Effect of trait heritability and nucleus size on the bias of estimates for genetic gain due to ignoring the selection effect in open nucleus breeding schemes

An open nucleus breeding scheme consisting a nucleus herd of 500 ewes and a commercial herd of 4500 ewes was simulated deterministically using the information of the population of Lori Bakhtiari sheep to investigate the effect of heritability and nucleus size on the bias of estimates for genetic gain. The objective function was 9 months body weight and the selection was based on own record. Different values for heritability in nucleus and commercial herd and also different nucleus sizes were considered and the expected genetic gain was estimated by ignoring or accounting for the selection effect and from the ratio of these two estimates the amount of bias of estimates for genetic gain was obtained. The results showed that when the heritability of the nucleus and the commercial herd were equal, the highest bias (18%) occurred. Increasing the heritability in nucleus and commercial herd from 0.1 to 0.9 showed the amount of bias was directly related to heritability and ranged from 10 to 36 %. When the heritability in nucleus was fixed at a given value, and the heritability in commercial herd changed, the rate of bias altered, but only by 10 % indicating the amount of bias is not so sensitive to the heritability in the commercial herd. Results related to the changes in nucleus size indicated that the amount of bias changed between 10 to 20 %; therefore, effect of this factor on the bias is negligible. In general, the most important factor on the bias of genetic gain estimates, due to ignoring the effect of selection, is the heritability.