Simulating genetic progress for traits with expensive phenotyping

Abstract

Phenotyping costs in dairy cattle breeding exhibit significant variability across traits. While milk production is recorded routinely at only low costs, traits such as feed efficiency and methane emissions pose challenges due to their expensive measurement requirements. This study leveraged the real-size digital twin of the Geno breeding program for the Norwegian Red dairy cattle breed to simulate genetic progress following ten years of selective breeding, particularly targeting traits demanding costly phenotyping. Multiple scenarios were simulated, varying in the number of phenotypes recorded, economic weight, and genetic correlation between the trait and total merit index. Our results highlight the importance of genetic correlation in achieving progress for traits with expensive phenotypes recorded at a limited scale. Increasing economic weight and the number of phenotypes increased genetic progress. Thus, there is an indirect indication that traits with low phenotyping costs and high correlation to expensive phenotypes should be prioritized when selecting for genetic improvement of a trait with expensive phenotypes. However, precise phenotypes are required for accurately estimating genetic correlations between traits with expensive phenotypes and traits with cheap phenotyping.