Utilisationde l' azote-15 et de la spectrometrie dans le proche infrarouge pour la determination de la degradabilite reelle in situ et in vitro des matieres azotees des fourrages.

Ruminal degradability of feed protein is a fundamental parameter in most of the new feeding systems, especially the French and Dutch ones. At present, the nylon-bag methods represents the most direct way to estimate this parameter (in sacco or in situ method). Bags containing the test diet are incubated in the rumen for variable periods. The nitrogen effective degradability (ED) is then determined from this kinetic study and from the rate of outflow of protein from the rumen. This technique is subject to criticism owing to its low repeatability. Moreover, the in situ method is laborious, very expensive and requires ruminally fistulated animals. Therefore, the aims of this study were firstly, to improve the nylon-bag technique, and secondly, to develop accurate and inexpensive routine methods in order to estimate the nitrogen ED of fresh and preserved forages. The microbial colonization of feed residues in sacco, which was known to be very important as far as forages are concerned, led to underestimate the nitrogen ED. It has been studied here by the isotope dilution technique using either nitrogen-15 labelled forages or nitrogen-15 labelled micro-organisms. It seems, despite a re-use of the feed nitrogen-15 by particle associated bacteria inside the bags, that using uniformly labelled forages or particles associated bacteria as the reference microbial sample gave similar values. While free bacteria gave lower values for microbial contamination of forage residues in sacco. In order to estimate nitrogen microbial contamination of feed residues in the bags with a rapid method, near infrared spectroscopy (NIRS) was used with 478 reference samples previously incubated in the rumen. The prediction model gave interesting results showing that NIRS had a potential to predict microbial contamination of in sacco forage residues (grasses: r=0.92 and SEcv=9.02; legumes: r=0.85 and SEcv=9.93). Moreover, taking into account microbial nitrogen content of in sacco residues decreased drastically variability between the animals. It is also shown that feeding animals with a mixed diet (hay plus concentrate 2-1 W/W) or with fresh grass produced similar values for nitrogen ED and that freeze-drying represented the most appropriate drying method for forage samples. Besides, an enzymic method was adapted to Fibertec system in order to estimate enzymic nitrogen degradability (ENZ.D) of 51 fresh and preserved forages. The in vitro values were closely related to in situ values (ED in situ = 2.39 + 0.984 ENZ.D in vitro, r=0.96 and syx=1.67).