Helena Soares Vasconcelos

Abstract

Abstract
The aim of this work was to investigate the potential of Fourier transform infrared spectroscopy (FTIR) to detect and predict the bacterial load of salmon fillets (Salmo salar) stored at 3, 8 and 30 degrees C under three packaging conditions: air packaging (AP) and two modified atmospheres constituted by a mixture of 50%N-2/40%CO2/10%O-2 with lemon juice (MAPL) and without lemon juice (MAP). Fresh salmon samples were periodically examined for total viable counts (TVC), specific spoilage organisms (SSO) counts, pH, FTIR and sensory assessment of freshness. Principal components analysis (PCA) allowed identification of the wavenumbers potentially correlated with the spoilage process. Linear discriminant analysis (LDA) of infrared spectral data was performed to support sensory data and to accurately identify samples freshness. The effect of the packaging atmospheres was assessed by microbial enumeration and LDA was used to determine sample packaging from the measured infrared spectra. It was verified that modified atmospheres can decrease significantly the bacterial load of fresh salmon. Lemon juice combined with MAP showed a more pronounced delay in the growth of Brochothrix thermosphacta, Photobacterium phosphoreum, psychrotrophs and H2S producers. Partial least squares regression (PLS-R) allowed estimates of TVC and psychrotrophs, lactic acid bacteria, molds and yeasts, Brochothrix thermosphacta, Enterobacteriaceae, Pseudomonas spp. and H2S producer counts from the infrared spectral data: For TVC, the root mean square error of prediction (RMSEP) value was 0.78 log cfu g(-1) for an external set of samples. According to the results, FTIR can be used as a reliable, accurate and fast method for real time freshness evaluation of salmon fillets stored under different temperatures and packaging atmospheres.

Abstract
The aim of this work was to evaluate the potential of Fourier transform infrared (FTIR) spectroscopy as a rapid and accurate technique to detect and predict the onset of spoilage in fresh chicken breast fillets stored at 3, 8, and 30 A degrees C. Chicken breasts were excised from carcasses at 6 h post-mortem; cut in fillets; packed in air; stored at 3, 8, and 30 A(0)C; and periodically examined for FTIR, pH, microbiological analysis, and sensory assessment of freshness. Partial least squares regression allowed estimations of total viable counts (TVC), lactic acid bacteria (LAB), Pseudomonas spp., Brochothrix thermosphacta, Enterobacteriaceae counts and pH, based on FTIR spectral data. Analysis of an external set of samples allowed the evaluation of the predictability of the method. The correlation coefficients (R-2) for prediction were 0.798, 0.832, 0.789, 0.810, 0.857, and 0.880, and the room mean square error of prediction were 0.789, 0.658, 0.715, 0.701, 0.756 log cfu g(-1) and 0.479 for TVC, LAB, Pseudomonas spp., B. thermosphacta, Enterobacteriaceae, and pH, respectively. The spectroscopic variables that can be linked and used by the models to predict the spoilage/freshness of the samples, pH, and microbial counts were the absorbency values of 375 wave numbers from 1,700 to 950 cm(-1). A principal component analysis led to the conclusion that the wave numbers that ranges from 1,408 to 1,370 cm(-1) and from 1,320 to 1,305 cm(-1) are strongly connected to changes during spoilage. These wave numbers are linked to amides and amines and may be considered potential wave numbers associated with the biochemical changes during spoilage. Discriminant analysis of spectral data was successfully applied to support sensory data and to accurately bound samples freshness. According to the results presented, it is possible to conclude that FTIR spectroscopy can be used as a reliable, accurate, and fast method for real time freshness evaluation of chicken breast fillets during storage.