Evaluation of Genotype x Environment Interaction and Stability of Grain Yield and Related Yield Components in Pearl Millet (Pennisetum glaucum (L.) R.Br.)

Thirty six pearl millet genotypes were evaluated in randomized complete block design with two \nreplications during 2011/2012 at two locations to study the magnitude of genotype by environment \ninteraction for yield and yield related traits and identify the most stable high yielding genotypes. \nANOVA of data at individual locations revealed significant differences among genotypes at Marigat \nand Koibatek for all yield components. Combined mean analysis of variance showed that the Genotype \nand location main effects and the genotype by environment interaction were highly \nsignificant (P≤0.01) for grain yield and other traits, indicating differential response of genotypes \nacross testing locations and the need for stability analysis. Marigat was the most suitable \nenvironment and gave highest mean grain yield of 3620 kg/ha. The lowest yield 870 Kg/ha was \nrecorded at Koibatek. Genotypes EUP 32, EUP 35, EUP 19 and EUP 10 produced high mean yield \nof 3530, 3080, 2690 and 2590 kg/ha respectively. The lowest grain 1290 kg/ha was obtained from \ngenotype EUP 4.Based on the parameters of stability, three stable (widely adapted) and high \nyielding genotypes (EUP 34, EUP 18, and EUP 9) were identified. They also out-yield the standard \nopen pollinated variety (OPV) check, Kat PM2. Genotypes EUP 32 was the highest yielding across \nall sites followed by EUP 35 and could be recommended for further multi-location evaluation in \nwarmer environment and possible release for commercial production. The findings of this study \nshowed that pearl millet hybrids have high potential for commercial production in Kenya than the \nOPVs. The ANOVA results showed that the effects of environments, genotypes and genotype x \nenvironment interaction (GE) were important in trait expression and performance of genotypes. In \naddition, it was observed that amount of rainfall received at both vegetative and post-anthesis \nphases and temperature had an effect on grain yield. The GGE biplot analysis characterised the \nenvironments in terms of stability and productivity, where Marigat was the best for grain yield; \nimplying that environment-specific selection should be adopted. Genotypes EUP 34, EUP 18, and \nEUP 9 were the best performing since they out yielded the standard OPV check. These stable high \nyielding genotypes can be evaluated further in varied agro-ecologies and recommended for release \nas commercial hybrid varieties in ASALs of Kenya.