1. A method for estimating the proportions of bacterial- and protozoal-N in the total non-ammonia-N reaching the lower gut of the ruminant under steady-state conditions was evaluated. Three trials using two different diets were conducted with a Holstein steer equipped with a rumen cannula and duodenal re-entrant cannulas.
2. An intraruminal primed infusion of (15nh4)2so4 was administered for 68 h during each trial. Bacteria and protozoa samples were isolated from rumen fluid at approximately 6 h intervals during each infusion period. Total non-ammonia-N was isolated from duodenal digesta samples taken at approximately the same times. All of these samples were analysed for 15N enrichment. A computer program was used to fit equations to the 15N-enrichment curves of bacterial- and protozoal-N. Models of both bacterial- and protozoal-N kinetics consisted of a small pool which equilibrated rapidly with rumen NH3 and a large pool with a fractional turnover rate of 0.045–0.070/h for bacterial-N and 0.056–0.069/h for protozoal-N.
3. Abomasal fluid turnover was estimated by a single injection of polyethylene glycol (molecular weight 4000) into the rumen followed by sampling of rumen fluid and duodenal digesta.
4. Estimates of abomasal fluid turnover, bacterial-N turnover, and protozoal-N turnover were entered into an equation which was adjusted by computer iteration to fit the 15n-enrichment curve of duodenal digesta non-NH3-N generated from each (15nh4)2so4 infusion period. The computer fit of this equation to the observed results gave estimates of 0.39–0.45 and 0.22–0.41 for the proportion of duodenal non-NH3-N derived from bacterial-N and protozoal-N respectively.
5. This method is potentially useful in estimating microbial protein passage to the lower gut in ruminants. Sampling digesta from the omasum rather than the duodenum would simplify the method and possibly increase the reliability of the estimates.