Production of Atmospheric DMS by Oceanic Phytoplankton

Dimethylsulfide (DMS) is the most abundant form of volatile sulfur (S) in the ocean and is the main source of biogenic reduced S to the global atmosphere

(Andreae and Crutzen, 1997). The sea-to-air flux of S due to DMS is currently estimated to be in the range (15-33) Tg S yr-1, constituting about 40% of the total atmospheric sulfate burden (Chin and Jacob, 1996). At the hemispheric scale (Gondwe et al., 2003) estimate that seawater DMS contributes 43% of the mean annual column burden of non-sea-salt sulfate (nss-SO2-) in the relatively pristine southern hemisphere, but only 9% in the northern hemisphere, where anthropogenic sulfur sources are overwhelming.

During its synthesis and cycling in the upper ocean, DMS is ventilated to the atmosphere, where it is rapidly oxidized to form nss-SO4- and meth-anesulfonate (MSA) aerosols. Sulfate aerosols (of both biogenic and anthropogenic origin) play an important role in the earth's radiation balance both directly through scattering, absorption, and reflection of solar and terrestrial radiation, and indirectly, by modifying cloud microphysical properties (Charlson et al., 1992). The flux of DMS from the ocean to the atmosphere is an important concern for atmospheric modelers since the net effect of DMS is believed to be a cooling effect for the global climate (Kiene, 1999). While wind forcing is known to control the piston pumping velocity of DMS gas across the air-sea interface (Liss and Slater, 1974), the percent yield of DMS from DMSP (100 X DMSproduction Rate/DMSPConsumption Rate) is well correlated to MLDs (Simo and Pedros-Allo, 1999). Combining this relationship with the global climatologies of MLDs produces a map demonstrating the seasonal and global variability of this efficiency (Fig. 5). Combining these estimates with estimates of the air-sea flux of DMS (Bates et al., 1987a) may have some potential for gaining additional insight in the level of net DMS production in the ocean. In addition to seasonal variations in DMS production, ecosystem changes in the Pacific equatorial regions are now known to undergo spatial and temporal variations that are linked to larger scale climate variations such as ENSO (Karl et al., 1995). Such variability should also impact the production level of DMS and its associated air-sea flux.

Various phytoplankton species synthesize differing amounts of dime-thylsulfoniopropionate (DMSP), the precursor to DMS. The function of DMSP in algal physiology seems to be varied, and it is thought to act as an osmolyte, a cryoprotectant, and also relieve oxidative stress in the algal cell (Kirst et al., 1991; Liss et al., 1993; Stefels, 2000; Sunda et al., 2002). In general, coccolithophorids and small flagellates have higher intracellular concentrations of DMSP.

It is well established that the oceans are highly supersaturated in DMS with respect to atmospheric concentrations (Barnard et al., 1982; Liss et al., 1993). In fact, DMS is so supersaturated that recent field experiments on air-sea CO2 gas flux have used the flux of DMS as a proxy signal to sparameterize the physical process of air-sea gas transfer (Dacey, private communication, 2003). Vertical profiles of DMS in the Sargasso Sea (Dacey

DMS Yield Climatologies

Figure 5: DMS percent yield climatologies estimated using observed MLD climatologies from the NODC XBT data set and the Simo and Pedros-Allo (1999) DMS yield relationship (For colour version, see Colour Plate Section).

DMS Yield Climatologies

Figure 5: DMS percent yield climatologies estimated using observed MLD climatologies from the NODC XBT data set and the Simo and Pedros-Allo (1999) DMS yield relationship (For colour version, see Colour Plate Section).

et al., 1998) show a marked subsurface 10 m maximum. A comparison of the depth-integrated annual cycle of DMS, chlorophyll, and primary production rates in this region shows that DMS concentrations peak in late summer (August) when both chlorophyll and primary production rates are lower than their earlier spring maximum values - suggesting that DMS production by phytoplankton is not directly linked to photosynthetic processes and may be due to release from grazing by zooplankton (Leck et al., 1990). However, correlations between algal biomass and DMS concentrations have been found for dinoflagellate and coccolithophore blooms (Leck et al., 1990). Because of these obvious complexities, attempts to model the production of DMS (Gabric et al., 1993) in ecosystem models have included both direct (primary production) and indirect (grazing) sources. Refer to Lee et al. (1999) for a recent review of DMS in aquatic environments.

Shaw (1983) and then Charlson et al. (1987) [notably called the "CLAW Hypothesis,'' derived from the first letters of the author's last names], postulated links between DMS, atmospheric sulfate aerosols, and global climate. It was hypothesized that global warming would be accompanied by an increase in primary production, and biogenic production of DMS-derived sulfate aerosols leading to increased scattering, more cloud condensation nuclei (CCN), and brighter clouds. Such changes in the atmosphere's radiative budget would cool the earth's surface and thus stabilize climate against perturbations due to greenhouse warming. While phytoplankton is the protagonists in this feedback loop, recent advances in understanding the complex cycle of DMS suggest that it is the entire marine food web (Fig. 6) that determines net DMS production and not just algal taxonomy (Simo, 2001).

The emission of DMS and aerosol particle concentrations is well correlated across varying latitudes and seasons (Bates et al., 1987b). However, Schwartz (1988), in a comparison between southern (SH) and northern hemisphere (NH) cloud albedo records, argues that the CLAW hypothesis is not valid since the anthropogenically introduced sulfur aerosols in the NH should have created a noted increase in cloud albedo over the past century and none were observed. The debate on the CLAW hypothesis continues to date (Sherwood and Idso, 2003).

The DMS-climate feedback hypothesis has stimulated a very significant research effort. Several large-scale studies inspired by the International Global Atmospheric Chemistry program (IGAC) have addressed aspects

Dmsp Model
Figure 6: Conceptual model of the cycling of DMSP and DMS in the upper ocean.

of the DMS-aerosol-climate connection, including ASTEX/MAGE (Huebert et al., 1996), ACE-1 (Bates et al., 1998), and AOE-91 (Leck et al., 1996). A global database of DMS seawater concentrations and fluxes has been compiled (Kettle et al., 1999; Kettle and Andreae, 2000), and more recently, a simple empirical algorithm relating DMS seawater concentration to the oceanic MLD and surface chlorophyll concentration has also been derived (Simo and Dachs, 2002).

Notwithstanding this progress, the quantitative evaluation of the DMS-climate hypothesis remains a daunting challenge. This is due in part to the need to integrate knowledge across the traditional disciplinary boundaries of ecology, oceanography, and atmospheric science but also due to our incomplete understanding of the DMS marine production cycle.

General circulation models predict the planet's mean temperature will increase under the "business as usual'' scenario (Houghton et al., 1996). The most recent estimate of average warming for a doubling of CO2 is 3.3 7 0.8 °K (Grassl, 2000). However, there is strong spatial variation in this perturbation, with large temperature and salinity changes predicted to occur in the polar oceans (e.g., Hirst, 1999). The associated warming and salinity reduction is generally accompanied by a shallowing of the oceanic mixed layer, and stronger illumination of the upper water column, both of which can affect the food-web dynamics and consequently DMS production (Gabric et al., 2001a). It is pertinent to note that the Simo and Dachs' algorithm employs an inverse relation between MLD and DMS concentration, suggesting that DMS seawater concentration is likely to increase under global warming.

Attempts to assess the direction and magnitude of the DMS-climate feedback (Foley et al., 1991; Lawrence, 1993; Gabric et al., 2001b) indicate a small-to-moderate negative feedback on climate (stabilizing), with magnitude of order 10-30%, and considerable regional variability. The results of the use of GCM data to force a DMS model in the Antarctic Ocean under a global warming scenario suggests that significant perturbation to the DMS flux will occur at high latitudes (Gabric et al., 2003). Fully coupled climate and biogeochemistry models (Joos et al., 1999; Cox et al., 2000) are the next step in further unraveling the DMS-climate link.

0 0

Post a comment