sign in sign up
<<
Home , Biogeochemical cycle, Marine prokaryotes, Nitrification, Nitrogen cycle, Nitrogen fixation, Phosphorus cycle

or collective redistirbution other or collective or means this reposting, of machine, is by photocopy article only anypermitted of with portion the approval The O of This article has been published in Oceanography published been This has article Special Issue Feature , Volume 4, a quarterly 22, Number The O journal of

Nutrient Cycles and ceanography Society. Marine Microbes in a © CO -Enriched 2009 by The O 2 ceanography Society. A By David A. Hutchins, Margaret R. Mulholland, and Feixue Fu ceanography Society. Send all correspondence to:ceanography Society. Send [email protected] Th e O or ll rights reserved. P reserved. ll rights ermission is granted to in teaching copy this and research. for use article R Box 1931, R Box ceanography Society, PO epublication, systemmatic reproduction, ockville, MD 20849-1931, U ockville, S A .

128 Oceanography Vol.22, No.4 Abstract

The ocean cycle is tightly linked with the cycles of the major elements CO2 using sunlight as an source, , , and silicon. It is therefore likely that enrichment of the ocean while chemoautotrophic and with anthropogenic CO2 and attendant acidification will have large consequences do the same thing using chem- for marine nutrient , and for the microbes that mediate many key ical sources of energy. Together, these nutrient transformations. The best available evidence suggests that the two broadly defined functional groups may respond strongly to higher CO2 through increases in global N2 fixation and provide the raw materials and energy possibly , as well as potential decreases in . These trends that fuel all of the ocean’s food webs. could cause nitrification to become a nitrogen cycle “bottleneck,” by increasing the For autotrophic groups that depend on flux of N2 fixed into while decreasing the fraction being oxidized to for a living, rising partial and . The consequences could include reduced supplies of oxidized of CO2 (pCO2) in the ocean nitrogen substrates to denitrifiers, lower levels of nitrate-supported new primary may actually represent an opportunity production, and expansion of the regenerated production system accompanied by rather than a misfortune. The term shifts in current communities. The phosphorus and silicon cycles seem “” carries undeniably less likely to be directly affected by enhanced CO2 conditions, but will undoubtedly negative connotations, and indeed it is respond indirectly to changing carbon and nitrogen biogeochemistry. A review of hard to envision how most animals and culture experiments that examined the effects of increased CO2 on elemental ratios microbial could benefit of phytoplankton suggests that for most and , C:N and N:P from lowered ocean pH. However, ratios will either remain at Redfield values or increase substantially. Natural dissolved inorganic carbon enrich- CO2 manipulation experiments show much more mixed outcomes, with ment is an equally accurate description both increases and decreases in C:N and N:P ratios reported at future CO2 levels. We of anthropogenic input of CO2 to the conclude our review with projections of overall trends in the cycles of nitrogen, phos- ocean, and this terminology evokes a phorus, and silicon over the next century as they respond to the steady accumulation much different image of how communi- of -fuel-derived CO2 in a rapidly changing ocean. ties might respond. Much of our view of ocean acidification as an unmitigated stressor for marine comes from Introduction nitrogen, phosphorus, and silicon. well-justified concern over its negative The accumulation of fossil-fuel-derived Consequently, the ongoing and accel- impacts on and other calcifying

CO2 in the and consequent ocean erating enrichment of with (Kleypas et al., 1999; Feely acidification represents an unprec- anthropogenic CO2 will likely have far- et al., 2004; Fabry et al., 2008). It is edented human perturbation of ocean reaching effects not only on the carbon important to remember, though, that chemistry on a global scale. A funda- cycle, but also on the cycles of these for many of the that power mental theme of modern oceanography biologically required . the cycles of carbon and nutrients in the that dates back to the classic work of Many of the key pathways and ocean and support virtually all marine

Redfield (1958) is a recognition of the transformations in ocean nutrient biological , the term “CO2 intimate interweaving of the marine cycles are mediated by autotrophs. fertilization” may be more appropriate. of carbon with the Photoautotrophs, including cyanobac- There are several comprehensive recent cycles of the major nutrient elements teria and eukaryotic phytoplankton, fix reviews on the effects of changing pCO2

Oceanography December 2009 129 THE MARINE NITROGEN CYCLE – on phytoplankton , including +5 NO3 Riebesell (2004) and Rost et al. (2008). +4 tion In this article we explore what is n currently known about the implica- +3 – Nitri ca NO2 tions of ocean acidification and CO2 e fertilization for the biogeochemical +2 Denitri catioNO cycles of nitrogen, phosphorus, and Assimilation silicon, and for the keystone marine +1 N2O organisms that mediate them, including photoautotrophs, chemoautotrophs, and 0 N2 Oxidation Stat Oxidation heterotrophs. Along with the rapidly

-1 Nitri cation mounting concern about ocean acidifi- N 2 F ixation cation, the marine science community -2 has produced a flood of new informa- Assimilation tion from experimental studies in both Organic -3 NH3 the laboratory and the field. These + Nitrogen NH 4 Degradation novel results are helping to provide the perspective we need to make educated Figure 1. Major chemical forms and transformations of nitrogen in the ocean. The guesses in response to such questions as: various chemical forms of nitrogen are plotted versus their oxidation states. Here, we consider the potential effects of increased oceanp CO2 on three of the critical How will crucial processes in the marine transformations within the N cycle: N2 fixation (red arrow), nitrification (blue nitrogen cycle, including nitrogen fixa- arrows), and denitrification (green arrow). tion, nitrification, and denitrification, respond to ocean acidification? Will the cycles of phosphorus and silicon change in response to anthropogenic CO2 CO2 enrichment and concomitant ocean physiology and biological community inputs, and if so, how? How will elevated acidification. We acknowledge that our structure could have major impacts on pCO2 affect the C:N:P stoichiometry of projections are speculative, as we recog- ocean biogeochemical cycles. phytoplankton in the ocean? Finally, we nize that many large gaps remain in this New nitrogen enters the ocean conclude this review with a projection developing picture, and there is a need through the fixation of atmospheric of how nutrient biogeochemistry might for additional work to fill these gaps. dinitrogen (N2) by marine N2 fixers, be altered in the future in response to also known as (Figure 1, The Nitrogen Cycle red arrow). Nitrogen losses from the David A. Hutchins ([email protected]) The potential for ocean acidification ocean are through organisms that – is Professor of Marine Environmental to affect multiple processes within reduce oxidized N (primarily NO3) to

Biology, Department of Biological the marine nitrogen cycle is obvious, N2 through classical denitrification, Sciences, University of Southern because this complex cycle involves an anaerobic process (Figure 1, green California, Los Angeles, CA, USA. compounds that are stable across eight arrow). Further losses of fixed N from the Margaret R. Mulholland is Associate oxidation states (Figure 1). Microbes ocean occur through anaerobic ammo- + Professor, Department of Ocean, , are the primary mediators of the cycling nium (NH4) oxidation via so-called and Atmospheric Sciences, Old Dominion of nitrogen between compounds, and “anammox.” Once input into the ocean, University, Norfolk, VA, USA. Feixue Fu is through their the nitrogen nitrogen is cycled through microbial and Research Assistant Professor, Department and carbon cycles are inextricably interactions. Nitrogen taken up of Biological Sciences, University of linked. Consequently, CO2 enrichment into particulate material and microbial Southern California, Los Angeles, CA, USA. or acidification-related shifts in both cell is released and regenerated

130 Oceanography Vol.22, No.4 a as ammonium and dissolved organic 120 compounds. In aerobic waters, and archaea can then oxidize tes (%) 100 ammonium, producing first nitrite Ra – – 80 (NO2) and then nitrate (NO3) (Figure 1, blue arrows). Here, we evaluate the state ixation

F 60 of our current knowledge about how 2 ocean uptake of anthropogenic CO2 40 and attendant acidification could affect three of the fundamental nitrogen cycle 20 Increase in N processes shown in Figure 1: N2 fixation 0 (red arrow), nitrification (blue arrows), 1 2 3 4 5 6 7 and denitrification (green arrow). b 1600 100 N2 Fixation 1400

The effects of rising 2CO on N2 fixa-

) 1200 90 (ppm) 2 tion by marine cyanobacteria have -1 richodesmium probably been better documented than T 1000 p CO

kg yr 80 any other part of the marine nitrogen 9 800 cycle, although our knowledge is still (x 10 limited to only a few taxa. Most studies by ixation F 70 600 2 to date have focused on the ubiquitous 400 subtropical and tropical 60

Trichodesmium, which is amenable to Global N 200 culture and believed to be responsible 2000 2050 2100 2150 2200 Year for up to 50% of total marine N2 fixation (Mahaffey et al., 2005). Four recently Figure 2. (a) Enhanced N2 fixation rates of cultured marine cyanobac- published studies have presented similar teria at projected future pCO2 (750–1000 ppm), expressed as a percent results showing that increasing pCO (%) increase over N2 fixation rates at currentp CO2 (370–400 ppm). 2 Data are taken from experiments using: (1) and (2) stimulates N2 fixation by strain GBR () grown at 29°C or cultures (Hutchins et al., 2007; Levitan 25°C across a range of 380–750 ppm CO2 (Hutchins et al., 2007); (3) T. erythraeum strain IMS 101 grown at both 25°C and 29°C across et al., 2007; Barcelos e Ramos et al., a range of 380–750 ppm CO2 (Hutchins et al., 2007); (4) T. erythraeum 2007; Kranz et al., 2009). The reported strain IMS 101 grown at 25°C across a range of 380–750 ppm CO2 (Barcelos e Ramos et al., 2007); (5) T. erythraeum strain IMS 101 grown increases in rates of N2 fixation from at 25°C across a range of 400–900 ppm CO2 (Levitan et al., 2007); present-day pCO2 (375–380 ppm) to (6) T. erythraeum strain IMS 101 grown at 25°C across a range of projected year 2100 pCO2 (~ 750 ppm) 370–1000 ppm CO2, (Kranz et al., 2009); and (7) range from 35% to 65%, although strain WH8501 grown at 28°C across a range of 380–750 ppm CO2 (Fu et al., 2008). (b) Potential future increases in annual global N2 Levitan et al. (2007) reported an increase fixation byTrichodesmium spp. (left axis), based on maximum (blue) in N2 fixation of 121% between 400 and and minimum (red) levels of CO2 stimulation observed in laboratory 900 ppm (Figure 2a). Several of these cultures by Hutchins et al. (2007). Estimated atmospheric pCO2 (right axis) is plotted in green. CO2 data from Royal Society (2005) culture studies also observed similar degrees of stimulation of Trichodesmium carbon fixation and/or growth rates by increased CO2.

Oceanography December 2009 131 The magnitude of these N2 and carbon In addition, there is still very little environment and are more common in fixation responses of Trichodesmium information available regarding CO2 and coastal systems. Therefore, represents the largest physiological effects on marine N2 fixers other than it is not clear that the Czerny et al. response yet reported for any marine Trichodesmium. Only one study has (2009) results have large microbial group (including phyto- examined the response of N2-fixing implications, except in specific envi- plankton) to projected pCO2 changes. If unicellular cyanobacteria to elevated ronments such as the Baltic Sea where these culture results can be extrapolated CO2, despite the fact that this group harmful blooms of these organisms can to the environment, Trichodesmium and is now thought to fix at least as much occur at fairly low salinities. As for the perhaps other N2-fixing cyanobacteria nitrogen as Trichodesmium on a global other important groups of marine N2 seem poised to be among the biggest basis (Zehr et al., 2001; Montoya fixers, such as unicellular cyanobac- “winners” in the future anthropo- et al., 2004). This investigation, using teria, /diazotroph associations, genically acidified ocean. Such major cultures of the unicellular diazotroph /diazotroph symbioses, and increases in diazotrophy due to CO2 Crocosphaera watsonii, demonstrated N2-fixing heterotrophic eubacteria or fertilization of the ocean could have that N2 fixation rates were enhanced by archaea, we are still at the rudimentary immense implications for the marine 40% when cultures were grown at pCO2 stages of understanding and quanti- nitrogen cycle over the next century. levels of 750 ppm relative to those grown fying the contributions of these groups

However, a closer examination of at 380 ppm (Figure 2a; Fu et al., 2008a). to marine N2 fixation and assessing the data presented in Figure 2a reveals Thus, the general trends observed where N2 fixation occurs in the ocean that our picture of CO2 impacts on were very similar to those reported for (Carpenter and Capone, 2008). Little or marine diazotrophy is still far from Trichodesmium. nothing at all is known about how they complete. To date, all of the published In contrast to the results of these will react to future pCO2 changes. Trichodesmium culture studies have used studies, a recent culture study by Czerny The other limitation to our under- only one species, T. erythraeum, and et al. (2009) of the bloom-forming standing of overall ocean acidification we know that physiological responses heterocystous cyanobacterium or CO2 enrichment effects on 2N fixa- sometimes vary greatly between even spumigena from the brackish Baltic Sea tion stems from the fact that all of closely related species. Consequently, found large decreases in cell division the published work so far has been the responses of other common species rates and slight decreases in N2 fixation from laboratory culture studies. Field within this to high pCO2 remain rates as pCO2 increased. These results evidence is still lacking for stimulation completely unknown. In fact, with the provide a cautionary note, in that we of N2 fixation rates by elevated pCO2 in exception of one study that compared cannot assume that all diazotrophs Trichodesmium or other diazotrophs. We T. erythraeum isolates from the Atlantic will respond favorably or uniformly to recently carried out a set of three prelim- and Pacific (Hutchins et al., 2007), all of higher CO2 levels. These authors suggest inary incubation experiments using the currently published data come from that this contrasting CO2 response by natural Trichodesmium colonies collected a single T. erythraeum culture strain, N. spumigena could be due to deleterious from a bloom along the west coast of Trichodesmium IMS 101 (Figure 2a). effects of lower pH on the transport of Florida. The colonies were collected with This commonly cultivated “lab weed” fixed nitrogen compounds from the net tows, then incubated on deck for four was isolated about 20 years ago from site of N2 fixation in the to to six daylight hours in filtered seawater North Carolina coastal waters (Prufert- neighboring vegetative cells. bubbled with 380 ppm or 750 ppm

Bebout et al., 1993). Consequently, it Estuaries and brackish waters may CO2/air mixtures. Rate measure- is unclear how representative these experience more dramatic pH changes ments from these incubations using 15 studies are, given the broad diversity of than the open ocean, due to their lower N2 tracer techniques (Mulholland uncultivated Trichodesmium species and buffering capacities (Najjar et al., in and Bernhardt, 2005) demonstrated growing in the tropical and press). Heterocystous cyanobacteria that increased pCO2 stimulated N2 fixa- subtropical . generally do not occur in the marine tion rates by 6%, 21%, and 41% above

132 Oceanography Vol.22, No.4 rates measured in colonies maintained Trichodesmium alone by the end of this Although Hutchins et al. (2007) at ambient pCO2 (Figure 3). The very century. Using the maximum values found that N2 fixation rates of rapid (hours) responses of these natural obtained from our culture experiments Trichodesmium increased with pCO2

Trichodesmium samples suggest that CO2 in this calculation boosts our estimate up to about 750 ppm, N2 fixation rates 9 -1 was a proximate limitation on N2 fixa- by another 20 x 10 kg N yr to nearly did not continue to rise as pCO2 was tion rates, in a manner very analogous to 100 x 109 kg N yr-1 by the year 2100 further elevated to 1250 and 1500 ppm. a limiting nutrient. The short time frame (Figure 2b). Because current estimates For this reason, we believe that the of their responses further suggests that of total global N2 fixation by all oceanic increase in the total amount of potential it was unlikely that the colonies needed diazotrophs combined are about Trichodesmium N2 fixation could level to induce new cellular biochemical 100–200 x 109 kg N yr-1 (Galloway et al., off by about the year 2100 (Figure 2b). systems to take advantage of the CO2 2004), this suggests the possibility that This saturation response also emphasizes enrichment. To our knowledge, these are by 2100, CO2 fertilization could increase the similarity of Trichodesmium pCO2 the first field data supporting the results new N inputs by Trichodesmium alone responses to classical phytoplankton from the laboratory studies referenced to levels approaching those estimated nutrient limitation kinetics; examination above, but much more evidence from for the entire marine diazotrophic of the Hutchins et al. (2007) data set natural diazotroph populations will be community today. Active growth and suggests an N2 fixation half-saturation needed before we can robustly assess the N2 fixation by Trichodesmium are constant of about 330 ppm CO2. The potential importance of CO2 stimulation restricted to waters > 20°C (Carpenter CO2 concentration of the atmosphere of marine N2 fixation for global biogeo- and Capone, 2008), so expansion of their passed this level during the 1970s chemical cycles. subtropical into higher latitudes (Royal Society, 2005). If we assume that the results of due to sea surface warming will tend to Of course, the major assumption previous laboratory culture experi- further increase this estimate as well. behind these calculations of future ments can legitimately be scaled up Adding in the contributions of unicel- trends is that other limiting factors will to the whole ocean, we can calculate lular cyanobacteria and other groups not constrain Trichodesmium N2 fixation the potential increase in future new N of N2 fixers might greatly augment this before the CO2 response saturates. There inputs by Trichodesmium due to CO2 estimate, if CO2 increases also stimulate has been a lively debate in the literature fertilization. Current annual global N2 N2 fixation by these less-understood about the relative importance of fixation by Trichodesmium spp. has been marine diazotrophs. and phosphorus as limiting nutrients for estimated to be about 60 x 109 kg N yr-1 (Mahaffey et al., 2005). Calculations using estimates of predicted atmospheric pCO2 increases over the next two 12 ) Figure 3. N2 fixation rates of collected centuries (Royal Society, 2005), along -1 380 ppm natural Trichodesmium spp. colonies hr from the in three sepa- with the N2 fixation rate relationships -1 10 L 750 ppm rate experiments, incubated for four determined across a range of pCO2 for to six hours at current pCO2 (380 ppm, the Pacific Trichodesmium isolate GBR 8 black) and projected year 2100 pCO2 (for Great Barrier Reef) by Hutchins (750 ppm, red). Error bars are the te (µmol N N (µmol te standard deviations of triplicate bottles et al. (2007), yield the potential trends Ra 6 in each treatment. N2 fixation rates at illustrated in Figure 2b. They suggest elevated pCO2 in these three experi- 4 ments were increased by 6%, 21%, and that the minimum levels of CO2 ixation

F 41% above rates at ambient pCO2. stimulation of N fixation we observed 2 2 N in cultures would result in a global 2 1 2 3 9 -1 increase of about 20 x 10 kg N yr , Natural Trichodesmium 9 -1 or a total of 80 x 10 kg N yr fixed by CO2 Incubation Experiments

Oceanography December 2009 133 diazotrophs in the present-day ocean and CO2 fixation and growth rates, traditional Liebig-type single limiting

(Hutchins and Fu, 2008), and adding this suggests a form of independent nutrient relationship, in which CO2 can pCO2 into the mix as another potentially co-limitation between the two (Saito stimulate N2 fixation only if iron limita- limiting factor is likely to muddy these et al., 2008). Trichodesmium popula- tion is first relieved. Thus, it is possible waters even further. It is therefore crucial tions that are apparently P-limited at that iron limitation may constrain the that we understand how the availability present-day pCO2 (Wu et al., 2000; projected future increases in global of these two other potentially limiting Sanudo-Wilhelmy et al., 2001) may in N2 fixation suggested in Figure 2b. nutrients may interact with observed fact be co-limited by CO2 as well, and Clearly, more experiments are needed

CO2 effects on 2N fixers. thus future CO2-mediated increases to examine responses of a wider variety

The interaction of CO2 with P in N2 fixation could happen regardless of marine diazotrophs to the interac- has been tested using a factorial of limiting concentrations of P in the tive effects of changing pCO2 and other experimental approach in which both central gyre . potentially limiting factors, such as iron, P-limited and P-replete Trichodesmium In contrast, Fu et al. (2008a) found a phosphorus, light, and temperature, cultures were grown at two pCO2 levels very different relationship between pCO2 before we can make confident predic-

(380 and 750 ppm; Hutchins et al., and iron availability for Crocosphaera. tions about the future of N2 fixation in a

2007). These results were intriguing, N2 fixation rates scaled with pCO2 in high-CO2 ocean. in that higher pCO2 still stimulated N2 iron-replete cultures (Figure 4a), but in fixation rates even in severely P-limited iron-limited cultures, increasing pCO2 Nitrification cultures. Because either or had no effect on rates of N2 fixation The physiology of nitrifying proteobac-

CO2 additions resulted in higher N2 (Figure 4b). This result suggests a more teria and archaea could be directly influenced by ocean acidification in at least two ways. There is evidence suggesting that the actual enzymatic substrate during the first step of nitri-

) Figure 4. N2 fixation rates of cultures fication by may be NH3, -1 a) Fe-replete of the unicellular cyanobacterium rather than NH+ (Ward, 2008). Because hr 3e-8 4

-1 Crocosphaera grown at 190 ppm the /ammonium (NH /NH+) CO2 (black), 380 ppm CO2 (red), and 3 4 750 ppm CO (green), under both buffer system in seawater has a pK of 2e-8 2 a (a) Fe-replete and (b) Fe-limited about 9.19 (Millero, 2007), the projected growth conditions. N2 fixation rates reduction in seawater pH over the are linearly correlated with pCO2 1e-8 in Fe-replete cultures, but are unaf- next century will reduce the fraction

ixation (nmol cell fected by pCO2 changes in Fe-limited

F of NH3 by nearly 50%, from ~ 6% to 2 cultures. From Fu et al. (2008a) ~ 3% (Bange, 2008). This suggests a 0 possible negative impact of higher )N

-1 b) Fe-limited 3e-8 acidity on marine nitrification rates, hr

-1 due to reductions in available substrate concentrations. 2e-8 On the other hand, because nitrifiers are autotrophs, like many phytoplankton 1e-8 they could also benefit from the

ixation (nmol cell increased availability of CO2 in seawater. F 2

N Almost nothing is presently known 0 190 380 750 about the physiology of inorganic- pCO2 (ppm) carbon-concentrating mechanisms in

134 Oceanography Vol.22, No.4 marine nitrifying chemoautotrophs or calculate changes in nitrification at Denitrification the potential for carbon limitation of projected year 2100 pCO2 levels, we esti- At present, there seems to be little these organisms. The ultimate cause mate a possible reduction in the global reason to expect ocean acidification of CO2 limitation in many prokary- median ammonium-specific oxidation to have direct consequences for the otic and eukaryotic phytoplankton at rate in the surface ocean of 0.195 d-1 physiology of most marine denitrifying present-day pCO2 is the low affinity (from a data compilation by Yool et al., eubacteria and archaea. Denitrification -1 of the Rubisco for CO2 (Rost 2007) to about 0.171 d by the end of and anammox occur in anaerobic envi- et al., 2008). Consequently, nitrifying this century. Using this same pH/nitri- ronments, and these areas in the ocean should be equally vulner- fication relationship (Huesemann et al., typically already have very elevated able to CO2 limitation, because both the 2002), a modeling study by Blackford pCO2 and low pH values. These condi-

NH3-oxidizing genus Nitrosomonas and and Gilbert (2007) projected a 20% tions, coupled with the slow ventilation – the NO2-oxidizing genus reduction in North Sea nitrification rates rates of most marine sediments and also fix carbon using the Calvin cycle at atmospheric CO2 concentrations of suboxic zones, suggest (Ward, 2008). Thus, there is the possi- 1000 ppm. Their model output suggested that near-term effects of increasing bility that the CO2 fertilization effect that the ratio of nitrate:nitrate + ammo- atmospheric pCO2 on their biology and could positively affect the carbon fixation nium in the region would decrease by chemistry may be limited. However, rates of marine nitrifiers, as it does for 5–10% as a consequence. Blackford and the microbes capable of various modes some photoautotrophs. Gilbert (2007) discussed the potential of denitrification are extremely phylo- In fact, the limited experimental ecosystem implications of this change, genetically diverse (Devol, 2008), and evidence currently available suggests including shifts in phytoplankton these assorted groups could respond in that lower pH probably has a net inhibi- community structure and reductions in different ways to ocean acidification. tory effect on nitrification. Huesemann the fluxes of oxidized nitrogen species For instance, nitrate reduction rates by et al. (2002) measured nitrification that support denitrification. some CO2-fixing, hydrogenotrophic rates in two seawater samples in which New preliminary work measuring in sewage sludge they simulated the effects of delib- nitrification rates in both open ocean are inhibited when elevated pCO2 causes erate ocean disposal of waste CO2 by and coastal marine environments as pH values to drop below 6 (Ghafari et al., bubbling seawater with CO2-amended a function of realistically anticipated 2009). However, to our knowledge there air. Although they examined a pH range pCO2 increases was presented by Beman are no studies demonstrating effects of over 0.5 unit increments that extended et al. (2008) at the American Society of realistically expected future ocean pH much lower than any foreseeable levels and Oceanography (ASLO) levels on denitrification rates of ecologi- of “passive” anthropogenic acidification, meeting in Nice, France. They found that cally important marine . they found that ammonium oxidation pCO2 increases from 390 to 750 ppm had Indirect effects of ocean acidifica- rates showed an almost linear decrease generally negative effects on nitrification tion on marine denitrification seem from pH 8 down to pH 6.5, at which rates, and that the of both much more probable, such as those that point nitrification was nearly 100% proteobacterial and archaeal nitrifiers may be driven by changes in the fluxes inhibited. Although the pH increments was lower in samples incubated at the of organic carbon that support their used by Huesemann et al. (2002) were higher pCO2. More studies are needed heterotrophic metabolisms, and/or the coarse, because the relationship was before we will be able to make reliable supply of oxidized nitrogen substrates linear between pH 8 and 6.5, it may be quantitative predictions of changes in for respiration. A recent modeling study justifiable to use it to extrapolate to the nitrification rates in a high CO2 ocean, by Oschlies et al. (2008) examined the smaller pH changes expected over the but such ongoing work promises to help possible effects of CO2-driven changes next century from acidification (roughly shed new light on the future responses in the C:N stoichiometry of phyto- -0.3 to -0.5 units or so). of this crucial component of the plankton on dissolved levels in Using this linear relationship to marine nitrogen cycle. the tropical ocean. Based on the results

Oceanography December 2009 135 of a mesocosm experiment that demon- The Oschlies et al. (2008) modeling discussed above, would greatly increase strated higher C:N export ratios at high results were very sensitive to the the ultimate source of fixed nitrogen to

CO2 (Riebesell et al., 2007, see below), assumed changes in flux C:N denitrifiers. In contrast, the potential they predict a global expansion of stoichiometry, and these authors actually decreases in nitrification rates at lower suboxic water volume by up to 50% over predicted a contrary trend of increased pH previously mentioned would tend the next 100 years as this “excess” organic dissolved oxygen levels in the tropical to decrease the availability and supply carbon is remineralized in subsurface thermocline if carbon export remained of oxidized nitrogen compounds to serve as substrates for denitrification. Although it seems that denitrification will almost certainly respond strongly to an acidifying ocean, these large uncer- …the ongoing and accelerating enrichment tainties suggest that more information of seawater with anthropogenic CO will will be needed to reliably determine the 2 magnitude and even the likely direction li kely have far-reaching effects not only on of many of these responses. “the , but also on the cycles of… biologically required nutrients [N, P, Si]. The P and Si cycles Compared to nitrogen, relatively little research effort has been directed toward

understanding the effects of high CO2 on the marine . The dominant form of inorganic phosphorus ” 2– waters. If this prediction proves correct, at Redfield ratios in the high-CO2 ocean. in seawater at pH 8 is HPO4 (~ 87%), – and the areal extent of denitrifying zones Because there are quite a number of but the fraction of H2PO4 will increase greatly expands, loss of fixed nitrogen contrasting and sometimes conflicting marginally with future acidification. The from the ocean due to denitrification reports on how phytoplankton elemental implications of this chemical speciation could increase dramatically over just stoichiometry may change in response to shift for phosphorus bioavailability are

decades in a high-CO2 ocean. Evidence higher pCO2 (see below), we suggest that uncertain (Hutchins and Fu, 2008). The has been found in the sedimentary the jury is still out on this hypothesis only study we are aware of that exam- nitrogen isotope record for large about future acidification-driven higher ined rates of phosphate uptake (using increases in suboxic zone denitrification C:N ratios resulting in significantly the isotopic tracer 33P) and potential during glacial/interglacial transitions increased global denitrification. dissolved organic phosphorus utiliza- (Altabet et al., 2002). Because deglacia- Basic changes in other critical tion (by assaying alkaline

tions are a time when atmospheric pCO2 processes both within and external activity) as a function of pCO2 is that of increases rapidly, this evidence seems to the ocean N cycle could have big Tanaka et al. (2008). They measured these consistent with the scenario proposed implications for trends in ocean deni- P cycling parameters in a Norwegian by Oschlies et al. (2008). However, most trification over the next century as well. fjord mesocosm experiment at 350, 700,

paleoceanographic studies attribute Several investigators have predicted and 1050 ppm CO2, and although a few the expansion of suboxic zones during large increases in ocean suboxia as a minor trends emerged, they found no

deglaciation to factors such as enhanced result of future sea surface warming statistically significant effects of pCO2 on export due to changing ocean circulation and intensified stratification (Sarmiento P biogeochemistry over the course of a patterns (Robinson et al., 2007), rather et al., 1998; Matear and Hirst, 2003; 24-day incubation period.

than to a direct effect of rising pCO2 Gnanadesikan et al., 2007). CO2-driven Gervais and Riebesell (2001) grew

on carbon export. enhancement of ocean N2 fixation, as the common bloom-forming diatom

136 Oceanography Vol.22, No.4 Skeletonema costatum in P-replete and examined acidification effects on diatom community have been reported in the P-limited cultures and found no differ- silicification, and these few suggest that Ross Sea, with large, heavily silicified ences in C:P ratios at pH values of 8.0 and direct pCO2 impacts are generally small. centric replacing smaller, lightly

8.6, although they did not examine lower Milligan et al. (2004) saw no difference silicified pennates at high pCO2 (Tortell pH values relevant to anthropogenic between the Si quotas of the diatom et al., 2008; Feng et al., in press). In ocean acidification. Burkhardt et al. Thalassiosira weissflogii when cultures another natural community experiment

(1999) also found no major impacts or were grown at 370 or 750 ppm CO2, during the North Atlantic , consistent influence of low pH on the although the quota increased somewhat Feng et al. (2009) combined two pCO2

C:P ratios of six diatoms and a dino- at very low (100 ppm) pCO2. Similarly, treatments (375 and 750 ppm) with two flagellate. Fu et al. (2008b) showed that our own unpublished results using a temperature treatments (ambient and cellular P quotas in the harmful bloom variety of pennate and centric diatoms ambient + 4°C) and saw lower particu- flagellateHeterosigma akashiwo were isolated from a range of marine envi- late Si:N ratios due to reduced diatom unchanged by increasing pCO2 from ronments suggest that large changes in abundance at the higher temperature, 375 to 750 ppm, while those of the diatom silicification are unlikely to occur but no effect on silicification due to Prorocentrum minimum between present-day and projected year pCO2. Similarly, Bellerby et al. (2008) were slightly lower at the higher pCO2. 2100 pCO2 levels. The biggest effect of found that silicate drawdown was

Likewise, Hutchins et al. (2007) and Fu pCO2 that Milligan et al. (2004) observed virtually identical in 350-, 700-, and et al. (2007) found that a similar pCO2 was not on live diatom cells but on 1050-ppm CO2 treatments in the same increase had no effect on the P cell quotas dissolution rates of empty silica frustules; Norwegian mesocosm experiment in of the cyanobacteria Trichodesmium dissolution rates were much higher at which Tanaka et al. (2008) examined erythraeum, , and 750 ppm CO2 than at 370 ppm. They phosphorus utilization (see above). . In contrast, Czerny et al. were uncertain about the reasons for (2009) documented a slight increasing these low-pH-enhanced silica dissolu- Phytoplankton Elemental trend in cellular P quotas with higher tion rates, but their results do suggest Stoichiometry pCO2 for the estuarine cyanobacterium the possibility that increasing pCO2 The elemental ratios of phytoplankton Nodularia spumigena. At this point, the could enhance Si remineralization rates represent the pivotal intersection of available evidence suggests negligible or from sinking particles, and thus poten- the cycles of carbon, nitrogen, and minor effects of projected future changes tially cause shoaling of silicate vertical phosphorus in the ocean. Because of in pCO2 on most phytoplankton phos- profiles and increased Si availability the central role that algal C:N:P ratios phorus requirements. in surface waters. play in our understanding of global The story for silicon is similar to Although the evidence for direct ocean biogeochemistry (Redfield, that of phosphorus. The dominant effects is sparse, changing pCO2 could 1958), the effects of changing pCO2 on form of dissolved silicon in the ocean certainly affect silicon biogeochemistry phytoplankton stoichiometry have been is silicic acid (H4SiO4), which, like indirectly through shifts in algal commu- examined in many experiments dating phosphorus, comprises a buffer system nity structure. A natural community back at least a decade, to some of the in seawater. The relatively basic pKa pCO2 manipulation experiment in the earliest publications in the relatively values for the dissociation of H4SiO4 equatorial Pacific showed lower dissolved new field of ocean acidification studies

(pKa1 = 9.84, pKa2 = 13.2) mean that it is Si:N consumption ratios at 150 ppm (Burkhardt et al., 1999). Despite this nearly all present in the fully protonated CO2 compared to 750 ppm (Tortell fairly large body of results, there is still form at pH 8. For this reason, further et al., 2002). These low ratios were no consensus on whether phytoplankton decreases in seawater pH are unlikely to attributed to a community shift away elemental ratios are likely to be altered substantially shift the chemical specia- from diatoms and toward Phaeocystis in a systematic, predictable manner in a tion of dissolved silicon. spp. at the lower pCO2. CO2-induced future acidified ocean. Only a handful of studies have changes within the diatom We gathered a collection of the

Oceanography December 2009 137 available experimental data on how pCO changes the elemental ratios of 2 2 both cultured phytoplankton and mixed a) Cultured Cyanobacteria C : N

p CO 40 natural phytoplankton communities C : P

(Figure 5). These data include published uture 30 studies, as well as some of our unpub- lished data. We attempted as much as 20 possible to constrain our summary resent to F to studies that examined C:N and/ or N:P ratio changes from present day rom P 10 to roughly expected year 2100 pCO2 levels (between ~ 380 and ~ 750 ppm), 0 although a few of these studies used % Change F Picocyanobacteria Diazotrophs: Crocosphaera and Trichodesmium other concentrations ranging from 1 2 3 4 5 6 7 8 9

150 to 2000 ppm. The unialgal culture 2 40 b) Cultured Eukaryotic Phytoplankton data offer a glimpse of how the stoichi- p CO ometry of individual phytoplankton 30 uture species may respond to changing pCO2 20 under carefully controlled laboratory conditions (Figure 5a,b). The field data 10 are from experiments carried out using resent to F 0

entire natural plankton assemblages, rom P and so these results integrate the net -10 responses and species composition shifts Prymnesiophytes: Harmful Bloom Phaeocystis and and of many algal species as well as any -20 Diatoms Raphidophytes % Change F other organisms (zooplankton, bacteria) 10 11 12 13 14 15 16 17 18 19 20 21 22 present in the incubations (Figure 5c). 2 The details of each experiment differ 30 c) Natural Communitities in important ways, including different p CO marine regimes, algal taxa, pCO2 levels,

uture 20 and incubation methodology, as well as other co-variables examined, such 10 as light and nutrients. Therefore, the resent to F absolute magnitude of C:N and N:P ratio 0 changes are not directly comparable rom P acifi c -10 m between experiments. Nevertheless, oast tica

this summary offers some interesting th Atlantic -20 Norwegian Offshore and Neritic US East

Nor Spring Bloo Ross Sea, Antarc P Equatorial Fjord Korean C Bering Sea Coast insights into what the major directions % Change F and trends in phytoplankton elemental 23 24 25 26 27 28 29 30 31 32 33 34 ratios could potentially be as the ocean continues to take up anthropogenic CO2 over the next century, as well as some cautionary notes about interpretation of experimental results.

138 Oceanography Vol.22, No.4 Figure 5. A compilation of data from the literature as well as some of the authors’ own unpublished Cyanobacteria Cultures experimental results on how increasing pCO from past or present-day levels (150­–420 ppm) 2 Figure 5a presents the collected data to projected future levels (750–2000 ppm) affects the C:N ratios (black) and N:P ratios (red) of important phytoplankton groups in cultures and in mixed natural communities. The trends in from laboratory cultures of marine these elemental ratios are presented as the percent (%) change from present-day to future pCO2; cyanobacteria on stoichiometry changes positive bars indicate increasing ratios, negative bars indicate declining ratios, and a circle symbol on the 0% line indicates no change in C:N (black) or N:P (red). Note that not every study measured in response to higher pCO2. A trend that both parameters. is immediately apparent is that in all of these studies, cyanobacterial C:N and (a) CO -mediated changes in C:N and N:P ratios of cultured marine cyanobacteria, including: 2 N:P ratios either increased or remained • 1 and 2, Synechococcus and Prochlorococcus grown at 380 and 750 ppm CO2 (Fu et al., 2007) • 3 and 4, Fe-replete and Fe-limited Crocosphaera grown at 380 and 750 ppm CO2 (Fu et al., 2008a) the same with increased pCO2; there • 5 and 6, P-replete and P-limited Trichodesmium grown at 380 and 750 ppm CO2 are no reports of decreases in either (Hutchins et al., 2007) of these ratios. Fu et al. (2007) found • 7, Trichodesmium grown at 400 and 900 ppm CO2 (Levitan et al., 2007) • 8, Trichodesmium grown at 380 and 750 ppm CO2 (Barcelos e Ramos et al., 2007) that both C:N and N:P ratios increased • 9, Trichodesmium grown at 370 and 1000 ppm CO (Kranz et al., 2009) 2 (by 9% and 29%) in Synechococcus, but

(b) CO2-mediated changes in C:N and N:P ratios of cultured marine eukaryotic phyto- neither ratio responded to changing plankton, including: pCO2 in Prochlorococcus. The C:N ratios • 10 and 11, the diatoms Asterionella grown at 430 and 820 ppm CO and Skeletonema grown 2 of Fe-replete cultures of the unicellular at 400 and 720 pCO2 (Burkhardt et al., 1999) • 12 and 13, the Antarctic diatom and the Antarctic colonial prymnesiophyte diazotroph Crocosphaera increased with

Phaeocystis grown at 430 and 820 ppm CO2 (recent work of author Fu and colleagues) higher pCO2 (by 8%) but N:P ratios did • 14 and 15, the coccolithophorid Emiliania huxleyi grown under low and high light at 375 and not, while both ratios were unchanged in 750 ppm pCO2 (Feng et al., 2008) • 16, the coccolithophorid Emiliania huxleyi grown at 490 and 750 ppm pCO2 (Iglesias-Rodriguez Fe-limited treatments (Fu et al., 2008a). et al., 2008) Among the laboratory studies using • 17 and 18, a noncalcifying strain of the coccolithophorid Emiliania huxleyi grown under low Trichodesmium, Hutchins et al. (2007) and high light at 360 and 2000 ppm CO2 (Leonardos and Geider, 2005) • 19 and 20, the toxic raphidophyte Heterosigma and the dinoflagellateProrocentrum grown at reported proportionately identical large 375 and 750 ppm pCO (Fu et al., 2008b) 2 N:P ratio increases in both P-replete and • 21 and 22, P-replete and P-limited cultures of the toxic dinoflagellate Karlodinium grown at

430 and 745 ppm CO2 (recent work of author Fu and colleagues) P-limited cultures (18%), similar to the 26% increase seen by Barcelos e Ramos (c) CO2-mediated changes in C:N and N:P ratios of incubated natural phytoplankton commu- nities, including: et al. (2007). Neither Levitan et al. (2007) • 23, a shipboard continuous culture incubation during the North Atlantic spring bloom at nor Kranz et al. (2009) reported N:P 390 and 690 ppm CO2 (Feng et al., 2009) ratios, and among the four studies, only • 24, a shipboard continuous culture incubation in the Ross Sea, , at 380 and 750 ppm Levitan et al. (2007) found increases in CO2 (Feng et al., in press) • 25, a shipboard semicontinuous culture incubation in the equatorial Pacific at 150 and 750 ppm C:N ratios with higher pCO2 (Figure 5a). CO2 (Tortell et al., 2002) • 26 and 27, mesocosm batch culture incubations in a Norwegian fjord at 350 and 700 ppm CO 2 Eukaryotic Phytoplankton and 410 and 710 ppm pCO2 (Riebesell et al., 2007; Engel et al., 2005) • 28, a mesocosm batch incubation in Korean coastal waters at 400 and 750 ppm CO2 Cultures (Kim et al., 2006) Figure 5b shows a compendium of • 29 and 30, shipboard continuous culture incubations in Bering Sea shelf waters at 10°C and 15°C, and 31 and 32, in Bering Sea offshore waters at 10°C and 15°C, all incubated at 370 and results from laboratory CO2 experiments 750 ppm pCO2 (unpublished data from the study presented in Hare et al., 2007) using eukaryotic phytoplankton. As with • 33 and 34, two semicontinuous incubations of a US East Coast estuarine phytoplankton commu- the cyanobacteria, most studies using nity at 380 and 750 ppm pCO2 (unpublished data of author Fu and colleagues) eukaryotic photoautotrophs found either increases or no change in C:N and N:P

ratios at higher pCO2 levels, but there were several exceptions. Burkhardt et al. (1999) saw that the C:N ratios of one diatom increased by 5%, while that of

Oceanography December 2009 139 another decreased by 5%; several other (2009, in press) found no changes in temperatures in the neritic experiment, species examined had similar variable either C:N or N:P ratios as a function while they were unchanged at 10°C and outcomes (data not shown), and this of pCO2 in two shipboard continuous decreased at 15°C (-18%) in the offshore study did not report N:P ratios. Our culture incubations carried out using incubation (Figure 5c). Finally, our unpublished results using Antarctic a /diatom community unpublished results from two separate isolates found no change in C:N ratios during the North Atlantic spring bloom, semi-continuous experiments carried in the chain-forming centric diatom and a diatom/Phaeocystis assemblage out in estuarine waters of the Delaware Chaetoceros, but a large decrease in N:P in the Ross Sea, Antarctica (Figure 5c). Inland Bays along the US East Coast ratios (-20%). This same investigation Tortell et al. (2002) used differences in showed that higher pCO2 depressed also documented decreases in C:N ratios nutrient drawdown to calculate much community C:N ratios in both cases (-11%) coupled with increases in N:P lower N:P utilization ratios (-18%) at (-5% and -13%), while N:P ratios were ratios (22%) in the ecologically dominant high pCO2, coinciding with a shift from either steady or increased (22%). colonial prymnesiophyte Phaeocystis diatoms to Phaeocystis in an equato- The results of these collected inves- antarctica (Figure 5b). Three studies rial Pacific assemblage. A mesocosm tigations suggest several trends. With a using the cosmopolitan coccolithophorid experiment using a diatom-dominated few exceptions, both the C:N and N:P Emiliania huxleyi all showed increases in assemblage from a Norwegian fjord ratios of unialgal cultures of cyanobac- C:N ratios, except that both Leonardos showed an 8% increase in C:N ratios but teria and eukaryotic usually and Geider (2005) and Feng et al. (2008) no change in N:P ratios; these results increased, or did not change, when found that C:N ratios in light-limited were used to suggest large potential pCO2 was elevated from present-day to cultures were unaffected by pCO2. N:P shifts in future global C:N export stoi- projected future levels. In contrast, incu- ratios in E. huxleyi did not respond to chiometry (Riebesell et al., 2007) and the bations of natural populations showed changing pCO2 in any of these studies. consequent increases in ocean suboxia a much greater degree of variability Experiments using harmful bloom flag- discussed in the denitrification section in both the direction and magnitude ellates found increases in both C:N (9%) above (Oschlies et al., 2008). A previous of elemental stoichiometry changes. and N:P (29%) in the toxic estuarine experiment using the same mesocosm These differences could be due in part raphidophyte Heterosigma akashiwo, but facility, but with an assemblage domi- to differences in incubation methods, no change in either ratio in the co-occur- nated instead by coccolithophores (Engel which have included small volume ring dinoflagellate, Prorocentrum et al., 2005), showed a small increase in shipboard continuous cultures similar minimum (Fu et al., 2008b). The toxic C:N (3%) but a large decrease in N:P in principle to laboratory chemostats dinoflagellate Karlodinium veneficum (-12%). Another mesocosm experi- (Hare et al., 2007; Feng et al., 2009, in exhibited increased N:P ratios at higher ment in Korean coastal waters found press), semi-continuous incubations in pCO2 in both P-replete and P-limited no response of C:N ratios to increasing which part of the seawater is removed cultures, but unaltered C:N ratios in both pCO2 but did not report N:P ratios (Kim periodically and replaced with fresh treatments (Figure 5b; recent work of et al., 2006). Two continuous culture filtered water (Tortell et al., 2002; recent author Fu and colleagues). incubations with communities collected work of author Fu and colleagues), and from the continental shelf and offshore very large volume mesocosms in which Natural Assemblage Experiments waters in the Bering Sea (Hare et al., plankton are grown in “batch” mode The collection of natural assemblage 2007), and carried out at both ambient until nutrients are exhausted (Engel experimental results suggests a much (10°C) and elevated (15°C) temperatures, et al., 2005; Kim et al., 2006; Riebesell more mixed picture in terms of the found large decreases in C:N ratios at et al., 2007). It is also highly likely direction of changes in C:N and N:P higher pCO2 levels in three out of four that the major differences in plankton with pCO2; both increases and decreases cases (C:N increased in the offshore, community composition in the various in these ratios have been commonly ambient temperature experiment). In the regimes where experiments have been reported (Figure 5c). Feng et al. same study, N:P ratios increased at both conducted, ranging from the equator to

140 Oceanography Vol.22, No.4 the polar seas and from estuaries to the also perturbed. It is universally recog- evidence that we have available at this remote open ocean, are also responsible nized that the cycle of carbon in the time, we project that there may be large for the variable results from these natural ocean is inextricably tied to the cycles increases in global N2 fixation rates assemblage incubations. Large varia- of the major nutrient elements. While over the next century. This large influx tions between experiments in biological the details and even the directions of of new nitrogen will be due to the CO2 factors like zooplankton , algal particular biogeochemical responses fertilization effect, and to diazotrophic interspecific , and bacterial are also almost certainly the case. Overall, though, it is clear that extrapolation of the results from any particular natural assemblage experi- Like any major ecological shift…our global ment to other regimes, or to the global perturbation of ocean carbonate chemistry is ocean in general, needs to be undertaken with extreme caution and with these not all negative. It will favor some organisms major qualifiers in mind. “ while selecting against others. Conclusions Marine scientists agree that the ocean has mitigated some of the climatic effects of anthropogenic CO2 emissions are still being debated, some plausible range expansion in a warmer ocean.” by absorbing a large fraction of the outlines of future marine nutrient cycles At the same time, global denitrifica- fossil-fuel-derived carbon emitted to are beginning to emerge. tion could also increase substantially, the atmosphere. We now also recognize following increases in ocean suboxia due that ocean ecosystems are paying a The Future Nitrogen Cycle to changes in ocean physics and surface steep price for this service through the The available evidence suggests that the ocean productivity. Oxygen deple- resulting acidification of seawater. Like ocean nitrogen cycle may be altered tion and denitrification could possibly any major ecological shift, though, our more than any other nutrient biogeo- be greatly enhanced through non- global perturbation of ocean carbonate chemical cycle in response to ocean Redfieldian “carbon overconsumption” chemistry is not all negative. It will favor acidification and CO2 fertilization. by phytoplankton growing at high pCO2, some organisms while selecting against The numerous biologically mediated although we suggest that the evidence others. In , environmental changes transformations within this cycle for the latter is still not conclusive (see that spell disaster for one species always seem especially likely to be sensitive below). Interestingly, the predicted represent an opportunity for another. to changes in the acid/base balance of century-time-scale increases in N2 fixa- Ocean food webs will persist by shifting seawater. Many key organisms such as tion (35–65%; Figure 2) are of about the to accommodate the changed conditions, phytoplankton and nitrifiers are auto- same magnitude as the 50% increase in although there is certainly no guarantee trophs, and so may be directly influenced denitrification predicted by the Oschlies that the human species will be happy by increasing CO2 availability. Other et al. (2008) model. Although both of with the results. critical microbial components of the these estimates certainly have large Considering the comprehensive nitrogen cycle such as denitrifiers and uncertainties, if both input and output chemical and biological changes now anammox organisms are heterotrophic, fluxes accelerate roughly equally, there underway as anthropogenic CO2 and so could be secondarily affected by could be a faster throughput but a rela- continues to accumulate in the ocean, changes in both the quantity and quality tively unchanged total fixed nitrogen it would be surprising indeed if linked of organic carbon reaching them. inventory in the ocean. marine biogeochemical cycles were not Based on the best experimental One factor that may prevent such a

Oceanography December 2009 141 balanced but faster spinning nitrogen to go; if nitrification is progressively of the at the expense cycle scenario is the apparent nega- inhibited by acidification, more ammo- of nitrification and new production may tive effect of lower pH on nitrification. nium will likely be diverted into the well negatively impact higher-trophic- Although the deep ocean will not experi- algal assimilatory pathway (Figure 1). level production and fisheries harvests in ence significant anthropogenic acidifica- Annamox could also possibly help to the future acidified ocean. tion for some time (Royal Society, 2005), consume some of this excess ammonium recent estimates place a large fraction of supply, but because anaerobic ammo- Interactions of Other Factors With Ocean Acidification Another large uncertainty is the

interaction of rising pCO2 with other Considering the comprehensive chemical global change processes such as global and biological changes now underway as warming, stratification, changes in ocean circulation and ice cover, and shifts in a nthropogenic CO2 continues to accumulate light and nutrient availability (Boyd et al., “in the ocean, it would be surprising indeed in press). For example, the interactions if linked marine biogeochemical cycles were between CO2 fertilization of and growth limitation by phos- not also perturbed. phorus and/or iron are still very much unresolved. The availability of P is likely to be reduced with intensified stratifica- tion of the future surface ocean. The same is true of iron in zones global nitrification within near-surface” nium oxidation also requires supplies of where phytoplankton receive most of waters just below the euphotic zone oxidized nitrite, it is also vulnerable to their inputs from below, but future trends (Yool et al., 2007). These shallow nitri- acidification-driven decreases in nitri- in iron supplies from continental aero-

fying organisms will almost certainly fication (Figure 1). This scenario would sols in the central gyres where most N2 encounter strong acidification impacts ultimately reduce surface ocean nitrate fixation occurs are more uncertain. Some within this century. Thus, inhibition by concentrations and nitrate-supported modeling studies predict changes in rising acidity could turn nitrification into new , while precipitation patterns over the continents a nitrogen cycle bottleneck, constricting increasing the relative importance of the that will reduce iron inputs to the ocean the flow of fixed N inputs from nitrogen ammonium-fueled regenerated produc- from dust (Mahowald and Luo, 2003), fixation to N losses through denitrifica- tion system. Such an outcome would while other investigators predict poten- tion and annamox. Decreases in nitrifi- undoubtedly have large consequences tially increased aeolian iron supply to cation rates at low pH could negate the for phytoplankton community structure, the ocean from accelerating agricultural trends discussed above that would have as many species have preferences for (Takeda and Tsuda, 2005) otherwise tended to increase denitrifica- specific forms of nitrogen (Mulholland or industrial (Sedwick et al., tion, because the supplies of nitrate that and Lomas 2008). One prediction would 2007). Although we have not extensively are required by denitrifiers as a respira- be increased global biomass of regener- considered iron or other micronutrients

tory would be substan- ated production specialists like pico- in this review, higher pCO2 could alter tially reduced. cyanobacteria and nanoflagellates, and ocean iron cycles through direct changes The potentially increased fluxes of reduced importance of new production in phytoplankton iron requirements

ammonium coming from CO2-fertilized specialists like diatoms. These ammo- (Figure 3; Fu et al., 2008a), and through nitrogen fixers will have only two ways nium-driven increases in the relative size pH-driven changes in the chemical

142 Oceanography Vol.22, No.4 speciation of organically complexed iron problematic. Higher global average algal natural selection and to occur. (Millero et al., 2009). There is a clear N:P ratios would tend to deplete the The even shorter time frame of most need for more holistic experimental and ocean inventory of fixed N relative to P, natural community experiments means modeling studies that attempt to incor- thus favoring increased nitrogen fixation they probably are not able to fully accli- porate feedbacks between rising ocean above and beyond the direct effects of mate to elevated pCO2, and so are best pCO2 and acidity, and all of these other CO2 fertilization. Higher algal C:N ratios interpreted in terms of the proximate concurrent and often linked environ- would provide negative feedback on responses of present-day communities mental change variables. rising atmospheric pCO2 (Riebesell et al., to pCO2 changes, in a manner analo- 2007), but could also potentially lead to gous to a diagnostic iron or nutrient Future Phosphorus worldwide expansion of oxygen-depleted addition experiment. and Silicon Cycles zones (Oschlies et al., 2008). Classic natural selection is driven At present, the P and Si cycles seem less Not surprisingly, the collected results by environmental changes that reduce likely to be directly affected by rising of experiments that manipulated pCO2 an ’s chances of survival or ocean pCO2, but they undoubtedly will in complex natural plankton assemblages reproductive fitness; thus, marine react indirectly to the expected changes are not quite as clear-cut as those from microbes that are negatively impacted in the C and N cycles. Phytoplankton monospecific culture studies. In fact, by higher pCO2 (such as the nitrifiers community composition changes may C:N ratios increased with higher pCO2 and calcifiers) may indeed evolve more be especially important in determining in only three of the natural community rapidly in an acidifying ocean. It is not the way the silicon cycle will respond data sets we examined; in five cases they clear, though, how natural selection in a in the future ocean. The apparent lack declined, and in three experiments they changing environment may affect those of a strong influence of pCO2 on most were unchanged. Trends in N:P ratios organisms that are instead favored by algal cellular P requirements should were similarly mixed. We need to be higher pCO2 (such as Trichodesmium). allow oceanographers to use P as a very cautious about making broad global In fact, the few studies that have looked relatively conservative denominator to generalizations about stoichiometry at the long-term responses (thousands which changing C and N quotas can changes from any particular experi- of generations) of phytoplankton to be compared; fortuitously, we already ment, and much more information will high pCO2 have found no evidence for routinely use P in this way as a normal- be needed to confidently extrapolate the evolution of new traits (Collins and izer in the . incubation results to century-scale, Bell, 2006). However, through random ocean-wide trends. mutational degradation, some cell lines Future Algal Elemental Ratios in these studies did lose existing traits and Nutrient Stoichiometry Evolution, Adaptation, that operate only at low pCO2, such as Far-reaching impacts on future nutrient and Changing Biology in high-affinity dissolved inorganic carbon cycles and biological productivity are the Future Ocean acquisition systems. This observation to be expected if algal elemental ratios Another thing that cannot be reliably raises the possibility that some marine are systematically influenced by pCO2, predicted from the short-term experi- microbes may adapt to a high pCO2 as some studies suggest. Our survey of ments discussed here is how populations ocean in ways that will make them less laboratory culture data suggests that the of biogeochemically critical marine competitive if human remediation efforts C:N and N:P ratios of most cyanobac- microbes may adapt and evolve in are able to eventually lower atmospheric teria and many eukaryotic phyto- response to higher pCO2. Most previous pCO2 once again. More of these types of plankton will either remain near Redfield laboratory CO2 experiments have been experiments examining the long-term values, or increase as pCO2 rises during careful to fully acclimate the cultures responses of key marine functional the coming century. These responses to the experimental conditions over groups are needed before we will be able appear to be species-specific, making multiple generations, but still have not to fully understand how nutrient biogeo- community and ecosystem predictions been carried out long enough for true chemistry and biology will respond

Oceanography December 2009 143 to the combined impacts of acidifica- Boyd, P.W., R. Strzepek, R., F.-X. Fu, and D.A. Fu, F-X., M.R. Mulholland, N.S. Garcia, A. Beck, Hutchins. In press. Environmental control of open P.W. Bernhardt, M.E. Warner, S.A. Sañudo- tion and CO2 fertilization in a rapidly ocean phytoplankton groups: Now and in the Wilhelmy, and D.A. Hutchins. 2008a. Interactions changing ocean. future. Limnology and Oceanography. between changing pCO2, N2 fixation, and Fe Burkhardt, S., I. Zondervan, and U. Riebesell. 1999. limitation in the marine unicellular cyanobacte-

Effect of CO2 concentration on C:N:P ratio in rium Crocosphaera. Limnology and Oceanography Acknowledgements marine phytoplankton: A species comparison. 53:2,472–2,484. We thank P.W. Bernhardt, N. Garcia, Limnology and Oceanography 44(3):683–690. Fu, F.-X., M.E. Warner, Y. Zhang, Y. Feng, and Carpenter, E.J., and D.G. Capone. 2008. Nitrogen D.A. Hutchins. 2007. Effects of increased M. Warner, and C. Heil for their help fixation. Pp. 141–198 in Nitrogen in the Marine temperature and CO2 on , growth with our field research, and D. Bronk Environment, 2nd ed. D.G. Capone, D.A. Bronk, and elemental ratios of marine Synechococcus M.R. Mulholland, and E.J. Carpenter, eds, Elsevier and Prochlorococcus (Cyanobacteria). Journal of and an anonymous reviewer for their Press, Amsterdam. Phycology 43:485–496. helpful comments. Grant support was Collins, S., and G. Bell. 2006. Evolution of natural Fu, F.-X., Y. Zhang, M.E. Warner, Y. Feng, and provided by NSF OCE 0722337 and algal populations at elevated CO2. Ecological D.A. Hutchins. 2008b. A comparison of future Letters 9:129–135. increased CO2 and temperature effects on 0825319 and EPA ECOHAB R83-3221 Czerny, J., J. Barcelos e Ramos, and U. Riebesell. 2009. sympatric Heterosigma akashiwo and Prorocentrum to D.A. Hutchins, NSF OCE 0722395 Influence of elevated CO2 concentrations on cell minimum. Harmful Algae 7, doi:10.1016/ division and nitrogen fixation rates in the bloom- j.hal.2007.05.006. to M.R. Mulholland, and NSF OCE forming cyanobacterium Nodularia spumigena. Galloway, J.N., F.J. Dentener, D.G. Capone, E.W. Boyer, 0850730 to F.-X. Fu. Special thanks Biogeosciences 6:1,865–1,875. R.W. Howarth, S.P. Seitzinger, G.P. Asner, to C. Baranyk for contributing her Devol, A.H. 2008. Denitrification including C.C. Cleveland, P.A. Green, E.A. Holland, and anammox. Pp. 263–302 in Nitrogen in the Marine others. 2004. Nitrogen cycles: Past, present, and original artwork. Environment, 2nd ed. D.G. Capone, D.A. Bronk, future. Biogeochemistry 70:153–226. M.R. Mulholland, and E.J. Carpenter, eds, Elsevier Gervais, F., and U. Riebesell. 2001. Effect of phos- Press, Amsterdam. phorus limitation on elemental composition and References Engel, A., I. Zondervan, K. Aerts, L. Beaufort, stable carbon isotope fractionation in a marine Altabet, M.A., M.J. Higginson, and D.W. Murray. A. Benthien, L. Chou, B. Delille, J.-P. Gattuso, diatom growing under different CO2 concentra- 2002. The effect of millennial-scale changes in J. Harlay, C. Heemann, and others. 2005. Testing tions. Limnology and Oceanography 46(3):497–504. Arabian Sea denitrification on atmospheric CO2. the direct effect of CO2 concentration on a bloom Gnanadesikan, A., J.L. Russell, and F. Zang. 2007. How Nature 415:159–162. of the coccolithophorid Emiliania huxleyi in meso- does the ocean ventilation change under global Bange, H.W. 2008. Gaseous nitrogen compounds (NO, cosm experiments. Limnology and Oceanography warming? Ocean Science 3:43–53. N2O, N2, NH3) in the ocean. Pp. 51–94 in Nitrogen 50(2):493–507. Ghafari, S., M. Hasan, and M.K. Aroua. 2009. Effect nd in the Marine Environment, 2 ed. D.G. Capone, Fabry, V.J., C. Langdon, W.M. Balch, A.G. Dickson, of and bicarbonate as inorganic D.A. Bronk, M.R. Mulholland, and E.J. Carpenter, R.A. Feely, B. Hales, D.A. Hutchins, J.A. Kleypas, carbon sources on growth and adaptation of auto- eds, Elsevier Press, Amsterdam. and C.L. Sabine. 2008. Ocean acidification’s hydrogenotrophic denitrifying bacteria. Journal of Barcelos e Ramos, J.B.E., H. Biswas, K.G. Schulz, effects on marine ecosystems and biogeochem- Hazardous Materials 162:1,507–1,513. J. LaRoche, and U. Riebesell. 2007. Effect istry. Eos, Transactions, American Geophysical Hare, C.E., K. Leblanc, G.R. DiTullio, R.M. Kudela, of rising atmospheric carbon dioxide on Union 89:143–144. Y. Zhang, P.A. Lee, S. Riseman, P.D. Tortell, and the marine nitrogen fixer Trichodesmium. Feely, R.A., C.L. Sabine, K. Lee, W. Berelson, D.A. Hutchins. 2007. Consequences of increased Global Biogeochemical Cycles 21, GB2028, J. Kleypas, V.J. Fabry, and F.J. Millero. 2004. Impact temperature and CO2 for algal community doi:10.1029/2006GB002898. of anthropogenic CO2 on the CaCO3 system in the structure and biogeochemistry in the Bering Sea. Bellerby, R.G.J., K.G. Schulz, U. Riebesell, C. Neill, . Science 305:362–366. Marine Progress Series 352:9–16. G. Nondal, E. Heegaard, T. Johannessen, and Feng, Y., W.E. Warner, Y. Zhang, J. Sun, F.-X. Fu, Huesemann, M.H., A.D. Skillman, and E.A. Crecelius. K.R. Brown. 2008. community and D.A. Hutchins. 2008. Interactive effects 2002. The inhibition of marine nitrification by carbon and nutrient uptake stoichiometry under of increased pCO2, temperature and irra- ocean disposal of carbon dioxide. varying ocean acidification during the PeECE III diance on the marine coccolithophore Bulletin 44:142–148. experiment. Biogeosciences 5:1,517–1,527. Emiliania huxleyi (Prymnesiophyceae). Hutchins, D.A., and F.-X. Fu. 2008. Linking the Beman, J.M., C.E. Chow, B.N. Popp, J.A. Fuhrman, European Journal of Phycology 43:87–98, oceanic biogeochemistry of iron and phosphorus Y. Feng, and D.A. Hutchins. 2008. Alteration of doi:10.1080/09670260701664674. with the marine nitrogen cycle. Pp. 1,627–1,653 oceanic nitrification under elevated carbon dioxide Feng, Y., C.E. Hare, K. Leblanc, J. Rose, Y. Zhang, in Nitrogen in the Marine Environment, 2nd ed. concentrations. Paper presented at ASLO meeting, G.R. DiTullio, P.A. Lee, S.W. Wilhelm, J.M. Rowe, D.G. Capone, D.A. Bronk, M.R. Mulholland, and Nice, France, January 25–30, 2008. Abstract J. Sun, and others. 2009. The effects of increased E.J. Carpenter, eds, Elsevier Press, Amsterdam. available online at: http://www.sgmeet.com/aslo/ pCO2 and temperature on the North Atlantic Hutchins, D.A., F.-X. Fu, Y. Zhang, M.E. Warner, nice2009/viewabstract2.asp?AbstractID=5886 Spring Bloom. I. The phytoplankton community Y. Feng, K. Portune, P.W. Bernhardt, and M.R. (accessed November 17, 2009). and biogeochemical response. Marine Ecology Mulholland. 2007. CO2 control of Trichodesmium Blackford, J.C., and F.J. Gilbert. 2007. pH variability Progress Series 388:13–25. N2 fixation, photosynthesis, growth rates, and and CO2 induced acidification in the North Sea. Feng, Y., C.E. Hare, J.M. Rose, S.M. Handy, elemental ratios: Implications for past, present, Journal of Marine Systems 64:229–241. G.R. DiTullio, P.A. Lee, W.O. Smith Jr., J. Peloquin, and future ocean biogeochemistry. Limnology and S. Tozzi, J. Sun, and others. In Press. Interactive Oceanography 52:1,293–1,304.

effects of iron, irradiance, and CO2 on Ross Sea phytoplankton. Deep-Sea Research Part I.

144 Oceanography Vol.22, No.4 Iglesias-Rodriguez, M.D., P.R. Halloran, R.E.M. Mulholland, M.R., and P.W. Bernhardt. 2005. The Sarmiento, J.L., T.M.C. Hughes, R.J. Stouffer, and Rickaby, I.R. Hall, E. Colmenero-Hidalgo, effect of growth rate, phosphorus concentration, S. Manabe. 1998. Simulated response of the ocean

J.R. Gittins, D.R.H. Green, T. Tyrrell, S.J. Gibbs, and temperature on N2 fixation, carbon fixa- carbon cycle to anthropogenic climate warming. P. von Dassow, and others. 2008. Phytoplankton tion, and nitrogen release in continuous cultures Nature 393:245–249, doi:10.1038/30455.

calcification in a high-CO2 world. Science of Trichodesmium IMS101. Limnology and Sedwick, P.N., E.R. Sholkovitz, and T.M. Church. 320(5874):336–340, doi:10.1126/science.1154122. Oceanography 50:839–849. 2007. Impact of anthropogenic combus- Kleypas, J.A., R.W. Buddemeier, D. Archer, Najjar, R.G., C.R. Pyke, M.B. Adams, D. Breitburg, tion emissions on the fractional of J.-P. Gattuso, C. Langdon, and B.N. Opdyke. M. Kemp, C. Hershner, R. Howarth, iron: Evidence from the Sargasso Sea. 1999. Geochemical consequences of increased M. Mulholland, M. Paolisso, D. Secor, K. Sellner, Geophysics Geosystems 8, Q10Q06, atmospheric carbon dioxide on reefs. D. Wardrop, and R. Wood. In press. Potential doi:10.1029/2007GC001586. Science 284:118–120. climate-change impacts on the Chesapeake Bay. Takeda, S., and A. Tsuda. 2005. An in situ iron- Kim, J.-M., K. Lee, K. Shin, J.-H. Kang, H.-W. Lee, Estuarine, Coastal, and Shelf Science. enrichment in the western subarctic Pacific M. Kim, P.-G. Jang, and M.-C. Jang. 2006. The Oschlies, A., K.G. Schulz, U. Riebesell, and (SEEDS): Introduction and summary. Progress in

effect of seawater CO2 concentration on growth A. Schmittner. 2008. Simulated 21st century’s Oceanography 64:95–109.

of a natural phytoplankton assemblage in a increase in oceanic suboxia by CO2-enhanced Tanaka, T., T.F. Thingstad, T. Løvdal, H.-P. Grossart, controlled mesocosm experiment. Limnology and biotic carbon export. Global Biogeochemical A. Larsen, M. Allgaier, M. Meyerhofer, K.L. Schulz, Oceanography 51(4):1,629–1,636. Cycles 22, GB4008, doi:10.1029/2007GB003147. J. Wohlers, E. Zollner, and U. Riebesell. 2008. Kranz, S.A., D. Sültemeyer, K.-U. Richter, and B. Rost. Prufert-Bebout, L., H.W. Paerl, and C. Lassen. 1993. Availability of phosphate for phytoplankton 2009. Carbon acquisition by Trichodesmium: Growth, nitrogen fixation, and spectral attenuation and bacteria and of glucose for bacteria at

The effect of pCO2 and diurnal changes. 2009. in cultivated Trichodesmium species. Applied and different pCO2 levels in a mesocosm study. Limnology and Oceanography 54(2):548–559. Environmental 59:1,367–1,375. Biogeosciences 5:669–678. Leonardos, N., and R.J. Geider. 2005. Elevated atmo- Redfield, A.C. 1958. The biological control of Tortell, P.D., G.R. DiTullio, D.M. Sigman, and F.M.M.

spheric carbon dioxide increases organic carbon chemical factors in the environment. American Morel. 2002. CO2 effects on taxonomic composi- fixation by Emiliania huxleyi (Haptophyta) under Scientist 46(3):205–221. tion and nutrient utilization in an Equatorial

nutrient-limited high-light conditions. Journal of Riebesell, U. 2004. Effects of CO2 enrichment on Pacific phytoplankton assemblage. Marine Ecology Phycology 41:1,196–1,203. marine phytoplankton. Journal of Oceanography Progress Series 236:37–43. Levitan, O., G. Rosenberg, I. Setlik, E. Setlikova, 60(4):719–729. Tortell, P.D., C.D. Payne, Y. Li, S. Trimborn, B. Rost, J. Grigel, J. Klepetar, O. Prasil, and I. Berman- Riebesell, U., K.G. Schulz, R.G.J. Bellerby, M. Botros, W.O. Smith, C. Riesselman, R.B. Dunbar,

Frank. 2007. Elevated CO2 enhances nitrogen P. Fritsche, M. Meyerhofer, C. Neill, G. Nondal, P. Sedwick, and G.R. DiTullio. 2008. CO2 fixation and growth in the marine cyanobacterium A. Oschlies, J. Wohlers, and E. Zollner. 2007. sensitivity of Southern Ocean phytoplankton. Trichodesmium. Global Change Biology 13:531–538. Enhanced biological carbon consumption in a Geophysical Research Letters 35, L04605,

Mahaffey, C., A.F. Michaels, and D.G. Capone. 2005. high CO2 ocean. Nature 450:545–548, doi:10.1038/ doi:10.1029/2007GL032583.

The conundrum of marine N2 fixation. American nature06267. Ward, B.B. 2008. Nitrification in marine systems. Journal of Science 305:546–595. Robinson, R.S., A. Mix, and P. Martinez. 2007. Pp. 199–262 in Nitrogen in the Marine Mahowald, N.M., and C. Luo. 2003. A less dusty Southern Ocean control on the extent of denitrifi- Environment, 2nd ed. D.G. Capone, D.A. Bronk, future? Geophysical Research Letters 30(17), 1903, cation in the southeast Pacific over the last 70 ky. M.R. Mulholland, and E.J. Carpenter, eds, Elsevier doi:10.1029/2003GL017880. Quaternary Science Review 26:2,001–2,212. Press, Amsterdam. Matear, R.J., and R.C. Hirst. 2003. Long-term Rost, B., I. Zondervan, and D. Wolf-Gladrow. 2008. Wu, J., W. Sunda, E.A. Boyle, and D.M. Karl. 2000. changes in dissolved oxygen concentrations in Sensitivity of phytoplankton to future changes in Phosphate depletion in the western North Atlantic the ocean caused by protracted global warming. ocean carbonate chemistry: Current knowledge, Ocean. Science 289:759–762. Global Biogeochemical Cycles 17(4):1,125, contradictions and research directions. Marine Yool, A., A.P. Martin, C. Fernandez, and D.R. Clark. doi:10.1029/2002GB001997. Ecology Progress Series 373:227–237. 2007. The significance of nitrification for oceanic Millero, F.J. 2007. The marine inorganic carbon cycle. The Royal Society. 2005. Ocean acidification due to new production. Nature 447:1,000–1,002, Chemical Reviews 107:308–341. increasing atmospheric carbon dioxide. Policy doi:10.1038/nature05885. Millero, F.J., R. Woosley, B. DiTrolio, and J. Waters. document 12/05 ISBN 0 85403 617 2. Abstract Zehr, J.P., J.B. Waterbury, P.J. Turner, J.P. Montoya, 2009. Effect of ocean acidification on the speciation available online at: http://www.royalsoc.ac.uk/ E. Omoregie, G.F. Steward, A. Hansen, and of metals in seawater. Oceanography 22(4):72–85. Report_WF.aspx?pageid=9633&terms=Ocean+ D.M. Karl. 2001. Unicellular cyanobacteria

Milligan, A.J., D.E. Varela, M.A. Brzezinski, and acidification+due+to+increasing+atmospheric+ fix N2 in the subtropical North Pacific Ocean. F.M.M. Morel. 2004. Dynamics of silicon metabo- carbon+dioxide (accessed November 17, 2009). Nature 412:635–638. lism and silicon isotopic discrimination in a Saito, M.A., T.J. Goepfert, and J.T. Ritt. 2008. Some

marine diatom as a function of pCO2. Limnology thoughts on the concept of colimitation: Three and Oceanography 49(2):322–329. definitions and the importance of bioavailability. Montoya, J.P., C.M. Holl, J.P. Zehr, A. Hansen, T.A. Limnology and Oceanography 53:276–290. Villareal, and D.G. Capone. 2004. High rates of Sañudo-Wilhelmy, S.A., A.B. Kustka, C.J. Gobler,

N2 fixation by unicellular diazotrophs in the olig- D.A. Hutchins, M. Yang, K. Lwiza, J. Burns, otrophic Pacific Ocean. Nature 430:1,027–1,032. D.G. Capone, J.A. Raven, and E.J. Carpenter. Mulholland, M.R., and M.W. Lomas. 2008. Nitrogen 2001. Phosphorus limitation of nitrogen fixation uptake and assimilation. Pp. 303–385 in Nitrogen by Trichodesmium in the central North Atlantic in the Marine Environment, 2nd ed. D.G. Capone, Ocean. Nature 411:66–69. D.A. Bronk, M.R. Mulholland, and E.J. Carpenter, eds, Elsevier Press, Amsterdam.

Oceanography December 2009 145