摘要：The goal of this article is to give some guidelines when modeling farmers’ rotation choices through optimization models. To improve the accuracy of such models, researchers can i) sophisticate the utility function or ii) specify the production function and the constraints of the model. Based on an interactive approach involving farmers, a preliminary discrete determinist model is built and tested under changing crops prices. Then, two discrete stochastic modeling approaches are compared; in the first one, yield risk is accounted as main source of income variability and, in the second one, risk is incorporated as a stochastic constraint of monthly inaccessible field days. Results show that risk aversion little affects rotation choice. A stochastic labour constraint accounting for field inaccessibility has considerable more impact on crops choice, especially in presence of imperfect labour market.
