Biogeochemical cycling of chromium and chromium isotopes in the sub-tropical North Atlantic Ocean

Chromium (Cr) is a redox-sensitive element and because Cr isotopes are fractionated by redox and/or biological processes, the Cr isotopic composition of ancient marine sediments may be used to infer changes in past seawater oxygenation or biological productivity. While there appears to be a 'global correlation' between the dissolved Cr concentration and Cr isotopic composition of seawater, there is ongoing debate about the relative importance of external sources and internal cycling on shaping the distribution of dissolved Cr that needs to be resolved to validate the efficacy of using Cr isotopes as a paleo proxy. Here, we present full water column depth profiles of total dissolved Cr (Cr(VI)+Cr(III)) and dissolved Cr isotopes (delta Cr-53), together with ancillary data, for three stations along a transect (GEOTRACES GApr08) across the sub-tropical North Atlantic. Concentrations of dissolved Cr ranged between 1.84 and 2.63 nmol kg(-1), and delta Cr-53 values varied from 1.06 to 1.42 parts per thousand. Although atmospheric dust, hydrothermal vents and seabed sediments have the potential to modify the distribution of Cr in the oceans, based on our observations, there is no clear evidence for substantial input of Cr from these sources in our study region although benthic inputs of Cr may be locally important in the vicinity of hydrothermal vents. Subsurface waters (below the surface mixed layer to 700 m water depth) were very slightly depleted in Cr (by up to similar to 0.4 nmol kg(-1)), and very slightly enriched in heavy Cr isotopes (by up to similar to 0.14 parts per thousand), relative to deeper waters and the lowest Cr concentrations and highest delta Cr-53 values coincided with lowest concentrations of colloidal (0.02 to 0.2 mu m size fraction) Fe. We found no direct evidence for biological uptake of dissolved Cr in the oligotrophic euphotic zone or removal of Cr in modestly oxygen depleted waters (O-2 concentrations similar to 130 mu mol kg(-1)). Rather, we suggest removal of Cr (probably in the form of Cr(III)) in subsurface waters is associated with the formation of colloid aggregates of Fe-(oxyhydr)oxides. This process is likely enhanced by the high lithogenic particle load in this region, and represents a previously unrecognized export flux of Cr. Regeneration of Cr in deeper waters leads to subtly increased levels of Cr alongside decreased delta Cr-53 values at individual sites, but this trend is more obvious at the global scale, with delta Cr-53 values decreasing with increasing radiocarbon age of deep waters, from 1.16 +/- 0.10 parts per thousand (1SD, n=11) in deep Atlantic waters to 0.77 +/- 0.10 parts per thousand (1SD, n=25) in deep Pacific waters. Removal of relatively isotopically light Cr from subsurface waters onto particulate material and regeneration of this Cr back into the dissolved phase in deep waters partly accounts for the systematic relationship between delta Cr-53 and Cr concentrations in seawater discussed by other studies.