Estimation of energy value in hay

Abstract

The concentration of net energy for lactation (NEL) of 100 hay samples that were collected in 1999 and 2000 was estimated by the use of in vitro method, which was based on the gas produced during the incubation of samples with rumen liquor. Values were compared to the estimates obtained by the procedures which are usual for Slovenian laboratories, i.e. on the basis of chemical composition and digestibility coefficients from tables or by the use of regression equations from literature. It was established that NEL concentrations that were estimated on the basis of DLG (1982) tables deviated from in vitro estimated values for 0.22 ± 0.35 MJ kg^–1 dry matter (DM). Estimates on the basis of recently published tables and equations deviated from in vitro estimated values for 0.25 ± 0.31 MJ (DLG, 1991), 0.20 ± 0.33 MJ (DLG, 1997) and 0.16 ± 0.33 MJ (GfE, 1998) kg^–1 DM. The equation according to Stekar et al. (1994) deviated on average from in vitro values for 0.01 ± 0.34. Regression analysis showed that all estimates deviated systematically from in vitro estimated values. For the estimation of low quality samples from extensively managed grasslands which contain less than 5.2 MJ NEL kg^–1 DM the usage of old German tables (DLG, 1982) can be recommended while for samples which contain more than 5.2 MJ NEL kg^–1 DM recently published tables (DLG, 1997) and official German equations (GfE, 1998) are more suitable. For average samples from Slovenian farms the highest accuracy can be achieved by the use of the equation proposed by Stekar et al. (1994). Relatively high average individual deviations from in vitro estimated values (0.27–0.32 MJ NEL kg^–1 DM) were characteristic for all estimates. Maximal individual deviations exceeded 1 MJ NEL kg^–1 DM. New regression equations for estimation of metabolizable energy concentration and NEL were proposed. It was found out that by the use of new equations deviations from the in vitro estimated values can be reduced by 10 to 50 %. Unexplained deviations of about 0.2 MJ NEL kg^–1 DM cannot be avoided by the means of classical Weende analysis.