Contribution of Fish to the Marine Inorganic CarbonCycleR. W. Wilson, et al.Science 323, 359 (2009);DOI: 10.1126/science.1157972The following resources related to this article are available online (this information is current as of January 21, 2009 ):Supporting Online Material can be found 5912/359/DC1A list of selected additional articles on the Science Web sites related to this article can befound 5912/359#related-contentThis article cites 32 articles, 7 of which can be accessed for 3/5912/359#otherarticlesThis article appears in the following subject cgi/collection/oceansInformation about obtaining reprints of this article or about obtaining permission to reproducethis article in whole or in part can be found Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright2009 by the American Association for the Advancement of Science; all rights reserved. The title Science is aregistered trademark of AAAS.Downloaded from on January 21, 2009Updated information and services, including high-resolution figures, can be found in the onlineversion of this article 5912/359

REPORTSReferences and Notes1. H. C. Urey, Geochim. Cosmochim. Acta 1, 209 (1951).2. N. F. Ness, K. W. Behannon, C. S. Scearce,S. C. Cantarano, J. Geophys. Res. 72, 5769 (1967).3. P. Dyal, C. W. Packer, C. P. Sonett, Science 169, 762(1970).4. S. K. Runcorn et al., Science 167, 697 (1970).5. J. S. Halekas, R. P. Lin, D. L. Mitchell, Meteorit. Planet.Sci. 38, 565 (2003).6. D. L. Mitchell et al., Icarus 194, 401 (2008).7. N. Sugiura, Y. M. Wu, D. W. Strangway, G. W. Pearce,L. A. Taylor, Proc. Lunar Planet. Sci. Conf. 10, 2189(1979).8. L. J. Srnka, Proc. Lunar Sci. Conf. 8, 785 (1977).9. D. A. Crawford, P. H. Schultz, Int. J. Impact Eng. 23, 169(1999).10. L. L. Hood, N. A. Artemieva, Icarus 193, 485 (2008).11. R. R. Doell, C. S. Gromme, A. N. Thorpe, F. E. Senftle,Science 167, 695 (1970).12. S. K. Rucorn, Nature 275, 430 (1978).13. M. Fuller, S. M. Cisowski, in Geomagnetism, vol. 2, J. A.Jacobs, Ed. (Academic Press, New York, 1987), pp. 307–456.14. K. P. Lawrence, C. L. Johnson, L. Tauxe, J. Gee, Phys.Earth Planet Int. 10.1016/j.pepi.2008.05.007 (2008).15. D. L. Shuster, B. P. Weiss, Science 309, 594 (2005).16. D. J. Stevenson, Rep. Prog. Phys. 46, 555 (1983).17. R. Gooley, R. Brett, J. R. Smyth, J. Warner, Geochim.Cosmochim. Acta 38, 1329 (1974).18. R. F. Dymek, A. L. Albee, A. A. Chodos, Proc. Lunar Sci.Conf. 6, 301 (1975).19. G. L. Nord, Proc. Lunar Sci. Conf. 7, 1875 (1976).20. D. E. Wilhelms, “The Geologic History of the Moon”[Professional Paper 1348, U.S. Geological Survey (USGS),Government Printing Office, Washington, DC, 1987].21. E. W. Wolfe, “The Geologic Investigation of theTaurus-Littrow Valley: Apollo 17 Landing Site”(Professional Paper 1080, Government Printing Office,Washington, DC, 1981).22. J. C. Huneke, G. J. Wasserburg, Lunar Sci. VI, 417 (1975).23. G. W. Lugmair, K. Marti, J. P. Kurtz, N. B. Scheinin,Proc. Lunar Sci. Conf. 7, 2009 (1976).24. W. R. Premo, M. Tatsumoto, Proc. Lunar Planet. Sci. Conf.22, 381 (1992).25. L. Husain, O. A. Schaeffer, Geophys. Res. Lett. 2, 29(1975).26. D. D. Bogard, L. E. Nyquist, B. M. Bansal, H. Wiesmann,C.-Y. Shih, Earth Planet. Sci. Lett. 26, 69 (1975).27. D. A. Papanastassiou, G. J. Wasserburg, Proc. Lunar Sci.Conf. 7, 2035 (1976).28. G. Ryder, M. D. Norman, R. A. Score, Proc. Lunar Planet.Sci. Conf. 11, 471 (1980).29. R. M. Bozorth, Ferromagnetism. (IEEE Press, New York,1951), pp. 968.30. G. W. Pearce, D. W. Strangway, “Apollo 16: PreliminaryScience Report” (SP-315, NASA, 1972), chap. 7C,pp. 7–55.31. J. Gattacceca, P. Rochette, Earth Planet. Sci. Lett. 227,377 (2004).32. J. Gattacceca et al., Phys Earth Planet Inter 166,1 (2008).33. J. Pohl, A. Eckstaller, Lunar Planet. Sci. 12, 851 (1981).34. A. Bischoff, D. Stoffler, Eur. J. Mineral. 4, 707 (1992).35. N. Artemieva, B. Ivanov, Icarus 171, 84 (2004).36. L. Carporzen, S. A. Gilder, R. J. Hart, Nature 435, 198(2005).37. Because some kamacite would have exsolved from taeniteduring slow-cooling in the deep lunar crust, the HCcomponent is probably a mixture of nonshock TRM andphase-transformation crystallization remanentmagnetization (SOM text).38. I. S. McCallum, J. M. Schwartz, J. Geophys. Res. 106,27969 (2001).39. I. S. McCallum et al., Geochim. Cosmochim. Acta 70,6068 (2006).40. D. H. Lindsley, D. J. Andersen, Proc. Lunar Planet. Sci.Conf. 13, A887 (1983).41. C. T. Herzberg, Lunar Planet. Sci. 10, 537 (1979).42. R. H. Hewins, J. I. Goldstein, Lunar Planet. Sci. 6, 356(1975).43. K. Righter, A. J. Campbell, M. Humayun, Geochim.Cosmochim. Acta 69, 3145 (2005).44. A. Meibom et al., Science 288, 839 (2000).45. J. R. Smyth, Proc. Lunar Planet. Sci. Conf. 17, E91(1986).46. D. Braddy, I. D. Hutcheon, P. B. Price, Proc. Lunar Sci.Conf. 6, 3587 (1975).47. C. W. Naeser, H. Faul, J. Geophys. Res. 74, 705 (1969).48. R. A. Ketcham, R. A. Donelick, W. D. Carlson, Am.Mineral. 84, 1235 (1999).49. R. B. Merrill, M. W. M. McElhinny, The Magnetic Fieldof the Earth: Paleomagnetism, the Core, and the DeepMantle (Academic Press, San Diego, 1998), p. 531.50. It is plausible that a small ( 10 cm), highly localized,magmatic dike could heat a sample for short timescales;however, such an event would be highly fortuitous, andeven more fortuitous to have taken place simultaneouslywith an impact event.51. D. W. Collinson, Surv. Geophys. 14, 89 (1993).52. D. R. Stegman, M. A. Jellinek, S. A. Zatman,J. R. Baumgardner, M. A. Richards, Nature 421, 143(2003).53. M. A. Wieczorek et al., Rev. Mineral. Geochem. 60, 221(2006).54. J. R. Williams, D. H. Boggs, C. F. Yoder, J. T. Ratcliff,J. Geophys. Res. 106, 27933 (2001).55. S. Goossens, K. Matsumoto, Geophys. Res. Lett. 35,L02204 (2008).56. H. Fechtig, S. T. Kalbitzer, in Potassium Argon Dating,O. A. Schaeffer, J. Zahringer, Eds. (Springer-Verlag,New York, 1966), pp. 68–107.57. We thank the Johnson Space Center staff and theCuration and Analysis Planning Team forExtraterrestrial Materials for allocating 76535;V. Fernandes for insights into lunar 40Ar/39Argeochronology; I. S. McCallum for discussions aboutthermobarometry; S. Slotznick and S. Pedersen for helpwith the paleomagnetic analyses; M. Zuber andT. Bosak for suggestions; and K. Willis for administrativehelp. B.P.W., D.L.S., and I.G.-B. thank the NASA LunarAdvanced Science and Exploration Research Program;B.P.W. thanks the Charles E. Reed Faculty InitiativesFund for support; and D.L.S. thanks the Ann andGordon Getty Foundation.Downloaded from on January 21, 2009( 350-km radius) partially liquid core. Furthermore, the field that magnetized 76535, which is 300 million years older than that recorded by allpreviously studied lunar samples, is from theearly epoch when the Moon would have mostlikely had a convecting core due to enhanced heatflow and a possible cumulate overturn event (52).Finally, the NRM in 76535 indicates that minimum paleointensities were of order microteslas,consistent with the theoretical expectations for alunar core dynamo (53). Our data and theseconsiderations suggest that at 4.2 Ga, the Moonpossessed a dynamo field, and by implication aconvecting metallic core.Supporting Online 12/356/DC1SOM TextFigures S1 to S12Tables S1 to S3References6 October 2008; accepted 3 December 200810.1126/science.1166804Contribution of Fish to the MarineInorganic Carbon CycleR. W. Wilson,1* F. J. Millero,2* J. R. Taylor,2 P. J. Walsh,2,3 V. Christensen,4S. Jennings,5 M. Grosell2*Oceanic production of calcium carbonate is conventionally attributed to marine plankton(coccolithophores and foraminifera). Here we report that marine fish produce precipitatedcarbonates within their intestines and excrete these at high rates. When combined with estimatesof global fish biomass, this suggests that marine fish contribute 3 to 15% of total oceaniccarbonate production. Fish carbonates have a higher magnesium content and solubility thantraditional sources, yielding faster dissolution with depth. This may explain up to a quarter of theincrease in titratable alkalinity within 1000 meters of the ocean surface, a controversialphenomenon that has puzzled oceanographers for decades. We also predict that fish carbonateproduction may rise in response to future environmental changes in carbon dioxide, and thusbecome an increasingly important component of the inorganic carbon cycle.he inorganic half of the marine carboncycle includes biogenic reaction of seawater calcium (Ca2 ) with bicarbonate(HCO3 ), producing insoluble calcium carbonate(CaCO3) in the process of calcification (1):TCa2 2HCO3 CaCO3 CO2 H2OThe vast majority of oceanic calcification isby planktonic organisms (2). Coccolithophoresare considered to be the major contributor, butforaminifera are also included in global carbonatebudgets (3). Upon death, their carbonate “skeletons” are released and rapidly sink to deeperocean layers. Based on observations and models,estimates of global production of new CaCO3www.sciencemag.orgSCIENCEVOL 323range from 0.7 to 1.4 Pg CaCO3-C year 1 (4–7)(Fig. 1).It is less widely known that all marine teleosts(bony fish) produce and excrete carbonate pre1School of Biosciences, University of Exeter, Exeter EX4 4PS,UK. 2Rosenstiel School of Marine and Atmospheric Science,University of Miami, Miami, FL 33149–1098, USA. 3Universityof Ottawa, Ottawa, ON K1N 6N5, Canada. 4Fisheries Centre,University of British Columbia, Vancouver, BC V6T 1Z4,Canada. 5Centre for Environment, Fisheries and AquacultureScience, Lowestoft, and School of Environmental Sciences,University of East Anglia, Norwich NR4 7TJ, UK.*To whom correspondence should be addressed. E-mail: [email protected] (R.W.W.), [email protected] (F.J.M.),and [email protected] (M.G.)16 JANUARY 2009359

360ranging from 3.2 1012 to 8.9 1012 mol year 1(0.04 to 0.11 Pg of CaCO3-C year 1). This rangeaccounts for 2.7 to 15.4% of estimates for totalglobal new CaCO3 production in the surfaceoceans.Several potentially biasing assumptions aremade in these calculations, but we adopted aconservative approach that, if anything, underestimates fish carbonate production. Adoptingthe more liberal of these realistic assumptionswould yield estimates almost three times as high,i.e., 9 to 45% of total global new CaCO3 production (see Supporting Online Material for detailsof the above calculations and assumptions). Despite this conservatism, our estimate shows thatfish are a major but previously unrecognized sourceof oceanic carbonate and contribute substantiallyto the marine inorganic carbon cycle (Fig. 1).An important question following from thisdiscovery is how the nature and fate of piscinecarbonates compares with those from traditionally accepted sources. At the higher pressure andcolder temperatures of the deep ocean, seawaterbecomes undersaturated with respect to CaCO3,leading to dissolution as it sinks, in a reversal ofreaction 1; thus, the concentration of dissolvedHCO3 and CO32 increases with depth [measured as an increase in the total alkalinity (TA) ofseawater]. Pelagic CaCO3 particles from traditional sources are predicted to dissolve once theyreach the chemical lysoclines for either calcite( 4300 and 750 m) or aragonite ( 1500 and500 m), respectively, in the North Atlantic andPacific Oceans (1, 17–20). However, contrary tothis view, recent carbonate budgets suggest thatthe majority (50 to 71%) of carbonates exportedfrom surface waters dissolve at much shallowerdepths (4, 5, 21). This results in an increase in TAfrom 2400 mM to 2480 and 2500 mM at 1000-mdepth, in the North Atlantic and Pacific oceans,respectively (1) (Fig. 3), a controversial phenomenon that has puzzled oceanographers for decades (7).The causes of CaCO3 dissolution above thelysocline (7) are subject to debate and have beenattributed to (i) dissolution in zooplankton guts(22–26); (ii) dissolution in microenvironmentswhere bacterial oxidation of organic matter enhances this process (27); and (iii) dissolution ofmore soluble forms of CaCO3, including pteropods and high-magnesium calcite (28, 29). However, dissolution in copepod guts can account foronly a small portion of the increase of TA (27).The sharp increase in TA in the Pacific indicatesthat a more soluble phase may be dissolving(28, 29), such as high-magnesium calcites thatare twice as soluble as aragonite (30, 31). Wesuggest that a large portion of the increasing TAin surface waters is indeed related to the dissolution of high-magnesium calcites produced byfish. Given their high magnesium content (8)(fig. S1) and solubility, we predict that dissolution of piscine carbonates will make a majorcontribution (up to 26%) to the increase in TA inthe shallower oceanic depths and helps at leastpartially explain this currently perplexing observation (7) (Fig. 3).The above estimate is a global average forfish-derived carbonates and does not take intoaccount the potential for regional hot spots ofpiscine carbonate production (figs. S4 and S5).Indeed, 50% of fish biomass is predicted to occurGlobal New Production 0.7-1.4 Pg CaCO3–C year-101Depth (km)cipitates. Walsh et al. (8) originally suggestedthat this might be quantitatively significant on alarge scale, an idea not previously consideredwithin a global carbonate budget framework. Carbonate precipitates are excreted by fish via theintestine as a by-product of the osmoregulatoryrequirement to continuously drink calcium- andmagnesium-rich seawater, and they are producedwhether or not fish are feeding (9). As imbibedseawater passes through the intestine, it is alkalinized (to pH 8.5 to 9.2) along with substantialsecretion of HCO3 ions, typically reaching 50 to100 mM in gut fluid (8–11), well in excess ofconcentrations in seawater ( 2.5 mM). Theseconditions cause precipitation of imbibed Ca2 (and some Mg2 ) ions as insoluble carbonates(8–11). This process has physiological importance in facilitating water absorption by the gut(10), and it reduces calcium absorption, whichsecondarily protects the kidney by minimizingrenal stone formation (12). Carbonate precipitatesformed in the gut are excreted either within discrete mucus-coated tubes or pellets, or incorporated with feces when fish are feeding (8–10).The organic mucus-matrix is rapidly degraded innatural seawater, leaving only inorganic crystalsof CaCO3 with high magnesium content (Mg:Caratio ranging from 10 to 33 mol %) (8) (fig. S1).A striking visual indication of the high rate ofcarbonate production in marine fish is providedby x-rays of European flounder (Platichthys flesus)after acute transfer from fresh water (in which theydo not produce carbonates) to seawater (Fig. 2).Accumulations of the precipitates (more x-rayopaque than some of the surrounding bones) canbe seen forming inside the intestine within 3hours of fish initiating drinking after transfer.Excreted carbonates have been collected andtitrated to reveal production rates in the temperateEuropean flounder and subtropical Gulf toadfish(Opsanus beta) ranging from 18 to 40 mmol Cper kg of fish per hour (8–13). This range is explained by differences in metabolic rate, which aredetermined by body mass and temperature withina species, as well as by interspecific life-styledifferences. In aquatic organisms, mass-specificmetabolism scales inversely with body size, increasing 1.6-fold with every 10-fold decrease inbody mass, and increases exponentially with temperature typically by 1.83-fold for every 10 C rise(14). Thus, smaller fish at higher temperaturesproduce proportionally more carbonate per unitbody mass (fig. S2).To calculate the teleostean contribution tooceanic carbonate budgets requires knowledge ofglobal marine fish biomass. We used two entirelyindependent models to describe the size composition and abundance of marine fish across theglobal oceans, one by using a size-based macroecological approach (15) and the other by usingEcopath software (16). The fish biomass estimates generated for each size-class and the relevant average local sea temperatures were thencombined with individual fish carbonate excretion rates to predict global fish CaCO3 productionTA 2400 μMSedimenTA 2500 μMtatiFish Carbonates 0.04-0.11 Pg CaCO 3 -C year -1(High Magnesium calcites soluble at shallow depths)on2 1 km Depth:Total dissolution 0.41-0.5 Pg CaCO 3 -C year -1Sinking Flux 0.4 Pg CaCO3 -C year -134Dissolution of lowersolubility carbonates(calcites & aragonites)Lysocline( CaCO3 solubility)Se5dimentati onFig. 1. A modified schematic diagram of ocean CaCO3 budget showing the potential contributionof high-magnesium calcite produced by marine teleost fish. The fish images represent teleosts froma wide range of species and habitats, because all teleosts (but not elasmobranchs) are thought toproduce carbonates as part of their osmoregulatory strategy (8–13). All values except the fishproduction rate are previously published estimates for total global production or dissolution in theupper ocean (2, 5–7).16 JANUARY 2009VOL 323SCIENCEwww.sciencemag.orgDownloaded from on January 21, 2009REPORTS

REPORTSFig. 2. Digital x-rayphotographs of liveAEuropean flounderBones of pectoral(Platichthys flesus)fin (overlies theshowing formation ofVertebraeabdominal area)Flounder acclimatedgut carbonates in unfedto fresh waterfish after transfer fromfresh water to seawater.HeadNote the absence ofbones (apart from theoverlying pectoral fin)over the abdominal area(bounded by dashed line)Areacontainingwhere the viscera (invisceracluding intestine) aresituated. (A) Flounderacclimated to fresh water for 1 week to allowBclearance of previouslyproduced carbonatesfrom the intestine. (B)X-ray photo taken 3Flounder 3 hours after acutetransfer to sea waterhours after a freshwaterflounder was transferredto seawater. In seawater,the fish rapidly initiatesdrinking and high ratesof intestinal HCO3 secretion. This results inthe formation of CaCO3precipitates that formx-ray opaque structureswithin the intestine (indicated by solid white arrows). X-ray images were taken with Siemens multix-TOP x-ray equipment and aKonica regus computed radiography system.www.sciencemag.orgSCIENCEVOL 323We predict that production of carbonate precipitates by fish will accelerate as a result of bothincreasing seawater temperatures and CO2 concentrations. First, metabolic rate increases exponentially with temperature in ectothermic fish,thus increasing metabolic CO2 production andintestinal carbonate excretion at the individuallevel (fig. S7). However, for communities, themodel of Jennings et al. (15) suggests that community fish biomass will decrease with temperature (for a given rate of primary production) andthat this will offset the accompanying increase incarbonate production owing to temperature effects on individual metabolism. Second, risingambient levels of dissolved CO2 will cause acorresponding increase in CO2 partial pressuresin the blood of fish (34, 35). In vitro studies showthat increasing blood CO2 concentrations stimulate intestinal cells to produce more HCO3 (36),and thus intestinal excretion of precipitatedcarbonates is predicted to rise with ambient CO2.This contrasts with the commonly cited view thatCaCO3 production rates decrease in calcifyingmarine plankton and corals as ambient CO2increases [(2, 33); but see Supporting OnlineMaterial and (37)]. The biomineralization mechanisms in these organisms are not well understood (37) but are dependent upon the ambientconcentrations of CO32 or HCO3 in seawater,which change with pH as CO2 concentration increases (2). Distinct from this, fish use endogenous CO2 to produce HCO3 ions that rise tovery high concentrations within the microenvironment of the gut lumen (typically 50 to 100 mM)(8–11). Thus, the contribution of fish to marinecarbonate production seems likely to increase inthe future and become an even more importantcomponent of the inorganic carbon cycle.Downloaded from on January 21, 2009lution in the ocean, ignoring a subtle process thatfurther links fish production and distribution tooceanic acid-base chemistry. HCO3 ions secretedby intestinal cells into the intestinal lumen of fishare derived largely from metabolic CO2 reactingwith water within intestinal epithelial cells, underthe catalytic influence of carbonic anhydrase (11).This reaction produces H , which is exported intothe blood and ultimately excreted into the externalseawater via ion-transporting cells in the gills offish (12, 13). Thus, there is an anatomical separation of, and physical distinction between, theacid and base components of this reaction and itsexcretory products; i.e., insoluble CaCO3 excretedvia the gut, and dissolved H ions excreted via thegills. Furthermore, solid CaCO3 will rapidly sinkand only redissolve at depth (raising TA at thispoint), whereas H ions excreted via the gills willremain in the surface ocean (decreasing TA). Regular vertical migrations of many pelagic fish species, often daily and over several hundred meters,may complicate interpretation of the expectedacid-base effects, but the principle is worthnoting.Postindustrial oceanic acidification due toelevated atmospheric CO2 is now well recognized and is predicted to have major impacts oncalcifying organisms (33), raising questions abouthow such future environmental changes mayinfluence piscine global carbonate production.Normalized Total Alkalinity (μmol kg-1)2280023002320234023605001000Depth (m)in only 17% of ocean area (15) (fig. S4). Furthermore, such hot spots are largely found overcontinental shelves and in upwellings where thewater is mostly shallow (100 to 200 m deep).This raises the possibility that fish could be themajor source of carbonate production in the surface ocean in these areas. Also, dissolution of fishcarbonates at such shallow depths may not occurif the carbonates are buried within sediments.Thus, we suggest that the localized high production rates and fate of fish carbonates in some partsof the ocean (and correspondingly low production areas elsewhere) require further investigation. In addition, most carbonates collected insediment traps cannot be visually identified andaccurately assigned to traditional planktonicsources. Intriguingly, some of these collectedcarbonate particles strongly resemble those foundin the intestines of marine fish (fig. S3). TheMg:Ca ratio of fish carbonates (10 to 33 mol %)overlaps with the range for the finest-sized fraction ( 37 mm) of magnesian calcite particlescollected in sediment traps in the Sargasso Sea (9to 12 mol %) (32). At that time, this magnesiancalcite phase of carbonate was assumed to originate from bryozoan skeletons attached to floating Sargassum. It is now tempting to suggest thatfish may be the source of this carbonate phase.So far, we have concentrated on production ofcarbonates, their excretion, and potential disso-150020002500Aragonite Saturation3000Fig. 3. The normalized total alkalinity of seawateras a function of depth for North Atlantic Waters(30 N and 23 E) (18, 20).16 JANUARY 2009361

References and Notes1. F. J. Millero, Chemical Oceanography (CRC Press, BocaRaton, FL, ed. 3, 2006).2. R. A. Feely et al., Science 305, 362 (2004).3. R. Schiebel, Global Biogeochem. Cycles 16, 1065(2002).4. J. D. Milliman, A. W. Droxler, Geol. Rundsch. 85, 496(1996).5. K. Lee, Limnol. Oceanogr. 46, 1287 (2001).6. D. Iglesias-Rodriguez et al., Eos Trans. AGU 83, 365(2002).7. J. D. Milliman et al., Deep Sea Res. Part I Oceanogr. Res.Pap. 46, 1653 (1999).8. P. J. Walsh, P. Blackwelder, K. A. Gill, E. Danulat,T. P. Mommsen, Limnol. Oceanogr. 36, 1227 (1991).9. R. W. Wilson, K. M. Gilmour, R. P. Henry, C. M. Wood,J. Exp. Biol. 199, 2331 (1996).10. R. W. Wilson, J. M. Wilson, M. Grosell, Biochim. Biophys.Acta 1566, 182 (2002).11. M. Grosell, J. Exp. Biol. 209, 2813 (2006).12. R. W. Wilson, M. Grosell, Biochim. Biophys. Acta 1618,163 (2003).13. J. Genz, J. R. Taylor, M. Grosell, J. Exp. Biol. 211, 2327(2008).14. A. Clarke, N. M. Johnstone, J. Anim. Ecol. 68, 893 (1999).15. S. Jennings et al., Proc. R. Soc. London B. Biol. Sci. 275,1375 10.1098/rspb.2008.0192 (2008).16. V. Christensen et al., Models of the World’s Large MarineEcosystems. GEF/LME Global Project PromotingEcosystem-Based Approaches to Fisheries Conservationand Large Marine Ecosystems (IOC Technical Series No.80, UNESCO, 2008).17. W. S. Broecker, in The Fate of Fossil CO2 in the Oceans,N. R. Anderson, A. Malahoff, Eds. (Plenum, New York,1977), p. 207.18. R. A. Feely et al., Global Biogeochem. Cycles 16, 1144(2002).19. C. L. Sabine, R. M. Key, R. A. Feely, D. Greeley, GlobalBiogeochem. Cycles 16, 1067 (2002).20. S. Chung et al., Global Biogeochem. Cycles 17, 1093(2003).21. C. L. Sabine et al., in The Global Carbon Cycle:Integrating Humans, Climate, and the Natural World,C. B. Field, M. R. Raupach, Eds. (Island Press,Washington, DC, 2004), pp. 17–44.22. T. Takahashi, Spec. Publ. Cushman Found. Foraminiferal Res.13, 11 (1975).23. J. K. B. Bishop, J. C. Stepien, P. H. Wiebe, Prog.Oceanogr. 17, 1 (1986).24. R. P. Harris, Mar. Biol. (Berlin) 119, 431 (1994).25. P. Van der Wal, R. S. Kempers, M. J. W. Veldhuis, Mar.Ecol. Prog. Ser. 126, 247 (1995).26. D. W. Pond, R. P. Harris, C. A. Brownlee, Mar. Biol. (Berlin)123, 75 (1995).27. H. Jansen, D. A. Wolf-Gladrow, Mar. Ecol. Prog. Ser. 221,199 (2001).28. R. H. Byrne, J. G. Acker, P. R. Betzer, R. A. Feely, M. H.Cates, Nature 312, 321 (1984).29. R. A. Feely et al., Mar. Chem. 25, 227 (1988).30. J. W. Morse, F. T. Mackenzie, Geochemistry ofSedimentary Carbonates (Elsevier, New York, 1990).31. J. W. Morse, D. K. Gledhill, F. J. Millero, Geochim.Cosmochim. Acta 67, 2819 (2003).32. V. J. Fabry, W. G. Deuser, Deep-Sea Res. 38, 713(1991).33. J. C. Orr et al., Nature 437, 681 (2005).34. H. O. Pörtner, M. Langenbuch, A. Reipschläger,J. Oceanogr. 60, 705 (2004).35. B. A. Seibel, P. J. Walsh, Science 294, 319 (2001).36. M. Grosell et al., Am. J. Physiol. Regul. Integr. Comp.Physiol. 288, R936 (2005).37. V. J. Fabry, Science 320, 1020 (2008).38. R.W.W. acknowledges support from the UK Biotechnologyand Biological Sciences Research Council (awardsBB/D005108/1, BB/F009364/1, and ISIS 1766) and TheRoyal Society (award RSRG 24241). F.J.M. acknowledgesthe Oceanographic Section of the U.S. National ScienceFoundation (NSF). P.J.W. is supported by the NaturalMorphogenesis of Self-AssembledNanocrystalline Materials of BariumCarbonate and SilicaJuan Manuel García-Ruiz,1 Emilio Melero-García,1 Stephen T. Hyde2The precipitation of barium or strontium carbonates in alkaline silica-rich environments leads tocrystalline aggregates that have been named silica/carbonate biomorphs because their morphologyresembles that of primitive organisms. These aggregates are self-assembled materials of purelyinorganic origin, with an amorphous phase of silica intimately intertwined with a carbonatenanocrystalline phase. We propose a mechanism that explains all the morphologies describedfor biomorphs. Chemically coupled coprecipitation of carbonate and silica leads to fibrillationof the growing front and to laminar structures that experience curling at their growing rim. Thesecurls propagate in a surflike way along the rim of the laminae. We show that all observedmorphologies with smoothly varying positive or negative Gaussian curvatures can be explained bythe combined growth of counterpropagating curls and growing laminae.he theoretical morphology of classicalcrystals is well accommodated within conventional crystal growth theory, where thedevelopment of various crystal faces is accountedT1Laboratorio de Estudios Cristalográficos, Instituto Andaluz deCiencias de la Tierra, Consejo Superior de InvestígacìonesCientificas–Universidad de Granada, Avenida del Conocimiento,Parque Tecnológico, Ciencias de la Salud, 18100 Armilla,Spain. 2Department of Applied Mathematics, Research Schoolof Physical Sciences, Australian National University, Canberra,Australian Capital Territory 0200, Australia.362for by the relative crystallographic surface energies at the atomic scale, and the overall symmetryis imposed by the atomic-scale packing. The relation between nonequilibrium crystal shapes andtheir physical and chemical growth conditions isalso part of the general picture (1). In contrast,despite numerous observations over the years (2)that life is able to make precise, smooth, differentiable shapes made of polycrystalline minerals(shells, teeth, bones, etc.), we have a limited understanding of the morphogenetical mechanisms16 JANUARY 2009VOL 323SCIENCESciences and Engineering Research Council (NSERC) ofCanada and the Canada Research Chair program.V.C. acknowledges support from NSERC, the GlobalEnvironment Facility’s UNEP/UNESCO/IOC (IntergovernmentalOceanographic Commission) LME (Large Marine Ecosystem)activities, and from the Sea Around Us Project, initiatedand funded by The Pew Charitable Trusts. S.J. thanksthe European Commission and the UK Departmentof Environment, Food, and Rural Affairs for fundingsupport, and A. Clarke (British Antarctic Survey,Cambridge) for providing a compilation of fish oxygenconsumption data. M.G. and S.J. are supported by theNSF (awards 0416440, 0714024, and 0743903).We also thank referees for their insightful comments;K. Knapp (School of Physics, University of Exeter)for the fish x-ray images (Fig. 2); R. Walton (Universityof Warwick, UK) for the powder x-ray diffractionanalysis (fig. S1); H. Sto

DOI: 10.1126/science.1157972 Science 323, 359 (2009); R. W. Wilson, et al. Cycle Contribution of Fish to the Marine Inorganic Carbon (this information is current as of January 21, 2009 ):