THE EFFECT OF GENOTYPE FOR CAPPA-CASEIN O N ESTIMATED BREEDING VALUESOF ACTUAL BULLS OUT OF THE HOLSTEIN POPULATION FROM 4 COUNTRIES

The potential effect of the polymorphism, A < > B, of the cappa-casein locus on the estimated breeding values of 1931 genotyped A. I. sires from four subpopulations of the international Holstein breed was exami ned by means of an analysis of variance under a completely fixed model . With respect to the genotypes es the data was gathered by inquiring at A. I. studs and later completed using published estimated breeding values. The results can be summarized as follows. 1. In the analysis o f the entire data set partially using converted breeding values protei n content showed a highly significant increase from the homozygous AA to the heterozygous AB genotypes; BB genotypes showed the highest bree ding values for protein contents. Differences from the other two genot ypes were significant/highly significant. 2. Between countries/subpopu lations different effects of genotype were found for the breeding valu es. Supposedly the reasons for this are different breeding strategies or breeding goals that lead to differential preference of specific sir es and maternal grandsires of bulls. The comparison of heterozygous ge notypes with the homozygous AA-genotypes gave the following, statistic ally significant differences: in Germany: Increase of protein contents in Italy: Increase of protein yield and protein contents in Canada: I ncrease of protein contents and decrease of milk yield in the USA: Dec rease of fat contents 3. The effect of specific sires of bulls was exa mined in analyses with and without inclusion of their sons. The result s show that for Canada the observed decrease in milk yield when substi tuting the allel A by the allel B could solely be attributed to one si ngle sire of sons. Analogously, the decrease in fat contents for the U SA could be traced back to another sire of sons. 4. Due to the propert ies of the sample and the analysis using a simple model the results sh ould be interpreted with great care. However, since differences betwee n genotypes and between specific sires of sons at least partially were drastic, and since the size of the sample was relatively large, the p resent study should give useful hints for a further examination.