Comparison of methods to determine the energy value of feeds for ruminants

Objective. Compare different methods (MET) to calculate TND based on chemical analysis of diets and digestibility trials. Materials and methods. in vivo digestibility trials were determined using zebu animals fed hay (Cynodon dactylon L. Pers. cv. Tifton 85), corn and soybean meal. The in vitro digestibility followed the Tilley and Terry (1963) methodology (TT). Methods for estimating TND were: conventional (MET 1), calorimetric based on the combustion values quantified in a bomb calorimeter (MET 2), calorimetric based on table combustion heat values (MET 3), based on the in vivo e in vitro organic matter digestibility (MET 4), and based on the multi-compartment model by the NRC (2001) (MET 5). Data was analyzed by a completely randomized design and comparison of the average effect was conducted using Tukey test (α=5%). The Bland-Altman method was used to evaluate the exchangeability between TND values obtained from TT assay (MET 4) and MET 5. Results. The TND values estimated in vivo by MET 1, MET 3 y MET 4 showed no statistical difference (p>0.05) but differed from the value obtained by MET 2 (p<0.05), which was lower. TND values estimated by MET 5 were not exchangeable with those obtained by TT assay, which were above the ranges established by literature. Conclusions. TND estimate by MET 2 proved to be more reliable, because it did not include assumptions or omissions for energy values of the organic components. MET 5 proved to be promising for the calculation of TND.