The current research is undertaken in the high-risk environment of poor farmers in Mali (West Africa) where agricultural biodiversity is crucial and people are always aiming at being more efficient in their use of the natural resource base. Hence, maximising the contribution of agricultural biodiversity to sustainable livelihoods involves strengthening human and social capital in ways that support the management of the natural capital, including plant genetic resources: interdisciplinary study has proven to be an invaluable tool. This research deals with the structure of society, the way the farmers act, where and how decisions are made to manage genetic resources. According to a scaling-up methodology and a systemic approach, data was gathered and processed in a suitable spatial framework. Then dynamic inputs such as farmers' strategies and social network interactions are modeled using multi-agent systems. The first finding from multi-agent simulation is to provide a common understanding of farmers' seed management. The participatory modelling process allows an increase in knowledge and procedural skills by learning from experience with farmers.