Comparison of methods for forage digestibility determination

The objective was to evaluate the ability of in situ and in vitro gas production methods in determining the variation in grass and grass silage organic matter digestibility (OMD) compared to in vitro pepsin–cellulase solubility, which was considered as a reference method. The variation in OMD was obtained by varying grass species, stage of maturity at harvest and preservation method (dried vs. ensiled). Five grass species (Dactylis glomerata, Phleum pratense, Lolium perenne, Festuca arundinacea and Lolium multiflorum × Festuca arundinacea) harvested from primary growth at four dates in 2005 were analyzed for chemical composition and OMD using in vitro pepsin–cellulase solubility (OMDPC), in vitro gas production (OMDGP) and three in situ methods. When using the in situ methods, the OMD was derived from traditional sequential nylon bag incubations and calculated using one-compartment or two-compartment models (OMDIS1 and OMDIS2, respectively), or from indigestible neutral detergent fibre (iNDF) concentration after 12-d incubation using empirical relationships (OMDiNDF). L. perenne had the lowest content of cell wall fractions (P<0.05) and iNDF (P<0.01) and it resulted in the highest values of OMD (P<0.01) determined by all five tested methods. However the residuals between OMDPC and the other OMD methods were smallest in P. pratense (P<0.01). The OMD values obtained by different methods were highly correlated (r > 0.91). The OMDIS1 and OMDIS2 had lower (P<0.05) average values (0.654 and 0.622, respectively) than OMDPC, OMDGP and OMDiNDF (0.747, 0.724 and 0.723, respectively). In situ and in vitro gas production methods provided also digestion rate values (kd(IS) and kd(GP), respectively). Higher values were determined for kd(GP) (at average 0.0636/h) than for kd(IS) (at average 0.0364/h) (P<0.01). In residual analysis there was also a linear bias (−0.432; P<0.01) indicating that the difference in the digestion rate estimates increased with increasing kd(GP). The results suggest that the relationship between pepsin–cellulase solubility and in vivo OMD could vary between grass species, and probably specific correction equations are needed. The in situ method clearly underestimated the rate of digestion and OMD, but high correlations with other OMD estimates and digestion rate estimated using in vitro gas production suggest that the method can be useful in ranking the feeding value of forages.