Application of mathematical programming to evaluation of systems of beef cattle breeding

Abstract

Performance data on Angus, Charolais, Hereford and Jersey cows and their straightbred and F₁ progeny at the Texas A&M University Agricultural Research Center at McGregor, supplemented with estimates of genetic and phenotypic parameters from the literature, provided the basis for a linear programming model designed to facilitate examination of sources of variation for productive efficiency of beef cattle systems. The effects of specific differences in milk yield, fertility level, productive longevity and cow size and progeny growth were examined along with the effects of heterosis and complementarity associated with various breeding plans. Three straight breeding systems involving small (S), medium (M) and large (L) cows, with mature weights of 430, 500 and 600 kg, were used as a basis for the breeding systems considered. Four systems were employed to simulate the production of F₁ slaughter progeny: LS, SM, LM and SL by mating large (L) and small (S) sires to small (S), medium (M) and large (L) dams. All slaughter progeny in these vertically integrated systems were fed to a constant proportion of mature weight. Two nutritional regimes were employed to estimate and satisfy nutrient requirements. Regime I involved least cost rations balanced for DP, ME and DM; regime II was characterized by the use of pasture and harvested forage. The model included fixed costs, relationships between cow size and progeny growth, different attrition rates and milk yields for cows of different ages and other considerations deemed necessary to make it trustworthy. The criterion for evaluating efficiency was net return to a fixed investment for nutrients. Average daily milk yields of 3.0, 5.0 and 7.0 kg were compared. In regime I, 5.0 kg was 0.18% and 1.85% more efficient than 3.0 and 7.0 kg, whereas for II, with cows converting relatively inexpensive forage into milk, 7.0 kg was 6.56% and 2.17% more efficient than 3.0 and 5.0 kg. ...