantasnanoeukaryotes,andthatcyano- bacterialnumbersremainedrelativelylowinall experimentaltreatments.Theflowcytometrydata indicateincreasesinthenumbersofbothpicoeu- karyotesandnanoeukaroytesinthe+Fe,+Si, and+Fe+Sitreatments,relativetocontrol samples,withnanoeukaryotesshowingthegreat- estincreasesinrelativeabundance(C.Dubreuil, pers.comm.1999). Takentogether,thechemicalandbiological resultsofthisexperimentindicatethatiron additionincreasedalgalcommunitygrowth,pri- marilybysmallpennatediatoms,whichprobably accountedformostoftheaccumulatedbiomass, andmonads.Relativetoadditionofironalone,

3363P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–3349 control eF+ Station 1' - PFZ +Si (+Fe) +Fe+Si 5202 02 51 51 01 01 5 5 00 02468100246810 incubation time (d)incubation time (d) 201000 05751 00501 0525 00 02468100246810 incubation time (d)incubation time (d) 5002 4 051 3 001 2 05 1 00 02468100246810 incubation time (d) incubation time (d) 052 03 002 051 02 001 01 05 00 02468100246810 incubation time (d) incubation time (d) 0 P Fi m g a . x 3 v . s. R i e n s c u u lt b s at o i f on bo t t i t m le e - . in N cu ot b e at t i h o a n t e t x h p e er + im Si en tr ts ea f t o m r e t n h t e s P w F e Z re st c a o t n i t o a n m ( i S n t a a t ti e o d n w 1 it ) h , i s r h o o n w i i n ng thi N s + ex N p , er S i i, me C n h t l . a,POC,PON,BSi,P net and

P.N.Sedwicketal./Deep-SeaResearchII49(2002)3327–33497333 Fig.4.Abundancesof>2mmalgalcellscountedbylightmicroscopyintheinitialandfinalsamplesfromexperimentsat(a)thePFZ station,(b)theSAF/STFstation,and(c)theSTZstation. additionofbothironandsilicicacidhadno furthersignificanteffectonnutrientdrawdown, accumulationofbiomassasestimatedfromChla, POCandPON,orestimatednetprimaryproduc- tionrate,suggestingthatcommunitygrowthwas notimmediatelylimitedbysilicicaciddeficiency. However,theelevatedBSiandP max valuesand enhancedabundanceofdiatomsobservedinthe +Siand+Fe+Sitreatmentsindicatethataddi- tionofsilicicacid(incombinationwithiron)did influencephytoplanktoncommunitycomposition, biogenicsilicaproductionandgrosslight-satu- ratedrateofphotosynthesis,bypromotingthe growthofsmallpennatediatoms.Itiscuriousthat theFe-contaminated+Sitreatmentsachieved significantlyhigherconcentrationsofbiogenic silicaandgreaternumbersofdiatomsthanthe +Fe+Sitreatments,giventhatthe+Fe+Si treatmentscontainedthesameconcentrationof silicicacidasthe+Sitreatmentsandanadditional B 1.2nMiron.Wespeculatethatthecontaminant FeaddedwiththeSistocksolutionwascomplexed