Use of recombinant inbred lines of wheat for study of associations of high-molecular-weight glutenin subunit alleles to quantitative traits. 2. Milling and bread-baking quality

Recombinant inbred lines (RILs) derived by single plant descent to F8 from a hybrid of Anza, a low-quality cultivar, and Cajeme 71, a high-quality cultivar, differed in alleles at three high-molecular-weight glutenin (HMW-glu) seed storage protein loci. The 48 RILs were classified by SDS-PAGE for the Anza alleles Glu-Alc (null), Glu-B1b (subunits 7 + 8), and Glu-D1a (subunits 2 + 12) and for Cajeme 71 alleles Glu-A1a (subunit 1), Glu-B1i (subunits 17 + 18), and Glu-D1d (subunits 5 + 10). All RILs and parents were grown in a replicated field trial with three levels of nitrogen (N) fertilization. Additive and additive X additive gene effects for the three loci were detected by orthogonal comparisons of means for each of six wheat end-use quality traits. Each HMW-glu genotype was represented by three to ten RILs so that variability among RILs within each HMW-glu genotype could be examined. N effects were consistently small. All traits except flour yield were highly correlated with predictor traits studied earlier. Flour protein content, baking water absorption, dough mixing time, bread loaf volume, and bread loaf crumb score were all correlated, suggesting similar gene control for these traits; however, specific additive locus contributions were evident: alpha-B, for flour yield; alpha-B and alpha-D for flour protein; and alpha-B for absorption, but differing in sign; all three loci for mixing time, but alpha-B was negative; and all three loci were positively associated with loaf volume. Digenic epistatic effects were significant for flour yield (alpha-AD), flour protein (alpha-AB), and absorption and mixing time (alpha-AD, alpha-BD). Only flour yield showed a trigenic epistatic effect. Six of seven epistatic effects were negative, thus showing how progress in breeding for high quality may be impeded by interaction of genes which, by themselves, have strong positive additive effects. Considerable genetic variance among RILs within a HMW-glu genotype was detected for all traits, and the summation of cz effects accounted for a mean of 13% of the parental differences for the six traits examined in this study. Clearly, further resolution of the genetics of wheat quality would be desirable from a plant breeding point of view.