Genotype by environment analysis of rice (Oryza sativa L.) populations under drought stressed and well-watered environments

In Kenya, the key abiotic stress affecting rice production is drought stress which is experienced mainly during reproductive phase of the crop. This study evaluated the performance of Kenyan rice populations including 19 generation of crosses, 6 parental and 5 checks under well-watered and drought stressed environment with the aim of identifying the phenotypic traits that confer drought tolerance in rice. The 19 generation of crosses were in F3 when they were evaluated in the short rains season and later advanced to F4 in the long rains season. Nineteen generation of crosses rice (Oryza sativa L.) lines, six parental and five check lines were evaluated for response to drought under drought stressed and well-watered environment. The study was conducted over two seasons in the year 2016/2017 at Kenya Agricultural Livestock Research Organization (KALRO) -Mwea Centre. The experiment was set up in an alpha lattice design with three replications. Drought stress was imposed at panicle initiation by withholding irrigation till physiological maturity meanwhile the well-watered environment continued to enjoy the recommended irrigation regime from planting to physiological maturity. Yield data were scored for drought tolerance. AMMI analysis of variance for grain yield showed that genotypes from crosses of crosses of SARO5XNERICA11, NERICA2XSARO5 and NERICA15XSARO5 expressed high grain yield. AMMI stability Variance (ASV) showed genotypes NERICA15, Duorado Precoce and progenies from crosses of NERICA11XNERICA2, SARO5XKomboka and NERICA2XNERICA11 expressed high stability in both well-watered and drought-stressed environment. GGE analysis showed that Principal Components (PC1) and PC2 accounted for 96.46% and 3.54%, respectively. GGE biplots showed that genotypes from crosses SARO5XNERICA11 and NERICA15XSARO5 were the most stable and high yielding. GGE biplots ranked the rice lines as follows: those above average in performance, the stable ones, unstable and those below average in performance. SARO5XNERICA11 is worth of selection due to its high mean yield value and is stable across the mega-environments.