HOST-RESISTANCE TO STEWARTS DISEASE IN MAIZE

The effect of host genotype, cytoplasm and rating date on development of Stewart''s wilt (Erwinia stewartii [E. F. Smith]Dye) of corn (Z. mays L.) was studied. The 15 possible single crosses among 6 inbred line parents in both normal and Texas male-sterile cytoplasm were grown in a split-plot design, with the split on cytoplasms. In a diallel analysis, inbred parents B37, 33-16 and Mo17 had significant negative general combining ability (gca) effects for each of 3 rating dates. The parents B14A and N28 had significant positive gca effects, with B14A conditioning susceptibility in a dominant manner. Reaction to Stewart''s disease was largely additive in this set of crosses, as the gca mean square was 20 times the magnitude of the specific combining ability (sca) mean square. The conventional recurrent selection method should be effective in improving populations for resistance to Stewart''s wilt disease. The most resistant plants could be identified before anthesis when artificially inoculated at about the 12th leaf stage. Highly significant differences were found among the 15 genotypes, 3 rating dates and 2 cytoplasms, the differences between cytoplasms being the least important. Disease response interactions were significant for genotypes .times. rating date, cytoplasm .times. rating date, gca .times. rating date, and sca .times. rating date. The genes conditioning lesion development were inherited qualitatively by the Castle-Wright formula. Diallel mating schemes using multiple disease ratings over time should be useful in studying gene action affecting disease development or general resistance.