Comparison of selection indices to screen lowland rice for drought resistance

Summary Although many selection indices have been used to screen rices (Oryza sativa L.) for drought resistance, there has been little comparison of the relative merits of these indices. Research was conducted to compare drought resistance as estimated from grain yields, canopy-temperature-based stress indices, visual scoring, and uprooting force for 30 rice genotypes grown in the field with a puddled Maahas clay (Typic Tropaquept) and to evaluate traits related to drought resistance from nonstressed plants grown in the field and in aeroponic culture. Water deficit was imposed in the field by withholding irrigation from 45 to 75 days after transplanting compared to a continuously flooded control. Grain yields in the stress treatment were most strongly correlated with visual assessment of drought stress symptoms according to a standard evaluation system (r = 0.66). Canopy-temperature-based indices were also significantly correlated with grain yields of the stress treatment (r from –0.55 to –0.63). No trait of aeroponically grown plants was correlated with traits of stressed plants in the field. We conclude that visual scoring of stressed plants was the best method of screening for drought resistance, but if controlled water deficit cannot be imposed, then drought resistance may be estimated by measuring both uprooting force and grain yield.