|Boettcher, P. - UNIV. OF GUELPH|
Submitted to: American Dairy Science Association Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: June 1, 1999
Publication Date: N/A
Technical Abstract: US and Canadian Holstein populations were examined for evidence of X chromosomal inheritance. Because the X chromosome of a bull is only inherited by female offspring, discrepancies between the performance of a bull's daughters and the daughters of his sons could indicate X-inheritance. The following model was used to test for discrepancies: Sire ePTA - Parent Average PTA = Sire + Maternal Grandsire + Error. Thresholds for empirical significance tests were established with a simulation study of a population with two-stage selection. A biallelic model of X-inheritance was assumed. The critical value for the sire and maternal grandsire effects were approximately 2% and 4.5% of the total variance, respectively, but varied according to the polygenic variance of the trait. The power of the test increased as the size of the X-chromosome effect increased and as the polygenic variance and initial frequency of the favorable allele decreased. Sire PTA from the US and Canada were used in separate analyses to test for evidence of X-chromosomal inheritance in each population. For the US, PTA for milk, fat, and protein yields; somatic cell score; and productive life from 17,296 bulls from 246 sires were used for the test. Canadian data consisted of PTA for milk, fat, and protein yields for 9363 bulls from 236 sires. Productive life (US) was the only trait for which a significant (P < 0.05) effect of either sire or maternal grandsire was observed. For that trait, nearly 8.0% of variance in adjusted sire PTA was associated with sire. However, further analysis shed doubt upon whether these results were indicative of X-inheritance. The correlation coefficient between solutions for sire and maternal-grandsire effects was 0.40. A zero or negative correlation was expected in the presence of X-inheritance.