Genotype by environment interactions for grain yield in upland breeding lines of rice under direct seeded conditions across multi-environments

Eighteen rainfed upland breeding lines were evaluated under four environments in randomized complete block design with three replications. Data on grain yield were subjected to combined ANOVA and AMMI analysis. The AMMI model was applied, with additive effects for the 18 breeding lines of rice genotypes (G) and four seasons of testing (Environments = E), and multiplicative term for genotype × environment interaction. The AMMI statistical model has shown that more than 50% proportion of the total variation in grain yield was attributed due to genotype × environment interaction. Most of the breeding lines showed environment specificity. CR2897-11-1 and CR28818-1 were found to be favorable during 2011 (En. 4) and CR2908-12-5-1 and CR2899-7-1 were found to be favorable in the year 2010 (En. 3) with grain yield more than 3 t ha–1.