Strategies to Objectively Group Merino Flocks in Sheep Genetics

The Merino breeding population in Australia exhibits considerable diversity in objectives and breeding philosophies. It could be beneficial for both the analysis and reporting of the national genetic evaluation to objectively group flocks into logical subsets. This study evaluates techniques to cluster flocks into logical groups based on either estimated breeding values or genomic information. Principle component analyses were conducted using flock mean breeding values and the genomic relationship matrix. Using the flock mean breeding values, 6 clusters of flocks were identified with the first 2 principle components explaining 73% of the variation between flocks. The first principle component separated flocks based on overall productivity, with approximately equal emphasis across all traits. The second component separated flocks based on fleece weight, wrinkle and staple length. Less separation between flocks was apparent for flocks with below average fibre diameter. The principle components of the genomic relationship matrix were also strongly correlated with mean breeding values across the flocks. The lack of accurate Australian Sheep Breeding Values (ASBVs) for some traits and genomic information across some of the flocks is a limitation of this approach as it makes allocation of some flocks challenging.