Re-assessing the surface cycling of molybdenum and rhenium

We re-evaluate the cycling of molybdenum (Mo) and rhenium (Re) in the near-surface environment. World river average Mo and Re concentrations, initially based on a handful of rivers, are calculated using 38 rivers representing five continents, and 11 of 19 large-scale drainage regions. Our new river concentration estimates are 8.0 nmol kg⁻¹ (Mo), and 16.5 pmol kg⁻¹ (Re, natural + anthropogenic). The linear relationship of dissolved Re and in global rivers (R² = 0.76) indicates labile continental Re is predominantly hosted within sulfide minerals and reduced sediments; it also provides a means of correcting for the anthropogenic contribution of Re to world rivers using independent estimates of anthropogenic sulfate. Approximately 30% of Re in global rivers is anthropogenic, yielding a pre-anthropogenic world river average of 11.2 pmol Re kg⁻¹. The potential for anthropogenic contribution is also seen in the non-negligible Re concentrations in precipitation (0.03-5.9 pmol kg⁻¹), and the nmol kg⁻¹ level Re concentrations of mine waters. The linear Mo- relationship (R² = 0.69) indicates that the predominant source of Mo to rivers is the weathering of pyrite. An anthropogenic Mo correction was not done as anthropogenically-influenced samples do not display the unambiguous metal enrichment observed for Re. Metal concentrations in high temperature hydrothermal fluids from the Manus Basin indicate that calculated end-member fluids (i.e. Mg-free) yield negative Mo and Re concentrations, showing that Mo and Re can be removed more quickly than Mg during recharge. High temperature hydrothermal fluids are unimportant sinks relative to their river sources 0.4% (Mo), and 0.1% (pre-anthropogenic Re). We calculate new seawater response times of 4.4 × 10⁵ yr (τ_Mo) and 1.3 × 10⁵ yr (τ_Re, pre-anthropogenic).