Sire Ranking Based upon Purebred versus Crossbred Progeny Performance in Swine

The genetic correlation between the same trait measured in purebred and crossbred progeny of the same sire, and the correlation between sire ranks based upon purebred and crossbred progeny, were estimated using two data sets consisting of records on 2,258 purebred and 4,815 crossbred progeny of 32 Duroc, 22 Hampshire, 9 Landrace, 8 Spotted and 31 Yorkshire boars. Traits investigated were pig birth, 21- and 42-d weights (BW, W21, W42), postweaning average daily gain (ADG) and probed backfat thickness (BF). Variance component estimates were obtained by Henderson's Method 3, and sire evaluations for purebred and crossbred progeny were computed from mixed model analyses. Paternal half-sib estimates of heritability were, in general, similar based upon either purebred or crossbred progeny. Genetic correlations were estimated both from variance component estimates and as product-moment correlations of sire evaluations for purebred and crossbred progeny. The intra-class method resulted in consistently lower estimates, and the unreliability of this procedure is discussed. Genetic correlations calculated from sire evaluations, and correlations of ranks based upon these evaluations, averaged .26 and .19 for BW, .12 and .16 for W21, .55 and .50 for W42, .56 and .59 for ADG and .88 and .84 for BF, respectively. Comparison of the expected response in crossbred populations suggested a possible advantage of selection for specific combining ability over mass selection for BW and W21, but not for W42, ADG and BF.