Marine sedimentary uranium to barium ratios as a potential quantitative proxy for Pleistocene bottom water oxygen concentrations

Oxygen is essential for marine ecosystems, and it is linked by respiration to carbon storage in the deep ocean. Reconstructing oxygen concentrations in the past has been limited by the absence of quantitative, rather than qualitative, proxies, but several new (semi-) quantitative oxygen proxies have recently been developed. In this study we explore the possibility of adding bulk sedimentary uranium (U) to this list by normalizing it to barium (Ba). First, U/Ba and bottom water oxygen concentrations are compared on a global scale, using a core top database, in pelagic environments greater than 200 m water depth. Then, the relationships between U/Ba and bottom water oxygen are examined on smaller spatial scales: within each ocean basin and regionally within the Eastern Equatorial Pacific, the Arabian Sea, and Western Equatorial Atlantic. At this regional scale, where secondary influences on the behavior of both U and Ba may be more spatially uniform, empirical piecewise linear calibrations are developed and subsequently tested on downcore records. U/Ba-based oxygen reconstructions generally agree with those derived from previously published alkenone preservation and benthic foraminiferal surface porosity records. Several limitations to the utility of U/Ba as a proxy for oxygen have also been identified. The proxy should only be applied in the uppermost sedimentary intervals that contain porewater sulfate to minimize barite diagenesis, and phosphorus contents should be monitored for the potential influence of apatite on uranium content. U/Ba is more successful at recording oxygen concentrations during mean glacial and interglacial periods than during climate transitions, when the timing and amplitude may be more sensitive to burndown and smoothing. Conservative errors on the calibrations result in the greatest utility of U/Ba in regions with relatively high oxygen concentrations (e.g., >50 μmol/kg) and large oxygen variability (±10 s of μmol/kg). Even with these caveats, U/Ba is only one of two quantitative oxygen proxies potentially capable of recording variability above 50 μmol/kg, and further investigation into its functionality in different environmental settings is worthwhile in the endeavor to reconstruct the full marine range of oxygen concentrations in the past.