Résumé
It has been known for some time that sulfate-reducing organisms preferentially process light over heavy sulfur isotopes (32S relative to 34S) leading to H2S and sulfide minerals such as pyrite (FeS2) that are depleted in heavy sulfur isotopes (negative delta 34S). The isotopic difference between pyrites and contemporaneous sulphate-minerals (e.g. gypsum, CaSO4 or barite, BaSO4) in sediments can therefore be used to trace microbial sulfate-reduction over time, down to 2,500 Myr ago. In rocks older than ca. 2,500 Myr, a much smaller isotopic difference is generally preserved, which could either indicate that microbial sulfate-reduction was absent in the Early Archean or that microbial sulfate-reduction operated under low seawater sulfate concentration. An exception are the strongly negative delta 34S values of microscopic sulfides hosted in barite of the 3,490 million year old chert-barite deposit at North Pole (Dresser Formation, Western Australia). On the basis of this finding it has been suggested that the record of microbial sulfate-reduction extends into the Early Archean. We show here that the microscopic sulfides at North Pole have a mass independently-fractionated sulfur isotopic anomaly (Delta 33S) that differs from that of their host sulfate (barite, Philippot et al., 2007, Science, 317, 1534-1537). These microscopic sulfides cannot have been produced by sulfate reducing microbes, nor by abiologic processes that involve reduction of sulfate. Instead, we interpret the combined negative delta 34S and positive Delta 33S signature of these microscopic sulfides as evidence for the early existence of organisms that disproportionate elemental sulfur. Recognition that the microscopic sulfides show systematic positive Delta 33S anomalies indicates that the source of sulfur used by elemental sulfur disproportionators was almost exclusively derived from the atmosphere.