Application of the gas production technique to feed evaluation systems for ruminants

A range of feed evaluation techniques is available to predict the nutritional value of ruminant feedstuffs. The aim of this paper was to critically evaluate use of gas production (GP) data as inputs to current feed evaluation systems, as well as to mechanistic rumen models. Topics discussed include prediction of rumen degradation of feed, efficiency of microbial synthesis, and the profile of volatile fatty acids (VFA) produced. Derivation of GP models to permit calculation of extent of degradation of organic matter (OM) has been of great value, as well as to reveal assumptions underlying models to analyse GP data. Gas production techniques (GPT) have good potential to predict rumen OM degradation, in particular through provision of kinetic information, but the potential to provide parameters of degradation of OM components seems limited. Optimal use of GP data to predict microbial efficiency and VFA formed is best achieved when the possibilities, and limitations, of batch culture GPT are recognised. The mechanisms governing microbial efficiency and VFA molar proportions in the GPT are not necessarily applicable to in vivo situations. A simple model based on classic microbial growth and substrate utilisation equations for batch cultures is used to illustrate the need to carefully interpret microbial efficiency in batch culture for use in in vivo situations. The profile of VFA in GPTs is only a reflection of the dynamic VFA profile in vivo. Of crucial importance for estimation of OM degradation, microbial efficiency and amount and type of VFA formed is knowledge of the fractional passage rate of ingesta from the rumen. The value of the GPT is greatly enhanced in combination with mechanistic modelling. However, the role of the technique is not in giving direct predictions of nutrient supply