Résumé
The High Rate Algal Pond (HRAP) is an efficient treatment for controlling wastewater pollution by reducing the organic matter and the inorganic nutrient content. Deterministic modelling of temporal evolution of algae could provide a rational basis for pond management. An experimental HRAP was set up in Mèze (France) and sampled weekly over 24 months. Models simulating the evolution of chlorophyll
a concentration and nutrients (N and P) were constructed using Stella II software.
The seasonal pattern of chlorophyll
a concentrations results from the annual cycle of solar irradiance and temperature, whereas shorter trends (1 to 4 weeks) are dependent on the evolution of zooplankton groups.
The first difficulty is to determine the functional relationships of the phytoplankton and zooplankton groups. In the model the evolution of the phytoplankton taxa is considered to be dependent on (i) inherent parameters of phytoplankton taxa (mortality rate, growth rate, saturating light intensity, etc.) and on (ii) parameters of zooplankton taxa (filtration rate, size selectivity, etc.). To take all these taxa as state variables, and all the associated parameters, into account is impossible: to solve such a problem, we forced the evolution of the biomass of the phytoplankton and zooplankton taxa. This approach improves the agreement 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 simple ecosystems, using deterministic modelling as a predictive tool requires a full understanding of the exact biological succession and interaction processes.