A model study on the stability of the macrophyte-dominated state as affected by biological factors

The transition of shallow lake ecosystems between the clear-water, macrophyte-dominated state and the turbid state dominated by phytoplankton depends on both physico–chemical and biological factors. In this study, the impact of some of these interactions on the stability of the macrophyte-dominated state of a lake are studied by means of the integrated eutrophication model PCLake. The model describes phytoplankton, macrophytes and a simplified food web, within the framework of closed nutrient cycles. The aim of the study is to evaluate the impact of herbivory by birds and fish on the transition from clear to turbid state, including the influence of variability in other biological parameters. The model was applied to the data of a small, biomanipulated lake, dominated by macrophytes, showing signs of a transition back to the turbid state. Simulations were carried out for the lake as well as for an experimental situation where herbivory was impeded. A parameter variation study was performed for 10 parameters, affecting the zooplankton, fish and macrophytes behaviour, to determine the sensitivities and the model uncertainty. The model reproduced adequately the transition of the lake from phytoplankton dominance before the biomanipulation, via dominance of rooted perennial plants in the first years after it, to a state characterized by turion-forming plants in early summer and phytoplankton in autumn. It is shown that the probability of the transition back to phytoplankton dominance is mainly enhanced by herbivory by birds. This caused a shift towards inedible plant species with a shorter natural growing season, allowing the return of a phytoplankton bloom in autumn. If herbivory was impeded, this shift did not occur and phytoplankton remained low due to nitrogen limitation. The model results were quite sensitive to the zooplankton filtering rate and, in the presence of herbivory only, to the macrophytes growth parameters. The impact of the fish parameters showed to be less important. The model may be used to evaluate the relative importance of different assumptions or factors in the success of biomanipulation measures in lakes.