On the nonlinear relationship between dissolved cadmium and phosphate in the modern global ocean: Could chronic iron limitation of phytoplankton growth cause the kink?

I report two vertical profiles of dissolved cadmium (Cd) and phosphate (PO4) from the Bering Sea: one from a high-nutrient, low-chlorophyll (HNLC) area, in which phytoplankton growth is limited by iron (Fe) availability, and one in highly productive waters near the continental shelf, where Fe is sufficient. At both stations, dissolved Cd and PO4 display nutrient-like profiles and are strongly correlated with depth. The surface-water dissolved Cd : PO4 ratio in the Fe-limited HNLC (0.21 6 0.03 nmol mmol21) is significantly lower than the ratio in the productive Fe-replete station (0.31 6 0.02 nmol mmol21). A simple model based on the results of previously published laboratory culture studies by others and field incubation experiments with natural phytoplankton assemblages is proposed relating the availability of Fe to the Cd : phosphorus content of phytoplankton, the dissolved Cd : PO4 of ocean surface waters, and the slope of Cd : PO4 in the nutricline. The model is consistent with available data and suggests that the nonlinearity or kink in the global dissolved Cd versus PO4 relationship exists because of chronic Fe-limiting conditions in high-latitude HNLC areas in the modern ocean. For almost 30 yr, oceanographers have been aware of the strong correlation of dissolved Cd concentrations with those of PO4 in seawater (Boyle et al. 1976; Bruland et al. 1978; de Baar et al. 1994). The relationship between dissolved cadmium (Cd) and phosphate (PO4) in the world ocean is well characterized and nonlinear, with a pronounced break in the slope, hereafter referred to as a ‘‘kink,’’ at PO4 concentrations of ,1.3 mmol L21 (Table 1; Figs. 1, 2). The correlation implies that the vertical distribution of Cd is controlled by its uptake by phytoplankton in surface waters and sinking of particulate organic matter and subsequent remineralization at depth. The strength of the correlation has been applied along with records of Cd : calcium (Ca) ratios in fossil tests of foraminifera to reconstruct past deep-water nutrient distributions in the ocean (Boyle 1988). This approach was recently expanded to include surface waters in an attempt to elucidate how changes in upper ocean nutrient inventories, reflecting the efficiency of the biological pump, might relate to glacial–interglacial climate change (Elderfield and Rickaby 2000). One of the main criticisms of the accuracy of the above paleonutrient proxy is the extent to which one can be certain that the dissolved Cd versus PO4 relationship in the ocean has remained constant in time and space (Saager and de Baar 1993; de Baar et al. 1994). The global Cd versus PO4 relationship in the modern ocean is generally described by two distinct linear relationships: one for North Atlantic Ocean data with PO4 concentrations ,1.3 mmol L21 and, after the kink, one for the Indian–Southern–Pacific Ocean data (Löscher et al. 1997). Accepting the regional difference above, the ratio of dissolved Cd to PO4 is quite constant in waters of .1,000 m; however, significant variability exists in surface-water ratios (de Baar et al. 1994; Rutgers van der Loeff et al. 1997), especially in areas in which phytoplankton growth is known to be iron limited (Martin et al. 1989, 1990). These observations have led to the suggestion that Cd might be preferentially removed from surface waters relative to PO4 during uptake by phytoplankton (Saager and de Baar 1993; Löscher et al. 1998). At present, we lack a mechanistic explanation for these surface-water deviations in the dissolved Cd : PO4 ratio and for the pronounced regional differences in Cd : phosphorus (P) cycling indicated by the kink in the global dissolved Cd versus PO4 relationship. Differences in the relative removal of dissolved Cd and PO4 by phytoplankton during their growth in surface waters could help to explain the vertical and horizontal variation of seawater Cd : PO4 ratios. Our understanding of the factors that control algal uptake of Cd and their Cd : P ratios are based largely on the results of laboratory studies with cultured phytoplankton grown in trace metal ion-buffered artificial seawater. These studies have demonstrated that the Cd content of marine phytoplankton is largely controlled by aqueous Cd ion concentrations and is inversely related to manganese (Mn) and zinc (Zn) ion concentrations (Lee and Morel 1995; Sunda and Huntsman 1998), as well as the concentration of aqueous carbon dioxide in the growth media (Lane and Morel 2000). The few field studies with natural algal assemblages tend to 1 Corresponding author ([email protected]). Acknowledgments I thank the captain and crew of the R/V Kilo Moana for their support and assistance at sea and Ken Bruland for his invitation to participate on the 2003 research cruise. Without the logistical support of the Bruland Group and Geoff Smith’s expertise at sea, this study would not have been possible. Ros Rickaby generously provided a compilation of dissolved cadmium and phosphate data that helped in preparing this manuscript. I thank Maite Maldonado and two anonymous reviewers for their comments on an earlier version of the manuscript. This work was supported by the Natural Sciences and Engineering Research Council of Canada and the Canada Foundation for Innovation. Limnol. Oceanogr., 51(3), 2006, 1369–1380 E 2006, by the American Society of Limnology and Oceanography, Inc.