Abstract
Pedigree of an actual pig strain developed by selection over 7 generations was analyzed, in order to clarify factors to be taken into consideration in establishing a method for predicting the amount of inbreeding and effective size in the process of pig strain development. In the pedigree analysis, the average inbreeding coefficient and coancestry were calculated in each generation, and the observed effective population size was obtained from the increasing rate of coancestry. The observed effective size was compared with the following two expected values : the effective population size expected under random selection (NeR), which was computed by assuming independent binomial distributions of parental family size in the four gametic pathways, and the expected value (NeS1) accounting for the effect of selection on the parental family size, which was obtained from the observed variance and covariance of parental family sizes. NeR overestimated the observed value over all generations. Although the harmonic mean of NeS1 over generations gave a closer value to the observed one, the effective population size in the later generations was seriously overestimated. These results indicate that the prediction only with the variance and covariance of parental family sizes leads to an overestimation of the effective size of population under continuous selection. A method accounting for the effect of selection on family sizes not only in the parental generation but also in the further ascending generations should be required for predicting the effective size and inbreeding in closed breeding strains of pig.
