Derivation and Use of Genomic Enhanced Estimated Breeding Values for Selection in Sheep

The efficacy of selection programs in sheep depends on accurate evaluation of genetic merit. Globally, genetic evaluations are based on best linear unbiased prediction (BLUP) in which fixed effects and breeding values are estimated simultaneously. With genotyping becoming increasingly routine, pedigree, performance, and genomic data are being combined in single-step genomic BLUP (ssGBLUP) evaluations to generate genomic enhanced estimated breeding values (GEBV). Since GEBV are more accurate, younger animals can be selected more reliably; a corollary benefit is a shortened generation interval. Traits that are lowly heritable, expressed in one sex, measured late in life, or difficult and/or expensive to measure benefit most from genomic evaluation. Adoption of this technology has lagged in sheep most likely due to the higher cost of genotyping compared to the value of an animal. Nevertheless, genomics has been successfully incorporated in sheep breeding programs in several countries (e.g., Australia, France, New Zealand). Increases in accuracy of EBV of up to 60% have been reported. In 2021, the efficacy of genomic selection in a U.S. sheep breed (Katahdin) was tested by comparing predictive ability and bias of EBV using ssGBLUP and pedigree BLUP for weight and fecal egg counts at weaning. Approximately 5,000 animals, most born within the last 5 years, were genotyped with a 50k array. Nearly 70k body weights and 14k fecal egg counts were available. Genomic information improved predictive ability for both traits by as much as 10%, and reduced bias in the evaluation of fecal egg counts. Accuracies of EBV improved appreciably in younger genotyped animals (by as much as 1.81 times). An outcome was release of GEBV in the evaluation of U.S. Katahdin sheep. With integration of genomics into genetic evaluation, there is clear opportunity to augment the reliability of genetic evaluation and thereby rates of selection response in sheep.