Heterosis, genetic effects and value of F(2)s and doubled-haploid lines inbarley breeding

To study the potential of F-1, F-2 and doubled-haploid lines in six-row bar ley, three connected experiments were developed: i) study of a complete 8 x 8 diallel; ii) comparison of F-1, F-2 and their parents for three crosses from the diallel; ii) comparison of 53 doubled-haploid lines from a cross t o their parents, F-1 and F-2. In the diallel experiment, the average of all F-1 for grain yield was equal to the average of the parents. The best cros s was only 3.2 % better than the best parent. However, for main ear grain w eight mean heterosis was 13 %. It was negative (-11 %) for straw yield. The re was a great variation among crosses. General combining ability (GCA) was a main component of such a genetic variation for all traits studied. Resis tance to mildew, thousand grain weight and heading date were the most addit ive and grain and straw yield the least additive traits. The study of F-2 s howed that, for grain yield, heterosis was reduced more than expected. This is interpreted as being mainly the result of intergenotypic competition be tween F-2 plants. However, epistasis is possible, and it was detected for y ield and its components, by the comparison of DH lines to their mid-parent. The comparison of DH lines with F-1 showed that the genetic advantage obta ined in F-1 for grain yield and its components can be fixed at the level of doubled-haploid lines. Nevertheless, the possible negative effect of recom bination and epistasis can lead to the use of haplodiploidisation after the F-1 generation. ((C) Inra/Elsevier, Paris.).