Including Information from Records in Later Herds in Animal Model Evaluations

Open Access
Publisher Website
Google Scholar
Abstract
Computing strategies were developed to allow information from later herds to be included in animal model evaluations for cows that changed herds. Records from later herds were modeled as containing the same permanent environment and herd-sire interaction as did the cow''s records from the first herd. This allowed later herd records to contribute to estimation of these effects and made repeatability the same across herds as within herd. The national evaluation system was changed to store management group deviations from later herds and include them when processing information from the first herd. This method of including all lactation records for cows that changed herds was tested on United States data for Ayrshies, Brown Swiss, Guernseys, and Jerseys. Correlations between parent average and yield deviation generally increased slightly for daughters of cows with lactations in more than one herd compared with the corresponding correlations if lactations from later herds were excluded. For January 1990 USDA-DHIA genetic evaluations, records from later herds were included for all breeds, which eliminated the need for two evaluations for cows that changed herds (the main evaluation with data from only the first herd and a supplemental evaluation with all lactations through fifth included).