Anomalous molybdenum isotope trends in Upper Pennsylvanian euxinic facies: Significance for use of δ98Mo as a global marine redox proxy

The use of molybdenum isotope data (δ98Mo) from organic-rich shales to draw inferences concerning marine paleoredox conditions at a global scale is predicated upon the assumptions of (1) a residence time of Mo in seawater much greater than the ocean mixing time, and (2) quantitative removal of Mo from a strongly euxinic ([H2S]aq > 11 μM) water column to the sediment, thus preserving the seawater δ98Mo signature. In this study we analyze Mo isotopic variation in the Hushpuckney Shale, a 73-cm-thick unit representing the late transgressive to early regressive stages of a glacio-eustatic cyclothem (Swope Formation) deposited in the Late Pennsylvanian Midcontinent Sea (LPMS) of North America. The Hushpuckney can be subdivided into four stratigraphic zones of distinctive geochemical character. Zones I and III, which accumulated under weakly euxinic conditions, acquired relatively high δ98Mo values (+0.9 to +1.1‰), whereas Zone II, which accumulated under intensely euxinic conditions, acquired lower δ98Mo values (~+0.6‰). Zone IV, which accumulated under suboxic conditions in the water column, acquired the heaviest δ98Mo values (+1.1 to +1.8‰). These results contrast with the pattern of redox — δ98Mo covariation in modern marine environments, in which the heaviest δ98Mo values are found in the most intensely euxinic facies. luate three different hypotheses to account for the Mo isotopic patterns of the Hushpuckney Shale. One hypothesis, seawater–freshwater mixing, is rejected owing to isotopic mass balance considerations. A second hypothesis is a local control on δ98Mo by water-column redox cycling of Mn, with particulate Mn-oxyhydroxides adsorbing isotopically light Mo and transferring it to the sediment, a process that was most active during deposition of Zone II. The significance of this scenario is that euxinic black shales may not reliably record global seawater δ98Mo in areas where a Mn-particulate shuttle is operative. A third hypothesis is based on rapid secular variation of the Mo isotope composition of Late Pennsylvanian global seawater. In order to account for δ98Mo trends within the Hushpuckney Shale, seawater δ98Mo must have varied by ~ 1.2‰ at a ~ 100-kyr timescale, which would have been possible only if the residence time of Mo in Late Pennsylvanian seawater was < 100 kyr. Although both the second and third hypotheses are viable based on the present limited δ98Mo dataset, we discuss how each model might be tested through additional Mo isotope data.