Deliverable D2.5 Methodological tools (e.g. based on NIRS determination in faeces) for the rapid evaluation of variation in nutrient digestibility between animals

In growing animals, feed efficiency is often described from the ratio between growth and corresponding feed consumption during a certain period. Feed efficiency is determined by the efficiency at digestive level and by the efficiency at a metabolic level. The efficiency measured at digestive level is difficult to assess in a large number of animals. In pigs, its measurement requires to house animals in a digestibility cage during a prolonged period to measure feed intake and totally collect the faeces that should be analysed for the nutrient contents. Because of this, this method cannot be used in selection farms of breeders. The variability in digestive ability has always been ignored whereas it has been shown to exist when animals are fed a diet with a high fiber content. The objectives of this deliverable are to identify an alternative methodology to characterize the digestive ability in pigs. For this purpose, an experiment was conducted at UMR Pegase, INRA, France with 63 pigs from Large White, Pietrain, and Duroc breeds that were fed alternatively two diets differing in crude fiber content (3.1 and 8.5% of dry matter) during four periods of three weeks each. The diets were supplemented with three molecules that may be used as indigestible marker: silicone oil, the plastic resin Kynar®, and polyethylene glycol. At each period, digestibility was measured using the gold standard (i.e., measurement of feed intake, total collection of faeces, and dedicated lab analyses to calculate the digestibility coefficients of dry matter, organic matter, energy, N, and crude fiber). At the end of each period, a sample of faeces was also collected directly from the rectum of the pigs. All samples were also analysed by Near InfraRed Spectroscopy (NIRS). The digestive ability of pigs was variable during the growing period, especially when the diet contained a high level of fiber and the digestibility of this type of diet increased when body weight increased. The utilization of indigestible markers did not give satisfying results, because of low recovery in the faeces (plastic resin or polyethylene glycol) or because of large variability in the recovery (silicone oil). The NIRS prediction of digestibility coefficients from a sample collected directly in the faeces was adequate (bias of digestibility coefficients for dry matter, organic matter, energy, and N close to 0%) when pigs are heavier than 60 kg and when they are fed a diet with a high fiber content. This type of diet is susceptible to challenge the digestive ability of pigs and allows identifying specifically the poor digesters in a population using NIRS. The method of collecting a single sample of faeces directly from the rectum of the pig and analysed by NIRS can be used in selection farms of breeders to give information on the digestive ability of individual pigs in a population.