PREDICTION OF BREEDING VALUES WITH ADDITIVE ANIMAL-MODELS FOR CROSSESFROM 2 POPULATIONS

Recent developments in the theory of the covariance between relatives in crosses from 2 populations, under additive inheritance, are used to predict breeding values (BV) by best linear unbiased prediction (BLUP ) using animal models. The consequences of incorrectly specifying the covariance matrix of BV is discussed. The theory of the covariance bet ween relatives in crosses from 2 populations is extended for predictin g BV in models with multiple traits. A numerical example illustrates t he prediction procedures.