PREDICTION OF GENOTYPE PERFORMANCE FROM ANCESTRAL RELATIONSHIP IN OAT

Coefficient of parentage, the probability that homologous alleles from different genotypes are identical by descent, has been used to estimate genetic similarity, The objective of this study was to determine if the performance of oat (Avena sativa L.) genotypes can be predicted by their coefficient of parentage (rp) values with a particular ancestor. Grain yield, test weight, and height data were gathered for 99 breeding lines from annual Uniform Midseason Oat Performance Nursery (UMOPN) reports. Each line was included in those trials for at least 2 yr from 1979 to 1989. Least-squares means across environments were derived for each line for each performance variable, from analysis of variance. Coefficients of parentage were calculated for each UMOPN line paired with each of 487 ancestral breeding lines, cultivars, and landraces. For 40 of those ancestors, UMOPN lines were categorized as closely related (0.125 less than or equal to rp < 1) or distantly related (0 < rp < 0.125). The means of the two parentage categories for each performance variable were compared by a two-tailed t-test. For six ancestors, closely-related UMOPN lines showed increased test weights and depressed grain yields, when compared to distantly-related lines. The cultivar Hercules enhanced the test weight of its close relatives, with no significant decrease in grain yield. UMOPN lines closely related to 'Ogle' or 'Tyler' had higher yields but lower test weights, compared to distantly-related lines. The greatest enhancement in mean yield of closely- versus distantly-related lines was 254 kg ha(-1) for 'Tippecanoe'. However, 'Clintland 60' appeared to be the most important ancestor contributing to improved yield without reduced test weight.