LAUSR

The identification of stable and adaptable high yielding cowpeas ( Vigna unguiculata (L.) Walp.) and highly discriminative environments will be useful for elite cultivar development in South Africa. Two statistical models, the Additive main effects multiplicative interaction (AMMI) and the genotype, genotype by environment biplot analysis have been used extensively to identify superior genotypes and ideal testing environments. Hence, the objective of this study was to identify elite cowpea lines and testing environments using the two models to inform future cultivar development strategies. Fifteen cowpea genotypes were evaluated for yield performance and stability across 3 different locations during 2016 and 2017 growing seasons in South Africa. Genotype main effects were significant for grain yield, while genotype × environment interaction effect was insignificant for grain yield. However, genotypic perfomance for grain yield was significantly affected by seasonal variability. The AMMI analysis of variance showed that genotypes (G), environments (E) and their interaction were significant for grain yield. The G and GE effects accounted for about 10% of the total variation in grain yield, while the environment accounted for 66%. The high yielding genotypes, Vigna Onb, TVU-5431 and Kisumu mix, were adapted to Mafikeng and Potchefstroom sites. Kisumu mix, followed by Vigna Onb, were the most stable genotypes, while Veg cowpea 1, Veg cowpea dacama cream and Veg cowpea 2 were the least stable genotypes across the three sites and seasons. The analysis also showed that Environment 4 (Potchefstroom) was the most ideal site for cowpea production among the test sites.