Persistence of iron(II) in surface waters of the western subarctic Pacific

The distribution of dissolved iron(II) [Fe(II)] was studied in surface waters of the western subarctic Pacific during the Subarctic Pacific Iron Experiment for Ecosystem Dynamics Study-II (SEEDS II) iron enrichment experiment using highly sensitive flow injection-based luminol chemiluminescence. Vertical profiles of Fe(II) and total dissolved iron were measured outside of the fertilized patch to investigate the chemical speciation of iron in this high-nitrate low-chlorophyll (HNLC) region. Ambient total dissolved iron concentrations ranged from 50 pmol L21 to150 pmol L21 depending on depth and sampling times. Unexpectedly, Fe(II) accounted for up to half of total dissolved iron, with concentrations up to ,50 pmol L21. Fe(II) concentrations decreased exponentially with depth and were undetectable at depths below 50 m. There was no evidence of increased Fe(II) concentrations associated with the subsurface chlorophyll maximum, indicating that photolysis, rather than biological reduction of Fe(III), was the primary source of Fe(II). Because Fe(II) concentrations in the fertilized patch remained elevated for more than a week after enrichment, Fe(II) oxidation rates at near-ambient concentrations were measured. Indeed, the temperature-dependent Fe(II) oxidation rates were significantly slower than predicted by Fe(II) oxidation models and rates measured in ligand-free seawater. These findings suggest that Fe(II) binding ligands may exist in these HNLC waters, with conditional stability constants on the order of 108– 109 with respect to Fe2+. The accumulation of Fe(II) during daylight hours did not alleviate iron limitation of eukaryotic phytoplankton in these waters, contrary to expectations from recent iron uptake models. The constraint of carbon export by iron (Fe) supply in the high-nitrate low-chlorophyll (HNLC) regions of the Southern Ocean, equatorial Pacific, and subarctic Pacific is now well demonstrated, but the inability of diatoms and other eukaryotic phytoplankton to fully utilize the ambient iron pools in these waters is much less understood. The thresholds for diffusion-limited iron uptake, for even large pennate diatoms, are ,10 pmol L21 (Hudson and Morel 1993; Wells 2003), yet dissolved iron concentrations in HNLC regions are often an order of magnitude higher. The overwhelming (,99%) control of iron speciation by highaffinity organic chelators (see Rue and Bruland 1997) is believed to restrict iron availability to diatoms, however, this expectation stems from the assumption that iron speciation is at or near equilibrium in surface ocean waters. At equilibrium, inorganic iron concentrations are ,0.1 pmol L21 (Rue and Bruland 1997), a level too low to support the growth of large oceanic phytoplankton (Brand et al. 1983; Sunda and Huntsman 1995; Wells 2003), and thermochemical dissociation rates of these complexes are too slow to replenish the inorganic Fe(III) species sequestered by uptake (Hudson and Morel 1993; Wells and Trick 2004). The apparently ubiquitous excess of these strong iron-specific organic ligands in HNLC waters (Rue and Bruland 1997) challenges the current view that natural iron deposition events can transform phytoplankton communities, because even large aerosol inputs to the ocean cause only subnanomolar increases of dissolved iron in surface waters (Sedwick et al. unpubl.). Photochemical cycling of Fe(III) is known to occur in surface ocean waters (Kuma et al. 1992; Johnson et al. 2004; Miller et al. 1995), and the photochemical action spectrum for these transformations suggests they can occur deep into the photic zone (Wells et al. 1991; Laglera and Van Den Berg 2007). Despite this, the net effect of photolysis on iron speciation generally has been assumed to be small because of rapid reoxidation of photoproduced Fe(II) in oxic seawater (King et al. 1995). Kinetic models have suggested that steady-state inorganic iron [Fe(III) plus Fe(II)] concentrations increase to only a few percent of total dissolved iron under full sunlight (Sunda and Huntsman 1995; Rue and Bruland 1997). This expectation has been challenged recently, particularly in cold seawaters where Fe(II) oxidation kinetics are substantially slower (Croot et al. 2001). Moreover, the apparent use of reductive, high-affinity iron uptake systems by some diatoms (Maldonado and Price 2001; Wells et al. 2005) suggests that photoproduced Fe(II) may be readily accessible. However, the extent that these redox processes change iron speciation in natural seawater is not well understood. Fe(II) oxidation in natural waters is controlled largely by reactions with dissolved oxygen and hydrogen peroxide. In most open ocean surface waters, hydrogen peroxide concentrations are significantly below 200 nmol L21 1 Corresponding author ([email protected]).