Aquatic plants are varying dynamically with environmental processes which include physical and chemical factors as well as biological processes on different spatial and temporal scales. Physically-Based (PB) Models have proven to be quite sufficient to represent abiotic dynamic processes, but for biological processes, these models seem to lack the flexibility to handle local interactions amongst different species and their environment. Alternatively, new modelling techniques like Cellular Automata (CA) and Multi-Agent System (MAS) are giving promising results for precisely these reasons. A growing literature exists on the use of CA and MAS models for aquatic population dynamics modelling.

This study sets up two different types of models for the case of Lake Veluwe in the Netherlands: a Cellular Automaton (CA) model and a Multi-Agent System (MAS) model. Resulting spatial pattern dynamics are compared. The CA model is based on predefined discrete grids, with emphasis on local effects leading to spatial heterogeneity. However, it is relatively difficult to define the proper rules and/or functions for CA models under predefined grids. On the other hand, MAS can produce global behaviour by considering both local effects as well as the individual characteristics of e.g. plants, and are not restricted to predefined grids, exhibiting more flexibility than CA models. The optimal way of modelling in this case is demonstrated to be based on using a physically-based model for the abiotic environmental dynamics, then using MAS for modelling the interactions between different factors as well as aquatic population's biotic processes.