Yield Performance of the F2 Generation as a Predictor of Transgressive Segregation in Biparental Soybean Populations.

Past research has suggested that yield performance of the F2 generation of a biparental cross can be used to predict the productivity of lines that are derived with inbreeding.  This method of discrimination among possible breeding populations has not been widely used and never adequately tested by soybean breeders.  The objective of the experiments reported here was to determine whether inbred lines from biparental crosses that show F2 heterosis for yield have greater genetic variation and more transgressive segregation than lines from crosses with no heterosis.  Random F6 derived lines from six biparental crosses were tested together with their parents and a bulk of the F2 generation of each cross.  Of the six populations, the one with the largest estimate of genetic variance and the most number of transgressive segregating lines, was also the population with F2 performance significantly greater than the two parents.  If F2 bulk heterosis is due to dominant (i.e. more favorable) alleles and/or epistasis among those alleles, then that productivity can be fixed in an inbred line.  Thus, F2 heterosis could be used to successfully choose among a set of crosses, which to continue inbreeding and which to discard.