STRATEGY FOR ESTIMATION OF VARIANCE COMPONENTS FOR THE JOINT NORDIC YIELD EVALUATION

In 2006 Denmark, Finland and Sweden introduced a common genetic evaluation of yield traits for Red Cattle, Holstein and Jersey. The evaluation is based on test-day (TD) records from Denmark and Finland, and 305d records from Sweden. The applied meta-model accommodates the different data sources and consent to a unique set of breeding values across the three countries (Mantysaari et al., 2006). This is achieved by modelling a genetic correlation of unity across countries, but allowing different variances and heritabilities. Because in Finland content traits are measured bimonthly, the traits milk, protein, and fat yield are evaluated simultaneously. Considering the first three lactations as different traits, leads for the main breeds to a multiple-trait model with 27 traits. Currently the models are upgraded to replace the Swedish 305d records by TD records, which requires new variance components (VC) for the Swedish data. Moreover, used VC for the Finnish data were estimated in 1997, which made it obvious to update the VC for all the breeds and countries.