Limnol. Oceanogr., 44(7), 1999, 1609–1615

Substitution of the noniron protein flavodoxin for the iron–sulfur protein ferredoxin is an iron-stress response employed by a variety of unicellular organisms, including many phytoplankton. The relative abundance of these two proteins has been shown to vary with the severity of growth limitation by iron in marine diatoms. During the IronEx II mesoscale iron-enrichment experiment, large volume (100–600 liters) phytoplankton samples were collected for analysis of community ferredoxin and flavodoxin abundance using a high-pressure liquid chromatography (HPLC) technique. In addition, three pennate diatom species isolated from the fertilization-induced phytoplankton bloom were used for follow-up laboratory experiments, which examined their iron physiology. Prior to iron enrichment, biomass levels were insufficient to obtain any ferredoxin or flavodoxin signals. Measurements were successful after iron enrichment, with unexpected results. The strength of the HPLC signal tracked the biomass levels of the IronEx II phytoplankton bloom. Chromatographic peaks were evident on the fifth day following enrichment and persisted throughout the experiment before they declined and eventually disappeared following the last iron infusion. The main chromatographic peak was identified as flavodoxin; there was no evidence of ferredoxin in any of the samples. Pennate diatom clones isolated from the fertilization-induced bloom and grown in the laboratory retain the ability to make ferredoxin when iron-replete and induce flavodoxin when iron-stressed. When iron-limited, they are able to completely repress flavodoxin expression in about 1 d in response to iron resupply. Thus, the unexpected absence of ferredoxin and the persistence of flavodoxin during IronEx II, despite the observed increases in biomass and photosynthetic efficiency, suggest that the iron additions were insufficient to completely relieve physiological iron limitation. Ever since Martin proposed the ‘‘iron hypothesis,’’ considerable effort has been invested in the study of the relationship between iron availability and primary production (Martin 1990). Initial evidence for iron limitation of phytoplankton growth was derived primarily from the results of shipboard nutrient addition bottle bioassays (e.g., Buma et al. 1991; Coale 1991). More recently, two mesoscale ironfertilization experiments performed in the eastern equatorial Pacific provided direct proof of iron limitation of production in that area (Martin et al. 1994; Coale et al. 1996). Interpretation of bottle bioassay results is complicated by potential artifacts due to manipulation and enclosure of the natural population (e.g., Venrick et al. 1977), and mesoscale enrichments are logistically difficult and prohibitively expensive. The difficulties inherent in these nutrient addition methods illustrate the need for a test of iron limitation that does not entail excessive manipulation or incubation of the phytoplankton community. One of the most promising candidates for a specific assay of iron stress is the ferredoxin/flavodoxin system of proteins 1 Present address: University of Warwick, Department of Biological Sciences, Coventry CV4 7AL, Great Britain.