Genetic trends for fertility in Swedish Red cattle using different models

S. Eriksson¹, W.F. Fikse¹, H. Hansen Axelsson¹, K. Johansson²

¹Swedish University of Agricultural Sciences, P.O. Box 7023, SE-750 07 Uppsala, Sweden;
²Växa Sverige, P.O. Box 7023, SE-75007 Uppsala, Sweden

Unfavourable genetic trends in female fertility have been observed in dairy cattle, explained by unfavourable genetic correlations with milk production. Scandinavian Red breeds have long been selected for functional traits, and less deterioration of fertility has been observed. The aim of this study was to estimate genetic trends in Swedish Red dairy cattle for two female fertility traits, udder health traits and protein yield, using different models.

The fertility traits were number of inseminations per service period (NINS), and interval from calving to first insemination (CFI). Data for virgin heifers from 1998 and cows in first and second lactation from 1990 until 2007 was included. The number of observations per trait varied from 141,003 (udder conformation) to 1,046,206 (clinical mastitis). (Co)-variance components were estimated prior to the prediction of breeding values. Multi trait (MT) animal and sire models were used for estimation of variance and covariance components. For prediction of breeding values MT-animal models were used. Traits were analysed both within trait groups (e.g. fertility traits), as in the current Nordic evaluation, and using a full MT model with all functional traits and protein yield. Analyses were made with and without heifer data.

The resulting genetic trends for Swedish AI bulls were neutral or favourable for udder health traits, CFI, and for NINS in virgin heifers. Unfavourable trends were seen for NINS in first and second lactation. In cows, the genetic trend seemed unfavourable for somatic cell score, clinical mastitis and NINS in lactating cows. These estimated trends were less unfavourable when evaluations were done within trait groups compared with using the full MT-model. Excluding heifer data had a smaller than expected effect on the estimated genetic trends.