Produção de leite em função de níveis de suplementação com concentrados para vacas leiteiras pastejando capim-elefante

This study aimed to evaluate the effects of supplement amount on milk production and composition. Eight crossbred Holstein-Zebu cows (517±40 kg) were allotted in two 4 x 4 Latin squares, in four periods of 14 days. The experiment was conducted on pasture of Elephant grass (Pennisetum purpureum, Schum) in the rainy season, and the treatments included a control (mineral mixture) and supplement with 24% CP in dry matter at levels of 1.25, 2.5, and 5.0 kg/animal/day, based on corn meal, soybean meal, urea and mineral mixture. The experiment was analyzed as Latin square design including effects of treatment, Latin square, animal within Latin square and period. There was no treatment effect (P>0.05) for the evaluated variables, due to large animal per animal variation. However, by evaluating the mean of milk production as a function of supplementation, there was a trend for curvilinear response, following a Michaelis-Menten relationship of enzymatic systems and was explained by the following equation of Lineweaver-Burk: 1/Milk = 0.0125*(1/Suppl) + 0.0826; r2 = 1.00. The theoretical maximum milk production (1/a) was 12.1 kg/animal/day, and the amount of concentrate to cause half maximum milk production (b/a) was verified with 5.2% of that necessary to reach 95% maximum response. It can be concluded that the marginal increase in milk production reduces with increasing supplementation, in which the adequate level of supplementation is between 1.25 and 2.5 kg of concentrate/cow/day Objetivou-se avaliar os efeitos da quantidade de suplementos na produção e composição do leite. Foram utilizadas oito vacas mestiças Holandês-Zebu, com 517±40 kg de peso médio, distribuídas estatisticamente na forma de dois quadrados latinos 4 x 4, em quatro períodos de 14 dias. O experimento foi conduzido em pasto de capim-elefante (Pennisetum purpureum, Schum) na estação chuvosa, e os tratamentos incluíram um controle (mistura mineral) e suplementos com 24% de proteína bruta na matéria seca nos níveis de 1,25; 2,5; e 5,0 kg/animal/dia, baseado em fubá de milho, farelo de soja, uréia e mistura mineral. O experimento foi analisado em quadrado latino incluindo efeitos de tratamento, quadrado latino, animal dentro de quadrado latino e período. Não houve efeito de tratamento (P>0,05) para as variáveis analisadas, devido à grande variação de animal para animal. Entretanto, ao avaliar a média da produção de leite em função da suplementação, houve tendência de resposta curvilínea, seguindo o mo