Doubled haploids for studying the inheritance of quantitative characters.

By using a doubled-haploid population derived from F2 plants, additive and additive X additive genetic variances, as well as the number of segregating genes, can be estimated. An F2-derived doubled-haploid population may contain almost 50% more of the best recombinant than an F1-derived population. However, the best recombinant occurs in the same frequency in the two populations when there is no linkage between genes. The difference in the frequency of the best recombinant between F2- and F3-derived populations is small. This implies that the doubled-haploid method using F2 plants provides only slightly less opportunity for recombination than the conventional breeding methods of self-pollinating crops. In the absence of additive epistasis, a weighted mean of recombination values can be estimated using an F2-derived population and its parental lines. When additive epistasis is present, it can be estimated from doubled-haploid populations derived from two backcrosses. Studies on the linkage of quantitative characters are needed for determining whether doubled haploids should be produced from F2 or from F1 plants in a breeding program.