Résumé
The commercial success of a given dessert or cooking banana variety is based on the acceptability by the stakeholders: consumers and/or processors of the food chain. Unsuccessful commercial attempts for varietal adoption or the use of new processes can be due to the low consideration given to the consumer preferences (socio-economics of food) during varietal development or processing. Banana may be consumed raw at full ripeness for the dessert types or after being cooked at various stages of ripeness for the cooking types, and can also be fermented at full ripeness for brewing or bioethanol production. The study presents the collaborative investigations carried out at CIRAD and CIAT aiming at differentiating varieties, genetic groups (haploidy) or consumption groups (dessert and cooking). The statistical tests carried out help to validate the most suitable consumption mode for an unknown variety at various stages of ripeness according to the following criteria: i) Morphology: bunch shape and weight, finger size, diameter and weight, and pulp-to-peel ratio; ii) Physicochemical composition of the pulp: dry matter content, texture, pH, titratable acidity, ash, mineral content, soluble solids, starch, sugar, organic acid content and flavor; and iii) Functional properties of starch and flour by rapid viscosity analysis (RVA) and differential scanning calorimetry (DSC): pasting temperature, cooking ability, peak viscosity, onset temperature, gelatinization enthalpy and amylose content. Based on socio-economic surveys, the methods used for Musa screening helped to differentiate and/or select suitable varieties based on the prediction of their respective adoption by the stakeholders and users. The methods applied will contribute to the improvement of the breeding stategies while favouring the selection of optimally coupled variety and process for agro-industrial uses. (Résumé d'auteur)