Abstract

Chilean salmon culture is based on a high degree of artificial selection, which has had the tendency to increase the inbreeding (F). Three types of nonrandom mating were evaluated to control the inbreeding in two best linear unbiased prediction selected coho salmon (Oncorhynchus kisutch) populations (even and odd). These included compensatory mating on the basis of breeding values (C), modified compensatory mating (Cl) based on the family mean of breeding values, and mating that minimized the mean co-ancestry of the group selected (MC scheme). In the odd population, the MC scheme (F = 2.0%) reduced the increase in inbreeding of the next generation by 50% and 46% when compared with random mating of selected individuals with sib mating restricted (F = 3.9%) and with C (F = 3.7%), respectively. In the even population, the MC scheme reduced the increase in inbreeding by 14% compared with Cl (9.7 versus 11.2). In both populations, the MC scheme also reduced variance in inbreeding (even, 59%; odd, 39%). Thus, the MC scheme was more efficient in reducing the increase and variance of inbreeding, thus limiting the expression of inbreeding depression. Although the MC scheme was more time consuming, we recommend this scheme to carry out crosses in each generation.