Résumé
In the field of agent-based modelling (ABM) for renewable resources management, actors (researchers, experts, stakeholders) are facing complex systems where social and physical dynamics take place within various interwoven levels of organization. Using conceptual modelling as a tool to formalize such systems facilitates their understanding as well as the sharing of the point of view of the various actors. In this context, we present an approach based on a multi-agent formalism called Agent-Group-Role. In order to implement this approach, we have developed a graphic methodology based mostly on UML and called ORIGAMI. This methodology was built extensively on the basis of a toy model and has been tested during a training experiment. Though some terminological problems occurred, the methodology has proven efficient in representing intervolved levels of organisation and their articulations and in explicating usually hidden assumptions.