Abstract
A deterministic model was developed in 1998 to evaluate the concurrent effects of selection and crossbreeding on the rate of genetic gain and increases in productivity of New Zealand dairy cattle over the ensuing 25-yr period. The predictions of today's breed composition of the national dairy herd and genetic trends for body weight and lactation yields of milk, fat, and protein are compared with today's actual values. Selection was assumed to use an index that included live weight and lactation yields of milk, fat, and protein. Mating strategies involving the Holstein-Friesian (F), Jersey (J), and Ayrshire (A) breeds were evaluated. Effects of heterosis and age were included to calculate phenotypic live weight and yields of milk, fat, and protein per cow. At the time the model was developed, New Zealand had an across-breed evaluation system, but only straightbred bulls were used after progeny testing based on records of straightbred and crossbred daughters. The model predicted that if crossbred cows and bulls could be considered as bull parents, faster rates of genetic gain may result because of increased selection intensities in the cow to breed bull selection pathway. This scenario transpired, and the best bulls and cows for farm profit were used regardless of breed. Under that mating strategy for the 2018 birthyear, the model predicted the national breed composition would be 11% F, 34% J, 52% F×J, and 2% A; the actual breed composition was 36% F, 9% J, 53% F×J, and 1% A. The model-predicted annual genetic gains would be 16.7 L of milk, 1.2 kg of fat, 1.5 kg of protein, and -0.7 kg of body weight; the realized annual improvements were 13.6 L of milk, 1.31 kg of fat, 1.17 kg of protein, and -0.36 kg of body weight. Predicted long-term responses to selection can closely mirror realized improvements, confirming the value of modeling to inform animal breeding decision-making.