The experiments were conducted in the experimental area of the Department of Crop Production of the Faculty of Agronomy and Veterinary Sciences, São Paulo State University, Jaboticabal Campus. Table I shows the relationship between the cultivars used in the crosses and the reactions of each one to powdery mildew.
Enough seeds were cultivated from F2 to obtain about 60 F2 plants per cross, so that the evaluations of powdery mildew resistance could be completed. The seeds were cultivated in pots from March through September 2000, the most appropriate period for studying powdery mildew resistance. The plants were inoculated naturally as follows: right after sowing, pots containing plants of highly susceptible varieties (BR-16 and IAC Foscarim 31) and with powdery mildew symptoms were placed among the pots of the studied plants, so that natural infection occurred rapidly and regularly.
Two evaluations were performed in the populations, one 20 days after sowing (CV1) and the other after flowering (CV2). Plants were scored for diseases according to a 0-to-5 grade scale, as proposed by Yorinori (1997), as follows: 0 = no leaf symptoms, 1 = 10% of the leaf surface with symptoms, 2 = 11 to 25% of the leaf surface with symptoms, 3 = 26 to 50% of the leaf surface with symptoms, 4 = 51 to 75% of the leaf surface with symptoms, 5 = more than 75% of the leaf surface with symptoms. In the present study, plants scored as grades 0, 1, 2, and 3 were considered resistant (R) and plants scored as grades 4 and 5 were considered susceptible (S).
The chi-square test (c2) was used to study the genetics of resistance to powdery mildew for a 3:1 segregation, as suggested by Snedecor and Cochran (1989) and Ramalho et al. (1989), as follows:
where F0 = Observed frequency and FE = expected frequency.
The heterogeneity test was applied to the results of the chi-square test for the eight populations studied to check the consistency of the same results for the eight populations. The procedure was carried out by the method by Mather (1957).