The use of near-infrared reflectance spectroscopy (NIRS) in the prediction of chemical composition of freeze-dried egg yolk and discrimination between different n − 3 PUFA feeding sources

The study evaluated near-infrared reflectance spectroscopy (NIRS) to predict the physicochemical composition of freeze-dried egg yolk samples from laying hens fed with four different diets. Beside the control (C), the other three diets were enriched with different sources of n − 3 polyunsaturated fatty acids (PUFA): marine origin (NF), extruded linseed (EL) and ground linseed (GL). Furthermore, NIRS was used to classify the yolks according to hens’ feeding regime. Samples were analyzed chemically and scanned in reflectance mode between 1100 and 2498 nm of the near-infrared region. The calibration results for the 365 yolk samples showed that NIRS can be used to predict their chemical composition, but it also highlights some limitation probably related to problems in the reference method. The prediction of PUFA was accurate while the calibration for the major chemical components (protein, lipids and DM) needs to be improved. Partial least squares (PLS) discriminant analysis was developed to differentiate the yolks, which originated from hens fed with the different diets. All the yolks from hens fed the C and the NF diets, and 98.6% of the two linseed diets, were correctly classified. pH, cholesterol and CIE colour parameters were not successfully predicted; the latter because the visible region was not scanned and this suggests that colour attributes cannot be predicted from NIR spectra alone. It was concluded that NIRS could be used both for estimation of chemical composition in nutritional experiments and as a screening analytical control technique for n − 3 PUFA enriched eggs.