A digestion technique using a commercially-available crude cellulase preparation from Trichoderma viride is described. The enzyme showed cellulase, hemicellulase and proteolytic activity when tested on herbage or herbage polysaccharides. A high correlation (r = 0.92, P < 0.001, residual standard deviation 2.5) was found between a simple one stage enzyme digestion and in vivo dry matter digestibility for a range of grass species and varieties. In vitro dry matter digestibility and detergent fibre were also determined on the same samples and the results correlated with in vivo dry matter digestibility. The cellulase method is rapid and reproducible and particularly well suited for the evaluation of plant material in breeding programmes.

The composition of oat husk has been determined in (i) 11 diverse oat genotypes, (ii) current high-yielding spring and winter varieties from trials in various regions of Britain, and ( 5 ) samples obtained from two British oat mills. Husks were comprised mainly of cell wall ( > 83 %) with about equal quantities of cellulose and hemicellulose ( N 30-35 %). Lignin and ash contents were 2-10% and 3.5-9% respectively. Protein and oil contents were low (1.6-5 % and 1-2.2 %) and only a proportion of the protein (40-60%) and total oil (5-30%) were of apparent nutritional significance. Starch and water-soluble carbohydrates were very low ( < 2 % and < 1 %). Pepsin-cellulase digestibility was low (7-14%) and it was negatively correlated with lignin content in the 11 genotypes. There were significant genetic differences in composition within both the 11 genotypes and the trial varieties; however, the range was small and there were significant differences due to location among the trial varieties. Husk, which was handseparated from oat mill samples, was generally inferior in quality to the genotypes and varieties. However, due to groat contamination, industrially separated samples from the oat mills were similar in composition to those from other sources. It is confirmed that oat husk is a poor quality feedstuff and the most significant improvement in grain quality could be made by breeding for reduced husk content.

Cellulase preparations from different fungi differed markedly in their ability to solubilise herbage and cellulose; T. viride cellulase was the most active, solubilking 70% of cellulose paper in 24 h. The correlation of cellulase solubility with the in vivo and in vitro dry matter digestibility of grasses, and with the in vitro digestibility of legumes was markedly improved by pretreatment of the herbage with acid pepsin. Using the two stage technique, closely similar regression lines were obtained for predicting the in vitro digestibility of both grasses and legumes. Use of the pepsin treatment also enabled a less active Basidiomycete cellulase to be used with results very similar to those obtained with the T. viride enzyme. The technique is proposed as a more rapid, convenient and precise method of predicting digestibility than the usual in vitro procedure. IntroductionA high correlation has been previously established between the in vivo dry matter digestibility of grasses and their solubility in a buffered cellulase so1ution.l The cellulase used was a crude preparation from Trichoderma viride which also showed hemicellulase and protease activity ; treatment of the residue from the cellulase digestion with pepsin did not improve the correlation with in vivo digestibility.The application of cellulases from other fungal sources has been subsequently investigated and showed, in accord with other published work,2 that the cellulolytic activity of enzymes from different fungi varies within wide limits when plant material or native cellulose is used as a substrate. In the course of this work it was noted that pepsin pretreatment of herbage could influence the subsequent solubility in cellulase solutions and the accuracy of predicting in vivo digestibility. The present paper is concerned with the results of this investigation, viz. (a) the cellulolytic activity of different fungal cellulases and (b) the effect of pepsin treatment of herbage on subsequent solubility in different cellulase solutions and on the prediction of in vivo and in vitro digestibility of herbage.

and there were no significant differences between cv. RvP and Bb 1277.The results are discussed in relation to the breeding potential of the high nutritive value of the Po valley ecotype Bb 1277 and the use ofthe DMD and FI determinations for selecting for improved herbage quality in Italian ryegrass.The influence of harvesting date on the nutritive value of the Italian ryegrass cv. RvP from Belgium and of Bb 1277, an introduced ecotype from the Po valley, was investigated.Ten clonally replicated genotypes of each population were grown as closely spaced plants and two ramets of each genotype harvested every 4 d for a period of 28 d after inflorescence emergence. Dry matter digestibility (DMD) was estimated and fibrosity index (FI) determined on both the leaf and stem fractions, and cellulose and lignin concentrations were determined on the stem fractions.The nutritive value of the leaf in terms of both DMD and FI, which was the same for both cv. RvP and Bb 1277 throughout the period of harvesting, was always higher and declined at a slower rate with increasing maturity than that of the stem. The stem of Bb 1277 declined in DMD and increased in FI and cellulose concentration at a significantly slower rate than cv. RvP. Significantly higher levels of DMD were detected in the stem of Bb 1277 for harvests 5 to 8 and significantly lower levels of FI and cellulose concentration from harvests 4 to 8. It was not, however, possible to detect any significant diflerences in stem lignin concentration. Dry matter yields increased throughout the period of harvesting

No abstract