Sulfur isotopes as a tracer for biogenic sulfate reduction in natural environments: A link between modern and ancient ecosystems. Geologica Ultraiectina (316)

Sulfur isotopes have been widely used to trace the activity of sulfate reducing prokaryotes in modern and ancient geochemical settings and to estimate the role of this microbial metabolism in global sulfur cycling. Extensive pure culture data provide detailed insight into cellular mechanisms involved in microbial sulfate reduction. However, studies on natural communities of sulfate reducers are required for interpreting delta34S variations in sedimentary rocks, and these are much more limited. This thesis uses a flow-through reactor technique to fill this gap, and describes ranges in potential sulfate reduction rates (SRRs) and sulfur isotope fractionation effects (epsilon) for sediments sampled from four diverse modern geochemical settings including a brackish estuary (Schelde estuary, The Netherlands), a hypersaline soda lake (Mono Lake, California, USA), a shallow marine hydrothermal vent system (Vulcano, Italy) and a freshwater river (River Schelde, Belgium). Sediments were sampled as 2 cm thick slices using the flow-through reactor approach, designed to preserve the original physical, geochemical and microbial structure. The main experimental variables were incubation temperature (10, 20, 30, 40, 50, 60, 85 degree Celcius) and electron donor supply (natural substrate, lactate or acetate). Sulfate was supplied in excess. Data analysis was restricted to SSR and epsilon values obtained under steady state conditions. SRR varied from 5 to 179 nmol cm-3 h-1, with corresponding epsilon values of 5 to 43 per mil. The range in epsilon is similar to the total variability found in previous sediment incubation and pure culture experiments and is consistent with the standard isotope fractionation model of Rees (1973). Most experiments resulted in relatively small epsilon values (<20 per mil). Isotope fractionation was distinct at each sampling site and differences were most likely controlled by electron donor availability and microbial community size and structure, although salinity and cellular energetics may have also played a role. No overall clear SRR versus epsilon relationship was found for all sites, except within subsets of data. The range in epsilon was generally larger at low SRRs (