SOME DIFFICULTIES IN MODELING CHLOROPHYLL-A EVOLUTION IN A HIGH-RATE ALGAL POND ECOSYSTEM

The High Rate Algal Pond (HRAP) is an efficient treatment for controll ing wastewater pollution by reducing the organic matter and the inorga nic nutrient content. Deterministic modelling of temporal evolution of algae could provide a rational basis for pond management. An experime ntal HRAP was set up in Meze (France) and sampled weekly over 24 month s. Models simulating the evolution of chlorophyll a concentration and nutrients (N and P) were constructed using Stella II software. The sea sonal pattern of chlorophyll a concentrations results from the annual cycle of solar irradiance and temperature, whereas shorter trends (1 t o 4 weeks) are dependent on the evolution of zooplankton groups. The f irst difficulty is to determine the functional relationships of the ph ytoplankton and zooplankton groups. In the model the evolution of the phytoplankton taxa is considered to be dependent on (i) inherent param eters of phytoplankton taxa (mortality rate, growth rate, saturating l ight intensity, etc.) and on (ii) parameters of zooplankton taxa (filt ration rate, size selectivity, etc.). To take all these taxa as state variables, and all the associated parameters, into account is impossib le: to solve such a problem, we forced the evolution of the biomass of the phytoplankton and zooplankton taxa. This approach improves the ag reement between the simulated and observed chlorophyll a concentration values. The second difficulty concerns the determinism of appearance of the phytoplankton and zooplankton taxa used in the model: up to now we are only able to force these appearances. Thus, even in quite simp le ecosystems, using deterministic modelling as a predictive tool requ ires a full understanding of the exact biological succession and inter action processes.