Yield and stability of new cowpea varieties in Zimbabwe

Cowpea ( Vigna unguiculata [L.] Walp) is highly nutritious and fits well in smallholder crop-livestock farming systems of sub-Saharan Africa. Newly developed cowpea varieties need to be assessed for grain yield performance and stability across environments. Furthermore, a comprehensive understanding of the test environments is crucial in rationalising resources for breeding programmes. The objective of this study was to identify high yielding and stable cowpea genotypes, and determine the ideal (representative and discriminating) testing environments for cowpea in Zimbabwe. Fifteen cowpea genotypes were evaluated at 5 locations, for 2 years under rain-fed conditions. Combined analysis of variance for grain yield across locations and years showed highly significant (P < 0.001) genotype x environment interaction (GE). The variance component due to GE was 6 times larger than that of genotypes. Genotype plus GE biplot showed that environments studied were actually in one mega-environment; and Harare Research Station (HRS) was identified as desirable discriminating and representative environment. Genotypes G1, G11 and G3 were high yielding and more stable than the check varieties, G4 (CBC2) and G7 (CBC3). These genotypes are suitable for release; while HRS is recommended for culling inferior genotypes during early generation testing.Le ni\ue9b\ue9 ( Vigna unguiculata [L.] Walp) est hautement nutritif et bien adapt\ue9 au syst\ue8me de production associant culture-animal de l\u2019Afrique Sub-Saharienne. Des vari\ue9t\ue9s nouvellement d\ue9velopp\ue9esont besoin d\u2019\ueatre \ue9valu\ue9es pour leur performance en grain et leur stabilit\ue9 \ue0 travers des environnements. De plus, une bonne ma\ueetrise des environnements \ue0 tester en rationalisant les ressources pour les programmes d\u2019am\ue9lioration g\ue9n\ue9tique. L\u2019objective de cette \ue9tude \ue9tait d\u2019identifier les g\ue9notypes \ue0 haut rendement et stables, et de d\ue9terminer les environnements id\ue9aux (repr\ue9sentatifs et discriminants) pour le ni\ue9b\ue9 au Zimbabwe. Quinze g\ue9notypes du ni\ue9b\ue9 ont \ue9t\ue9 \ue9valu\ue9s dans 5 localit\ue9s, pendant deux ans sous des conditions pluviales. Analyse combin\ue9e de variance pour le rendement en grain \ue0 travers les\ua0localit\ue9s et ann\ue9es ont montr\ue9 une interaction g\ue9notype x environnement (GE) significative (P < 0.001). La composante de variance due \ue0 GE \ue9tait 6 fois plus grande que celle des g\ue9notypes. G\ue9notype plus GE biplot a montr\ue9 que les environnements \ue9tudi\ue9s \ue9taient \ue9videmment dans un m\ue9ga-environnement\ua0; et la station de recherche de Harare (HRS) \ue9tait identifi\ue9e comme un environnement d\ue9sirable discriminant et repr\ue9sentatif. Les g\ue9notypes G1, G11 et G3 \ue9taient \ue0 haut rendement et plus stable que les vari\ue9t\ue9s contr\uf4les, G4 (CBC2) et G7 (CBC3). Ces g\ue9notypes sont les plus appropri\ue9s pour d\ue9livrance, alors que HRS est recommand\ue9e pour faible s\ue9lection dans la g\ue9n\ue9ration pr\ue9coce