Potential used of near infrared reflectance spectroscopy to predict meat physico-chemical composition of guinea fowl (Numida meleagris) reared under different production systems

Near infrared reflectance spectroscopy (NIRS) was evaluated as a tool to predict the physico-chemical composition of samples of Guinea fowl (Numida meleagris) breast and thigh meat. Two different production systems were studied (confinement versus free-range) using 60 animals. The breast and thigh pieces were extracted from the carcass of each animal and analysed according to the official reference methods to determine the content in ash, fat, protein, WHC (water holding capacity), and DM (dry matter). All the samples were scanned to obtain their near infrared reflectance spectrum, using a 19-filter device that reads in the wavelength range of 1445–2348 nm. Multiple linear correlation (MLR) was used as a statistical model to predict the physico-chemical composition. The best prediction equations were obtained for the fat and protein calibrations, with SEc = 0.310 and RC2=0.961 for fat, and SEc = 0.640 and RC2=0.95 for protein. The validation of the equations was also good for fat and protein (SEvc = 0.2179 and 1-variance ratio (VR) = 0.8342, SEvc = 1.9609 and 1-VR = 0.7609, respectively). The worst prediction equations were for the WHC and ash content, with SEc = 1.49, RC2=0.759, SEvc = 4.1711, 1-VR = 0.392, and SEc = 0.030, RC2=0.899, SEvc = 0.3421, 1-VR = 0.4631, respectively.