The role of iron as a micronutrient in the Antarctic sea ice environment

It is now understood that the Southern Ocean is a high-nutrient, low-chlorophyll zone; production being limited by the micronutrient iron (Fe). The seasonal formation and subsequent melt of Antarctic sea ice covers an area of approximately 17 million km2, an area roughly twice the size of the Australian continent. Sea ice has the ability to store Fe at concentrations two orders of magnitude higher than in the underlying water column. Its formation can negatively influence the concentration of dissolved Fe in surface waters surrounding the continent by entrainment within the sea ice. However, during its melt, it can release this stored reserve of Fe into the underlying water column at a time that is coincidentally ideal for algal growth. During the winter, light-limited conditions prevail, however, during spring and into summer, stratification of the water column due to increased meltwater and shallowing of the mixed layer, combine with increased solar radiation, seeding of the water column with sea ice algae and the release of Fe. The result is large algal blooms that may significantly affect the regional carbon cycle with possible flow-on effects to global climate. However, considerable variability in the estimates of the size and relative contribution of these ice edge blooms exists. Fe can be sourced to Southern Ocean surface waters from sediment resuspension, aeolian dust deposition, hydrothermal plume waters or from extraterrestrial dust. However, there is growing evidence that sea-ice-entrained Fe may be one of the most bioavailable due to abundant organic complexation coupled with photo-oxidation either in situ or upon release into strongly-stratified melt waters. This work adds significantly to our understanding of the distribution and concentration of Fe in Antarctic sea ice; and is coupled with measurements of physical and biogeochemical variables. It is based on two independent research voyages into the Antarctic sea ice environment including one spatial study during the transition from winter to early spring, sampling pack and fast ice (SIPEX, Sept-Oct 2007) and one temporal study in fast ice off the coast of East Antarctica (Oct-Dec 2009). Before this thesis less than a handful of studies had sampled Antarctic sea ice in dedicated studies. In an area of 17 million km2 with characteristically high heterogeneity, this makes for an extremely patchy data set requiring more validation. Aside from adding to the global data set for Antarctic sea ice, brines, underlying sea water and snow, with measurements of dissolved, particulate and total dissolvable Fe, salinity, temperature, Chlorophyll a (Chl a), ice texture, stable oxygen isotope, exopolysaccharides, particulate and dissolved organic carbon and nitrogen and macronutrient variables, the main findings of this work are: 1) Validation that sea ice can concentrate Fe up to two orders of magnitude higher than in the underlying water column. 2) Exopolysaccharides are not directly correlated with proportional concentrations of Fe within sea ice however they can be with Chl a and particulate organic carbon. 3) Apparent dissolved Fe (dFe ‚àöv= brine volume fraction) and estimates of cellular carbon to Fe ratios suggest that during SIPEX, the sea ice microbial biota were not limited by dissolved Fe but rather may have been by nitrogen (NO2+NO3) or silicate (Si(OH)4). 4) Conversely, under-ice seawater algal communities may have been limited by dissolved Fe and/or light and grazing by zooplankton during SIPEX. 5) A significant inverse correlation between dissolved Fe and Chl a in the basal layers of pack ice, most likely indicates the active drawdown of dissolved Fe by the sea ice biota, combined with some fraction lost to the water column or converted to the particulate fraction. 6) During SIPEX, a markedly higher concentration of particulate Fe was observed at our fast ice site (0.96 ‚Äö- 214 nmol.L-1) relative to several pack ice sites (0.87 ‚Äö- 77.7 nmol.L-1) and comparison of particulate leachable Fe (plFe) with particulate Fe (pFe) indicated the highly refractory nature and therefore most likely sedimentary origin of this enrichment. 7) A high particulate-to-dissolved Fe ratio was observed at the fast ice site during SIPEX (285:1) relative to the highest observed in pack ice (23:1). This suggests a decoupling between the sources and/or sinks of the dissolved and particulate fractions, which is closely linked to the proximity of the continent and a release of shelf-derived pFe. 8) Preferential release of dissolved Fe (and not particulate Fe) into brines at all sites sampled with the sack-hole method (and therefore indicative of brine drainage), indicates the diffuse nature of the dissolved fraction. Furthermore, this indicates that there may be a temporal decoupling between the release of the dissolved and particulate fractions into the water column as sea ice becomes more permeable during the seasonal melt. 9) During a temporal study of fast ice off the coast of East Antarctica, we observed in high temporal resolution the release of each of the size fractions of Fe into the underlying water column. This information coupled with high-quality meteorological data gives valuable insight into the processes involved in the seasonal melt of sea ice and its release of Fe.