Paternal Half-Sib Correlations Between Pairs in the Same and Different Herds

Open Access
Publisher Website
Google Scholar
Abstract
Correlations were computed between first- lactation milk records expressed as devia- tions from herd-mate averages of pairs of Holstein paternal half-sibs in the same and in different herds. These were com- pared to the comparable daughter-dam re- gressions. After correcting the daughter- dam regression for environmental eom'ela- tion and change in variance from the dam to the daughter generation, both the daugh- ter-dam and paternal half-sib estinmtes of heritability were about .34, which suggests that maternal genetic variance is. unim- portant. This estimate is, however, higher than the .25 found from the paternal half- sib estimate from an analysis for variance components. The comparison of the half-sih correla- tions where the pair was in the same herd (.168) and in different herds (.086) sug- gests that the environmental correlation among paternal half-sibs in the same herds is about .08 of the total variance. There was no evidence that this environmental correlation is less for pairs of artificially sired than for naturally sired paternal half- sisters in the same herd. Several reports (7-10) have attempted to find reasons for the difference between herita- hility estimates from daughter-dam regression (.44) and paternal half-sib correlation (.25) reported by Bradford and Van Vleck (2) for first-lactation milk records of Holstein cows expressed as deviations from herd-mate aver- ages. Most heritability estimates from daugh- ter-dam regression in these data are about .40. The correlation between daughter and dam rec- ords in the same herd apparently accounts for about .02 of the difference (9). The increase in variance associated with time through an increase in mean production also can account for 10-12% of the total variance (7). Removal of these two sources of upward bias in the daughter-dam estimates gives a corrected esti- mate of about .34, which is still higher than the paternal half-sib estimate. The purpose of this study was to see if the Received for publication Augusl 18, 1965. variance component procedure used in estimat- ing the paternal half-sib correlation may be giving a lower estimate than the method which uses the product moment correlation between pairs of animals by the same sire. An estimate of' the environmental correlation between pa- ternal half-sibs was also obtained by comparing product nmment correlations between pairs in the same and different herds.