PISCATOR, an individual-based model to analyze the dynamics of lake fish communities

Unraveling the mechanisms that drive dynamics of multi-species fish communities is notoriously difficult. Not only are the interactions between fish populations complex, but also the functional niche of individual animals changes profoundly as they grow, making variation in size within populations and even within cohorts highly important to consider. Not surprisingly, traditional aggregated populations models have proved limited in their capacity to describe the dynamics of interacting fish species, and individual-based models have become popular for modeling fish populations. Nonetheless, the majority of the individual-based models describes either a single species or focus entirely on a certain life stage. We present the individual-based model Piscator, which describes a multi-species fish community and demonstrates techniques to deal with the inherent complexity of such a model. We propose a novel procedure for calibration and analysis, in which the complexity of the model is increased step-by-step. We also illustrate the use of a special Monte-Carlo sensitivity analysis to identify clusters of parameters that have roughly the same effects on the model results. As an example, we use the model to analyze a fishery experiment in the Frisian Lakes (The Netherlands). Despite high bream catches (40–50 kg ha −1 per year), it was observed that the seine fishery had unexpected little effect on the bream population. Our simulation results suggest that if one takes community feedbacks and climatic variability into account, this effect can be explained. The main cause was, besides a reduction of piscivory due to a simultaneous gill-net fishery, a coincidental strong year-class just before the fishery started. The strong development of this year-class could be explained by 3 subsequent warm years, whereas yearly variations in recruitment were less important. We also suggest that this relatively realistic model could play a role in ecological theory. It can be used to analyze the conditions for multi-year cycles and chaotic dynamics, phenomena that are usually predicted only from simple abstract models.