Stability of Proofs for High Predicted Difference Holstein Sires
暂无分享,去创建一个
Abstract Evaluations by modified contemporary comparisons of sires from United States Department of Agriculture sire summary lists were used to examine stability of sire evaluations for bulls with at least one predicted difference for milk exceeding 453 kg. Each bull's evaluation first exceeding 453 kg was compared to his most recent evaluation. Average decline within paired summaries was 21 kg for 236 available bulls and 28 kg for a subset of 73 bulls with 95% or higher Repeatability on most recent summary. Last available predicted difference with Repeatability equal to or exceeding 95% was linearly related to evaluations first exceeding 45 3 kg but quadratically related to change in Repeatability within paired summaries. Minimum expected last predicted difference occurred at Repeat-abilities between 70 and 79% with small and sizeable increases, respectively, projected for higher and lower Repeatabilities. Predicted difference at a given time was far more important in explaining differences in last available evaluations than was Repeatability at that time. Declines in predicted difference above 70% Repeatability may come partially from differences in genetic and managemental characteristics of herds in which sires are tested originally and herds in which they subsequently receive heavy use.