The North Atlantic spring bloom is one of the largest annual biological events in the ocean, and is characterized by dominance transitions from siliceous (diatoms) to calcareous (coccolithophores) algal groups. To study the effects of future global change on these phytoplankton and the biogeochemical cycles they mediate, a shipboard continuous culture experiment (Ecostat) was conducted in June 2005 during this transition period. Four treatments were examined: (1) 12°C and 390 ppm CO 2 (ambient control), (2) 12°C and 690 ppm CO 2 (high pCO 2 ), (3) 16°C and 390 ppm CO 2 (high temperature), and (4) 16°C and 690 ppm CO 2 ('greenhouse'). Nutrient availability in all treatments was designed to reproduce the low silicate conditions typical of this late stage of the bloom. Both elevated pCO 2 and temperature resulted in changes in phytoplankton community structure. Increased temperature promoted whole community photosynthesis and particulate organic carbon (POC) production rates per unit chlorophyll a. Despite much higher coccolithophore abundance in the greenhouse treatment, particulate inorganic carbon production (calcification) was significantly decreased by the combination of increased pCO 2 and temperature. Our experiments suggest that future trends during the bloom could include greatly reduced export of calcium carbonate relative to POC, thus providing a potential negative feedback to atmospheric CO 2 concentration. Other trends with potential climate feedback effects include decreased community biogenic silica to POC ratios at higher temperature. These shipboard experiments suggest the need to examine whether future pCO 2 and temperature increases on longer decadal timescales will similarly alter the biological and biogeochemical dynamics of the North Atlantic spring bloom.

[1] The FeCycle experiment provided an SF 6 labeled mesoscale patch of high-nitrate low-chlorophyll (HNLC) water in austral summer 2003. These labeled waters enabled a comparison of the inventory of particulate iron (PFe) in the 45-m-deep surface mixed layer with the concurrent downward export flux of PFe at depths of 80 and 120 m. The partitioning of PFe between four size fractions (0.2-2, 2-5, 5-20, and >20 mm) was assessed, and PFe was mainly found in the >20-mm size fraction throughout FeCycle. Estimates of the relative contribution of the biogenic and lithogenic components to PFe were based on an Al:Fe molar ratio (0.18) derived following analysis of dust/soil from the nearest source of aerosol Fe: the semi-arid regions of Australia. The lithogenic component dominated each of the four PFe size fractions, with medians ranging from 68 to 97% of PFe during the 10-day experiment. The Fe:C ratios for mixed-layer particles were $40 mmol/mol. PFe export was $300 nmol m À2 d À1 at 80 m depth representing a daily loss of $1% from the mixed-layer PFe inventory. There were pronounced increases in the Fe:C particulate ratios with depth, with a five-fold increase from the surface mixed layer to 80 m depth, consistent with scavenging of the remineralized Fe by sinking particles and concurrent solubilization and loss of particulate organic carbon. Significantly, the lithogenic fraction of the sinking PFe intercepted at both 80 m and 120 m was >40%; that is, there was an approximately twofold decrease in the proportion of lithogenic iron exported relative to that in the mixed-layer lithogenic iron inventory. This indicates that the transformation of lithogenic to biogenic PFe takes place in the mixed layer, prior to particles settling to depth. Moreover, the magnitude of lithogenic Fe supply from dust deposition into the waters southeast of New Zealand is comparable to that of the export of PFe from the mixed layer, suggesting that a large proportion of the deposited dust eventually exits the surface mixed layer as biogenic PFe in this HNLC region.

Global climate change is predicted to have large effects on the ocean that could cause shifts in current algal community structure, major nutrient cycles, and carbon export. The Bering Sea is already experiencing changes in sea surface temperature (SST), unprecedented algal blooms, and alterations to trophic level dynamics. We incubated phytoplankton communities from 2 Bering Sea regimes under conditions of elevated SST and/or partial pressure of carbon dioxide (pCO 2 ) similar to predicted values for 2100. In our 'greenhouse ocean' simulations, maximum biomass-normalized photosynthetic rates increased 2.6 to 3.5 times and community composition shifted away from diatoms and towards nanophytoplankton. These changes were driven largely by elevated temperature, with secondary effects from increased pCO 2 . If these results are indicative of future climate responses, community shifts towards nanophytoplankton dominance could reduce the ability of the Bering Sea to maintain the productive diatom-based food webs that currently support one of the world's most productive fisheries.

No abstract