Bioactive trace Metal distributions and Biogeochemical controls in the Southern ocean

explore regions aBStract. Extensive sampling in many regions of the Southern Ocean has demonstrated that surface water concentrations of dissolved Fe are low enough to limit phytoplankton growth. In contrast, there is currently no evidence that other bioactive elements (e.g., Mn, Zn, Co) are similarly limiting. Although atmospheric input of dissolved Fe to

was that the surface waters of the Gulf of Alaska contained such low levels of dissolved Fe that this limited the ability of phytoplankton to grow in these waters and to remove the readily available macronutrients, NO 3 , PO 4 , and Si(OH) 4 . This hypothesis, initially suggested by Gran (1931) and Hart (1934), was confirmed by showing that the addition of Fe to deckboard incubations resulted in increased chlorophyll a and depletion of NO 3 and PO 4 over that in the controls. The conclusion that the biological state of this area, which had long been recognized as a High Nutrient Low Chlorophyll (HNLC) region, was due to the scarcity of dissolved Fe, and the suggestion that the Southern Ocean, another HNLC region, was similarly Fe limited provoked a lively debate within the oceanographic community (see Chisholm and Morel, 1991).
Understanding the circumstances that lead to incomplete biological utilization of macronutrients in HNLC regions had been of great interest to oceanographers for some time for several reasons. One of the most pressing topical reasons was the relationship between HNLC regions and the global carbon cycle. The "fixing" of oceanic dissolved CO 2 into particulate organic carbon by phytoplankton in oceanic surface waters and its subsequent export via vertical particle settling results in a net transfer of carbon from the surface to the deep ocean. Because CO 2 in the atmosphere and dissolved CO 2 in the surface ocean are constantly equilibrating, this removal of surface carbon by the "biological pump" has the net effect of reducing atmospheric CO 2 levels. Thus, it had been recognized that the excess macronutrients in Southern Ocean surface waters represented a large inefficiency in the ocean's nutrient cycling, limiting the ocean's ability to take up atmospheric CO 2 . This realization led Martin (1990) to postulate the "glacial Fe hypothesis, " which explicitly suggested that enhanced Fe delivery to Southern Ocean surface waters as a result of increased atmospheric dust loads during the last glacial maximum could have been responsible for a 2-4 Gt C yr -1 increase in primary productivity, resulting in lowered atmospheric CO 2 . Martin's calculation that the addition of Fe to contemporary Southern Ocean surface waters could support the removal of 3 Gt C yr -1 provoked immediate interest among those who were seeking CO 2 sinks to ameliorate the effects of anthropogenic CO 2 production (see Chisholm and Morel, 1991 (Schaule and Patterson, 1977).
Although obtaining only one sample per deployment, this sampler showed, for the first time, that the vertical distributions of contamination-prone trace species were not highly variable as had been previously thought, but had "oceanographically consistent" distributions (Sclater et al., 1976).    (Hunter et al., 1996), a technique later extended to fully instrumented commercial rosettes . The results from US JGOFS work (Measures and Vink, 2001;Coale et al., 2005) showed that, despite the ~ 0.1 nM difference between the two data sets, there was general agreement that Fe concentra-     (Wedepohl, 1995). Duce et al. (1991) showed atmospheric deposition values ranging from > 20 g mineral dust m -2 yr -1 in regions near the Sahara to less than 0.1 g m -2 yr -1 over the remote Southern Ocean, which has little or no nearby ice-free land masses to generate mineral aerosols. However, these estimates were made using the mass of aerosols collected on sampling towers situated on land that were then extrapolated out over the ocean. In other words, there were no actual data from the ocean, and regional coverage was limited in oceanic areas without landmasses. Although there are now sampling programs that regularly collect aerosol samples from research vessels , the short duration of oceanic cruises means that these data represent only brief time periods, and thus, given the highly sporadic nature of dust deposition events, it is hard to generate reliable annual estimates from these extrapolated values. An alternate approach by Measures and Brown (1996) uses We convert this number to an estimate of annual dissolved Fe deposition by assuming the Fe content of mineral dust is equivalent to average crustal material (1 mmol g -1 Fe by weight; Wedepohl, 1995) and that 5% of this material will dissolve in the surface waters (Figure 3).

Fe iNput FroM Sea ice aNd iceBergS
Observation of blooms and elevated Fe levels near retreating ice edges has led several workers to suggest that the input of iron derived from melting sea ice may be an important factor in their development (Martin, 1990;Sedwick and DiTullio, 1997). However, melting sea ice also promotes water-column stratification and mixed-layer shoaling, which are also necessary preconditions for the initiation of spring blooms. It is now well established that the concentration of Fe in sea ice and icebergs may be up to two orders of magnitude higher than that measured in surrounding Southern Ocean waters, so that ice melt may constitute a non-negligible source of Fe into Southern Ocean surface waters in austral spring and summer (Lannuzel et al., 2007;Smith et al., 2007). However, estimating the relative importance of ice-derived Fe in the Southern Ocean is difficult because the concentration of Fe measured in sea ice and icebergs appears to be extremely variable from place to place, within the ice itself and also across seasons. For example, Fe in icebergs may vary from 4 to 600 nM (Lin et al., 2011), and Fe in sea ice was found to vary seasonally between 2 nM in summer and 18 nM in winter .

SuSpeNded SediMeNtS aS a Source oF diSSolVed Fe
The occurrence of anomalously productive regions within the HNLC Southern Ocean, notably downstream of islands and the Antarctic Peninsula, suggests that natural processes may be adding Fe to these regions (Sullivan et al., 1993). expeditions employed a multitracer approach using rare earth elements (Zhang et al., 2008) and short-lived radium isotope activities van Beek et al., 2008). These data suggest that the Fe enrichment in both regions mainly results from the lateral advection of water masses that were in contact with shallow lithogenic sediments of the island shelf systems rather than vertical or atmospheric inputs.
Recent data from the Drake Passage region (Hatta et al., 2008) (Morel et al., 1994) and that Mn limitation, albeit less clear, was detected at Mn levels less than 0.1 nM (Brand et al., 1983). These results, however, were never unequivocally confirmed during bottle incubation experiments of ambient Southern Ocean water subjected to enrichments of Zn (Scharek et al., 1997;Coale et al., 2003;Ellwood, 2004) or Mn (Martin et al., 1990a, Buma et al., 1991Scharek et al., 1997;Sedwick et al., 2000). In retrospect, the lack of evidence for Zn limitation of Southern Ocean waters during incubation experiments is not surprising because dissolved Zn in Southern Ocean surface waters is two to three orders of magnitude higher than the free Zn 2+ concentration (~ 2 pM) that limited growth of cultured coastal diatoms (Martin et al., 1990b;Morel et al., 1994;Sanudo-Wilhelmy et al., 2002;Coale et al., 2005;Ellwood, 2008;Croot et al., 2011). In addition, the most comprehensive Zn speciation data set from the Atlantic sector of the Southern Ocean indicates that most dissolved Zn in this region is labile, with free Zn 2+ concentrations greater than 0.1 nM   (Martin et al., 1990b;Coale et al., 2005;Ellwood, 2008;Croot et al., 2011). Dissolved Zn and Si are generally strongly correlated in the Southern Ocean, even though a physiological mechanism connecting Si and Zn uptake in marine diatoms remains elusive (Ellwood, 2008;Croot et al., 2011). However, given that diatoms are a dominant taxonomic group in the Southern Ocean, they potentially dictate the cycling of Zn in these waters (Ellwood, 2008).

Dissolved Mn concentrations
in the Southern Ocean range from 0.04-0.83 nM (Westerlund and Öhman, 1991;Sedwick et al., , 2000Middag et al., 2011), with values greater than 1 nM near productive shelves (Martin et al., 1990b) and in the vicinity of hydrothermal vent systems . Although dissolved Mn profiles do suggest biological uptake and regeneration in some Southern Ocean regions (Sedwick et al., , 2000Middag et al., 2011), there is still no convincing evidence of algal growth limitation in Southern Ocean waters due to Mn deficiency alone.
The biogeochemical cycle of dissolved Co is linked to that of Zn and Cd, because Co is capable of substituting for these metals in some carbonic anhydrase enzymes (Saito et al., 2010).
Published dissolved Co measurements in the Subantarctic Zone southwest of New Zealand (Ellwood, 2008) in the Ross Sea (Saito et al., 2010) and in the Atlantic sector of the Southern Ocean (Martin et al., 1990b;Bown et al., 2011) range from 5-85 pM. In the Ross Sea, Saito et al. (2010)  showed that the addition of Co alone did not yield significant phytoplankton growth over controls from the Ross Sea (Bertrand et al., 2007). However, it has been shown that the addition of Fe and the Co-containing vitamin B 12 together can enhance phytoplankton growth and alter community composition compared to Fe-only amendments (Bertrand et al., 2007). Vitamin B 12 is produced by specific prokaryotes and is required by most eukaryotic phytoplankton species (Bertrand et al., 2007)

Future reSearch
Mesoscale artificial Fe fertilization experiments have unequivocally confirmed that primary productivity in various HNLC regions is Fe limited (Martin et al., 1994;Boyd et al., 2004;Coale et al., 2004). However, the long-term impacts of such perturbations on ecosystem structure and whether they can export carbon to the ocean's interior have been difficult to ascertain due to the limited observational period that followed the initiation of the blooms. In addition, the intermittent infusion approach using acidified FeSO 4 that had to be adopted in these experiments to create and sustain the blooms is undoubtedly a limited analog of the large-scale Fe fertilization of the Southern Ocean that may have contributed to the reduced atmospheric CO 2 levels recorded at glacial maxima.