The genetic divergence of sixteen rice cultivars regarding their resistance to the rice stem bug, Tibraca limbativentris Stål, was estimated by multivariate analysis techniques. The experiment was carried out in greenhouse, in randomized block design with eight replications. Eight plant resistance related traits were evaluated. Genetic divergence was evaluated by multivariate procedures: generalized Mahalanobis (D²) distance, the Tocher's grouping optimization method and canonical variables. The most dissimilar cultivars were Bico Ganga and Marabá Branco, while Agulha and Branco Tardão were the most similar. Five clusters were formed by the Tocher's optimization method. Three canonic variables explained 88.5% of the observed variation. We concluded that the multivariate analysis techniques are suitable for analyzing the genetic divergence among rice cultivars, indicating Bico Ganga and Marabá Branco as the most promising for future breeding programs of resistance to the rice stem bug.
Plant breeding; rice stem bug; multivariate analysis
