The chemical and isotopic composition of sedimentary sulfide and reactive iron are widely used parameters for the discrimination of oxic versus euxinic water column conditions in the geological past. Here we test the applicability of this proxy in a systematic study of organic-rich shelf and slope sediments from the Benguela upwelling system. These sediments deposit under (1) sulfidic, (2) anoxic, but non-sulfidic, and (3) fully oxic conditions, but below an oxygen minimum zone. Due to the high accumulation rates of organic matter, sulfate reduction and formation of sedimentary sulfides are observed at the sediment-water interface in all three environments. We find that the overall isotope difference between dissolved sulfate and pyrite increases from 25 ‰ in sulfidic to 45 ‰ in anoxic to 65 ‰ in fully oxic bottom waters. These net isotope fractionations do not reflect the isotopic fractionation by sulfate-reducing bacteria alone and thus cannot be used to extract information on sulfate reduction rates. Reaction transport modeling indicates that in the case of the anoxic, but non-sulfidic bottom waters, the isotope fractionation by sulfate-reducing bacteria substantially exceeds the net effect of 45 ‰, and may be as high as 85 ‰. The isotope effect by sulfate-reducing bacteria, however, is reduced by superposition of normal isotope effects during the anaerobic oxidation of sulfide with nitrate, by the large sulfur-storing bacteria Beggiatoa and Thiomargarita sp., which may be as high 45 ‰.
The 34S-enrichment of dissolved sulfide in the zone of anaerobic methane oxidation varies between +14 and +19 ‰ vs. CDT and depends on the isotopic composition of the initial hydrogen sulfide formed at the sediment-water interface rather than the depth of the transition zone. Diagenetic overprinting by continuing sedimentary sulfide precipitation during burial is negligible in these sediments due to general iron limitation. Nevertheless, the isotopic composition of sulfides formed at the sediment-water interface is not preserved because of isotopic exchange of a fraction of the pyrite pool with coexisting dissolved sulfide. In these sediments, this exchange leads to 34S-enrichment in pyrite of about 15 ‰ relative to the initial isotope composition. Our observations indicate that buried sediments never preserve the initial isotope composition of sulfides formed at the sediment-water interface or in the water column. However, the isotopic imprints of oxic, anoxic, and sulfidic bottom water conditions are sufficiently distinct from each other that they remain preserved as relative isotopic differences in the isotopic composition of buried sedimentary sulfides.