Résumé
Process Based Models (PBM) are used to compute crop yield, they treat a field as a homogenous medium, total production is computed accurately, but there is little information on individual variability of plants, and number and size of organs. Functional Structural Plant Models (FSPM) simulate individual plant development and growth, they build plant architecture on the basis of biomass production and partitioning according to botanical rules, but generally lack the ability to upscale to the field level to address crop production. In the present study we combined crop biomass production of maize estimated by the PBM PILOTE with biomass partitioning and individual plant architecture generated by the FSPM GreenLab, in order to benefit from the field-level accuracy of the former and from the heterogeneity of individual architecture of the latter.