TY - JOUR
AU1 - Lee, A., J.
AU2 - Henderson, C., R.
AB - Summary In the use of simplified selection index procedures to evaluate cows on their milk yielding ability, it is assumed that repeatability and the variance of records are constant for all herds and lactations. New York DHIA records on grade and purebred Holstein cows were used to examine the validity of these assumptions, the effect of cow culling on estimation of the repeatability and the variance of records, and the practical effects of apparent violation of assumptions and biases in estimation on cow evaluation procedures. Though a slight heterogeneity across herds of both intraherd regressions of a later on an earlier record and the within herd variances of records was found, in view of the large sampling errors, it would not be practical to use separate variance and repeatability estimates for each herd. The large range in pooled intraherd regressions of a later on an earlier lactation record and in pooled within herd variances for different lactations was almost completely accounted for by the effects of cow culling on milk yield. After adjustment for the effects of cow culling, the within herd variance of milk yield records was ten percent smaller in first than in later lactations while the within herd covariances were remarkably similar. It is concluded that unbiased estimation of the repeatability and variance of milk yield records must take into account the effects of cow culling. Considering different variances and covariances for different lactations after adjustment for the effects of cow culling instead of assuming homogeneous variances and constant repeatability resulted in only slight changes in selection index weights and accuracy. The simplified selection index procedure of weighting the average of n milk yield records on a cow by nr/(1+(n−1)r) where repeatability, r, is in the range of .55 to .60 is recommended for the within herd evaluation of cows on their milk yielding ability or future production. This content is only available as a PDF. Author notes 1 Present address: Animal Research Institute, Ottawa, Canada. Copyright 1971 by American Society of Animal Science. American Society of Animal Science
TI - Effect of Cow Culling on Repeatability Estimates
JF - Journal of Animal Science
DO - 10.2527/jas1971.322167x
DA - 1971-02-01
UR - https://www.deepdyve.com/lp/oxford-university-press/effect-of-cow-culling-on-repeatability-estimates-gfQbkho1Ck
SP - 167
EP - 173
VL - 32
IS - 2
DP - DeepDyve
ER -