Dairy cattle mating plans at herd level using genomic information

Genotyping of dairy cattle can benefit farmers by increasing the accuracy of breeding values and improving mating plans at herd level. Validation studies on breeding values in this thesis revealed that for the vast majority of traits analysed in Holstein, Jersey and Nordic Red Dairy Cattle, genomically enhanced breeding values for virgin heifers were able to predict cow performance significantly more accurately than parent average breeding values. Linear programming was used to optimise matings based on economic scores for Red Dairy Cattle and Holstein, considering genetic level, semen cost, recessive genetic defects, and genetic relationship. For Holstein, we also studied polledness and beta-casein genotype. The mating results for Red Dairy Cattle showed that it was possible to reduce genetic relatedness between parents and eliminate expression of genetic defects with minimal effect on genetic level. Similar results were achieved for Holstein cattle, in which it was also possible to increase the frequency of polled and beta-casein genotype A2A2 offspring without negatively impacting other comparison criteria. Evaluation of the long-term impact of genomic mating allocations in a simulation study revealed that planning matings with genomic information at herd level involves important risk management decisions, e.g. a trade-off between using fewer bulls to increase the polled allele frequency more quickly and using more bulls to reduce the rate of inbreeding and the variation in carrier frequency for genetic defects