MILK PRODUCTION BY CROSSBRED COWS AS A FUNCTION OF SUPPLEMENTATION WITH ENERGY AND/OR PROTEIN CONCENTRATES AT PASTURE OR CONFINED

Abstract

Three experiments were conducted to evaluate the effects of energetic and/or protein supplementation on milk production and composition by cows, with the presence of calf at milking, in pasture or receiving chopped sugarcane or elephant grass, with access to mixture containing urea and mineral salt, during the dry season. In each experiment, eight crossbred Holstein-Zebu cows (480 to 540 kg) were distributed in two 4x4 Latin squares, in four periods of seven days. The treatments consisted of increasing levels of supplements containing 57% of corn and 43% of soybean meal – CSM (0, 1.5, 3.0 and 6.0 kg/cow/day), wheat meaddlings – WM (0.5, 1.0, 2.0 and 4.0 kg) and soybean meal – SM (0, 0.65, 1.3 and 2.6 kg). The concentrates were furnished twice a day, during milking time. The statistical analyses included the effects of treatments (polynomial regression), Latin square, animal inside Latin square and period, in addition to use of models of saturation kinetics. There was no effect of treatment for none of experiments, with mean ranging from 7.5 to 11.2 kg of milk/cow/day, although milk production presented curvilinear trend with concentrate levels, following the Michaelis-Menten model. The obtained equations of Lineweaver-Burk were: Exp. 1: 1/Milk = 0.0394*(1/(CSM)) + 0.0866; Exp. 2: 1/Milk = 0.0061*(1/WM) + 0.0917; Exp. 3: 1/Milk = 0.0226*(1/SM) + 0.0825. The theoretical maximum milk production (1/a) was 11.5, 10.9 and 12.1 kg/cow/day; the amount of concentrate to cause half Kmax (b/a = Ks) was 0.45, 0.06 and 0.27 kg; and the amount of concentrate to reach zero cost-benefit ratio was 1.7, 0.9 and 1.25 kg for CSM, WM and SM, respectively. The marginal increase in milk production by increase in supplementation was decreasing (0.94, 0.60 and 0.36 kg of milk/kg of CSM; 1.79, 0.28 and -0.11 kg of milk/kg of WM; and 1.57, 1.67 and 1.19 kg of milk/kg of SM). Models of saturation kinetics are more appropriate to make recommendations of nutrients in relation to traditional linear models.