Fate of riverine iron over estuarine salinity gradients

Sammanfattning: Rivers have traditionally not been considered important sources of bioavailable iron (Fe) to the marine waters, since most Fe is lost by salinity-induced aggregation and sedimentation during estuarine mixing. However, recent studies from boreal rivers found a remarkably high Fe stability, probably due to the interaction with organic matter. Recent studies have suggested that Fe speciation is a key factor, i.e. that Fe (oxy)hydroxides are effectively removed by aggregation processes, but that organic Fe complexes (Fe-OM) are less affected by increasing salinity. This hypothesis has been supported by indirect assessments of Fe speciation (based on molecular size and Fe:C ratios). One aim of the current thesis was to directly assess variability in Fe speciation in boreal river waters, by X-ray absorbance spectroscopy (XAS), and to examine the link to Fe stability, which was assessed by artificial seawater mixing experiments. Another aim was to explore what factors may control temporal and spatial variability in Fe speciation within and across rivers. It has further been proposed that variation in Fe isotopic composition reflects Fe speciation, which was tested in this thesis by subjecting a set of samples to both analyses. Finally, to test how colloidal size distributions in river waters vary with regards to surface charge and Fe speciation, XAS analysis was combined with dynamic light scattering and zeta potential. The combination of these methods provided the potential for a comprehensive picture of the Fe phases present in river water and how they react to increasing salinity. The overarching aim of this thesis was to gain a better understanding of what factors determine the fate of Fe from boreal rivers across estuarine salinity gradients. For this purpose, 10 rivers, from the north to the south of Sweden, with different catchment characteristics were considered. Among the river mouths a significant, but variable contribution of Fe-OM in relation to Fe (oxy)hydroxides was detected. That Fe (oxy)hydroxides are more affected by increasing salinity than Fe-OM was confirmed by selective removal of Fe (oxy)hydroxides due to salinity-induced aggregation. Moreover, the relative contribution of Fe-OM correlated well with Fe transport capacity (FeTC), i.e. supporting that organic complexation of Fe favors Fe export to open waters. However, Fe-OM complexes were also found in the aggregated fraction, illustrating that the control of Fe stability is not simply explained by the prevalence of the respective Fe phases alone. The Fe-OM contribution was more dominant further upstream in a catchment than at the river mouth, and also more prevalent during high-flow than at low-flow conditions. Consequently, the higher contribution of Fe-OM during spring, when much of the annual Fe discharge is taking place, resulted in a higher FeTC. Separation of salinity-induced aggregates from the fraction remaining in suspension revealed that Fe (oxy)hydroxide displayed lower δ56Fe values, and Fe-OM displayed higher δ56Fe values than the in situ Fe. This points to the possibility of inferring Fe speciation of the fraction that survives the estuarine mixing zone from isotope analysis. The combination of XAS and DLS analysis demonstrated the existence of three size distributions; Fe (oxy)hydroxide were observed both as nanoparticles (10-40 nm) with positive surface charge, and larger aggregates with OM interactions (300-900 nm). An intermediate (100-200 nm) and negatively charged distribution was inferred to contain Fe-OM. After increasing salinity, the smallest Fe (oxy)hydroxide nanoparticles were no longer detected. Interestingly, both the larger size distributions were still detected at high salinity. In all, the results from this thesis support that boreal rivers may provide significant amounts of bioavailable Fe to marine waters beyond the estuary, due to significant contributions of Fe-OM complexes. Moreover, the results illustrates that a division between small Fe-OM complexes that “survive” estuarine salinity gradients and large Fe (oxy)hydroxides that are aggregated and lost to the sediment, is too simplistic, since both phases can be found in a wide size range.