Four ewes fitted with ruminal and duodenal T-piece cannulae were each given six diets in a 6 × 4 factorial design. Diets or experimental treatments consisted of two ratios of forage: concentrate (700:150 (LC) and 400: 600 (HO). Forage was ammonia-treated straw and the concentrate was formulated with barley supplemented with one of three protein sources: sunflower meal, soya-bean meal or fish meal. Duodenal flows ofdigesta were estimated by the dual-phase technique using Co-EDTA and Yb acetate as liquid and solid markers. Microbial nitrogen (N) was estimated from the digesta flow of purine bases and 15N enrichment using as reference samples, bacterial isolates from the liquid (LAB) or solid (SAB) phase of rumen digesta.

Duodenal flow of purine bases (mmol/day) was lower on LC (12·9) than HC (17·7) diets but in both treatments it was depressed by fish meal (12·3) compared with either soya-bean (17·3) or sunflower meal (16·3) as supplements (s.e. 1·13). Urinary excretion of purine derivatives showed a similar trend, 8·6 v. III mmol/day in LC and HC respectively and 8·8 v. 10·4 and 10·5 mmol/day in fish meal, soya-bean and sunflower meal diets (s.e. 0·56), respectively. Variation in excretion of urinary purine derivatives was mainly associated with digestible organic matter intake with an average ratio of 1·7 (s.e. 0·11) mmol per 100 g digestible organic matter intake. Irrespective of the microbial marker used, microbial yield was higher in animals offered HC than in those offered LC and with soya-bean or sunflower meal compared with fish meal supplemented diets. The microbial purine bases/N (mmol/g) ratio varied between LAB (1·99, s.e. 0·092) and SAB (1·69, s.e. 0·071) isolates leading to different estimates of microbial-N yield (g) from duodenal purine bases (7·76 (s.e. 2·84) v. 9·13 (s.e. 3·24)), urinary excretion of allantoin (5·57 (s.e. 2·0) v. 6·57 (s.e. 2·03)) or total purine derivatives (6·43 (s.e. 2·39) v. 7·56 (s.e. 2·77)). Urinary excretion of allantoin or total purine derivatives provided consistently lower estimates of duodenal microbial-N than duodenal purine bases or 15N, although it closely reflected the pattern observed in direct measurements.

The purpose of the present study was first to assess the extent to which unlabelled ascorbate in the diet of guinea-pigs can exchange with labelled ascorbate within their organs when the dietary intake is varied over a wide range, and second to determine whether the retention of label might be used to assess either the amount of ascorbate intake or its biological availability where these are not known. The retention of [14C]ascorbate in the body and in various organs of guinea-pigs were, therefore, measured following a 13 d period of graded dietary intakes of ascorbate. It was found first, that the amount of label retained in each of the organs, 13 d after the initial dose of labelled ascorbate, was much more closely related to the amount of ascorbate intake after labelling than to the intake (and tissue ascorbate levels) before and at the time of labelling. Second, most of the individual internal organs exhibited a constant relationship between the specific activity at 13 d and the dietary intake, except for brain which was flushed to a smaller extent. Third, in agreement with several previous studies a high proportion of the radioactive label in the tissues was found to be still present in ascorbate. The specific activity of column-purified ascorbate was very similar to the estimated specific activity in the crude extract, which implies that it may be possible to estimate specific activities (or stable isotope enrichments) at certain sites without rigorous isolation procedures. Fourth, the amount of radioactivity appearing in the urine 2 d before killing the animals was correlated with the amount of ascorbate intake and with tissue specific activities, suggesting that intakes (or bioavailability) might be predicted from the patterns of label-appearance in the urine