ESTIMATING PAST SEAWATER REDOX FROM MO-ISOTOPE DATA: POTENTIAL AND PITFALLS

In recent years, interest has grown rapidly in how seawater oxygen concentrations both force and respond to ecological and environmental change, at present and in the past, with particular importance being placed on the relationship between seawater oxygenation and global warming. The motivation for this interest is clear: Oxygen solubility decreases as temperature increases, and present-day observations of diminishing seawater oxygen levels have parallels in Earth’s past, where so-called Oceanic Anoxic Events (OAEs) represent a relatively infrequent and extreme Earth system response to global warming.

The isotope composition of molybdenum (Mo) in the Earth’s crust and oceans can vary by a small but predictable degree as a result of redox and speciation processes. A picture of how the areal extent of highly reducing environments has varied over time may be obtained by analyzing well-preserved organic-rich mudrocks of known age, which can preserve the Mo-isotope composition of past seawater. A major requirement to using this proxy is that redox conditions at the time of deposition be well defined such that the effects of potential Mo-isotope fractionation during sediment accumulation can be constrained. A second important caveat is that the palaeogeographic setting of sediment deposition should be known.

We present and discuss the implications of Mo-isotope data from an interval spanning the late Paleocene and early Eocene, which encompasses the Paleocene-Eocene Thermal Maximum (PETM) c. 56 Ma ago. Our data are for contemporaneous sample suites representing both restricted basin (Arctic ocean) and open-ocean (Tethyan continental margin) environments. The results from the restricted Arctic basin reflect the Mo-isotope composition of contemporaneous seawater, whereas our data from present-day Guru-Fatima, which was formerly an open-ocean site on the Tethyan continental margin, are some 0.7 permil lower. These results are the first indication from the palaeo-record to suggest that this offset reflects a fixed fractionation between MoO₄^2- and MoO_xS_4-x^2- in anoxic and sulphidic porewaters, respectively. Furthermore, the results highlight the potential for producing incorrect estimates of the areal extent of past marine anoxia from Mo-isotope data if hydrographic conditions are not properly considered.