Diallel Analysis of Bacterial Wilt Resistance in Tomato Derived from Different Sources.

Bacterial wilt, caused by Ralstonia solanacearum, is a major constraint to tomato production in the tropics and subtropics. Most bacterial wilt-resistant tomato cultivars have not shown consistently high resistance levels over locations. The objective of this study was to determine whether combining resistance derived from different sources would result in F1 progenies with resistance greater than that of the parents. Five bacterial wilt-resistant tomato lines or accessions (CL5915, L285, CRA84, H7997, and GA219), each derived from different resistance sources, and a susceptible processing tomato line (UC204A) were crossed in all combinations without reciprocals. Parents, F1 progenies, and F2 progenies were evaluated in greenhouses at three locations (Taiwan, Philippines, and Indonesia) for percent survival 6 weeks after drench inoculation with virulent local strains of R. solanacearum. Percent survival means over locations were 17.4 to 83.0 for parents and F1 progeny and 16.2 to 75.0 for parents and F2 progeny. The percent survival means over locations of L285 × H7997 were highest among crosses in the F1 (83.0) and F2 (75.0) generations but were not significantly greater than that of H7997. Highly significant mean squares were found in the F1 and F2 progenies for general combining ability (GCA) and GCA × locations. Positive GCA effects over locations were detected for H7997, CRA84, and L285, indicating that progeny with those lines as parents showed bacterial wilt resistance that was greater than the average of all crosses. Only H7997, however, had positive GCA effects estimates at each location for each generation, and its GCA effects estimates over locations were significantly greater than those of the other parents in the F1 and F2 progenies. Among this set of parents, H7997 is the best source to develop bacterial wilt-resistant progeny. We did not observe statistically significant increases in resistance by combining different resistance sources. However, the presence of large GCA variances suggests that hybridization of parents that have high GCA for bacterial wilt resistance, such as H7997, CRA84, or L285, followed by selection in segregating populations might yield inbred progeny with resistance greater than that of the parents.