In contrast to Fe and Mn, concentrations of other essential trace metals such as Co, Cu, Ni and Zn in Antarctic seawater are usually higher or in the same range as those reported for other oceans (Tables 9 and 10). Martin et al. (1990)
found that concentrations of dissolved Co and Cu in the Drake Passage showed a minimal surface depletion when compared to profiles of the two metals in the Gulf of Alaska. In February and March 1991, concentrations of dissolved Ag, Cd, Cu, Fe, Ni and Zn in surface waters of the Weddell Sea shelf were higher than in surface waters of other marine environments (Sanudo-Wilhelmy et al. 2002). The results of this study suggest that elements with a strong coastal component, such as Ag, Cd, Co, Cu, Ni and Zn, may be exported from the Antarctic Peninsula to the open Weddell Sea. Furthermore, the inverse relationship between chlorophyll a concentrations and trace metal residence times suggests the importance of biological activity in cycling bioactive metals. Westerlund and Ohman (1991b) found that in the Weddell Sea, in contrast to other oceans, concentrations of trace metals are high throughout the whole water column. In general, Cd, Cu and Zn concentrations were slightly lower in surface waters, Ni only showed very small variations while Co concentrations clearly increased in surface waters. The melting of ice was considered responsible for the increase in Co concentrations, as comparisons between the elemental composition of Antarctic snow and seawater revealed that snow was enriched in Co and Pb, concentrations of Zn were similar, while those of Cd and Ni were higher in seawater. Tables 9 and 10 show that, except for higher values in the Gerlache Strait, concentrations of Co in water samples from different depths in neritic and offshore areas of the Southern Ocean are rather constant (about 26 pmol l1). However, Sanudo-Wilhelmy et al. (2002) reported median Co levels of 51 pmol l1 in surface waters from the Weddell Sea shelf. The higher values were attributed to input of Co from diagenetic processes occurring in shelf sediments, and to the mixing of upwelling waters with local shelf/upper slope waters.
Cobalt is essential for several species of phytoplankton, but its uptake is generally minimal and the element can be replaced by chemically similar metals such as Zn. Through laboratory experiments, Sunda and Huntsman (1995) found that the cyanobacterium Synechococcus bacillaris needs Co but not Zn for growth, the coccolithophore Emiliana huxleyi has a Co requirement which can be partly met by Zn, and diatoms of the genus Thalassiosira have Zn requirements which can be largely met by Co. The replacement of Co by Zn could explain why Antarctic surface waters are not usually remarkably depleted in Co. By reviewing literature data on Zn and Co concentrations in ocean surface waters, Sunda and Huntsman (1995) noted that Co depletion occurs only after Zn depletion (i.e. when Zn concentrations fall below 0.3 nmol l1).
Like Co concentrations, those of Cu and Ni in Southern Ocean surface waters (Table 9) are usually higher than typical ocean surface values (usually about 0.5 and 2 nmol l1 respectively; Bruland 1980), and they do not show depletion near the surface. Copper is one of the most studied metals in oceans because it is essential to life. However, if environmental bioavailability of Cu increases (e.g. in recently upwelled waters), it can become toxic, because an excessive absorption of the metal may inhibit enzymes which require other metallic elements. The behaviour of Cu in the marine environment is intermediate to that of nutrients (which are depleted in surface waters, and accumulate in deep and older waters) and scavenged elements (such as Al and Pb, which mostly derive from sources external to the ocean and show higher concentrations in surface and younger waters than in deep or older waters). The bioavailability and geochemical behaviour of Cu in seawater are largely affected by the formation of complexes with organic ligands. Capodaglio et al. (1994) studied Cu complexation in surface waters from Terra Nova Bay during three summer expeditions. Total Cu concentrations ranged between 0.5 and 4.8 nmol l-1, and showed a uniform spatial distribution and scarce variations during the three campaigns. Like in many other marine areas, two classes of Cu ligands were found: a stronger one with considerable spatio-temporal variations, and a weaker one with a rather homogenous distribution in the study area.
Sanudo-Wilhelmy et al. (2002) found that the Ag/Cu ratio in the Weddell Sea corresponds to that reported for the Pacific Ocean, and hypothesised that Weddell Sea surface waters may influence trace metal compositions in Pacific subsurface waters. Furthermore, Westerlund and Ohman (1991b) found that the Weddell Sea and the Pacific Ocean have similar nutrient/trace metal ratios. Indeed, as discussed in Chapter 3, intermediate/deep waters of the world ocean mainly originate in the Weddell Sea.
As in the case of Cu, spatial and vertical variations in Ni concentrations in different regions of the Southern Ocean are rather small and these are generally not significantly linked to nutrient concentrations. Nolting and de Baar (1994) measured higher Ni concentrations in surface waters from the Scotia Sea and Antarctic Confluence (Table 9). They hypothesised two possible sources of Ni: uptake from marginal sediments by waters flowing through the Gerlache Strait, and/or transport by deep Pacific waters, whose Ni contents are in the same range as those measured in the Scotia Sea.
Zinc is one of the most important essential micronutrients, because it is a component of nearly 300 enzymes and is involved in many metabolic processes (Vallee and Auld 1990). Morel et al. (1994) showed that carbon uptake by marine phytoplankton may be limited by Zn, which is a constituent of the metalloenzyme carbonic anhydrase. Because of such bioactivity, open-ocean surface waters typically have low Zn concentrations (0.1 nmol l-1; Bruland and Franks 1983), which are usually strongly correlated with those of Si and N (Bruland et al. 1991). It has therefore been hypothesised that Zn availability, like that of Fe, may limit oceanic primary productivity (Morel et al. 1994). Furthermore, while much Fe limitation is probably Fe-N co-limitation, that of Zn is probably Zn-C co-limitation (i.e. Zn may have a larger impact than Fe on the global C cycle).
Phytoplankton organisms have adopted different strategies to grow in environments with limited Zn availability. Coccolithophores reduce their Zn requirement and carbonic anhydrase activity by using CO2 formed from
HCO3- through calcification reaction. This makes the biological carbon pump particularly inefficient, because the export of CaCO3 regenerates CO2. On the contrary, other marine primary producers, such as some species of diatoms, can replace Zn in enzymatic sites with other chemically similar metals such as Cd and Co (Price and Morel 1990).
Biological uptake of Zn, depletion of its concentrations in surface waters at sub-nanomolar concentrations, and statistically significant relationships between Zn and Si concentrations have been reported in some highly productive regions of the Southern Ocean (e.g. Orren and Monteiro 1985; Nolting and de Baar 1994; Fitzwater et al. 2000). Zinc is known to be incorporated into diatom frustules (Ellwood and Hunter 1999), which carpet the seafloor of the Southern Ocean. The downward flux of diatoms, also called the "silicate pump", might be the main mechanism for transferring Zn and other trace metals from surface waters to bottom water. During the sedimentation of diatoms, trace metals can be scavenged on the outer surface of frustules; both trace metals and Si are released when the diatoms dissolve. However, relationships between trace metals and Si often show different slopes and intercepts in different regions and/or at different depths, indicating rather complex interactions and biogeochemical cycles. In some highly productive areas of the Ross Sea shelf, Fitzwater et al. (2000) found remarkable surface depletions of Zn (<1 nmol l-1), Co and Cd, while Si concentrations remained >30 |mol l-1. In areas where soluble Fe concentrations were also high, the relative depletion of Zn, Cd and Co in surface waters may indicate Zn stress or limitation.
6.2.3 The "Cadmium Anomaly" in the Southern Ocean
Among trace metals, Cd is one of the most studied in seawater profiles, and its global marine biogeochemistry is reasonably well understood. The distribution of Cd in oceans is commonly strongly correlated with phosphate concentrations. The Cd/PO4 ratio generally exhibits a linear relationship, which is important in understanding biogeochemical processes occurring in modern and ancient waters. Boyle (1988) plotted available high-quality datasets and suggested that the relationship between Cd and PO4 concentrations worldwide can be described through a general equation, consisting of two separate relations: one for phosphate concentrations<1.3 |mol l-1 (with virtually zero intercept, in all upper ocean waters and North Atlantic deep waters), and one for concentrations>1.3 |mol l-1 (mostly North Pacific and South Atlantic deep waters). According to the latter relationship, for a given value of PO4 in deep open-ocean waters, Cd concentrations are consistent within about ±7%.
More recently, de Baar et al. (1994) compiled a larger dataset of selected high-quality Cd and phosphate values for deep waters, revealing two data clusters with a statistically significant bimodality: deep North Atlantic versus deep Antarctic/Indo/Pacific waters. The two distinct biogeochemical provinces for Cd cycling were attributed to the scarce input of Cd- and phosphate-rich waters from the Weddell Sea to the deep North Atlantic. Evidence supporting the two Cd/PO4 relationships was also provided by Yeats et al. (1995) and Frew (1995). The latter author attributed the apparent global Cd-phosphate link to the formation of bottom waters with high Cd/PO4 ratios around the Antarctic continent. However, the question of whether the global Cd/PO4 ratio is better described by one or two different linear relationships is still a matter of debate. By adopting the same selection criterion and adding new data to the de Baar and co-workers selection, Löscher et al. (1997) found statistically significant differences between the North Atlantic and Indo-Pacific Oceans. However, when only two datasets with the smallest phosphate concentrations (PO4<1 |mol l-1) were excluded, the differences were no longer statistically significant. It follows that more accurate Cd and phosphate data are needed, especially in areas with low concentrations.
An important feature of Cd marine biogeochemistry is that the metal is incorporated into CaCO3 exoskeletons of marine organisms in concentrations reflecting those of the seawater in which the organism is growing (Boyle 1992). The Cd/Ca ratio in benthic shells indicates Cd concentrations in waters and those of phosphate, via the Cd/PO4 relationship. The biogeochemistry of Cd has therefore been used to study labile nutrient concentrations in oceans and to infer the palaeochemistry and palaeocirculation of bottom waters (e.g. Shen et al. 1987; Boyle 1992). However, Mackensen and Douglas (1989) showed that when dealing with several species of foraminifers, only one truly epibenthic species should be used to study bottom water palaeochemistry. The interpretation of results is complicated by the decoupling of Cd and phosphate during early diagenesis (Saager et al. 1992). In the Southern Ocean, for instance, 613C data suggest that deep waters were nutrient-enriched in the glacial period, whereas Cd data indicate only minor changes (Boyle 1992). According to several researchers (e.g. Hutchins and Bruland 1998; Takeda 1998), the use of opal accumulation rates as proxies to estimate past levels of productivity in the Southern Ocean may be biased. Indeed, in this ocean there is no evidence of higher opal accumulation rates during glacial eras, when the enhanced deposition of dust increased phytoplankton productivity (Boyle 1998). A better knowledge of Cd and 613C biogeochemistry in surface and deep Antarctic waters, particularly in areas where deep waters form, is probably necessary to clarify discrepancies and to gain a better understanding of relationships between ocean chemistry and circulation, and of atmospheric CO2 concentrations (Frew 1995).
Southern Ocean surface waters are exceptional because Cd (see Tables 9 and 10) and phosphate (about 2.0 |mol l-1) concentrations in these waters are much higher than in other ocean surface waters (often strongly depleted to levels as low as 0.01 nmol l-1 for Cd and less than 0.04 |mol kg-1 for phosphate;
e.g. Yeats and Campbell 1983). Furthermore, while maximum Cd concentrations in Pacific and North Atlantic waters are found at depths of about 1,000 m, maximum values in Antarctic waters are already reached at depths of 100-300 m (Table 9). The Cd/PO4 ratio in the Antarctic Circumpolar Current and in the northern portion and western rim of the Weddell Sea ranges from 0.50 to 0.58 (nmol l-1/|mol l-1; Nolting et al. 1991; Löscher et al. 1998; Sañudo-Wilhelmy et al. 2002). However, Westerlund and Öhman (1991b) reported a much lower ratio (0.31) for the southern portion of the Weddell Sea. In the upper 300 m of a transect in the Southern Ocean, Nolting and de Baar (1994) found different Cd/PO4 ratios in each of the studied areas: the Scotia Sea, Antarctic Confluence and Weddell Sea. These results suggest that within the Southern Ocean, which is a mosaic of marine subsystems, the Cd/PO4 ratio is susceptible to spatio-temporal variations likely due to phytoplankton growth in surface waters. Phytoplankton assemblages may be dominated by different species of diatoms or by Phaeocystis antarctica, which have species-specific macro- and micronutrient requirements. As discussed by Arrigo et al. (1999) for the N/P ratio, spatio-temporal differences in phytoplankton composition may contribute to variations in the Cd/PO4 ratio of Southern Ocean surface waters.
Scarponi et al. (2000) studied the summer evolution of Cd distribution in the water column at Wood Bay (Ross Sea) and found that in November, before ice melting and phytoplankton blooms, the vertical profile of Cd along the water column was characterised by uniform concentrations (mean 0.64 nmol l-1). These relatively high and uniform Cd concentrations were ascribed to the upwelling or vertical diffusion of Cd-enriched deep waters and to the scarce uptake or scavenging of this metal by organisms. Later in the season, with the onset of ice melting and the development of phytoplankton, Cd was strongly depleted in surface water to a depth of 50 m (minimum concentration 0.056 nmol l-1 at -10 m, at the end of January). In February, as a consequence of reduced phytoplankton uptake and shallow regeneration cycles, Cd concentrations in the upper 50 m began to increase.
Was this article helpful?