Role of iron-binding organic ligands in the distribution of dissolved iron in Antarctic sea ice

The availability of iron (Fe) is decisive for biochemical reactions involved in marine primary productivity and atmospheric carbon dioxide drawdown. Low Fe solubility and paucity of Fe sources lead to Fe limitation in Antarctic surface waters, which strongly constrains phytoplankton growth. This limitation is seasonally alleviated when sea ice melts, as sea ice is generally enriched in Fe compared to seawater. This natural fertilization event benefits both sympagic ice algae and pelagic phytoplankton. In seawater, the concentration of dissolved Fe (DFe) is controlled by iron-binding organic ligands (L), yet sea-ice environment is comparatively understudied. The first part of this thesis (Chapter 2) aimed to investigate how salinity and temperature may affect the physicochemical detection of L. The experimental design offered a comparison between artificial ligands used in sea-ice samples (1-nitroso-2-naphthol or NN) and in seawater samples (salicylaldoxime or SA, and 2-(2-thiazolylazo)-p-cresol or TAC), in order to define: 1) which artificial ligand is more appropriate for the determination of L in the sea-ice environment; 2) the fertilization potential of sea ice, with respect to L, allowing the comparison between sea-ice and seawater data. Within the salinity range considered (1 < S < 90), only SA and NN were successfully calibrated, whereas conditional stability constants were not achieved with TAC outside the 21 < S < 35 range. When titrating natural samples, only SA was able to detect DFe organic speciation parameters along the salinity range considered (5 < S < 78). The results, therefore, suggest that SA is the most suitable artificial ligand for the investigation of L in sea ice. In addition, a second experiment was performed, to understand if the common practice of titrating samples at room temperature, instead of at in-situ conditions, can affect the determination of the DFe complexing parameters. The titration of natural samples with both NN and SA at different temperatures (4 ºC and 20 ºC) showed no significant difference with respect to the DFe organic speciation parameters. However, Fe lability at 4 ºC was 40 ± 20% lower than at 20 ºC, which could be even lower in environments characterized by temperature lower than 4 ºC (i.e. polar oceans). The second part of this thesis (chapters 3 and 4) offered an ecological investigation of ligands concentration and potential drivers, across four Antarctic sea-ice datasets. Samples were collected in the Eastern (Mawson and Davis Seas) and Western (Weddell Sea) sectors of Antarctica, allowing for the first pan-Antarctic study on DFe organic complexation. The seasonal heterogeneity (winter, early and late spring, summer) of the 137 datapoints, and the suite of ancillary physical and biogeochemical data, allowed an illustration of the L cycle in sea ice. Results showed that DFe is > 99% organically complexed throughout the cycle of sea-ice formation and melt. Significant correlations between DFe and L were only observed from spring onwards, suggesting that different drivers played a role in the distribution of L throughout the year. Sea-ice physics and incorporation of organic matter of different origin (i.e. biological, upwelled, photochemical) led the autumn-winter L distribution, with ligands likely of humic-like composition, given the average conditional stability constant (11.5 ± 0.3). The winter-spring transition is strongly thought to mark a turning point for sea-ice L distribution. Biomass increase led to the highest L concentration (up to 74.6 nM) matching the bloom stage recorded at Davis. This fresh organic material was represented by algal-mediated exopolymeric substances (EPS), due to the average logK'Fe'L of 12.1 ± 0.5. With time and approaching summer, this organic material became more refractory, decreasing in the average logK' value to 11.8±0.7. Results would support the recently developed theory that sea-ice organisms produce and live within a biofilm matrix, made of EPS, which helps retain macro and micronutrients, and increases the solid fraction of sea ice that organisms can adhere to when sea ice melts in summer. In conclusion, the DFe distribution in sea ice is controlled by its organic complexation. The development of sea-ice microbial communities appears critical for the in-situ production first and remineralization after, of organic matter. These two pathways are equally important, as they provide inorganic and smaller organic material for 1) sea-ice organisms and 2) cryopelagic species in the seawater below. In view of the predicted reduction in sea-ice extent and duration, it is necessary to better understand the potential alteration of the hypothesised L cycle in sea ice, and how marine primary producers and related Southern Ocean food web may be affected