Genetic analysis for drought tolerance and yield stability in interspecific and oryza sativa L. rice germplasm.
Rice is currently an important staple food crop in Kenya. However, consumption continues to outstrip production. In spite of this, the country is endowed with untapped virgin land for rice production under rainfed upland and lowland ecologies. Nonetheless, drought, lack of modern improved farmer preferred cultivars, poor agronomic practices, and continued cultivation of low yielding late maturing landraces, are among the major challenges to rice production and expansion in the country. The objectives of this study were to: 1) document farmers’ desired traits in rice cultivars and perceptions of rice production constraints in coastal region of Kenya, 2) identify drought tolerant rice genotypes at reproductive growth stage among the popular landraces, local cultivars, and exotic interspecific and Oryza sativa L. lines, 3) determine the inheritance of earliness and combining ability effects for phenological and morphological traits in rice under drought and no drought stress conditions, 4) determine the combining ability effects for grain yield and related traits in rice under drought and no drought stress conditions, 5) assess the heritability, correlation and the direct and indirect effect of phenological, morphological and yield component characters on grain yield and 6) estimate the magnitude of genotype x environment interaction (GEI) for grain yield in rice. The study period was between January 2013 and March 2015. Farmers’ desired traits in rice cultivars and perceptions of rice production constraints in coastal region of Kenya were assessed using formal household survey and participatory rural appraisal (PRA) methodology during 2013 and 2014. Data were collected from Msambweni Sub-county of Kwale County and Kaloleni sub-County of Kilifi County of coastal region of Kenya from a total of 326 respondents. The results established that farmers preferred high yielding, short duration and drought tolerant cultivars of medium height with white, long and bold grains. Preference for cultivars with good baking qualities was one of the unique traits that featured in this study. Important traits for a variety with good baking qualities were; white milled rice flour with low fat content, dough easy to work on, porous and does not stick on the baking tin while baking. Drought was ranked as the most important constraint and drought stress occurring at reproductive and grain filling stage was the most prevalent. These findings reveal that an opportunity exists in the coastal region to breed for high yielding, early maturing drought tolerant cultivars with white, long, bold, tasty and aromatic grains, good for confectionery purposes. Genetic variability for drought tolerance at reproductive stage was assessed among 21 rice genotypes comprising of 6 interspecific and 15 Oryza sativa genotypes. These were evaluated at KALRO-Mtwapa in pot experiments under two conditions of no drought and drought between April 2013 and March 2014. For each treatment, the experimental design was randomized complete block design with four replications and the plot size made up of 10 pots. Data collection included canopy temperature, relative leaf water content, leaf rolling and drying, days to 50% flowering, spikelet fertility and grain yield per plant. Considerable genetic variability for drought tolerance at reproductive stage existed among the interspecific and Oryza sativa L. rice lines. Two local Oryza sativa cultivars, Shingo la Mjakazi and Kitumbo were moderately drought tolerant. The interspecific genotype CT16323-CA-25-M was highly drought tolerant while NERICA 2 was moderately tolerant. These genotypes were identified as potential donors for drought tolerance at reproductive stage and may be used in breeding programmes aimed at developing drought tolerant cultivars for the rainfed lowland and upland ecologies in sub-Saharan Africa. Inheritance and combining ability effects for earliness, grain yield and its contributing traits were studied using 10 x 10 half diallel mating design. The 10 parents included five interspecific and five O. sativa L. lines. The resulting F1 progenies were advanced to F3 generation. The 45 F3 populations, 10 parents and one check were evaluated in 7 x 8 alpha lattice design with two replications under three no drought and one random managed drought stress conditions at three sites in coastal region of Kenya. Inheritance of earliness was found to be conditioned by non-additive gene action under random drought conditions and additive gene action under no drought conditions. One interspecific line CT16323-CA-25-M and one O. sativa line, Vandana, consistently exhibited desirable general combining ability for earliness under drought and no drought conditions. In addition, Vandana, was a good general combiner for grain yield and spikelet fertility under no drought conditions. Across environments, the line Dourado precoce had the best general combining ability effect for a thousand grain weight. The interspecific line, NERICA 2, was the best for number of grains per panicle while NERICA 1 had the best general combining ability effects for heavy panicle weight and weight of grains per panicle, and good grain phenotypic acceptability. Therefore these parents (CT16323-CA-25-M, Vandana, Dourado precoce, NERICA 2 and NERICA 1) may be hybridized with the intent of selecting promising genotypes within the segregating generations. The best F3 populations with desirable specific combining ability effects were CT16323-CA-25-M x Vandana and Duorado x Vandana combining short duration with increased plant height and higher tiller number and NERICA 1 x NERICA 2 combining moderate drought tolerance index with desirable alleles for high yields, high thousand grain weight, heavy panicle weight, heavy grains per panicle and a good grain phenotypic acceptability. Early generation testing in these crosses was recommended to identify plants with desirable characters that may be advanced to homozygosity followed by selection of best pure lines for release in the region. The narrow sense heritability estimates for earliness based on days to heading was high (67%), indicating predominance of additive gene action while that for grain yield was low (0.1%) implying predominance of non-additive gene action. Direct effects on grain yield were significant and positive for number of productive tillers per plant (P = 0.71), panicle weight (P = 0.66) and spikelet fertility (P = 0.49). However, the heritability estimates for number of productive tillers per plant (29%) were moderate, and low for panicle weight (0.7%) and spikelet fertility (4%) limiting their use in early generation selections. A thousand grain weight had a high narrow sense heritability (82%) and positive indirect effect (P =0.44) on grain yield via panicle weight indicating that improvement of grain yield may begin in early generations by indirectly selecting for high thousand grain weight via heavy panicle weight. The stability analysis of the 45 F3 populations and their parents over four environments using the AMMI and GGE biplot models showed that ranking of the genotypes changed across environments. This revealed a crossover type of genotype by environment interaction. Both AMMI and GGE biplot analyses showed that the four environments fell into three mega environments and identified G37 (Luyin 46 x IR55423-01) as the most high yielding genotype. However they differed on the most stable and high yielding genotype across the test environments. The AMMI analysis showed that G41 (NERICA-L-25 x Vandana) followed by G1 (NERICA 1 x NERICA 2) and G34 (CT16323-CA-25-M x Vandana) were the most stable and high yielding genotypes. In contrast, the GGE biplot showed that G39 (Luyin 46 x IR74371-54-1-1) followed by G40 (NERICA-L-25 x IR55423-01) were the most stable and high yielding genotypes. Overall, this study provided valuable information that will help in setting and prioritization of breeding goals and objectives aimed at breeding for early maturing, farmer preferred cultivars, tolerant to drought stress at rice reproductive stage.