Models to support cropping plan and crop rotation decisions. A review

Farmers must yearly allocate fields to different crops and choose crop management options. Far from being obvious, these decisions are critical because they modify farm productivity and profitability in the short and long run. To support farmers and efficiently allocate scarce resources, decision support models are developed. Decision support models are mainly based on two concepts, the cropping plan and the crop rotation decisions. These decisions concern crop choice, crop spatial distribution within the farmland and crop temporal successions over years. Decisions are at the core of the farm management. Decisions have strong impacts on resource use efficiency and on environmental processes at both farm and landscape scales. We review here more than 120 references where cropping plan and crop rotation decision concepts were incorporated into models. Our aim is to review how these two concepts have been formalised and used in agronomic, economic and land-use studies. We found that cropping plan decisions selection and design have been done using many approaches based on different objectives and handled at very different scales. The main results show that (1) cropping plan design decisions have mainly been tackled as a static concept, i.e. as if they were a single decision made only once a year or once a rotation; (2) modelling the achievement of a suitable cropping plan is often based on a single monetary criterion optimization procedure instead of a multi-criteria assessment; and (3) when considered, uncertainty of information is defined as stochastic factors or probability of occurrence, but this probability is kept static whatever the knowledge of the dynamic evolution of various constraints. We argue that cropping plan and crop rotation decisions are on the contrary dynamic processes incorporated into a succession of other planned and adaptive decisions made at annual and long-term horizons. For supporting farmers in their decisions, new cropping plan decision models will require new modelling paradigm. A promising improvement could be reached by including explicitly the simulation of the farmers' decision-making processes, based on the simulation of the decision-making processes rather than on single normative approaches.