Faba bean adaptation to autumn sowing under European climates
Faba bean (Vicia faba L.) is an important grain legume widely grown as a spring crop to avoid frost damage. However, there is interest in winter types for the expected benefits in grain yield as compared to spring ones. In the current experiments, we compared field performance in autumn sowings of 15 faba bean winter-type cultivars that were sown in two consecutive autumns in 12 climatically contrasting sites in Austria, France, Germany, Spain, and the UK. GGE biplot analyses (genotype plus genotype-byenvironment interaction) were conducted to evaluate yield performance of faba bean genotypes and identification of mega-environments. Crossover genotypeenvironment was large and mainly due to the geoclimatic area. GGE biplot allowed identification of three mega-environments, namely continental, oceanic, and Mediterranean. Due to the climatic diversity of the environments, no cultivar performed well in all environments. Cultivars Clipper, Castel, Target, Wizard, and Gabl-107 performed well in oceanic mega-environment, whereas cultivars Castel, HIX, and Target performed well in continental mega-environment. None of the studied cultivars were suited to Mediterranean environments, and only Irena was able to give some modest yield at Cordoba. The average tester coordinate (defined by the average of first and second principal components of all environments) allowed to evaluate cultivars for their yielding ability and stability and to evaluate environments for their discriminating ability and to be more representative of the mega-environment. Thus, Wizard and Gabl-107 were the highest yielding cultivars being relatively stable over oceanic and continental environments. In contrast, the cultivars Irena and Divine yielded poorly at all environments. The results support the specific breeding for each major geoclimatic zone based on distinct genetic bases and selection environments.© INRA and Springer-Verlag, France 2012. ; Financial support by the European Union EUFABA project (ref. QLK5-CT2002-02307) is acknowledged. ; Peer Reviewed