The fate of added iron during a mesoscale fertilisation experiment in the Southern Ocean

The first Southern Ocean Iron RElease Experiment (SOIREE) was performed during February 1999 in Antarctic waters south of Australia (61°S, 140°E), in order to verify whether iron supply controls the magnitude of phytoplankton production in this high nutrient low chlorophyll (HNLC) region. This paper describes iron distributions in the upper ocean during our 13-day site occupation, and presents a pelagic iron budget to account for the observed losses of dissolved and total iron from waters of the fertilised patch. Iron concentrations were measured underway during daily transects through the patch and in vertical profiles of the 65-m mixed layer. High internal consistency was noted between data obtained using contrasting sampling and analytical techniques. A pre-infusion survey confirmed the extremely low ambient dissolved (0.1 nM) and total (0.4 nM) iron concentrations. The initial enrichment elevated the dissolved iron concentration to 2.7 nM. Thereafter, dissolved iron was rapidly depleted inside the patch to 0.2–0.3 nM, necessitating three re-infusions. A distinct biological response was observed in iron-fertilised waters, relative to outside the patch, unequivocally confirming that iron limits phytoplankton growth rates and biomass at this site in summer. Our budget describing the fate of the added iron demonstrates that horizontal dispersion of fertilised waters (resulting in a quadrupling of the areal extent of the patch) and abiotic particle scavenging accounted for most of the decreases in iron concentrations inside the patch (31–58% and 12–49% of added iron, respectively). The magnitude of these loss processes altered towards the end of SOIREE, and on days 12–13 dissolved (1.1 nM) and total (2.3 nM) iron concentrations remained elevated compared to surrounding waters. At this time, the biogenic iron pool (0.1 nM) accounted for only 1–2% of the total added iron. Large pennate diatoms (>20 μm) and autotrophic flagellates (2–20 μm) were the dominant algal groups in the patch, taking up the added iron and representing 13% and 39% of the biogenic iron pool, respectively. Iron regeneration by grazers was tightly coupled to uptake by phytoplankton and bacteria, indicating that biological Fe cycling within the bloom was self-sustaining. A concurrent increase in the concentration of iron-binding ligands on days 11–12 probably retained dissolved iron within the mixed layer. Ocean colour satellite images in late March suggest that the bloom was still actively