Prediction of metabolisable energy content of commercial total mixed rations (TMR) for lactating dairy cows based on gas production measured into two TMR fractions

The aim was to estimate the ME content of TMR samples for lactating dairy cows based on 24 h gas production (GP) measured in two TMR fractions obtained by sieving that were considered proxy representations of their forage and concentrate contents. Samples of TMR for lactating cows, collected at 15 Italian dairy farms, were dried and fractionated using an apparatus composed of four sieves (mesh diam. of 6.00, 2.36, 1.18, 0.60 mm) and a bottom pan. The particles retained by the 6.00, 2.36 and 1.18 mm sieves were bulked to create a fraction named “Long” (L), while particles retained by the 0.60 mm sieve and the bottom pan created a fraction named “Fine”(F). Samples of whole TMR and of F and L fractions were fermented in rumen fluid (220 mg of dried sample into graduated 100 ml glass syringes with 30 ml of diluted rumen fluid) to measure GP at 0, 1, 2, 4, 6, 8, 12, 24, 48 and 72 h of incubation, and GP at 24 h was used in multiple equations to predict the ME content. The L fraction was the main part of the TMR (633 ± 61.7 g/kg DM) and contained more aNDFom (391 vs 185 g/kg DM; P < 0.01), less starch (213 vs 394 g/kg DM; P < 0.01) and less CP (132 vs 181 g/kg DM; P < 0.01) than the F fraction. The 24 h GP measured on the F fraction was 17% higher (P < 0.01) than in the L fraction lending to a 25% increase in the predicted ME content of the F fraction (12.1 vs 9.7 MJ/kg DM, P < 0.01). When the GP of the two fractions were summed on the basis of their relative weights in each TMR to obtain a calculated GP of the whole TMR, this did not differ from that directly measured on the original TMR sample. The ME values of the TMR had substantial variability among samples and the highest ME values measured for the F and L fractions were 13.23 and 10.82 MJ/kg DM, respectively. The inclusion of these F and L fraction values in TMR samples that had the lowest ME concentrations (e.g. 10.0–10.4 MJ/kg DM), as an hypothetical consequence of an improvement in nutritive quality of the dietary feed components, resulted in a calculated ME increment of about 0.5–0.6 and 0.8–1.0 MJ/kg DM, respectively. These simulations suggest appreciable ME variations in cases of substituting nutritionally extreme forages and/or concentrates with corresponding feeds of different nutritive quality. © 2017 Elsevier B.V.