Nir Spectroscopy and Electronic Nose Evaluation on Live Rabbits and on the Meat of Rabbits Fed Increasing Levels of Chia (Salvia Hispanica L.) Seeds

Abstract: Three groups of ten young rabbits each received a diet enriched with Chia (Salvia hispanica L.) seeds at 0% (C0), 10% (C10) and 15% (C15), respectively. At the end of the experimental period, which lasted 35 days, all the rabbits were slaughtered. The Longissimus Dorsi (LD) muscle and perirenal fat samples were collected 24 h post mortem from each carcass and analysed with a GC method for the Fatty Acid (FA) profiles and their indexes. In vivo Spectroscopy, was conducted using a portable UV-Vis. NIR spectrophotometer (Model LSP LabSpec-Pro; 350-2500 nm). The LD muscle specimens (2×2 cm long) were fixed in 95% ethanol, stored for 3 days and finally scanned after the tissues had been exposed for 2 h. LD muscle samples were also analysed raw and cooked using a ten-MOS Electronic Nose (EN) device (AIRSENSE). Discrimination of the individuals between couple of groups, fitted with 1 or 2 dummy values, was performed through a Partial Least Square Discriminant Analysis using WinISI II software. The cross-validated R2 coefficients were retained to compare the matrix distances that were clustered in a Hierarchical Analysis. The in vivo and ethanol LD specimen NIR spectra and EN profiles were calibrated with a set of 81 variables. The average cluster, based on the 81 variables, clearly separated the two treated groups from the control group (av.ge R2 of the matrix = 0.854). The in vivo NIR evaluation was similar to the previous one, but at a low level (R2 = 0.316), while the ethanol muscle specimen highlighted the same pattern, in particular at a higher level (R2 = 0.374). The EN evaluation confirmed the differences for the raw muscle (R2 = 0.443), which were then reduced after cooking (R2 = 0.137). The NIRS applied to the live experimental rabbits showed that this feed experiment produced real differences between the groups. The NIRS applied to the muscle tissues prepared with ethanol showed meat quality traits that were also evaluated by a panel test. In conclusion, EN offer significant knowledge, which normally can only be achieved by a trained panel. The digital spectra can be linked to lipo-oxidation of the intramuscular fat and to a wide set of laboratory analyses, but only for very useful indications and not for purely analytical purposes (0.40<R2<0.70).