Prediction of the energy value of pig feeds by enzymatic cell wall analysis

Abstract

The purpose of this study performed in connection with the general research programme on prediction of the energy value of compound feeds for pigs was to estimate the advantage of a new rapid enzymatic method for cell wall analysis developed by INRA (WICW : waterinsoluble cell wall). The experiment involved 43 complex diets containing various types of dietary fibres. The feed energy value (DE) was measured in castrated male pigs of the same weight range (about 40 kg) using four animals per diet (a total of 172 digestive balances). With the aim of increasing the accuracy of the analytical data and the impact of prediction equations, the experimental feeds underwent several analyses in various laboratories. Main results :

1 - On the basis of classical criteria for choice of equations for prediction of feed energy value (residual standard deviation, R²), the models involving the WICW variable seemed to be the most accurate. However, they did not show a very marked advantage relative to prediction models involving crude fibre or cell wall residues according to Van Soest.

2 - When combining the errors related to the models (RSD) and the errors associated with the analyses (SLAB), equations with WICW always proved to be better than those using the other cell wall indicators.

3 - With the best models using WICW, the total uncertainty of equations for prediction of digestible energy (equation + analyses) did not exceed 75 kcal per kg dry matter. Equation 25 (Lipids, WICW,Ashes, Crude Protein) proved to be the most interesting.

4 - As the choice of the best predictors depends to a large extent on the accuracy of analysis, further large-scale investigations are required for better determining the reproductibility of the WICW method as compared to other techniques for plant cell wall analysis and confirm its interest for prediction of the energy value of pig feeds.