Controls on Southern Ocean phytoplankton production : a systems approach

Plankton ecosystems play a significant part in the global cycle of carbon by contributing ~40% of total annual primary productivity. Changes are expected in response to warming, increased ocean stratification, altered supply of iron as a limiting micronutrient, and ecosystem structures. To diagnose these effects, numerical models of plankton ecology have become a key approach. This thesis examines the Nutrient Phytoplankton Zooplankton Detritus (NPZD) ecosystem model of Oschlies & Garcon [1999], and especially the modifications required to simulate phytoplankton biomass in naturally iron-fertilized Southern Ocean waters. This provides perspectives on the sensitivity of Southern Ocean ecosystems to global change. The NPZD model was used to simulate the phytoplankton bloom that forms in natural iron-fertilized waters of the Kerguelen plateau and adjacent High Nutrient Low Chlorophyll (HNLC) waters. Sensitivity to mixed layer depth and temperature was found to be negligible in comparison to zooplankton grazing, which also strongly modulated the ability of increased iron availability to affect biomass via phytoplankton growth rates. Direct comparison with satellite chlorophyll estimates showed that the base model could not achieve chlorophyll concentrations greater than ~0.6 mg m\(^{-3}\), in comparison to the ~2.7 mg m\(^{-3}\) of chlorophyll observed in the Kerguelen bloom. Incorporating an indirect simulation of iron-fertilization provided only a small improvement. An increase in zooplankton grazing (by a factor of 7) was required to achieve the high peak bloom biomass. To match the seasonal cycle and both summer maximum and winter minimum biomass concentrations, it was also necessary to impose seasonal cycles in key photosynthetic parameters (α - initial slope of the photosynthesis/irradiance curve, and μ\(_{max}\) - maximal growth rate). The same model was used to examine the response of plankton ecosystems to temporary iron fertilizations. The model was able to simulate the phase and amplitude patterns of the SOIREE bloom in Antarctic waters, given increases in α and μmax similar to those observed in the field (a factor of 2.5 over 13 days). Simulating a series of SOIREE-like experiments at different times of the seasonal cycle showed that while blooms induced in spring resulted in high surface chlorophyll levels, only blooms induced in summer resulted in both high chlorophyll and high potential carbon export. These findings also emphasized the importance of low overwintering biomass of phytoplankton and zooplankton to the spring biomass peak. Expanding the simulations to all latitudes of the Southern Ocean revealed distinct variations in the intensity of responses to iron fertilization, and suggested that the optimal time and location for iron fertilization in terms of potential carbon export is early summer in the Subantarctic zone