Genetic improvement in Coho salmon (Oncorhynchus kisutch). I: Selection response and inbreeding depression on harvest weight

In this study the genetic response to selection and the effects of inbreeding on harvest weight in two populations (which spawned as even and odd years classes) of Coho salmon are described. Artificial selection was performed using the best linear unbiased prediction (BLUP) of breeding values obtained from an animal model. Both populations were also selected for early spawning date using independent culling levels; these results are presented in a companion paper. The estimation of genetic parameters was based on the phenotypic records of 12,208 animals randomly sampled at harvest time (random group). These fish represented a mean of 64% and 48% of all animals with phenotypic records in the even and odd year classes, respectively. The narrow sense heritability estimated for harvest weight was high and very similar in the two populations, 0.39 ± 0.03 in the even year class and 0.40 ± 0.04 in the odd year class. Due to the mating design, the genetic selection differentials were 2.3 times greater in males than in females. The mean genetic selection response obtained was high in both populations, 383.2 g or + 1.26 σ and 302.4 g or + 1.23 σ per generation in the even and odd year classes, respectively. This is equivalent, on the average, to an increase in weight of 13.9% per generation compared to the base population, or 10.2% per generation with respect to the difference between successive generations. After the 4th generation of selection the mean inbreeding level was 9.5% in the even year class and 4.4% in the odd year class, which are close to preliminary estimates based on the records of the random group plus a group of animals with high harvest weight (Gallardo, J.A., García, X., Lhorente, J.P., Neira, R., 2004b. Effects of nonrandom mating schemes to delay the inbreeding accumulation in cultured populations of Coho salmon (Oncorhynchus kisutch). Can. J. Fish. Aqua. Sci. 61, 547-553.). The estimated effects of inbreeding depression on harvest weight were negative, but not significantly different from 0 in either the even year class (b_f = - 7.04 g/ΔF, s.e. = 3.9; Mean = 4118 g; inbreeding depression per 10% of ΔF was - 1.7%; P = 0.06) or the odd year class (b_f = - 4.8 g/ΔF, s.e. = 0.33; Mean = 3243 g; inbreeding depression per 10% of ΔF was - 1.5%; P = 0.33). Other programs in salmon have selected for more than one character, however, there are still no estimates of economic weights published in salmon breeding programs, which are necessary to calculate the total genetic merit for these animals.