The relation between wheat lines derived from the F2, F3, F4 and F5 generations for grain yield and harvest index

Summary

The relationships between the F2, F3, F4 and F5 generations for grain yield were determined using random, pedigreed lines derived from each generation. The lines from two crosses were grown in plots at two sites over two years. In the first year, only F2 and F3 derived lines were available, but in the second year the F2 to F5 were grown.

Correlations between lines in one generation and the mean of lines derived from them in a following generation increased as the generations were advanced. Correlations between consecutive generations were higher than those between generations two or three apart. Correlations between F2 and F5 derived lines, which indicate the effectiveness of selecting F2 lines, varied from 0.10 ns to o.49^** when lines from both generations were grown in the same environment. Correlations between years of lines from the same or different generations were low and often non-significant.

Harvest index was measured on the F2 and F3 derived lines at the one site in the first year. Selection for improvement of grain yield using harvest index was no more effective than selection for yield directly, when considered across years.

It is conclued that, while gains in yield can be achieved by selecting for yield in early generations, a foremost consideration needs to be the influence of different sites and years on the effectiveness of selection.