Genotype-by-environment interaction analysis of rice (Oryza spp.) yield in a floodplain ecosystem in West Africa

Abstract

Abstract West Africa has large areas of river floodplains, most of which are not currently used for farmland. Rice ( Oryza spp.) is a promising crop for farming in floodplains because of its high adaptability to a wide range of water environments. On the other hand, there is great variation in soil fertility and water availability even in a small area within a floodplain. Hence, we evaluated 27 rice genotypes in four fields in three years in a floodplain of the Northern Region of Ghana to investigate genotype × environment (G × E) interactions for rice yield and to identify stable, high-yielding genotypes. The genotypes consisted of O. sativa , O. glaberrima and New Rice for Africa (NERICA), and many were selected for their reported submergence resistance because of the anticipated submergence damage in the floodplain. There were large variations in yield, which ranged from 0.14 to 5.35 t ha −1 depending on the location within a floodplain, genotype and year, and there were significant genotype, environment and G × E interaction effects on yield, accounting for 24.8%, 20.2%, and 28.2%, respectively, of the total variation. The results suggested that selection of suitable location with high soil fertility and low risk of submergence is necessary to achieve high yield in a floodplain. In addition, early sowing would be effective high-yielding crop management, which reduced the risk of submergence-induced damage just after sowing and secured sufficient growth duration to achieve high yield. Genotype IR42 showed the highest average yield among environments, but its yield stability was low. On the other hand, several genotypes including Amankwatia, a local aromatic cultivar adapted to irrigated and lowland environments, and IRBL9-W[RL], a blast-tolerant variety containing the Sub1 gene for submergence tolerance, showed high, stable yield. To put these results to practical use in other floodplain areas in West Africa, physiological mechanisms causing G × E interaction for rice yield should be further studied.