6.04 the biological pump - imedea divulga csic-uib€¦ · impacts on the geochemistry of many...

6.04The Biological PumpC. L. de la Rocha

University of Cambridge, UK

NOMENCLATURE 84

6.04.1 INTRODUCTION 84

6.04.2 DESCRIPTION OF THE BIOLOGICAL PUMP 846.04.2.1 Photosynthesis and Nutrient Uptake 86

6.04.2.1.1 Levels of primary production 876.04.2.1.2 Patterns in time and space 876.04.2.1.3 Nutrient limitation 88

6.04.2.2 Flocculation and Sinking 886.04.2.2.1 Marine snow 886.04.2.2.2 Aggregation and exopolymers 896.04.2.2.3 Sinking 89

6.04.2.3 Particle Decomposition and Repackaging 906.04.2.3.1 Zooplankton grazing 906.04.2.3.2 Bacterial hydrolysis 906.04.2.3.3 Geochemistry of decomposition 90

6.04.2.4 Sedimentation and Burial 916.04.2.5 Dissolved Organic Matter 916.04.2.6 New, Export, and Regenerated Production 92

6.04.3 IMPACT OF THE BIOLOGICAL PUMP ON GEOCHEMICAL CYCLING 926.04.3.1 Macronutrients 92

6.04.3.1.1 Carbon 926.04.3.1.2 Nitrogen 926.04.3.1.3 Phosphorus 936.04.3.1.4 Silicon 94

6.04.3.2 Trace Elements 966.04.3.2.1 Barium 966.04.3.2.2 Zinc 966.04.3.2.3 Cadmium 986.04.3.2.4 Iron 98

6.04.4 QUANTIFYING THE BIOLOGICAL PUMP 996.04.4.1 Measurement of New Production 996.04.4.2 Measurement of Particle Flux 100

6.04.4.2.1 Sediment traps 1006.04.4.2.2 Particle-reactive nuclides 1016.04.4.2.3 Oxygen utilization rates 101

6.04.5 THE EFFICIENCY OF THE BIOLOGICAL PUMP 1026.04.5.1 Altering the Efficiency of the Biological Pump 102

6.04.5.1.1 In HNLC areas 1026.04.5.1.2 Through changes in community composition 1046.04.5.1.3 By varying the C : N : P ratios of sinking material 1046.04.5.1.4 By enhancing particle transport 105

6.04.6 THE BIOLOGICAL PUMP IN THE IMMEDIATE FUTURE 1056.04.6.1 Response to Increased CO2 1056.04.6.2 Response to Agricultural Runoff 106

6.04.6.2.1 Shift towards silicon limitation 1066.04.6.2.2 Shifts in export production and deep ocean C : N : P 106

6.04.6.3 Carbon Sequestration via Ocean Fertilization and the Biological Pump 106

ACKNOWLEDGMENTS 107

REFERENCES 107

83

NOMENCLATURE

H2CO3* dissolved CO2 þ H2CO3 (mM)

JCorgflux of organic C to depth

(g C m22 yr21)z depth (m)PP primary production (g C m22 yr21)D diffusivity of CO2 in seawater (m2 s21)r radius of phytoplankton cell (mm)Ce extracellular CO2 concentration (mM)Ci intracellular CO2 concentration (mM)k0 rate constant for HCO3 ! CO2 (s21)FCO2

flux of CO2 to cell surface (mmol s21)t residence time (yr)P

CO2 total CO2 (mM)

6.04.1 INTRODUCTION

Despite having residence times (t) that exceedthe ,1,000 yr mixing time of the ocean (Broeckerand Peng, 1982), many dissolved constituents ofseawater have distributions that vary with depthand from place to place. For instance, silicicacid (t ¼ 1.5 £ 104 yr), nitrate (t ¼ 3,000 yr),phosphate (t ¼ (1–5) £ 104 yr), and dissolvedinorganic carbon (DIC; t ¼ 8.3 £ 104 yr) aregenerally present in low concentrations in surfacewaters and at much higher concentrations belowthe thermocline (Figure 1). Additionally, theirconcentrations are higher in older deep watersthan they are in the younger waters of the deepsea (Figure 2). This is the general distributionexhibited by elements and compounds takingpart in biological processes in the ocean andis generally referred to as a “nutrient-type”distribution.

Both the lateral and vertical gradients in theconcentrations of nutrients result from “thebiological pump” (Figure 3). Dissolved inorganicmaterials (e.g., CO2, NO3

2, PO432, Si(OH)4) are

fixed into particulate organic matter (carbo-hydrates, lipids, proteins) and biominerals (silicaand calcium carbonate) by phytoplankton insurface waters. Some of these particles aresubsequently transported, by sinking, into thedeep. The bulk of the organic material andbiominerals decomposes in the upper ocean viadissolution, zooplankton grazing, and microbialhydrolysis, but a significant supply of materialdoes survive to reach the deep sea and sediments.Thus, just as biological uptake removes certaindissolved inorganic materials in surface waters,the decomposition of sinking biogenic particlesprovides a source of dissolved inorganic materialto deeper waters. Thus, deeper waters containhigher concentrations of biologically utilizedmaterials than do surface waters. Older deeperwaters contain higher concentrations of bumscompared to newly formed deep waters or surfacewaters.

One side-effect of the biological pump is thatCO2 is shunted from the surface ocean and intothe deep sea, thus lowering the amount in theatmosphere. For many years it has beenrecognized that pre-Industrial CO2 levels in theatmosphere were about one-third of what theywould be in the absence of a biological pump(Broecker, 1982). It is also known that thebiological pump is not operating at its fullcapacity. In so-called “high-nutrient, low-chlorophyll” (HNLC) areas of the ocean, aconsiderable portion of the nutrients suppliedto the surface waters is not utilized in support ofprimary production, most likely due to thelimitation of phytoplankton growth by aninadequate supply of trace elements (e.g., Martinand Fitzwater, 1988). The possibility that thebiological pump in HNLC areas might bestimulated by massive additions of iron bothartificially as a means of removing anthropogenicCO2 from the atmosphere and naturally as acause for the lower glacial atmospheric CO2

levels (Martin, 1990) is the focus of muchresearch and debate (e.g., Martin et al., 1994;Coale et al., 1996; Boyd et al., 2000; Watsonet al., 2000).

Although the biological pump is most popu-larly known for its impact on the cycling ofcarbon and major nutrients, it also has profoundimpacts on the geochemistry of many otherelements and compounds. The biological pumpheavily influences the cycling, concentrations,and residence times of trace elements—such ascadmium, germanium, zinc, nickel, iron, arsenic,selenium—through their incorporation intoorganic matter and biominerals (Bruland, 1980;Azam and Volcani, 1981; Elderfield andRickaby, 2000). Scavenging by sinking biogenicparticles and precipitation of materials in themicroenvironment of organic aggregates andfecal pellets plays a large role in the marinegeochemistry of elements such as barium,thorium, protactinium, beryllium, rare earthelements (REEs), and yttrium (Dehairs et al.,1980; Anderson et al., 1990; Buesseleret al., 1992; Kumar et al., 1993; Zhang andNozaki, 1996). Even major elements in seawatersuch as Ca2þ and Sr2þ display slight surfacedepletions (Broecker and Peng, 1982; deVilliers, 1999) as a result of the biologicalpump, despite their long respective oceanicresidence times of 1 Myr and 5 Myr (Broeckerand Peng, 1982; Elderfield and Schultz, 1996).

6.04.2 DESCRIPTION OF THE BIOLOGICALPUMP

The biological pump can be sectioned intoseveral major steps: the production of organic

The Biological Pump84

matter and biominerals in surface waters, thesinking of these particles to the deep, and thedecomposition of the settling (or settled) particles.In general, phytoplankton in surface waters take upDIC and nutrients. Carbon is fixed into organicmaterial via photosynthesis and, together withnitrogen, phosphorus, and trace elements, form thecarbohydrates, lipids, and proteins, which allcomprise bulk organic matter. Once formed, thisorganic matter faces the immediate possibility ofdecomposition back to CO2, phosphate, ammonia,and other dissolved nutrients through consumptionby herbivorous zooplankton and degradation by

bacteria. In fact, most of the primary productionformed will be recycled within the upper hundredfew meters of the water column (Martin et al.,1987). Some portion of the primary productionwill, however, be exported to deeper waters oreven to the sediments before decomposition andmay escape remineralization entirely and remainin the sedimentary reservoir.

It is worth taking a closer look at the varioussteps in the biological pump (Figure 3). Rates,overall amounts, and the distribution and char-acter of materials produced, transported, anddecomposed vary wildly within the ocean.

Figure 1 Depth profiles of: (a)P

CO2, (b) dissolved CO2, (c) silicic acid, (d) nitrate, and (e) phosphate from theIndian Ocean (278 40 S, 568 580 E; GEOSECS Station 427) (source Weiss et al., 1983).

Description of the Biological Pump 85

6.04.2.1 Photosynthesis and Nutrient Uptake

In the initial step of the biological pump,phytoplankton in sunlit surface waters convertCO2 into organic matter via photosynthesis:

CO2 þ H2O þ light –! CH2O þ O2 ð1Þ

The first stable product of carbon fixation by theenzyme, ribulose bisphosphate carboxylase(Rubisco), is glyceraldehyde 3-phosphate, a 3-Csugar. This 3-C sugar is fed into biosyntheticpathways and forms the basis for all organiccompounds produced by photosynthetic organ-isms. Fixed carbon and major and trace elements

Figure 2 Nitrate concentrations along the great ocean conveyor at 2,000 m depth (source Levitus et al., 1994, byway of the LDEO/IRI Data Library).

Figure 3 Diagram of the biological pump (after OCTET workshop report).


such as hydrogen, nitrogen, phosphorus, calcium,silicon, iron, zinc, cadmium, magnesium, iodine,selenium, and molybdenum are used for thesynthesis of carbohydrates, lipids, proteins, bio-minerals, amino acids, enzymes, DNA, and othernecessary biochemicals.

Besides carbon, the two main components ofphytoplankton organic matter are nitrogen andphosphorus, in the average molar proportion of106C:16N:1P (Redfield, 1934, 1958) known as theRedfield ratio. In addition, diatoms, by virtue ofdepositing opal (amorphous, hydrated silica) intheir cell wall, have an average C:Si ratio of 8(Brzezinski, 1985), although this ratio may varyfrom at least 3 to 40 depending on the conditionsof light, temperature, and nutrient availability(Harrison et al., 1977; Brzezinski, 1985). Cocco-lithophorids produce scales (coccoliths) made ofCaCO3 and contain 20–100 mmol of CaCO3 permol of organic carbon (Paasche, 1999).

Phytoplankton particulate matter (organic andbiomineralized) contains many trace elements.The most abundant are magnesium, cadmium,iron, calcium, barium, copper, nickel, zinc, andaluminum (Table 1), which are important con-stituents of enzymes, pigments, and structuralmaterials. Carbonic anhydrase requires zinc orcadmium (Price and Morel, 1990; Lane andMorel, 2000), nitrate reductase requires iron(Geider and LaRoche, 1994), and chlorophyllcontains magnesium. Additionally, elements suchas sodium, magnesium, phosphorus, chlorine,potassium, and calcium may be present as ions

within cells and are important for osmoregulationand the maintenance of charge balance (e.g.,Fagerbakke et al., 1999).

A wide variety of ions may be adsorbed onto thesurfaces of biogenic particles. The removal anddeposition of particle-reactive elements such asthorium (Buesseler et al., 1992) and protactinium(Kumar et al., 1993) have been shown to correlatewith the primary production of particles in theocean. Additionally, thorium has been shown tocomplex with colloidal, surface-reactive polysac-charides (Quigley et al., 2002).

6.04.2.1.1 Levels of primary production

The amount of primary production carried out inthe oceans each year has been estimated from oceancolor satellite data and shipboard 14C incubationsto be 140 g C m22 for a total of 50–60 Pg C(4–5 Pmol C) fixed in the surface ocean each year(Shuskina, 1985; Martin et al., 1987; Field et al.,1998). This represents roughly half of the globalannual 105 Pg C fixed each year (Field et al., 1998),despite the fact that marine phytoplankton compriseless than 1% of the total photosynthetic biomass onEarth. Extrapolation from Redfield ratios suggeststhe incorporation 0.6–0.8 Pmol N, 40–50 Tmol Pinto biogenic particles each year in association withmarine primary production. From the proportion ofprimary production carried out by diatoms and theaverage Si:C ratio of diatoms, silica production ratesmay be calculated to be 200–280 Tmol Si yr21

(Nelson et al., 1995; Treguer et al., 1995).

6.04.2.1.2 Patterns in time and space

Rates of primary productivity in upwellingregions of the ocean outpace those ofnon-upwelling coastal regions, which in turn aregreater than rates in the oligotrophic open ocean(Figure 4; Ryther, 1969; Martin et al., 1987).Open-ocean primary production levels are,130 g C m22 yr21, whereas in nonupwellingcoastal areas and upwelling zones they are250 g C m22 yr21 and 420 g C m22 yr21, respect-ively (Martin et al., 1987). However, because theopen ocean constitutes 90% of the area of theocean, the bulk (80%) of the ocean’s annualcarbon fixation occurs there rather than in coastaland upwelling regions.

Different types of phytoplankton dominateprimary production in the different marineregimes. Diatoms perform roughly 75% of theprimary production that occurs in upwelling andcoastal regions of the ocean but less than 35% ofthat taking place in the open ocean (Nelson et al.,1995). Phytoplankton biomass and primary pro-duction in the open ocean are dominated instead

Table 1 Elemental composition of marinephytoplankton from cultures and plankton tows.

Element Element : C ratio(mol : mol)

References

N 0.15 bSi (diatoms only) 0.13 cP 0.009 bCa 0.03 d,eFe 2.3 £ 1026–1.8 £ 1023 d, e, fZn 6 £ 1025 d,eAl 1 £ 1024 d,eCu a3 £ 1026–0.006 d,eNi a2 £ 1025–0.006 eCd a5 £ 1027–0.005 d,eMn a4 £ 1026–0.004 d,eBa a1 £ 1025–0.01 d,eMg a0.02 dNa a0.1 dSr a8 £ 1025 dTi a1 £ 1025 dCr a2 £ 1026 d

a Calculated from dry weight data using an average phytoplanktonC content on a dry weight basis of 50%. b Redfield (1958).c Brzezinski (1985). d Martin and Knauer (1973).e Collier and Edmond (1984). f Sunda and Huntsman (1995a).


by prokaryotic picoplankton (Chisholm et al.,1988; Liu et al., 1999; Steinberg et al., 2001).

Outside of the tropics, levels of marine primaryproductivity vary systematically throughout theyear (Heinrich, 1962). Standing stocks of phyto-plankton and levels of primary production peak inthe spring following the onset of water columnstratification and the increase in available light.Depletion of nutrients in the stratified watercolumn in summer inhibits phytoplankton growthand grazing by zooplankton reduces standingstocks. Some areas may experience a small bloomof phytoplankton in the autumn when light levelsare still adequate and the onset of winter convec-tion and overturning injects nutrients into theeuphotic zone.

6.04.2.1.3 Nutrient limitation

The upper limit of primary production is set bythe supply of nutrients (nitrogen, phosphorus,silicon, iron) to the euphotic zone. Nitrogen inputsto the surface ocean may limit the primaryproductivity of the whole ocean over short time-scales. Over timescales approaching and exceed-ing the (1 – 5) £ 104 yr residence time ofphosphorus (Ruttenberg, 1993; Filippelli andDelaney, 1996), its inputs limit global oceanprimary productivity (Tyrrell, 1999).

Regionally and for different types of phyto-plankton, the limitation of both the rate andoverall amount of primary production is morevaried. Major nutrients in HNLC areas of theocean (such as the Southern Ocean, the Equatorial

Pacific, and the North Pacific subarctic) are nevercompletely consumed in support of primaryproduction, because low levels of iron limitphytoplankton growth (Martin and Fitzwater,1988; Martin et al., 1994; Coale et al., 1996;Boyd et al., 2000). Diatoms, which, unlike otherdominant members of the phytoplankton, requiresilicon for growth, are often limited by lowconcentrations of silicic acid in surface waters(Brzezinski and Nelson, 1996; Nelson and Dortch,1996). Growth of diazotrophic (N2 fixing) phyto-plankton such as the cyanobacteria, Trichodes-mium, will be more susceptible to phosphorus andiron limitation, of course, than to nitrogenlimitation. Even the concentration of dissolvedCO2 in seawater (especially in the midst of aphytoplankton bloom) may limit instantaneousrates, although not ultimate levels, of primaryproduction (Riebesell et al., 1993; Wolf-Gladrowet al., 1999).

6.04.2.2 Flocculation and Sinking

6.04.2.2.1 Marine snow

The primary formation of biogenic particles inthe euphotic zone represents the maximumamount of material that may be transportedinto the deep ocean or sediments. In practice,however, less than half of these particles survivezooplankton grazing and microbial attack longenough to be exported from the euphotic zone,and only a few percent endure to settle into thedeep ocean and sediments (Martin et al., 1987).Material that reaches the deep ocean andseafloor does so not as individual phytoplankton

Figure 4 Distribution of primary production in the ocean (source ICMS, Rutgers University).


cells slowly meandering down towards thebottom, but rather arrives as larger, rapidlysinking particles (McCave, 1975; Suess, 1980;Billett et al., 1983; Fowler and Knauer, 1986;Alldredge and Gotschalk, 1989) that havetraversed the distance between surface anddeep in a matter of days (Billett et al., 1983;Asper et al., 1992).

These larger particles, known collectively as“marine snow” (Alldredge and Silver, 1988), areformed either by zooplankton, which producemucous feeding structures and fecal pellets, or bythe physical coagulation of smaller particles(McCave, 1984; Alldredge et al., 1993). Coagu-lation is the more important of the two formationpathways. The bulk of the organic materialreaching the deep sea does so as aggregatedphytoplankton that has not been ingested byzooplankton (Billett et al., 1983; Turner, 2002).

Sinking rates of marine snow are orders ofmagnitude greater than those of unaggregatedphytoplankton cells (Smayda, 1970; Shanks andTrent, 1980; Alldredge and Gotschalk, 1989). Theshorter transit time from surface to bottom foraggregated particles results in the enhancedtransport of carbon, nitrogen, phosphorus, silicon,and other materials to the deep sea and sedimentsdespite the fact that marine snow particles are sitesof elevated rates of decomposition and nutrientregeneration. Intense colonization and hydrolysisof the particles by bacteria (Smith et al., 1992;Bidle and Azam, 1999, 2001) and breakup andconsumption of the particles by zooplankton(Steinberg, 1995; Dilling et al., 1998; Dillingand Alldredge, 2000) reduce the vertical flux ofmaterials to the seafloor.

6.04.2.2.2 Aggregation and exopolymers

Coagulation requires the success of twoactivities: the collision of particles and theirsubsequent joining to form an aggregate. In theocean, particles collide due to processes such asshear, Brownian motion, and differential settling(Kepkay, 1994). The probability of particlesattaching following a collision is controlled bythe physical and chemical properties of theparticles’ surfaces (Alldredge and Jackson,1995). The probability of sticking is greatlyenhanced by exopolymers produced by phyto-plankton and bacteria (Alldredge and McGillivary,1991; Alldredge et al., 1993; Passow, 2000;Engel, 2000).

These exopolymers, known as transparent exo-polymer particles (TEPs; Passow, 2002), turn outto be important for the transport of material to thedeep. The formation of rapidly sinking aggregatesis controlled more by TEP abundance than byphytoplankton concentrations (Logan et al., 1995)

and TEP is required for the aggregation and sedi-mentation of diatoms out of the water column(Passow et al., 2001).

Little is known about the chemical character-istics of the polymers responsible for particleaggregation in marine systems. They are com-prised of acidic polysaccharides (Alldredge et al.,1993; Mopper et al., 1995) and proteins (Long andAzam, 1996). The carbohydrate component ofTEP contains glucose, mannose, arabinose,xylose, galactose, rhamnose, glucuronate, andO-methylated sugars (Janse et al., 1996; Hollowayand Cowen, 1997), and is generally rich indeoxysugars (Mopper et al., 1995). Very littleelse is known about the specific composition ofTEP, and virtually nothing is known of itsstructural characteristics (Holloway and Cowen,1997; Schumann and Rentsch, 1998; Engel andPassow, 2001).

Exopolymer particles are formed from dis-solved organic matter (DOM) and continue toscavenge DOM as they grow, providing amechanism for the biological pumping of DOMinto the deep sea (Engel and Passow, 2001). TEPalso contains carbon and nitrogen in proportionsexceeding Redfield ratios (Mari et al., 2001; Engelet al., 2002), providing a mechanism for pumpingof carbon in excess of what would be predictedfrom the availability of nitrogen.

6.04.2.2.3 Sinking

Sinking rates of solitary phytoplankton cells areonly about a meter per day (Smayda, 1970).Particles that sink this slowly require over a yearto reach the benthos of the relatively shallowcontinental shelf, and ten years to reach theabyssal ocean floor. Given the rapid rates ofmicrobial decomposition of organic material inthe ocean and the abundance of zooplanktongrazers, it is virtually impossible for such a slowlysinking particle to reach the seafloor.

Sinking rates of marine snow, however, aregreater than 100 m d21 (Shanks and Trent, 1980;Alldredge and Gotschalk, 1989). Transit time to thedeep in this case is days to weeks, which agrees withobservations of a close temporal coupling betweensurface production and seafloor sedimentation (e.g.,Billett et al., 1983; Asper et al., 1992).

It may easily be argued then that particle flux iscontrolled by rates of particle aggregation andsinking, perhaps more than it is controlled byoverall levels of primary production. For example,year-to-year variability in carbon export to deepwaters correlates more strongly with the size ofthe dominant primary producer than with year-to-year variations in levels of carbon fixation (Boydand Newton, 1995).


6.04.2.3 Particle Decomposition andRepackaging

Organic matter in the ocean rapidly decomposesand there is intense recycling of elements evenwithin the euphotic zone. The flux of particulateorganic carbon (POC) in the ocean decreasesexponentially with depth below the euphotic zone(Figure 5; Martin et al., 1987). New productionconstitutes, on average, only 20% of the totalprimary production in the sea (Harrison, 1990;Laws et al., 2000). Mediating this decompositionand recycling are zooplankton and heterotrophicbacteria (Cho and Azam, 1988; Smith et al., 1992;Steinberg, 1995; Dilling et al., 1998; Dilling andAlldredge, 2000). Bacteria and zooplanktondiminish the sinking particulate flux by bothconsuming particles and converting them back toCO2 and dissolved materials, and by convertinglarge, sinking particles into smaller particles withreduced or nonexistent sinking rates.

Although the bulk of particles are brokendown in the surface ocean, midwater processesare also important. Midwater decompositionof sinking particles deflates the regional vari-ability in fluxes of POM. The POC flux range,0.5–12 g C m22 yr21, among different regions inthe Atlantic at 125 m is compressed by 85% to0.5–2.4 g C m22 yr21 by a depth of 3,000 m dueto the biological consumption and repackaging ofparticles at depth (Anita et al., 2001).

6.04.2.3.1 Zooplankton grazing

Zooplankton may reduce the sinking flux ofbiogenic particles in the ocean in two ways. Thefirst is by grazing upon particles which reduces thetotal amount of particulate organic material(POM) in the water column and shifts itsoccurrence from large, fast-sinking aggregates tosmaller fecal pellets which constitute only a minor

portion of the sinking organic flux (Turner, 2002).The second way they reduce the particle flux is byactively breaking up aggregates into smallerparticles. At stations off Southern California, forinstance, the average overnight increase in thenumber of aggregates per liter by 15% wasattributable to the fragmentation of larger particlesby swimming euphausiids (Dilling and Alldredge,2000).

The relative impact of zooplankton grazing onprimary production decreases with increasingproduction levels; the proportion of primaryproduction that is consumed by zooplanktondecreases exponentially as productivity levelsincrease (Calbet, 2001). This supports the obser-vation that the ratio of export production to totalproduction is higher in areas of high productivity.Globally, ,12% of marine primary production, or5.5 Pg C (0.5 Pmol C), is consumed by mesozoo-plankton each year (Calbet, 2001).

6.04.2.3.2 Bacterial hydrolysis

Bacterial hydrolysis plays a major role in thedecomposition of sinking and suspended matter inthe ocean. Bacterial organic carbon has beenobserved to make up over 40% of the total POC inthe water column, and the proportion of thesinking flux of carbon utilized by bacteria maybe equal to 40–80% of the surface primaryproduction in near-shore areas (Cho and Azam,1988). Marine aggregates have been shown tocontain high concentrations of hydrolytic exo-enzymes, such as proteases, and polysaccharidaseslike glucosidases (Smith et al., 1992). Turnovertimes of organic components of marine aggregatesdue to hydrolysis may be short, on the order offraction of a day to a few days (Smith et al., 1992).Bacterial proteases have also been shown toenhance the dissolution rates of biogenic silica indiatom aggregates (Bidle and Azam, 2001). Muchof the hydrolyzed material is not taken up by thebacteria attached to the particles but instead joinsthe pool of DOM present in the water (Smith et al.,1992).

6.04.2.3.3 Geochemistry of decomposition

Ingestion of POM by zooplankton results in therespiration and excretion of a portion of that POMas CO2, NH4, dissolved organic nitrogen (DON),phosphate, and dissolved organic phosphorus(DOP). Assimilation efficiencies of organic matterfor zooplankton grazing on phytoplankton rangefrom 10% to 40% (Ryther, 1969; Michaels andSilver, 1988). Zooplankton do not assimilatesignificant quantities of silicon from diatomsconsumed, leaving regeneration of silicic acid tobe mediated strictly by opal dissolution rates.

Figure 5 Sinking fluxes of C in the open ocean. Curveshown is the exponential relationship (flux ¼ 1.53(z/100)20.858) shown for carbon flux in the northeast

Pacific (after Martin et al., 1987).


6.04.2.4 Sedimentation and Burial

Of the 50–60 Pg C (4–5 Pmol C) fixed intoorganic material in the surface ocean each year(Shuskina, 1985; Martin et al., 1987; Field et al.,1998) and the 16 Pg C (1.3 Pmol C) exported tothe deep sea (Falkowski et al., 2000), only,0.16 Pg C (0.13 Pmol C) reach the seafloor(Hedges and Keil, 1995). Each year roughly0.16 Pg C (13 Tmol C) are actually preserved inocean sediments (Hedges and Keil, 1995). Thus,the rate of accumulation of organic carbon byocean sediments is 0.5% of that of carbon fixationin the surface ocean.

The accumulation of sedimentary organiccarbon in the ocean varies greatly from placeto place, with continental margin sedimentsaccounting for the bulk of the organic carbonbuildup (Hedges and Keil, 1995; de Haas et al.,2002). About 94% of the sedimentary organiccarbon that is preserved in the oceans is buriedon continental shelves and slopes (Hedges andKeil, 1995). This leaves only 6% of the totalsedimentary organic carbon to accumulate in theopen ocean. Since the open ocean, due to its vastarea, plays host to 80% of the annual primaryproduction, the accumulation of only 6% of theorganic carbon suggests an overall preservationefficiency of 0.02% in the open ocean. On thecontinental shelves and slopes, this preservationefficiency is, by comparison, large at 1.4%.

Many reasons have been suggested for theregional differences in the accumulation andpreservation of organic carbon in the sediments.Differences in the flux of organic particles to theseafloor due to differences in overhead primaryproduction levels, rates of aggregation and sink-ing, and depth of the water column may contributeto a higher preservation efficiency. The oxygencontent of bottom waters has also been suggestedas important, although a correlation betweenburial efficiency and bottom water oxygen con-centration is not seen (Hedges and Keil, 1995) andrates of organic matter hydrolysis by bacteria maystill be high, even under anoxic conditions (e.g.,Arnosti et al., 1994). The long-term preservationof organic material in sediments may be tied to thesorption of the organic molecules to mineralsurfaces (Hedges and Keil, 1995), although thenature of the associations and the rates at whichthey occur have not been closely detailed.

6.04.2.5 Dissolved Organic Matter

DOM has not been as intensively studied asother aspects of the biological pump perhapsbecause DOM does not sink. However, DOM doesplay an active role in the biological pump in atleast three ways. Much DOM can be utilizedbiologically and may directly provide phosphorus

and nitrogen for primary production (Clark et al.,1998; Zehr and Ward, 2002). DOM may assembleinto colloidal and particulate material that can sinkas well as scavenge other material to form marinesnow (Alldredge et al., 1993; Kepkay, 1994; Chinet al., 1998). DOM is also a large reservoir ofcarbon in the ocean, containing at least an order ofmagnitude more carbon than the other organiccarbon reservoirs in the ocean (Kepkay, 1994).

Although the origins of DOM have not beenfully detailed, phytoplankton can serve as thedominant source of DOM to the ocean. Activelygrowing phytoplanktons secrete DOM (Biddandaand Benner, 1997; Soendergaard et al., 2000;Teira et al., 2001) and the polysaccharide com-position of phytoplankton exudates resembles thatof the high molecular weight fraction of DOM(Aluwihare and Repeta, 1999). PhytoplanktonDOM is also released during grazing by zoo-plankton (Strom et al., 1997). Organic matter,such as mucus, on phytoplankton cell surfacesmay also be hydrolyzed by bacteria and releasedas DOM (Smith et al., 1995).

The exact composition of marine DOM isunknown. It has, as of early 2000s, been shownto contain carbohydrates, which consist largelyof polysaccharides, and amino acids, amides (suchas chitin), phosphorus esters, and phosphonates(Benner et al., 1992; McCarthy et al., 1997; Clarket al., 1998; Amon et al., 2001). Microbialdegradation could play a role in setting thecomposition of DOM in the ocean (Amon et al.,2001).

One feature that has great relevance to theimportance of DOM to the carbon cycle is itsenrichment in carbon over the Redfield proportion.The C : N : P ratios of high molecular weight DOMare on the order of 350 : 20 : 1 (Kolowith et al.,2001). Carbon enrichment is also observed forbulk DOM (Kaehler and Koeve, 2001). Part of thisenrichment in carbon may be due to the enhancedremineralization of phosphorus and nitrogen fromDOM (Clark et al., 1998; Kolowith et al., 2001), assuggested by an increase in the C : P ratio of DOMwith depth (Kolowith et al., 2001). DOM may alsojust simply be produced with high C : N and C : Pratios. TEPs which form from DOM precursors(Alldredge et al., 1993; Chin et al., 1998) havehigh C : N ratios (Engel and Passow, 2001;Mari et al., 2001). The polysaccharides thateventually form TEP are exuded by phytoplanktonand may represent excess photosynthate (Engel,2002), carbohydrates, and lipids formed whennutrient limitation shuts off the supply of, forexample, the nitrogen needed for the synthesis ofnitrogen-containing compounds such as proteins(Morris, 1981).

The importance of the carbon-enriched DOMpool as a reservoir in the carbon cycle hinges uponboth the turnover time and amount of the carbon


therein. Estimates for the amount of dissolvedorganic carbon (DOC) in the ocean vary, althoughan estimate places the size of the pool at 200 Pg C(Kepkay, 1994), which is more comparable to the750 Pg of carbon present in the atmosphere than itis to the 3.6 £ 104 Pg C deep-sea reservoir of DIC(Sundquist, 1993). The average age of marineDOM is ,6,000 yr (Williams and Druffel, 1987).The bulk (,70%) of the DOM is low molecularweight (Benner et al., 1992) and relatively resistantto microbial degradation (Bauer et al., 1992; Amonand Benner, 1994). High molecular weight com-pounds that are quickly turned over by bacterialdecomposition (Amon and Benner, 1994) make upthe remaining 30% of the DOM pool.

6.04.2.6 New, Export, and RegeneratedProduction

Not all of the primary production in the oceanfeeds carbon into the biological pump. The vastportion of carbon fixed globally each year in theeuphotic zone is remineralized by zooplanktonand bacteria in the euphotic zone and convertedstraight back to CO2 and dissolved nutrients.These recycled nutrients may then be used to fuelfurther carbon fixation.

It has long been recognized (Dugdale andGoering, 1967) that a portion of the primaryproduction (regenerated production) is supportedby nutrients regenerated in the euphotic zone, andanother portion (new production) is supported bynutrients imported into the euphotic zone throughupwelling, river inputs, nitrogen fixation, oratmospheric deposition. The ratio of new to totalprimary production in the ocean, known as thef-ratio (Eppley and Peterson, 1979), is generallyhigher in upwelling environments than it is inoligotrophic regions of the ocean (Harrison, 1990;Laws et al., 2000). On average, ,20% of the totalglobal marine primary production is new pro-duction, although the range of values from regionto region is ,0.07–0.7 (Laws et al., 2000).

6.04.3 IMPACT OF THE BIOLOGICAL PUMPON GEOCHEMICAL CYCLING

6.04.3.1 Macronutrients

6.04.3.1.1 Carbon

The influence of the biological pump on thedistribution of DIC carbon in the ocean mayserve as a rough model for the influence it hason the distribution of a score of other elements.Low concentrations of DIC are observed insurface waters (Figure 1) due to the uptake ofdissolved CO2 (and perhaps HCO3

2; Raven,1997) by phytoplankton. Concentrations of

dissolved CO2 increase most rapidly justbelow the euphotic zone, associated with thebulk of the decomposition of POC (Martin et al.,1987; Anita et al., 2001). Deep-water concen-trations of dissolved CO2 are higher thansurface-water concentrations, and older deepwaters contain more dissolved CO2 than youngerones (Broecker and Peng, 1982).

Much of the current scientific interest in thebiological pump revolves around the impact it hason levels of CO2 in the atmosphere and,subsequently, on climate. This biological fixationof carbon into organic matter through photosyn-thesis lowers the concentration of dissolved CO2

in surface waters and thus allows for the influx ofCO2 from the atmosphere. This fixed carbon maythen be exported to deeper waters or the sedimentsbefore it decomposes back to CO2, maintainingthe observed gradient in dissolved CO2 concen-trations between waters of the surface and deep(Figure 1). Atmospheric concentrations of CO2

are thus lower for the given size of the oceanicDIC reservoir than they would be in the absence ofthis biological transport of carbon to the deep. Ifall life in the ocean were to die off and the oceanand atmosphere came to equilibrium with respectto CO2, concentrations of CO2 in the atmospherewould rise by ,140 matm (Broecker, 1982),which is a remarkable 50% of the pre-Industrialinterglacial value of 280 matm.

Details of the influence of the biological pumpon the distribution and cycling of carbon in theocean and the controlling factors are discussedlater in this chapter. We briefly discuss therelationship of the biological pump to the cyclingof other elements. This is by no means anexhaustive overview of the biological shufflingof elements throughout the ocean, but instead ahighlight of several elements of particular bio-geochemical interest.

6.04.3.1.2 Nitrogen

Of all the elements playing an important role inthe regulation of the biological pump, nitrogen isthe one with the most complex biologicallymediated cycling. Nitrogenous species takingpart in productivity range from N2, which may befixed into a more universally biologically availableform by nitrogen fixing bacteria, to NO3

2, whichfollows from the production of NO2

2 from NH4þ

through nitrification (Figure 6). The denitrificationpathway sequentially results in the transformationof NO2

2 to N2O and N2. In addition to inorganicnitrogen species are dissolved organic forms, suchas amides, urea, free amino acids, amines(McCarthy et al., 1997), which can be utilizedbiologically, although to varying degrees.

Most of the primary production in the ocean issupported by dissolved inorganic nitrogen (DIN)


that has recycled in the euphotic zone, assuggested by the average marine f-ratio of 0.2(Laws et al., 2000). This further suggests that thebulk of the primary production in the ocean relieson NH4

þ, since that is the predominant recycledform of nitrogen. Extrapolating from the RedfieldC : N ratio of 6.6 and the average f-ratio of 0.2 andthe overall estimate of primary production in theocean each year of 4–5 Pmol C (Shuskina, 1985;Martin et al., 1987; Field et al., 1998) yields0.7 Pmol of particulate organic nitrogen (PON)produced each year, 0.1 Pmol of which is exportedout of the euphotic zone.

Nitrogen fixation. The two aspects of thenitrogen cycle having the greatest impact on thebiological pump are nitrogen fixation and deni-trification. The first provides a mechanism fordrawing on the extensive atmospheric pool of N2

gas in support of primary production. The secondprovides a pathway for DIN to be converted backto N2 gas and removed from the ocean system.

In the ocean ,28 Tg N (2 Tmol N) are fixedeach year (Gruber and Sarmiento, 1997). Nitrogenfixation accounts for about half of the new nitrogenused in primary production (Karl et al., 1997). Onlyprokaryotic organisms can fix nitrogen, leavingthis, at least in the ocean, in the hands of thecyanobacteria and out of the hands of eukaryoticalgae such as diatoms, dinoflagellates, and cocco-lithophorids (unless they are hosting diazotrophicsymbionts). Until recently it was believed that thefilamentous, colony-forming cyanobacterium Tri-chodesmium and cyanobacterial symbionts in

diatoms were responsible for the bulk of thenitrogen fixation occurring in the ocean (Caponeet al., 1997). However, direct measurements of therates of nitrogen fixation by Trichodesmium,coupled with knowledge of their distribution andabundance, fell significantly short of nitrogenfixation rates calculated from geochemical budgets(Gruber and Sarmiento, 1997). It has now beendiscovered that free-living, unicellular cyanobac-teria are expressing the genes for the nitrogen-fixing enzyme, nitrogenase (Zehr et al., 2001), andmay be contributing considerably to fixed nitrogenbudgets in the ocean.

It has been suggested that rates of nitrogenfixation in the modern ocean are limited by theavailability of iron. The iron requirement ofTrichodesmium, however, turns out to be muchlower than previously estimated. Instead, it is theavailability of phosphate that controls the upperlimit of nitrogen fixation in the modern ocean(Sanudo-Wilhelmy et al., 2001).

The demonstration of the phosphorus limitationof nitrogen fixation by cyanobacteria supports thenotion that over geologic time phosphorus ulti-mately limits productivity (Tyrrell, 1999). Whencyanobacteria face a shortfall of nitrogen, they fixnitrogen to meet their demands. This influx of newnitrogen to nitrogen-limited systems allows themto draw down levels of phosphate until the systemis phosphorus limited. Under phosphorus limi-tation, nitrogen fixation is curtailed (Sanudo-Wilhelmy et al., 2001). Thus, while nitrogenmay often be limiting to instantaneous rates ofcarbon fixation, as is frequently the case in themodern ocean, over long time periods the input ofphosphorus to the oceans sets the upper attainablelimit for net primary production (Tyrrell, 1999).

The control of nitrogen fixation by phosphorusavailability makes the nitrogen cycle stand outagainst all of the other global biogeochemicalcycles. In an astounding twist, the biologicaldemand for one element relative to the othercontrols the input flux of one of the elements.Confirmation of this comes in the form of Redfieldratios, since the N : P ratio in both the seawaterreservoir and the marine organic matter output issame (16 : 1) and significantly greater than unity.Due to mass balance requirements, assumingsteady state over long timescales, the only waythis can be the case is if the ratio of N : P inputs tothe ocean is also 16 : 1. It is far more likely that theequality of N : P ratios of inputs, outputs, andreservoir is due to the grand control of nitrogenfixation by phosphate than it is likely to be due toremarkable coincidence.

6.04.3.1.3 Phosphorus

Although there is a tendency to considerdissolved inorganic phosphorus (DIP, measured

Figure 6 Microbial transformations of the nitrogencycle. Pathways depicted are: 1—N2 fixation; 2—DINassimilation; 3—ammonium regeneration; 4—nitrifica-tion; 5—nitrate/nitrite reduction; and 6—denitrification.

Impact of the Biological Pump on Geochemical Cycling 93

as soluble reactive phosphate; Strickland andParsons, 1968) as being simply PO4

32, DIP existsas a considerable number of species. At seawaterpH, DIP is predominantly H2PO4

2 (87%) and only12% PO4

32 and 1% H2PO42 (Greenwood and

Earnshaw, 1984). There are also numerousdissolved organic forms of phosphorus that aretaken up by phytoplankton and used to fuelprimary production in the ocean.

One interesting aspect of the phosphorus cycleis that, unlike the cases for the other major nutrientelements in the ocean, the phosphorus cycle facesthe complexity of containing sinks that are notmediated by biological activities. For example,dissolved phosphate may scavenge onto iron ormanganese oxyhydroxide particles associatedwith hydrothermal activity or react with basaltduring the circulation of water through mid-oceanridge hydrothermal systems (Follmi, 1996). Theexact values are poorly known, but it is estimatedthat the scavenging of phosphate by the hydro-thermal oxyhydroxides may constitute up to 50%of the removal flux of phosphorus from the ocean(e.g., Froelich et al., 1982; Berner et al., 1993).Another large inorganic sink for dissolved phos-phorus is the precipitation of authigenic phos-phate, which account for somewhere betweenabout 10% and 40% of the removal flux. Removalof phosphorus as sedimenting POC by comparisonis thought to be 20–50% of the output ofphosphorus from the ocean.

Roughly 5 Tg P (0.2 Tmol P) are removed fromthe ocean each year as both organic and inorganicphases. This is in reasonable balance with theroughly 5 Tg of reactive phosphorus being broughtin to the system each year naturally. However,these natural sources constitute only half of themodern-day input of phosphorus to the ocean(Froelich et al., 1982), anthropogenic inputshaving doubled the annual phosphate flux.

A doubling of phosphorus inputs to the oceancould have a significant impact on the biologicalpump, although it should be noted that the ocean isalready nitrogen limited and further inputs ofphosphorus will only exacerbate this. Shifts inN : P ratios of surface waters may alter thestructure of the phytoplankton community. Forexample, Phaeocystis grows poorly under high-phosphate conditions and may see its numbersdeclining.

6.04.3.1.4 Silicon

Dissolved silicic acid is required for the growthof diatoms, which deposit opal (amorphous,hydrated silica) in their cell wall and dominatethe production of opal in the modern-day ocean(Lisitzin, 1972). Silica (240 Tmol) is produced bydiatoms in the surface ocean each year (Nelsonet al., 1995; Treguer et al., 1995). Total inputs of

dissolved silicic acid to the euphotic zone are, bycomparison, 120 Tmol Si, mostly from rivers(5 Tmol Si) and upwelling (115 Tmol Si; Tregueret al., 1995). Half of the biogenic silica producedeach year dissolves in the upper 100 m of thewater column (Nelson et al., 1995), and a further47% dissolves in the deep ocean and seafloor, fora net deposition of 6 Tmol (3% of surfaceproduction) each year (Treguer et al., 1995).

Opal accumulation on the seafloor. Opal preser-vation efficiencies are generally highest in pro-ductive environments (e.g., 6% in the permanentlyopen ocean zone of the Southern Ocean versus 0.4%in the oligotrophic North Atlantic). Given numberssuch as these, the traditional view of opal accumu-lation in the sediments is that it is closely linked toopal production in overlying waters (e.g., Broeckerand Peng, 1982), but many factors besides opalproduction govern opal preservation. Opal dissol-ution rates more than double for every 10 8C rise intemperature (Kamitani, 1982). Aggregation mayreduce rates of silicon regeneration from diatoms(Bidle and Azam, 2001). Additionally, the fractionof produced silica that reaches the seabed correlateswith high seasonality and high ratios of carbonexport to production (Pondaven et al., 2000), whichalso tend to be areas where organic matter is formedin blooms and is transported quickly to thesediments as large aggregates.

The impact of the appearance of the diatoms onthe marine silica cycle. The silica cycle is anexcellent example of how much the biologicalpump can impact concentration and distributionof elements in seawater. Presently, surfaceconcentrations of silicic acid are low: the overallaverage silicic acid content of ocean waters isonly 70 mM (Treguer et al., 1995). Prior to theappearance of the diatoms in the Early Tertiary(Tappan and Loeblich, 1973), oceanic concen-trations of silicic acid ranged ,1,000 mM Si. Thesponges and radiolarians controlling the oceanicinventory of silicic acid at that time possessedneither the numbers nor the need to draw downsilicic acid concentrations. The increased output ofbiogenic silica from the ocean associated with theascension of diatoms, with their high affinity forsilicic acid and high cellular requirements forsilicon, resulted in a precipitous drop in the silicicacid content of ocean waters over the LateCretaceous and Paleocene (Figure 7), stabilizingin the Eocene to the ,100 mM values that haveheld ever since (Siever, 1991).

Since sponges and radiolarians are not greatplayers in particle flux, the rise of the diatoms musthave profoundly altered the partitioning of silicicacid between surface and deep. The familiarnutrient-type distribution may have only existedfor the last 50–100 million years. The approximate14-fold drop in silicic acid concentration alsosuggests that the residence time of silicon in


the ocean dropped from 2 £ 105 yr to today’s1.5 £ 104 yr.

Excessive pumping of silicon. Silicic acid, byvirtue of being regenerated from silica instead offrom relatively labile POM, is regenerated moredeeply than the other major nutrients (Figure 1;Dugdale et al., 1995). This decoupling betweensilicic acid and the other nutrients, combined withthe fact that not all phytoplankton utilize silicon,results in there being no Redfield relationshipbetween silicon and carbon, nitrogen, or phos-phorus. Upwelled waters contain silicic acid and

nitrate close to the ,1 : 1 molar ratio of utilizationby diatoms, but Si : N ratios increase with depthbelow the euphotic zone (Figure 8).

Impact of iron on silicon pumping. Iron limitationof diatoms increases the pumping of silicon (relativeto nitrogen and carbon) to deeper waters. Iron-limited diatoms are inhibited in their utilization ofnitrogen as a result of the iron requirement of nitratereductase (Geider and LaRoche, 1994). However,iron-limited diatoms continue to take up silicic acid,although at lowered rates (De La Rocha et al., 2000).As a result, the Si : N ratios of iron-limited diatoms

Figure 7 Estimated average marine concentrations of silicic acid over the Phanerozoic (after Siever, 1991).

Figure 8 Ratios of silicic acid to nitrate with depth in the Pacific. Profiles shown are from: (a) the NE Pacific WOCEstation 66 (478 330 N, 1458 330 W); (b) HOT station Aloha (228 450 N, 1588 000 W); (c) SE Pacific WOCE station 288

(388 600 S, 888 000 W); and (d) SE Pacific WOCE station 241 (538 200 S, 768 360 W).


may be as high as 2 or 3 (Takeda, 1998; Hutchinsand Bruland, 1998), much higher than the 0.8 ofnutrient-replete diatoms (Brzezinski, 1985).

6.04.3.2 Trace Elements

The biological pump influences, to varyingdegrees, the distribution of many elements inseawater besides carbon, nitrogen, phosphorus,and silicon. Barium, cadmium, germanium, zinc,nickel, iron, selenium, yttrium, and many of theREEs show depth distributions that very closelyresemble profiles of the major nutrients. Addition-ally, beryllium, scandium, titanium, copper,zirconium, and radium have profiles whereconcentrations increase with depth, although thecorrespondence of these profiles with nutrientprofiles is not as tight (Nozaki, 1997).

6.04.3.2.1 Barium

Vertical profiles of dissolved barium (Ba2þ) inthe ocean resemble profiles of silicic acid andalkalinity (Figure 9; Lea and Boyle, 1989; Jeandelet al., 1996), suggesting that biological processesstrongly influence barium distributions throughoutthe ocean. However, the strict incorporation ofbarium into biogenic materials is not the dominantmeans of Ba2þ removal from ocean waters.Despite the similarity between the profilesof Ba2þ and Si(OH)4 which suggests a commonremoval phase, the amount of barium incorporatedinto diatom opal (,9 £ 1026 mol Ba per molSi; Shemesh et al., 1988) cannot account forthe 2 £ 1024 mol Ba2þ per mol Si(OH)4 slope(Jeandel et al., 1996) in the ocean. Ba2þ appearsinstead to be mainly removed from seawater asbarite (BaSO4) formed in association with opaland decaying organic material (Dehairs et al.,1980; Bishop, 1988). The exact mechanism forbarite precipitation is unknown, but it is thoughtthat it forms in the SO4

22-enriched microenviron-ments of decaying particles that may be, thus,supersaturated with respect to barite (Dehairset al., 1980).

Although the marine budget of barium is onlyapproximately known, it does appear to be bothbalanced and controlled by biogenic particleformation. Approximately 35 Gmol of Ba2þ areremoved from surface waters every year (Dehairset al., 1980). Of this 35 Gmol, 60% settles as bariteand the rest is incorporated into or adsorbed ontophases such as CaCO3 or SiO2 (Dehairs et al., 1980;Dymond et al., 1992). Barium (10–25 Gmol) isburied on the seafloor each year (Dehairs et al.,1980).

Because barite forms in association with organicmaterial, there is a tight correlation (R 6 ¼ 0.93)

between the sedimentary fluxes of carbon and barite(Dymond et al., 1992). Thus, barite accumulationrates have been used to infer past levels of exportproduction in the ocean (e.g., Dymond et al., 1992;Paytan et al., 1996).

6.04.3.2.2 Zinc

The profiles of dissolved zinc in the ocean arealso similar to those of Si(OH)4 (Figure 10;Bruland, 1980), but as with barium, the mainremoval phase for zinc is not opal. Less than 3%of the zinc taken up by diatoms is deposited intheir opaline cell wall (Ellwood and Hunter,2000), and the Zn : Si ratio of acid-leached opalis much lower than that of dissolved Zn andSi(OH)4 in the water column (Bruland, 1980;Collier and Edmond, 1984). Instead, most of

Figure 9 Profiles of Ba2þ, Si(OH)4, and alkalinity inthe Indian Ocean (068 090 S, 508 550 E) (source Jeandel

et al., 1996).


Figure 10 Profiles of Zn, Cd2þ, Ni, Cu, Si(OH)4, NO32, and PO4

2 with depth in the North Pacific (after Bruland,1980).


the zinc removed from the surface ocean is boundup in POM.

Because zinc is required for the growth ofphytoplankton, its availability affects the biologi-cal pump. Although zinc limitation of anentire phytoplankton community has never beendemonstrated, levels of dissolved zinc are oftenlow enough to limit many taxa (Morel et al., 1994;Sunda and Huntsman, 1995b; Timmermans et al.,2001). Zinc is an integral part of the enzyme,carbonic anhydrase (Morel et al., 1994), whichhelps maintain an efficient supply of CO2 toRubisco.

The impact of low concentrations of zinc onphytoplankton growth varies. Some phytoplank-ton substitute cobalt for zinc in many enzymes(Price and Morel, 1990; Sunda and Huntsman,1995b; Timmermans et al., 2001) and canmaintain maximal growth rates at low levels ofzinc. Calcification aids in the acquisition of DIC,so calcareous phytoplankton, such as the cocco-lithophorids, are not as dependent on carbonicanhydrase (and therefore zinc) to maintain highrates of carbon fixation (Sunda and Huntsman,1995b). Thus, while low levels of zinc may notcurtail overall levels of primary production, theymay shift the phytoplankton community structureaway from the diatoms and towards the cocco-lithophorids (Morel et al., 1994; Sunda andHuntsman, 1995b; Timmermans et al., 2001),which will greatly impact the ratio of organic C toCaCO3 of particles sinking to the deep sea.

6.04.3.2.3 Cadmium

Like barium and zinc, cadmium also shows anutrient-like distribution in the ocean (Figure 10;Bruland, 1980; Loscher et al., 1998), more closelymirroring those of the labile nutrients, NO3

2 andPO4

2, than those of Si(OH)4 and alkalinity.Cadmium is taken up by phytoplankton andincorporated into organic material, accountingfor the similarity of its profile to that of NO3

2 andPO4

2. Cadmium may also be adsorbed onto thesurfaces of phytoplankton (Collier and Edmond,1984).

Cadmium is taken up by phytoplankton slightlypreferentially to PO4

2 (Loscher et al., 1998;Elderfield and Rickaby, 2000). Waters with lowPO4

2 concentrations thus have lower Cd : PO42

ratios (0.1 nmol Cd per mmol PO42) than waters

with higher PO42 concentrations where Cd : PO4

2

may approach 0.4 nmol Cd per mmol PO42

(Elderfield and Rickaby, 2000). Additionally,surface water Cd : PO4

2 ratios drop over thedevelopment of the spring bloom in the SouthernOcean (Loscher et al., 1998).

Cadmium also regenerates preferentially fromdecomposing particles (Knauer and Martin, 1981;

Collier and Edmond, 1984). More labile com-ponents of decaying particles have higher Cd : Pratios than bulk decaying particles (Knauer andMartin, 1981). Box models of cadmium andphosphorus cycling also require enhanced regen-eration of cadmium from particles for the replica-tion of observed cadmium distributions in theocean (Collier and Edmond, 1984).

Cadmium is generally considered to be toxic toorganisms and how the marine phytoplanktonutilize their cadmium is unknown. Cadmium maysubstitute for zinc in carbonic anhydrase at timeswhen zinc is limiting (Price and Morel, 1990;Lane and Morel, 2000). It is possible thatcadmium may play a role in polyphosphatebodies, a form of cellular storage of phosphorusthat has been shown to contain significantquantities of elements such as calcium, zinc, andmagnesium (Ruiz et al., 2001).

The similarity between cadmium profiles andnutrient profiles has been used as a means ofreconstructing past patterns of primary productionand nutrient cycling. Several attempts have beenmade to reconstruct phosphate concentrationsfrom the Cd : Ca ratio of foraminifera (Boyle,1988; Elderfield and Rickaby, 2000). Foramini-fera substitute Cd2þ for Ca2þ in the lattice of theircalcite tests, and the ratio of Cd : Ca incorporationvaries with the Cd : Ca ratio of seawater (Boyle,1988). The Cd : Ca ratio of foraminifera in theSouthern Ocean suggests that phosphorus insurface waters was not as heavily utilized byphytoplankton during the last glacial maximum(LGM), suggesting lower levels of primaryproduction relative to nutrient flux into theeuphotic zone at that time (Elderfield andRickaby, 2000).

6.04.3.2.4 Iron

Of all the trace elements whose distributions areaffected by the biological pump, iron is the one thathas the most profound impact on the workings of thebiological pump. Iron plays a key role in many of thecrucial enzymes in biological systems, such assuperoxide dismutase, ferredoxin, and nitratereductase (Geider and LaRoche, 1994) that evolvedat a time when the oceans were low in oxygen and,therefore, high in dissolved iron. As a result marinephytoplankton have a heavier demand for ironrelative to the present-day availability of dissolvediron in the ocean. Growth of phytoplankton and ratesof photosynthesis are frequently limited by the lackof iron in surface waters (Martin and Fitzwater,1989).

One of the major inputs of iron to the oceancomes from the dissolution of Fe(II) from wind-borne continental dust deposited on the surface ofthe ocean (Zhuang et al., 1990). In oxygenicenvironments, such as surface ocean waters, Fe(II)


will quickly be oxidized to the insoluble formFe(III) and removed from seawater. Fe(II) is alsotaken up by phytoplankton as well as complexed byligands exuded into the water by marine organismsto prevent its precipitation as Fe(III).

Profiles of dissolved iron in seawater show theinfluence of both biotic and abiotic processes. Atstations in the Northeast Pacific, dissolved ironconcentrations are low in surface waters, reflect-ing biological uptake. Iron concentrations alsoshow peak values at depth, corresponding tothe oxygen minimum zone (Martin and Gordon,1988), suggesting the abiotic reduction of Fe(III)back to the soluble Fe(II).

Vast tracks of the ocean, such as the equatorialPacific, the Northeast Pacific subarctic, theSouthern Ocean, and even parts of the Californiaupwelling zone do not have sufficient supplies ofiron to fully support phytoplankton growth (Martinand Fitzwater, 1988; Martin et al., 1994; Coaleet al., 1996; Hutchins and Bruland, 1998; Boydet al., 2000). In these areas, macronutrients such asnitrogen and phosphorus are rarely depleted.Attention has turned to these HNLC areas as siteswhere further primary production (and the associ-ated drawdown of atmospheric CO2) could occur.Increased supplies of dust stimulating the biologi-cal pump in HNLC regions may be responsible forlow atmospheric CO2 concentrations during gla-cial times (Martin and Gordon, 1988; Watson et al.,2000). Artificially stimulating the biological pumpby seeding HNLC areas with chelated iron has beenproposed as a means of pumping the 90 matm ofCO2 in the atmosphere (put there by humans) intothe deep sea, although there is not much consensusas to the effectiveness of such an endeavor(Chisholm et al., 2001).

6.04.4 QUANTIFYING THE BIOLOGICALPUMP

There are many different ways to quantify thebiological pump. Total levels of carbon fixation insurface waters may be estimated in bottleincubations from the uptake of 14CO2 by phyto-plankton (Steemann Nielsen, 1952) or from thedeviation of oxygen isotopes from their terrestrialmass fractionation line (Luz and Barkan, 2000).At the other end of the biological pump,sedimentary accumulation rates of organic carbonmay be measured. In between, the impact of thefeeding and vertical migration activities of mid-water organisms on the flux of particles may beinvestigated (e.g., Steinberg, 1995; Dilling et al.,1998). However, because of the current interest inthe CO2 pumping capacity of the biological pump,we concentrate here on the methods used toestimate export production and particle flux fromthe euphotic zone.

Particle flux is frequently extrapolated from themeasurement of new production in surface waters(Dugdale and Goering, 1967). Export of POM outof surface waters or into the deep sea may also beestimated directly through its collection in sedi-ment traps (e.g., Martin et al., 1987; Anita et al.,2001). Export or sedimentation of POM may alsobe estimated from disequilibria between twonuclides (e.g., 238U and 234Th, and 230Th and231Pa) that are scavenged by particles of differentdegrees (Buesseler et al., 1992; Kumar et al.,1993; Francois et al., 1997).

Some methods for quantifying the biologicalpump focus the relationship between the biologi-cal pump and CO2. POC flux measurements orestimates of nutrient removal from surface watersmay be used in conjunction with various oceanmodels to estimate the impact of the biologicalpump on atmospheric concentrations of CO2

(Sarmiento and Toggweiler, 1984; Sarmientoand Orr, 1991). Others have focused on theimportance of the ratio of POC to CaCO3 to thesequestering of CO2 in the ocean (Anita et al.,2001; Buitenhuis et al., 2001).

6.04.4.1 Measurement of New Production

The method in most widespread use forquantifying the biological pump hinges upon theideas that the surface ocean is at steady state onannual timescales with respect to the nitrogenbudget and that nitrogen predominantly limitsphytoplankton growth in the ocean. In such asystem, the amount of productivity that isexported from the euphotic zone must be equalto the amount of productivity that is fuelled by theinput of allocthonous or “new” nitrogen to theeuphotic zone (Dugdale and Goering, 1967).According to this definition, “new production” isone that is supported from dissolved nitrogenupwelled into the euphotic zone or fixed from N2

into PON, and export production is taken to beequal to new production. For the sake ofmeasurement, the above definition of new pro-duction has been simplified even further. Experi-mentally, new production is taken to be equal tothe production that uses NO3

2 (and NO22 also but

more in theory than in practice) as its nitrogensource, as opposed to NH4

þ or any of the organicforms of nitrogen.

It should be pointed out that measurement ofnew production is the measurement of themaximum amount of productivity that can beexported without running the system down withrespect to the annual supply of nutrients to theeuphotic zone. It may not always be appropriate toassume that new production and export productionare equal; the fluxes will be equal only in systemsthat are not evolving. The validity of theassumption that new production equals export

Quantifying the Biological Pump 99

production depends on to what degree and overwhat timescale the nitrogen cycle in surfacewaters is at steady state.

At the moment the nitrogen cycle in the ocean isnot at steady state on the decadal timescale. Inthe last few decades, the N : P ratio of many oceanicwaters has changed (Pahlow and Riebesell, 2000;Emerson et al., 2001) and near-shore waters haveshifted towards silicon limitation from nitrogenlimitation (Conley et al., 1993) due to anthropo-genic inputs of DIN to these systems and anincrease in the extent of nitrogen fixation. On verylong timescales, the nitrogen cycle is not at steadystate either, but responds to changes in the oceanicinventory of phosphorus, which ultimately governsthe rate of nitrogen fixation (Tyrrell, 1999). Theseimbalances may not be large enough on a yearlytimescale to affect the estimate of exportproduction from new production, but noassessment of this has been made.

Another assumption open to question is that therate of NO3

2 uptake adequately represents the rateof uptake of new forms of nitrogen. Thissimplification has come about for two reasons.The first is that the uptake rate of NO3

2 byphytoplankton can be measured with reasonableease by tracking the uptake of 15N-labeled NO3

2

(Dugdale and Goering, 1967). The second is thatNO3

2 is not produced in the euphotic zone to anysignificant degree and so its presence there canonly be as a result of upwelling or atmosphericdeposition.

The form of DIN released during the death,decay, and grazing of phytoplankton is NH4

þ,which is also the most easily utilizable form of DINto phytoplankton. Oxidized forms of DIN, such asNO3

2 and NO222, must be reduced to NH4

þ bynitrifying bacteria such as Nitrosomonas, Nitro-bacter, Nitrospira, and Proteobacteria (Zehr andWard, 2002) prior to assimilation by phytoplank-ton. The classical view is that nitrification does notoccur in the euphotic zone due to the inhibition ofnitrifying bacteria by light (Zehr and Ward, 2002).Eukaryotic phytoplankton are also thought tooutcompete nitrifying bacteria for the supplies ofNH4

þ in surface waters. Thus, NO32 found in the

euphotic zone must have had its origins outside ofthe euphotic zone, in deeper waters, in rivers oragricultural runoff, or from atmospheric depositionand is taken as the sole representative of newnitrogen in the euphotic zone (Dugdale andGoering, 1967).

Of course, NO32 is not likely to be the only form

of allocthonous nitrogen in the euphotic zone.Concentrations of NH4

þ in upwelled water are notzero, although they are much lower than those ofNO3

2. DON may also serve as a significant sourceof new nitrogen to the euphotic zone.

There is one last word of caution concerning theuse of NO3

2 uptake to estimate the transport of CO2

to depth via the biological pump. Measurement ofnew production divulges no information concerningthe depth of decomposition of the POM formed orthe ratio of POC to CaCO3 of the exported particles(Anita et al., 2001). For instance, CO2 from materialdecomposed beneath the euphotic zone but abovethe maximum depth of winter mixing will beventilated straight back out the atmosphere.CaCO3 formation, a feature that is not common toall phytoplankton, diminishes the efficiency of CO2

drawdown with primary production (see below;Buitenhuis et al. (2001)).

6.04.4.2 Measurement of Particle Flux

Means more direct than the measurement ofnew production exist for the estimation of particlefluxes into the deep. Moored or free-floating trapsmay be used to collect sinking particles (e.g.,Martin et al., 1987). Alternatively, particle fluxmay be estimated from particle-reactive nuclides(e.g., Buesseler et al., 1992). Particle flux mayalso be estimated from the consumption of oxygen(associated with the decomposition of sinkingPOM) in waters below the surface layer (e.g.,Jenkins, 1982).

6.04.4.2.1 Sediment traps

The collection of particles in sediment traps,while perhaps the most direct way of measuringthe sinking flux of POM, is a method not free froma certain amount of controversy. Sediment trapsboth over-collect and under-collect particles.Comparison of 234Th accumulating in a suite ofsediment traps with 234Th fluxes expected fromU–Th disequilibria in the upper 300 m of theocean suggested that the particle collectionefficiency of these traps ranged from 10% to1,000% (Buesseler, 1991). Traps deployed in thedeep ocean also show a considerable variability intrapping efficiencies (e.g., Scholten et al., 2001).Despite the magnitude of these biases, there is nogenerally applied method for correcting fluxesusing particle-reactive isotopes (Anita et al.,2001).

Zooplankton actively swimming into sedimenttraps also serve as a source of error in fluxmeasurements. It is impossible to differentiatethese “swimmers” from zooplankton that havesettled passively into the cup as part of the sinkingPOM flux. Swimmers may constitute as much asa quarter of the POC collected by the trap(Steinberg et al., 1998) and are generally removedfrom trap material prior to analysis. This intro-duces minimal error into the trap estimates of POCflux, as detrital zooplankton likely only comprise


,2% of the total organic matter sinking into thetraps (Steinberg et al., 1998).

6.04.4.2.2 Particle-reactive nuclides

Radionuclides in the uranium decay series serveas useful tracers of particle flux. One type of thesetracers consists of a soluble parent nuclide and aparticle-reactive daughter. These soluble nuclide–particle-reactive pairs include 238U – 234Th,234U–230Th, and 235U–231Pa. The half-life of theparent exceeds the mixing time of the ocean andits distribution throughout the ocean is uniform.Once the soluble parent isotope decays to theparticle-reactive daughter, the daughter is sca-venged onto particulate material.

In systems with no particle scavenging, theactivities of the parent and daughter nuclide willbe in secular equilibrium. What is seen instead isthat the activities of 234Th, 230Th, and 231Pa arelower in surface waters than those of their parents(Figure 11). The difference in the activities ofparent and daughter is a measure of the uptake ofdaughter onto particles (Buesseler et al., 1992).With the help of a model of particle scavenging,fluxes of the particle-reactive daughter may beestimated from its vertical distribution (e.g., Coaleand Bruland, 1985; Buesseler et al., 1992). If theratio of the particle-reactive nuclide to POC orPON is known, then the calculated flux of nuclidecan be converted to an estimate of particle flux(Buesseler et al., 1992).

Relative estimates of particle flux may also bemade from the ratio of two particle-reactivenuclides, such as 230Th and 231Pa, which arescavenged onto particles to different degrees.The half-lives of these isotopes are muchlarger than their residences times in the ocean(,104 yr versus tens to hundreds of years), andthus there is no significant radioactive decay thatoccurs in the water column. The extent of thescavenging of 231Pa, which is not as particle

reactive as 230Th, is highest in areas of highparticle flux (Anderson et al., 1990). Thus,sediments in high flux areas exhibit 231Pa/230Thratios in excess of the initial production ratio of0.093, and sediments accumulating slowly exhibitratios less than 0.093 (Anderson et al., 1990).231Pa/230Th ratios have been used to infer changesin productivity and sediment accumulation ratesbetween the present-day interglacial and theLGM, ,2 £ 104 yr ago (Kumar et al., 1993;Francois et al., 1997).

6.04.4.2.3 Oxygen utilization rates

The distribution of oxygen in ocean waterscontains information about primary production.For example, the amount of excess oxygenpresent in the seasonal thermocline in the Pacificwas long ago used to suggest that 14C-basedestimates of primary production were severelyunderestimating levels of primary production inthe ocean (Shulenberger and Reid, 1981). Oxygendeficiencies in deeper waters have been used toestimate levels of export production (Jenkins,1982).

The supplies of oxygen to waters below theeuphotic zone are primarily physical: advectionand mixing. Removal of oxygen from these waterstakes place through the oxidation of organicmatter (Equations (2) and (3)):

O2 þ CH2O ! CO2 þ H2O ð2Þ

2O2 þ NH3 þ OH2 ! NO23 þ 2H2O ð3Þ

By measuring rates of ventilation and the degreeof oxygen undersaturation in deeper waters, anestimate of the rates of oxygen utilization (OUR)may be made and integrated to yield the totalamount of oxygen consumed beneath the euphoticzone each year (Jenkins, 1982). From this numbera flux of POC and PON may be calculated.Estimates of export production based on OURs arereasonably in line with the amount of newproduction that could be supported from measuredfluxes of NO3

2 into surface waters (Jenkins, 1988).One advantage of using the OUR method for

estimating export fluxes is that it integrates overlarger spatial and temporal scales than doestimates based on sediment traps and nuclidefluxes. Also unlike the estimates of new pro-duction based on NO3

2 uptake, OURs are directlycoupled to the recycling of CO2 via particledecomposition and thus a more direct measure ofthe impact of the biological pump on atmosphericCO2. In practice, however, the measurement ofNO3

2 uptake is less technically challenging and iscarried out much more frequently than areestimates from nuclides and OURs.

Figure 11 Profiles of 234Th (circles) and 238U(diamonds) in the upper ocean (after Buesseler, 1991).

Quantifying the Biological Pump 101

6.04.5 THE EFFICIENCY OF THEBIOLOGICAL PUMP

6.04.5.1 Altering the Efficiency of the BiologicalPump

There is much talk concerning the “efficiency”of the biological pump. Is it pumping as muchcarbon to the deep sea as it could be? The generalconsensus is that it is not operating at its fullcapacity, and this is generally meant to imply thatglobally, the nitrate flux into the euphotic zone isnot fully consumed in support of marine primaryproduction. For example, Broecker (1982) hassuggested that if all of the nutrients supplied toocean surface waters were consumed by phyto-plankton, the atmospheric CO2 content woulddrop by ,130 ppmv.

A simplistic estimate of how much more CO2

the biological pump could draw out of theatmosphere gives a false impression of ourunderstanding of the system. One hint that thereis a considerable decoupling of primary pro-duction and nutrient drawdown from levels ofexport production is the regional variability in theratio of export to total production. Levels ofexport production are higher in more highlyproductive areas than in low-productivity areas,as would be expected. However, a greaterproportion of the total primary production isexported from mesotrophic and eutrophic regionsthan from oligotrophic areas of the ocean. Theratio of export production to total production maybe as low as 0.12 in oligotrophic areas but as highas 0.4 in highly productive regimes (Anita et al.,2001).

There is a lesson to be learnt here with regard tospurring the biological pump into action. Most of thetalk regarding increasing the efficiency of thebiological pump focuses only on productivity andnutrient drawdown, but there is much more to thebiological pump than nutrient drawdown. The mainquestion asked is: Can we stimulate an increase inprimary production (and by extension exportproduction) by adding iron to the ocean? From thepoint of view of ascertaining whether or notincreased glacial dust could have stimulated thebiological pump and could have been responsiblefor the lower levels of CO2 then present in theatmosphere, this is an appropriate question. Butfrom the point of view of effectively sequesteringanthropogenic carbon in the deep sea via thebiological pump, this is a very narrow approach.Carbon sequestration via the biological pump reallycannot be considered as a practical activity until wehave a better predictive understanding of not onlyprimary production but also particle aggregationand sinking, zooplankton grazing and microbialhydrolysis, and organic carbon to CaCO3 pro-duction and rain rates. These are the factors that

control both the proportion and total amount oforganic carbon pumped to the deep sea.

6.04.5.1.1 In HNLC areas

The biological pump in HNLC areas is notoperating at full efficiency on at least two counts.Phytoplankton growth is curtailed by the lack ofavailability of trace metals such as iron, and soconcentrations of the major nutrients, nitrogen andphosphorus, are not drawn down and carbonfixation does not occur to its maximum possibleextent (Martin and Fitzwater, 1988). In addition,iron limitation heavily impacts the larger phyto-plankton, such as diatoms, that are important toparticle flux. When a phytoplankton community isreleased from iron limitation, diatom growth isstimulated more strongly than the growth of theother phytoplankton (Cavender-Bares et al., 1999;Lam et al., 2001).

Given all of this, the addition of iron to HNLCwaters should result in higher levels of carbonfixation, increased growth of diatoms, localdrawdown of CO2, and enhanced export of carbonto the deep sea. Iron-addition experiments inbottles and in situ on the mesoscale unequivocallysupport the first two points. Addition of iron toHNLC waters results in madly blooming phyto-plankton (Martin and Fitzwater, 1989; Martinet al., 1994; Coale et al., 1996; Boyd et al., 2000;Strass, 2002). Chlorophyll concentrations mayquadruple (e.g., Strass, 2002) and carbon fixationrates may triple (e.g., Coale et al., 1996) over thefirst few days after iron addition (Figure 12).

Alternatively, support for the drawdown ofCO2 following iron addition has been mixed. InIronEx-I, the first mesoscale iron experiment, ironaddition did not result in a marked drop of CO2 con-centrations in surface waters of the equatorialPacific (Watson et al., 1994). Some CO2 was drawndown during IronEx-II, also in the equatorialPacific, but the amount removed from surfacewaters was still not enough to prevent theserecently upwelled, high CO2 waters from out-gassing CO2 to the atmosphere (Coale et al., 1996).The Southern Ocean, however, is a better ecosys-tem than the equatorial Pacific for stimulating CO2

drawdown with iron. Both large-scale iron-addition experiments that have been carried outthere (SOIREE and EisenEx-1) produced a signifi-cant lowering (25–30 matm) of the CO2 content ofsurface waters (Boyd et al., 2000; Watson et al.,2000; Strass, 2002).

The one key critical point that all of the iron-enrichment experiments have failed to show is anincrease in the export of carbon to the deep sea. Eventhe Southern Ocean experiments, where a CO2

drawdown occurred, did not result in an increase inexport flux. Sediment traps set out at a depth of


100 m (below the fertilized patch) in the SOIREEexperiment collected a similar amount of POC totraps stationed outside of the patch (Boyd et al.,2000). In the EisenEx-1 experiment, there was noincrease in sinking POC over the three-weekwindow of the experiment (Strass, 2002).

There are many possible reasons for the lack of anobserved increase in export flux in the mesoscale

iron-addition experiments. The first is that there isno increase in export flux, but instead an increase inthe flow of carbon through the microbial loop(microzooplankton and bacteria) which wouldresult in the regeneration of CO2 and nutrients inthe euphotic zone. Microzooplankton biomass andbacterial activities were seen to increase inmany of the mesoscale iron-addition experiments

Figure 12 Responses to Fe addition to HNLC waters: chlorophyll, carbon fixation, NO32, and PCO2

(after Boyd et al.,2000; Watson et al., 2000).

The Efficiency of the Biological Pump 103

(e.g., Boyd et al., 2000; Strass, 2002), suggestingthat some fraction of the carbon fixed through ironaddition was being shunted into the microbial loopinstead of being pumped into deeper waters.Alternatively, the sediment traps used may simplyhave not been deployed for long enough or in theright location to catch the carbon sinking out as aresult of the iron-induced bloom.

6.04.5.1.2 Through changes in communitycomposition

Changes in community composition shouldprofoundly affect the efficiency of the biologicalpump. Shifting productivity from small cells,such as photosynthetic picoplankton, which donot efficiently sink, to large cells, such as diatoms,capable of aggregating into particles that sink athundreds of meters a day will result in thepumping of more carbon to the deep. However,there is finite amount of productivity that may beexported without running the nutrients in thesystem down to zero. In a steady-state system,shifting the community structure cannot increasethe total amount of export production above thelevel of new production. However, a shift incommunity structure may export carbon deeperinto the water column before it decomposes,thus delaying the return of that carbon tothe surface ocean by a significant number ofyears.

A shift in community composition may also beimportant to the biological pump if it is from acalcareous to a noncalcareous phytoplankton, asprecipitation of CaCO3 diminishes the ocean’sability to hold dissolved CO2. DIC (

PCO2) is

present in seawater as several species, dissolvedCO2, carbonic acid, and the dissociated forms,bicarbonate ion, and carbonate ion:

XCO2 ¼ H2COp

3 þ HCO23 þ CO22

3 ð4Þ

(where H2CO3* represents the sum of dissolved

CO2 and carbonic acid). The ratio of the threespecies varies largely with pH, with the undisso-ciated forms dominating only at low pH andHCO3

2 favored at the typical seawater pH of8.3 (Zeebe and Wolf-Gladrow, 2001). Thesaturation concentration of H2CO3

* is controlledin large part by the amount of HCO3

2 and CO322

present in solution. Removal of HCO32 or

CO322 during CaCO3 formation results in a

lowering of the saturation concentration ofdissolved CO2, and therefore the outgassing ofCO2 from solution to the atmosphere.

Precipitation of CaCO3 both produces CO2

(Equation (5)):

Ca2þ þ 2HCO23 ! CaCO3 þ H2O þ CO2 ð5Þ

and lowers the alkalinity (approximately equal to2[CO3

22] þ [HCO32]) of the surface waters (Zeebe

and Wolf-Gladrow, 2001). Loss of alkalinitydecreases the capacity of surface waters to playhost to dissolved CO2. In the surface ocean, underan atmosphere with a CO2 partial pressure of350 matm (19 matm less than the 2001 values;Keeling and Whorf, 2001), the precipitation of1 mol of CaCO3 results in the release of 0.6 molof CO2 concentrations (Ware et al., 1992;Frankignoulle et al., 1994).

The coccolithophorid, Emiliania huxleyi, con-tains, on average, 0.433 mol of CaCO3 (present asscales covering the cell) for every mole of organiccarbon (Buitenhuis et al., 2001). Thus, for everymole of CO2 fixed into organic matterby coccolithophorids such as E. huxleyi, roughly0.26 mol of CO2 will be released due to theformation of CaCO3. A shift, therefore, fromexport production being carried out by cocco-lithophorids to a noncalcareous phytoplanktonsuch as diatoms will result in an increaseddrawdown of carbon for the same amount ofprimary production. The fact that carbon fixationby phytoplankton that calcify is only 75% aseffective at removing CO2 as carbon fixation doneby phytoplankton that do not precipitate CaCO3

complicates estimates of CO2 drawdown from thesurface ocean POC export flux (Anita et al., 2001).

6.04.5.1.3 By varying the C : N : P ratios ofsinking material

The C : N : P ratios of sinking POM are notfixed; phosphorus and nitrogen are preferentiallyregenerated from sinking POM, which allows forthe sequestering of more carbon in the deep oceanper unit nitrogen or phosphorus than if Redfieldratios remain unaltered during the decompositionof organic matter. It was remineralization ratioscalculated from DIC and nutrient concentrationsalong isopycnals that initially suggested that theC : N of sinking particles might be higher than thevalue of 6.6 proposed by Redfield (Takahashiet al., 1985). Sediment trap material also showedthat the C : N and C : P ratios of sinking particlesincrease with depth, from near Redfield values of6.6 and 106 at the base of the euphotic zone to,11 and 180 by 5,000 m (Martin et al., 1987).

Since the above initial observations, evidencehas mounted for the preferential remineralizationof nitrogen and phosphorus out of POM relative tocarbon. Ocean models with continuous verticalresolution support the preferential release ofnitrogen and phosphorus from sinking POM inline with the estimates of Martin et al. (Shaffer,1996). Bacteria have been shown to more rapidlydegrade PON than POC (Verity et al., 2000).Increases in the C : N and C : P of sinking particles


have also been observed at station ALOHA nearHawaii (Christian et al., 1997). Differences in theC : N ratio of suspended POC (6.4) and lessbuoyant POC trapped at the pycnocline (8.6) inthe Kattegat, just east of the North Sea, furthersuggest that nitrogen is remineralized morerapidly than carbon at significant levels even inthe euphotic zone (Olesen and Lundsgaard, 1995).

6.04.5.1.4 By enhancing particle transport

While sediment trap evidence supports the ideathat the flux of organic carbon (JC org) to depth (z)can be estimated from levels of primary pro-duction (PP) (Equation (6)):

JC org ¼ 0:1PP1:77z20:68 ð6Þ

the correlation between POC flux and overlyinglevels of primary production (R 2 ¼ 0.53; Anitaet al., 2001) is not strong. The ratio of POC flux (at125 m) to primary production also shows greatvariability from region to region, ranging from0.08 to 0.38 in the Atlantic Ocean alone (Anitaet al., 2001).

There are many reasons that an increaseddownward transport of POC out of the euphoticzone does not directly follow from an increase inproductivity. Export fluxes are also controlled bythe packaging of smaller more numerous particlesinto larger less numerous aggregates and by therate at which the resulting aggregates sink. Thedepth to which POC exported from the euphoticzone is transported must be determined by thebalance between the carbon content of the particle,the rate of microbial hydrolysis, and the sinkingrate of the particle. Carbon fixed by marinediatoms, which are large and adept at aggregating,stands a better chance at being exported to thedeep than carbon fixed by very small, unsinkingpicoplankton.

The mode of productivity is also crucial to thesinking flux of POC. Systems exporting POC inpulses (e.g., following the periodic formation ofphytoplankton blooms in temperate and highlatitude areas) export a much higher proportionof their total production than do systems, such asthe oligotrophic central gyres, where carbon fixationis a more static, steady process (Lampitt and Anita,1997). This results partly from the influence ofparticle concentration on aggregation and partlyfrom zooplankton population dynamics. In temper-ate areas, zooplankton reproduction cannot beginuntil after the onset of the spring phytoplanktonbloom, leading to a lag in the increase inzooplankton population (and subsequently in thegrazing pressure exerted by zooplankton) behindthat of the increase in phytoplankton numbers(Heinrich, 1962) allowing for the aggregation and

sinking of intact phytoplankton from the euphoticzone. In lower latitude, oligotrophic-area standingstocks of phytoplankton and zooplankton are not asdecoupled in time and there is no clear windowwithin which a high density of phytoplankton cellsmay aggregate, escape predation, and sink to thedeep.

6.04.6 THE BIOLOGICAL PUMP IN THEIMMEDIATE FUTURE

What may we expect out of the biological pumpin the future? We cannot currently predict howocean biology will respond to climate warming(Sarmiento et al., 1998), but there are otherquestions that we may begin to ask. Will, forinstance, the biological pump respond to the risein surface ocean concentrations of DIC tied to therise in atmospheric CO2 fuelled by anthropogenicemissions? What impact will inputs of agriculturalfertilizers have on the biological pump? What rolewill artificial stimulation of the biological pumpthrough deliberate ocean fertilization play in thesequestration of excess CO2?

6.04.6.1 Response to Increased CO2

Will the biological pump respond to theincrease CO2? Increasing concentrations of CO2

do not appear to have the significant impact on theC : N or C : P ratios of phytoplankton (Burkhardtet al., 1999) that is necessary if CO2 is to stimulateproductivity in the ocean. However, possibly therate of carbon-rich exudation by phytoplanktonand the production of TEP will rise withincreasing CO2 concentrations (Engel, 2002). Itmay also be possible for CO2 levels to increase theproportion of carbon diverted into the biologicalpump at the expense of grazing and microbial foodwebs. Riebesell et al. (1993) suggested that carbonfixation by phytoplankton may be rate limited bythe diffusive flux of CO2 to the cell which deliversCO2 too slowly relative to NO3

2 and PO42 to take

up carbon, nitrogen, and phosphorus in Redfieldproportions.

The diffusive flux of CO2 to the surface of aphytoplankton cell (FCO2

) is controlled not only bythe diffusivity of CO2 (D) and the radius of the cell(r), but the concentration gradient of CO2 betweenthe aqueous medium and the interior of the cell(Ce–Ci) and the rate constant for the conversion ofHCO3

2 to CO2 (k0) as the carbonate systemequilibrates following CO2 removal (Equation(7)):

FCO2¼ 24prD 1 þ

rffiffiffiffiffiffiD=k0

p

� �ðCe 2 CiÞ ð7Þ

The Biological Pump in the Immediate Future 105

If the diffusive flux of CO2 is controlling thedelivery of CO2 to the carbon-fixing enzyme,Rubisco, then an increase in the dissolved CO2

content of seawater should stimulate phytoplank-ton growth rates. In particular, an increase indissolved CO2 concentrations should stimulategrowth rates in the larger (i.e., aggregate forming)size classes and thus should also result in ashunting of carbon out of the mouths of zoo-plankton and into the biological pump (Riebesellet al., 1993).

It is not clear that the acquisition of carbon isthe rate-limiting step for photosynthesis even inthe larger phytoplankton cells. Phytoplanktonare not limited to the passive diffusion of CO2

to the cell surface for carbon acquisition.HCO3

2 may be taken up directly by phyto-plankton and used as a source of CO2 forphotosynthesis (Korb et al., 1997; Nimer et al.,1997; Raven, 1997; Tortell et al., 1997). Theconsiderably greater abundance of HCO3

2 thandissolved CO2 in seawater would imply thatcarbon-fixation rates of marine phytoplanktonare not CO2 limited, and increasing concen-trations of CO2 will not have an impact onrates of primary production.

6.04.6.2 Response to Agricultural Runoff

Inputs of agricultural fertilizers are havingmore than one impact on the biological pump.A shift in the NO3

2: PO42: Si(OH)4 of natural

waters is causing a shift in the phytoplanktoncommunity structure that should impact thebiological cycling of carbon in aquatic systems(Conley et al., 1993). Additionally, a recentshift in the C : N : P ratios of deeper waters andan increase in export production have beenobserved for the northern hemisphere oceans(Pahlow and Riebesell, 2000).

6.04.6.2.1 Shift towards silicon limitation

Since the 1850s, changes in land use patterns,human population density, and extent of the useof fertilizers have resulted in an increased flux ofNO3

2 and PO42 to rivers, lakes, and the coastal

ocean (Conley et al., 1993) and even to the openocean through atmospheric deposition, forexample, of anthropogenic nitrous oxides (Pahlowand Riebesell, 2000). At the same time, largefreshwater systems, such as the North AmericanGreat Lakes, and coastal areas, such as theAdriatic and Baltic Seas and the MississippiRiver plume, have been shifting away from nitro-gen or phosphorus limitation and towards siliconlimitation (Conley et al., 1993; Nelson andDortch, 1996). The extra productivity fuelled onthe extra nitrogen and phosphorus has resulted in

the removal of silicon from these systems and ashift in the phytoplankton community structureaway from diatoms. Given that rivers are signifi-cant sources of new NO3

2, PO42, and Si(OH)4 to

the global ocean, this effect is expected to spreadfurther throughout the sea.

A shift in community structure associated withsilicic acid depletion may sharply reduce theamount of carbon delivered to deep waters andsediments via the biological pump. Diatoms,which are the only major phytoplankton grouprequiring silicic acid, are relatively large andnotable for aggregating and sinking. Diatoms feeda greater portion of the organic matter theyproduce into the biological pump than do mostother classes of phytoplankton. Currently, diatomsperform more than 75% of the primary productionthat occurs in high nutrient and coastal regions ofthe ocean (Nelson et al., 1995), the exact areas thatwill be impacted by this input of additional NO3

2

and PO42 and the exact areas where carbon tends

to make it down into the sediments.One possible further impact of the decline of the

diatoms lies in the identity of their probablysuccessor. If a CaCO3-producing phytoplankter,such as coccolithophorids, steps in to utilize theNO3

2 and PO42 the silicon-starved diatoms leave

behind, the CO2 pumping efficiency of thebiological pump will decline (Robertson et al.,1994) even if the same amount of organic carboncontinues to be removed to the deep sea andsediments each year.

6.04.6.2.2 Shifts in export production anddeep ocean C : N : P

There is evidence for anthropogenic pertur-bations increasing the biological pumping ofcarbon into deep waters. In both the North Pacificand North Atlantic, for example, deep-water N : Pratios have increased since the early 1950s(Pahlow and Riebesell, 2000) possibly due toatmospheric deposition of anthropogenic nitrogenand a subsequent shift towards phosphoruslimitation in these areas. Concomitant with therise in N : P is an increase in the apparent oxygenutilization (AOU) in both the North Atlanticand North Pacific, which suggests that levels ofexport production have also increased. This isestimated to have resulted in the increased oceanicsequestration of 0.2 Pg C yr21 (Pahlow andRiebesell, 2000).

6.04.6.3 Carbon Sequestration via OceanFertilization and the Biological Pump

There have been many calls to sequester anthro-pogenic CO2 in the deep ocean by stimulating


primary production through the addition ofliterally tons of iron to HNLC surface waters.Patents have been taken out on the idea (e.g.,Markels, 2001) and, counter to the recommen-dations of the American Society of Limnologyand Oceanography (http://www.aslo.org/policy/docs/oceanfertsummary.pdf), several companieshave been established to dispense carbon creditsto industries willing to pay for ocean fertilization(Chisholm et al., 2001). However, it remainsunclear whether ocean fertilization will eversuccessfully transport carbon in the deep sea,how long the transported carbon might remain outof contact with the atmosphere, and what side-effect large-scale fertilization will have onmarine geochemistry and ecology (Fuhrman andCapone, 1991; Peng and Broecker, 1991a,b;Sarmiento and Orr, 1991; Chisholm et al., 2001;Lenes et al., 2001).

A perusal of the literature suggests that carbonsequestration by iron fertilization is not a panaceafor the anthropogenic carbon emissions thatincreased atmospheric CO2 from 280 matm atthe start of the industrial revolution to 369 matmby December 2000 (Keeling and Whorf, 2001).Ocean models suggest that enhanced ocean uptakeof carbon with iron fertilization of the AntarcticOcean will at best draw down atmospheric CO2 by70 matm if carried out continuously for a century(Peng and Broecker, 1991a,b) and damp theannual anthropogenic CO2 input to the atmos-phere by less than 30% of current annual emissionlevels (Joos et al., 1991).

Large-scale iron fertilization will have side-effects. If iron fertilization resulted in a completedrawdown of the nutrients available in theSouthern Ocean, for example, the average O2

content of deep waters will drop by 4–12%(Sarmiento and Orr, 1991), with areas of anoxiacropping up in the Antarctic (Peng and Broecker,1991b) and Indian Oceans (Sarmiento and Orr,1991). Even small-scale patches of anoxia wouldhave a profound negative impact on the survivaland distribution of metazoan fauna in the oceanand alter the balance of microbial transformationsof nitrogen between reduced and oxidized phases(Fuhrman and Capone, 1991).

ACKNOWLEDGMENTS

The author thanks U. Passow for a helpfulreview, A. Engel, U. Passow, and A. Wischmeyerfor insightful conversation, A. Alldredge,R. Collier, and L. Dilling for promptly sendingreprints, A. Knoll and A. Yool for pointing out thelong-term silica cycle, W. Berelson, H. Elderfieldand N. McCave for helpful tips, and R. Bangerand S. Bishop for finding everything the authorcould not.

REFERENCES

Alldredge A. L. and Gotschalk C. C. (1989) Direct obser-vations of the mass flocculation of diatom blooms:characteristics, settling velocities and formation of diatomaggregates. Deep-Sea Res. 36, 159–171.

Alldredge A. L. and Jackson G. A. (1995) Aggregation inmarine systems. Deep-Sea Res. II 42, 1–7.

Alldredge A. L. and McGillivary P. (1991) The attachmentprobabilities of marine snow and their implications forparticle coagulation in the ocean. Deep-Sea Res. 38,431–443.

Alldredge A. L. and Silver M. W. (1988) Characteristics,dynamics, and significance of marine snow. Prog. Ocea-nogr. 20, 41–82.

Alldredge A. L., Passow U., and Logan B. E. (1993) Theabundance and significance of a class of large, transparentorganic particles in the ocean. Deep-Sea Res. 40,1131–1140.

Aluwihare L. I. and Repeta D. J. (1999) A comparison of thechemical characteristics of oceanic DOM and extracellularDOM produced by marine algae. Mar. Ecol. Prog. Ser. 186,105–117.

Amon R. M. W. and Benner R. (1994) Rapid cycling of high-molecular-weight dissolved organic matter in the ocean.Nature 369, 549–552.

Amon R. M. W., Fitznar H.-P., and Benner R. (2001) Linkagesamong the bioreactivity, chemical composition, and diage-netic state of marine dissolved organic matter. Limnol.Oceanogr. 46, 287–297.

Anderson R. F., Lao Y., Broecker W. S., Trumbore S. E.,Hofman H. J., and Wolfli W. (1990) Boundary scavenging inthe Pacific Ocean: a comparison of 10Be and 231Pa. EarthPlanet. Sci. Lett. 96, 287–304.

Anita A. N., et al. (2001) Basin-wide particulate carbon flux inthe Atlantic Ocean: regional export patterns and potential foratmospheric CO2 sequestration. Global Biogeochem. Cycles15, 845–862.

Arnosti C., Repeta D. J., and Blough N. V. (1994) Rapidbacterial degradation of polysaccharides in anoxic marinesystems. Geochim. Cosmochim. Acta 58, 2639–2652.

Asper V. L., Deuser W. G., Knauer G. A., and Lohrenz S. E.(1992) Rapid coupling of sinking particles fluxesbetween surface and deep ocean waters. Nature 357,670–672.

Azam F. and Volcani B. E. (1981) Germanium–siliconinteractions in biological systems. In Silicon and SiliceousStructures in Biological Systems (eds. T. L. Simpson andB. E. Volcani). Springer, pp. 69–93.

Bauer J. E., Williams P. M., and Druffel E. R. M. (1992) 14Cactivity of dissolved organic carbon fractions in the north-central Pacific and Sargasso Sea. Nature 357, 667–670.

Benner R., Pakulski J. D., McCarthy M., Hedges J. I., andHatcher P. G. (1992) Bulk chemical characteristics ofdissolved organic matter in the ocean. Science 255,1561–1564.

Berner R. A., Ruttenberg K. C., Ingall E. D., and Rao J. L. (1993)The nature of phosphorus burial in modern marine sediments.In Interactions of C, N, P and S Biochemical Cycles andGlobal Change (eds. R. Wollast, F. T. Mackenzie, and L.Chou). Springer, pp. 365–378.

Biddanda B. and Benner R. (1997) Carbon, nitrogen, andcarbohydrate fluxes during the production of particulate anddissolved organic matter by marine phytoplankton. Limnol.Oceanogr. 42, 506–518.

Bidle K. D. and Azam F. (1999) Accelerated dissolution ofdiatom silica by natural marine bacterial assemblages.Nature 397, 508–512.

Bidle K. D. and Azam F. (2001) Bacterial control of siliconregeneration from diatom detritus: significance of bacterialectohydrolases and specie identity. Limnol. Oceanogr. 46,1606–1623.

References 107

http://www.aslo.org/policy/docs/oceanfertsummary.pdf

http://www.aslo.org/policy/docs/oceanfertsummary.pdf

Billett D. S. M., Lampitt R. S., Rice A. L., and Mantoura R. F.C. (1983) Seasonal sedimentation of phytoplankton to thedeep sea benthos. Nature 302, 520–522.

Bishop J. K. B. (1988) The barite-opal-organic carbon associ-ation in oceanic particulate matter. Nature 332, 341–343.

Boyd P. and Newton P. (1995) Evidence of the potentialinfluence of planktonic community structure on the inter-annual variability of particulate organic carbon flux. Deep-Sea Res. I 42, 619–639.

Boyd P. W., et al. (2000) A mesoscale phytoplankton bloom inthe polar Southern Ocean stimulated by iron fertilization.Nature 407, 695–702.

Boyle E. A. (1988) Cadmium: chemical tracer of deepwaterpaleoceanography. Paleoceanography 3, 471–489.

Broecker W. S. (1982) Ocean chemistry during glacial time.Geochim. Cosmochim. Acta 46, 1689–1705.

Broecker W. S. and Peng T.-H. (1982) Tracers in the Sea.Eldigio Press.

Bruland K. W. (1980) Oceanographic distributions of cad-mium, zinc, nickel, and copper in the North Pacific. EarthPlanet. Sci. Lett. 47, 176–198.

Brzezinski M. A. (1985) The Si : C : N ratio of marine diatoms:interspecific variability and the effect of some environmentalvariables. J. Phycol. 21, 347–357.

Brzezinski M. A. and Nelson D. M. (1996) Chronic substratelimitation of silicic acid uptake in the western Sargasso Sea.Deep-Sea Res. II 43, 437–453.

Buesseler K. O. (1991) Do upper-ocean sediment traps provideand accurate record of particle flux? Nature 353, 420–423.

Buesseler K. O., Bacon M. P., Cochran J. K., and Livingston H. D.(1992) Carbon and nitrogen export during the JGOFS NorthAtlantic Bloom Experiment estimated from 234Th : 238U dis-equilibria. Deep-Sea Res. 39, 1115–1137.

Buitenhuis E. T., van der Wal P., and de Baar H. J. W. (2001)Blooms of Emiliania huxleyi are sinks of atmospheric carbondioxide: a field and mesocosm study derived simulation.Global Biogeochem. Cycles 15, 577–587.

Burkhardt S., Zondervan I., and Riebesell U. (1999) Effect ofCO2 concentration on C : N : P ratio in marine phytoplank-ton: a species comparison.

Calbet A. (2001) Mesozooplankton grazing effect on primaryproduction: a global comparative analysis in marineecosystems. Limnol. Oceanogr. 46, 1824–1830.

Capone D. G., Zehr J. P., Paerl H. W., Bergman B., andCarpenter E. J. (1997) Trichodesmium: a globally significantmarine cyanobacterium. Science 276, 1221–1229.

Cavender-Bares K. K., Mann E. L., Chisholm S. W., OndrusekM. E., and Bidigare R. R. (1999) Differential response ofequatorial Pacific phytoplankton to iron fertilization. Limnol.Oceanogr. 44, 237–246.

Chin W. C., Orellana M. V., and Verdugo P. (1998)Spontaneous assembly of marine dissolved organic matterinto polymer gels. Nature 391, 568–572.

Chisholm S. W., Falkowski P. G., and Cullen J. J. (2001) Dis-crediting ocean fertilization. Science 294, 309–310.

Chisholm S. W., Olson R. J., Zettler E. R., Goericke R.,Waterbury J. B., and Welschmeyer N. A. (1988) A novelfree-living prokaryote abundant in the oceanic euphoticzone. Nature 334, 340–343.

Cho B. C. and Azam F. (1988) Major role of bacterial inbiogeochemical fluxes in the ocean’s interior. Nature 332,441–443.

Christian J. R., Lewis M. R., and Karl D. M. (1997) Verticalfluxes of carbon, nitrogen, and phosphorus in the NorthPacific Subtropical Gyre near Hawaii. J. Geophys. Res. 102,15667–15677.

Clark L. L., Ingall E. D., and Benner R. (1998) Marinephosphorus is selectively remineralized. Nature 393, 428.

Coale K. H. and Bruland K. W. (1985) 234Th:238U disequilibriawithin the California current. Limnol. Oceanogr. 30, 22–33.

Coale K. H., et al. (1996) A massive phytoplankton bloominduced by an ecosystem-scale iron fertilization experimentin the equatorial Pacific Ocean. Nature 383, 495–501.

Collier R. and Edmond J. (1984) The trace elementgeochemistry of marine biogenic particulate matter. Prog.Oceanogr. 13, 113–199.

Conley D. J., Schelske C. L., and Stoermer E. F. (1993)Modification of the biogeochemical cycle of silica witheutrophication. Mar. Ecol. Prog. Ser. 101, 179–192.

de Haas H., van Weering T. C. E., and de Stigter H. (2002)Organic carbon in shelf seas: sinks or sources, processes andproducts. Cont. Shelf Res. 22, 691–717.

Dehairs F., Chesselet R., and Jedwab J. (1980)Discrete suspended particles of barite and the bariumcycle in the open ocean. Earth Planet. Sci. Lett. 49,528–550.

De La Rocha C. L., Hutchins D. A., Brzezinski M. A., andZhang Y. (2000) Effects of iron and zinc deficiency onelemental composition and silica production by diatoms.Mar. Ecol. Prog. Ser. 195, 71–79.

de Villiers S. (1999) Seawater strontium and Sr/Ca variabilityin the Atlantic and Pacific oceans. Earth Planet Sci. Lett.171, 623–634.

Dilling L. and Alldredge A. L. (2000) Fragmentation of marinesnow by swimming macrozooplankton: a new processimpacting carbon cycling in the sea. Deep-Sea Res. I 47,1227–1245.

Dilling L., Wilson J., Steinberg D., and Alldredge A. (1998)Feeding by the euphausiid Euphausia pacifica and thecopepod Calanus pacificus on marine snow. Mar. Ecol.Prog. Ser. 170, 189–201.

Dugdale R. C. and Goering J. J. (1967) Uptake of new andregenerated forms of nitrogen in primary productivity.Limnol. Oceanogr. 12, 196–206.

Dugdale R. C., Wilkerson F. P., and Minas H. J. (1995) Therole of a silicate pump in driving new production. Deep-SeaRes. I 42, 697–719.

Dymond J., Suess E., and Lyle M. (1992) Barium in deep-seasediment: a geochemical proxy for paleoproductivity.Paleoceanography 7, 163–181.

Elderfield H. and Rickaby R. E. M. (2000) Oceanic Cd/Pratio and nutrient utilization in the glacial Southern Ocean.Nature 405, 305–310.

Elderfield H. E. and Schultz A. (1996) Mid-ocean ridgehydrothermal fluxes and the chemical composition of theocean. Ann. Rev. Earth Planet. Sci. 24, 191–224.

Ellwood M. J. and Hunter K. A. (2000) The incorporation ofzinc and iron into the frustule of the marine diatomThalassiosira pseudonana. Limnol. Oceanogr. 45,1517–1524.

Emerson S., Mecking S., and Abell J. (2001) The biologicalpump in the subtropical North Pacific Ocean: nutrientsources, Redfield ratios, and recent changes. GlobalBiogeochem. Cycles 15, 535–554.

Engel A. (2000) The role of transparent exopolymer particles(TEP) in the increase in apparent particle stickiness (alpha)during the decline of a diatom bloom. J. Plankton Res. 22,485–497.

Engel A. (2002) Direct relationship between CO2 uptake andtransparent exopolymer particle production in naturalphytoplankton. J. Plankton Res. 24, 49–53.

Engel A. and Passow U. (2001) Carbon and nitrogen content oftransparent exopolymer particles (TEP) in relation to theirAlcian Blue absorption. Mar. Ecol. Prog. Ser. 219, 1–10.

Engel A., Goldthwait S., Passow U., and Alldredge A. (2002)Temporal decoupling of carbon and nitrogen dynamicsin a mesocosm diatom bloom. Limnol. Oceanogr. 47,753–761.

Eppley R. W. and Peterson B. J. (1979) Particulate organicmatter lux and planktonic new production in the deep ocean.Nature 282, 677–680.

Fagerbakke K. M., Norland S., and Heldal M. (1999) Theinorganic ion content of native aquatic bacteria. Can.J. Microbiol. 45, 304–311.

Falkowski P., et al. (2000) The global carbon cycle: a test ofour knowledge of Earth as a system. Science 290, 291–296.


Field C. B., Behrenfeld M. J., Randerson J. T., and FalkowskiP. (1998) Primary production of the biosphere: integratingterrestrial and oceanic components. Science 281, 237–240.

Filippelli G. M. and Delaney M. L. (1996) Phosphorusgeochemistry of equatorial Pacific sediments. Geochim.Cosmochim. Acta 60, 1479–1495.

Follmi K. B. (1996) The phosphorus cycle, phosphogenesisand marine phosphate-rich deposits. Earth Sci. Rev. 40,55–124.

Fowler S. W. and Knauer G. A. (1986) Role of large particles inthe transport of elements and organic compounds throughthe oceanic water column. Prog. Oceanogr. 16, 147–194.

Francois R., Altabet M. A., Yu E.-F., Sigman D. M., BaconM. P., Frank M., Bohrmann G., Bareille G., and LabeyrieL. D. (1997) Contribution of Southern Ocean surface-waterstratification to low atmospheric CO2 concentrations duringthe last glacial period. Nature 389, 929–935.

Frankignoulle M., Canon C., and Gattuso J.-P. (1994) Marinecalcification as a source of carbon dioxide: positive feedbackof increasing atmospheric CO2. Limnol. Oceanogr. 39,458–462.

Froelich P. N., Bender M. L., Luedtke N. A., Heath G. R., andDe Vries T. (1982) The marine phosphorus cycle. Am. J. Sci.282, 474–511.

Fuhrman J. A. and Capone D. G. (1991) Possible biogeochem-ical consequences of ocean fertilization. Limnol. Oceanogr.36, 1951–1959.

Geider R. J. and LaRoche J. L. (1994) The role of iron inphytoplankton photosynthesis, and the portential for ironlimitation of primary productivity in the sea. Photosyn. Res.39, 275–301.

Greenwood N. N. and Earnshaw A. (1984) Chemistry of theElements. Pergamon.

Gruber N. and Sarmiento J. L. (1997) Global patterns of marinenitrogen fixation and denitrification. Global Biogeochem.Cycles 11, 235–266.

Harrison P. J., Conway H. L., Holmes R. W., and Davis C. O.(1977) Marine diatoms grown in chemostats under silicate orammonium limitation: III. Cellular chemical compositionand morphology of Chaetoceros debilis, Skeletonemacostatum, and Thalassiosira gravida. Mar. Biol. 43, 19–31.

Harrison W. G. (1990) Nitrogen utilization in chlorophyll andprimary productivity maximum layers: an analysis based onthe f-ratio. Mar. Ecol. Prog. Ser. 60, 85–90.

Hedges J. I. and Keil R. G. (1995) Sedimentary organic matterpreservation: an assessment and speculative synthesis. Mar.Chem. 49, 81–115.

Heinrich A. K. (1962) The life histories of planktonic animalsand seasonal cycles of plankton communities in the ocean.J. Cons. Int. Explor. Mer. 27, 15–24.

Holloway C. F. and Cowen J. P. (1997) Development of ascanning confocal laser microscopic technique to examinethe structure and composition of marine snow. Limnol.Oceanogr. 42, 1340–1352.

Hutchins D. A. and Bruland K. W. (1998) Iron-limited diatomgrowth and Si : N uptake ratios in a coastal upwellingregime. Nature 393, 561–564.

Janse I., vanRijssel M., Gottschal J. C., Lancelot C., andGieskes W. W. C. (1996) Carbonhydrates in the North Seaduring spring blooms of Phaeocystis: a specific fingerprint.Aquat. Microb. Ecol. 10, 97–103.

Jeandel C., Dupre B., Lebaron G., Monnin C., and Minster J.-F.(1996) Longitudinal distributions of dissolved barium, silicaand alkalinity in the western and southern Indian Ocean.Deep-Sea Res. I 43, 1–31.

Jenkins W. J. (1982) Oxygen utilization rates in North Atlanticsubtropical gyre and primary production in oligotrophicsystems. Nature 300, 246–248.

Jenkins W. J. (1988) Nitrate flux into the euphotic zone nearBermuda. Nature 331, 521–523.

Joos F. L., Sarmiento J. L., and Sigenthaler U. (1991) Estimatesof the effect of Southern Ocean iron fertilization onatmospheric CO2 concentrations. Nature 349, 772–775.

Kaehler P. and Koeve W. (2001) Marine dissolved organicmatter: can its C : N ratio explain carbon overconsumption?Deep-Sea Res. I 48, 49–62.

Kamitani A. (1982) Dissolution rates of silica from diatomsdecomposing at various temperatures. Mar. Biol. 68, 91–96.

Karl D. M., Letelier R., Tupas L., Dore J., Christian J., andHebel D. (1997) The role of nitrogen fixation in biogeo-chemical cycling in the subtropical North Pacific Ocean.Nature 388, 533–538.

Keeling C. D. and Whorf T. P. (2001) Atmospheric CO2

records from sites in the SiO air sampling network.In Trends: A Compendium of Data on Global Change.US Department of Energy, Oak Ridge National Laboratory,Carbon Dioxide Information Analysis Center, Oak Ridge,TN.

Kepkay P. E. (1994) Particle aggregation and the biologicalreactivity of colloids. Mar. Ecol. Prog. Ser. 109, 293–304.

Knauer G. A. and Martin J. H. (1981) Phosphorus and cadmiumcycling in northeast Pacific waters. J. Mar. Res. 39, 65–76.

Kolowith L. C., Ingall E. D., and Benner R. (2001)Composition and cycling of marine organic phosphorus.Limnol. Oceanogr. 46, 309–321.

Korb R. A., Saville P. J., Johnston A. M., and Raven J. A.(1997) Sources of inorganic carbon for photosynthesis bythree species of marine diatom. J. Phycol. 33, 433–440.

Kumar N., Gwiazda R., Anderson R. F., and Froelich P. N.(1993) 231Pa/230Th ratios in sediments as a proxy for pastchanges in Southern Ocean productivity. Nature 362,45–48.

Lam P. J., Tortell P. D., and Morel F. M. M. (2001) Differentialeffects of iron additions on organic and inorganic carbonproduction by phytoplankton. Limnol. Oceanogr. 46,1199–1202.

Lampitt R. and Anita A. N. (1997) Particle flux in deep seas:regional characteristics and temporal variability. Deep-SeaRes. I 44, 1377–1403.

Lane T. W. and Morel F. M. M. (2000) A biological functionfor cadmium in marine diatoms. Proc. Natl Acad. Sci. USA97, 4627–4631.

Laws E. A., Falkowski P. G., Smith W. O., Jr., Ducklow H.,and McCarthy J. J. (2000) Temperature effects on exportproduction in the open ocean. Global Biogeochem. Cycles14, 1231–1246.

Lea D. W. and Boyle E. (1989) Barium content of benthicforaminifera controlled by bottom water composition.Nature 338, 751–753.

Lenes J. M., et al. (2001) Iron fertilization and theTrichodesmium response on the West Florida shelf. Limnol.Oceanogr. 46, 1261–1277.

Levitus S., Burgett R., and Boyer T. (1994) World Ocean Atlas1994 Volume 3: Nutrients. US Department of Commerce.

Lisitzin A. P. (1972) Sedimentation in the World Ocean. Soc.Econ. Paleontol. Mineral. Spec. Publ. 17, 218pp.

Liu H., Landry M. R., Vaulot D., and Campbell L. (1999)Prochlorococcus growth rates in the central equatorialPacific: an application of the fmax approach. J. Geophys.Res. 104, 3391–3399.

Logan B. E., Passow U., Alldredge A. L., Grossart H. P.,and Simon M. (1995) Rapid formation and sedimentationof large aggregates is predictable from coagulation rates(half-lives) of transparent exopolymer particles (TEP).Deep-Sea Res. II 42, 203–214.

Long R. A. and Azam F. (1996) Abundant protein-containingparticles in the sea. Aquat. Microb. Ecol. 10, 213–221.

Loscher B. M., de Jong J. T. M., and de Baar H. J. W. (1998)The distribution and preferential biological uptake ofcadmium at 68W in the Southern Ocean. Mar. Chem. 62,259–286.

Luz B. and Barkan E. (2000) Assessment of oceanicproductivity with the triple-isotope composition of dissolvedoxygen. Science 288, 2028–2031.

Mari X., Beauvais S., Lemee R., and Pedrotti M. L. (2001)Non-reedfield C : N ratio of transparent exopolymeric

References 109

particles in the northwestern Mediterranean Sea. Limnol.Oceanogr. 46, 1831–1836.

Markels M., Jr. (2001) Method of sequestering carbon dixoidewith spiral fertilization. US Patent No. 6,200,530.

Martin J. H. (1990) Glacial-interglacial CO2 change: the ironhypothesis. Paleoceanography 5, 1–13.

Martin J. H. and Fitzwater S. E. (1988) Iron deficiency limitsphytoplankton growth in the north-east Pacific subarctic.Nature 331, 341–343.

Martin J. H. and Gordon R. M. (1988) Northeast Pacific irondistributions in relation to phytoplankton productivity.Deep-Sea Res. 35, 177–196.

Martin J. H. and Knauer G. A. (1973) The elementalcomposition of plankton. Geochim. Cosmochim. Acta 37,1639–1653.

Martin J. H., Knauer G. A., Karl D. M., and Broenkow W. W.(1987) VERTEX: carbon cycling in the northeast Pacific.Deep-Sea Res. 34, 267–285.

Martin J. H., et al. (1994) Testing the iron hypothesis inecosystems of the equatorial Pacific Ocean. Nature 371,123–129.

McCarthy M., Pratum T., Hedges J., and Benner R. (1997)Chemical composition of dissolved organic nitrogen inthe ocean. Nature 390, 150–154.

McCave I. N. (1975) Vertical flux of particles in the ocean.Deep-Sea Res. 22, 491–502.

McCave I. N. (1984) Size spectra and aggregation of suspendedparticles in the deep ocean. Deep-Sea Res. 31, 329–352.

Michaels A. F. and Silver M. W. (1988) Primary production,sinking fluxes and the microbial food web. Deep-Sea Res.35, 473–490.

Mopper K., Zhou J. A., Ramana K. S., Passow U., DamH. G., and Drapeau D. T. (1995) The role of surface-active carbohydrates in the flocculation of a diatombloom in a mesocosm. Deep-Sea Res. II 42, 47–73.

Morel F. M. M., Reinfelder J. R., Roberts S. B., ChamberlainC. P., Lee J. G., and Yee D. (1994) Zinc and carbon co-limitation of marine phytoplankton. Nature 369, 740–742.

Morris I. (1981) Photosynthetic products, physiological state,and phytoplankton growth. Can. J. Fish. Aquat. Sci. 210,83–102.

Nelson D. M. and Dortch Q. (1996) Silicic acid depletion andsilicon limitation in the plume of the Mississippi River:evidence from kinetic studies in spring and summer. Mar.Ecol. Prog. Ser. 136, 163–178.

Nelson D. M., Treguer P., Brzezinski M. A., Leynaert A., andQueguiner B. (1995) Production and dissolution of biogenicsilica in the ocean: revised global estimates, comparisonwith regional data and relationship to biogenic sedimen-tation. Global Biogeochem. Cycles 9, 359–372.

Nimer N. A., Iglesias-Rodriguez M. D., and Merrett M. J.(1997) Bicarbonate utilization by marine phytoplanktonspecies. J. Phycol. 33, 625–631.

Nozaki T. (1997) A fresh look at element distribution in theNorth Pacific. EOS Trans., AGU Electron. Suppl http://www.agu.org/eos_elec/97025e.html.

Olesen M. and Lundsgaard C. (1995) Seasonal sedimentationof autochthonous material from the euphotic zone of acoastal system. Estuar. Coast. Shelf Sci. 41, 475–490.

Paasche E. (1999) Reduced coccolith calcite production underlight-limited growth: a comparative study of three clones ofEmiliania huxleyi (Prymnesiophyceae). Phycologia 38,508–516.

Pahlow M. and Riebesell U. (2000) Temporal trends in deepocean Redfield ratios. Science 287, 831–833.

Passow U. (2000) Formation of transparent exopolymerparticles, TEP, from dissolved precursor material. Mar.Ecol. Prog. Ser. 192, 1–11.

Passow U. (2002) Transparent exopolymer particles (TEP) inaquatic environments. Prog. Oceanogr. 55, 287–333.

Passow U., Shipe R. F., Murray A., Pak D., Brzezinski M. A.,and Alldredge A. L. (2001) The origin of transparent

exopolymer particles (TEP) and their role in the sedimen-tation of particulate matter. Cont. Shelf Res. 21, 327–346.

Paytan A., Kastner M., and Chavez F. P. (1996) Glacial tointerglacial fluctuations in productivity in the EquatorialPacific as indicated by marine barite. Science 274,1355–1357.

Peng T.-H. and Broecker W. S. (1991a) Dynamical limitationson the Antarctic iron fertilization strategy. Nature 349,227–229.

Peng T.-H. and Broecker W. S. (1991b) Factors limiting thereduction of atmospheric CO2 by iron fertilization. Limnol.Oceanogr. 36, 1919–1927.

Pondaven P., Ragueneau O., Treguer P., Hauvespre A.,Dezileau L., and Reyss J. L. (2000) Resolving the “opalparadox” in the Southern Ocean. Nature 405, 168–172.

Price N. M. and Morel F. M. M. (1990) Cadmium and cobaltsubstitution for zinc in a marine diatom. Nature 344,658–660.

Quigley M. S., Santschi P. H., Hung C.-C., Guo L., andHoneyman B. D. (2002) Importance of acid polysaccharidesfor 234Th complexation to marine organic matter. Limnol.Oceanogr. 47, 367–377.

Raven J. A. (1997) Inorganic carbon acquisition by marineautotrophs. Adv. Bot. Res. 27, 85–209.

Redfield A. C. (1934) On the proportions of organic derivativesin sea water and their relation to the composition ofplankton. In James Johnstone Memorial Volume (ed.R. J. Daniel). University Press of Liverpool, pp. 177–192.

Redfield A. C. (1958) The biological control of chemicalfactors in the environment. Am. Sci. 46, 205–221.

Riebesell U., Wolf-Gladrow D. A., and Smetacek V. (1993)Carbon dioxide limitation of marine phytoplankton growthrates. Nature 361, 249–251.

Robertson J. E., Robinson C., Turner D. M., Holligan P.,Watson A. J., Boyd P., Fernandez E., and Finsh M. (1994)The impact of a coccolithophore bloom on oceanic carbonuptake in the northeast Atlantic during summer 1991. Deep-Sea Res. I 41, 297–314.

Ruiz F. A., Marchesini N., Seufferheld M., and Docampo R.(2001) The polyphosphate bodies of Chlamydomonasreinhardtii possess a proton-pumping pyrophosphatase andare similar to acidocalcisomes. J. Biol. Chem. 276,46196–46203.

Ruttenberg K. C. (1993) Reassessment of the oceanic residencetime of phosphorus. Chem. Geol. 107, 405–409.

Ryther J. H. (1969) Photosynthesis and fish production in thesea. Science 166, 72–76.

Sanudo-Wilhelmy S. A., Kustka A. B., Gobler C. J., HutchinsD. A., Yang M., Lwiza K., Burns J., Capone D. G., RavenJ. A., and Carpenter E. J. (2001) Phosphorus limitation ofnitrogen fixation by Trichodesmium in the central AtlanticOcean. Nature 411, 66–69.

Sarmiento J. L. and Orr J. C. (1991) Three-dimensionalsimulations of the impact of Southern Ocean nutrientdepletion on atmospheric CO2 and ocean chemistry. Limnol.Oceanogr. 36, 1928–1950.

Sarmiento J. L. and Toggweiler J. R. (1984) A new model forthe role of the oceans in determining atmospheric PCO2

.Nature 308, 621–624.

Sarmiento J. L., Hughes T. M. C., Stouffer R. J., and Manabe S.(1998) Simulated response of the ocean carbon cycle toanthropogenic climate warming. Nature 393, 245–249.

Scholten J. C., et al. (2001) Trapping efficiency of sedimenttraps from the deep eastern North Atlantic: 230Th calibration.Deep-Sea Res. II 48, 243–268.

Schumann R. and Rentsch D. (1998) Staining particulateorganic matter with DTAF—a fluorescence dye for carbo-hydrates and protein: a new approach and application of a2D image analysis system. Mar. Ecol. Prog. Ser. 163,77–88.

Shaffer G. (1996) Biogeochemical cycling in the global ocean:2. New production, Redfield ratios, and remineralization inthe organic pump. J. Geophys. Res. 101, 3723–3745.


Shanks A. L. and Trent J. D. (1980) Marine snow: sinking ratesand potential role in vertical flux. Deep-Sea Res. 27A,137–143.

Shemesh A., Mortlock R. A., Smith R. J., and Froelich P. N.(1988) Determination of Ge/Si in marine siliceous micro-fossils: separation, cleaning and dissolution of diatoms andradiolaria. Mar. Chem. 25, 305–323.

Shulenberger E. and Reid J. L. (1981) The Pacific shallowoxygen maximum, deep chlorophyll maximum, and primaryproductivity, reconsidered. Deep-Sea Res. 28A, 901–919.

Shuskina E. A. (1985) Production of principal ecologicalgroups of plankton in the epipelagic zone of the ocean.Oceanology 25, 653–658.

Siever R. (1991) Silica in the Oceans: biological–geochemicalinterplay. In Scientists on Gaia (eds. S. Schneider and P.Boston). MIT Press, pp. 287–295.

Smayda T. J. (1970) The suspension and sinking ofphytoplankton in the sea. Oceanogr. Mar. Biol. Ann. Rev.8, 353–414.

Smith D. C., Simon M., Alldredge A. L., and Azam F. (1992)Intense hydrolytic enzyme activity on marine aggregates andimplications for rapid particle dissolution. Nature 359,139–142.

Smith D. C., Steward G. F., Long R. A., and Azam F. (1995)Bacterial mediation of carbon fluxes during a diatom bloomin a mesocosm. Deep-Sea Res. II 42, 75–97.

Soendergaard M., Williams P. J. L., Cauwet G., Riemann B.,Robinson C., Terzic S., Woodward E. M. S., and Worm J.(2000) Net accumulation and flux of dissolved organiccarbon and dissolved organic nitrogen in marine planktoncommunities. Limnol. Oceanogr. 45, 1097–1111.

Steemann Nielsen E. (1952) The use of radioactive carbon(14C) for measuring organic production in the sea. J. Cons.Int. Explor. Mer. 18, 117–140.

Steinberg D. K. (1995) Diet of copepods (Scolpalatum vorax)associated with mesopelagic detritus (giant larvaceanhouses) in Monterey Bay, California Mar. Biol. 122,571–584.

Steinberg D. K., Pilskaln C. H., and Silver M. W. (1998)Contribution of zooplankton associated with detritus tosediment trap “swimmer” carbon in Monterey Bay, Califor-nia, USA. Mar. Ecol. Prog. Ser. 164, 157–166.

Steinberg D. K., Carlson C. A., Bates N. R., Johnson R. J.,Michaels A. F., and Knap A. P. (2001) Overview of the USJGOFS Bermuda Atlantic Time-series Study (BATS): adecade-scale look at ocean biology and biogeochemistry.Deep-Sea Res. II 48, 1405–1447.

Strass V. H. (2002) EisenEx-1: test of the iron hypothesis in aSouthern Ocean eddy. EOS: Trans., AGU Abstracts of 2002Ocean Sciences Meeting.

Strickland J. D. H. and Parsons T. R. (1968) A PracticalHandbook of Seawater Analysis. Fisheries Research Boardof Canada.

Strom S. L., Benner R., Ziegler S., and Dagg M. J. (1997)Planktonic grazers are a potentially important source ofmarine dissolved organic carbon. Limnol. Oceanogr. 42,1364–1374.

Suess E. (1980) Particulate organic carbon flux in the oceans-surface productivity and oxygen utilization. Nature 288,260–263.

Sunda W. G. and Huntsman S. A. (1995a) Iron uptake andgrowth limitation in oceanic and coastal phytoplankton.Mar. Chem. 50, 189–206.

Sunda W. G. and Huntsman S. A. (1995b) Cobalt and zincinterreplacement in marinephytoplankton: biological andgeochemical implications. Limnol. Oceanogr. 40,1404–1417.

Sundquist E. T. (1993) The global carbon dioxide budget.Science 259, 934–941.

Takahashi T., Broecker W. S., and Langer S. (1985) Redfieldratio based on chemical data from isopycnal surfaces.J. Geophys. Res. 90, 6907–6924.

Takeda S. (1998) Influence of iron availability on nutrientconsumption ratio of diatoms in oceanic waters. Nature 393,774–777.

Tappan H. and Loeblich A. R., Jr. (1973) Evolution of theoceanic plankton. Earth Sci. Rev. 9, 207–240.

Teira E., Pazo M. J., Serret P., and Fernandez E. (2001)Dissolved organic carbon production by microbial popu-lations in the Atlantic Ocean. Limnol. Oceanogr. 46,1370–1377.

Timmermans K. R., Snoek J., Gerringa L. J. A., Zondervan I.,and de Baar H. J. W. (2001) Not all eukaryotic algae canreplace zinc with cobalt: Chaetoceros calcitrans (Bacillar-iophyceae) versus Emiliania huxleyi (Prymnesiophyceae).Limnol. Oceanogr. 46, 699–703.

Tortell P. D., Reinfelder J. R., and Morel F. M. M. (1997)Active uptake of bicarbonate by diatoms. Nature 390,243–244.

Treguer P., Nelson D. M., Van Bennekom A. J., DeMaster D. J.,Leynaert A., and Queguiner B. (1995) The silica balance inthe world ocean: a reestimate. Science 268, 375–379.

Turner J. T. (2002) Zooplankton fecal pellets, marine snow andsinking phytoplankton blooms. Aquat. Microbiol. Ecol. 27,57–102.

Tyrrell T. (1999) The relative influences of nitrogen andphosphorus on oceanic primary production. Nature 400,525–531.

Verity P. G., Williams S. C., and Hong Y. (2000) Formation,degradation, and mass: volume ratios of detritus derivedfrom decaying phytoplankton. Mar. Ecol. Prog. Ser. 207,53–68.

Ware J. R., Smith S. V., and Reaka-Kudla M. L. (1992) Coralreefs: sources or sinks of atmospheric CO2. Coral Reefs 11,127–130.

Watson A. J., et al. (1994) Minimal effect of iron fertilizationon sea-surface carbon dioxide concentrations. Nature 371,143–145.

Watson A. J., Bakker D. C. E., Ridgwell A. J., Poyd P. W., andLaw C. S. (2000) Effect of iron supply on Southern OceanCO2 uptake and implications for glacial atmospheric CO2.Nature 407, 730–733.

Weiss R. F., Broecker W. S., Craig H., and Spencer D. (1983)GEOSECS Indian Ocean Expedition: Volume 5. Hydro-graphic Data 1977–1978. National Science Foundation.

Williams P. M. and Druffel E. R. M. (1987) Radiocarbon indissolved organic matter in the central north Pacific Ocean.Nature 330, 246–248.

Wolf-Gladrow D. A., Riebesell U., Burkhardt S., and Bijma J.(1999) Direct effects of CO2 concentration on growth andisotopic composition of marine plankton. Tellus (B Chem.Phys. Meteorol.) 51B, 176–461.

Zeebe R. E. and Wolf-Gladrow D. (2001) CO2 in Seawater:Equilibrium, Kinetics, Isotopes. Elsevier.

Zehr J. P. and Ward B. B. (2002) Nitrogen cycling in the ocean:new perpsectives on processes and paradigms. Appl.Environ. Microbiol. 68, 1015–1024.

Zehr J. P., Waterbury J. B., Turner P. J., Montoya J. P.,Omoregie E., Steward G. F., Hansen A., and Karl D. M.(2001) Unicellular cyanobacteria fix N2 in the subtropicalNorth Pacific Ocean. Nature 412, 635–638.

Zhang J. and Nozaki Y. (1996) Rare earth elements and yttriumin seawater: ICP_MS determinations in the East Caroline,Coral Sea, and Southern Fiji basins of the western PacificOcean. Geochim. Cosmochim. Acta 60, 4631–4644.

Zhuang G., Duce R. A., and Kester D. R. (1990) Thedissolution of atmospheric iron in surface seawater of theopen ocean. J. Geophys. Res. 95, 16207–16216.

q 2003, Elsevier Ltd. All rights reservedNo part of this publication may be reproduced, stored in a retrieval system ortransmitted in any form or by any means, electronic, mechanical, photocopying,recording or otherwise, without prior written permission of the Publisher.

Treatise on GeochemistryISBN (set): 0-08-043751-6

Volume 6; (ISBN: 0-08-044341-9); pp. 83–111

References 111

6.04 the biological pump - imedea divulga csic-uib€¦ · impacts on the geochemistry of many...

Documents