An evaluation of bias in estimated breeding values for weaning weight in Angus beef cattle field records. III. Estimates of bias due to nonrandom mating.

Field records from 15 herds involved in the Angus Herd Improvement Records program of the American Angus Association indicated that positive assortative mating for weaning weight estimated breeding value (EBV) was being practiced. For each dam, a separate EBV was calculated from information on each progeny with all other progeny data deleted. Individual, paternal and maternal half-sib data available in all years were included in each calculation and thus, were held constant. Bias in cow EBV caused by nonrandom mating was measured as the pooled within dam regression of dam EBV on progeny's sire EBV deviated from the contemporary site average EBV. Across herds, the average estimated bias associated with positive assortative mating ranged from .05 to .08 ratio units.