Comparison of Early Generation Yield Testing and a Single Seed Descent Procedure in Wheat Breeding

Abstract

Two breeding procedures were compared in two spring wheat ( Triticum aestivum L. em. Thell.) crosses. Both procedures were started from the same selected F 2 plants in each cross. In one procedure, a single seed was taken from each F 2 plant to produce the F 3 generation, the process was repeated for the F 4 . The F 5 lines were then produced from the seed of individual F 4 plants (single seed descent method, SSD). In the second procedure, the F 3 lines were yield‐tested and a pedigree method followed in the F 3 and F 4 (YT method). Each F 4 line was bulked to provide seed for a yield test in the F 5 . The F 6 SSD lines and F 5 YT lines were then compared in yield tests. The mean yield of the YT lines was significantly higher in both crosses, mainly because there were fewer, very low yielding YT lines. The correlations between the F 3 and F 5 yields of the YT lines were significant in both crosses, but were not particularly large (0.29 and 0.14). The yield of the F 5 YT lines, derived from the highest yielding 20% of the F 3 lines, was compared with the highest‐yielding 20% of the Fo SSD lines. In each cross, the SSD lines were at least as good as the YT lines. Although selection based on F 3 yields would have had some effect, it is doubtful that it was worth the labor involved. The SSD procedure appears to have considerable merit