A comparison of the neutral detergent-cellulase method with other laboratory methods for predicting the digestibility in vivo of maize silages from three European countries

Four populations of maize silage have been studied with the objective of comparing the ability of the neutral detergent-cellulase plus amylase (NCD) procedure with various fibre methods and starch to predict digestible organic matter content in dry matter (DOMD). The four populations comprised 16 samples examined at the ADAS Feed Evaluation Unit (Population F), 16 samples from the Rowett Research Institute (Population R), 50 samples studied at the National Institute of Animal Nutrition, Ghent, Belgium (Population B) and 24 samples from the Research Institute for Livestock Feeding and Nutrition, Lelystad, Netherlands (Population D). The NCD method provided the best relationship with DOMD in vivo. The overall relationship was: DOMD (g kg−1 DM) =235 + 0.650 NCD (g kg−1 DM), R2=64.4%, r.s.d.=25.8 the NCD-based relationship for Population R had a significantly (P < 0.001) different intercept, overall, the NCD method was considerably superior at predicting DOMD in vivo than all the fibre methods examined and starch content. The reason for the behaviour of Population R cannot be explained but its retention in the overall relationship was shown to have little effect on predicted values.