Three ruminally and duodenally cannulated non-lactating Finnish Ayrshire cows were used to investigate ruminal and intestinal digestion of cell-wall carbohydrates by a combined in situ method. Five grasses cut at 10 d intervals were incubated in the rumen for 0, 6, 12, 24, 48, 72 and 96 h, and the undegraded residues were exposed to intestinal digestion. With advancing maturity of grass both the rate and extent of cell-wall digestion decreased. At early stages of growth the decreases were faster for the rate of digestion and at late stages of growth for the extent of digestion. Applying a passage rate of 0.02/h in one compartmental rumen model resulted in digestibility values markedly lower than typically observed in vivo. However, applying a rumen model incorporating a selective retention of particles and time-dependent release of particles from the non-escapable pool resulted in much higher digestibility values. Recovery of lignin after 96 h ruminal incubation with a subsequent mobile-bag incubation was very low (from 244 to 460 mg/g). Intestinal disappearance of neutral-detergent fibre (NDF) and hemicellulose decreased with advancing maturity of grass and with increasing length of preceding ruminal incubation period, i.e. with decreasing potential digestibility of the material. Disappearance of hemicellulose was much greater than that of cellulose for intact grasses but the difference diminished with increasing length of preceding rumen incubation period. On average, 195 mg/g of potentially digestible NDF disappeared from the mobile bags in the intestines. The post-ruminal digestion as a proportion of the total NDF digestibility varied between 0.034 and 0.058. Despite methodological problems both in ruminal in situ and intestinal mobile bag techniques, these methods can be used to investigate ruminal and intestinal cell-wall digestion and to partition cell-wall digestibility between ruminal and post-ruminal digestion providing that appropriate rumen models are used.

To study the importance of site of recovery and other factors related to mobile-bag (MB) digestion values, two consecutive experiments in which diets were applied in a 3 × 3 Latin square design, were carried out with cannulated growing heifers. In Expt 1, several types of experimental feed were exposed to intestinal digestion in mobile bags made of two cloth types and filled with intact or rumen-undegradable (RUD) feed material to be recovered either from the ileum (IB) or faeces (FB). In Expt 2, mean retention time (MRT) of Yb-labelled digesta particles within the intestine and in vivo digestibility of diets were measured. With vegetable concentrates, FB resulted generally in slight overestimation of small-intestinal dry matter and N digestion, while with meat-and-bone meal no difference between FB and IB was found. The respective N digestibility of RUD late-cut silage was clearly underestimated as measured from FB. The disappearance of neutral-detergent fibre (NDF) of all feeds under test was higher from FB than from IB. It was not possible to isolate the influence of the large intestine on the MB values by changing bag cloth type. Irrespective of the longer retention time of bags and longer MRT of Yb in the intestine on a low as compared with a high level of feeding, only NDF disappearance of feeds increased due to lower feeding level. Altering the diet type to increase large-intestinal fermentation, as indicated in vivo, usually had no effect on the MB values. It is concluded that the site of collection of bags does not practically affect small-intestinal digestion values of feed N, unless the feed is rich in fibre.