ROTAT, a tool for systematically generating crop rotations

This paper reports part of a methodology for a model-based exploration of land use motivated by the lack of sustainability of small farming systems in southern Uruguay. Explorative land use studies aim to gain insight into future possibilities for agricultural development. They support strategic thinking during the design of new farming systems. The crop rotation plays a central role in a farming system and represents a logical starting point in the design process. The combination and sequence of crop species determine characteristics of farming systems such as crop yields, soil erosion, occurrence of soil-borne pests, diseases and weeds, and dynamics of nitrogen and labour. Here, we present a software tool called ROTAT, designed for generating crop rotations based on agronomic criteria in a transparent manner. The program combines crops from a predefined list to generate all possible rotations. The full factorial number of possible combinations of crops is limited by a number of filters controlled by the user. These filters are designed to eliminate crop successions which are agronomically unfeasible and for farm-specific reasons not practical or desirable. The filters represent expert knowledge in a quantitative and explicit way. The use of this computer program as a stand-alone tool in the process of designing crop rotations is illustrated with a published case study from an ecological pilot farm in Flevoland (The Netherlands). Using this software we were able to design 840 rotations based on the same crops and designing criteria that were used for the example farm. Many of these rotations might be interesting alternatives to the one actually implemented. Coupled with a sound procedure to evaluate the performance of such a large number of rotations 'a priori', ROTAT can reduce the risk of ignoring promising options and the arbitrariness present in previous studies dealing with design of rotations. The usefulness Of ROTAT for designing production activities in explorative land use studies based on linear programming is discussed. (C) 2002 Elsevier Science B.V. All rights reserved.