The use of a cellulase technique to predict digestibility, metabolizable and net energy of forages

Abstract

With samples of 24 maize silages, 21 grass silages and 18 grass hays the 2N HCl-cellulase-pepsin method and several versions of the pepsin-cellulase method were compared for their accuracy in predicting in vivo organic matter digestibility. The rumen fluid-pepsin in vitro method was used as reference. As none of the tested methods proved to be the best for all forage types, the procedure with the best uniformity was chosen for further study. It was originally developed for concentrates and includes three steps: (1) pepsin in 0.1 N HCl at 40°C for 24 h; (2) heating the same solution at 80°C for 45 min to hydrolyse starch; (3) cellulase (from Trichoderma viride) at 40°C for 24 h. With this technique cellulase-digestible organic matter of the dry matter ( DOMD) was determined on 50 maize silages, 50 grass silages, 24 grass hays and 16 straws for which composition and digestibility in vitro and in vivo were known. By means of linear regression analysis prediction equations for in vivo DOMD, metabolizable and net energy were derived. The best equations for net energy based on cellulase DOMD showed coefficients of variation of 3.3, 4.2 and 2.6% for maize silage, grass silage and grass hay, respectively, and of 2.7, 3.5 and 2.6%, respectively, for those based on in vitro DOMD. Prediction of feeding value of straw was very imprecise. In spite of the slightly lower accuracy, the cellulase technique selected has operational advantages over the rumen-fluid in vitro method for routine feed evaluation.