Received: 18/06/2000 © Copyright 2000 Accepted: 06/11/2000 – 1 – http://www.csu.edu.au/ci/vol07/lparro01/ http://www.csu.edu.au/ci/ Volume 7 Incorporating Complexity in Ecosystem Modelling Lael Parrott Dept. of Agricultural and Biosystems Engineering McGill University Quebec Canada Email: laelparrott@hotmail.com Robert Kok Dept. of Agricultural and Biosystems Engineering McGill University Quebec Canada Email: kok@macdonald.mcgill.ca Abstract This article is a discussion of how the field of ecosystem modelling is being affected by the adoption of ideas arising from complex system studies. The modelling process is presented as being composed of four stages, starting with a modelled system, which is then depicted in turn by conceptual, representational, and computational models. The way that each of these stages is affected by conceptualizing an ecosystem as a complex system is discussed, with reference to current trends in ecosystem modelling. In particular, the recent emphasis on the use of “object- based” models, in which an ecosystem is represented at a high level of resolution as a collection of many thousands of interacting components, is presented as being an obvious method by which to reproduce complex dynamics in computer-based simulation. Three main types of object-based models are identified: individual-based models, agent-based models, and cellular automata. These three types are discussed with reference to examples from the literature. Several different computational approaches and programming platforms that are applicable to object-based ecosystem modelling are then reviewed. The paper’s target audience is anyone who desires, for teaching or research purposes, an overview of the various modelling methods (and their advantages and disadvantages) that are currently being used to represent ecosystems in the context of complexity research.