Iron organic speciation during the LOHAFEX experiment: Iron ligands release under biomass control by copepod grazing

Abstract

The LOHAFEX iron fertilization experiment consisted in the fertilization of the closed core of a cyclonic eddy located south of the Antarctic Polar Front in the Atlantic sector of the Southern Ocean. This eddy was characterized by high nitrate and low silicate concentrations. Despite a 2.5 fold increase of the chlorophyll-a (Chl-a) concentrations, the composition of the biological community did not change. Phytoplankton biomass was mostly formed by small autotrophic flagellates whereas zooplankton biomass was mostly comprised by the large copepod Calanus simillimus. Efficient recycling of copepod fecal pellets (the main component of the downward flux of organic matter) in the upper 100-150 m of the water column prevented any significant deep export of particulate organic carbon (POC). Before fertilization, dissolved iron (DFe) concentrations in the upper 200 m were low, but not depleted, at ~0.2 nM. High DFe concentrations appeared scattered from day 14 onwards as a result of the grazing activity. A second fertilization on day 21 had no significant effect on the DFe and Chl-a standing stocks. Work with unfiltered samples using different acidification protocols revealed that, by midway of LOHAFEX, rapid recycling of iron-replenished copepod fecal pellets explained the source of bioavailable iron that prolonged the duration of the bloom for many weeks. Here we present the evolution of the organic speciation of iron in the upper 200 m of the water column during LOHAFEX by a Competing Ligand Equilibrium method using voltammetry. During the first 12 days of the experiment, ligands of an affinity for iron similar to the ligands found before fertilization (logK'Fe'L~11.9) accumulated in fertilized waters mostly in the upper 80 m (from ~1 nM to ~2.5 nM). The restriction of ligand accumulation to the depth of Chl-a penetration points to exudation by the growing autotrophic population as the initial source of ligands. From day 5 onwards, we found in many samples a new class of ligands (L1) characterized by a significant higher conditional stability constant than the background complexation (logK'Fe'L1~12.9). During the middle section of the experiment (days 12 to 25) the accumulation of overall ligands and specifically L1, reached an upper limit in surface waters (at ~3 nM). Overall ligands and L1 accumulation was also observed below the mixed layer depth indicating that grazing was the process behind ligand release. During the last 10 days of the experiment liga�