A data-set with 47 treatment means (N = 211) was compiled from research institutions in Denmark, Norway, and Sweden in order to develop a prediction equation for enteric methane (CH4) emissions from dairy cows. The aim was to implement the equation in the Nordic feed evaluation system NorFor. The equation should therefore be based on input variables available in NorFor. The best equation to predict CH4 (MJ/d) was based on dry matter intake (DMI, kg/d), and content of (g/kg DM) fatty acids (FA), crude protein (CP), and neutral detergent fiber (NDF). The equation was CH4 = 1.36 (±0.10) × DMI – 0.125 (±0.039) × FA – 0.02 (±0.012) × CP + 0.017 (±0.005) × NDF (RMSE = 3.00 MJ CH4/d; CV = 13.8%; R2 = 0.77), where RMSE is the root mean square error and CV is the coefficient of variation. However, CP was on the borderline of being significant and did not quantitatively explain much variation in CH4 emission. Based on the present research, we concluded, therefore, that the equation CH4 = 1.23 (±0.08) × DMI – 0.145 (±0.039) × FA + 0.012 (±0.005) × NDF (RMSE = 3.10 MJ CH4/d; CV = 14.3%; R2 = 0.75) is most suited for being implemented in NorFor. However, the ability of the proposed equation to predict enteric methane emissions is uncertain until evaluated on an independent data-set.
Acta Agriculturae Scandinavica. Section A. Animal Science, 2013, Vol 63, Issue 3, p. 126-130