In‐vitro evaluation of european diets for pigs. Prediction of the organic matter digestibility by an enzymic method or by chemical analysis

The in‐vitro method of Metz and van der Meer simulates the digestion in pigs by consecutive incubations with pepsin/pancreatin and cellulase. This method was studied as an alternative to prediction of the in‐vivo organic matter digestibility by chemical analysis. After calibration with samples of known in‐vivo digestibility, the in‐vitro method enabled prediction of the digestibility coefficient of the organic matter obtained in trials with growing pigs. In this study, 89 samples of diets with known in‐vivo organic matter digestibility were analysed by the enzymic method. The linear relationship between digestibility of organic matter in vivo (dO) and in vitro (vOp) for all samples had a correlation coefficient ( r ) of 0.92 and a residual standard deviation (RSD) of 2.0. The natural logarithm of crude fibre (InXF) as an additional variable in the regression model improved the prediction ( r = 0.95; RSD = 1.7). This indicated that the in‐vitro method does not fully simulate fibre digestion in the pig. For the Dutch diets the prediction of dO from the measured chemical composition of the raw materials and the digestibility coefficients from the Dutch Tables of Feedstuffs showed a similar correlation to prediction with in‐vitro methods. However, the deviation of the average predicted value from the average in‐vivo value was 3.7 units when digestible nutrients were used and only 0.3 units when the in‐vitro method was used.