In order to conciliate agricultural production and ecosystem protection, managing the spatial and temporal organization of crops at the landscape level could be an interesting option. Decisions farmers make for their cropping plan at the farm level are among the underlying controlling factors of this organization. The main aim of this thesis was to analyse how the modelling of farmer cropping plan decisions in their spatial and temporal dimensions could explain the spatial crop organization at the landscape level. In the Niort plain (Poitou-Charentes region), we first identified (i) landscape regularities in crop organization and (ii) farmer decision rules at the farm level. The comparison of these two elements confirmed that regularities in spatial and temporal crop organization could be explained by generic farmer decisions. Based on 12 on-farm surveys, we then built the DYSPALLOC model. This conceptual model can simulate the spatial crop organization from one year to another at the farm level, through the spatial and temporal representation of the farmer cropping plan decision process. This model, suitable for both arable and mixed crop-livestock farms, gave the possibility to specify the rules for defining plot limits inside the farming territory. We defined three types of plots based on their boubdaries: elementary islets, permanent plots and temporary plots. The evaluation of the model showed an 83% rate of success in the crop allocation to plots. A virtual experimentation also validated the concepts of permanent and temporary plots. We finally used the model at the level of a landscape composed of several farms, using input data built on the basis of spatial data-bases. We demonstrated that the spatial crop organization was closer to the real one when simulated taking farmer cropping plan decisions into account than when simulated randomly, even when introducing agronomic constraints. This work could thus support the coordination of individual farmer cropping plan decisions at the landscape level for building spatial crop organization favourable to ecosystem services.