The effect of genetic strain of Holstein Friesian dairy cows in a high input grass-based system on production traits

Phenotypic expressions for traits of economic and welfare importance in domestic animals are as a consequence of the genotype of the animal and the environment in which that genotype performs. An interaction between the genotype and environment can occur when the relative expression of the genotype changes with different environments, and is referred to as a genotype-by-environment (G × E) interaction (Falconer and Mackay, 1996). Genotype-by-environment interactions occur as the genes or the emphasis of genes associated with the performance of a trait change in different environments. Genotype-by-environment interactions can cause issues for animal breeders as performances of genotypes in one environment may not translate to the same performance in a different environment leading to a lower accuracy of selection which could reduce the rate of genetic gain.
Currently in Ireland, there is one total merit index used for breeding dairy replacements, the Economic Breeding Index (EBI), that consists of relevant traits to the dairy industry that are weighted by their economic importance. This is used to select dairy cows and bulls to be parents of the next generation. The EBI was designed to optimise profitability for a low-input spring-calving grass-based system (Berry et al., 2007), indicated by the high weighting of fertility traits in the index (ICBF, 2018). Within Ireland however, there exists a wide variation in dairy production systems from low-input and output systems to high-input and output systems. The adequacy of using the EBI to select dairy animals across these range of dairy production systems has been investigated as part of this thesis. Previous work has suggested that using a total merit index outside of the environment it is designed for may lead to decreases in genetic gain (Mulder et al., 2006). When a selection index is used across a range of environments, the ability of sire PTA to predict daughter performance changes (Kearney et al., 2004b; Fahey et al., 2007). These changes in environment along with a range of other factors that can impact the ability of sire PTA to predict daughter performance have been assessed as part of this thesis also. Data consisted of 1,643,474 records from 670,559 cows from 7,389 herds across Ireland. The milk production of herds varied considerably and this variation in milk production was used to group herds into quartiles to indicate differences in environment.
The objectives of this thesis were to 1) determine if G × E interactions between within-herd 305-d milk yield quartiles for milk production traits exists 2) to assess if the sires’ PTA is a suitable tool across a range of dairy production systems in Ireland to predict daughter performance and 3) to assess how certain factors can affect the ability of sire PTA to predict daughter performance for milk production traits.