Effects of solar radiation on dimethylsulfide cycling in the western Atlantic OceanIntroductionMaterials and methodsShipboard sample collectionFree-floating drifter arrayDeckboard incubationsDMS analysis35S methodsOptical determinationsIntegrated light dose at depthDMS sea-air fluxInterpolating measured rate constants with depthResultsDMS photolysisPhotolysis vs. total DMS consumptionBiological consumption of DMSDMS loss process comparison with depthChanges in DMS concentrations as a function of light doseDiscussionAcknowledgementsReferencesDeep-Sea Research I 53 (2006) 136–153Effects of solar radiation on dimethylsulfide cycling in thewestern Atlantic OceanD.A. Toolea,, D. Slezakb,c, R.P. Kieneb,c, D.J. Kieberd, D.A. SiegelaaInstitute for Computational Earth System Science, University of California, Santa Barbara, CA 93106, USAbDepartment of Marine Sciences, University of South Alabama, Mobile, AL 36688, USAcDauphin Island Sea Lab, Dauphin Island, AL 36528, USAdChemistry Department, College of Environmental Science and Forestry, State University of New York, Syracuse, NY 13210, USAReceived 19 November 2004; received in revised form 29 July 2005; accepted 2 September 2005Available online 9 November 2005AbstractThe influence of solar radiation on springtime rates of photochemical and biological consumption of dimethylsulfide(DMS) in surface waters from the western Atlantic Ocean was examined by exposing 0.2 mm filtered and unfiltered surfaceseawater to natural sunlight at five depths in the upper 30 m. Parallel deck incubations of 0.2 mm filtered seawater undervarious long-pass optical filters were also carried out to aid in assessing the wavelength dependence of DMS photolysis.DMS photolysis rate constants for mid-day exposure (10:30–17:30 local time) to surface irradiance ranged from 0.026 to0.086 h1and were highest in coastal and shelf waters. Photolysis rate constants decreased with increasing irradiationdepth, in accordance with the attenuation of ultraviolet radiation (UVR, 280–400 nm). Total DMS consumption rates(photochemical+biological) in unfiltered surface samples also decreased with increasing incubation depth and were largerthan photolysis rates at nearly all depths and all stations. The decrease in photolysis rate constants with exposure depthwas mirrored by biological DMS consumption rate constants that were severely inhibited at surface irradiances, andapproached or exceeded dark rate constants at deeper exposure depths. Photolysis rates were 2–19 times greater thanestimated biological consumption rates in the surface light exposed samples, while biological consumption rates weresignificantly larger than photolysis rates at incubation depths below the 1% light level for UV–B radiation (280–320 nm).Total DMS loss rates increased up to nine-fold with UVR exposure, but changes in DMS concentrations were not stronglycorrelated to light dose, presumably due to parallel, light-mediated DMS production processes. The primary loss processfor DMS depended mainly on the depth interval considered and the attenuation of UVR; in general, photochemicalremoval dominated shallow layers characterized by high UV–B intensities, whereas biological removal dominated indeeper layers where UV–B was absent, but UV–A (320–400 nm) and visible (400–700 nm) light fluxes were still relativelyhigh. These results demonstrate that UVR exposure significantly influences the spatial and temporal pattern of DMSproduction and loss processes, and ultimately the DMS flux to the atmosphere.r 2005 Elsevier Ltd. All rights reserved.Keywords: Sulfur compounds; Dimethylsulfide; Ultraviolet radiation; Photochemistry; Rate constants; Light attenuation; North Atlantic;Sargasso Sea; Gulf of MaineARTICLE IN PRESSwww.elsevier.com/locate/dsr0967-0637/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.dsr.2005.09.003Corresponding author. Present address: Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution,Woods Hole, MA 02543, USA. Tel.:+1 508 289 3552; fax: +1 508 457 2193.E-mail address: [email protected] (D.A. Toole).1. IntroductionThe biogeochemical cycling of reduced sulfurcompounds between the upper water column andthe marine atmosphere has been implicated in acloud-albedo feedback loop (e.g. Charlson et al.,1987; Shaw, 1983). This hypothesized climateregulation mechanism suggests that as upper-oceantemperatures and radiation exposure increase, mar-ine communities may respond by increasing produc-tion of the volatile organic sulfur compounddimethylsulfide (DMS) and its precursor dimethyl-sulfoniopropionate (DMSP). The atmospheric oxi-dation products of DMS are sulfate aerosols whichcan function as new cloud condensation nuclei orpromote the growth of existing condensation parti-cles thereby directly, and indirectly, reducing thesolar radiation reaching the ocean surface (Ayersand Gillett, 2000; Charlson et al., 2000). The sea-airventilation of biologically produced DMS alsocontributes to the background levels of troposphericsulfate aerosols in the remote marine atmosphere(Andreae and Crutzen, 1997), and is potentiallyinvolved in the autocatalytic activation of ozone-destroying halogens (Ayers and Gillett, 2000),reinforcing potential radiative and climatic impacts.As a result, alterations in the oceanic biogeochemicalcycling of DMS, and the resulting flux to theatmosphere, could alter the radiative properties ofthe atmosphere. Understanding how the biogeo-chemical cycling of reduced sulfur compounds, andultimately their emission to the atmosphere, varies inresponse to physical forcing such as light and mixingis critical to assessing the feedback mechanisms putforth in the climate regulation hypothesis.DMS concentrations are the product of acomplex and dynamic network of production andloss processes that span the entir e foodweb andoccur on a variety of scales including physical andchemical cycling mechanisms (e.g. Simo´, 2004).Very few studies have investigated the mechanismsinvolved in how light affects DMS cycling processesand concentrations. Most efforts have focused onphytoplankton intracellular DMS and DMSP pro-duction (e.g. Slezak and Herndl, 2003; Sunda et al.,2002; van Rijssel and Gieskes, 2002) or the impactof variations in mixed layer depth (e.g. Simo´andDachs, 2002; Simo´and Pedro´s-Alio´, 1999a). Mixedlayer DMS is removed via three primary mechan-isms: photolysis (e.g. Brimblecombe and