Forty rice (Oryza sativa L.) cultivars and breeding lines used in the International Rice Research Institute (IRRI) upland rice breeding program were evaluated in adjacent weed-free and weedy trials in aerobic soil conditions during the wet seasons of 2001, 2002, and 2003. The objectives of this study were to investigate genetic variability in weed suppression and yield and to identify traits that could be used as selection criteria for improved weed competitiveness. Correlations among and heritability (H) of agronomic traits and early vigor were estimated in weedy and weed-free trials. Regression analysis was performed to predict weedy yield and weed biomass. Cultivars differed widely in the growth of weed biomass they permitted (126¿296 g m¿2) and in yield under competition (0.5¿2.5 Mg ha¿1). Cultivar yield, duration, biomass, harvest index, height, and vegetative vigor under weed-free conditions were closely correlated with the same traits measured under weedy conditions. Weedy yield and weed biomass were both moderately heritable (H = 0.55 and 0.38 for means estimated from single-year, three replicate trial, respectively) and genetically correlated with each other (r = ¿0.84). Weed-free yield and vigor at two weeks after seeding (WAS) were moderately heritable (H = 0.68 and 0.38 for means estimated from a single-year, three replicate trial, respectively) and were highly genetically correlated with weedy yield (r = 1.00 and 0.88, respectively) and weed biomass (r = ¿0.89 and ¿0.67, respectively). Vegetative vigor at two WAS and grain yield measured under weed-free conditions explained a combined 87% of cultivar variation in weedy yield and 40% in weed biomass. Indirect selection on these two traits was predicted to be efficient for improving yield under weed competition and weed-suppressive ability of aerobic rice.
- junglerice echinochloa-colona
- upland rice
- suppression ability