Quantification of algal iron requirements in the Subantarctic Southern Ocean (Indian sector)

Shipboard iron-addition incubation experiments were carried out in the Indian sector of the Subantarctic Southern Ocean during the Antares-IV campaign in late January-February 1999. The aim of these experiments was to estimate the dissolved iron requirements of the native phytoplankton community in this oceanic region, in order to improve the parameterisation of iron as a limiting nutrient in a coupled 1D physical-biogeochemical ocean model. The experiments were conducted with plankton collected from the upper water column (similar to 20 m depth) at three sites in the Crozet Basin between 43-46degreesS and 61-65degreesE: (1) the Polar Front Zone (PFZ, dissolved Fe=0.33nM), (2) the confluence of the Subantarctic and Subtropical Fronts (SAF/STF, dissolved Fe = 0.29 nM), and (3) the southern Subtropical Zone (STZ, dissolved Fe = 0.09 nM). Experimental results from each site indicate that algal community growth rates varied as a function of added iron concentration. Monod saturation functions fitted to the experimental data yield estimates for the community half-saturation constant for growth with respect to iron (K(mu)) of 0.41-0.45 nM Fe (PFZ), 0.055-0.086 nM Fe (SAF/STF) and 0.092-0.093 nM Fe (STZ, with macronutrients, added), each of which has an estimated uncertainty of +/- 20 %. The K(mu) estimate for the SAF/STF site reflects the mixed algal assemblage (diatoms+ nanoplankton + dinoflagellates) that grew in the experimental incubations, whereas the K(mu) estimates for the PFZ and STZ sites probably reflect the Fe requirements of the small pennate diatoms such as Pseudo-nitzschia spp., which dominated the algal biomass produced in these experiments. The fact that there are significant differences between the K(mu) estimates for the PFZ and STZ sites suggests that similar diatom assemblages may have quite different iron requirements, perhaps due to differences in environmental conditions (e.g., light, macronutrient levels). We also examine the sensitivity of a one-dimensional coupled physical-biogeochemical model to the choice of K(mu) for iron, using time-series observations from the KERFIX station close to the Polar Front. The model was best able to simulate the KERFIX observations using a diatom K(mu) value of similar to0.1 nM Fe, which is considerably less than our experimental estimate of similar to0.4 nM Fe for the PFZ. This discrepancy probably reflects differences in the iron requirements of diatom populations immediately north and south of the Polar Front in the Kerguelen region, due to differences in diatom species composition, availability of light and silicic acid, or the environmental and physiological histories of the diatom populations. (C) 2002 Elsevier Science Ltd. All rights reserved.