Interrelationships in the digestion of sheep given barley grain diets containing either fish meal, soya-bean meal or urea

Three sheep, each fitted with a ruminal cannula and duodenal re-entrant cannulae were given three isonitrogenous, isoenergetic diets in a Latin Square design. Each diet contained approximately 60% DM as barley grain and 40% of total N as either white fish meal, soya-bean meal or urea. These diets were fed continuously and supplied about 28 g N/day. Diets containing such large amounts of barley grain usually produce wide variations in the rumen volatile fatty acid (VFA) proportions and these have been correlated with various other digestive characteristics. Several ruminal and duodenal components were measured in this study and interrelationships between them sought. The molar proportions of VFA varied widely from 45-67% for acetate, 13-48% for propionate and 7-23% for butyrate. Rumen pH was positively correlated with acetate (P less than 0.01), but negatively so with propionate (P less than 0.01) and butyrate (P less than 0.1). The numbers of rumen ciliate protozoa also varied widely and were related to rumen pH (P less than 0.05) and VFA proportions; positively to acetate (P less than 0.001) and butyrate (P less than 0.01) but negatively to propionate (P less than 0.001). Duodenal N was always less than fed N. The mean composition of this duodenal N was 10.1% ammonia-N, 6.7% RNA-N, 79.0% amino acid-N and 7.3% was unaccounted for. Efficiencies of synthesis of microbial and bacterial crude protein (derived from 35S and 2,6-diaminopimelic acid data) ranged from 10.5 to 42.2 g microbial N per kg apparently digested organic matter (ADOM) and 5.0 to 27.9 g bacterial N per kg ADOM. Division of VFA patterns into either propionate or butyrate type fermentations suggested several further interrelationships. No relationship was established between these fermentation patterns and the extent or efficiency of microbial crude protein synthesis. Possible interrelationships between different digestive characteristics are discussed and a plea made for both more extensive and intensive experimentation before such designation of cause and effect can be assigned.