Bias Due to Corrective Matings in Sire Evaluations for Calving Ease

Abstract

Calving ease data for 57,051 Holstein-Friesian births from 1980 to 1983 were analyzed for evidence of bias due to corrective matings in sire evaluations. Two models for evaluating sires were compared. Model 1 contained effects of herd-year-season, size of cow-parity-sex of calf, and sire of calf, and Model 2 was the same as Model 1 with the addition of a sire of the dam effect. Data were separated into three subsets according to year of birth such that the first subset contained the first year of data, the second subset included both first and second year, and the last subset was the entire data set. Sire proofs from each subset and by each model were partitioned to determine the influence of sire of dam effects on sire of calf evaluations. Negative correlations from −.25 to −.33 between sire of calf proofs and effects of size of cow-parity-sex of calf indicated that phenotypic nonrandom mating was removed by size of cow-parity effects in the model. Near zero correlations between sire of calf proofs and sire of dam effects suggested little nonrandom mating on a genetic basis was taking place. Correlations of sire of calf proofs over the three subsets were greater than expected.