Reports Natural prokaryotic communities collected at 14 stations between 1000 to 4200 m in the Pacific and Atlantic Ocean (fig. S1) were exposed to ambient, 2, 5 and 10-fold concentrations of Jesús M. Arrieta,1,2* Eva Mayol,1 Roberta L. Hansman,3 Gerhard J. natural DOC collected from their original location by solid phase Herndl,3,4 Thorsten Dittmar,5 Carlos M. Duarte1,2,6 extraction (6, 7) and incubated at 1 Department of Global Change Research, Institut Mediterrani d’Estudis Avançats (IMEDEA), Consejo Superior de in situ temperature. A consistent 2 Investigaciones Científicas (CSIC)/Universidad de las Islas Baleares (UIB), 07190 Esporles, Spain. Red Sea Research increase in prokaryotic abunCenter, King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Kingdom of Saudi Arabia. 3 Department of Limnology and Bio-Oceanography, Division Bio-Oceanography, University of Vienna, Althanstr. 14, 1090 dance over time was observed in Vienna, Austria. 4Department of Biological Oceanography, Royal Netherlands Institute for Sea Research (NIOZ), 1790AB response to increasing concenDen Burg, Netherlands. 5Research Group for Marine Geochemistry (ICBM-MPI Bridging Group), Institute for Chemistry trations of DOC in all 14 experi6 and Biology of the Marine Environment (ICBM), Carl von Ossietzky University Oldenburg, Germany. The UWA Oceans ments (Fig. 1 and fig. S2). Institute, The University of Western Australia, Crawley, WA, Australia. Maximum prokaryotic abun*Corresponding author. E-mail: [email protected] dances obtained at ~10-fold DOC Oceanic dissolved organic carbon (DOC) is the second largest reservoir of concentrations were 3.6 to 11.7 organic carbon on Earth. About 72% of the global DOC inventory is stored times higher than those obin deep oceanic layers for years to centuries, supporting the current view served in the corresponding conthat it consists of materials resistant to microbial degradation. An trols (Fig. 1 and fig. S2). alternative hypothesis is that deep-water DOC consists of many different, Unamended controls showed intrinsically labile compounds at concentrations too low to compensate for much lower, sometimes undethe metabolic costs associated to their utilization. Here we present tectable, prokaryotic growth, experimental evidence showing that low concentrations rather than comparable to the values obrecalcitrance preclude consumption of a significant fraction of DOC leading served in deep layers of the to slow microbial growth in the deep ocean. These findings demonstrate an ocean in other studies (8, 9), alternative mechanism for the long-term storage of labile DOC in the deep whereas specific growth rates in ocean, which has been hitherto largely ignored. the higher DOC enrichments showed values up to 0.4 d−1 typiThe accepted paradigm is that recalcitrant DOC is ubiqui- cal of productive surface waters (Fig. 2 and fig. S3). No sigtous in the ocean and makes up the bulk of the DOC pool at nificant differences (t test, P > 0.05) in prokaryotic growth depths >1000 m and at DOC concentrations below 42 μmol were observed between unamended controls and extraction C L−1 (1). However, most of the components of the recalci- controls, confirming that the observed growth was due to trant DOC pool remain unidentified (1, 2) and there is little the materials extracted from seawater and not the result of evidence of structural properties that could make these a contamination with labile organics during the extraction compounds unavailable to microbial degradation. Converse- procedure (fig. S4). The solid phase extraction method used ly, the dilution hypothesis (3, 4) postulates that most organ- to concentrate natural DOM may have introduced some ic substrates in the deep ocean are labile but cannot be used compositional bias toward small and polar compounds by prokaryotes at concentrations below the levels matching while losing a significant part of the DOM pool (extraction the energetic investment required for their uptake and deg- efficiency ~40%), but this does not change the fact that inradation. An early study (5) tested the dilution hypothesis creasing the concentration of the extractable components of by looking for microbial consumption in concentrates of natural DOC resulted in enhanced prokaryotic growth. natural DOC from deep waters but found no significant Chemical alterations of DOC such as the disruption of suchanges in DOC concentrations after a two month incuba- pramolecular arrangements (10) or mild hydrolysis protion. Those results led to the conclusion that deep-water duced during the concentration procedure are an unlikely DOC is composed of recalcitrant molecules and therefore to explanation for the observed response since the treatments the dismissal of the dilution hypothesis. Here we revisit the where DOC concentration was doubled by adding concendilution hypothesis using a simple experimental approach, trated DOC showed little or no enhancement of prokaryotic similar to that used by Barber in 1968 (5) but using method- growth as compared to controls. Hence, we validated the ologies not available at that time. Specifically, we tested the dilution hypothesis tested, showing that dilution limits C hypothesis that no significant increase in prokaryotic utilization in the deep ocean. growth should be detectable when increasing DOC concenSpecific growth rates increased with DOC concentrations trations, as expected if deep oceanic DOC were structurally following a classical Monod model (R2 0.71 to 0.98) at all the refractory. locations studied (Fig. 2, fig. S3, and table S2) further con/ sciencemag.org/content/early/recent / 19 March 2015 / Page 1 / 10.1126/science.1258955
Downloaded from www.sciencemag.org on March 20, 2015
Dilution limits dissolved organic carbon utilization in the deep ocean
firming the hypothesis that prokaryotic growth in deep waters is limited by the concentration of DOC. In 9 out of 14 experiments the initial in situ DOC concentration was low enough to capture the lower part of the curve and thus, to give an estimate of the minimum DOC concentration necessary to support prokaryotic maintenance metabolism (Fig. 2, fig. S3, table S2). According to these estimates, concentrations of natural DOC below 30.7 ± 5.4 μmol C L−1 (average ± SE, n = 9) would not be sufficient to support prokaryotic metabolism, a value not significantly different (t test, P > 0.05) from the lowest concentrations of DOC around 34 μmol C L−1 reported for the deep ocean (11). Prokaryotic growth efficiency (PGE) in the unamended controls was always lower than 3% similar to the values reported for deep-water masses (9). No statistically significant differences among treatments were detected in our PGE estimates due to the accumulation of errors propagating from the original measurements (one-way ANOVA, P > 0.05). However, growth efficiency estimates show a consistent tendency to increase with increasing DOC concentration in all the experiments (fig. S5) suggesting a positive effect of concentration that could be related to a relief from substrate limitation (12). An increasing growth efficiency with increasing DOC concentrations cannot be explained if the bulk of the DOC components were structurally recalcitrant. The differences between our results and those reported by Barber in 1968 are probably due to a combination of methodological improvements. We used only dissolved materials while Barber used a concentrate of everything with nominal size > 500 Da including particulate matter. Thus, true DOC concentrations in Barber’s experiments were probably not as high as intended. Also, a large fraction of marine DOC consists of molecules < 500 Da (13, 14) (fig. S7), thus, the composition was biased toward higher molecular weight compounds in Barber’s experiments, meaning that the molar enrichment in his 5-fold C concentration treatment was probably much lower than in our experiments. In the four experiments carried out in the North Atlantic, incubations were kept on the ship for an additional month after the cruise until the ship returned to the harbor (fig. S2, K to N). Prokaryotic abundance remained essentially constant during the additional month in all but one of the experiments (M) where a second phase of intense growth was observed in the most concentrated treatments (fig. S2). While it is unclear why this happened, we can rule out contamination since growth occurred in all replicates and a pronounced decrease in DOC was found in these samples (fig. S6). The appearance of large cells in the flow cytograms indicating significant growth of heterotrophic protists (15) in the late stages of the more concentrated cultures and the fact that the prokaryotic carbon demand inferred from the increase in abundance and growth efficiency was much lower than the measured decrease in DOC concentration indicate that labile DOC was still being consumed at the end of the experiments even when no increase in prokaryotic
abundance was detectable due to enhanced prokaryotic mortality. A corollary to the dilution hypothesis is that bulk DOC is composed of a large diversity of individual molecules. Indeed, molecular characterization by Fourier-transform ion cyclotron resonance mass spectrometry (FT-ICR-MS) (13, 14) (fig. S7) shows that marine DOC is composed of a very large variety of small molecules (200-700 Da) present in very low concentrations and adding up to a large pool of apparently recalcitrant DOC. We characterized the changes caused by microbial activity on the molecular composition of deep oceanic DOM in two additional experiments (O and P) in unamended seawater and in samples where the concentration of in situ DOC was raised to approximately 5 times the original concentration. Molecular characterization by FTICR-MS showed a highly reproducible pattern of microbial utilization among replicate independent incubations (fig. S8) affecting a large number of different molecules (fig. S9). Significant utilization (relative signal in replicate incubations significantly lower than in the corresponding initial samples; t test P < 0.05) was found for 2,095 and 1,753 different compounds in the controls and for 2,846 and 936 different compounds in the 5x concentrated samples for experiments O and P, respectively. Consistent differences could be observed between the set of molecules utilized in concentrated versus unamended samples resulting in a total of 3,950 different molecules consumed in either the concentrated samples or the controls for experiment O and 2,140 in experiment P. Moreover, the sum of the normalized signal of all the peaks in the FT-ICR-MS fingerprint where significant consumption was detected in either concentrates or controls was >70% in experiment O and >40% in experiment P, indicating that a major fraction of the original DOC consisted of labile compounds. The dilution hypothesis provides an alternative framework explaining observations of the apparent recalcitrance of DOC and lends a physiological meaning to the operationally defined “semi-labile” and “semi-refractory” fractions (16, 17). We hypothesize that, under the dilution hypothesis, very heterogeneous mixtures of labile compounds appear semirefractory while increasingly less diverse DOM assemblages containing larger concentrations of some substrates will present higher microbial growth and DOC turnover rates resulting in increasing degrees of apparent lability. The microbial generation of apparently recalcitrant material (18) from labile substrates in a process recently dubbed the “microbial carbon pump” (19) can also be explained by the dilution hypothesis. Microbial utilization of abundant, labile compounds results in hundreds of different metabolites (20), which are subsequently consumed down to the lowest utilizable concentration. This mechanism explains observations of relatively concentrated, labile materials being transformed into apparently recalcitrant matter by microbial consumption (18) but does not necessarily imply the formation of structurally recalcitrant molecules. Indeed, “recal-
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 2 / 10.1126/science.1258955
citrant” DOC is not defined structurally, but operationally, as the DOC pool remaining after long experimental incubations or as the fraction transported in an apparently conservative manner with the ocean circulation (1). Thus, the dilution hypothesis severely limits the feasibility of geoengineering efforts to enhance carbon storage in the deep ocean (21) using the microbial carbon pump. FT-ICR-MS characterization of DOC from different oceans (13, 14, 22, 23) and also from this study (fig. S5) shows no indication of prevalent, intrinsically recalcitrant compounds accumulating in significant amounts. Conversely, FT-ICR-MS data show that oceanic DOC is a complex mixture of minute quantities of thousands of organic molecules in good agreement with the dilution hypothesis. Mean radiocarbon ages of deep oceanic DOC in the range of 4,000-6,000 years have been considered as evidence for its recalcitrant nature (24, 25). However, these are average ages of a pool containing a mixture of very old molecules >12 ka but also featuring a large proportion of contemporary materials (26). Moreover, elevated radiocarbon ages only demonstrate that these old molecules are not being newly produced at any significant rate, since that would lower their isotopic age, but does not necessarily imply that they are structurally recalcitrant. Furthermore, it is unlikely that natural organic molecules can accumulate in the ocean in significant concentrations and remain recalcitrant or be preserved for millennia when degradation pathways for novel synthetic pollutants evolve soon after these compounds are released in nature (27). While there might be a truly recalcitrant component in deep oceanic DOC, our results clearly show that the concentration of individual labile molecules is a major factor limiting the utilization of a significant fraction of deep oceanic DOC. These results provide, therefore, a robust and parsimonious explanation for the long-term preservation of labile DOC into one of the largest reservoirs of organic carbon on Earth, opening a new avenue in our understanding of the global carbon cycle. REFERENCES AND NOTES 1. D. A. Hansell, Recalcitrant dissolved organic carbon fractions. Annu. Rev. Mar. Sci. 5, 421–445 (2013). Medline doi:10.1146/annurev-marine-120710-100757 2. E. B. Kujawinski, The impact of microbial metabolism on marine dissolved organic matter. Annu. Rev. Mar. Sci. 3, 567–599 (2011). Medline doi:10.1146/annurevmarine-120308-081003 3. H. W. Jannasch, Growth of marine bacteria at limiting concentrations of organic carbon in seawater. Limnol. Oceanogr. 12, 264–271 (1967). doi:10.4319/lo.1967.12.2.0264 4. H. W. Jannasch, The microbial turnover of carbon in the deep-sea environment. Global Planet. Change 9, 289–295 (1994). doi:10.1016/0921-8181(94)90022-1 5. R. T. Barber, Dissolved organic carbon from deep waters resists microbial oxidation. Nature 220, 274–275 (1968). Medline doi:10.1038/220274a0 6. Materials and methods are available as supplementary materials on Science Online. 7. T. Dittmar, B. Koch, N. Hertkorn, G. Kattner, A simple and efficient method for the solid-phase extraction of dissolved organic matter (SPE-DOM) from seawater. Limnol. Oceanogr. Methods 6, 230–235 (2008). doi:10.4319/lom.2008.6.230 8. D. L. Kirchman, X. A. G. Morán, H. Ducklow, Microbial growth in the polar oceans role of temperature and potential impact of climate change. Nat. Rev. Microbiol.
7, 451–459 (2009). Medline 9. T. Reinthaler, H. van Aken, C. Veth, J. Arístegui, C. Robinson, P. J. B. Williams, P. Lebaron, G. J. Herndl, Prokaryotic respiration and production in the meso- and bathypelagic realm of the Eastern and Western North Atlantic basin. Limnol. Oceanogr. 51, 1262–1273 (2006). doi:10.4319/lo.2006.51.3.1262 10. A. Nebbioso, A. Piccolo, Molecular characterization of dissolved organic matter (DOM): A critical review. Anal. Bioanal. Chem. 405, 109–124 (2013). Medline doi:10.1007/s00216-012-6363-2 11. D. A. Hansell, C. A. Carlson, D. J. Repeta, R. Schlitzer, Dissolved organic matter in the ocean: A controversy stimulates new insights. Oceanography (Wash. D.C.) 22, 202–211 (2009). doi:10.5670/oceanog.2009.109 12. A. Konopka, Microbial physiological state at low growth rate in natural and engineered ecosystems. Curr. Opin. Microbiol. 3, 244–247 (2000). Medline doi:10.1016/S1369-5274(00)00083-7 13. G. Kattner, M. Simon, B. Koch, in Microbial Carbon Pump in the Ocean, N. Jiao, F. Azam, S Sanders, Eds. (Science/AAAS, Washington, DC 2011), pp. 60–61. 14. T. Dittmar, J. Paeng, A heat-induced molecular signature in marine dissolved organic matter. Nat. Geosci. 2, 175–179 (2009). doi:10.1038/ngeo440 15. M. V. Zubkov, P. H. Burkill, J. N. Topping, Flow cytometric enumeration of DNAstained oceanic planktonic protists. J. Plankton Res. 29, 79–86 (2007). doi:10.1093/plankt/fbl059 16. C. A. Carlson, H. W. Ducklow, A. F. Michaels, Annual flux of dissolved organic carbon from the euphotic zone in the northwestern Sargasso Sea. Nature 371, 405–408 (1994). doi:10.1038/371405a0 17. J. H. Sharp, K. Yoshiyama, A. E. Parker, M. C. Schwartz, S. E. Curless, A. Y. Beauregard, J. E. Ossolinski, A. R. Davis, A Biogeochemical View of Estuarine Eutrophication: Seasonal and Spatial Trends and Correlations in the Delaware Estuary. Estuaries Coasts 32, 1023–1043 (2009). doi:10.1007/s12237-0099210-8 18. H. Ogawa, Y. Amagai, I. Koike, K. Kaiser, R. Benner, Production of refractory dissolved organic matter by bacteria. Science 292, 917–920 (2001). Medline doi:10.1126/science.1057627 19. N. Jiao, G. J. Herndl, D. A. Hansell, R. Benner, G. Kattner, S. W. Wilhelm, D. L. Kirchman, M. G. Weinbauer, T. Luo, F. Chen, F. Azam, Microbial production of recalcitrant dissolved organic matter: Long-term carbon storage in the global ocean. Nat. Rev. Microbiol. 8, 593–599 (2010). Medline doi:10.1038/nrmicro2386 20. R. P. Maharjan, S. Seeto, T. Ferenci, Divergence and redundancy of transport and metabolic rate-yield strategies in a single Escherichia coli population. J. Bacteriol. 189, 2350–2358 (2007). Medline doi:10.1128/JB.01414-06 21. R. Stone, The invisible hand behind a vast carbon reservoir. Science 328, 1476– 1477 (2010). Medline doi:10.1126/science.328.5985.1476 22. R. Flerus, O. J. Lechtenfeld, B. P. Koch, S. L. McCallister, P. Schmitt-Kopplin, R. Benner, K. Kaiser, G. Kattner, A molecular perspective on the ageing of marine dissolved organic matter. Biogeosciences 9, 1935–1955 (2012). doi:10.5194/bg9-1935-2012 23. O. J. Lechtenfeld, G. Kattner, R. Flerus, S. L. McCallister, P. Schmitt-Kopplin, B. P. Koch, Molecular transformation and degradation of refractory dissolved organic matter in the Atlantic and Southern Ocean. Geochim. Cosmochim. Acta 126, 321–337 (2014). doi:10.1016/j.gca.2013.11.009 24. P. M. Williams, E. R. M. Druffel, Radiocarbon in dissolved organic matter in the central North Pacific Ocean. Nature 330, 246–248 (1987). doi:10.1038/330246a0 25. J. E. Bauer, A. Dennis, Hansell, Craig A. Carlson, in Biogeochemistry of Marine Dissolved Organic Matter (Academic Press, San Diego, 2002), pp. 405–453. 26. C. L. Follett, D. J. Repeta, D. H. Rothman, L. Xu, C. Santinelli, Hidden cycle of dissolved organic carbon in the deep ocean. Proc. Natl. Acad. Sci. U.S.A. 111, 16706–16711 (2014). Medline doi:10.1073/pnas.1407445111 27. S. D. Copley, Evolution of a metabolic pathway for degradation of a toxic xenobiotic: The patchwork approach. Trends Biochem. Sci. 25, 261–265 (2000). Medline doi:10.1016/S0968-0004(00)01562-0 28. R. Benner, M. Strom, A critical evaluation of the analytical blank associated with DOC measurements by high-temperature catalytic oxidation. Mar. Chem. 41, 153–160 (1993). doi:10.1016/0304-4203(93)90113-3 29. D. Marie, F. Partensky, S. Jacquet, D. Vaulot, Enumeration and Cell Cycle Analysis of Natural Populations of Marine Picoplankton by Flow Cytometry Using the Nucleic Acid Stain SYBR Green I. Appl. Environ. Microbiol. 63, 186–193 (1997). Medline
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 3 / 10.1126/science.1258955
30. R. Fukuda, H. Ogawa, T. Nagata, I. Koike, Direct determination of carbon and nitrogen contents of natural bacterial assemblages in marine environments. Appl. Environ. Microbiol. 64, 3352–3358 (1998). Medline 31. J. Baranyi, C. Pin, Estimating bacterial growth parameters by means of detection times. Appl. Environ. Microbiol. 65, 732–736 (1999). Medline 32. M. E. Tros, T. N. Bosma, G. Schraa, A. J. Zehnder, Measurement of minimum substrate concentration (Smin) in a recycling fermentor and its prediction from the kinetic parameters of Pseudomonas strain B13 from batch and chemostat cultures. Appl. Environ. Microbiol. 62, 3655–3661 (1996). Medline 33. R. Suzuki, H. Shimodaira, Pvclust: An R package for assessing the uncertainty in hierarchical clustering. Bioinformatics 22, 1540–1542 (2006). Medline doi:10.1093/bioinformatics/btl117 34. J. Oksanen et al., vegan: Community Ecology Package (2013; http://CRAN.Rproject.org/package=vegan). 35. R Development Core Team, R: A language and environment for statistical computing.R Foundation for Statistical Computing. Vienna, Austria http://www.R-project.org (2014). ACKNOWLEDGMENTS This is a contribution to the MALASPINA Expedition 2010 project, funded by the CONSOLIDER-Ingenio 2010 program of the from the Spanish Ministry of Economy and Competitiveness (Ref. CSD2008-00077). J.M.A. was supported by a “Ramón y Cajal” research fellowship from the Spanish Ministry of Economy and Competitiveness. E.M. was supported by a fellowship from the JAE program of CSIC. G.J.H. and R.H. were supported by the Austrian Science Fund (FWF) projects: I486-B09 and P23234-B11 and by the European Research Council under the European Community’s Seventh Framework Programme (FP7/20072013) / ERC grant agreement No. 268595 (MEDEA project). We thank A. Dorsett for assistance with DOC analyses, participants in the Malaspina Expedition and the crews of the BIO Hespérides, and RV Pelagia and the personnel of the Marine Technology Unit of CSIC (UTM) for their invaluable support. Original datasets are available online at http://digital.csic.es/handle/10261/111563. Author contributions: JMA designed the experimental setup, carried out part of the experiments, measured prokaryotic abundance, analyzed the data and wrote the manuscript. EM carried out part of the experiments and data analysis. CMD designed the Malaspina 2010 Expedition and was responsible for DOC analyses and, together with GJH contributed to the design of the experiments and discussion of results. RH and TD analyzed the FT-ICR-MS samples. All authors discussed the results and contributed to the manuscript. SUPPLEMENTARY MATERIALS www.sciencemag.org/content/science.1258955/DC1 Materials and Methods Figs. S1 to S9 Tables S1 and S2 References (28–35) 18 July 2014; accepted 4 March 2015 Published online 19 March 2015 10.1126/science.1258955
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 4 / 10.1126/science.1258955
Fig. 1. Prokaryotic abundance in experimental treatments containing approximately 2, 5, and 10 times the in situ DOC concentration vs controls containing unnamended seawater. Error bars represent the standard error of the mean of triplicate cultures. Only the 5 experiments carried out in the North Pacific are represented. The results of the 14 experiments are shown in supplemental figure S2.
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 5 / 10.1126/science.1258955
Fig. 2. Specific growth rates at different concentrations of DOC. Horizontal error bars represent the standard deviation of the mean initial DOC concentration measured in the triplicate cultures and vertical error bars represent the standard deviation of the mean of the specific growth rates estimated for each of the triplicate cultures. Only the 5 experiments carried out in the North Pacific are shown. The results of the 14 experiments are available as supplemental figure S3.
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 6 / 10.1126/science.1258955
Downloaded from www.sciencemag.org on March 20, 2015
Dilution limits dissolved organic carbon utilization in the deep ocean
firming the hypothesis that prokaryotic growth in deep waters is limited by the concentration of DOC. In 9 out of 14 experiments the initial in situ DOC concentration was low enough to capture the lower part of the curve and thus, to give an estimate of the minimum DOC concentration necessary to support prokaryotic maintenance metabolism (Fig. 2, fig. S3, table S2). According to these estimates, concentrations of natural DOC below 30.7 ± 5.4 μmol C L−1 (average ± SE, n = 9) would not be sufficient to support prokaryotic metabolism, a value not significantly different (t test, P > 0.05) from the lowest concentrations of DOC around 34 μmol C L−1 reported for the deep ocean (11). Prokaryotic growth efficiency (PGE) in the unamended controls was always lower than 3% similar to the values reported for deep-water masses (9). No statistically significant differences among treatments were detected in our PGE estimates due to the accumulation of errors propagating from the original measurements (one-way ANOVA, P > 0.05). However, growth efficiency estimates show a consistent tendency to increase with increasing DOC concentration in all the experiments (fig. S5) suggesting a positive effect of concentration that could be related to a relief from substrate limitation (12). An increasing growth efficiency with increasing DOC concentrations cannot be explained if the bulk of the DOC components were structurally recalcitrant. The differences between our results and those reported by Barber in 1968 are probably due to a combination of methodological improvements. We used only dissolved materials while Barber used a concentrate of everything with nominal size > 500 Da including particulate matter. Thus, true DOC concentrations in Barber’s experiments were probably not as high as intended. Also, a large fraction of marine DOC consists of molecules < 500 Da (13, 14) (fig. S7), thus, the composition was biased toward higher molecular weight compounds in Barber’s experiments, meaning that the molar enrichment in his 5-fold C concentration treatment was probably much lower than in our experiments. In the four experiments carried out in the North Atlantic, incubations were kept on the ship for an additional month after the cruise until the ship returned to the harbor (fig. S2, K to N). Prokaryotic abundance remained essentially constant during the additional month in all but one of the experiments (M) where a second phase of intense growth was observed in the most concentrated treatments (fig. S2). While it is unclear why this happened, we can rule out contamination since growth occurred in all replicates and a pronounced decrease in DOC was found in these samples (fig. S6). The appearance of large cells in the flow cytograms indicating significant growth of heterotrophic protists (15) in the late stages of the more concentrated cultures and the fact that the prokaryotic carbon demand inferred from the increase in abundance and growth efficiency was much lower than the measured decrease in DOC concentration indicate that labile DOC was still being consumed at the end of the experiments even when no increase in prokaryotic
abundance was detectable due to enhanced prokaryotic mortality. A corollary to the dilution hypothesis is that bulk DOC is composed of a large diversity of individual molecules. Indeed, molecular characterization by Fourier-transform ion cyclotron resonance mass spectrometry (FT-ICR-MS) (13, 14) (fig. S7) shows that marine DOC is composed of a very large variety of small molecules (200-700 Da) present in very low concentrations and adding up to a large pool of apparently recalcitrant DOC. We characterized the changes caused by microbial activity on the molecular composition of deep oceanic DOM in two additional experiments (O and P) in unamended seawater and in samples where the concentration of in situ DOC was raised to approximately 5 times the original concentration. Molecular characterization by FTICR-MS showed a highly reproducible pattern of microbial utilization among replicate independent incubations (fig. S8) affecting a large number of different molecules (fig. S9). Significant utilization (relative signal in replicate incubations significantly lower than in the corresponding initial samples; t test P < 0.05) was found for 2,095 and 1,753 different compounds in the controls and for 2,846 and 936 different compounds in the 5x concentrated samples for experiments O and P, respectively. Consistent differences could be observed between the set of molecules utilized in concentrated versus unamended samples resulting in a total of 3,950 different molecules consumed in either the concentrated samples or the controls for experiment O and 2,140 in experiment P. Moreover, the sum of the normalized signal of all the peaks in the FT-ICR-MS fingerprint where significant consumption was detected in either concentrates or controls was >70% in experiment O and >40% in experiment P, indicating that a major fraction of the original DOC consisted of labile compounds. The dilution hypothesis provides an alternative framework explaining observations of the apparent recalcitrance of DOC and lends a physiological meaning to the operationally defined “semi-labile” and “semi-refractory” fractions (16, 17). We hypothesize that, under the dilution hypothesis, very heterogeneous mixtures of labile compounds appear semirefractory while increasingly less diverse DOM assemblages containing larger concentrations of some substrates will present higher microbial growth and DOC turnover rates resulting in increasing degrees of apparent lability. The microbial generation of apparently recalcitrant material (18) from labile substrates in a process recently dubbed the “microbial carbon pump” (19) can also be explained by the dilution hypothesis. Microbial utilization of abundant, labile compounds results in hundreds of different metabolites (20), which are subsequently consumed down to the lowest utilizable concentration. This mechanism explains observations of relatively concentrated, labile materials being transformed into apparently recalcitrant matter by microbial consumption (18) but does not necessarily imply the formation of structurally recalcitrant molecules. Indeed, “recal-
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 2 / 10.1126/science.1258955
citrant” DOC is not defined structurally, but operationally, as the DOC pool remaining after long experimental incubations or as the fraction transported in an apparently conservative manner with the ocean circulation (1). Thus, the dilution hypothesis severely limits the feasibility of geoengineering efforts to enhance carbon storage in the deep ocean (21) using the microbial carbon pump. FT-ICR-MS characterization of DOC from different oceans (13, 14, 22, 23) and also from this study (fig. S5) shows no indication of prevalent, intrinsically recalcitrant compounds accumulating in significant amounts. Conversely, FT-ICR-MS data show that oceanic DOC is a complex mixture of minute quantities of thousands of organic molecules in good agreement with the dilution hypothesis. Mean radiocarbon ages of deep oceanic DOC in the range of 4,000-6,000 years have been considered as evidence for its recalcitrant nature (24, 25). However, these are average ages of a pool containing a mixture of very old molecules >12 ka but also featuring a large proportion of contemporary materials (26). Moreover, elevated radiocarbon ages only demonstrate that these old molecules are not being newly produced at any significant rate, since that would lower their isotopic age, but does not necessarily imply that they are structurally recalcitrant. Furthermore, it is unlikely that natural organic molecules can accumulate in the ocean in significant concentrations and remain recalcitrant or be preserved for millennia when degradation pathways for novel synthetic pollutants evolve soon after these compounds are released in nature (27). While there might be a truly recalcitrant component in deep oceanic DOC, our results clearly show that the concentration of individual labile molecules is a major factor limiting the utilization of a significant fraction of deep oceanic DOC. These results provide, therefore, a robust and parsimonious explanation for the long-term preservation of labile DOC into one of the largest reservoirs of organic carbon on Earth, opening a new avenue in our understanding of the global carbon cycle. REFERENCES AND NOTES 1. D. A. Hansell, Recalcitrant dissolved organic carbon fractions. Annu. Rev. Mar. Sci. 5, 421–445 (2013). Medline doi:10.1146/annurev-marine-120710-100757 2. E. B. Kujawinski, The impact of microbial metabolism on marine dissolved organic matter. Annu. Rev. Mar. Sci. 3, 567–599 (2011). Medline doi:10.1146/annurevmarine-120308-081003 3. H. W. Jannasch, Growth of marine bacteria at limiting concentrations of organic carbon in seawater. Limnol. Oceanogr. 12, 264–271 (1967). doi:10.4319/lo.1967.12.2.0264 4. H. W. Jannasch, The microbial turnover of carbon in the deep-sea environment. Global Planet. Change 9, 289–295 (1994). doi:10.1016/0921-8181(94)90022-1 5. R. T. Barber, Dissolved organic carbon from deep waters resists microbial oxidation. Nature 220, 274–275 (1968). Medline doi:10.1038/220274a0 6. Materials and methods are available as supplementary materials on Science Online. 7. T. Dittmar, B. Koch, N. Hertkorn, G. Kattner, A simple and efficient method for the solid-phase extraction of dissolved organic matter (SPE-DOM) from seawater. Limnol. Oceanogr. Methods 6, 230–235 (2008). doi:10.4319/lom.2008.6.230 8. D. L. Kirchman, X. A. G. Morán, H. Ducklow, Microbial growth in the polar oceans role of temperature and potential impact of climate change. Nat. Rev. Microbiol.
7, 451–459 (2009). Medline 9. T. Reinthaler, H. van Aken, C. Veth, J. Arístegui, C. Robinson, P. J. B. Williams, P. Lebaron, G. J. Herndl, Prokaryotic respiration and production in the meso- and bathypelagic realm of the Eastern and Western North Atlantic basin. Limnol. Oceanogr. 51, 1262–1273 (2006). doi:10.4319/lo.2006.51.3.1262 10. A. Nebbioso, A. Piccolo, Molecular characterization of dissolved organic matter (DOM): A critical review. Anal. Bioanal. Chem. 405, 109–124 (2013). Medline doi:10.1007/s00216-012-6363-2 11. D. A. Hansell, C. A. Carlson, D. J. Repeta, R. Schlitzer, Dissolved organic matter in the ocean: A controversy stimulates new insights. Oceanography (Wash. D.C.) 22, 202–211 (2009). doi:10.5670/oceanog.2009.109 12. A. Konopka, Microbial physiological state at low growth rate in natural and engineered ecosystems. Curr. Opin. Microbiol. 3, 244–247 (2000). Medline doi:10.1016/S1369-5274(00)00083-7 13. G. Kattner, M. Simon, B. Koch, in Microbial Carbon Pump in the Ocean, N. Jiao, F. Azam, S Sanders, Eds. (Science/AAAS, Washington, DC 2011), pp. 60–61. 14. T. Dittmar, J. Paeng, A heat-induced molecular signature in marine dissolved organic matter. Nat. Geosci. 2, 175–179 (2009). doi:10.1038/ngeo440 15. M. V. Zubkov, P. H. Burkill, J. N. Topping, Flow cytometric enumeration of DNAstained oceanic planktonic protists. J. Plankton Res. 29, 79–86 (2007). doi:10.1093/plankt/fbl059 16. C. A. Carlson, H. W. Ducklow, A. F. Michaels, Annual flux of dissolved organic carbon from the euphotic zone in the northwestern Sargasso Sea. Nature 371, 405–408 (1994). doi:10.1038/371405a0 17. J. H. Sharp, K. Yoshiyama, A. E. Parker, M. C. Schwartz, S. E. Curless, A. Y. Beauregard, J. E. Ossolinski, A. R. Davis, A Biogeochemical View of Estuarine Eutrophication: Seasonal and Spatial Trends and Correlations in the Delaware Estuary. Estuaries Coasts 32, 1023–1043 (2009). doi:10.1007/s12237-0099210-8 18. H. Ogawa, Y. Amagai, I. Koike, K. Kaiser, R. Benner, Production of refractory dissolved organic matter by bacteria. Science 292, 917–920 (2001). Medline doi:10.1126/science.1057627 19. N. Jiao, G. J. Herndl, D. A. Hansell, R. Benner, G. Kattner, S. W. Wilhelm, D. L. Kirchman, M. G. Weinbauer, T. Luo, F. Chen, F. Azam, Microbial production of recalcitrant dissolved organic matter: Long-term carbon storage in the global ocean. Nat. Rev. Microbiol. 8, 593–599 (2010). Medline doi:10.1038/nrmicro2386 20. R. P. Maharjan, S. Seeto, T. Ferenci, Divergence and redundancy of transport and metabolic rate-yield strategies in a single Escherichia coli population. J. Bacteriol. 189, 2350–2358 (2007). Medline doi:10.1128/JB.01414-06 21. R. Stone, The invisible hand behind a vast carbon reservoir. Science 328, 1476– 1477 (2010). Medline doi:10.1126/science.328.5985.1476 22. R. Flerus, O. J. Lechtenfeld, B. P. Koch, S. L. McCallister, P. Schmitt-Kopplin, R. Benner, K. Kaiser, G. Kattner, A molecular perspective on the ageing of marine dissolved organic matter. Biogeosciences 9, 1935–1955 (2012). doi:10.5194/bg9-1935-2012 23. O. J. Lechtenfeld, G. Kattner, R. Flerus, S. L. McCallister, P. Schmitt-Kopplin, B. P. Koch, Molecular transformation and degradation of refractory dissolved organic matter in the Atlantic and Southern Ocean. Geochim. Cosmochim. Acta 126, 321–337 (2014). doi:10.1016/j.gca.2013.11.009 24. P. M. Williams, E. R. M. Druffel, Radiocarbon in dissolved organic matter in the central North Pacific Ocean. Nature 330, 246–248 (1987). doi:10.1038/330246a0 25. J. E. Bauer, A. Dennis, Hansell, Craig A. Carlson, in Biogeochemistry of Marine Dissolved Organic Matter (Academic Press, San Diego, 2002), pp. 405–453. 26. C. L. Follett, D. J. Repeta, D. H. Rothman, L. Xu, C. Santinelli, Hidden cycle of dissolved organic carbon in the deep ocean. Proc. Natl. Acad. Sci. U.S.A. 111, 16706–16711 (2014). Medline doi:10.1073/pnas.1407445111 27. S. D. Copley, Evolution of a metabolic pathway for degradation of a toxic xenobiotic: The patchwork approach. Trends Biochem. Sci. 25, 261–265 (2000). Medline doi:10.1016/S0968-0004(00)01562-0 28. R. Benner, M. Strom, A critical evaluation of the analytical blank associated with DOC measurements by high-temperature catalytic oxidation. Mar. Chem. 41, 153–160 (1993). doi:10.1016/0304-4203(93)90113-3 29. D. Marie, F. Partensky, S. Jacquet, D. Vaulot, Enumeration and Cell Cycle Analysis of Natural Populations of Marine Picoplankton by Flow Cytometry Using the Nucleic Acid Stain SYBR Green I. Appl. Environ. Microbiol. 63, 186–193 (1997). Medline
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 3 / 10.1126/science.1258955
30. R. Fukuda, H. Ogawa, T. Nagata, I. Koike, Direct determination of carbon and nitrogen contents of natural bacterial assemblages in marine environments. Appl. Environ. Microbiol. 64, 3352–3358 (1998). Medline 31. J. Baranyi, C. Pin, Estimating bacterial growth parameters by means of detection times. Appl. Environ. Microbiol. 65, 732–736 (1999). Medline 32. M. E. Tros, T. N. Bosma, G. Schraa, A. J. Zehnder, Measurement of minimum substrate concentration (Smin) in a recycling fermentor and its prediction from the kinetic parameters of Pseudomonas strain B13 from batch and chemostat cultures. Appl. Environ. Microbiol. 62, 3655–3661 (1996). Medline 33. R. Suzuki, H. Shimodaira, Pvclust: An R package for assessing the uncertainty in hierarchical clustering. Bioinformatics 22, 1540–1542 (2006). Medline doi:10.1093/bioinformatics/btl117 34. J. Oksanen et al., vegan: Community Ecology Package (2013; http://CRAN.Rproject.org/package=vegan). 35. R Development Core Team, R: A language and environment for statistical computing.R Foundation for Statistical Computing. Vienna, Austria http://www.R-project.org (2014). ACKNOWLEDGMENTS This is a contribution to the MALASPINA Expedition 2010 project, funded by the CONSOLIDER-Ingenio 2010 program of the from the Spanish Ministry of Economy and Competitiveness (Ref. CSD2008-00077). J.M.A. was supported by a “Ramón y Cajal” research fellowship from the Spanish Ministry of Economy and Competitiveness. E.M. was supported by a fellowship from the JAE program of CSIC. G.J.H. and R.H. were supported by the Austrian Science Fund (FWF) projects: I486-B09 and P23234-B11 and by the European Research Council under the European Community’s Seventh Framework Programme (FP7/20072013) / ERC grant agreement No. 268595 (MEDEA project). We thank A. Dorsett for assistance with DOC analyses, participants in the Malaspina Expedition and the crews of the BIO Hespérides, and RV Pelagia and the personnel of the Marine Technology Unit of CSIC (UTM) for their invaluable support. Original datasets are available online at http://digital.csic.es/handle/10261/111563. Author contributions: JMA designed the experimental setup, carried out part of the experiments, measured prokaryotic abundance, analyzed the data and wrote the manuscript. EM carried out part of the experiments and data analysis. CMD designed the Malaspina 2010 Expedition and was responsible for DOC analyses and, together with GJH contributed to the design of the experiments and discussion of results. RH and TD analyzed the FT-ICR-MS samples. All authors discussed the results and contributed to the manuscript. SUPPLEMENTARY MATERIALS www.sciencemag.org/content/science.1258955/DC1 Materials and Methods Figs. S1 to S9 Tables S1 and S2 References (28–35) 18 July 2014; accepted 4 March 2015 Published online 19 March 2015 10.1126/science.1258955
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 4 / 10.1126/science.1258955
Fig. 1. Prokaryotic abundance in experimental treatments containing approximately 2, 5, and 10 times the in situ DOC concentration vs controls containing unnamended seawater. Error bars represent the standard error of the mean of triplicate cultures. Only the 5 experiments carried out in the North Pacific are represented. The results of the 14 experiments are shown in supplemental figure S2.
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 5 / 10.1126/science.1258955
Fig. 2. Specific growth rates at different concentrations of DOC. Horizontal error bars represent the standard deviation of the mean initial DOC concentration measured in the triplicate cultures and vertical error bars represent the standard deviation of the mean of the specific growth rates estimated for each of the triplicate cultures. Only the 5 experiments carried out in the North Pacific are shown. The results of the 14 experiments are available as supplemental figure S3.
/ sciencemag.org/content/early/recent / 19 March 2015 / Page 6 / 10.1126/science.1258955
