The first definition we could give to model is quite general : a model is any representation of the real world or what it is supposed to be at a given moment ; it clearly depends on the knowledge about reality or part of it ; as an abstraction it is necessary based on concepts and assumptions. This paper deals with mathematical modls only and among them, those which try to simulate biological mechanisms (for about 15 years however, other mathematicial approaches have been developping such as adaptative strategies, adapted from economics). Simulation alloxs to study dynamical behaviour of the elements of the ecosystem (which we denoted the real world) defined by the modeliser. Then two main purposes may be reached by the model : - prediction and management of the evolution of the ecosystem - test of the validity of our knowledge and hypothesis ; consequently model needs both mathematical and biological information.