BOARD-INVITED REVIEW: Efficiency of converting digestible energy to metabolizable energy and reevaluation of the California Net Energy System maintenance requirements and equations for predicting dietary net energy values for beef cattle.

For the past several decades, nutrient requirement systems for beef cattle in North America have recommended that dietary ME can be calculated as dietary DE × 0.82, but considerable published data suggest a variable relationship between DE and ME. We reviewed the literature and tabulated the results of 23 respiration calorimetry studies (87 treatment mean data points), in which measurements of fecal, urinary, and gaseous energy were determined with beef cattle (bulls, steers, and heifers) and growing dairy cattle. Mixed-model regression analyses to adjust for the effects of the citation from which the data were obtained suggested a strong linear relationship between ME and DE (Mcal/kg of DM; ME = 0.9611 × DE - 0.2999; = 0.986, root mean square error [RMSE] = 0.048, < 0.001 for intercept, slope ≠ 0). Analysis of residuals from this simple linear regression equation indicated high correlations of residuals with other dietary components, and a slight increase in precision was obtained when dietary CP, ether extract, and starch (% of DM) concentrations were included in a multiple linear regression equation (citation-adjusted = 0.992, RMSE = 0.039). Using the simple linear relationship, we reevaluated the original data used to develop the California Net Energy System (CNES) for beef cattle by recalculating ME intake and heat production and regressing the logarithm of heat production on ME intake (both per BW, kg daily). The resulting intercept and slope of the recalculated data did not differ ( ≥ 0.34) from those reported for the original analyses of the CNES data, suggesting that use of the linear equation for calculating ME concentration was consistent with NEm and NEg values as derived in the CNES. Nonetheless, because the cubic equations recommended by the NRC to calculate dietary NEm and NEg from ME were based on conversion of DE to ME using 0.82, these equations were mathematically recalculated to account for the linear relationship between DE and ME. Overall, our review and analyses suggested that there is a strong linear relationship between DE and ME, which seems to be consistent across a wide range of dietary conditions, cattle types, and levels of intake. Applying this linear relationship to predict ME concentrations agreed with the original CNES calculations for NE requirements, thereby allowing the development of new equations for predicting dietary NEm and NEg values from ME.