Pelagic iron cycling during the subtropical spring bloom, east of New Zealand

Abstract

Upper ocean cycling of dissolved and particulate iron was investigated within an eddy during a study of the annual subtropical phytoplankton bloom east of New Zealand in the austral spring of 2008. During this GEOTRACES process study, dissolved iron surface mixed layer concentrations were initially high at ~0.6nmolkg-1 and declined to ~0.03nmolkg-1 due to biological consumption during a diatom-dominated bloom. The consequent iron limitation of the phytoplankton assemblage resulted in the decline and downward export of the bloom. Particulate iron concentrations were high in the surface mixed layer and varied between 3.9nmolL-1 and 12.1nmolL-1 during the onset and export of the bloom, respectively. The particulate iron and manganese results, along with a tracer model for the establishment of the eddy, suggest that the likely origin of the upper ocean particulate and dissolved iron pools within the eddy was from the continental margins along eastern New Zealand. Iron to aluminium (Fe:Al) ratios for suspended particulate material collected using McLane pumps deployed at 100m and 200m were significantly higher (range 0.16 to 2.90) than Fe:Al ratios for sinking particulate material intercepted using free-floating sediment traps (range 0.19 to 0.23) at corresponding deployment depths to the pumps. Based on the particulate Fe:Al ratios obtained for McLane pump collected samples, greater than 70% iron within the mixed layer was biogenic, indicating that either the resident biota are efficiently retaining iron or iron is associated with organic detritus. Iron budgets, based on the dissolved, suspended and sinking particulate iron datasets, were constructed for the evolution and subsequent decline phases of the diatom bloom. The budgets reveal that the turnover time for dissolved iron within the mixed layer was on the order of days, suggesting that iron was rapidly exchanged from the dissolved pool to the particulate biogenic and lithogenic pools. In contrast, the residence time for iron in the particulate biogenic pool was in the order of 5-8months indicating that iron was rapidly recycled but strongly retained by the biological community compared to the pelagic residence time for lithogenic iron, which was around 10-24days. The regeneration of iron from biogenic particles was highest immediately below the euphotic zone (~50m) and decreased with depth. Estimated regeneration fluxes for dissolved iron released from biogenic particles at 100m depth were 0.34±0.26nmolm-3d-1 during the development phase of the bloom and 7.1±3.9nmolm-3d-1 at the peak of the bloom, when the overall export flux for particulate iron had increased 2-fold. Our results suggest a dynamic balance between iron regeneration from particulate organic material below the euphotic zone and the lability of this organic matter. Finally, the different Fe:Al ratios obtained for suspended and sinking particulate matter indicate that these pools are biogeochemically different. Future GEOTRACES process study work should focus on determining the biogeochemical differences of iron between these pools because this will assist with interpreting GEOTRACES sections and evaluating the impact of iron biogeochemistry on pelagic production and remineralisation within the global ocean.