Comparison of Testcrosses Produced from F2 and First Backcross Populations in Maize

Abstract

Single crosses and backcrosses from established lines are importeat parental sources for developing new inbred lines in hybrid maize (Zea mays L.) breeding. The objectives of this study were (i) compare, both theoretically and experimentally, the testcross properties of F2 vs. first backcross populations; and (ii) to outline criteria for choosing between these two types of base populations in recycling breeding programs. Two homozygous lines (P1, P2) and individual plants from their F2 and first backcross (BC1, BC2) generations were mated to an unrelated single‐cross tester. A total of 186 testcross progenies were evaluated for grain and forage traits at three and two locations, respectively. The five generations (P1, P2, F2, BC1, BC2) differed significantly (P < 0.05) in their testcross means (X̄) for grain (GY) and forage (FY) yield. A nonepistatic model accounted for 92 to 99% of the variation among testcross generation means for five of the six traits examined, yet the presence of epistasis was indicated in several instances. In accordance with theory, the ratio of BC (the pooled estimate of BC1 and BC2) to F2 genotypic variance among testcross progenies (σ2g) was close to 0.50 for all traits but one. Heterogeneous σ2g estimates in BC1 and BC2 were found for GY, FY, and grain dry matter content, suggesting the presence of epistasis. Predicted selection response for testcross performance (ΔG) was greatest in the F2. Testcross progenies of the F2, BC1, and BC2 generations displayed similar phenotypic and genotypic correlations as expected by theory. The “usefulness” criterion U = x̄ + ΔG was employed to assess the breeding prospects of each generation. With weak selection, the better backcross had superior usefulness for four traits, whereas with strong selection the F2 had highest usefulness for all traits except GY.