Data from Nichols Farm, Anita , Iowa, collected from 1975 through 1980, were used. Angus Herd Improvement Records with birth, weaning and yearling weights were obtained for 717 bulls by 80 sires. Also available, from Hawkeye Breeders Service, Des Moines , Iowa, were measures of sperm concentration and motility, percentage live sperm, number of sperm/ml, semen score and quality and scrotal circumference on the bulls as yearlings . Heritabilities and genetic and phenotypic correlations were estimated for the above traits from variance and covariance components calculated by restricted maximum likelihood. Estimates of heritability for birth, weaning and yearling weight were .70, .46 and .49, respectively. Heritability estimates for the semen traits were low (0 to .24); the estimate for scrotal circumference was .36. In general, genetic and phenotypic correlations among semen traits were high and favorable, with absolute values ranging between .49 and 1.11. Genetic correlations between semen traits and scrotal circumference were low; corresponding phenotypic correlations were moderate. Semen traits were lowly correlated phenotypically with growth traits (-.08 to .08) and genetic correlation estimates were variable. Estimates of phenotypic and genetic correlations suggested that yearling scrotal circumference was more closely related to yearling weight than to either birth or weaning weight. Estimates of phenotypic and genetic correlations between scrotal circumference and yearling weight were .26 and .68, respectively.

Yamada's method of estimating genetic co-variances between traits measured in different experimental units is discussed. It is shown that if the data are unbalanced, this method gives biased estimates of genetic covariances unless the traits have identical genetic and residual variances. An alternative unbiased procedure is suggested.