Marine biogeochemical cycling of cadmium and its isotopes: Studies of the South Pacific Ocean, Mediterranean Sea and Black Sea

Cadmium (Cd) is utilised by phytoplankton in the ocean, especially under zinc (Zn)- and iron (Fe)-limited conditions, and is a crucial yet underconstrained component of the ocean’s biological pump, regulating ocean-atmosphere CO2 levels, and thereby global climate. Moreover, the depth distribution of Cd in the oceans mimics that of the major nutrient phosphate (PO4), therefore a well correlated relationship between oceanic Cd and PO4 is widely used to infer past nutrient cycling in the ocean based on the Cd/Ca ratios of foraminifera. However, spatial variabilities in the Cd-PO4 relationship question the reliability of this proxy. Therefore, an improved understanding of the biogeochemical behaviour of Cd in the oceans is needed. In this regard, the Cd stable isotope system offers the potential to provide new insights into the oceanic cycling of Cd because each process imparts a unique isotopic signature to the water column. In the last decade, improved analytical techniques have enabled the natural Cd isotopic shifts in the marine system to be unravelled in some areas of the world’s oceans. The results to date for high-Cd waters indicate that the Cd isotope system offers potential as a proxy of past productivity changes by providing further evidence for the biological importance of Cd. However, observations for low-Cd waters are limited to a few isolated data points, and additional datasets for Cd-depleted, as well as Cd-replete, waters are needed to further constrain Cd’s biogeochemical role in the oceans. In this study, paired dissolved Cd concentration and Cd isotopic measurements were performed with high-precision on waters collected during the GEOTRACES GP-13 zonal section in the South Pacific Ocean, and waters and sediments sampled during the GA04N section in the Black and Mediterranean Seas, using multiple-collector ICP-MS (MC-ICPMS) and techniques in double-spiking. Both oceanic regions are characterized by extremely low sub-picomolar (pM) concentrations of Cd across some depth gradients, requiring the analysis of sub-nanogram quantities of Cd that are 20-fold smaller than typical Cd load sizes (10-15 ng). To this end, new analytical protocols were devised, and the robustness of these analytical methods was evaluated by undertaking a series of different experiments before measuring the GP-13 and GA04N samples. Reliable Cd isotopic measurements could be obtained for samples containing as little as 0.4 ng of Cd with a precision of ± 3.5 Ɛ, while larger Cd load sizes yielded a precision of ± 0.5 Ɛ. The South Pacific subtropical gyre is the most oligotrophic gyre in the global ocean and a unique area to study the Cd isotope systematics associated with phytoplankton productivity under ultra-low nutrient concentrations, including those for Cd, that exist in this region. The Cd concentration and Cd isotopic composition were negatively correlated across the thermocline depth range, extending from 1500 to 150 m depth, due to active biological uptake in surface waters, and the preferential uptake of lighter Cd isotopes over heavier isotopes by phytoplankton, with remineralisation of Cd at depth. This process is best described by Cd isotope fractionation under open system conditions with continuous Cd replenishment and a fractionation factor of 1.0006 ± 0.0002, in contrast to closed system conditions without Cd replenishment that are typically assumed for the open ocean settings. Additionally, an unexpected positive correlation was observed between Cd concentration and Cd isotopic composition in the upper mixed layer of the South Pacific Ocean, from 150 to 15 m depth, suggesting the dominant role of mixing with Cd sourced from atmospheric deposition in these oligotrophic waters. In the surface waters of the South Pacific Ocean, the Cd isotopic signature was positively correlated with dissolved manganese (Mn) and Fe and negatively correlated with dissolved Zn as observed previously in the high-Cd Southern Ocean and suggests the influence of different uptake pathways (Mn and Zn uptake systems) on Cd uptake and Cd isotopic composition, even under ultra-low Cd levels. The permanently anoxic Black Sea is an ideal natural laboratory to study the behaviour of Cd and its isotopes under the different redox zonations that exist in the water column under oxygen-deficient conditions. In the upper oxic zone of the Black Sea, from the surface to 70 m depth, biological utilisation of Cd modulates the Cd isotope systematics. The observed Cd isotopic fractionation in this zone is defined equally well by both open and closed system fractionation conditions with fractionation factors of 1.0008 ± 0.0002 and 1.0005 ± 0.0001 respectively. In the nitrogenous zone, between 70 and 100 m depth, the Cd isotopic composition was approximately constant, despite a significant 10-fold depletion of Cd with depth, most likely due to Cd adsorption to Mn and Fe oxide particles in this sub-oxic zone. The Cd isotopic composition shifted towards heavier values below 100 m depth due to CdS precipitation in the deep sulphidic layer of the Black Sea, which occurs in the presence of free sulphide in the water column, and produced a fractionation factor of 1.0003 ± 0.0002. This implies that the Cd isotopic composition of the oxygen-lean Archean ocean, spanning the first 2 billion years of Earth’s history, may have been isotopically heavier than today, indicating that the Cd isotope system may be a valuable tracer of anoxic combined with sulphidic levels in the past oceans. Surface Cd concentrations are enriched in the Mediterranean Sea compared to the surface concentrations of the Atlantic and Pacific Oceans. The depth distribution of Cd in the northern Mediterranean Sea shows significant variations between the eastern and western Mediterranean basins (EMED and WMED, respectively). In the WMED, Cd follows a nutrient-type depth distribution. However, a better correlation of Cd with salinity compared to the major nutrient PO4 suggests that mixing is likely to be the main factor controlling the vertical Cd distribution in the WMED. In contrast, the Cd depth distribution is homogenous in the EMED and dissolved Cd does not show any significant correlation with either salinity or PO4. The similar endmember Cd concentrations of different deep water masses existing in the EMED might explain the homogenous dissolved Cd distribution. These results will inform a planned Cd isotope study of the Mediterranean Sea.