Vertical distributions of iron-(III) complexing ligands in the Southern Ocean

Abstract

Electrochemically derived iron speciation data from four vertical profiles to 1000 m depth were obtained during the SAZ-Sense voyage to offshore waters south of Australia in summer (January/February, 2007). The dual aims of this study were firstly to devise a new operational definition to represent the ‘complexing capacity’, or total concentration of iron-complexing ligands, and subsequently derive vertical profiles of these ligand classes. Secondly, to compare the vertical trends for each ligand class with vertical distributions in oceanic properties thought to control ligand production (i.e. siderophores produced by bacteria and particle remineralisation). Based on simulated ligand titrations, we operationally defined Σ L as the overall class of ligands, which represents all iron-complexing ligands detectable under the analytical conditions chosen. The stability constant of ΣL is a weighted average for these ligands. The ligand titration data suggests the presence of an excess of iron-complexing ligands throughout the water column with an average concentration of [Σ L]=0.75±0.20 nM ( n=47), and an average stability constant of log K FeS L , Fe 3 + =21.50±0.24 ( n=47). Here, based on the range of observed stability constants we define a distinctly different class of extremely strong ligands ( L 1) to be the ligand class with a stability constant of log K Fe L 1 , Fe 3 + ≥22 , whereas Σ L ranged from 21.00 to 21.95 for log K Fe L 1 , Fe 3 + . L 1 had an average concentration and stability constant of 0.42±0.10 nM ( n=14) and 22.97±0.48 ( n=14), respectively. L 1 was only found in three of the four depth profiles, and was restricted to the upper ocean (i.e. <200 m depth), whereas Σ L was observed at all sampling depths down to 1000 m. Heterotrophic bacterial abundances (a proxy for siderophore production) were always the highest in the surface mixed layer (50–72 m depth for the 4 stations) then decreased sharply, whereas POC downward flux (a proxy for remineralisation) was greatest below the surface mixed layer then decreased exponentially. It has been suggested that siderophores control L 1 production whereas the remainder of Σ L may be set by particle breakdown (Hunter and Boyd, 2007). Hence we should expect some vertical partitioning of L 1 (present<70 m depth) and Σ L (present over the water column). However, profiles at all stations in subtropical, subantarctic, and polar waters exhibited distinguishable concentrations of L 1 to 200 m depth (i.e. straddling both regions of putative L 1 and Σ L production). There remain issues with the separation of different ligand classes, such that since [ L 1]≤ [Fe] , deeper in the water column, the concentration of L 1 cannot be resolved, and hence the provenance of both L 1 and Σ L cannot be clearly assigned.