No CrossRef data available.
Heterogeneity of variance between herds for milk production traits
Published online by Cambridge University Press: 24 November 2017
Extract
One assumption made by most users of the BLUP (Best Linear Unbiased Prediction) method to predict breeding values, is that variances are homogeneous over herds or herd-year-seasons (HYS). In dairy cattle there is abundant evidence, however, of heterogeneity of variance across herds or herd groups (see e.g. Hill et al. 1983 and Brotherstone and Hill, 1986, for U.K. studies). The effect of ignoring heterogenous variances between herds on prediction of breeding values for bulls may be small when using a sire model, if sires were tested across many herd-variance groups. Loss in efficiency may be larger when sires are tested in few herds, or, for cows, when the genetic evaluation is for bulls and cows simultaneously (with an animal model (AM)). The aim of this study was to calculate individual herd parameter estimates to investigate heterogeneity of within herd variance in the U.K. dairy population. The investigated trait was fat yield and the estimations were carried out using a REML (Restricted Maximum Likelihood; Patterson and Thompson, 1971) program written by Karin Meyer (Meyer, 1989).
- Type
- Sheep and cattle breeding
- Information
- Copyright
- Copyright © The British Society of Animal Production 1991