Accounting for Pretest Environment and Selection in Estimating Breeding Values for Station-Tested Beef Bulls

Publisher Website
Google Scholar
Abstract
Station-tested beef bulls were evaluated for postweaning average daily gain using a single-trait best linear unbiased prediction (BLUP) animal model and a two-trait BLUP animal model incorporating information on the correlated trait of average daily gain to weaning. The model for average daily gain to weaning contained breed and herd-year as fixed effects, linear and quadratic covariates of age at weaning and a random animal effect. The model for test-station average daily gain contained breed and test group as fixed effects and a random animal effect. Records on both growth traits were available on 5,082 Aberdeen Angus, Charolais, Hereford, Limousin, Shorthorn and Simmental bulls tested in Canada. Preweaning average daily gain records were available on 14,939 more untested bull calves from herd-years represented by at least one tested bull. Paternal half-sib information was incorporated into both the single and two-trait analyses. The two-trait model accounted for the selection of bulls entering test stations on the basis of phenotypic performance to weaning and for pretest environmental carry-over and the resulting effects on performance in the test station. Estimated breeding values from the theoretically best model, a two-trait BLUP animal model with maternal grandsire (MGS) information, were correlated with evaluations from other models. Correlations were .84, .87, .88 and .92, respectively, between the best model and in index based on a simple contemporary comparison, single-trait BLUP, single-trait BLUP with MGS information, and two-trait BLUP (without MGS) models. The two-trait MGS model and an index based on a simple contemporary comparison, single-trait BLUP, single-trait BLUP with MGS information, and two-trait BLUP (without MGS) models. The two-trait MGS of the effects of preselection and environmental carry-over. Significant reranking of bulls can be expected with use of the two-trait evaluation. The application of such an approach in existing testing programs is recommended. Copyright © 1988. American Society of Animal Science. Copyright 1988 by American Society of Animal Science

This publication has 0 references indexed in Scilit: