Digestive and metabolic utilisation of dairy cows supplemented with concentrates characterised by different carbohydrates

A respiration chamber experiment was conducted with 12 Brown Swiss dairy cows fed six concentrates with diverse carbohydrate composition over three experimental periods (n = 6). The complete diets consisted of forage and concentrate (1:1). The concentrates were characterised by individual feedstuffs (oat hulls, soybean hulls, apple pulp, Jerusalem artichoke tubers, molasses, wheat), which were either rich in lignified fibre, non-lignified fibre, pectins, fructans, sugars or starch. On basis of analysed proximate contents and tabulated digestibility values, the complete diets were formulated according to Swiss standards to be isoenergetic (net energy for lactation; NEL) and to meet the requirements of the cows for NEL and absorbable protein at the duodenum. In spite of this attempt, the oat hulls diet, being characterised by a high content of lignified fibre (hemicelluloses and cellulose), was significantly inferior in many respects to the other five experimental diets. This included digestibility and energy metabolisability as well as trends toward lower milk yield and energy retention (milk and body stores). Elevated levels of milk urea and proportions of urine-N relative to N-intake suggested that the oat hulls diet provided the least amount of fermentable energy for rumen microbial protein synthesis. The other five diets did not vary that much in their effect on energy turnover, except in urinary energy loss, which was highest with the Jerusalem artichoke diet. Both diets rich in sugars (molasses and Jerusalem artichoke) were superior in energy utilisation compared to the two diets rich in fibre (oat hulls and soybean hulls); however, this did not result in clear differences in energy retention. Contents of NEL, as measured in conjunction with energy balance, clearly differed from the estimates based on tabulated values and those obtained by regressions based on proximate contents, while regressions based on measured digestibilities gave more reliable results (r = 0.68).In conclusion, estimation of the true net energy content of concentrates characterised by less frequently used carbohydrates with different properties is unsatisfactory when only approaches relying on analysed proximate contents and tabulated digestibilities are applied.