Microscale C-O-S isotope signatures record microbial sulfate reduction in ancient deep subsurface pyrite–calcite–barite assemblages

Abstract. Fracture-coating pyrite from deep within the Fennoscandian Shield records the largest 34S-enrichment observed on Earth to date (δ34S = +147 ‰) – likely the result of late-stage Rayleigh distillation during closed-system microbial sulfate reduction (MSR). This implies even heavier sulfur isotope values for the complementary sulfate reservoir during pyrite formation, possibly recorded in coeval sulfate minerals such as barite [BaSO4]. However, barite has been poorly explored as an archive of ancient deep subsurface biosignatures. Here, we compiled published microscale δ34Spyrite, δ18Ocalcite, and δ13Ccalcite data with new secondary ion mass spectrometry (SIMS) δ34Sbarite and δ18Obarite analyses from two localities on the Fennoscandian Shield (Forsmark and Laxemar/Äspö, Southeastern Sweden), aiming to constrain how barite records the history of microbial processes. Comparison between the δ34Spyrite range across both localities (−53.9 to +131.7 ‰) with the δ34Sbarite range (+7.6 to +52.0 ‰) demonstrates that the sulfate reservoir corresponding to extremely 34S-enriched pyrite is not recorded in barite. We identified two groups of barite distinguished by their distinct δ18O/δ34S trends, which is proposed to record different MSR-related processes based on cogenetic δ13Ccalcite data. Although there is overlap between the metabolic processes recorded in both groups, the steeper trending Group 1 was dominantly associated with sulfate-dependent anaerobic oxidation of methane (AOM), whereas the shallower trending Group 2 was dominated by organoclastic sulfate reduction (OSR). The different trends likely resulted from an interplay of MSR pathways (AOM vs. OSR), as well as variations in sulfate reduction rates (SRR) and fractionation-related isotope enrichment (34ε), attributed to paleoenvironmental ratios of sulfate to electron donor abundances. Lower sulfate/electron donor ratios favored methanogenesis and AOM at lower SRR, whereas higher ratios inhibited methanogenesis and favored OSR at higher SRR. A preservation bias against extremely 34S-enriched barite due to undersaturation at high degrees of Rayleigh distillation likely explains its absence in the deep subsurface. Our study highlights the need for microscale multiple stable isotope signatures in fracture-hosted mineral assemblages to understand metabolic processes in the ancient deep biosphere, while stressing that these records are strongly affected by local hydrogeochemical conditions.