Contribution of storage and structural polysaccharides to the fermentationprocess and nutritive value of lucerne ensiled alone or mixed with cereal grains

The low content of water soluble carbohydrates (WSC) and the high buffering capacity usually constrains lucerne (Medicago sativa) ensilability. Howeve r, there are other sources of readily fermentable substrate and positive ef fects of adding cereal grains to wet silages have been reported. An experim ent was carried out to study the contribution of hemicellulose, cellulose a nd natural or added starch to the silage fermentation process, and in vitro digestibility and gas production. A crop of lucerne (10% blooming) was cho pped, inoculated with lactic acid bacteria, and ensiled in microsilos with four treatments: alone (T0), or with 50 kg/t of sucrose (SU), barley (BRL), or maize (MZ). Microsilos were incubated for 47 days. Polysaccharide recov ery was determined by the mass balance technique, and starch by dacron bag. Except for SU, all treatments were apparently limited by the amount of WSC needed to sustain good silage fermentation. As expected, SU produced an ex cellent fermentation but despite initial WSC differences, there were no sig nificant differences with BRL for total fatty acids, lactic and acetic acid , total acidity, pH and NH3-N. Fermentation characteristics of MZ were gene rally worse than BRL's and similar to T0's ones. Hemicellulose was extensiv ely hydrolysed (T0: 59.5%; MZ: 38.1%; BRL: 48.8%; and SU: 46.8%). Cellulose was almost quantitatively recovered and starch recovery averaged 70% for T 0 and BRL, and approximately 100% for MZ and SU. The type of cereal added a ffected starch recovery using the dacron ba technique (MZ: 77%; BRL: 47%; P < 0.05). True dry matter digestibility was the same for all treatments (fr esh lucerne: 81.3%; T0: 80.7%; MZ: 81.8%; BRL: 80.5%; and SU: 82.1%), and n eutral detergent fibre digestibility (NDF-D) of silages was reduced in rela tion to the fresh material (fresh lucerne: 55.1%; T0: 48.8%; MZ: 51.9%; BRL : 45.3%; and SU: 46.7%). Although all silages showed a rapid in vitro ferme ntation pattern, gas production (48 h) of SU and BRL (265.4 and 260.5 ml/g DM) was greater than T0 and MZ (246.0 and 244.6 ml/g DM) (P < 0.05). It can be concluded that starch hydrolysis depends on the type of grain, and that starch from barley and legume cell wall components could supply additional fermentable substrate to the ensiling process. These data also showed that ensiling reduced NDF-D and suggest that labile NDF fractions removed durin g the ensiling process were later accessible during the in vitro fermentati on. (C) 2001 Elsevier Science B.V. All rights reserved.