Blazy, Jean-Marc, Ozier-Lafontaine, Harry, Doré, Thierry, Thomas, Alban, and Wery, Jacques
Agricultural systems, 2009 June, v. 101, no. 1-2, p. 30-41.
crop management systems, set of constraints, crop rotation, good agricultural practices, agroecosystems, crop management, agroecological zones, geographical variation, farming systems, innovation adoption, Musa, farm income, bananas, farms, cultivars, intercropping, pesticides, and conservation practices
Includes references Prototyping methods are useful for designing alternative crop management systems (CMS), but they are usually considered mainly at the field level and are poorly equipped to take into account farm diversity in terms of economic, social and natural constraints. This may limit the likelihood of adoption of alternative CMS. This paper proposes a two-step methodological framework that takes into account the diversity of farms in the prototyping of new CMS. The first step of the framework is to design a farm typology that is able to characterise the diversity of current CMS in terms of their technical nature, farming context, and performance. We define a farming context as a set of characteristics at the farm level that are likely to influence the structure and the biophysical and economic performances of CMS at the field level. The typology is designed from a statistical analysis of a set of descriptive variables collected for a sample of farms. The second step of the framework uses this farm typology in a specific work agenda with a panel of experts in order to design a set of alternative CMS prototypes. Prototypes, and the objectives they are meant to reach, are designed by analysing current CMS performances and by mobilising agro-ecological knowledge to identify technical ways to improve them. The typological characterisation of farm diversity then makes it possible to define a set of constraints (SOC) for each farm type. Once the prototypes are designed, the experts use a compatibility indicator (CI) ranging from zero to one to measure the level of compatibility between each prototype and the SOC of each farm type. We define a prototype as compatible with a farm type if the prototype is not likely to increase the constraints of a farm of that type (CI=1). Using this indicator in a process of iterative improvement of prototype design makes it possible to define a set of a priori compatible prototypes for each farm type. The aim of this paper is to present our methodological framework and its application for prototyping new banana crop management systems in Guadeloupe. The six-class typology we developed corresponded to a great diversity of CMS with contrasting performances and farming contexts. The method led to the prototyping of 16 innovative CMS, involving different modalities of intercropping, patterns of pesticide use, choices of hybrid cultivars, and rotations with cover or cash crops. Finally, we discuss the limits of our methodological framework and its utility for improving prototyping approaches.