9633711 WELSCHMEYER Uncertainties in rates of oceanic primary production have been disputed for years. Although, the problem has been the subject of much review and experimentation, a discrepancy still exists between estimates of primary production based on non-manipulative biogeochemical calculations and those based on short-term 14C incubation experiments. We renew the theoretical, but untested, arguments that grazer-induced losses of metabolized carbon result in measurable underestimates of the true rate of primary production in standard 14C experiments. A model is elaborated which predicts significant (two-fold) underestimates of production using realistic assumptions commensurate with modern views of marine food web structure. The model clearly identifies the importance of the timing of diel microzooplankton grazing as it affects the outcome of routine 12 h or 24 h incubations. Preliminary experiments, both in the lab and in the field, suggest the model is indeed realistic, and that underestimates of production are easily demonstrated, even in short 12 h daylight incubations. This project will undertake extensive laboratory experimentation designed to establish the stochiometry of metabolized 14C-loss against known rates of microzooplnakton grazing rates and respiration. The model will thus be empirically parameterized, allowing predictive estimates of 14C-production errors to be made under in situ conditions of temperature, grazing and phytoplankton growth. A series of testable hypotheses is generated for field samples, and a complementary plan of field experiments is proposed, which will quantify grazer-induced production errors along a gradient of coastal-oligotrophic sites, including the JGOFS Hawaii time series station. An increasing gradient in production errors is expected as the natural community of grazers changes from coastal to oligotrophic plankton assemblages. This research project will result in a fully-tested analysis of grazer-induced errors in primary production experimen ts, including recommendations for optimized methodology , and will provide realistic constraints on error limits which may pertain to contemporary primary production data. A potentially significant portion of the oceanic productivity dilemma may be resolved as a result of this project. ***
