Abstract
Plant breeders evaluate their germplasm/ accessions in multi-environments to study their performance and adaptation to a specific or general environment. The problem of genotype-by-environment (GxE) interactions in the interpretation of multilocational trial analysis and the prediction of genotype performance can be eased with the adoption of the Additive Main Effects and Multiplicative Interaction (AMMI) model analysis. The AMMI model was used in the present study to evaluate 13 promising hybrid rice genotypes established at five different locations in Bangladesh to identify stable and adaptable genotypes, determine the magnitude of GxE interaction and identify factors contributing to the GxE interaction pattern. Analysis of variance showed that the effects of environments, genotypes and GxE were highly significant for plant height, days to maturity, panicle length and grain yield. BRRI1A/BRRI12R, II32A/BRRI10R and the standard check variety BRRI dhan29, which display a semi dwarf plant stature, were stable over all the environments. The promising hybrids BRRI1A/BRRI12R, BRRI10A/BRRI15R, BRRI9A/BRRI15R and BRRI hybrid dhan3 were stable across the environments with shorter growth duration. The hybrids II32A/BRRI15R, II32A/BRRI16R II32A/BRRI10R and BRRI hybrid dhan3 were also stable over all the environments with long panicle. Genotypes BRRI10A/BRRI12R, II32A/BRRI15R, II32A/BRRI10R, BRRI9A/BRRI15R, BRRI hybrid dhan2 and BRRI hybrid dhan3 were high-yielding and stable. While, the genotypes II32A/BRRI16R and II32A/BRRI12R were high-yielding but unstable. Wide variations in soil moisture availability were identified as a major causal factor of the interactions observed. Among the five locations, Barisal could be considered to be the optimum site for the selection of broad-based improved hybrid rice genotypes.
