Assessment of factors Influencing in vitro gas and methane production by meta-analysis

The analysis of specific factors affecting in vitro gas production (GP; ml/g DM), methane (CH(4); ml/g DM), and CH(4)/GP was carried out by a meta-analysis approach using 274 results from 39 papers published over 12 years. Only feeds or diets incubated without additives were considered. The common factors accounted were: the pressure in the GP equipment used (PR; constant or increasing), the incubation time (IT; 24 or ≥48 h), the collection of rumen fluid (CRF; from fasted or fed animals), the donor species (DS; sheep or cattle), and the nitrogen (N) in the buffer solution (presence or absence). In addition to these dichotomous factors (0 or 1), the NDF content (g/kg), the rumen fluid/buffer ratio (ml/ml), the rumen fluid/feed sample ratio (ml/g DM), the buffer/feed sample ratio (ml/g DM), and the buffered rumen fluid/feed sample ratio (BRF/FS, ml/g DM) were considered. However, with the exception of NDF (391.6±175.3 g/kg), only the BRF/FS (129.7±61.4 ml/g DM) was considered as potential factor to be accounted after a preliminary analysis, being the most important source of variation. A final stepwise analysis was carried out using the 5 dichotomous factors, NDF and BRF/FS contents to assess the relative magnitude of each factor on GP, CH(4) and CH(4)/GP. The most relevant factors analysed were NDF and IT for GP and CH(4) (sum of partial R^2 of 0.44 and 0.35, resp.), followed by CRF and BRF/FS (sum of partial R^2 of 0.06 and 0.09, resp.). The increasing PR was the fifth factor entering the model (partial R2 of 0.01), but affected only CH(4). The relative magnitude of factors affecting the GP/CH(4) was in order: CRF, BRF/FS, PR, and IT (R2=0.18). The analysis carried out showed that DS and N in the buffer did not influence the analysed variables. As expected, the increase of IT and the reduction of NDF enhanced the GP and CH(4). Looking at methodological factors, rumen fluid from fed animals showed greater GP and CH(4) (+37.7 and +10.6 ml/g DM, resp.) than using fluid from fasted animals. For a unit increase in BRF/FS (ml/g DM), a corresponding increase of 0.32 and 0.14 ml/g DM in GP and CH(4), resp., was observed. The increasing PR enhances CH(4) production of 4.5 ml/g DM, as compared to constant PR in the GP equipment. Results of this meta-analysis suggest that a strong standardization of methodological procedures (accumulation of gas at constant or increasing PR, CRF, IT, and BRF/FS) is needed to compare results of different trials.needed to compare results of different trials.