Iron availability is suggested to be a primary factor limiting nitrogen fixation in the oceans. This hypothesis is principally based on cost-benefit analyses of iron quotas in the dominant nitrogen-fixing cyanobacteria, Trichodesmium spp., in the contemporary oceans. Although previous studies with Trichodesmium have indicated that iron availability enhanced nitrogen fixation and photosynthesis, no clear relationship has been reported between cellular iron quotas and nitrogen fixation. We re-examined the proposed link between iron availability and nitrogen fixation in laboratory isolates and natural populations collected from coastal waters north of Australia. In laboratory cultures grown under iron-limiting conditions, we measured a decline in cellular iron quotas, photochemical quantum yields, the relative abundance of photosystem I to photosystem II reaction centers, and rates of nitrogen fixation. Nitrogen fixation displayed a critical threshold of the dissolved sum of total inorganic Fe species ([FeЈ]) of ca. log[FeЈ] ϭ Ϫ9.7. Field populations of Trichodesmium, collected during bloom conditions, showed high iron quotas consistent with high nitrogen fixation rates. Using seasonal maps of aeolian iron fluxes and model-derived maps of surface water total dissolved Fe, we calculated the potential of nitrogen fixation by Trichodesmium in the global ocean. Our results suggest that in 75% of the global ocean, iron availability limits nitrogen fixation by this organism. Given present trends in the hydrological cycle, we suggest that iron fluxes will be even more limiting in the coming century.Nitrogen fixation by planktonic prokaryotes is a major source of new nitrogen for the oceans (Capone et al. 1997;. In all cyanobacteria, the enzyme responsible for this process, nitrogenase, consists of two proteins, an iron 1 Corresponding author (irfrank@imcs.rutgers.edu). AcknowledgmentsWe thank J. and Y. Reinfelder for their help with the tracer experiments, iron solubility, and MINEQLϩ calculations; H. M. Geller and N. Goldman from the Department of Pharmacology, Robert Wood Johnson Medical School, for help and access to the confocal laser; P. Ludden for the antibody to nitrogenase; D. Kolber for generating the global maps; K. Wyman for invaluable lab assistance; and D. Capone, E. Carpenter, and the crew of the R/V Ewing for enabling our field sampling. We also thank J. Prospero, F. M. M. Morel, and two anonymous reviewers for their comments.

The conventional model of iron uptake in marine eukaryotic phytoplankton-the FeЈ model-suggests a dependency of uptake rate on the concentration of unchelated iron species (FeЈ), and not the concentration of total iron or iron chelated with organic ligands. However, iron in seawater is bound by strong organic ligands that buffer such low FeЈ concentrations that they should not support phytoplankton growth. Studies that show uptake and extracellular reduction of siderophore-bound iron by diatoms and provide indications that the iron uptake system of phytoplankton may be similar to that of yeast in which extracellular reduction is a prerequisite for uptake, call for revisions of the FeЈ model. In this paper we propose a new model for iron uptake by diatoms in which extracellular reduction of all Fe species is a necessary step for iron acquisition. Experiments verifying the predictions of the model are presented. In particular we show data supporting the fact that Fe(II) is formed as an intermediate during Fe uptake from all experimental media, including those buffered by Fe(III)EDTA. This model reconciles the standing FeЈ model with new data and concepts on reduction of iron chelates and provides a convenient framework for designing and interpreting iron uptake experiments in a variety of natural and artificial media.A major role for iron in marine phytoplankton physiology, ecosystem structure, and the ocean carbon cycle is emerging from numerous oceanographic studies (e.g., Wells et al. 1994;Maldonado et al. 2001). Trace metal clean procedures and many innovative analytical techniques have provided accurate measurements of oceanic iron concentrations and are beginning to reveal the speciation and cycling of iron in the ocean (e.g., Johnson et al. 1997;Rue and Bruland 1997). In order to understand the import of these findings for phytoplankton physiology and ecology, we must understand the relationship between uptake mechanisms of phytoplankton and the concentration and speciation of iron in seawater. The conventional model of iron uptake by marine eukaryotic phytoplankton-the FeЈ model-suggests a dependency of uptake rate on the concentration of unchelated

No abstract

In many aquatic environments the essential micronutrient iron is predominantly complexed by a heterogeneous pool of strong organic chelators. Research on iron uptake mechanisms of cyanobacteria inhabiting these environments has focused on endogenous siderophore production and internalization. However, as many cyanobacterial species do not produce siderophores, alternative Fe acquisition mechanisms must exist. Here we present a study of the iron uptake pathways in the unicellular, planktonic, non-siderophore producing strain Synechocystis sp. PCC 6803. By applying trace metal clean techniques and a chemically controlled growth medium we obtained reliable and reproducible short-term (radioactive assays) and long-term (growth experiments) iron uptake rates. We found that Synechocystis 6803 is capable of acquiring iron from exogenous ferrisiderophores (Ferrioxamine-B, FeAerobactin) and that unchelated, inorganic Fe is a highly available source of iron. Inhibition of iron uptake by the Fe(II)-specific ligand, ferrozine, indicated that reduction of both inorganic iron and ferrisiderophore complexes occurs before transport through the plasma membrane. Measurements of iron reduction rates and the inhibitory effect of ferrozine on growth supported this conclusion. The reduction-based uptake strategy is well suited for acquiring iron from multiple complexes in dilute aquatic environments and may play an important role in other cyanobacterial strains.