This work aimed to investigate the effects of 2 levels of N (low or high) and 2 energy sources (starch or fiber) on N partitioning, N ruminal metabolism, and digestion in dairy cows. Four Holstein cows were used in a 4 × 4 Latin square design. The 4 cows (on average, 662 ± 62 kg and at 71 ± 10 d in milk at the beginning of the experiment) were fitted with rumen, proximal duodenum, and terminal ileum cannula. The cows received 4 diets having the same forage proportion on a DM basis. The high level of N supply met 110% of the protein requirements of cows with an adequate supply in rumen-degradable N. The low level covered 80% of these requirements with a shortage in rumen-degradable N. Energy sources differed by their nature (i.e., starch from barley, corn, and wheat or fiber from soybean hulls and dehydrated beet pulp). Duodenal digesta flow was determined using YbCl3 as a marker. Microbial duodenal N flow was determined using purine and pyrimidine bases as markers from liquid-associated bacteria and mixed bacteria samples. Microbial N flow and efficiency of microbial protein synthesis, calculated using mixed bacteria as a reference microbial sample, were not significantly modified by the N level (P = 0.19 and 0.29, respectively) and the energy source of the diet (P = 0.11 and 0.08, respectively). Total tract apparent digestibility of OM and total tact digestibility of NDF were lower at the low N level (P = 0.006 and 0.007, respectively). Total tract apparent digestibility of OM tended to be greater (P = 0.08) with high-starch diets than with high-fiber diets. Total tact digestibility of NDF was greater (P< 0.001) with high-fiber diets than with high-starch diets. Duodenal N flow was less (P = 0.001) at the low N level than high N level and tended to be greater (P = 0.06) with high-starch diets than with high-fiber diets. Daily output of N in urine was less (P < 0.001) at the low N level than at the high N level. Daily output of N in feces did not differ between low and high N levels (P = 0.24) and between high-starch and high-fiber diets (P = 0.17). Milk yield and protein yield were less (P = 0.002 and P = 0.013, respectively) at the low N level than at the high N level. Milk fat yield tended to be less (P = 0.09) at the low N level than at the high N level and with high-starch than with high-fiber diets (P = 0.06). In conclusion, a large reduction in dietary N led to reduced N excretion in urine and decreased milk production but did not affect N excretion in feces or microbial protein synthesis.

No abstract