Grain yield stability of cowpea genotypes in contrasting environmental conditions in Southern Mozambique

In Southern region of Mozambique crop yields are very unpredictable mostly due to high intra and inter-seasonal and spatial rainfall variability. Assessing yield stability in this region is crucial for identifying best performing genotypes across the region and best environment for selection. A study was conducted to assess yield stability of 24 cowpea genotypes to variable environmental conditions in Southern region of Mozambique. The experiments were alpha-lattice designs arranged in 4 rows and 6 columns with 4 replications. Detection of GxE was done using the Residual Maximum Likelihood and its quantification and stability analysis were performed using GGE biplot from GenStat 15.0. Genotype by environment interaction was present and it was two times higher than the genotype effect. This GxE was of cross-over type indicating that genotypes performed differently to the change in environment. Genotypes IT00K-96, UC-CB46 e IT98K-1105-5 were high yielding and adapted to all environments. Of the total of environments used to evaluate varieties, Nhacoongo was both discriminative and representative environment. The environments on which the genotypes were evaluated were grouped into two mega-environments; the first mega-environment comprised Nhacoongo and Inharrime, and the second comprised Bilene, Chimondzo and Ricatla. The study reveled that in Southern Mozambique, yield can be predicted using genotypes such as IT00K-96, UC-CB46 and IT98K-1105-5. In addition, selection should be conducted at Nhacoongo.