Genetic evaluation of Holstein Friesian sires for daughter condition-scorechanges using a random regression model

In dairy cattle type classification schemes, heifers are condition scored ( CS) only once during their first lactation. Although genetic analysis of co ndition-score changes is not possible using an animal model, the data can b e analysed as repeated observations on the sire. CS records for 100 078 Holstein Friesian heifers, the progeny of 797 sires, were available. Sires differed in the shape of the regression of mean daug hter CS on stage of lactation at both the phenotypic and genetic level. Gen etic analysis was carried out using a random regression model (RRM) which c an account for differences between sires in the shape of the CS curves. CS curves for individual sires were modelled using a cubic polynomial. Heritability estimates for CS at each stage of lactation generally increase d through the lactation from 0.20 in stage 2 (days in milk 31 to 60) to 0.2 8 in later lactation stages. Genetic correlations between CS at different s tages were generally high (>0.80), with the exception of correlations with stage 1 (days in milk 1 to 30) which decreased to 0.63 with stages 6 and 7, suggesting that CS at stage 1 is under different biological control from C S at other stages of the lactation. Using RRM, sire estimated breeding valu es (EBVs) for CS at each stage of the lactation were estimated. Sire rankin gs on EBV at each stage were seen to change through early mid and later lac tation stages.