Correlated genetic trend in the environmental variability of weight traits in mice

Genetic trends for weight traits and their environmental variability were analysed in an experimental mice population selected during 17 generations to increase weight gain by comparing three selection methods: classic selection with random mating (Method A), classic weighted selection with random mating (Method B) and classic selection with minimum coancestry mating (Method C). Males were selected based on their own phenotypic records for WG. The analysis involved three traits: weight at 21 days (W21), weight at 42 days (W42) and weight gain between 21 and 42 days (WG). Genetic trends were obtained by averaging, within generations, the breeding values obtained for the traits and their environmental variability under a classical animal model assuming that the environmental variance is homogeneous and an alternative model assuming the heterogeneous environmental variance is partly under genetic control. All the genetic trends were positive for the traits and negative for their environmental variability but the trend in phenotypic variances was steady showing that the model analysing the environmental variability failed to separate correctly the genetic from the systematic effects under an artificial selection scenario. The higher additive genetic variance estimated under the heterogeneity model led to higher genetic trends when this model was used, thus changing the order of the preferred methods of selection moving Method B form intermediate to be the worst. The results also showed that correlated changes in the variability of weight gain and related traits originated as a consequence of selection process in the trait, but these changes do not seem to be unfavourable for the animals since the scale effect tends to compensate the correlated reduction in variability of these traits.