1. An alternative statistical technique, based on analysis of the 75th percentile, was used to examine teenagers' diets. The technique was developed to identify personal characteristics that affected the prevalence of high energy and nutrient intakes. It provided useful information not obtainable from methods based on average intakes.

2. The study sample was a representative sample of 1055 high-school students in Tasmania, Australia, aged 11–16 years.

3. Among girls, the prevalence of high intakes of energy and several nutrients was not affected by any of the fifteen characteristics studied. (In this context, ‘high’ was defined as above the 75th percentile for girls.) Fatter, heavier girls were less likely to have high intakes of carbohydrate, iron and niacin-equivalent, and also of meat and ‘empty-energy’ foods. Girls from small families were less likely to have high calcium intakes, and girls from goverment schools less likely to have high thiamin intakes.

4. Among boys, age generally had the greatest influence on the prevalence of high intakes (‘high’ being defined as above the 75th percentile for boys). Older boys were more likely to have high intakes of energy and most nutrients, as well as meat, cereals and fruit and vegetables. These age-related changes were, however, rather asynchronous. High fat intakes were more common among boys wbo had drunk more than five glasses of alcoholic beverages in the preceding week. High Ca intakes were more frequent among those who exercised vigorously, and high vitamin A intakes less common among the sons of well-educated mothers.

1. The impact of frame-size categories in weight-height tables was studied by comparing the efficiency of the body-mass index (weight/height2 (W/H2)) and weight adjusted for body-height and a body-diameter, W/(H2Dp), in predicting body fatness.

2. Body-weight, body-height, six body-diameters and four skinfold thicknesses were measured in ninety-five men and seventy women, aged between 23 and 35 years. Percentage of body fat was calculated from skinfold thicknesses using regression equations according to Durnin & Womersley (1974).

3. The inclusion of a body-diameter increased the explained variation of body fatness from 57% to 62% (knee) and 63% (shoulder) in men and from 63% to 69% (knee) in women.

4. It can be concluded that in the present population the efficiency of the prediction of percentage of body fat was not improved markedly by the inclusion of a body-diameter in the body-mass index, thus giving no support for the inclusion of frame-size categories in weight-height tables.

1. A weaning food commonly used in Nigeria was simulated by mixing processed sorghum (Sorghum bicolor) with skim-milk powder (830: 170, w/w).

2. Replacing 310 g/kg sorghum with processed cowpeas (Vigna unguiculata) resulted in an increase in protein content from 96 to 113 g/kg and an increase in biological value of the protein from 0.74 to 0.87.

3. The two mixtures were compared with an established commercial baby food by a panel of Nigerian mothers and all three foods were found to be equally acceptable.

4. Processing the cowpeas by pressure cooking followed by roller-drying reduced the trypsin-inhibitor content to minimum levels.

5. It is concluded that the protein content and quality of the weaning mixture popularly used in Nigeria could be improved by partially replacing the sorghum with cowpeas.

1. In a previous study (Kindt et al. 1983, 1984) it was assumed that a protein hydrolysate, devoid of phenylalanine, together with intact protein as given to children with phenylketonuria (PKU), was equivalent to egg or milk protein. One group of children was given this ‘PKU protein’ in amounts corresponding to the Joint FAO/WHO ad hoc Expert Committee (1973) recommendations. The results indicated that the Joint FAO/WHO ad hoc Expert Committee (1973) recommended levels of protein intake were marginal.

2. The purpose of the present study was to evaluate whether the quality of the protein hydrolysate, together with intact protein (‘PKU protein’), is equivalent to egg or milk protein. This was done using a rat bioassay. Four protein sources were used: (1) egg protein, (2) protein hydrolysate, (3) protein hydrolysate diluted with non-essential amino acids, (4) protein hydrolysate mixed with food proteins (‘PKU protein’), comparable with the diet previously used (Kindt et al. 1983, 1984).

3. The results indicated that the ‘PKU protein’ was of very high qua1ity: net protein utilization (NPU) > 90. The protein hydrolysate alone and the protein hydrolysate diluted with non-essential amino acids gave a NPU > 80.

4. The conclusion drawn from the present study is that the ‘PKU protein’, as used in the treatment of children with PKU, is equivalent to egg or milk protein. This supported the view that the Joint FAO/WHO ad hoc Expert Committee (1973) recommended levels of intake were marginal.

1. Diets containing isolated soya-bean protein induce lower levels of serum cholesterol in animals than diets containing casein. Experiments were conducted to investigate whether differences in digestibility of the proteins might explain this effect.

2. At pH 8 with pancreatic enzymes or intestinal peptidase, soya-bean protein was hydrolysed in vitro much less rapidly than casein. However, with pepsin (EC 3. 4. 23. 1) at acid pH, soya-bean protein was hydrolysed more rapidly than casein.

3. These differences in digestibility may be due to pH-dependent changes in solubility of the proteins. Casein and soya-bean protein were most soluble at alkaline and acid pH respectively.

4. Heat treatment of the proteins resulted in lower solubilities and digestibilities. Sonication of soya-bean protein at pH 7.8 increased solubility but only slightly raised digestibility.

5. When fed to rabbits, enzymically hydrolysed soya-bean protein induced a 2.3-fold higher concentration of serum cholesterol than did intact soya-bean protein. The hypocholesterolaemic effect of soya-bean protein may be partly attributable to its low solubility and digestibility at alkaline pH.

1. A low-iodine diet consisting of maize, peas (Pisum sativum), torula yeast, meat meal, maize oil and added vitamins, minerals and amino acids was given to eight pairs of adult, common cotton-eared marmosets (Caliithrix jacchus jacchus). Eight control pairs were given the same diet to which potassium iodate was added. Both groups also received low-I apple and deionized water.

2. The diet provided adequate nutrition, as confirmed by the maintenance of body-weight and good health.

3. In the I-deficient marmosets the concentration of plasma thyroxine was decreased from 140.1 nmol/l to 22.4 nmol/l and thyroid-stimulating hormone increased significantly from 1.8 ng/ml to 9.0 ng/ml compared with control marmosets, thereby indicating severe I deficiency.

4. Compared with newborn offspring from control marmosets, the thyroid glands from the I-deficient offspring showed (a) an increase in weight, (b) a decrease in I content and (c), on histological examination, hyperplasia, hypertrophy and a total absence of colloid material in the follicles.

1. Estimates of nutrient intakes based on food purchasing records modelled on the National Food Survey (NFS) were compared with nutrient intakes calculated from food consumption records based on a semi-weighed method, a combination of weighing and household measurement techniques. Of eighty-two families in Cambridge who completed the study, thirty-two were volunteers and fifty were from a random sample in which the co-operation rate was 73%.

2. The estimated energy and nutrient contents of the Cambridge food purchases were very similar to those reported by the NFS for families of similar composition and income. The energy intakes obtained by the semi-weighed method were compared with results from twenty-five studies of energy intake based on quantitative measurements of food consumption: there was no evidence to suggest that the semi-weighed method consistently under- or overestimated intakes in the Cambridge subjects.

3. Purchases adjusted to allow for waste and consumption of food by visitors contained significantly more energy, protein, carbohydrate, calcium, iron and dietary fibre than measured home food consumption. There was no significant difference in the nutrient content of purchases and consumption per 4.184 MJ (1000 kcal), with the exception of Fe and ascorbic acid.

4. Measured wastage of edible food in thirty-one families averaged 3.2% of purchases. Estimate of wastage in all eighty-two families was 3.8%, and consumption of food by visitors accounted for 3.0% of purchases.

5. The excess of purchases over measured home food energy intake is probably accounted for by a net increase in larder stocks rather than wastage, consumption of food by visitors, or under-recording of intakes.

1. Breath hydrogen and methane were determined by gas chromatography in eleven normal individuals given a low-fibre, mixed diet (control) and after ingestion of 20–50 g Palatinit®/d, an equimolar mixture of D-glucosyl-α(1 → 1)-D-mannitol and D-glucosyl-α(1 → 6)-D-glucitol (Isomalt®).

2. A linear relation was found (r 0.85; P < 0.001) between the amount of Palatinit ingested and breath H2 per 10 h in subjects who did not exhale methane. If methane was formed in addition to H2, the sum of both gases followed a linear dose-effect relation.

3. The mouth-to-caecum time, indicated by the first increase in breath H2, after ingestion, was shortened by about half, yet no sign of diarrhoea was observed. Stool weight and stool frequency did not change significantly.

4. The linear relation between a dose of 20–50 g Palatinit and exhalation of H2 (eventually plus methane) indicated that a relatively constant fraction of the dose given underwent cleavage and absorption in the small intestine, the remainder being transported into the large bowel. Microbial gas formation in the colon as well as the fractional transfer of these gases into the expiratory air occurred at fixed proportions, thus allowing an insight into colonic microbial contributions to carbohydrate utilization in the human large bowel.

1. The metabolism of 65Zn administered intramuscularly (Expt 1) or enterally (Expt 2) at the beginning of pregnancy in rats given a control or marginal-Zn diet was measured. In Expt 2 a comparison was also made between pregnant and non-pregnant rats. The loss of 65Zn (assumed to represent labile body Zn) was markedly reduced in animals fed on a marginal-Zn diet compared with controls, and this effect occurred very rapidly, i.e. within 48 h of introducing the marginal-Zn diet. Pregnancy itself had a much less important effect on 65Zn turnover than diet. Transfer of 65Zn to the litter was significantly greater in the animals fed on a marginal-Zn diet compared with controls but total Zn transfer was reduced.

2. The effect of length of time on a marginal-Zn diet on fetal growth and composition was examined. Compared with controls, fetal weight was significantly greater in litters from mothers fed on a marginal-Zn diet during the last 4, 7 or 14 d of pregnancy, but no different in litters from mothers fed on a marginal-Zn diet throughout pregnancy. There were no differences in the proportions of protein or fat in the fetuses from mothers fed on the control or marginal-Zn diets but the Zn concentration was lower in litters from mothers fed on a marginal-Zn diet during part or ail of the pregnancy when compared with controls.

3. The transfer of 65Zn from mothers to litters during birth and the first 3 d of lactation was measured. There were no differences in maternal or litter 65Zn just before or just after birth, but within 72 h maternal 65Zn had significantly decreased and litter 65Zn increased. Increases in litter size were associated with greater total litter 65Zn but reduced individual fetal 65Zn.

4. These experiments demonstrate the importance of an adequate daily supply of Zn during pregnancy. Although there is room for adaptation to a marginal-Zn intake (by reducing Zn excretion) the maintenance of Zn homeostasis is only possible in the absence of other forms of stress, such as pregnancy, to the body. The consequence of insufficient Zn at times of rapid fetal growth on carbohydrate and lipid metabolism warrants further investigation.

1. The effects of a 100 g/kg diet substitution of bagasse on the body-weight gain, food consumption and faecal dry weight of mice given a high-sucrose diet and on the activities of hepatic glucose-6-phosphate dehydrogenase (EC 1. 1. 1. 49), 6-phosphogluconate dehydrogenase (EC I. I. I. 44), malate dehydrogenase (oxaloacetatedecarboxylating) (NADP+) (EC I. I. I. 40), ATP-citrate (pro-3S) lyase (EC 4. 1. 3.8), 6-phosphofructokinase (EC 2. 7. 1. II), pyruvate kinase (EC 2.7. 1. 40) and fructose-1,6-bisphosphatase (EC 3. 1. 3. II) were studied.

2. Bagasse had no effect on body-weight gain, food consumption or faecal dry weight.

3. Bagasse decreased the activities of glucose-6-phosphate dehydrogenase, 6-phosphogluconate dehydrogenase and phosphofructokinase expressed on a wet weight basis and on a protein basis.

4. Bagasse decreased the activities of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydro-genase expressed on a body-weight basis.

5. These results suggest that bagasse decreases the flux through some pathways of hepatic lipogenesis when mice are given high-sucrose diets.

1. Thirty-three male, weanling guinea-pigs were divided into two groups. One group was fed on a zinc-deficient (ZnD) diet (1.25 mg Zn/kg diet) for 45 d and then divided into two subgroups: subgroup 1 continued on diet ZnD while subgroup 2 was fed on a Zn-repleted diet (100 mg Zn/kg diet) for 15 d. The second group was fed on a Zn-adequate diet (50 mg Zn/kg diet) for 60 d.

2. Zn deficiency could be produced within 21 d as evidenced by clinical signs and reduction in serum Zn concentration.

3. Clinical signs exhibited were listlessness, scabby lesions on skin near the foot pads, mild alopecia and a reduction in body-weight gain. No effect was observed on food intake.

4. Significant decreases in packed cell volume and total leucocyte count from 48 d onwards accompanied by absolute lymphocytopenia and relative neutrophilia were observed.

5. Supplementation of Zn in the Zn-repleted group resulted in marked, though incomplete, improvement with regard to serum Zn concentration, clinical signs and haematological changes.

1. The role of bacterial, dietary and intestinal phytases (EC 3. 1. 3. 8) in the hydrolysis of phytate was investigated in the golden hamster and rat by assaying phytase in the small intestine and by measuring the disappearance of phytate from the stomach and large intestine, using chromium oxide as an insoluble solid-phase marker.

2. It was confirmed that an active phytase was present in the proximal third of the small intestine of the rat but the enzyme was undetectable in the hamster.

3. Extensive bacterial breakdown of phytate occurred in the pregastric pouch and true stomach of the hamster with both phytase-containing and phytase-free diets, with phytate digestibilities in the true stomach ranging from 0.69–0.90, confirming that the hamster can be regarded as a pseudo-ruminant.

4. With a phytase-free diet, the digestibility of phytate in the stomach of the rat was very low (0.05) but with a wheat-based diet substantial breakdown of phytate occurred (digestibility up to 0.49), presumably under the influence of the cereal phytase.

5. Intestinal phytase did not appear to be of great significance in the rat but some further hydrolysis of the residual phytate probably occurred in the large intestine of both species by bacterial phytase.

1. A technique for measuring gastric emptying in growing pigs by complete removal of digesta through a gastric cannula is described.

2. Four pigs were fitted with gastric cannulas and each was used in three trials.

3. The effects of level of feeding (trial 1), cellulose (C), maize-oil (MO) or sucrose (SU) supplementation (trial 2) and the level of water intake (trial 3) on gastric emptying of digesta, dry matter (DM) and nitrogen from a barley-weatings-soya-bean (B) diet were measured during 4 h after the morning feed.

4. In trial 1, pigs were given 0.66, 0.83, 1. 00 or 1. 17 times their standard level of feeding. As the level of feeding rose, so the weight of digesta, DM and N emptied in the first hour after feeding increased. This trend continued to some extent in the second hour, but no effects of level of feeding were seen in the third and fourth hours.

5. In trial 2, maize-oil addition to the diet significantly reduced the gastric-emptying rate of DM in the second hour after feeding, compared with the rates for diet C. The rate of N emptying was significantly slower for diets MO and SU than for diet C.

6. In trial 3 there were no significant effects of water intake level (1. 75, 2. 50 and 3. 25 times the weight of diet) on the rate of DM or N emptying from the stomach. The rate of digesta (and thus of water) emptying in the first hour after feeding increased significantly as the water intake rose.

7. It was concluded that because the pattern of gastric emptying was very similar despite large differences in nutritional inputs, an important property of the process appeared to be resilience.

1. Sheep fitted with re-entrant canulas in the proximal duodenum and terminal ileum were used to determine the amount of α-glucoside entering, and apparently disappearing from, the small intestine when either dried-grass or ground maize-based diets were fed. The fate of any α-glucoside entering the small intestine was studied by comparing the net disappearance of such a-glucoside from the small intestine with the absorption of glucose into the mesenteric venous blood.

2. Glucose absorption from the small intestine was measured in sheep equipped with catheters in the mesenteric vein and carotid artery. A continuous infusion of [6-3H]glucose was used to determine glucose utilization by the mesenteric-drained viscera and the whole-body glucose turnover rate (GTR).

3. The amounts of α-glucoside entering the small intestine when the dried-grass and maize-based diets were given were 13.9 (SE 1.5) and 95.4 (SE 16.2) g/24 h respectively; apparent digestibilities of such α-glucoside in the small intestine were 60 and 90% respectively.

4. The net absorption of glucose into the mesenteric venous blood was —2.03 (SE 1.20) and 19.28 (SE 0.75) mmol/h for the dried-grass and maize-based diets respectively. Similarly, total glucose absorption amounted to 1.52 (SE 1.35) and 23.33 (SE 1.86) mmol/h (equivalent to 7 and 101 g/24 h respectively). These values represented 83 and 11 1% of the a-glucoside apparently disappearing from the small intestine, determined using the re-entrant cannulated sheep.

5. Total glucose absorption represented 8 and 61% of the whole-body GTR for the dried-grass and maize-based diets respectively. Endogenous glucose production was significantly lower when the sheep were fed on the maize-based diet compared with the dried-grass diet.

6. The mesenteric-drained viscera metabolized a small amount of glucose, equivalent to 234 and 17% of the total glucose absorbed for the dried-grass and maize-based diets respectively.

7. It is concluded that a large proportion of the starch entering the small intestine of sheep given a maize-based diet is digested and absorbed as glucose, and thus contributes to the whole-body GTR.

1. The effect of an exogenous supply of glucose, provided by the digestion of maize starch in the small intestine, on endogenous glucose metabolism and insulin action was studied in sheep using the euglycaemic insulin clamp procedure.

2. Insulin was infused intravenously at rates of 0.2, 0.5, 1.0 and 6.0 mU/min per kg live weight for four consecutive periods in each of four sheep fed on dried-grass and maize-based diets. Glucose was also infused intravenously at a variable rate, sufficient to maintain the plasma glucose concentration at basal levels. Whole-body rates of glucose metabolism were determined using a continuous infusion of [6-3H]glucose.

3. From the resultinginsulin dose-response curves, it was observed that, when the sheep were fed on the dried-grass diet, the responsiveness of glucose metabolism to insulin was less than that reported for non-ruminants.

4. When fed the maize-based diet, the glucose metabolic clearance rates (MCR) observed during insulin infusions were significantly greater (P < 0.05) than those observed for the dried-grass diet. However, after correcting for the non-insulin-mediated glucose disposal, differences between diets were not significant.

5. The sensitivity of glucose utilization to insulin was not affected by diet. The plasma insulin concentrations causing half-maximal insulin-mediated glucose MCR were 103 (SE 21) and 85 (SE 11) mU/l for the dried-grass and maize-based diets respectively.

6. The sensitivity of endogenous glucose production to insulin was also unaffected by diet. The plasma insulin concentrations resulting in the suppression of endogenous glucose production to half the basal level were 80 (SE 26) and 89 (SE 29) mU/l for the dried-grass and maize-based diets respectively.

7. It is concluded that the observed increase in glucose utilization on the maize-based diet was due partly to a slight change in responsiveness to insulin and also partly to a change in the rate of non-insulin-mediated glucose disposal.

1. Rapeseed (Brussicu nupus), barley grain, ryegrass (Loliurnperenne) and barley straw were labelled with 15N as an internal marker. The dilution of 15N was used to estimate microbial N (RMN15N) in the feed residues in nylon bags incubated in the rumen for 5, 12 and 24 h. For comparative purposes, diaminopimelic acid (DAP) content of the residues was also determined and rumen microbial N (RMNDAP-N) in the feed residues calculated using DAP as a bacterial marker. The influence of two bag pore sizes (20 μm and 40 μm) with different sample sizes (1 g and 5 g respectively) was also studied.

2. For all feeds, the average disappearance of 15N was faster than that of total N, the difference between N and 15N disappearance being marked with barley, ryegrass and barley straw. The disappearance of microbially corrected dry matter (DM; correction calculated from the 15N values) was, accordingly, always faster than the uncorrected DM disappearance. Except for the bag pore/sample size effect for N disappearance, significant (P < 0.01–0.001) feed, pore/sample size and incubation-time effects were always found for the disappearance values.

3. Errors (%) resulting from the microbial contamination (calculated from the 15N values) in N-loss measurement with rapeseed, barley, ryegrass and barley straw, at 5, 12 and 24 h in 20 μm bags were respectively: –1.8, –3.9, –0.9; –3.8, –22.4, –3.8; –7.2, –4.1, –2.9; –164.5, –146.3, –204.6. In 40 μm bags the corresponding errors were respectively: –4.4, –1.2, –0.7; –26.1, –10.5, –3.9; –13.2, –6.4, –5.5; –221.2, –310.1, –1284.6.

4. The largest residual proportions of RMN15N, RMNDAP-N and DAP-N (% of total N) were found in barley straw, followed by barley, ryegrass and rapeseed, in that order. RMN15N (g/kg residual DM) followed the descending order: barley, ryegrass, straw, rapeseed. RMNDAP-N, (g/kg residual DM) and DAP (mg/kg residual DM) followed the descending order: barley, ryegrass, rapeseed and barley straw. Feed, pore/sample size and incubation-time effects were always significant (P < 0.001).

5. With 40 μm bags RMN15N values of barley, ryegrass and barley straw (expressed as % of total N or g/kg residual DM) were substantially higher than those of RMNDAP-N With 20 μm bags the RMN15N and RMNDAP-N values were generally quite close for these feeds. With rapeseed residues, RMN15N was clearly lower than RMNDAP-N with 20 μm bags, but only small differences were found with 40 μm bags.

6. A more diverse microbial colonization of feed samples was indicated in the 40 μm bags compared with the 20 μm bags. It was also noticed, consequently, that in most cases larger error resulted from the microbial contamination in 40 μm bags.

1. In a 4 x 4 Latin square experiment four cows were given, twice daily, diets consisting of (g/kg dry matter (DM)) 500 barley, 400 grass silage and 100 soya-bean meal. The diets were given at either 1.15 (L) or 2.3 (H) times maintenance energy requirements and the soya-bean meal was either untreated (U) or formaldehyde (HCH0)-treated (T).

2. The passage of digesta to the duodenum was estimated using chromic oxide as a flow marker; 35S was used to estimate the amount of microbial protein entering the small intestine. A microbial fraction was prepared by differential centrifugation from duodenal digesta. Samples of bacteria and of protozoa from rumen digesta were also prepared.

3. The total amino acid contents of feedingstuffs, duodenal digesta, duodenal microbial material, rumen bacteria and rumen protozoa were determined by ion-exchange chromatography. The D-alanine and D-glutamic acid contents of the samples were determined by gas–liquid chromatography.

4. The quantity of each amino acid entering the small intestine was significantly (P < 0,001) increased by increasing DM intake and tended to be increased by formaldehyde-treatment of the soya-bean meal. There were net losses of all amino acids across the forestomachs except for lysine, methione, o-alanine and D-glutamic acid for which there were net gains.

5. There were significant (P < 0.05) differences in amino acid composition between rumen bacteria and duodenal microbial material; differences in amino acid composition between rumen bacteria and rumen protozoa were also observed.

6. D-Alanine and D-glutamic acid were present in the silage but not in the barley or either of the soya-bean meals. All samples of microbes and digesta contained D-alanine and D-glutamic acid.

7. The use of D-ahine and D-glUtamiC acid as markers for microbial nitrogen entering the small intestine was assessed. Estimates of the quantities of microbial N entering the small intestine based on the D-alanine or D-glutamic acid contents of rumen bacteria or duodenal microbes were significantly higher than those determined using 35S as a marker.

1. The effect of peroral administration of xylitol on the absorption of iron and the activities of xanthine oxidase (EC 1. 2. 3. 2) and ferroxidase in rat duodenal wall was studied.

2. Adult male rats were given the basal diet containing 200 g xylitol/kg or the same diet containing no added carbohydrates for 8 weeks. Both feeding groups comprised twelve animals.

3. Xylitol significantly increased serum and liver Fe concentrations with a concomitant, significant increase in the duodenal xanthine oxidase activities, but caused a marginal increase in the duodenal ferroxidase activities.

4. In vitro, sugar alcohols reduced the binding rate of Fe to transferrin.

5. The xylitol-induced increase of Fe absorption may involve the following mechanism: the high intraluminal xylitol concentration of the xylitol-fed rats keeps Fe in the form of a soluble complex for a prolonged period of time, due to the slow absorption of xylitol. The polyol-Fe complex in turn induces xanthine oxidase and ferroxidase formation.

1. Nitrogen balance was studied in growing male SPF-rats fed on diets in which each essential amino acid was varied from zero to about 120% of optimum requirement. From the balance results, optimum and maintenance requirements were estimated as well as the efficiency of utilization of amino acids for growth and growth + maintenance.

2. N balance increased with increasing dietary level of the deficient amino acid; the response gradually diminished as the content of the amino acid approached optimum. At zero level of intake, negative N balance was found for all amino acids except histidine. The highest loss of body N was found in the sulphur-amino-acid-free diet and the lowest one in the lysine-free diet.

3. Maximal utilization of essential amino acids for growth was found at dietary levels corresponding to 30–60% of optimum requirement and ranged from about 0.65 to 0.85 except for S amino acids and histidine. The utilization of S amino acids was about 0.55 while that of histidine exceeded 1.0. The utilization of amino acids for growth-tmaintenance was maximal at the lowest levels of intake and gradually decreased as the dietary concentration of the limiting amino acid increased. At dietary levels near optimum the utilization was about 06–07, except for S amino acids where the utilization was less than 0.5.

1. The ability of ytterbium acetabe (Yb acetate) to fulfil the requirements of a particulate-phase digesta-flow marker in a dual-phase marker system, and of the indigestible acid-detergent-fibre fraction of the feed (IADF) to act as a digesta flow marker, were examined using six mature wether sheep given a diet of dried grass (1 kg dry matter (DM)/d).

2. CrEDTA was continuously infused (240 mg chromium/d) into the rumen of all sheep and Yb acetate was also continuously infused (100 mg Yb/d) into the rumen of three of the sheep. At this level of infusion the equilibrium concentration of Yb in rumen, duodenal and ileal digesta and in faeces could be reliably measured by atomic absorption spectrometry.

3. Estimates of faecal DM excretion based on either Yb or IADF did not differ (P > 0.05) from that determined by total collection, whereas estimates based on Cr were significantly (P < 0.05) lower. Urinary excretion accounted for 3.1% of the infused Cr but no Yb was detected in urine. Estimates of ileal DM flow, assuming total marker recovery, were similar (P > 0.05) with all three markers, whereas the estimate of duodenal DM flow based on IADF was lower (P < 0.05) than the estimates based on either Cr or Yb.

4. Compared with the infusion of Cr alone, the infusion of Cr and Yb had no effect (P > 0.05) on nutrient flows at the duodenum, ileum and in faeces nor on microbial degradative activity, volatile fatty acid production and N metabolism in the rumen.

5. Polyester bag and in vitro studies showed that pre-labelling the dried grass with up to 285 mg Yb>g DM did not affect its susceptibility to microbial degradation.

6. The Yb in rumen, duodenal and ileal digesta was predominantly (> 90%) associated with the particulate matter but was not uniformly distributed and its concentration increased as particle size decreased.

7. The use of CrEDTA and Yb acetate as a dual-phase marker system proved more reliable in estimating ‘true’ duodenal flow than the use of the individual markers when the digesta sample was unrepresentative.