6.10 BiologicalFluxesinthe andAtmospheric p CO2 D.Archer University ofChicago,IL, USA

6.10.1 INTRODUCTION 275 6.10.2 CHEMICAL REARRANGEMENT OF THE OF THE OCEAN 276 6.10.2.1 The Organic BiologicalPump 276 6.10.2.1.1 limitation andnewproduction 276 6.10.2.1.2 asalimitingmicronutrient 276 6.10.2.1.3 Measuringcarbon export 277 6.10.2.2 CaCO3 Production andExport 277 6.10.2.2.1 Calciteandaragonite 277 2 2 6.10.2.2.2 Temperatureand[CO3 ]regulation ofCaCO3 production 278 6.10.2.2.3 Latitudinaldistribution discrepancy 278 6.10.2.2.4Organic C/CaCO 3 production ratio 278 6.10.2.3 SiO2 Production andExport 278 6.10.2.4GeochemicalSignatureofthe BiologicalPump 279 6.10.2.4.1 Organic carbon redissolution depth 279 6.10.2.4.2 CaCO3 watercolumn redissolution 279 6.10.2.4.3 SiO2 sinkingandredissolution 280 6.10.2.4.4The ballast modelfor sinkingparticles 281

6.10.2.5DirectAtmospheric p CO2 Signatureofthe BiologicalPump 281 6.10.2.5.1 Organic carbon pump 281 6.10.3 CONTROLS OF MEAN OCEAN 282 6.10.3.1.1 Controls on organic matterburial 283 6.10.3.1.2 CaCO3 285 6.10.3.1.3 SiO2 286

6.10.3.1.4Indirectatmospheric p CO2 signatureofthe biologicalpump 286 6.10.4SUMMARY 287 REFERENCES 287

6.10.1 INTRODUCTION asnitrateandiron,the physicsofsinkingorganic matter,andthe production andredissolution of The basic outlinesfor the carbon cycleinthe phytoplankton companion CaCO3 and ocean,asitisrepresented inoceancarbon cycle SiO2 ,inthe watercolumn andinthe . models,weresummarized byBroeckerandPeng The carbon cycleistypically brokendown into (1982).Since then,ongoingfieldresearch has two components,bothofwhich will be summar- revised thatpicture, unearthingnewdegreesof ized here.The firstisthe biologicalpump, freedom andsensitivitiesinthe oceancarbon effectingthe redistributingofbiologically active cycle, which mayprovide cluestochangesinthe elements like carbon,nitrogen,andsilicon within behavior ofthe carbon cycleoverthe glacial the circulatingwaters ofthe ocean. The secondis cyclesanddeeperinto geologicaltime.These the ultimateremovaloftheseelements byburialin sensitivitiesinclude newideasabout the chem- . Thesetwo componentsofthe carbon istry andcyclingofphytoplankton nutrientssuch cycletogethercontrol the meanconcentrations of

275 276 BiologicalFluxesinthe OceanandAtmospheric p CO2 many chemicals inthe ocean,includingoceanpH such asdeposition from the atmosphere(asinthe andthe p CO2 ofthe atmosphere. caseofiron),or localproduction (asinnitrogen Thischapteroverlaps considerably withthe fixation).Traditionally,oceanographers have only two others from thisTreatiseIhavehad the distinguished “new” versus “recycled”primary pleasureto read:Chapters 6.04and6.19.My production usingthe different forms ofbiologi- chapterisdistinguished, Isuppose, bythe callyavailablenitrogen(EppleyandPeterson, perspectiveofageochemicaloceanmodeler, 1979). Whenphytoplankton degrade, theirnitro- attemptingto integratenewfieldobservations genisreleased inthe reducedform,asammonia, into the context ofthe oceancontrol ofthe p CO2 of which correspondstonitrogen’s chemicaloxi- the atmosphere. dation stateinproteins andDNA.Ammonia released withinthe euphotic zoneisquickly assimilated bythe next generation ofphytoplank- ton. Ammonia released below the euphotic zone 6.10.2 CHEMICAL REARRANGEMENT escapesthisfate, becausethe darkness precludes OF THE WATERS OF THE OCEAN . Instead, thisammonia gets oxi- 6.10.2.1 The Organic Carbon BiologicalPump dized bybacteria to nitrate.Therefore, most of the biologically availablenitrogenindeepwaters The cyclingofdissolved carbon andassociated ofthe oxic oceanisinthe form ofnitrate.Using chemicals (such asoxygen,nitrogen,andphos- thisdichotomy,oceanographers candistinguish phorus)inthe oceanisdrivenbythe livesand betweennewproduction,supported byupwelling deathsofphytoplankton.Sunlightisrequired asan from below,asthe uptake ofisotopicallylabeled energysource for photosynthesis,andisavailable nitrate, versus recycled production asthe uptake insufficient intensity only into the top 100 morso ofisotopicallylabeled ammonia.The ratioof ofthe ocean; the rest ofthe oceanisdark, or at newproduction to production hasbeencalled the least too darkto makealivingharvestingenergy f -ratiobybiologicaloceanographers (Dugdale fromlight. Thenetgeochemicaleffectof andGoering, 1967). Valuesofthe f -ratiorange photosynthesisinthe oceanistoconvert dissolved from 10% to 20% inwarm,oligotrophic, low- constituentsinto particles,piecesofsolid material nutrient open-oceanconditions,to 50% inblooms thatcantake leaveoftheirparent fluid bysinking. andnear-shoreconditions (EppleyandPeterson, Astheydo,theystripthe surface waters of 1979). However,the apparent correlation whichevercomponent locally limits the growth between f -ratioandprimary production maybe ofphytoplankton (typically nitrogenoriron), anartifactofdifferencesinsea-surface andcarrythe entiresuiteofbiologicallyuseful temperature(Laws etal .,2000). elementstodepth. Asthe particlessink, theydegrade inthe column,or theyreach the seafloor,to degrade withinthe sediments or to exitthe oceanbyburial. 6.10.2.1.2 Iron asalimitingmicronutrient Thiscombination ofremovalatthe sea surface andaddition atdepthmaintains higherconcen- MartinandFitzwater(1988)originally proposed trations ofbioactiveelements andcompoundsat thatthe availability ofthe micronutrient iron may depthinthe oceanthanatthe surface.Theshape be limitingphytoplankton production inremote ofadepthprofileofdissolved nitrogeninthe partsofthe worldocean. Thisproposalwas form ofnitrate, alimiting“fertilizer” or nutrient to convincingly provenbyaseriesofopen-ocean phytoplankton,resemblesthoseofotherbioactive fertilization experiments,inthe equatorialPacific (Coale etal .,1996; Martin etal .,1994)andinthe compounds,includingC,PO 4 ,Si, Fe, andCd.This process,fundamentaltothe oceancarbon cycle, (Boyd etal .,2000). Although hasbeencalled the biologicalpump (Broeckerand photosynthesisratesandchlorophyllconcen- Peng, 1982; see Chapter6.04). trations increased significantly inall cases,the ironfertilization wasless effectiveatstimulating carbon sinkingto depth, especially inthe Southern 6.10.2.1.1 Nutrient limitation andnew Ocean,wherethe added ironrecycled inthe production euphotic zoneuntilitdispersed bylateralmixing somemonthslater(Abraham etal .,2000). Several Phytoplankton insurface waters,upon com- attempts havebeenmadeto modelthe iron cyclein pletion oftheirearthly sojourns,sufferoneoftwo the ocean(ArcherandJohnson,2000; Fung etal ., fates. Eithertheyaredegraded andconsumed 2000;LefevreandWatson,1999).Although withinthe surface waters ofthe ocean,or they dust deposition delivers moreiron to the surface maysinkto the deepsea.Onaverage, the ofthe ocean(Fung etal .,2000),uncertainty removalofdissolved from the surface remains about the availability ofthatirontothe zoneto depthmustbe balanced bymixingup of phytoplankton. Ithasbeenargued thatdust dissolved nutrients from below,or othersources deposition mayplayonly aminor roleinregulating ChemicalRearrangement ofthe Waters ofthe Ocean 277 phytoplankton abundance (ArcherandJohnson, waters. Verticalconvection thus drivesexportof 2000; LefevreandWatson,1999). TheNorth sea-surface DOC to depth.Upwellingbringslow- PacificOcean,inparticular,appears to be a DOC watertothe sea surface,wherephotosyn- problemarea, withhigh deposition ofdust (Duce thesisbegins the production ofDOC.ThisDOC andTindale, 1991)failingto drivedepletion ofsea- canbe exported laterallyorback to the ocean surface nutrients. interior.Onaglobalmean,DOC export accounts for less thanhalfofthe exportoforganic matter from the surface ocean,andonly atiny fraction to 6.10.2.1.3 Measuringcarbon export the deepsea (Yamanaka andTujika, 1997). Inthe steadystate(thatistosay,on along Nutrient supply. Inspiteofthe developments of enough timeaverage),the uptake ofnewnutrients, sediment trapcalibration byradiogenic isotopes nitratefor example, must be balanced bythe andexport oforganic matterinthe suspended or removalofnutrients inexported organic matter. dissolved states,problems remaininbalancing Oceanographers havedefined aquotient called the carbon andnutrient exportagainst supplyby e -ratio,describingthe exportefficiencyofeupho- physicalfluid transport. Thisproblemismost tic zoneorganic matter,exportproduction/gross acuteinthe subtropicalgyres,intensely sampled production. Onlongenough timescales,the e -ratio intwo time-seriesJGOFS stations:onenear mustequalthe f -ratio. Inthisway,the biological Bermuda (BATS)andonenearHawaii (HOT). pump acts to limitthe ofthe ocean, Netverticalmotionsinthe upperoceanare pacingitaccordingto the overturningtimescaleof determined bythe divergence ofsurface currents, the oceancirculation. drivenbyfriction withthe wind.The direction Sediment traps andsinkingparticles. The andintensity ofthe resultingverticalmotion is measurement oforganic carbon exportfrom the determined bythe curl ofthe windstress,which surface oceanhasremained problematic over leadstonetupwellinginthe subpolargyres, the decades. Theworkhorseofthe oceanographic carryingnutrients verticallyinto the euphotic community for determiningsinkingparticulate zone, but downwardinthe subtropicalgyres, fluxesisthe moored sediment trap. Sediment traps supplyingno nutrients for photosynthesisatall areplagued bybiasesfrom hydrodynamicsand (Peixoto andOort,1992). Onasmallerspatial from “swimmers,”which actively enterthe open scale, eddiesandfronts leadto verticalmotion and collection cupsofthe traps to feed,either offsets inthe depthofthe constant-density consumingour signalorperhaps perishinginthe (isopycnal) surfaces(MahadevanandArcher, preservativeandaddingto it(Hedges etal .,1993). 2000; McGillicuddyandRobinson,1997; Swimmers areoperationally excluded bypicking OschliesandKoeve, 2000). Thesemesoscale themout individually. Hydrodynamic over-or motionsmaycarry nutrients verticallyinto the under-trappingcanbe detected bycomparingthe euphotic zone, particularlyinBermuda wherethe delivery rateofradiogenic thorium withits known proximity to the GulfStreamgenerateshigh levels source rateinthe watercolumn above(Buesseler, ofmesoscalemotions. Sea-surface helium con- 1991). Ingeneral,deepertraps (below 1,500 m) centrations (Jenkins,1988)seemtosupportthe arefoundto be unbiased byhydrodynamic idea ofverticaltransportbyphysicalfluid trappingerrors,whereasshallowertraps may motions. AtHawaii inparticular,nitrogenfix- undersamplethe truesinkingflux byafactor of ation,the production ofbiologically available 2ormore, on average (Yu etal .,2001). Thus, nitrogeninthe formofammonia from the historicalestimatesofeuphotic zonecarbon ex- generallyinert but ubiquitous N 2 gas,isthought port based on sediment traps mustbe evaluated to supply alarge fraction ofthe required nitrogen withcaution (Berger,1989; Martin etal .,1987). (Karl etal .,1997). Nutrientsmaybe mined from Export ofDOC. Anotherrecent revision ofour deeperwaters andactively transported into the understandingofcarbon exporthasbeenthe euphotic zonebydiatoms (Villareal etal .,1999). generation ofdissolved organic matter(DOM). Dissolved matterisunableto physically separate itselffrom its parent fluid, the wayparticulate 6.10.2.2 CaCO3 Production andExport organic matter(POM)doesbysinking.However, 6.10.2.2.1 Calciteandaragonite significant gradients inthe concentration ofDOC exist inthe ocean,so fluid transportthrough the Severalmineralphasesareproducedbyphyto- steady-stateDOC concentration fieldrepresents andexported to depthinthe ocean, the export ofsignificant quantitiesoforganic associated withandanalogous to the production matterinthe dissolved form. Concentrations of andexportoforganic matter. First amongtheseis DOCinthe deepoceanrange fall inarelatively CaCO3 ,comprised oftwo mineralphases: lowandnarrowrange,37–42 m M.DOC andthe moresolublearagonite(Milliman,1974; reacheshigherconcentrationsinthe surface Mucci, 1983).The ofcalcitedepends ocean,up to 80–100 m Minoligotrophic surface also on the concentration ofmagnesium;higher 278 BiologicalFluxesinthe OceanandAtmospheric p CO2 magnesium leadstohighersolubility (Plath etal ., deep-surface gradientsinalkalinity (mostly from 1980).The dominant mineralindeep-sea sedi- dissolved CaCO3 )andsomeproxy for organic ments islow-magnesium calcite, the least soluble carbon degradation,such asnitrateor totalCO2 form. Calciteisthermodynamically stableinthe corrected for CaCO3 (Li etal .,1969). However, surface oceaninmostparts ofthe world, anddown thisestimateissensitiveto the depthscaleof to severalkilometers depth(BroeckerandPeng, particledegradation inthe watercolumn. Organic 1982; Millero,1982). matterdegradesmorequickly thanCaCO3 dissolves,andso much ofthe organic matter producedmaydegradeinthe top fewhundred 2 2 6.10.2.2.2 Temperatureand[CO3 ] meters ofthe watercolumn,whereits impact regulation ofCaCO3 production on the abyss/surface chemicalgradients will be smallerthanif itdegraded inthe abyss itself. Ingeneral,CaCO3 sinkingfluxesinthe deepsea Arecent analysisofthe 100–200 mgradientsin tendto correlatewithfluxesoforganic matter phosphateandalkalinity fromWOCE data (DymondandLyle, 1993; KlaasandArcher, seems to indicateaproduction ratioof10 or 2002; Milliman,1993; TsunogaiandNoriki, more(Sarmiento etal .,submitted). Reconciling 1991). However,CaCO 3 production decreases the shallow-andwhole-oceangradients ofsoft- dramatically whensea-surface temperaturesdrop tissueandCaCO3 pump products will require below , 10 8 C(Honjo etal .,2000; Wefer etal ., someknowledge ofthe depthto which these 1990). ThiscouldbebecauseCO 2 gassolubility components bothsink. increaseswithdecreasingtemperature, driving 2 2 CO3 to alowerconcentration bypHequilibrium chemistry.Laboratory andcontrolled system growthexperiments haveobserved acorrelation 6.10.2.3 SiO2 Production andExport 2 2 betweenwater[CO3 ]andcalcification rates incoccolithophorids(Riebesell etal .,2000)and The secondmost important mineralphaseofthe incorals (Langdon etal .,2000).Shells of biologicalpump isbiogenic ,SiO2 .Opal foraminiferatendto be thickerfromhigher shellsareproduced bydiatoms,amarinealga, and 2 2 byradiolarians,aheterotrophic . Onalarge [CO 3 ](warmer) source waters (Barkerand Elderfield, 2002). However,indeeptraps there scale, production ofSiO2 islimited bythe isnosignificant correlation betweenthe rainratio availability ofdissolved silica, inthe form of andsea-surface temperature, once SST risesabove silicic , H 4 SiO4 . the critical10 8 Clevel(KlaasandArcher,2002). Ingeneral,diatomsappeartobe adapted to turbulent,unstableconditions (Margalef, 1978), andchangesinwatercolumn turbulence associ- ated withclimatemaybe responsiblefor decadal- 6.10.2.2.3 Latitudinaldistribution timescalechangesinthe relativeproduction rates discrepancy ofdiatomsandcoccolithophoridsinthe North Adiscrepancyexists betweensatelliteobser- Atlantic (Antia etal .,2001). The dynamicsof vations ofcalcification bycoccolithophoridsand growtharedocumented mostclearly inthe geochemicalestimatesofthe spatialdistribution NorthAtlantic, wheredeepwintertimemixing ofcalcification rate.Massiveblooms ofcocco- limits the growthofphytoplankton untilspring- lithophoridsareseenfrom space asthe organisms timestratification andthe springbloom. The shedtheirCaCO3 plates(coccoliths),which sequence ofevents thatfollows isthatdiatoms scatterlightinthe waterand, therefore, appear bloom mostquickly,strippingthe surface oceanof lightfrom space.Theseblooms arefoundinhigh dissolved silica, andpartially depletingitofthe latitudes,especially inthe NorthAtlantic (Brown soft-tissuenutrients nitrateandphosphate(Lochte andYoder,1994). Incontrast,geochemical etal .,1993). The samesequence isseeninthe estimatesofcalcification,such asthe sourcesof Southern Ocean,wherediatoms apparently inthe watercolumn ofthe ocean(Feely depletesurface waterofH4 SiO4 beforethe elimi- etal .,submitted;Sarmiento etal .,submitted),and nation ofdissolved nitrateandphosphate(Levitus, the distribution ofCaCO 3 on the seafloor(Archer, 1982; Levitus etal .,1993). arefoundin 1996a)tendto revealaworldwheretotal coastalwaters wherenutrients arereplete(Nelson calcification ismostintenseinlowerlatitudes. etal .,1995),andinthe bloomingplumeofiron fertilization experiments(Coale etal .,1996). Diatomexport from the surface oceanhasbeen correlated withorganic carbon export,andithas 6.10.2.2.4Organic C/CaCO 3 production ratio beenproposed thatafewlarge diatomsmight, infact,dominatethe exportoforganic carbon The classicalestimate, 4mol oforganic matter (Goldman,1987).(Iron mayplayanotherrole permoleofCaCO3 ,wasderived from the relative inthe , asithasbeenshown that ChemicalRearrangement ofthe Waters ofthe Ocean 279 the diatoms thatgrow iniron-limited Southern subsurface watermassesreturn to the sea surface Oceanconditions carry moresilica permoleof morequickly thanthe deepest waters ofthe ocean. organic carbon or nitrogen(Franck etal .,2000; Carbon andnutrients deposited inthe shallow Hutchins andBruland, 1998;Takeda,1998).) thermoclinewill,therefore, haveasmallerimpact Theseconsiderations haveled to amodel on the surface/deepfractionation ofthe oceanthan paradigmwhereinabundant supply ofsilica abyssalredissolution.The mineralphasescalcite mightstimulatethe diatoms,depletingavailable andopal,andto alesserextent , tendto sea-surface nitrateandthus starvingoutthe cocco- dissolvemoredeeplyinthe watercolumn. lithophoridswhowouldproduce CaCO3 .Honjo refers to the Atlantic asa“carbonateocean” and the Pacific asa“”oceanbased on acompa- 6.10.2.4.1 Organic carbon redissolution risonofAtlantic andPacificCaCO3 andSiO2 depth sinkingfluxes(Honjo,1996). Models ofocean biogeochemistry (Archer etal .,2000b;Heinze Martin etal .(1987) publishedaninitial etal .,1999; Matsumoto etal .,inpress)tendto formulation ofaredissolution curvefor organic include somesilicadependence for CaCO 3 carbon,which hasextremely beeninfluentialin production rates,based on the idea thatdiatoms the oceanmodelingcommunity. Theirformu- outcompetecoccolithophoridsgivensufficient lation wasbased on sediment trapdatafrom the dissolved silica.Indeepsediment traps,however, NorthEast Pacific Ocean(the VERTEX program) the sinkingflux ofsilica tendstocorrelate anddescribed the sinkingflux oforganic matteras positively withCaCO3 flux,ratherthannegatively apowerlaw: asonemightexpectif silica inhibits CaCO3 Flux z Flux 100 m z = 100 2 0 : 858 production (Figure1; KlaasandArcher,2002). ð Þ¼ ð Þð Þ Subsequent analysishasrevealed spatialhet- erogeneity inthe organic carbon sinkingflux 6.10.2.4GeochemicalSignatureofthe (Lutz etal .,inpress). Ingeneral,the fluxesinthe BiologicalPump interior ofthe oceanareless variablethanexport The intensity ofthe biologicalpump signature fluxesdetermined from the surface ocean; inother on the chemistry ofthe oceandepends,ofcourse, words,moreproductiveregions tendto be on the totalrateatwhich materialisexported from relatively less efficient atsinkingto depththan the surface euphotic zone, but italsodependson less productiveregions. Otheranalyseshave how deepitsinksbeforeitdegradesback to focused on the roleofheaviermaterials,such as dissolved constituents. Thisisbecauseshallower CaCO3 ,SiO2 ,andlithogenic clays,asballast,in transportingorganic mattertodepth(Armstrong etal .,2002; KlaasandArcher,2002)(see below).

6.10.2.4.2 CaCO3 watercolumn redissolution

Asignificant fraction ofthe CaCO3 producedin the surface oceanreachesthe seafloor,to dissolve or accumulateinthe sediments.Thetraditional viewhasalways beenthatwatercolumn CaCO3 dissolution isrelatively insignificant,consistent withits thermodynamic stability. Sediment trap results suggest thatCaCO3 survivesrelatively intactits tripthrough the watercolumn ofthe ocean. Measured fluxesofCaCO3 on multiple sediment traps atdifferent depthsonasingle mooringshow no cleartrendtowarddecreasing Figure1 Sediment trapsinkingfluxesofCaCO 3 CaCO3 flux withdepth(Figure2). Inaddition, plotted against SiO2 .The point ofthisplot isthat visualexamination ofCaCO3 shells captured in high fluxesofSiO2 tendto correlatewithhigh fluxes sediment traps doesnot revealany overwhel- ofCaCO3 ,except inregions ofcoldSST (sources Brewer etal .,1980; Deuser,1986; DymondandCollier, mingly obvioussigns ofdissolution withinthe 1988;DymondandLyle, 1993; Honjo,1990; Honjo watercolumn (Honjo,1976). However,three lines etal .,1982,1998;Ittekkot etal .,1991; Martin etal ., ofevidence suggest that,inspiteofsediment trap 1987; Milliman,1993; Noriki andTsunogai, 1986; data, aconsiderablefraction ofthe CaCO3 Pilskaln andHonjo,1987; Tsunogai andNoriki, 1991; producedinthe upperoceandissolvesasitsinks Wefer etal .,1982,1988). through the watercolumn. 280 BiologicalFluxesinthe OceanandAtmospheric p CO2 ocean,MunkandWunsch (1998)showed that boundary mixing, advected into the oceaninterior alongisopycnalsurfaces,isindistinguishable from morediffusemixingthroughout the interior. We maypresumethatthe samereasoningapplies to sourcesofalkalinity. CaCO3 budgets.Anotherargument comesfrom comparison ofmeasured calcification ratesinthe surface oceanwithratesofCaCO 3 accumulation insediments. Milliman etal .(2000) makeastrong casefor dissolution inthe watercolumn abovethe saturation horizon for CaCO3 ,based on acompa- rison ofeuphotic zonecalcification rateestimates (Balch etal .,2000; Balch andKilpatrick, 1996) andsediment accumulation ratesinregions wherethe overlyingwaterissupersaturated with respecttocalcite.Oceancarbon cyclemodels typically imposeasignificant watercolumn dissolution fraction ofCaCO3 rain(Archerand Maier-Reimer,1994;Orr etal .,2001).The mechanism for thisdissolution isunclear. The Feely results seemtoindicatethatthe flux of aragoniteinthe openoceanmaybe higherthan had previously beenassumed.However,theyalso Figure2 Sediment trapsinkingfluxesofCaCO3 . Multipletraps which shareacommon mooringare findasignificant dissolution flux uncorrelated connected withlines. The point ofthismessy plot isthat withthe aragonitesaturation horizon,which ispre- the sinkingflux ofCaCO3 doesnot appeartodecrease sumably calcite.Ithasbeenproposed thatcalcite withdepthinresponseto watercolumn dissolution mightbe dissolvinginthe guts ofzooplankton (sourcesBrewer etal .,1980; Deuser,1986; Dymond (JansenandWolf-Gladrow,2001),or within andCollier,1988;DymondandLyle, 1993; Honjo, acidic microenvironmentsinside marinesnow. 1990; Honjo etal .,1982,1998;Ittekkot etal .,1991; Planktonic shell thickness. Athirdindicator of Martin etal .,1987; Milliman,1993; Norikiand Tsunogai, 1986; Pilskaln andHonjo,1987; Tsunogai CaCO3 dissolution withinthe watercolumn andNoriki, 1991; Wefer etal .,1982,1988). comesfrom amoresubtleexamination ofthe shellsofforaminifera, the thickness oftheirwalls infact. Thistechniquewaspioneered byLohmann (1995)andfurtherpursued byBroeckerandClark Chemicalsignature .Onesuch lineofevidence (2001).Whattheyfindisthatthe shell thickness of arisesfrom the distribution ofalkalinity inthe forams increasesthroughout the first kilometerof watercolumn. UsingGeosecsdata, Fiadiero the watercolumn,anddecreasesbelow , 3km, (1980) concluded thatthe alkalinity distribution representativeofdissolution withinthe water inthe deepPacific Oceancouldonly be explained column. bywatercolumn dissolution ofCaCO 3 .More recently,the WOCE programglobaloceansurvey hasprovided us withanew,much moreextensive andhigher-quality globalsurveyofoceanchem- 6.10.2.4.3 SiO2 sinkingandredissolution istry. Feely etal .(submitted)haveused thesedata The dissolution ofSiO2 asitsinksthrough the to essentially confirm the Fiadiero findings. Inthe watercolumn isanotherongoingtopic ofresearch Pacific Ocean,theyconclude thatupto75%of anddiscussion. The Southern Oceanisacrucial the CaCO production flux couldbedegradingin 3 depocenterfor SiO2 indeep-sea sediments,inthe the watercolumn ofthe ocean. Theyfindtwo hot form ofabandofhigh-opalsediments known as zonesofCaCO 3 dissolution,onecorrelatingwith the opalbelt. Inthe viewofNelson etal .(1995), the aragonitesaturation horizon inthe depth the intensity ofopalburialinthislocation isnot range of1–2km,comprisingapproximately half dueto high ratesofopalproduction,but rathera ofthe totaldissolution flux,andanothersource in peculiarhigh efficiencyofopaltransfertothe the depthrange of2–5 km. The deepalkalinity seafloorinthisregion. Gnanadesikan etal .(1999) source iscomparableto calculated ratesof point to extremetemperaturesensitivity ofopal sedimentary dissolution (Archer,1996b),which dissolution kinetics,andfindthatanocean maybe difficult to distinguishfrom dissolution in circulation/silica cyclemodelisbetterableto the watercolumn. Addressingadifferent question, reproduce the observed distribution ofopalwhen thatofthe location ofdiapycnalmixinginthe thisdependence isincluded.Opaldissolvesfaster ChemicalRearrangement ofthe Waters ofthe Ocean 281 inwarmerwater,inpart explainingthe seafloor andcarbon export. Perhaps the particles,produced transport efficiencywithinthe coldSouthern inthe organic carbon to CaCO3 ratioof4–10,are Oceanwaters. Anadditionalfactor maybe the unableto sinkvery quickly untilorganic matter episodic natureofprimary production,which degradation decreasesthe organic ratiotothe precludesastandingcrop ofgrazers. 0.5–1.0 range seeninthe deepsea, atwhich point the express elevatortakesthemtothe seafloor.

6.10.2.4.4The ballast modelfor sinking particles 6.10.2.5DirectAtmospheric p CO2 Signature Oneofthe morerobustfeaturesofthe ofthe BiologicalPump sediment trapdatasetisthe constancyofthe 6.10.2.5.1 Organic carbon pump “organic carbon”:CaCO3 rainratiointhe .Although the absolutefluxesvary widely by The biologicalpump hasafirst-orderimpacton area andtime, the ratiooftheirfluxes(mol :mol) the CO2 concentration ofthe atmosphere, and seems quitestableinthe range of0.5–1.0.One thereforeon the climateofthe earth.The removal explanation for thisobservation,which has ofdissolved CO2 fromsurface waterduring significant implications for the ability ofthe photosynthesisdecreasesthe equilibrium partial oceantochange its pHbychangingthe rainratio, pressureofCO 2 inthe overlyingatmosphere(this iscalled the ballast model(Armstrong etal ., quantity isdefined asthe p CO2 ofthe surface 2002; KlaasandArcher,2002). The density of water). Changesinthe biologicalpump insome organic matterissimilarenough to thatofwater form maybe responsiblefor the glacial/intergla- thatorganic matter,byitself, wouldnot sinkat cial p CO2 cycles(Petit etal .,1999). The strongest all. Armstrong etal .(2002) did aregression of directimpactofthe biologicalpump on p CO2 is sediment trapfrom the equatorialPacific and drivenbythe depletion ofsurface waters of foundthatorganic carbon sinkingfluxescouldbe dissolved CO2 bythe production oforganic describedremarkably well bycombininga carbon (the “soft-tissue”pump). ballasted flux (asimplefunction ofthe ballast Timescale. Thetimescalefor sea-surface

flux atdepthtimessomeuniform“carrying depletion ofCO 2 to affectatmospheric p CO2 is coefficient”)witha“Free POC”component, thatofoceanrearrangement andatmosphere/ which degradesovera600 mdepthscaleinthe oceangasexchange equilibration:on the order surface ocean. KlaasandArcher(2002) extended ofhundredsofyears. Lookinginto the past for thisanalysistothe globalsediment trapdataset, biologicalpump-drivenchangesinthe p CO2 ofthe anddistinguished three forms ofballast:CaCO 3 , atmosphere, wewouldlookfor changesthatoccur , andlithogenic material,mainly on thistimescale(unless the efficiencyofthe terrestrialclays.Theyfoundthatthe carrying biologicalpump itselfischangingmoreslowly coefficient (grams organic carbon pergram thanthat). ballast)for CaCO3 wastwiceashigh asfor the Modelsensitivity.Oneofthe morerecent otherphases,aresult which isqualitatively surprisewasthe discovery thatthe sensitivity of consistent withthe greaterdensity ofCaCO 3 . atmospheric p CO2 to the biologicalpump isquite Whenthe globalfluxesofthe three ballast phases modelspecific (Archer etal .,2000a;Bacastow, aremultiplied bytheircarryingcoefficients,they 1996; Broecker etal .,1999). Inparticular,box foundthat80% ofthe organic matterflux to the models ofoceanchemistryaremoresensitiveto seafloorwascarried byCaCO3 .Thisexplains the efficiencyofthe biologicalpump thanare the relativeconstancyofthe rainratiointhe oceanmodels based on acontinuumrepresen- deepsea, anditalso seems to limitthe ability of tation ofcirculation andchemistry,most the oceancarbon cycleto vary thisrainratioby notably the generalcirculation models (GCMs) changing, e.g.,the relativeproductivitiesof (Figure3). Thehorizontalaxisofthisplot isthe diatomsandcoccolithophorids. Someecological inventoryofthe dissolved nutrient phosphorus,in 3 2 change inthe surface ocean,impactingthe sink- the form ofPO 4 ,withinthe top 50mofthe world ingflux ofCaCO 3 ,wouldmerely change the flux ocean. The sea-surface phosphorus inventory oforganic matterinconcert withthe changing dependsonabalance betweendelivery byup- CaCO3 flux. wellingandmixingofphosphorus from below, Thisidea issomewhatatoddswiththe common andremovalassinkingbiogenic particles. observation from nearthe sea surface,eitherfrom The reason for the discrepancyisthatmost of shallow sediment traps or from phytoplankton the availablesea-surface nutrients (which couldbe growthstudies,thatdiatomsdominatethe export depleted werethe phytoplankton moreefficient or ofcarbon from the surface ocean(Goldman, the circulation less energetic)areto be foundin 1987). However,thereisroom inthe “Free POC” the high latitudes,notably inthe Southern Ocean. component ofArmstrong etal .(2002) for diatoms Astartlingandseminalfindinginthe 1980s was 282 BiologicalFluxesinthe OceanandAtmospheric p CO2 surface thanitisabovebox models,atmospheric

p CO2 isalso less sensitiveto the soft-tissuepump, variations inwhich arefocused inthe high latitudesbythe availability ofnutrients. Con- versely,the effects ofthe CaCO3 pump seemtobe strongerinGCMsthaninbox models (Matsumoto etal .,inpress),perhaps becauseCaCO 3 pro- duction occurspredominatlyinwarm surface waters.The reasons for the discrepancyarestill unclear. Archer etal .(2000a)pointed to vertical diffusion,which dominatesthe residence timeof waterinthe surfacelayers ofthe ocean. Toggweiler(2003) pointstothe extent of equilibration ofsurface waters,inparticularin the high latitudes. Untilthe discrepancybetween box models andGCMsisresolved,itisdifficult to predictthe sensitivity ofthe realoceantochanges inthe soft-tissuebiologicalpump. Figure3 Steady-statemodelatmospheric p CO2 responseto the efficiencyofthe biologicalpump,as CaCO3 pump. The CaCO3 component ofthe 3 2 biologicalpump alsoaffects the p ofsurface indicated bythe inventory ofPO 4 inthe top 50mof CO2 the ocean(source Archer etal .,2000a). waters andthereforeofthe atmosphere.CaCO3 production depletesthe parent waterofdissolved 2 2 carbonateion,CO 3 ,which actually drivesthe thatchangestothe nutrient concentration inthe dissolved CO2 concentration up bythe reaction high-latitude surface oceaninasimpletypeof CO CO2 2 H O 2HCO2 1 oceanchemistrymodels called box models havea 2 þ 3 þ 2 $ 3 ð Þ disproportionately large impactonthe steady- The CaCO3 pump thereforeacts to increasethe stateCO concentration ofthe atmosphere(Knox 2 p CO2 ofthe atmosphere, counteracting, to asmall andMcElroy,1984;Sarmiento andToggweiler, extent,the effectofthe soft-tissuepump. This 1984;SiegenthalerandWenk, 1984). The low- seemingparadoxicalbehavior isaresult ofthe pH latitude surface oceanisconversely less effective chemistry ofthe carbonatebuffersystem. Dis- atdeterminingthe p CO2 ofthe atmosphere. solved CO2 ,inits hydrated form H 2 CO3 ,isan Abox modeliscomprised ofsomesmall acid, anditexists inbufferchemistry equilibrium 2 2 numberofoceanreservoirs,typically three to a withcarbonateion,CO 3 ,the basesalt form of 2 2 fewtens. Thechemicalcharacteristicsofthe sea- H 2 CO3 .The removalofCO 3 bycalcification waterwitheach box istakentobe homogeneous acidifiesthe parent seawater,shiftingthe carbon- (well mixed). Waterflow inthe oceanisdescribed atebuffersystemequilibrium towardhigherCO2 asfluxesbetweenthe boxes; giveninunits ofSv concentration. Aconverseofthisprocess isthe 6 2 1 (Sverdrups,10 ms ),thesetypicallyrange to dissolution ofCaCO3 on the seafloororincoral valuesless than100 Sv. Aflow betweentwo reefsinresponseto the invasion offossilfuelCO2 boxescarrieswithitthe chemicalsignatureofthe into the ocean(Archer etal .,1989b, 1997; Walker source box; thisnumericaltechniqueiscalled andKasting, 1992). Onemightnaively expect “upstream” differencing.Gasexchange acts to coralreefstogrow largerwiththe increased pull sea-surface gasconcentrations,such asCO2 availability ofcarbon to buildthemfrom,but the 14 and CO2 ,towardequilibrium valueswiththe pHequilibrium chemistryoverridesthe absolute atmosphere.Thefluxesimposed betweenboxes abundance ofdissolved carbon. aredetermined bytuningvarious tracers,usually 14 C, towardobserved values. Morerecently,however,itwasfoundthat 6.10.3 CONTROLS OF MEAN OCEAN GCMshaveagreatersensitivity to low-latitude CHEMISTRY forcing(Bacastow,1996; Broecker etal .,1999), andthereforelowersensitivity to high-latitude The meanconcentrations ofconstituentsof forcing(Archer etal .,2000a). AGCM aims to seawateraredetermined not bysimpledistillation represent the chemicalandphysicalfieldsofthe ofriverwaterbut bytheirvarious mechanismsof oceaninacontinuous way,sampled on afinite removalfrom the ocean. The dominant cation in difference grid (discretized). Flows across the grid riverwater,e.g.,iscalcium from weatheringof aredetermined bybalancingthe forcesofinertia, carbonateandsilicaterocks,whereasthe domi- gravity,stress,friction,andthe rotation ofthe nant cation inthe oceanissodium,becausethere

Earth.Becauseatmospheric p CO2 valuesabove areno efficient loss mechanismsfor sodium GCMsareless sensitiveto the high-latitude sea analogous to the formation ofCaCO3 asaloss Controls ofMeanOceanChemistry 283 mechanism for (Holland, 1978). The equilibrium withthe phase.The chemistry cycleofCaCO3 inthe oceanisofparticular ofthe overlyingwaterservesasanupper interest,asitdeterminesthe pHofthe oceanand boundary condition for thisequation. the p CO2 ofthe atmosphere, on timescalesof thousandstohundredsofthousandsofyears. 6.10.3.1.1 Controlsonorganic matterburial Organic matterburialcanbe considered anet source ofoxygentothe atmosphere, andmech- Hypsometry. Globally,the burialoforganic anismsfor regulatingatmospheric O 2 focus on the matterinsedimentsisaccomplishedlargely mechanicsofburyingorganic matterinsediments withinarelatively small area ofthe seafloor,in (CappellenandIngall,1996; Walker,1977). coastalwaters ofthe continentalshelfandslope. The redissolution or burialoforganic matterin Thereareseveralreasons for this. First,ratesof sediments isadecision thatismade jointly bythe tendto be higherincoastal physicsofdiffusion,chemistry oforganic matter waters,inpart becauseofcoastalupwelling, and oxidation,andthe biologywhich mediatesthe inpart becauseofriverineandrecyclingsources chemicalreactions. The concentration profileofa ofnutrients such asiron(Johnson etal .,2001). soluteinsediment porewaterisgoverned bythe Accumulation ratesofabiotic materialsuch as diffusion equation,which canbe writtenmost clayminerals areorders ofmagnitude higherin simplyas near-shoresediments thaninthe abyssalocean (Middelburg etal .,1997; Tromp etal .,1995). This › C › › C 0 D j tendstopromoteorganic carbon burial. Second, › t ¼ ¼ › z M › z þ C the efficiencyoforganic mattertransfertothe sediments ishigherincoastalwaters. The water where C isasoluteconcentration, D M isa column itselfisthinner,minimizingdegradation moleculardiffusion coefficient,and j C isa ofsinkingmaterial. Also,coastalphytoplankton production or comsumption rate.Whenapplied themselvestendto be largerthantheirpelagic to the distribution ofdissolved ,thisexpre- cousins,increasingtheirexport efficiency(see ssion results inageneric oxygenprofilewithinthe Chapter6.04). The highest ratesofprimary sediments (e.g.,Figure4(a)).Atany depthinthe production tendto be foundon continentalshelves sediments,the downwarddiffusiveflux ofoxygen or just alongthe shelf/slopebreak, wherecoastal isbalanced bythe integrated consumption below upwellingismanifested moststrongly. Results of thatdepth.Whenapplied to the concentration ofa the SEEP programindicatethatmost ofthe 2 2 dissolution productsuch asH 4 SiO4 ofCO 3 ,the organic matterproducedon the sloperespires diffusion equation results inageneric profileas there(Biscaye etal .,1994). However,organic illustrated inFigure4(b). Withdepth, the pore carbon concentrations insandyshelfsediments watersbecomeincreasingly isolated from areoftenratherlow. Measured sediment respir- diffusivecontactwiththe overlyingwater,and ation ratesoftenreach amaximum somewhat the concentration ofthe soluteapproaches furtheroffshore, infiner-grained continentalslope sediments thatdepositinthislower-energysetting (Jahnke etal .,1990). Although carbon export from the shelfto the slopemaybe anegligible term inthe shelfcarbon budget,itisbynomeans trivialtothe budgetofthe slope, analogous to crumbsfrom the high table. Degradation ratekinetics. Becausemost ofthe chemicalandphysicalprocessesthatcontrol organic matterdegradation aredetermined by biology,mostofthemaredetermined for model- ingpurposesempirically. The firstorderof business isarateconstant for the degradation reaction itself: OrgC electron acceptor þ CO electron donor ! 2 þ Rate k OrgC ¼ ð Þ where k isthe rateconstant,inunits ofinverse time.Ithasbeenfoundthatdegradation ratesare

Figure4 Modelprofilesof(a)O 2 and(b)H 4 SiO4 independent ofthe concentration ofthe electron insurface sediments. From the Mudsmodel(Archer acceptor (e.g.,O2 )untilthe concentration of etal .,2001). O 2 reachessomecriticallow level(, m M). 284 BiologicalFluxesinthe OceanandAtmospheric p CO2 Thispsuedo-zero-orderdependence iscalled canalso be treated asaconstant withinthe Monodkinetics(Devol,1978). Thesekinetic bioturbated layer. Tomakethingsevenmore rateformulations becomeimportant if the rate complicated,bioturbation ratesderived from constantsfor organic degradation differfor radionuclidesseemtodependon the lifetimeof different electron acceptors. Oxic degradation,in the radionuclide, withfastermixingfor the particular,appears to be fasterthansulfate shorter-lived tracer(Fornes etal .,1999; Smith degradation,but itisconfined to the upperoxic etal .,1993). zoneofthe sediment (Figure5). Pore-waterirrigation.Pore-waterirrigation is Rateconstants for organic degradation canbe the exchange ofsolutechemistry withoverlying estimated inthe fieldfrom pore-waterdata(e.g., waterresultingfrom the action ofbenthic macro- high-resolution microelectrode oxygenprofiles fauna, whopump overlyingwaterthrough their (Hales etal .1993))orfrom laboratory incubations burrows withinthe sediment. Ratesofirrigation (for sulfaterespiration inparticular,Weber etal . canbe determined bycomparingtotaloxygen (2001)). Theserateconstants vary widely,with uptakerates,from benthic flux chambers,with valuesorders ofmagnitude higherinshelfand diffusiveuptakerates,frommicroelectrodes, slopethaninabyssalsediments.The rateconstants which areassumed to be independent ofany havebeenparametrized asafunction oftotal irrigation influence.Ingeneral,pore-waterirriga- sediment accumulation rate(Middelburg etal ., tion becomesnegligible, or atleast unmeasurable, 1997; Tromp etal .,1995),organic carbon rain inabyssalsediments (ArcherandDevol,1992; (Archer etal .,2001),or the meanage ofthe Archer etal .,2001; Reimers andSmith, 1986). sedimentary organic matter(Emerson,1985). Oxygen. Atraditionalviewofoxygendepen- Infact,thereisnot simplyonehomogeneous dence on organic carbon concentrations of kindoforganic matterinsediments,but rather sedimentswaschallengedinthe 1980s by modelingandfieldstudiesareableto resolveat laboratory measurements ofthe degradation of 2 2 least two typesoforganic matterwithdegradation freshphytoplankton usingeitherO 2 or SO4 asan kineticsdifferingbyseveralorders ofmagnitude electron acceptor (HenrichsandReeburgh, 1987). (Berner,1980). However,degradation rateconstantsderived from Bioturbation. Otherbiologically mediated rel- fielddataappeartodiverge astheyslow down in evant processesinclude bioturbation,the physical the deep(Tromp etal .,1995),to valuesthatdiffer mixingofsolid sediment,andpore-waterirriga- byfiveor sixorders ofmagnitude (Figure5). tion,anenhancement ofsolutetransport by Calvert andPederson (1993) compared organic benthic fauna.Bioturbation issoubiquitous that carbon concentrations betweenoxic andanoxic its absence insedimentary deposits (indicated by (BlackSea)settingsandconcluded thatoxygen finelaminae)istakenasstrongevidence ofanoxia plays littlepart inorganic carbon preservation. duringdeposition. Ingeneral,the rateofbioturba- However,Archer etal .(2001) summarized data tion scalessimilarly to thatoforganic carbon from awide variety oflocations,andfound respiration,increasinginshallowerwater. The dependence bothon oxygenconcentration and depthofbioturbation appears to be much less on organic carbon raintothe seafloor. variable(Boudreau,1994, 1998). Thebioturbation Protection byphysicaladsorption. Athird rateappears to decreasesmoothly withdepth viewpoint comesfrom the observation thatthe (Archer etal .,2001; MartinandSayles,1990),but organic matterconcentration ofsediment appears

Figure5 Model-tuned respiration rateconstants for various electron acceptors (O 2 ,NO 2 ,Mn,Fe, andS),andfor two fractions oforganic matterraintothe seafloor (fasterfraction shown for O 2 only). Dependence on organic carbon rainwasbased on datafrom TothandLehrman(1977),Tromp etal .(1995),Westrich andBerner(1984)tuned to reproduce sediment pore-waterandsolid phasefieldmeasurements (Archer etal.,2001). Controls ofMeanOceanChemistry 285 to correlateinversely withthe grainsizeofthe (Arakaki andMucci, 1995;BernerandMorse, sediment,insuch awayastomaintainaconstant 1974). The totalloss ofsolid-phaseCaCO3 is ratiooforganic matterconcentration to the surface determined asthe sum ofthesereactions,and area ofthe sediment grains (HedgesandKeil, thereforeexhibits acomplicated dependence on 1995). Keil etal .(1994)wereableto chemically the saturation stateofthe fluid.Thisbehavior has extractorganic matterfrom the surfacesofclays, beenparametrized ashigh-orderkinetic ratelaw andfoundthatits reactivity wasincreased ofthe form significantly bythisextraction. We notealso the R k 1 2 V n surprisingsuccess ofthe Mudssediment modelat ¼ ð Þ predictingorganic carbon concentration insedi- where k isarateconstant, V isthe saturation state ments,withno consideration oforganic adsorp- 2 2 2 Ca þ CO3 = K sp0 ,and n isapotentially tion effects atall. ðno¼½nunityŠ½ dissolutŠ ionÞ rateorder(Keir,1980; Globalsedimentary budget. Finally,on aglobal Morse, 1979). However,the numericalvalueof scale, wepoint to aninterestingmodelbyBerner the rateorder(the exponent n above)isdifficult,if andCanfield(1989),which tookasagivenan not impossible, to extractfrom fielddata(Archer organic carbon andreducedsulfur concentration etal .,1989a;HalesandEmerson,1996,1997a), ofsediments ofdifferent type(nearshoresands, andsensitiveto the imprecision ofthe saturation offshoreclays,abyssalclays,etc.). Inthisview,a statefor CaCO3 inlaboratory measurements netsource or sinkofO2 to the atmospherecould (HalesandEmerson,1997b). Ingeneral,ithas be drivenbyrearrangement ofthe sedimentary beenfoundthatthe rateconstant (the pre- mass ofthe Earth, from abyssalclays to near- exponentialfactor k above)variesbyorders of shoreclays,for example. magnitude from place to place inthe ocean(Hales andEmerson,1996,1997a),but less so if linear dissolution kineticsareused.We shouldalso note thatlaboratory determinations ofthe rateconstant 6.10.3.1.2 CaCO3 tendto exceedfielddeterminations bytwo to three Dissolution kinetics. Incontrast to the degra- orders ofmagnitude. dation oforganic matter,the dissolution ofCaCO 3 Pore-waterpH. The dissolution ofCaCO3 in insediments reliesonabiotic dissolution which sediments isalso complicated byinteraction with ought,inprinciple, to be moreconservativeacross organic matterdiagenesisvia the pHequilibrium regions ofthe ocean. Acomplication isthat chemistry ofseawater(reaction (1),above). The CaCO3 dissolution into seawateriscomprised of addition ofdissolved CO2 from organic carbon multiplechemicalreactions such as degradation shifts the equilibrium to the right, 2 2 decreasingthe concentration ofCO 3 inthe pore 2 2 2 2 . CO3 H þ Ca þ HCO3 waters. Figure6shows amodel-generated con- þ ! þ 2 2 2 centration profileofCO 2 ,HCO3 ,andCO 3 in and sediment porewaters abovethe saturation horizon 2 2 for CaCO3 (i.e.,wherethe [CO 3 ]ofthe . Caþ H CO . CO HO CaHCOþ þ 2 3 ! 3 þ 3 overlyingwaterishigherthanthe equilibrium

2 2 2 Figure6 ModelprofilesofCO 2 ,HCO3 ,andCO 3 insurface sediments. From the Mudsmodel(Archer etal .,2001). 286 BiologicalFluxesinthe OceanandAtmospheric p CO2

valuewithrespecttocalcite). Sediment chemistry 6.10.3.1.4Indirectatmospheric p CO2 models predictthatrespiration accounts for a signatureofthe biologicalpump significant dissolution flux from sediments,e.g., 20–40% ofthe calciteraintothe seafloormaybe The biologicalpumpalsohasanindirect mechanism ofaffectingthe p ofthe atmos- dissolving, evenwherethe overlyingwateris CO2 saturated withrespecttocalcite(Archer,1991; phere, called CaCO3 compensation. The mechan- Emerson andArcher,1990). ism isbased on the constraint thatonlongenough The implication ofarolefor respiration asa timescales,the inputsandoutputs ofalkalinity to the oceanmust balance.The input arisesfrom driverfor CaCO3 dissolution isthatany hypo- theticalchange inthe ratiooforganic carbon to weatheringofcarbonateandsilicaterocksonland, calciterainingto the seafloormayimpactthe andremovalisbythe production andsedimen- globalburialrateofcalcite, andthus (bya tation ofCaCO3 .The crucialcomponent ofthis mechanism to be explained below) the steady- mechanism isthatthe fraction ofCaCO3 produced statepHofthe ocean(ArcherandMaier-Reimer, thatisburied dependsonthe acidity ofthe ocean. 1994). If input ofalkalinity exceedsoutput,e.g.,the buildup ofalkalinity inthe oceanpulls the pHto highervalues,allowingagreaterfraction ofthe CaCO3 producedto be buried (BroeckerandPeng, 6.10.3.1.3 SiO2 1982,1987).The systemapproachesequilibrium on atimescaleofroughly 5,000 years (Archer Onthe face ofit,the dissolution ofSiO2 from sediments oughttobe simplerthanthatoforganic etal .,1997). Becausethe mechanism relieson changingthe meanchemistry ofthe oceanby matter,which isdrivenbiologically,or CaCO3 , withits couplingto pore-waterpHandorganic interaction withthe sediments,thismechanism carbon respiration. The dissolution reaction for hasalso beencalled the “opensystem” response (andthe directeffectdescribed abovehas SiO2 issimply correspondingly beencalled the closed system SiO H O H SiO aq response(Sigman etal .,1998)). The linkto 2 þ 2 ! 4 4 ð Þ atmospheric p CO arisesfrom the dependence of Although H SiO (aq) couldbeaffected bythe 2 4 4 the dissolved CO2 concentration on seawaterpH. dissociation reaction AspHrises,less andless ofthe dissolved carbon 2 isfoundinthe acidform CO2 . H 4 SiO4 H þ H 3 SiO4 $ þ . CaCO3 compensation provides severalways inwhich the p ofthe atmosphere the equilibrium constant for thisreaction makesit CO2 insignificant inseawater. couldbeaffected bythe biologicalpump. Most However,opaldiagenesisinsediments,despite obviously,ifthe delivery rateofalkalinity to the its apparent simplicity,isnot withoutmystery. oceanfrom weatheringwereto increase, thena The asymptotic concentration ofsilicic acidin greaterfraction ofCaCO3 production wouldhave porewaters incontactwithmeasurablesolid- to be buried to balance weathering.Thiswould driveoceanpHup,decreasingatmospheric p phaseopaloughttoreflectthe solubility ofopal. CO2 Instead,whatisobserved aregenerally lower (Berger,1982; Opdyke andWalker,1992). One concentrations thanthe laboratory-measured equi- mechanism for increasingweatheringduring librium values,withatendencytowardcorrelation glacialtimeisthe exposureofshallow-water withthe productivity ofthe surfaceocean. reefsandcarbonatebanks,which thereforetrans- Asymptotic inSouthern Oceansediments form from significant depocenters ofCaCO3 from reaches600 m Mandup,whileinthe subtropical the ocean(Kleypas,1997; Milliman,1993)to oligotrophic oceanasymptotic silicon isonly significant weatheringsources(GibbsandKump, 150–300 m M(Schink etal .,1974). Archer etal . 1994;MunhovenandFrancois,1993). This (1993) argued thatdynamically,withthe frame- mechanism musthaveplayed someroleinglacial workofthe diffusivepore-watersystem,the p CO2 variations,although the timingofevents at equilibrium silicon hasagreaterimpactonopal the deglacialtion precludesaprimary sea-level preservation thanthe rateconstant for dissolution. driverof p CO2 (BroeckerandHenderson,1998). Recent workhasfoundacorrelation betweenthe Rainratio. Somewhatless obviously,a asymptotic (equilibrium)silicon concentration decreaseinthe production ofCaCO 3 would andthe concentration ofaluminuminthe mandateanincreaseinthe burialfraction that sediment solid phase(CappellenandQiu,1997; mustbe maintained inordertoachievethroughput Dixit etal .,2001),rationalizingthe decreasing balance.Third, CaCO3 dissolution insediments is solubility asafunction ofthe formation of promoted bythe oxic degradation oforganic aluminosilicatesonthe opalsurface.Similar matter(Emerson andBender,1981).Therefore, an behavior hasbeendocumented inthe laboratory increaseinorganic matterraintothe seafloor (Lewin,1961). couldultimately drivepHtowardthe basic to References 287

the :regionalexport patterns andpotentialfor compensate, decreasingatmospheric p CO2 .Both the directandthe indirecteffects oforganic atmospheric CO2 sequestration. GlobalBiogeochem. Cycles 15 ,845–862. carbon andCaCO 3 raindependon the ratioof Arakaki T.andMucci A.(1995)Acontinuous andmechanistic organic carbon to CaCO3 produced.Thisratio representation ofcalcitereaction-controlled kineticsin thereforehasaname, the “rainratio.”The effects dilutesolutions at25 8 Cat1atm totalpressure. Aquat. oforganic carbon andCaCO 3 on CaCO3 com- Chem. 1 ,105–130. pensation arecollectively called the rain-ratio ArcherD.E.(1991) Modelingthe calcitelysocline. J.Geophys. Res. 96(C9),17037–17050. hypothesis(ArcherandMaier-Reimer,1994). ArcherD.E.(1996a)Anatlasofthe distribution ofcalcium carbonateinsediments ofthe deepsea. GlobalBiogeochem. Cycles 10,159–174. 6.10.4SUMMARY ArcherD.E.(1996b)Adata-drivenmodelofthe globalcalcite lysocline. GlobalBiogeochem. Cycles 10,511–526. Insummary,someadvancesinthe biogeo- ArcherD.E.andDevol A.(1992) Benthic oxygenfluxesonthe chemicalcyclesinthe oceanhavesuggested Washingtonshelfandslope: acomparison of insitu microelectrode andchamberflux measurements. Limnol. severalnewsensitivities,ways inwhich the past or Oceanogr. 37,614–629. futureclimatechange couldimpactthe p CO2 ArcherD.E.andJohnson K.(2000) Amodelofthe iron cyclein concentration ofthe atmosphere.Someofthese the ocean. GlobalBiogeochem. Cycles 14 ,269–279. mechanismsactonthe 103 -yeartimescaleof ArcherD.E.andMaier-ReimerE.(1994)Effectofdeep-sea sedimentary calcitepreservation on atmospheric CO2 ocean/atmosphere p CO equilibration,such asthe 2 concentration. Nature 367,260–264. effectofiron on the biologicalpump. Thiseffectis ArcherD.,Emerson S.,andReimers C.(1989a)Dissolution of minimized if (Section 6.10.2.1.2)iron ispre- calciteindeep-sea sediments:pHandO2 microelectrode dominantly recycled withinthe oceanratherthan results. Geochim. Cosmochim. Acta 53 ,2831–2846. supplied bydeposition ofcontinentaldust,and ArcherD.,Emerson S.,andSmithC.R.(1989b)Direct measurement ofthe diffusivesublayeratthe deepsea floor (Section 6.10.2.5.1)if the realoceanbehavesasa usingoxygenmicroelectrodes. Nature 340 ,623–626. generalcirculation modelratherthanasabox ArcherD.,LyleM.,Rodgers K.,andFroelich P.(1993) What model,inthe waythatthe high-latitude surface controls opalpreservation intropicaldeep-sea sediments. Paleoceanography 8 ,7–21. oceaninteracts withthe p CO2 ofthe atmosphere. Otherpotentialsensitivitiesinclude adecrease ArcherD.,Kheshgi H.,andMaier-RiemerE.(1997) Multiple timescalesfor neutralization offossilfuelCO2 . Geophys. incalcification withacidity ofsurface waters Res. Lett. 24 ,405–408. (Section 6.10.2.2.2)andanincreaseinwater ArcherD.,EshelG.,WinguthA.,BroeckerW.S., columnredissolution ofCaCO 3 (Section Pierrehumbert R.T.,TobisM.,andJacobR.(2000a) Atmospheric p sensitivity to the biologicalpump inthe 6.10.2.4.2). Alternatively,CaCO 3 production CO2 mayincreasewithanincreaseinthe stratification ocean. GlobalBiogeochem. Cycles 14 ,1219–1230. ArcherD.E.,WinguthA.,Lea D.,andMahowaldN.(2000b) ofthe watercolumn inafutureglobalwarming Whatcaused the glacial/interglacialatmospheric p CO2 world(Section 6.10.2.3;Antia etal .,2001),or cycles? Rev. Geophys. 38 ,159–189. perturbations indust deposition dueto changesin ArcherD.E.,MorfordJ.L.,andEmerson S.(2001) Amodelof landuse(Sections 6.10.2.1.2 and6.10.2.3). suboxic sedimentary diagenesissuitablefor automatic OntimescalesofCaCO compensation (Section tuningandgridded globaldomains. GlobalBiogeochem. 3 Cycles (inpress). 6.10.3.1.4),any changesincalcification rate, or ArmstrongR.A.,Lee C.,HedgesJ.I.,HonjoS.,andWakeham the ratioofcalciteto organic carbon reachingthe S.G.(2002) Anewmechanistic modelfor organic carbon fluxesinthe oceanbased on the quantitativeassociation of seafloor,haveanamplified effectonthe p CO2 of the atmosphere.Thiseffectmaybe minimized if POC withballast minerals. Deep-Sea Res. II 49 ,219–236. organic matterrequiresballasting, predominately Bacastow R.B.(1996) The effectoftemperaturechange ofthe warm surface waters ofthe on atmospheric CO2 . fromCaCO3 ,to reachthe seafloor (Section GlobalBiogeochem. Cycles 10,319–334. 6.10.2.4.4). Theimpactofirondeposition on Balch W.M.,DrapeauD.T.,andFritz J.J.(2000) Monsoonal Si :C ratios ofdiatoms mayaffectatmospheric forcingofcalcification inthe ArabianSea. Deep-Sea Res. II 47 ,1301–1337. p CO (Matsumoto etal .,inpress),ifanincreasein 2 BarkerS.andElderfieldH.(2002) Foraminiferalcalcification silicon availability depressescalcification,an responseto glacial–interglacialchangesinatmospheric effectwhich isnot clearlyseeninsediment trap CO2 . Science 297,833–836. data(KlaasandArcher,2002). Balch W.M.andKilpatrick K.(1996) Calcification ratesinthe equatorialPacific along140 8 W. Deep-Sea Res. II 43 (4–6), 971–994. BergerW.H.(1982) DeglacialCO2 buildup:constraints on the REFERENCES coralreef model. Palaeogeogr. Palaeoclimatol. Palaeoecol. 40 ,235–253. AbrahamE.R.,LawC.L.,BoydP.W.,LavenderS.J., BergerW.H.(1989) Globalmaps ofoceanproductivity. In MaldonadoM.T.,andBowleW.R.(2000) Importance of Productivity ofthe Ocean:Present andPast (eds. S.V.S. stirringinthe development ofaniron-fertilized phytoplank- W.H.Berger,andG.Wefer). Wiley,pp. 429–453. ton bloom. Nature 407,727–703. BernerR.A.(1980) Aratemodelfor organic matter Antia A.N.,KoeveW.,FischerG.,Blanz T.,Schulz-BullD., decomposition duringbacterialsulfatereduction inmarine ScholtenJ.,NeuerS.,KremlingK.,Kuss J.,Peinert R., sediments. In Biogeochemistry ofOrganic Matteratthe Hebbeln D.,Bathmann U.,ConteM.,FehnerU.,and Sediment–WaterInterface.Comm. Natl. RechercheScien- ZeitzschelB.(2001) Basin-wide particulatecarbon flux in tifique, France. 288 BiologicalFluxesinthe OceanandAtmospheric p CO2 BernerR.A.andCanfieldD.E.(1989) Anewmodelfor DixitS.,CappellenP.V.,andBennekom A.J.V.(2001) atmospheric oxygenoverPhanerozoic time. Am. J.Sci. 289 , Processescontrollingsolubility ofbiogenic silica andpore 333–361. waterbuild-up ofsilicic acid inmarinesediments. Mar. BernerR.A.andMorseJ.W.(1974)Dissolution kineticsof Chem. 73,333–352. calcium carbonateinsea waterIV.Theory ofcalcite Duce R.A.andTindaleN.W.(1991) Atmospheric transport of dissolution. Am. J.Sci. 274 ,107–134. iron andits deposition inthe ocean. Limnol. Oceanogr. 36, BiscayeP.E.,Flagg C.N.,andFalkowski P.G.(1994)The 1715–1726. shelfedge exchange processesexperiment,SEEP-II: an DugdaleR.C.andGoeringJ.J.(1967) Uptake ofnewand introduction to hypotheses,results andconclusions. Deep- regenerated forms ofnitrogeninprimary productivity. Sea Res. II 41 ,231–252. Limnol. Oceanogr. 12,196–207. BoudreauB.P.(1994)Isburialvelocity amasterparameter DymondJ.andCollierR.(1988)Biogenic particlefluxesin for bioturbation? Geochim. Cosmochim. Acta 58, the EquatorialPacific: evidence for bothhigh andlow 1243–1249. productivity duringthe 1982–1983ElNino. Global BoudreauB.P.(1998)Meanmixed depthofsediments: Biogeochem. Cycles 2 ,129–137. the whereforeandthe why. Limnol. Oceanogr. 43 , DymondJ.andLyleM.(1993) Particlefluxesinthe oceanand 524–526. implicationsfor sourcesandpreservation ofoceansediments. BoydP.W.,Watson A.J.,LawC.S.,AbrahamE.R.,Trull T., In GeomaterialFluxes,GlacialtoRecent (eds. W.Heyand Murdoch R.,BakkerD.C.E.,Bowie A.R.,BuesselerK.O., T.Usselman). NationalAcademy ofSciences,Washington, ChangH.,CharetteM.,Croot P.,DowningK.,FrewR.,Gall DC. M.,HadfieldM.,Hall J.,HarveyM.,Jameson G.,LaRoche Emerson S.(1985)Organic carbon preservation inmarine J.,LiddicoatM.,LingR.,MaldonadoM.T.,McKayR.M., NodderS.,PickmereS.,PridmoreR.,Rintous S.,Safi K., sediments. In The Carbon CycleandAtmospheric CO2 : Sutton P.,Strzepek R.,TannabergerK.,TurnerS.,WaiteA., NaturalVariations ArcheantoPresent (ed.B.Sundquist). andZeldisJ.(2000) Amesoscalephytoplankton bloom in AmericanGeophysicalUnion,Washington,DC. the polarSouthern Oceanstimulated byiron fertilization. Emerson S.andBenderM.L.(1981) Carbon fluxesatthe Nature 407,695–702. sediment–waterinterface ofthe deepsea: calcium BrewerP.G.,Nozaki Y.,SpencerD.W.,andFleerA.P.(1980) preservation. J.Mar. Res. 39,139–162. Sediment trapexperimentsinthe deepNorthAtlantic: Emerson S.andArcherD.(1990) isotopic andelementalfluxes. 38 (4),703–728. preservation inthe ocean. Phil. Trans. Roy. Soc.London A BroeckerW.S.andClarkE.(2001)Anevaluation of 331,29–41. Lohmann’s foraminiferaweightdissolution index. Paleo- EppleyR.W.andPeterson B.J.(1979) Particulateorganic ceanography 16,531–534. matterflux andplanktonic newproduction inthe deepocean. BroeckerW.S.andHenderson G.(1998)The sequence of Nature 282 ,677–680. event surroundingtermination II andtheirimplicationsfor Feely R.A.,SabineC.L.,Lee K.,Millero F.J.,LambM.F., the causeofglacial–interglacialCO2 changes. Paleoceano- GreeleyD.,BullisterJ.L.,KeyR.M.,PengT.-H.,Kozyr A., graphy 13,352–364. Ono T.,andWongC.S. Insitu calcium carbonate BroeckerW.S.andPengT.H.(1982) Tracers inthe Sea. dissolution inthe Pacific Ocean. GlobalBiogeochem. Cycles EldigioPress,Palisades,NY. (submitted for publication). BroeckerW.S.andPengT.H.(1987) The roleofCaCO3 Fiadeiro M.(1980) The alkalinity ofthe deepPacific. Earth compensation inthe glacialtointerglacialatmospheric CO2 . Sci.Lett. 49 ,499–505. change. GlobalBiogeochem. Cycles 1 ,15–29. FornesW.L.,DeMasterD.J.,LevinL.A.,andBlairN.E. BroeckerW.,Lynch-Steiglitz J.,ArcherD.,Hofmann M., (1999) Bioturbation andparticletransport inCalifornia slope Maier-ReimerE.,MarchalO.,StockerT.,andGruberN. sediments:aradiochemicalapproach. J.Mar. Res. 57 , (1999) How strongisthe Harvardton–Bearconstraint? 335–355. GlobalBiogeochem. Cycles 13,817–821. Franck V.M.,Brzezinski M.A.,CoaleK.H.,andNelson D.M. Brown C.W.andYoderJ.A.(1994)Coccolithophorid blooms (2000) Iron andsilicic acid concentrations regulateSiuptake inthe globalocean. J.Geophys. Res. 99,7467–7482. northandsouthofthe PolarFrontalZoneinthe Pacific BuesselerK.O.(1991) Doupper-oceansediment traps provide Sectorofthe Southern Ocean. Deep-Sea Res. II 47 , anaccuraterecordofparticleflux? Nature 353 ,420–423. 3315–3338. Calvert S.E.andPadersenT.F.(1993) ofrecent FungI.Y.,Mayn S.K.,TegenI.,DoneyS.C.,JohnJ.G.,and oxic andanoxic marinesediments:implicationsfor the Bishop J.K.B.(2000) Iron supply anddemandinthe upper geologicalrecord. Mar. Geol. 113,67–88. ocean. GlobalBiogeochem. Cycles 14 ,281–295. CappellenP.V.andIngall E.(1996) Redox stabilization ofthe GibbsM.andKump L.R.(1994)Globalchemicalerosion atmosphereandoceans byphosphorus-limited marine duringthe last glacialmaximum andthe present:sensitivity productivity. Science 271,493–495. or changesinlithologyandhydrology. Paleoceanography 9 , CappellenP.V.andQiuL.(1997) Biogenic silica dissolution 529–543. insediments ofthe Southern Ocean:I.Solubility. Deep-Sea GnanadesikanA.(1999) Aglobalmodelofsilica cycling: Res. II 44,1109–1128. CoaleK.H.,Johnson K.S.,FitzwaterS.E.,Gordon R.M., sensitivity to eddyparameterization andremineralization. TannerS.,ChavezF.P.,FerioliL.,Sakamoto C.,Rogers P., GlobalBiogeochem. Cycles 13,199–220. Millero F.,SteinbergP.,NightingaleP.,CooperD.,Cochlan GoldmanJ.C.(1987) Spatialandtemporaldiscontinuities W.P.,Landry M.R.,Constantinou J.,RollwagenG., ofbiologicalprocessesinpelagic surface waters.In TrasvinaA.,andKudelaR.(1996) Amassivephytoplankton TowardsaTheory on BiologicalandPhysicalProcesses bloom induced byanecosystem-scaleiron fertilization inthe WorldOcean (ed.R.B.J.). Reidal,Dordrecht,The experiment inthe equatorialPacific Ocean. Nature 383 , Netherlands. 495–501. HalesB.andEmerson S.(1996) Calcitedissolution insediment DeuserW.G.(1986) Seasonalandinterannualvariations in ofthe Ontong-JavaPlateau: insitu measurementsof deep-waterparticlefluxesinthe SargassoSeaand porewaterO 2 andpH. GlobalBiogeochem. Cycles 10, theirrelation to surface hydrography. Deep-Sea Res. 33, 527–541. 225–246. HalesB.andEmerson S.(1997a)Calcitedissolution in Devol A.H.(1978)Bacterialoxygenuptake kineticsasrelated sedimentsofthe CearaRise: insitu measurementsof to biologicalprocessesinoxygendeficient zonesofthe porewaterO 2 ,pH, andCO 2 (aq). Geochim. Cosmochim. oceans. Deep-Sea Res. 25 ,137–146. Acta 61,501–514. References 289

HalesB.andEmerson S.(1997b)Evidence insupport offirst- deepsea: implicationsfor the rainratio. GlobalBiogeochem. orderdissolution kineticsofcalciteinseawater. Earth Cycles 16,doi:10.1029/2001GB001765. Planet. Sci.Lett. 148,317–327. KleypasJ.A.(1997) Modeled estimatesofglobalreef habitat HalesB.,Emerson S.,andArcherD.(1993) Respiration and andcarbonateproduction since the last glacialmaximum. dissolution inthe sediments ofthe western NorthAtlantic: Paleoceanography 12,533–545. estimatesfrom models of insitu microelectrode measure- Knox F.andMcElroy M.(1984)Change inatmospheric CO2 : ments ofporewateroxygenandpH. Deep-Sea Res. 41 , influence ofthe marinebiotaathigh latitude. J.Geophys. 695–719. Res. 89 ,4629–4637. HedgesJ.I.andKeilR.G.(1995)Sedimentary organic matter Langdon C.,Takahashi T.,SweeneyC.,ChipmanD.,Goddard preservation—anassessment andspeculativesynthesis. J.,MarubiniF.,AcevesH.,Barnett H.,andAtkinson M.J. Mar. Chem. 49 ,81–115. (2000) Effectofcalcium carbonatesaturation stateon the HedgesJ.I.,Lee C.,WakehamS.G.,HernesP.J.,andPeterson calcification rateofanexperimentalreef. GlobalBiogeo- M.L.(1993) Effects ofpoisons andpreservativesonthe chem. Cycles 14 ,639–654. fluxesandelementalcomposition ofsediment trapmaterials. Laws E.A.,Falkowski P.G.,W.O.S, Jr.,Ducklow H.,and J.Mar. Res. 51 ,651–668. McCarthyJ.J.(2000)Temperatureeffects on export HeinzeC.,Maier-ReimerE.,WinguthA.M.E.,andArcherD. production inthe openocean. GlobalBiogeochem. Cycles (1999) Aglobaloceanic sediment modelfor longterm 14 ,1231–1246. climatestudies. GlobalBiogeochem. Cycles 13,221–250. LefevreN.andWatson A.(1999) Modelingthe geochemical HenrichsS.M.andReeburgh W.S.(1987) Anaerobic cycleofiron inthe oceans andits impactonatmospheric mineralization ofmarinesediment organic matter:rates CO2 concentrations. GlobalBiogeochem. Cycles 13, andthe roleofanaerobic processesinthe oceanic carbon 727–736. economy. GeomicrobiologyJ. 5 ,191–237. Levitus S.(1982) ClimatologicalAtlasofthe WorldOcean .US HollandH.D.(1978) The Chemistry ofthe Atmosphereand Government PrintingOffice, Washington,DC, 173pp. Oceans,351pp. Wiley,NewYork. Levitus S.,ConkrightM.E.,Reid J.L.,NajjarR.G.,and HonjoS.(1976) Coccoliths:production,transportation,and MantylaA.(1993) Distribution ofnitrate, , and . Mar. Micropaleo. 1 ,65–79. silicateinthe world’s oceans. Prog.Oceanogr. 31, HonjoS.(1990) Particlefluxesandmodern sedimentation inthe 245–273. polaroceans. In PolarOceanography:Part B.Chemistry, LewinJ.C.(1961) The dissolution ofsilica from diatom walls. Biology,andGeology .Academic Press,pp. 687–739. Geochim. Cosmochim. Acta 21,182–198. HonjoS.(1996) Fluxesofparticlestothe interior ofthe open Li Y.-H.,Takahashi T.,andBroeckerW.S.(1969) Degree of oceans. In ParticleFlux inthe Ocean (eds. V.Ittekkot,P. saturation ofCaCO3 inthe Oceans. J.Geophys. Res. 74 , Aschauffer,S.Honjo,andP.Depetris). Wiley,NewYork. 5507. HonjoS.,ManganiniS.J.,andColeJ.J.(1982) Sedimentation LochteK.,Ducklow H.W.,FashamM.J.R.,andStienenC. ofbiogenic matterinthe deepocean. Deep-Sea Res. 29(5A), (1993) Plankton succession andcarbon cyclingat47 8 N 609–625. 20 8 Wduringthe JGOFS NorthAtlantic Bloom Experiment. HonjoS.,ManganiniS.J.,andWeferG.(1998)Annualparticle Deep-Sea Res. II 40 ,91–114. flux andawinteroutburst ofsedimentation inthe northern Lohmann G.P.(1995)Amodelfor variation inthe chemistry NorwegianSea. Deep-Sea Res. 8 ,1223–1234. ofplanktonic foraminiferadueto secondary HonjoS.,FrancoisR.,ManganiniS.,DymondJ.,andCollierR. calcification andselectivedissolution. Paleoceanography (2000) Particlefluxestothe interior ofthe Southern Oceanin 10,445–458. the Western Pacific sector along170 W. Deep-Sea Res. II 47 , Lutz M.,DunbarR.,andCaldeiraK.Regionalvariability inthe 3521–3548. verticalflux ofparticulateorganic carbon inthe ocean Hutchins D.A.andBrulandK.W.(1998)Iron-limited diatom interior. GlobalBiogeochem. Cycles (inpress). growthandSi:Nuptake ratios incoastalupwellingregime. MahadevanA.andArcherD.(2000) Effectofmesoscale Nature 393,561–564. circulation on nutrient supply to the upperocean. J.Geophys. Ittekkot V.,NairR.R.,andHonjoS.(1991) Enhanced particle Res. 105 ,1209–1225. fluxesinBayofBengalinduced byinjection offreshwater. Margalef R.(1978)Life-forms ofphytoplankton assurvival Nature 351 ,385–387. alternativesinanunstableenvironment. Oceanol. Acta 1 , Jahnke R.A.,Reimers C.E.,andCravenD.B.(1990) 493–509. Intensification ofrecyclingoforganic matteratthe seafloor MartinJ.H.andFitzwaterS.E.(1988)Iron deficiencylimits nearoceanmargins. Nature 348,50–54. photoplankton growthinthe north-east Pacific subarctic. JansenH.andWolf-GladrowD.A.(2001) Carbonate Nature 331,341–343. dissolution incopepodguts:anumericalmodel. Mar. MartinJ.H.,KnauerG.A.,Karl D.M.,andBroenkow W.M. Ecol. Prog.Ser. 221,199–207. (1987) VERTEX: carbon cyclinginthe northeast Pacific. Jenkins W.J.(1988)Nitrateflux into the euphotic zonenear Deep-Sea Res. 34 (2),267–285. Bermuda. Nature 331,521–524. MartinW.R.andSaylesF.L.(1990) Seafloor diagenetic Johnson K.S.,ChavezF.P.,ElrodV.A.,FitzwaterS.E., fluxes,WoodsHoleOceanographicInstitution. PenningtonJ.T.,Buck K.R.,andWalz P.M.(2001) The MartinJ.,CoaleK.H.,Johnson K.S.,FitzwaterS.E.,Gordon annualcycleofiron andthe biologicalresponseincentral R.M.,TannerS.J.,HunterC.N.,ElrodV.E.,Nowicki J.L., California coastalwaters. Geophys.Res. Lett. 28 , ColeyT.L.,BarberR.T.,LindleyS.,Watson A.J.,Scoy 1247–1250. K.V.,LawC.S.,LiddicoatM.I.,LingR.,Stanton T., Karl D.,LetelierR.,TupasL.,DoreJ.,ChristianJ.,andHebel StockelJ.,Collins C.,Anderson A.,BidigareR.,Ondrusek D.(1997) The roleofnitrogenfixation ingeoceochemical M.,LatasaM.,Millero F.J.,Lee K.,YaoW.,ZhangJ.Z., cyclinginthe subtropicalNorthPacific Ocean. Nature 388, Friederich G.,Sakamoto C.,ChavezF.,Buck K.,KolberZ., 533–538. GreeneR.,Falkowski P.,Chisolm S.W.,Hoge F.,SwiftR., KeilR.G.,Montlucon D.B.,PrahlF.G.,andHedgesJ.I. YungelJ.,TurnerS.,NightingaleP.,Hatton A.,Liss P.,and (1994)Sorptivepreservation oflabileorganic matterin TindaleN.W.(1994)Testingthe ironhypothesisin marinesediments. Nature 370,549–552. ecosystems ofthe equatorialPacific Ocean. Nature 371, KeirR.S.(1980) The dissolution kineticsofbiogenic calcium 123–129. carbonatesinseawater. Geochim. Cosmochim. Acta 44, MatsumotoK.,SarmientoJ.L.,andBrzezinski M.A.Silicic 241–252. acid “leakage”from the Southern Oceanasapossible

KlaasC.andArcherD.E.(2002) Association ofsinking mechanism for explainingglacialatmospheric p CO2 . Global organic matterwithvarious typesofmineralballast inthe Biogeochem. Cycles (inpress). 290 BiologicalFluxesinthe OceanandAtmospheric p CO2 McGillicuddyD.J.andRobinson A.R.(1997) Eddyinduced Riebesell U.,ZondervanI.,Rost B.,Tortell P.D.,Zeebe R.E., nutrient supply andnewproduction inthe Sargasso Sea. andMorelF.M.M.(2000) Reduced calcification ofmarine Deep-Sea Res. I 44,1427–1450. plankton inresponseto increased atmospheric CO2 . Nature MiddelburgJ.J.,Soetaert K.,andHermanP.M.J.(1997) 407,364–367. Empiricalrelationsfor useinglobaldiagenetic models. Sarmiento J.L.andToggweilerR.(1984)Anewmodelfor the Deep-Sea Res. I 44,327–344. roleofthe oceans indeterminingatmospheric p CO2 . Nature Millero F.J.(1982) The effectofpressureon the solubility of 308 ,621–624. minerals inwaterandseawater. Geochim. Cosmochim. Acta Sarmiento J.L.,DunneJ.,GnanadesikanA.,KeyR.M., 46 ,11–22. MatsumotoK.,andSlaterR.Anewestimateofthe CaCO3 to MillimanJ.D.(1974) MarineCarbonates .Springer,Heidel- organic carbon export ratio. GlobalBiogeochem. Cycles berg, 375pp. (submitted for publication). MillimanJ.D.(1993) Production andaccumulation ofcalcium SchinkD.R.,FanningK.A.,andPilson M.E.Q.(1974) carbonateinthe ocean:budgetofanon-steadystate. Global Dissolved silica inthe upperporewaters ofthe Atlantic Biogeochem. Cycles 7 ,927–957. Oceanfloor. JGR 79(15),2243–2250. MillimanJ.D.,Troy P.J.,Balch W.M.,Adams A.K.,Li SiegenthalerU.andWenkT.(1984)Rapid atmospheric Y.-H.,andMackenzie F.T.(2000) Biologically mediated CO2 variationsandoceancirculation. Nature 308 , dissolution ofcalcium carbonateabovethe chemical 624–626. lysocline? Deep-Sea Res. I 46 ,1653–1699. SigmanD.M.,McCorkleD.C.,andMartinW.R.(1998)The MorseJ.W.(1979)Thekineticsofcalcium carbonate calcitelysoclineasaconstraint on glacial/interglaciallow dissolution andprecipitation,227. latitude production changes. GlobalBiogeochem. Cycles 12, Mucci A.(1983) The solubility ofcalciteandaragonitein 409–428. seawateratvarioussalinities,temperatures,andone SmithC.R.,PopeR.H.,DeMasterD.J.,andMagaardL. atmospheretotalpressure. Am. J.Sci. 283 ,780–799. (1993) Age-dependent mixingofdeepsea sediments. MunhovenG.andFrancoisL.M.(1993) Glacial–interglacial Geochim. Cosmochim. Acta 57 ,1473–1488. changesincontinentalweathering: possibleimplications for Takeda S.(1998)Influence ofiron availability on nutrient atmospheric CO2 .In Carbon Cyclinginthe GlacialOcean: consumption ratioofdiatoms inoceanic waters. Nature 393, Constraints on the Ocean’s RoleinGlobalChange (ed.R. 774–777. Zahn). Springer,Berlin. ToggweilerJ.R.,GnanadesikanA.,Carson S.,MurnaneR.,and MunkW.andWunsch C.(1998)Abyssalrecipes:II. Sarmiento J.L.(2003) Representation ofthe carbon cyclein Energeticsoftidalandwindmixing. Deep-Sea Res. I 45, box models andGCMs:1. Solubility pump. GlobalBiogeo- 1977–2010. chem. Cycles 17(1),1026,doi: 10.1029/2001GB001401. Nelson D.M.,TreguerP.,Brzezinski M.A.,Laynaert A., TothD.J.andLehrmanA.(1977) Organic matterreactivity andQueguinerB.(1995)Production anddissolution of andsedimentation ratesinthe ocean. Am. J.Sci. 277, biogenic silica inthe ocean:revised globalestimates, 465–485. comparisonwithregionaldataandrelationshipto Tromp T.K.,CappellenP.V.,andKeyR.M.(1995)Aglobal biogenic sedimentation. GlobalBiogeochem. Cycles 9 , modelfor the early diagenesisoforganic carbon andorganic 359–372. phosphorusinmarinesediments. Geochim. Cosmochim. Noriki S.andTsunogai S.(1986) Particulatefluxesand Acta 59 ,1259–1284. major componentsofsettlingparticlesfrom TsunogaiS.andNorikiS.(1991)Particulatefluxesof sediment trapexperimentsinthe Pacific Ocean. 33(7), carbonateandorganic carbon inthe ocean. Isthe marine 903–912. biologicalactivity workingasasinkofthe atmospheric Opdyke B.N.andWalkerJ.C.G.(1992) Return ofthe coralreef hypothesis:basintoshelfpartitioningof carbon? Tellus 43B ,256–266. VillarealT.V.,PilskainC.,Brzezinski M.,LipschultzF., CaCO3 andits effectonatmospheric p CO2 . 20, 733–736. DennetM.,andGardnerG.B.(1999) Upwardtransport of Orr J.C.,Maier-ReimerE.,andMikolajewiczU.(2001) oceanic nitratebymigratingdiatom mats. Nature 397, Estimatesofanthropogenic carbon uptake from four three- 423–425. dimensionalglobaloceanmodels. GlobalBiogeochem. WalkerJ.C.G.(1977) Evolution ofthe Atmosphere . Cycles 15 ,43–60. Macmillan,NewYork, 318pp. OschliesA.andKoeveW.(2000) Aneddy-premitting WalkerJ.C.G.andKastingJ.F.(1992) Effects offueland coupled physical–biologicalmodelofthe NorthAtlantic forest conservation on futurelevels ofatmospheric carbon 2. Ecosystemdynamicsandcomparisonwithsatelliteand dioxide. Palaeogeogr. Palaeoclimat. Palaeoecol. (Global JGOFS localstudiesdata. GlobalBiogeochem. Cycles 14 , andPlanetary Change Section) 97,151–189. 499–523. WeberA.,Riess W.,WenzhoeferF.,andJorgensenB.B. Peixoto J.P.andOort A.H.(1992) PhysicsofClimate . (2001) Sulfatereduction inBlack Seasediments: AmericanInstituteofPhysics,NewYork, 520pp. insitu andlaboratory radiotracermeasurementsfrom PetitJ.R.,JouzelJ.,RaynaudD.,Barkov N.I.,BarnolaJ.-M., the shelfto 2,000 mdepth. Deep-Sea Res. I 48, BasileI.,BenderM.,ChappellazJ.,DavisM.,DelaygueG., 2073–2096. DelmotteM.,Kotlyakov V.M.,LegrandM.,Lipenkov V.Y., WeferG.,Suess E.,BalzerW.,LiebezeitG.,MullerP.J., Lorius C.,PepinL.,Ritz C.,SaltzmanE.,andStievenardM. UngererC.A.,andZenkW.(1982) Fluxesof (1999) Climateandatmospheric history ofthe past 420,000 biogenic componentsfromsediment trapdeployment years from the Vostokice core, Antarctica. Nature 399, incircumpolarwaters ofthe Drake Passage. Nature 299, 429–436. 145–147. Pilskaln C.A.andHonjoS.(1987) The fecalpelletfraction of WeferG.,FischerG.,FuettererD.,andGersonde R.(1988) biogeochemicalparticlefluxestothe deepsea. Global Seasonalparticleflux inthe BransfieldStrait,Antarctica. Biogeochem. Cycles 1 (1),31–48. Deep-Sea Res. 35 ,891–898. PlathD.C.,Johnson K.S.,andPytkowiczR.M.(1980) The WeferG.,FischerR.,FuttererD.K.,Gersonde R.,HonjoS., solubility ofcalcite—probably containingmagnesium—in andOstermann D.(1990) Particlesedimentation and seawater. Mar. Chem. 10,9–29. productivity inAntarctic waters ofthe Atlantic sector. Reimers C.E.andSmithK.L.J.(1986) Reconcilingmeasured 363–379. andpredicted fluxesofoxygenacross the deep Westrich J.T.andBernerR.A.(1984)The roleofsedimentary sea sediment–waterinterface. Limnol.Oceanogr. 31, organic matterinbacterialsulfatereduction:the Gmodel 305–318. tested. Limnol. Oceanogr. 29,236–249. References 291

Yamanaka Y.andTujika E.(1997) Roleofdissolved YuE.-F.,FrancoisR.,Bacon M.P.,HonjoS.,FleerA.P., organic matterinthe marinebiogeochemicalcycle: ManganiniS.J.,LoeffM.M.R.v.d.,andIttekot V.(2001) studiesusinganoceanbiogeochemicalgeneral Trappingefficiencyofbottom-tethered sediment traps circulation model. GlobalBiogeochem. Cycles 11, estimated fromthe intercepted fluxesof230 Th and 599–612. 231 Pa. Deep-Sea Res. I 48,865–8891.

q 2003,ElsevierLtd.All rights reserved TreatiseOnGeochemistry Nopart ofthispublication maybe reproduced, stored inaretrievalsystemor ISBN (set):0-08-043751-6 transmitted inany form or byany means,electronic, mechanical,photocopying, recordingor otherwise, without prior writtenpermission ofthe Publisher. Volume6; (ISBN: 0-08-044341-9); pp. 275–291