Production of atmospheric sulfur by oceanic plankton: biogeochemical, ecological and evolutionary links

Opinion

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

Production of atmospheric sulfur by oceanic plankton: biogeochemical, ecological and evolutionary links Rafel Simó Biological production of the volatile compound dimethylsulfide in the ocean is the main natural source of tropospheric sulfur on a global scale, with important consequences for the radiative balance of the Earth. In the late 1980s, a Gaian feedback link between marine phytoplankton and climate through the release of atmospheric sulfur was hypothesized. However, the idea of microalgae producing a substance that could regulate climate has been criticized on the basis of its evolutionary feasibility. Recent advances have shown that volatile sulfur is a result of ecological interactions and transformation processes through planktonic food webs. It is, therefore, not only phytoplankton biomass, taxonomy or activity, but also food-web structure and dynamics that drive the oceanic production of atmospheric sulfur.Accordingly, the viewpoint on the ecological and evolutionary basis of this amazing marine biota–atmosphere link is changing.

Rafel Simó Dept of Marine Biology and Oceanography, Institut de Ciències del Mar (CSIC), Pg Joan de Borbó s/n, 08039 Barcelona, Catalonia, Spain. e-mail: [email protected]

Dimethylsulfide (DMS, see Glossary) is the most abundant form of volatile sulfur in the ocean. Since Lovelock postulated in 1972 that DMS could account for the ‘missing’ global flux of gaseous sulfur from the oceans to the atmosphere1, some 150 oceanographic cruises have compiled data on its ubiquity and supersaturation in surface seawater2. Today, DMS is well recognized as the main natural source of reduced sulfur to the global troposphere3. The most recent estimates4 of its emission flux range from 15 × 1012 to 33 × 1012 g S y−1, enough to make a major contribution to the atmospheric sulfur burden (Table 1) and, therefore, to the chemistry and radiative behavior of the atmosphere (Box 1). Introduced in the late 1980s, the idea of marine microbiota being principal factors in atmospheric chemistry and climate regulation encouraged the study of the links between plankton and atmospheric sulfur, study that has since yielded some 1000 papers in the scientific literature. Over the years, the interest in the topic has not only increased but has also diversified. DMS is produced by the enzymatic breakdown of dimethylsulfoniopropionate (DMSP), which is an abundant compound in phytoplankton.

287

DMSP represents a major fraction of organic sulfur in marine particles (mainly microorganisms) and is a major sulfur carrier for the transfer of this element through the marine food web (Box 1). Processes driving the synthesis, fluxes and transformations of DMSP and DMS are being unveiled5 because such research is not only being addressed from a global biogeochemistry perspective but also from the perspective of other disciplines, such as cell physiology, marine ecology and chemistry. What controls the planktonic production of atmospheric sulfur? Phytoplankton, sulfur and climate: the CLAW hypothesis

It was originally thought that phytoplankton produce DMS directly, and many authors still imply this when describing the oceanic biogeochemical sulfur cycle. Factors controlling phytoplankton activity (e.g. light, temperature and nutrients) were therefore expected to control DMS production. In 1987, this view provided the basis for the hypothesis of a feedback between oceanic phytoplankton and climate6, with a GAIAN (or geophysiological) thinking of the Earth in the background7,8. The CLAW HYPOTHESIS suggests that, because the radiation budget of the Earth is sensitive to the density of CLOUD CONDENSATION NUCLEI, and the major source of these nuclei over the oceans is DMS, ‘biological (algal) regulation of the climate is possible through the effects of temperature and sunlight on phytoplankton population and DMS production’6 (Fig. 1). The word ‘regulation’ implies not only ‘influence’ but also ‘feedback’ between DMS production and cloud ALBEDO, although the actual sign (positive or negative) of the feedback loop requires further experimental test because of the existing uncertainty on the response of biological DMS production to changes in radiation. Evolutionary concerns on climate regulation by phytoplankton

One of the main criticisms of the CLAW hypothesis (and, by extension, of Gaia) was its evolutionary feasibility9. If phytoplankton were the producers of DMS, the climate feedback hypothesis was apparently based on the altruism of algal populations. The authors of the original CLAW paper recognized that climate regulation would imply an improbable altruistic behavior of phytoplankton for the biosphere; therefore they proposed as possible ‘selfish’ explanations that cloud feeding by DMS would favor phytoplankton by enhancing the return of nitrogen (N, a generally limiting resource in the surface ocean) through rainfall and by lessening harmful UV input. However, not all phytoplankton species (not even all clones within species) produce equivalent amounts of the DMS precursor; moreover, many organisms other than DMS producers also have a high N demand and are sensitive to UV. Hence, the supposed benefits of DMS production would be shared with non-DMS producers, and DMS

http://tree.trends.com 0169–5347/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02152-8

Opinion

288

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

Table 1. Contribution of natural and man-made sources to sulfur emissions and burden in the global atmosphere Source

Global sulfur emissionsa (TgS y–1) (Mean, range)

Contribution to emissions (%)

Contribution to sulfate burdenb (%)

Man-made

70 (60–100)

70

37

Volcanic

7 (4–16)

7

18

Biogenicc

22 (15–50)

23

42d

aAfter

Refs 4, 50, 51. to the integrated SO42– content of the entire atmosphere column, after Ref. 52. cIncludes all terrestrial and oceanic biogenic emissions, of which >90% is oceanic DMS. dThe biogenic (mostly DMS) contribution to sulfate burden averages 42% but varies greatly among large regions: <10% Northern mid-latitude continents; <50–70% Extratropical oceans NH; >80% Tropics and SH. NH: Northern Hemisphere; SH: Southern Hemisphere. bContribution

producers would still appear to be altruistic. The concern was that altruism would fail at an evolutionary scale9,10. In phytoplankton, how would a gene persist that encodes a compound (with associated energetic costs) that benefits not only the producing species (or clone) but also all plankton and the entire biosphere? A recent attempt to overcome this conceptual altruism in the CLAW hypothesis was made by Hamilton and Lenton11. Local changes in radiative balance induced by microalgal blooms through DMS emission leads to increasing local winds and enhances the aerial dispersion of cells of the DMS-producer species or clones. If true, adaptation (i.e. selection between clonal patches) is consistent with phytoplankton influence on climate12,13. However, there are several problems with this idea. First, a large fraction of DMS emissions occur in oligotrophic

regions with non-blooming, mixed assemblages of phytoplankton species. Conditions for dispersal of DMS-producer cells would also disperse cells of competitor species, so it is difficult to see how DMScaused winds could confer advantage on the DMS producers. Furthermore, it is improbable that the atmospheric oxidation of DMS, which requires hours or days, would preferentially cause radiation and wind changes in the air parcel over the DMSproducing patch. Finally, the main problem is that the bulk of DMS production is not a direct action of phytoplankton, but results from a complex set of food-web interactions on which selection should act at organization levels higher than the population level. There is, however, no need to invoke either altruism or selfish adaptation related directly to DMS to explain the exhalation of atmospheric sulfur by plankton on an evolutionary basis. First, phytoplankton do not have metabolic pathways for the direct synthesis of DMS, instead they produce large amounts of non-volatile DMSP, a compound that performs several important physiological and ecological functions of benefit to the producer (Box 2). The evolution of these functions could be the driving force for evolutionary selection of DMSP synthesis in many phytoplankton10, which gives rise to climate-active DMS. DMS could therefore be viewed as ‘only’ a by- or waste product of planktonic DMSP production and degradation10,14,15, through interactions within the food web. DMSP production in microalgae: physiological acclimation or evolutionary adaptation?

According to current knowledge, any link between the algal cell and atmospheric DMS is via DMSP. There

Box 1. Biogeochemical roles of DMS and its biological precursor Dimethylsulfide (DMS) and its biological precursor dimethylsulfoniopropionate (DMSP) play important roles in global sulfur biogeochemistry: • Oceanic emission of DMS represents the main natural source of atmospheric sulfur.This emission flux is nearly enough to balance the global sulfur budget, compensating for a sulfur deficiency on the continents that is a result of runoff losses to the oceana. • DMS emission from the global ocean equals one-third of global anthropogenic sulfur emissions. However, as the atmospheric lifetime of DMS sulfur is longer than that of anthropogenic sulfur, oceans contribute about 40% of the total sulfur burden of the atmosphere. Over oceanic regions remote from continents (such as most of the Southern Hemisphere), the vast majority of atmospheric sulfur comes

http://tree.trends.com

from oceanic DMS. Oxidation of DMS in the marine atmosphere produces sulfur AEROSOLS that, either directly or by acting as cloud condensation nuclei, scatter solar radiation thereby influencing the radiative balance of the Earth.This has been suggested to be part of one of the major self-regulated feedback mechanisms linking global biosphere and climateb,c, with a present-day cooling capability of ~3.8 K (Ref. d). • Because it is an abundant component in many phytoplankton taxa, is valuable as compatible solute, and because it contains a methylated, reduced sulfur moiety, DMSP is very prone to microbial degradation and very appetizing for bacteria and grazers. It therefore represents a major carrier for sulfur transfer through microbial food webs and organic sulfur cycling in the pelagic oceane.

• One of the products of microbial DMSP degradation, methanethiol, is very reactive with trace metals and could, therefore, affect metal availability and chemistry in seawatere. Box Glossary Aerosol: airborne particle of any origin.

References a Lovelock, J.E. et al. (1972) Atmospheric sulphur and the natural sulphur cycle. Nature 237, 452–453 b Charlson, R.J. et al. (1987) Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326, 655–661 c Andreae, M.O. and Crutzen, P.J. (1997) Atmospheric aerosols: biogeochemical sources and role in atmospheric chemistry. Science 276, 1052–1058 d Watson, A.J. and Liss, P.S. (1998) Marine biological controls on climate via the carbon and sulphur geochemical cycles. Philos. Trans. R. Soc. London Ser. B 353, 41–51 e Kiene, R.P. et al. (2000) New and important roles for DMSP in marine microbial communities. J. Sea Res. 43, 209–224

Opinion

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

289

Box 2. Functions of DMSP in marine phytoplankton

CCN Atmospheric acidity

Albedo

Solar radiation (heat, light, UV)

Sulfate and sulfonate aerosol

Sulfur transport to continents

Temperature Wind speed DMSg

Oceanic mixing layer

DMSaq Planktonic food web

TRENDS in Ecology & Evolution

Fig. 1. The feedback system linking oceanic plankton and climate through the production of atmospheric sulfur.The original CLAW hypothesis postulated that production of dimethylsulfide (DMSaq) by phytoplankton, and its subsequent ventilation (DMSg) and oxidation in the atmosphere, feeds cloud condensation nuclei (CCN) in marine stratus, thereby increasing cloud albedo. If the consequent reduction in solar irradiance made phytoplankton produce less DMS, then a negative feedback would operate, thus stabilizing climate. Although this hypothesis is still conjecture, recent advances suggest that it is not only phytoplankton but the whole food web that releases DMS and that the response of net DMS production to changes in solar radiation might operate through the profound effects of surface vertical mixing on oceanic biogeochemistry and food-web dynamics.

are several possible factors that control phytoplankton production of DMSP: Taxonomy. Some species of phytoplankton produce more DMSP than do others. As a general rule, haptophytes (including coccolithophorids and small flagellates) and small dinoflagellates produce more, or have higher intracellular concentrations of, DMSP than do diatoms16,17. Nitrogen availability. Andreae18 first suggested that the relative production of DMSP and its N analog, COMPATIBLE SOLUTE Gly betaine (GBT), depends on contemporaneous N availability, so that the algae switch between the synthesis and accumulation of GBT (under N repletion) and DMSP (under N deficiency). This suggestion was later supported by the discovery of a transamination step in the DMSP synthesis pathway in the coccolithophorid Emiliania huxleyi19. In recent experimental work with batch and chemostat cultures20, however, a reciprocal (‘switching’) relationship with DMSP content was not evident even though GBT levels varied with a varying N supply. It might be that evolutionary adaptation to N availability had a greater influence on DMSP synthesis by algae than did short-term acclimation to N availability. This would explain the taxonomic patterns observed: most diatoms have evolved to be more adapted to conditions of N repletion and they synthesize less DMSP per cell volume16 (they probably synthesize more GBT or other N osmolytes); conversely, small haptophytes (e.g. coccolithophorids) and many small dinoflagellates are typical of more N-deficient conditions, so they have evolved to produce more http://tree.trends.com

Dimethylsulfoniopropionate (DMSP) has emerged as a fascinating compound as an increasing number of essential physiological and ecological functions besides being precursor of dimethylsulfide (DMS) have been revealed or suggested: • It is a compatible solute involved in osmoprotection in algae and bacteriaa–d and cryoprotection in algaea–d. • It acts as a methyl donor in metabolic reactionsb. • Its synthesis and exudation by unicellular algae might represent an overflow mechanism for excess reduced sulfur and reducing power under conditions of unbalanced growthd. • It is the precursor of cues for chemosensory attraction (DMS is the chemical signal) between algae and bacteriae and between zooplankton and birdsf, and chemosensory deterrence (acrylate is the chemical signal) between algae and microzooplankton grazersg. References a Malin, G. and Kirst, G.O. (1997) Algal production of dimethyl sulfide and its atmospheric role. J. Phycol. 33, 889–896 b Kiene, R.P. et al., eds (1996) Biological and Environmental Chemistry of DMSP and Related Sulfonium Compounds, Plenum Press c Welsh, D.T. (2000) Ecological significance of compatible solute accumulation by micro-organisms: from single cells to global climate. FEMS Microbiol. Rev. 24, 263–290 d Stefels, J. (2000) Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43, 183–197 e Zimmer-Faust, R.K. et al. (1996) Bacterial chemotaxis and its potential role in marine dimethylsulfide production and biogeochemical sulfur cycling. Limnol. Oceanogr. 1330–1334 f Nevitt, G.A. et al. (1995) Dimethyl sulphide as a foraging cue for Antarctic Procellariiform seabirds. Nature 376, 680–682

DMSP. This implies that the probability of finding higher levels of DMSP is greater under conditions of N depletion because of phytoplankton succession, and one should not expect strong shifts in DMSP levels in response to very short pulses of N supply. A significant exception to this hypothetical rule is the haptophyte Phaeocystis, which forms dense coastal blooms in highnitrate, silicate-deficient conditions, but which exhibits high intracellular DMSP concentrations. Light. It is still unclear whether DMSP biosynthesis is directly light dependent. Although dark uptake of exogenous sulfate is common in microalgae, there is little evidence for dark DMSP production21. A recent study has found that the DMSP synthesis rate in oceanic phytoplankton was a diurnal process that was strictly proportional to 14C fixation (photosynthesis) (R. Simó et al., unpublished). There is also no clear answer as to whether DMSP production responds to

Opinion

290

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

Sea surface Viral attack Autolysis Exudation Algae DMSP

Assimilation after grazing

Ventilation Bacteria Nonvolatile sulfur

Assimilation into protein DMSP dissolved Cell lysis Release

Photo-oxidation DMSP lyase (bacterial and algal)

DMS

Bacteria

Protozoans Copepods DMSP in fecal and detrital material Sinking

TRENDS in Ecology & Evolution

Fig. 2. The fate of phytoplanktonic dimethylsulfoniopropionate (DMSP) in seawater. DMSP is released from the cell through exudation and, most importantly, by cell lysis and grazing. A fraction is assimilated by grazers, whereas the rest is either cleaved to dimethylsulfide (DMS) by algal and bacterial DMSP lyases, or used by heterotrophic bacteria through other pathways. DMSP acts both as a source of carbon and a source of sulfur for methionine and subsequent protein synthesis. DMS is also consumed by bacteria, and lost through photooxidation. Only a small fraction escapes this tight cycling and vents to the atmosphere. DMSP cycling is an integral part of the food web, and foodweb dynamics controls DMS emission to the atmosphere.

changes in sunlight intensity; experimental observations of intracellular DMSP content under different radiation treatments have given ambiguous results5,21. Light and nutrient availability. Stefels21 has recently hypothesized that DMSP production is an overflow mechanism for when growth is unbalanced by scarcity of nutrients and there is a need for dissipating excess energy and excess reduced sulfur. Because DMSP is synthesized from Met through the loss of the N moiety, its production would save and regenerate intracellular N while sulfate assimilation, carbon fixation and Met synthesis keep going under nutrientlimited conditions. Also, because intracellular DMSP levels are much higher than those of Met and its precursor Cys, production of DMSP would keep the concentrations of these two amino acids low, thereby preventing possible inhibitory feedback mechanisms from occurring. Even the low DMSP exudation rates estimated in healthy phytoplankton (1–10% d−1; Ref. 22) would be enough to serve as physiological overflow mechanism. Extracellular DMSP LYASE, either ectoplasmatic or in bacteria surrounding the algal cell, would maintain enough of a membrane gradient to facilitate this exudation. Stefels’ hypothesis agrees with Margalef’s notion23 that, because photon catching, which provides reducing power to the cell, cannot be discontinued completely, the synthesis of relatively simple, N- or phosphorus (P)-poor molecules is stimulated when essential nutrients are scarce and the cell cannot invest in protein and division. These carbon–energy overflow substances might evolve through natural selection to be useful in the cell and become constituents of, for example, auxiliary structures or defense mechanisms23. DMSP turnover and DMS production

Now we understand a little more about why and how many microalgae contain DMSP. However, the path http://tree.trends.com

linking phytoplankton to DMS through DMSP is not straightforward. Attempts to correlate DMS to phytoplankton biomass (chlorophyll a concentration) or activity (primary production) on large spatial or temporal scales have failed2. This might be not only because of the taxonomy dependence of DMSP production in algae, but also because most DMSP breakdown into DMS requires DMSP being released extracellularly. Exudation by healthy algae represents only a small fraction of DMSP release, which occurs mostly through cell lysis (autolysis and viral attack) and grazing24–27. In grazing, a fraction of the algal DMSP is assimilated by the grazer26,27. Some of the released DMSP is then converted into DMS by algal and/or bacterial DMSP lyases, either free in solution, attached to particles or in the guts and vacuoles of grazers. A large fraction of DMSP is, however, utilized through other pathways that do not produce DMS. Particularly, DMSP demethylation and demethiolation represents a source of reduced sulfur that can be assimilated readily into protein by microorganisms at a lower energetic cost28,29 (Fig. 2). A novel insight into what controls DMS production can be obtained by splitting the production flux into two terms: DMS production flux = DMSP flux × DMS yield where the DMSP flux is the DMSP consumption rate and the DMS yield is the DMS production efficiency from DMSP consumption (i.e. DMS production rate × 100/DMSP consumption rate). DMSP flux. DMSP is very labile, with turnover times of between a few hours and one to two days. It is released and consumed at high rates and does not accumulate in large amounts in dissolved form in seawater. The DMSP flux is, therefore, related to the rate of phytoplankton loss, which is influenced by herbivore grazing pressure, viruses, competition for resources and environmental (e.g. nutrients, turbulence and radiation) stress, and is also related to the growth efficiency and the sulfur demand of grazers. The DMSP flux is, therefore, part of food-web dynamics. DMS production yield. This term has been addressed recently from two different perspectives. According to Kiene et al.30, DMS production yield would result from the combination of seawater DMSP concentration and the bacterial sulfur demand. In my opinion, DMSP concentration should be redrawn as the ‘DMSP flux’ defined above. In any case, this relates to phytoplankton taxonomy and loss rate. Bacterial sulfur demand is assumed to depend on heterotrophic bacterial production. The higher their sulfur demand, the more DMSP the bacteria will use as a sulfur source for biomass production, and the less will be wasted as DMS. Thus, low DMSP concentration (flux) with high sulfur demand would lead to higher DMSP assimilation and lower DMS yield, whereas high DMSP with low sulfur demand

Opinion

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

291

Box 3.The bacterial switch Marine bacteria are involved in both dimethylsulfide (DMS)- and non-DMSproducing degradation pathways of dimethylsulfoniopropionate (DMSP). Bacteria therefore exert a controlling role on DMS production efficiency by switching DMSP degradation towards more or less DMS production. Once this switching role is recognized, two questions arise that are crucial to our understanding of the quantitative contribution of bacteria to net DMS production: 1 Is DMSP a (universal) substrate for most phylotypes in heterotrophic bacterioplankton?The recent observation of a slight correlation between turnover of dissolved DMSP and total heterotrophic bacterial activitya points to an affirmative answer. 2 Is DMSP use a characteristic of only a few bacterial metabolisms? Recent work by González and colleaguesb,c provides a series of evidences that Roseobacter (a lineage of α-Proteobacteria, which are very abundant in coastal and open-sea waters) are involved in DMSP degradation.

Perhaps the answer is both: only a few lineages of bacteria process DMSP, but they are so abundant that they account for a large fraction of total bacterial activity. The combined techniques of radioactive 35S-DMSP tracingd and FLUORESCENT IN-SITU HYBRIDIZATION, using either MICROAUTORADIOGRAPHYe or FLOW-CYTOMETRY SORTINGf,g, will help answer these questions. Gene probing for the enzyme DMSP lyase also opens interesting perspectives for allocating DMS production among bacterioplankton. Box Glossary Flow-cytometry sorting: technique by which cells (e.g. unicellular planktonic organisms) are sorted by size or fluorescence and recovered using a flow cytometer. Fluorescence is either associated with cell pigments or generated by protein and DNA stains or metabolic and phylogenetic probes. Fluorescent in-situ hybridization: method by which fluorescent DNA probes are used to locate certain phylogenetic groups among the microorganisms of unperturbed natural communities. Microautoradiography: method by which a radioactively marked substrate is tracked as it is incorporated by microorganisms, by the signal left on a photographic emulsion.

would lead to higher DMS yield. In this ‘DMSP availability hypothesis’, the DMSP concentration or flux occurs in the bulk dissolved phase. Recent work31 suggests that bacteria growing attached to, or closely around algal cells might be exposed to very high local levels of DMSP, which would lead to DMS yields that are higher than those inferred from bulk seawater measurements. An independent, empirical hypothesis by Simó and Pedrós-Alió32 suggests that the DMS production yield varies non-linearly with MIXING LAYER depths of the ocean surface, probably because of differential photoinhibitory effects of UV on phytoplankton and bacteria. According to this ‘vertical mixing hypothesis’, very shallow mixing exposes microorganisms to higher levels of harmful UV-B, and heterotrophic bacteria are damaged more, because of lack of protective pigments. Shallow mixing, therefore, inhibits bacterial DMSP–sulfur assimilation and favors higher DMS yields. Both hypotheses are compatible and complementary. Both highlight that DMS production is a function not only of the flux at which its precursor is released from primary producers, but also of the efficiency at which the precursor is converted into DMS. The phytoplankton–bacterioplankton coupling is crucial for this efficiency, because bacteria possess the principal switch for a more or less efficient DMS production (Box 3). http://tree.trends.com

References a Kiene, R.P. and Linn, L.J. (2000) Distribution and turnover of dissolved DMSP and its relationship with bacterial production and dimethylsulfide in the Gulf of Mexico. Limnol. Oceanogr. 45, 849–861 b González, J.M. et al. (1999) Transformation of sulfur compounds by an abundant lineage of marine bacteria in the alpha-subclass of the class Proteobacteria. Appl. Environ. Microbiol. 65, 3810–3819 c González, J.M. et al. (2000) Bacterial community structure associated to a DMSPproducing North Atlantic algal bloom. Appl. Environ. Microbiol. 66, 4237–4246 d Kiene, R.P. and Linn, L.J. (2000) The fate of dissolved dimethylsulfoniopropionate (DMSP) in seawater: tracer studies using 35S-DMSP. Geochim. Cosmochim. Acta 64, 2797–2810 e Lee, N. et al. (1999) Combination of fluorescent in situ hybridization and microautoradiography – A new tool for structure-function analyses in microbial ecology. Appl. Environ. Microbiol. 65, 1289–1297 f Servais, P.C. et al. (1999) Coupling bacterial activity measurements with cell sorting by flow cytometry. Microb. Ecol. 38, 180–189 g Gasol, J.M. and del Giorgio, P.A. (2000) Using flow cytometry for counting natural planktonic bacteria and understanding the structure of planktonic bacterial communities. Sci. Mar. 64, 197–224

DMS consumption

Once produced, DMS is consumed biologically33 within one to several days. In oxygenated seawater, methylotrophic bacteria are the most probable consumers33, although several metabolisms are possible. With the exception of very particular conditions, DMS concentrations are generally constrained2 to between 0.5 and 10 nM. Such a narrow range suggests that losses occur in tight coupling with production processes. Indeed, not only biological consumption but also significant photolysis and ventilation rates have been found coupled to DMS production in the open ocean34,35. Most studies show that bacteria are a major sink for DMS. Therefore, because bacterioplankton are involved in both DMSP and DMS utilization (Fig. 2), factors controlling bacterial activity36 (such as UV-B radiation, temperature, nutrients and dissolved organic matter quality) ultimately also play a role in controlling DMS concentration. Food-web dynamics: the key to sulfur venting from the ocean

The recent advances reviewed here reinforce the old37 but often forgotten idea that DMS ‘net production’ (for the flux of DMS made available to air–sea exchange) is related to food-web dynamics and not just to phytoplankton taxonomy, physiology and activity. This explains why global attempts to correlate

292

Opinion

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

directly DMS concentration with chlorophyll a levels have failed2. It might also explain why, in the Equatorial Pacific over a period of 15 years, DMS concentrations during El Niño events were not significantly different from those in the background years38, although important reductions in chlorophyll a levels and primary production accompanied such events39. Shifts in the food web towards more microbial-dominated webs (fewer large cells, less particle sedimentation and high production–grazing coupling) would have compensated for the apparent loss of potential DMS generation through the decrease in chlorophyll a and primary production, with higher biomass-specific DMS production. A similar food-web related explanation has been given for the ‘summer DMS paradox’: In the large oligotrophic regions of the global surface ocean, maximum DMS concentrations occur in summer when sunlight is maximum but surface chlorophyll a concentrations are at their annual minimum32. Again, higher efficiency in food-web DMS production together with lower DMS consumption would compensate for a lower DMS generation potential. The ‘summer DMS paradox’ supports the CLAW hypothesis because stronger sunlight and higher temperatures correspond with higher DMS concentrations (Fig. 1). However, this apparent link between food-web DMS production and incident radiation flux reinforces the controversy of how DMS might have evolved to be a biogenic product for climate regulation. If it was hard to explain that a phytoplankton population (individual or species) might produce something for the good of the community, it is even harder to think that members of the whole plankton community (constituted by parts that even compete among themselves for resources) ‘agree’ to produce something for the good of them all and the biosphere. This could be regarded as a fully Gaian mechanism in that a community with a stabilizing feedback (i.e. with an environment-altering trait that benefits the community, for example, moisture retention and nutrient recycling within rainforests) would tend to persist and spread through evolution by selection at the ecosystem level12. An alternative explanation, based only on the selfish adaptation of individuals or populations can be given. In summer, phytoplankton succession favors populations more adapted to N deficiency, populations that would therefore synthesize more DMSP per individual. Coupling between primary production, algal lysis and grazing by microzooplankton is tighter in summer and involves a larger fraction of primary production (because of smaller cell size and consequent reduced sinking). This would lead to higher DMSP flux per biomass in the surface layer. Shallow mixing causes bacterial activity to be reduced by UV-B damage and to be limited by the lack of nutrients. This would decrease DMSP assimilation, increase DMS production and decrease DMS consumption, all leading to higher net http://tree.trends.com

DMS production. All these effects ‘at the individual or population level’ would result in higher net DMS production in summer, when incident radiation is higher. The negative feedback response of the oceanic DMS emission to changes in incident solar radiation (suggested in the CLAW hypothesis) is, therefore, feasible through changes in vertical mixing and water-column light regime, and without necessarily calling for altruism or natural selection at the ecosystem level. Prospects

To improve our understanding of the role that biogenic volatile sulfur from the ocean plays in global biogeochemistry and climate, we should prioritize research on: • Further understanding the links between DMSP, DMS and food-web structure and dynamics. A way to achieve this is by introducing temporal and/or spatial variability in the study of the processes of DMS cycling. Most studies in the open ocean have provided snapshots of the DMS cycle, which are useful for revealing potential couplings or uncouplings. However, these results should be used with caution when extrapolated to longer timescales. Only the observation of the changes in process rates through seasons, through short-term (e.g. storms) and long-term (e.g. El Niño) overturning events, through BIOGEOCHEMICAL PROVINCES or MESOSCALE HYDRODYNAMIC STRUCTURES will help unravel the dynamics of the sulfur cycle in relation to a dynamic environment and biota. This is relevant to research on how the ocean responds as a sulfur source to global warming. Climate warming has the potential to alter food-web structure and dynamics40, thereby impacting the net air–sea exchange of gases41. This seems clear for CO2 and O2, and should be the same for DMS. • Opening the ‘black box’ of microbial metabolisms involved in the production and consumption of volatile sulfur, so that we are able to identify which microorganism does what and ask what controls their distribution and activity. As outlined in Box 3, a first step towards such identification will be provided by molecular techniques for bacterial ‘functional diversity’ that are currently being developed. • Being able to define ‘DMS-producing pelagic ecosystems’ (i.e. community structures and dynamics favorable to net DMS production): what they are like, how they work, when and where they occur, and in response to what physical FORCINGS. This should make use of other concepts such as those of ‘biogeochemical provinces’42, ‘phytoplankton functional groups’ (e.g. coccolithophorids, silico-phytoplankton and N2 fixers), ‘pelagic food webs’43 or ‘biogeochemical guilds’13, adapted or combined. Will such DMSproducing ecosystems be characterized by key measurable (or estimable) parameters, either in situ or from remote sensors? Chemotaxonomic

Opinion

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

293

Glossary Albedo: measurement of the ability of a surface to reflect light; in a global sense, the proportion of solar radiation reaching the Earth that is reflected back into space. It plays an important role in regulating the heat balance of the globe. White surfaces, such as clouds and ice or snow cover, have a high contribution to regional and planetary albedo. Biogeochemical provinces: 56 regions of the global oceans defined by A. Longhurst on the basis of differentiated ocean currents, fronts, topography and recurrent features in the sea surface chlorophyll a field. CLAW hypothesis: hypothesis released in 1987 by Charlson, Lovelock, Andreae and Warren, which states that regulation of climate by oceanic phytoplankton is possible through the production of dimethylsulfide (DMS).This DMS oxidates in the atmosphere, changes cloud albedo and, therefore, affects temperature and sunlight.These, in turn, have feedback effects on phytoplankton population and DMS production.The link between phytoplankton and climate made by the CLAW hypothesis has been proposed as one of the testable processes in the GAIATHEORY. Cloud condensation nuclei (CCN): tiny airborne particles (generally of submicronic size) that serve as nuclei for water condensation in clouds.The amount and size of CCN determines the optical depth of the cloud and, therefore, the cloud albedo. Hygroscopic sulfate aerosols formed upon atmospheric oxidation of DMS are excellent CCN. Compatible solute: a cytoplasmatic organic solute that occurs at concentrations high enough to be active but that is compatible with metabolism and has no inhibitory effects (unlike inorganic ions). Major functions of compatible solutes are protecting proteins and stabilizing membranes

Acknowledgements I thank all who provided encouragement and fruitful discussions, particularly C. PedrósAlió, J.M. Gasol, J. Grimalt, G. Malin, M. Estrada, X. Moran, R. Massana, C. Marrasé, R. Kiene and S. Archer. Valuable comments by D.Wilkinson and an anonymous reviewer helped improve the article. A brief encounter with the late M. Keller, whose work has been a fundamental source of thoughts for this paper, was very enlightening.

under conditions of high ionic strength (osmoprotection) and freezing temperatures (cryoprotection). DMS: dimethylsulfide (H3C-S-CH3), the most abundant and ubiquitous volatile sulfur compound in marine environments, where it occurs at concentrations of up to 300 nM (usually 1–10 nM) as a product of the enzymatic cleavage of dimethylsulfoniopropionate (DMSP). DMSP: dimethylsulfoniopropionate [(H3C)2-S+CH2CH2COO−], abundant intracellular component in many phytoplankton taxa, where it acts as a compatible solute. It breaks down enzymatically into DMS and acrylate. DMSP lyase: enzyme that cleaves DMSP. It occurs in some algae, protozoa, bacteria and fungi, either intracellularly or ectoplasmatically. Forcing: change to an environmental variable induced by an environmental factor; for example, radiative forcing caused by changing concentrations of aerosols or CO2; for example wind speed forcing on ocean mixing. Gaia Theory: first released as hypothesis by J.E. Lovelock in the 1970s as the notion of the biosphere as an adaptive control system that can maintain the Earth in homeostasis that is comfortable for life. It evolved into theory as a numerical basis was introduced. In its latest formulations, the theory proposes that life and the abiotic Earth have evolved together (as the single entity Gaia) with emerging regulatory feedback mechanisms that have kept surface environmental conditions in some degree of homeostasis. Lovelock named the science of Gaia geophysiology. Although the GaiaTheory is still controversial, it has been influential in the rise of Earth system science.

analysis of pigments is being used as a powerful descriptive tool, but it covers only phytoplankton and misses the dynamics of the food web. Unsurprisingly, pigment distributions are insufficient to explain the temporal variability of DMS concentration44. To obtain the ‘ecosystem fingerprints’ I am suggesting, taxonomic descriptors should be combined with dynamic descriptors such as the new:total primary production ratio (f ratio), autotrophic:heterotrophic biomass ratio (A:H), phytoplankton production:plankton community respiration ratio (P:R), or grazed:total primary production ratio. The increasing capabilities of satellites (particularly through algorithm developments) make them a very helpful tool because they provide variables needed for the largescale estimates of some of these dynamic descriptors. • Understanding the short-term effects of meteorological forcing (total solar radiation, UV intensity and quality, and wind speed) on DMSproducing communities and processes34,35,45,46. • Developing tools for predicting present and future DMS distributions in the surface ocean. Plankton–DMS models currently in developmental stages47–49 need to be fed with better (and more comprehensive) experimental data and combined with general circulation models to make them http://tree.trends.com

GBT: glycine betaine [(H3C)3-N+-CH2COO−], intracellular component in some phytoplankton, where it acts as a compatible solute. It is considered to be among the most effective and widely used compatible solutes found in nature. Mesoscale hydrodynamic structures: eddies, filaments and fronts formed by waters of different densities and motions that introduce instability to otherwise homogeneous water masses, with consequent decoupling in food webs and biogeochemical processes. Mixing layer: surface layer of the water column of the oceans. It has a homogeneous density, and is ‘separated’ from the water beneath by a density gradient (pycnocline). Although it is generally known as mixed layer, some authors distinguish between the ‘mixed’ layer (i.e. the layer above a strong pycnocline dictated by the seasonality of the heat flux and by major currents; it can either be well mixed by turbulence or in transient stagnant conditions evolved from past mixing) and the ‘mixing’ layer (i.e. the layer actually in turbulent mixing; its thickness changes in the short term as a rapid response to changes in heat flux and wind speed). The depth of the mixed–mixing layer governs, among other things, the exposure of dissolved substances and planktonic organisms to solar radiation. SOLAS: Surface Ocean–Lower Atmosphere Study: a new international program associated with the International Geosphere–Biosphere Program (IGBP). Its goal is to achieve quantitative understanding of ocean–atmosphere interactions and feedback, and how this coupled system affects and is affected by climate and environmental change.

applicable to regional or global scales. Again, modeling requires deep understanding of how the system works. A short cut to by-pass modeling would be to find empirical algorithms between DMS concentration and data from satellites or oceanographic atlas. The launching of the new international program SOLAS (Surface Ocean–Lower Atmosphere Study, http://www.ifm.uni-kiel.de/ch/solas/main.html), one of the major goals of which is the study of marine sulfur emissions, will help coordinate the efforts designed to address the challenges I have presented here. The provocative hypothesis of an active feedback link between marine biota and climate through the production of atmospheric sulfur has yielded unprecedented progress in our understanding of the dynamics of oceanic sulfur and its implications for the physiology of individuals and ecosystems. Although fundamental gaps in our knowledge still remain, it must be recognized that, today, DMS and DMSP are perhaps the organic substances for which distribution and transformation processes in the ocean are the best known. Taking advantage of the huge research effort made so far, the DMS/DMSP story offers a unique opportunity to link atmospheric chemistry and climate to the ecology and evolution of marine plankton.

294

Opinion

References 1 Lovelock, J.E. et al. (1972) Atmospheric sulphur and the natural sulphur cycle. Nature 237, 452–453 2 Kettle, A.J. et al. (1999) A global database of sea surface dimethylsulfide (DMS) measurements and a procedure to predict sea surface DMS as a function of latitude, longitude, and month. Global Biogeochem. Cycles 13, 399–444 3 Andreae, M.O. and Crutzen, P.J. (1997) Atmospheric aerosols: biogeochemical sources and role in atmospheric chemistry. Science 276, 1052–1058 4 Kettle, A.J. and Andreae, M.O. (2000) Flux of dimethylsulfide from the oceans: a comparison of updated data sets and flux models. J. Geophys. Res. 105, 26793–26808 5 Malin, G. and Kirst, G.O. (1997) Algal production of dimethyl sulfide and its atmospheric role. J. Phycol. 33, 889–896 6 Charlson, R.J. et al. (1987) Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326, 655–661 7 Lovelock, J.E., ed. (1995) The Ages of Gaia (2nd edn), Oxford University Press 8 Lovelock, J.E., ed. (2000) Homage to Gaia, Oxford University Press 9 Dawkins, R., ed.(1982) The Extended Phenotype, Oxford University Press 10 Kaldeira, K. (1989) Evolutionary pressures on planktonic production of atmospheric sulphur. Nature 337, 732–734 11 Hamilton, W.D. and Lenton, T.M. (1998) Spora and Gaia: how microbes fly with their clouds. Ethol. Ecol. Evol. 10, 1–16 12 Wilkinson, D.M. (1998) Is Gaia really conventional ecology? Oikos 84, 533–536 13 Lenton, T.M. (1998) Gaia and natural selection. Nature 394, 439–447 14 Volk, T., ed. (1998) Gaia’s Body: Toward a Physiology of Earth, Springer-Verlag 15 Wilkinson, D.M. (2000) Homeostatic Gaia: an ecologist’s perspective on the possibility of regulation. In Abstracts Guide of the 2nd Chapman Conference on the Gaia Hypothesis, pp. 11–12 16 Keller, M.D. et al. (1989) Dimethyl sulfide production in marine phytoplankton. In Biogenic Sulfur in the Environment (Saltzman, E. and Cooper, W.J., eds), pp. 167–182, American Chemical Society 17 Liss, P.S. et al. (1993) Production of DMS by marine phytoplankton. In Dimethylsulphide: Oceans, Atmosphere and Climate (Restelli, G. and Angeletti, G., eds), pp. 1–14, Commission of the European Communities – Kluwer 18 Andreae, M.O. (1986) The ocean as a source of atmospheric sulfur compounds. In The Role of Sea-Air Exchange in Geochemical Cycling (Buat-Menard, P., ed.), pp. 331–362, Reidel 19 Gage, D.A. et al. (1997) A new route for synthesis of dimethylsulphoniopropionate in marine algae. Nature 387, 891–894 20 Keller, M.D. et al. (1999) Production of glycine betaine and dimethylsulfoniopropionate in marine phytoplankton. Ι. Batch cultures. ΙΙ. N-limited chemostat cultures. Mar. Biol. 135, 237–257 21 Stefels, J. (2000) Physiological aspects of the production and conversion of DMSP in marine algae and higher plants. J. Sea Res. 43, 183–197 22 Laroche, D. et al. (1999) DMSP synthesis and exudation in phytoplankton: a modeling approach. Mar. Ecol. Prog. Ser. 180, 37–49

http://tree.trends.com

TRENDS in Ecology & Evolution Vol.16 No.6 June 2001

23 Margalef, R. (1997) Our Biosphere, Excellence in Ecology 10 (Kinne, O., ed), Ecology Institute 24 Hill, R.W. et al. (1998) Virus-mediated total release of dimethylsulfoniopropionate from marine phytoplankton: a potential climate process. Aquat. Microb. Ecol. 14, 1–6 25 Malin, G. et al. (1998) Elevated production of dimethylsulfide resulting from viral infection of cultures of Phaeocystis pouchetii. Limnol. Oceanogr. 43, 1389–1393 26 Tang, K. et al. (1999) Dimethylsulfoniopropionate (DMSP) in marine copepods and its relation with diets and salinity. Mar. Ecol. Prog. Ser. 179, 71–79 27 Archer, S.D. et al. (2001) A dilution approach to quantify the production of dissolved dimethylsulphoniopropionate and dimethyl sulphide due to microzooplankton herbivory. Aquat. Microb. Ecol. 23, 131–145 28 Kiene, R.P. et al. (1999) Dimethylsulfoniopropionate and methanethiol are important precursors of methionine and protein-sulfur in marine bacterioplankton. Appl. Environ. Microbiol. 65, 4549–4558 29 Kiene, R.P. and Linn, L.J. (2000) The fate of dissolved dimethylsulfoniopropionate (DMSP) in seawater: tracer studies using 35S-DMSP. Geochim. Cosmochim. Acta 64, 2797–2810 30 Kiene, R.P. et al. (2000) New and important roles for DMSP in marine microbial communities. J. Sea Res. 43, 209–224 31 Scarratt, M. et al. (2000) Particle size-fractionated kinetics of DMS production: where does DMSP cleavage occur at the microscale? J. Sea Res. 43, 245–252 32 Simó, R. and Pedrós-Alió, C. (1999) Role of vertical mixing in controlling the oceanic production of dimethyl sulphide. Nature 402, 396–399 33 Kiene, R.P. (1993) Microbial sources and sinks for methylated sulfur compounds in the marine environment. In Microbial Growth on C1 Compounds (Murrell, J.C. and Kelly, D.P., eds), pp. 15–33, Intercept 34 Kieber, D.J. et al. (1996) Impact of dimethylsulfide photochemistry on methyl sulfur cycling in the Equatorial Pacific Ocean. J. Geophys. Res. 101, 3715–3722 35 Simó, R. and Pedrós-Alió, C. (1999) Short-term variability in the open ocean cycle of dimethylsulfide. Global Biogeochem. Cycles 13, 1173–1181 36 Kirchman, D., ed. (2000) Microbial Ecology of the Oceans, J. Wiley & Sons 37 Belviso, S. et al. (1990) Production of dimethylsulfonium propionate (DMSP) and dimethylsulfide (DMS) by a microbial food web. Limnol. Oceanogr. 35, 1810–1821

38 Bates, T.S. and Quinn, PK. (1997) Dimethylsulfide (DMS) in the equatorial Pacific Ocean (1982 to 1996): Evidence of a climate feedback? Geophys. Res. Lett. 24, 861–864 39 Chavez, F.P. et al. (1999) Biological and chemical response of the Equatorial Pacific Ocean to the 1997–98 El Niño. Science 286, 2126–2131 40 Petchey, O.L. et al. (1999) Environmental warming alters food-web structure and ecosystem function. Nature 402, 69–72 41 del Giorgio, P.A. et al. (1999) Linking planktonic biomass and metabolism to net gas fluxes in northern temperate lakes. Ecology 80, 1422–1431 42 Longhurst, A. (1995) Seasonal cycles of pelagic production and consumption. Prog. Oceanogr. 36, 77–167 43 Legendre, L. and Rassoulzadegan, F. (1995) Plankton and nutrient dynamics in marine waters. Ophelia 41, 153–172 44 Uher, G. et al. (2000) Distribution and air–sea exchange of dimethyl sulphide at the European western continental margin. Mar. Chem. 69, 277–300 45 Sakka, A. et al. (1997) Effects of reduced ultraviolet radiation on aqueous concentrations of dimethylsulfoniopropionate and dimethylsulfide during a microcosm study in the lower St Lawrence Estuary. Mar. Ecol. Prog. Ser. 149, 227–238 46 Slezak, D. et al. (2000) Effects of solar radiation on the production and consumption of DMSP and DMS in the upper water column. Abstract for the 7th European Marine Microbiology Symposium, Amsterdam, September 2000 47 Gabric, A. et al. (1993) Modeling the production of dimethylsulfide during a phytoplankton bloom. J. Geophys. Res. 98, 22805–22816 48 van den Berg, A.J. et al. (1996) Model structure and analysis of dimethylsulphide (DMS) production in the southern North Sea, considering phytoplankton dimethylsulphoniopropionate(DMSP lyase and eutrophication effects. Mar. Ecol. Prog. Ser. 145, 233–244 49 Jodwalis, C.M. et al. (2000) Modeling of dimethyl sulfide ocean mixing, biological production, and sea-to-air flux for high latitudes. J. Geophys. Res. 105, 14387–14399 50 Lelieveld, J. et al. (1997) Terrestrial sources and distribution of atmospheric sulphur. Philos. Trans. R. Soc. London Ser. B 352, 149–158 51 Rodhe, H. (1999) Human impact on the atmospheric sulfur balance. Tellus A-B 51, 110–122 52 Chin, M. and Jacob, D.J. (1996) Anthropogenic and natural contributions to tropospheric sulfate: a global model analysis. J. Geophys. Res. 101, 18691–18699

New In Brief Section Would you like to supply us with timely, newsworthy information that we can pass on to our News Editors for the In Brief section? If so, please contact: The Assistant Editor [email protected] and state the information in fewer than 200 words