Apparent digestibility of forage kale (cv. Maris Kestrel) was compared with that of mixed perennial ryegrass (0.85) and clover (0.15) herbage in 2 experiments. In a 3rd experiment measurements were made of the partition of digestion between the reticulo-rumen and post-ruminal regions of the digestive tract in sheep fed kale, using intra-ruminal infusion of inert ruthenium phenanthroline and Cr-EDTA markers and time sequence sampling from the duodenum. Results in experiment 3 were compared with predicted values for a range of ruminant diets, including digestion of fresh perennial ryegrass and white clover. In experiments 1 and 2 the ratio of readily fermentable to structural carbohydrate (CHO) was much higher in kale (2.6-3.2) than in mixed ryegrass and clover herbage (0.6), and showed little difference between kale leaf and stem components. Apparent digestibilities of organic matter (OM), cellulose and hemicellulose were 80-90% for kale diets, varied little with changes in leaf/stem ratio, and were generally slightly higher than for mixed ryegrass and clover herbage. Calculated metabolizable energy (ME) concentrations were 12.0 and 11.7 MJ ME/kg D.M. for kale and mixed ryegrass and clover, respectively. Despite large intakes of water-soluble CHO and pectins in kale-fed sheep in experiment 3, duodenal flows were small and, respectively, 92 and 98% of the total amounts digested were digested in the rumen. Ruminal cellulose digestion (95% total digested) was as predicted for normal ruminant diets, but less hemicellulose was digested in the rumen (71% total digested) than predicted. Total N flow at the duodenum was 5 g/day less than N intake, and, respectively, 27 and 66% of the N apparently digested was accounted for as total N absorption of amino acids from the small intestine was less for the kale diet (14% ME intake) than for diets of fresh perennial ryegrass and white clover (20% ME intake). Possible reasons are that 10% more of the OM ingested was digested in the rumens of sheep fed kale than was predicted, that the ratio of dietary total N/100 g digestible OM was much less for kale (3.7) than for perennial ryegrass and white clover (5.6) diets, and that protein formed a lower proportion of the total N in kale diets (75%) than in diets of ryegrass and clover (90%). Despite 3.3 g of S-methyl-L-cysteine sulfoxide (SMCO) being consumed per day, none was detected in either rumen or duodenal contents or in feces. Thus, complete conversion of SMCO to dimethyl disulfide occurred in the rumen, dimethyl disulfide may have caused the depression in ruminal digestion of hemicellulose, and the digestion products of kale are deficient in absorbed amino acids relative to other forms of energy, thus explaining responses to essential amino acid supplementation in lambs grazing kale.
