The Effect of Linkage and Genetic Grouping on the Accuracy of Across-Flock Genetic Evaluation in Australian Merino Sheep

Abstract

The aim of this research was to characterize the significance of linkage and genetic grouping, and to develop better and alternative models of genetic grouping for across-flock evaluations in the Australian Merino sheep industry. Most livestock industry datasets used for genetic evaluation have missing pedigree and performance data in varying amounts. The implicit assumption underlying an animal model evaluation that all base animals have similar genetic merit with a common variance will not hold true in a majority of cases. Not accounting for previous selection on base animals will bias estimated breeding values (EBVs). The use of outside sires having different population means and with either incomplete or no pedigree information is a common practice which may affect the mean breeding value of animals in a flock. To account for the differences in the pedigree and data recording, and differences in the genetic means of base populations, genetic groups are generally included in the evaluation procedure. In Merino evaluations by Sheep Genetics in Australia, genetic groups are included in the model to account for differences in the genetic means of base populations of different flocks. The ongoing improvement in the models of genetic evaluation will lead to further developments of current methods used for accounting for missing data and pedigree information. The recent advancements in molecular genetics techniques (for example, DNA finger printing and gene mapping) and simultaneous reduction in the costs associated with their widespread use in ascertaining parentage in different livestock industries, may in future reduce the need for grouping animals. However, some animals will always need to be grouped as base populations have varying means for traits of interest and pedigree can only be traced back for a few generations in farm animals.