Search

Twenty-four multiparous and fifteen first lactation Holstein cows averaging 263 days in milk and weighing 614 kg were fed diets adequate or deficient in ruminal nitrogen (N), based on predictions of the Cornell Net Carbohydrate and Protein System (CNCPS). After adjustment to a low crude protein (CP) total mixed rations (TMR; 12.6% CP), the cows were allocated to 13 blocks based on lactation number, milk production, body condition score, and body weight. Within each block, cows were randomly assigned to one of the 3 treatment (TRT) diets (9.4, 11.1 and 14.1% CP for TRT 1, 2, and 3, respectively). All diets contained the same proportion of high moisture corn, chopped grass hay, and minerals, with urea substituted for corn silage as needed to reach the three CP levels. The TRT diets were then fed to the cows for 4 wk. Milk production was significantly affected by TRT: 15.5, 18.8, and 21.7 kg/d for TRT diets 1, 2, and 3, respectively. DMI was increased significantly as the percentage of CP increased from 9.4 to 14.1% CP: 17.6, 20.0, and 21.2 kg/d for TRT diets 1,2, and 3, respectively. CNCPS predictions for production (with and without the N adjustment for ruminal N deficiency) of metabolizable protein (MP) allowable milk were compared with observed milk production. Using the average individual weekly cow data from all 3 TRT, we found that the CNCPS accounted for 72 and 68% of the variation in MP allowable milk without and with the N deficiency adjustment, respectively. The overall mean bias without the N adjustment was 3.3 kg of milk (over prediction model bias of 14.6%, P < 0.001), and the N adjustment reduced the model over-prediction bias to 0.01 kg of milk (P = 0.96).