The evolution of the Earth surface sulfur reservoir

Abstract

The surface sulfur reservoir is in intimate contact with the mantle. Over long time scales, exchange with the mantle has influenced the surface reservoir size and possibly its isotopic composition. Processes delivering sulfur to the Earth surface from the mantle include volcanic outgassing, hydrothermal input, and ocean crust weathering. The sulfide fixed in ocean crust as a consequence of hydrothermal sulfate reduction, and subduction of sedimentary sulfides, represent return pathways of sulfur to the mantle. The importance of these different pathways in influencing the size of the surface sulfur reservoir depends on the particulars of ocean and atmosphere chemistry. During times of banded iron formation when the oceans contained dissolved iron, sulfide from submarine hydrothermal activity was precipitated on the seafloor and subsequently subducted back into the mantle and, therefore, had little impact on the surface sulfur reservoir size. With sulfidic ocean bottom water conditions, which may have occurred through long stretches of the Mesoproterozoic and Neoproterozoic, significant amounts of sulfide is subducted into the mantle. When the oceans are oxic, sulfide subduction is unimportant, and an additional source, ocean crust weathering, delivers sulfur to the Earth surface. Thus, under oxic conditions the surface environment accumulates sulfur, and probably has for most of the last 700 million years. Mass balance modeling suggests that the surface sulfur reservoir may have peaked in size in the early Mesoproterozoic, declined to a minimum in the Neoproterozoic, and increased to its present size through the Phanerozoic. The exchange of sulfur between the mantle and the surface environment can also influence the isotopic composition of the surface reservoir. Modeling shows that the subduction of 34S-depleted sulfur through the Mesoproterzoic could have significantly increased the average δ34S of the surface reservoir into the late Neoproterozoic. The preserved isotope record through the Neoproterozoic is well out of balance, with the average δ34S for sulfate and sulfide both exceeding the modern crustal average. This imbalance could be explained, at least partly, if the crustal average was more 34S-enriched than at present, as the modeling presented here suggests.