Genomic selection was introduced in dairy cattle breeding programmes in the 2010’s. This innovation accelerated genetic progress by increasing the accuracy of breeding values for young selection candidates and shortening generation intervals. However, since genomic selection was introduced, increases in inbreeding rates per year and per generation have been observed in dairy cattle populations. Loss of genetic diversity can lead to an increased frequency of harmful genotypes and inbreeding depression in traits important in dairy cattle breeding programmes. Moreover, loss of genetic diversity compromises a population’s ability to adapt to changing environmental conditions and selection objectives.
The aim of this thesis was to explore the impact of genomic selection on the genomic architecture of the Finnish Ayrshire dairy cattle breed. Specifically, the objectives were to assess the effect of genomic selection on genetic diversity, identify genomic regions under the influence of selection, and examine associations between genetic variants influenced by genomic selection and milk production and fertility traits.
The dataset comprised over 2.7 million Finnish Ayrshire with pedigree information and more than 77,000 Finnish Ayrshire with genotypes. Additional data included over 40,000 genotyped heifers and cows with records on milk production and fertility traits. The introduction of genomic selection was assumed to have occurred between 2012 and 2014. Inbreeding coefficients were estimated based on pedigree data and runs of homozygosity (ROHs). Genetic diversity was assessed based on inbreeding rate and effective population size using the inbreeding estimates. Genomic regions under the influence of selection were identified using complementary approaches: i) regions with high ROH frequency (ROH
islands), ii) regions with changes in ROH frequency (∆ROH islands), iii) decay of extended haplotype homozygosity, and iv) generation proxy selection mapping. Genetic differentiation following the introduction of genomic selection was assessed using the Hudson estimator of Wright’s Fst. Finally, a genome-wide association study was performed, including the genetic variants responding to genomic selection.
Average inbreeding was similar before, during, and after the introduction of genomic selection, ranging from 0.019 to 0.023 for pedigree-based estimates and from 0.055 to 0.057 for ROH-based estimates. However, an increasing trend in median ROH length and the frequency of ROHs over 10 Mb was observed during the genomic selection introduction, suggesting greater genomic relatedness due to selection and inbreeding. Subsequently, the
effective population size estimated with pedigree information decreased from 160–180 individuals to 140–150 individuals after genomic selection introduction. Despite this decline, the current effective population size of Finnish Ayrshire is above the Food and Agriculture Organization of the United Nations’ recommended minimum of 50–100 individuals. Selection signatures overlapped genomic regions and genes related to, for iv example, dairy-cattle milk production and milk composition, fertility, and growth. Thirteen genetic variants responding to genomic selection were associated with heifer fertility.
In conclusion, the effective population size of Finnish Ayrshire is at a sustainable level. Genomic selection has allowed efficient selection of genetic variants with desirable effects on heifer fertility. Nevertheless, genetic diversity in the Finnish Ayrshire is at risk. The effective population size is decreasing, and homozygosity arising from common ancestry is increasing in genomic regions related to, for example, milk production and fertility traits.
To maintain sufficient genetic diversity, it is essential to control relatedness among breeding individuals and monitor genomic inbreeding and ROH evolution. Optimal contribution selection maximizes genetic gain while restricting inbreeding increase. Future research should evaluate the feasibility of incorporating genomic information into optimal contribution selection, particularly accounting for increased homozygosity in specific genomic regions in the Finnish Ayrshire.
M.Sc. Katri Sarviaho, defended her doctoral dissertation entitled "The effect of genomic selection on the genetic architecture of Finnish Ayrshire cattle" at the Faculty of Agriculture and Forestry, University of Helsinkiin Raisio-Sali (B2), Latokartanonkaari 7 on February 20th, 2026, at 1 p.m. Dr. Martin Johnsson from SLU, Sweden served as the opponent and Professor Pekka Uimari as the custos.
The dissertation is available in electronic form in E-thesis service at