Evaluating Selection Strategies Based on Inbreeding Load via Simulation.

Abstract

Inbreeding load is the fraction of the mutation load that is due to recessive action of mutations. This load is only expressed in individual's inbred offspring. As such, inbreeding load contributes to genetic heterogeneity among individuals and can thus be used as a criterion of selection. Future parents could be selected based on their estimated breeding values and their estimated inbreeding load. Here, we evaluated the effectiveness of selection strategies involving inbreeding load in a dairy sheep breeding scheme. We evaluated seven strategies with a stochastic simulation across 15 generations of genetic evaluation and selection. The strategies differed in the criteria of selection (only estimated breeding values, only estimated inbreeding load, both or random) and mate strategies minimising inbreeding load, adjusting estimated breeding values by the expected future inbreeding or random. The strategies were compared in terms of genetic gain, pedigree-based inbreeding coefficients, rate of inbreeding, effective population size, inbreeding depression estimates, and accuracy of selection. Results showed that among the seven strategies, there was no significant difference between the five in which selection was based on estimated breeding values over the 15 years of the simulated dairy sheep breeding scheme. As expected, selection only on the estimated inbreeding load and the random selection and random mating strategy had significantly lower genetic gain than the five strategies in which selection was based on estimated breeding values. To conclude, selecting animals using estimated inbreeding load is feasible. However, the magnitude of inbreeding load effects and its accuracy did not show a clear practical interest. Furthermore, the benefit of using inbreeding load in mate allocation strategies was negligible.