Predictive Ecological Modeling of Harmful Algal Blooms

We have thus far constructed a series of coupled ecological/physical models to predict the origin and fate of harmful algal blooms of the toxic dinoflagellate, Gymnodinium breve, on the West Florida shelf We find that (1) the maximal population growth rate of G. breve must be similar to0.80 day(-1) during initiation of a red tide, but as little as 0.08 day(-1) during its decay, (2) diatoms dominate when estuarine and shelf-break supplies of nitrate are made available to a model community of small and large diatoms, coccoid cyanophytes and Trichodesmium, non-toxic and red-tide dinoflagellates, microflagellates, and coccolithophores, (3) a numerical recipe for large red tides of G. breve requires DON supplies, mediated by iron-starved, nitrogen-fixers, while small blooms may persist on sediment sources of DON, (4) selective grazing must be exerted on the non-toxic dinoflagellates, and (5) vertical migration of G. breve in relation to seasonal changes of summer downwelling and fall/winter upwelling flow fields determines the duration and intensity of red tide landfalls along the barrier islands and beaches of the west coast of Florida, once other losses are specified. Given poorly known initial and boundary conditions and the expense of shipboard monitoring programs, however, bio-optical moorings or remote sensors are the most likely sources of model validation and improved HAB forecasts. Accordingly, the bio-optical implications of our ecological models must be included in future simulation analyses of HABs on both the West Florida shelf and within other coastal regions. Of particular importance for initiation of the coupled biophysical models is an improved understanding of the relationship of remotely sensed surface signals to shade-adapted dinoflagellates, aggregating below the first optical depth of the water column.