Controlling complexity in individual-based models of aquatic vegetation and fish communities

Individual-based models (IBM's) have become a popular tool in ecology during the last ten years. These models have individual animals or plants as basic units and generate patterns of distribution or abundance resulting from the interaction between individuals. This approach has important advantages. To mention few:

• The biological principle that each individual is different can be incorporated, resulting in a higher realism.
  
• Parameters needed in the models as well as the predicted variables are typically of the type measured by experimental biologists.
  
• Model behavior is often rather robust to variation in formulation of the processes.
  

For these reasons the expectations were high for the contribution of these models to ecological theory. However, a recently published review of 50 individual-based animal population models concludes that the results of the first ten years have been disappointing. The main reason is that most individual-based models are very complex and special techniques needed to cope with this complexity have only occasionally been applied. In this thesis, two individual-based models are presented analyzed with a suite of special and often novel techniques. The result is an enhanced understanding of some mechanisms that may regulate the dynamics of fish communities and aquatic macrophytes, but also a strategy to deal with complexity of this kind of models which has evolved along the way.
This strategy consists of three phases: 'scrutinizing', 'simplifying' and 'synthesizing'. The first step is a thorough analysis of the model behavior with respect to a selected set of parameters ('scrutinizing'). Secondly, similar analyses are done with several simplified versions of the model ('simplifying'). In this step, relationships between state variables or species that may potentially cause incomprehensible behavior, are replaced by fixed values or highly simplified relations. The last step is to explain the differences between the full and the simplified versions and to discuss the results in the light of the existing ecological theory, field patterns or other models ('synthesizing'). It is argued that this way of combining analyses of simple and more elaborate models is a powerful way to gain understanding of complex systems.