Milk production records of Braunvieh cows of Tirol region in Austria were used to construct different selection indices and subindices. Data on 46456 lactation records including 16569 paternal half-sisters representing 1556 sires were used to estimate the genetic and phenotypic variation and covariation of 305- day milk, fat and protein yield traits. Analysis was carried out using a mixed model including effects of year and month of calving, parity, days open, grazing region and interaction of year by parity (as fixed effects) and sires and cows-within-sires (as random effects). Seven indices of selection for improving yield traits of cows were constructed involving all combinations of two or three traits. Subindices were derived as if the three or two sources of information were used to select for just one trait, viz, to select for milk or fat or protein yield. Large differences in index coefficients (b's) for protein yield against yields of milk or fat were observed. Rate of genetic progress in aggregate genotype decreased when protein yield was dropped and consequently, a considerable genetic improvement for cow productivity might be achieved through selection for protein yield. Milk yield contributed about 50% of the total economic-genetic gain while yields of fat plus protein contributed the remainder, 50%. Th expected economic-genetic gain per generation in yield traits was large in all of the indices and subindices constructed; estimates ranging between 414 to 424 kg for milk, 16.7 to 17.5 kg for fat and 14.1 to 15.2 kg for protein. Correlations between the index or subindex constructed and each individual trait in the aggregate genotype were relatively high and ranged from 0.62 to 0.74. A reduced index or subindex based on protein yield with either milk or fat yield could be practically applied efficiently to improve the productivity of Braunvieh cows under the local Austrian conditions. |
