What controls dissolved iron concentrations in the world ocean?

Abstract

Dissolved (< 0.4 μm) iron has been measured in 354 samples at 30 stations in the North and South Pacific, Southern Ocean and North Atlantic by the Trace Metals Laboratory at Moss Landing Marine Laboratories. These stations are all more than 50 km from a continental margin. The global distribution of dissolved iron, which is derived from these profiles, is remarkable for several reasons. The dissolved iron profiles have a uniform shape with a nutrient-like profile at each station. Concentrations at the surface are all < 0.2 nmol kg −1 and average 0.07 nmol kg −1 . Below 500 m, the average concentration is 0.76 nmol kg −1 . The largest value in the data set is 1.38 nmol kg −1 . There is no inter-ocean fractionation, which is unique for an element with a nutrient-like profile. Published estimates of the iron residence time are on the order of 100 to 200 yr, indicative of rapid removal. Other elements with such short residence times are characterized by vertical profiles that decrease with depth and deep concentrations that decrease with age as water passes from the Atlantic to the Pacific. This is not the case for iron. The largest horizontal changes in dissolved iron are observed in gradients from the continental margin. There is only a factor of three difference between the minimum (0.4 nmol kg −1 ) and maximum (1.3 nmol kg −1 ) value in the data set at a depth near 750 m, where variability is at a maximum. The minimum concentrations are found at stations in the remote central Pacific and the maximum values occur at stations adjacent to the continental margin. The major source of iron in the deep sea is generally aeolian deposition. Integrated (surface to 500 m) concentrations of iron at each station are only weakly correlated with the aeolian iron deposition flux, however. This contrasts with other elements such as lead that also have strong atmospheric sources. These observations lead us to conclude that the nutrient-like profile is maintained by a mechanism that reduces the scavenging rate of dissolved iron at concentrations less than 0.6 nmol kg −1 . This mechanism may be complexation by strong iron binding ligands, which have been found in both the Atlantic and Pacific at concentrations near 0.6 nM. This apparent solubility would act to diminish inter-ocean fractionation. It would allow a nutrient-like profile to develop before scavenging began to remove iron. In order to test the concept, we developed a numerical model to make quantitative predictions of dissolved iron concentrations from place to place. The dissolved iron source in the ocean interior is remineralization from sinking particulate organic matter. Scavenging removes dissolved iron only at concentrations greater than the apparent solubility. The only geographically variable parameter in the model is the export flux of carbon from the surface layer, which carries iron with it. The model generated dissolved iron profiles, based on measured or estimated values of the carbon export flux, are in remarkable agreement with the observed profiles at all stations from the North Atlantic through the Southern Ocean to the North Pacific.