Description: An in situ iron enrichment experiment was carried out in the Southern Ocean Polar Frontal Zone and fertilized a patch of water within an eddy of the Antarctic Circumpolar Current (EisenEx, Nov. 2000). During the experiment, a physical speciation technique was used for iron analysis in order to understand the changes in iron distribution and size-fractionations, including soluble Fe (〈200 kDa), colloidal Fe (200 kDa-0.2 µm) and labile particle Fe (〉0.2 µm), throughout the development of the phytoplankton bloom. Prior to the first infusion of iron, dissolved (〈0.2 µm) iron concentrations in the ambient surface seawater were extremely low (0.06±0.015 nM) with colloidal iron being a minor fraction. For the iron addition, an acidified FeSO4 solution was released three times over a 23-day period to the eddy. High levels of dissolved iron concentrations (2.0±1.1 nM) were measured in the surface water until 4 days after the first iron infusion. After every iron infusion, when high iron concentrations were observed before storm events, there was a significant correlation between colloidal and dissolved iron concentrations ([Colloidal Fe]=0.7627[Dissolved Fe]+0.0519, R2=0.9346). These results indicate that a roughly constant proportion of colloidal vs. dissolved iron was observed after iron infusion (~76%). Storm events caused a significant decrease in iron concentrations (〈0.61 nM in dissolved iron) and changed the proportions of the three iron size-fractions (soluble, colloidal and labile particle). The changes in each iron size-fraction indicate that colloidal iron was eliminated from surface mixed layer more easily than particulate and soluble fractions. Therefore, particle and soluble iron efficiently remain in the mixed layer, probably due to the presence of suspended particles and naturally dissolved organic ligands. Our data suggest that iron removal through colloidal aggregation during phytoplankton bloom should be considered in the oceanic iron cycle.