During the 13 day Southern Ocean Iron RElease Experiment (SOIREE), dissolved iron concentrations decreased rapidly following each of three iron-enrichments, but remained high (>1 nM, up to 80% as FeII) after the fourth and final enrichment on day 8. The former trend was mainly due to dilution (spreading of iron-fertilized waters) and particle scavenging. The latter may only be explained by a joint production-maintenance mechanism; photoreduction is the only candidate process able to produce sufficiently high FeII, but as such levels persisted overnight (8 hr dark period)-ten times the half-life for this species-a maintenance mechanism (complexation of FeII) is required, and is supported by evidence of increased ligand concentrations on day 12. The source of these ligands and their affinity for FeII is not known. This retention of iron probably permitted the longevity of this bloom raising fundamental questions about iron cycling in HNLC (High Nitrate Low Chlorophyll) Polar waters. ¿ 2001 American Geophysical Union