Clarification and Reevaluation of Population-Based Diallel Analyses

摘要: Plant breeders and geneticists often use diallel mating designs to obtain genetic information about a trait of interest from fixed or randomly chosen set parental lines. Diallel analyses broad-based populations have frequently been conducted by means three presented Gardner Eberhart in 1966. The original paper used sequential model fitting estimates effects corresponding sums squares. This approach, although having long history, suffers shortcomings which led confusion what hypotheses the actually test. objectives this were delineate clearly all models implicitly required perform Analyses II III, present explicit formulas for terms population are fundamental unambiguous. While developing effects, we discovered typographic error associated with variety example Analysis II. Our results also indicate that III effect rather nonintuitive both biologically genetically, incorporate multipliers functions number parents. Another specific result shows varietal obtained Eberhart's unconstrained estimates, while those constrained assumption no heterosis. These implications interpretation such effects. A SAS computer program analyzing diallels among according is reported.