Azevêdo et al., 2011. Rev. Bras. Zootec., 40 (2): 391-402

The objective of this work was to evaluate the equations adopted by the NRC and Detmann to estimate the energy content of agricultural and agroindustrial byproducts. It was evaluated byproducts of pineapple, cocoa, palm oil, beans, sunflower, guava, cassava bark, cassava stem, raw cassava, papaya, mango, passion fruit, turnip and corn gluten meal. All animals received corn silage and the mixture urea/ammonium sulfate (9:1) so rations with the two levels of each byproduct remained isonitrogenous. Among the models for digestible crude protein prediction (CPad), Detmann 2 model presented difference by the hypotheses test for joint null while estimates by the NRC and Detmann 1 models were similar to those obtained by in vivo observations. For digestible ether extract (EEad), digestible non-fibrous carbohydrates (NFCad) and digestible neutral detergent fiber (NDFd), the estimates obtained by the model proposed by the NRC differed, by the test of significance of hypotheses of joint null, to in vitro observations and the best adjustment was obtained by the model suggested by Detmann, having as a consequence, the value. However, among all the digestible fractions, NDFd presented the greatest dispersion of points regarded to equality line. NRC and Detmann models can be considered as efficient in predicting total digestible nutrients (TDN) and digestible energy (DE) because the estimates do not differ from the observed values, however they also present great dispersion of the points regarded to equality line. It is recommended to use the models proposed by Detmann for predicting digestible fractions of crude protein, ether extract and non fibrous carbohydrates on agriculture and agroindustrial byproducts.