Submitted to: International Symposium on Animal Production
Publication Type: Abstract Only
Publication Acceptance Date: September 5, 1997
Publication Date: N/A
Technical Abstract: Genetic evaluations in the US currently are computed with a state-of-the- art animal model, which enables data from all relatives of each cow with yield records to contribute to her evaluation and her sire's. Unknown- parent genetic groups allow recognition of improved genetic level over time and differences between unknown sires and dams. Records from the 1st 5 lactations are included using a repeatability model. A 1st lactation is required for a record to affect relatives' evaluations. Advances in computer technology and evaluation methodology have allowed improvements to include (1) using records from later herds for cows that change herds, (2) accounting for projected records' reduced genetic variance, (3) adjusting for heterogeneous variance, (4) including age and parity in the model as well as using multiplicative adjustments, (5) accounting for inbreeding when forming the relationship matrix inverse, (6) incorporating Canadian evaluations, (7) calculating somatic cell score and productive life evaluations, and (8) including unsupervised records subject to additional editing. Evaluations are distributed primarily through industry cooperators and the internet. A test-day model is planned so that environmental influences can be estimated more accurately by defining their effects on a test-day basis. Within-herd analysis estimates test-day and herd-specific age and season effects. The model accounts for genetic differences in lactation curve shape and maturity rate by across-herd multitrait analysis. Older yield data are included through correlations with test-day traits. Individual genetic estimates by lactation stage and parity allow more precise definition of genetic indexes. Implementation of a test-day model is anticipated by the year 2000.