Critical review of chemical and enzymatic methods for the estimation of nutritive value in roughages

Abstract

The methods used classically for estimating the nutritive value of temperate forages (grasses and legumes) have to be corrected to take into account the specificity of roughages (crop residues, numerous tropical species, trees and shrubs), which may affect the results of assays. Tropical forages have their own physical and chemical features linked to genetic factors. For example, they may contain tannins liable to form protein-tannin complexes and to inhibit certain enzymes. In general, the nutritive values of roughages cannot be predicted as accurately as those of temperate forages. For tannin-free temperate and tropical forages, in vivo organic matter digestibility can be predicted by chemical methods based on the determination of cell wall contents or lignin, but these methods are less accurate than enzymatic methods (cellulase). Digestible protein content (DCP) is closely linked to crude protein content (CP) in temperate grassland legumes and in tropical grasses. To determine DCP of tropical herbaceous or woody dicotyledons, cell wall nitrogen (ADIN: acid detergent insoluble nitrogen), which closely corresponds to indigestible nitrogen content, also has to be determined. Tannins, which are present in many dicotyledons, also modify the overall digestibility of nitrogen, but their effects depend on the types of tannins present and on other factors, and are difficult to predict. Prediction of PDI (Protéines réellement digestibles dans l'Intestin grêle, i.e. digestible protein in the small intestine) requires knowing the theoretical degradability of protein in the rumen (Deg) and the true digestibility of dietary proteins (dr) in the small intestine. The methods of Deg prediction using commercially available proteases permit a useful classification of temperate and tropical forages, but must be used carefully especially in tannin-containing forages. Chemical and enzymatic methods of dr prediction (pepsin-pancreatin) have not been widely used for temperate or tropical forages. Tannin determination provides a qualitative interpretation of differences in the nutritive value between tannin-containing forages. However, none of the numerous methods tested to date can be considered as predictive of the nutritive value. This is partly due to the diversity of the phenol groups present and their methods of determination and to interactions with other digestible nutrients, which makes it difficult to predict the effects of tannins on digestion. For these forages, Van Soest's fractionation corrected for tannin content is a reliable predictive model for energy content, while NDIN (neutral detergent insoluble nitrogen) and ADIN determinations provide an initial approach to degradation mechanisms occurring in the rumen and intestine.