A comparison of some laboratory techniques for the estimation of the digestibility of the organic matter in forage samples.

The 194 samples of forage were of 6 groups; grass or other forage dried, fresh or ensiled. Workers from 5 institutes analysed some of the samples by their usual methods, 10 methods altogether. Digestibility estimated with sheep was regressed on the other results and correlation coefficients and residual standard deviations were calculated; those were closer for groups of forages than for all forages together. The least accurate predictor of digestibility in vivo was crude fibre content, which is much used in the Netherlands. The best method for each fodder group was the method of Tilley and Terry (NAR 34, 156) as modified by Koelen, Kemmink and Dijkstra (De bepaling van de voederwaarde van ruwvoeders met behulp van de in vitroverteerbaarheid. Internal Report, Institute for Animal Feeding and Nutrition Research 'Hoorn' (1969), No. 27, available in English). It took least work but required sheep rumen fluid. For all groups together, the cell-wall constituents method of Gaillard and Nijkamp (NAR 38, 8034) was better for 54 samples but the residual standard deviation was higher than within groups; it required more labour, but not rumen fluid. (Abstract retrieved from CAB Abstracts by CABI’s permission