Simulating the potential of genomic selection in dairy cattle

Abstract

Genomic selection was simulated in a large dairy cattle population in comparison with a breeding scheme based on bull progeny-testing. The maximum increase of genetic gain (about 80% without cost increase) was obtained when bull sires, in much larger number, were selected only based on genomic information, without progeny-testing. In spite of a shorter generation interval, annual inbreeding and coancestry rates were decreased because these bull sires were much more numerous than in the reference scheme. The general recommendation to the AI industry is to pick-up a quite large batch of commercial sires (for breeding regular females) within the set of genomic bull sires, in total contrast with the conventional practice where few bull sires are a selected set of commercial sires, themselves in limited number because selected after progeny-testing. These results obtained in 2009 paved the way for a sustainable genomic selection of French dairy cattle, and are still up to date. Indeed, it is critical for selection and dairy cattle production sustainability to maintain a high enough effective size of the breeds to limit the inbreeding trend. An overview of the first years of activity shows that these recommendations have been well respected on average.