Generalized Least Squares Estimation of a Genotypic Covariance Matrix

Abstract

A generalized least squares approach is used to estimate the genotypic covariance matrix of weaning weight, feed consumption, average daily gain and retail product for beef cattle. In this procedure, the elements of 'within-sires' and 'among-sires' sample covariance matrices are the observations. A satisfactory fit to these data is obtained with a model in which two genotypic factors are present. The model can be used to determine economic values for traits considered in breeding programs.