Development of Selection Criteria and Strategies for Organic Rice Breeding

Organic agriculture is still largely dependent on the varieties produced in the conventional system, which may or may not be suitable for the practice. A two-year study was conducted to assess the performance of various rice genotypes under different input systems [conventional (CS), organic (OS), and zero (ZS)] and in various farmers’ organic practices. The study examined genotypes by organic management practices (GxE) interaction and investigated the need for a separate breeding program for organic rice. This study is very useful to plant breeders in developing rice varieties suited for organic production system. Analysis of grain yield revealed a highly significant genotypic variance in each input system and GxE variance. The response of genotypes differs on each input system. Genotypic and phenotypic correlation analysis between OS and CS showed low correlation. The low indirect correlated response implies that selection for organic rice is effective if conducted under OS rather than CS. These support the need for a separate breeding program for organic rice varieties. The GxE interaction for grain yield across farmers’ organic management practices showed a highly significant effect, indicating a very variable response of genotypes across management practices, which emphasized the importance of developing varieties specific for the unique practices of organic farmers. Direct phenotypic selection for grain yield would not be very effective due to the moderately low broad sense heritability (H) values (53–56%). Path analysis showed that the number of spikelets has the highest direct effect (0.70) to grain yield during the wet season, whereas plant height (0.47) and biomass (0.46) hold sway during the dry season. Direct phenotypic selection for these traits, given their high H value and genetic advance, will significantly contribute to grain yield but seasonal variation should be considered.