Genetic evaluation of dairy goats for milk and fat yield with an animal model.

An animal model was applied to evaluate 120,073 Alpine, LaMancha, Nubian, Saanen, and Toggenburg bucks and does. Parities higher than six were excluded. The model included fixed herd-year-season and random herd-sire interaction, permanent environmental breeding value, and residual effects. Breeding value included additive genetic value and fixed accumulated group effect. A doe's records in different herds were accommodated by predicting a separate permanent environmental effect for each herd. Lactations were weighted according to lactation length. Does of all breeds were used as contemporaries. Breed differences were accounted for with a grouping strategy that grouped unknown parents by breed, sex of parent, sex of progeny, and birth year of progeny. Evaluations for milk and fat were computed as separate traits during the same processing. To achieve three-digit accuracy, 100 iterations were completed, each requiring slightly over 4 s of central processing unit time on a Cray X-MP/48 computer. Predicted producing ability was computed as sum of predicted breeding value, herd-sire interaction, and permanent environmental effects for use in ranking does on expected yield in next lactation. Evaluations of goats served as a test for the animal model evaluation system under development for dairy cattle.