EFFECTS OF MATING STRUCTURE IN PUREBRED POPULATIONS ON THE ESTIMATIONOF CROSSBREEDING PARAMETERS

Usually a simple additive-dominance model is considered to be adequate for the explanation of the genetic basis of heterosis. From this mode l a linear relationship is expected between the heterosis and the degr ee of heterozygosity of crossbreds. Assuming Hardy-Weinberg equilibriu m in parental populations, the increase in heterozygosity of the F1-ge neration can be calculated from the squared difference of the parental allele frequencies. This increase is halved in secondary crossbred ge nerations such as F2 or backcrosses. If the parental generations are s ubdivided into subpopulations, there will be whereas the allele freque ncies for the population as a whole remain unchanged. Therefore, for m ost traits the mean of such a subdivided population is expected to be reduced due to inbreeding depression. When two parent populations stru ctured into subunits are crossed, the reduction of heterosis in advanc ed crossbred generations will not follow the classical additive-domina nce model. The results of crossbreeding experiments with structured pa rent populations may then cause misleading genetic interpretations and may lead to wrong decisions concerning the expected efficiency of cro ssbreeding programmes. The paper deals with the parametrization used t o describe the performance of crossbred generations when reversible in breeding due to subdivided parental populations is present. To combine the effects of inbreeding and crossbreeding in a comprehensive model, the genetic model of Dickerson (1969) was extended by a direct inbree ding parameter and an epistasis parameter in backcross generations. Th is extended model was checked by a crossbreeding experiment with the J apanese quail.