Identification and Quantification of Lactic Acid Bacteria in a Water-Based Matrix with Near-Infrared Spectroscopy and Multivariate Regression Modeling

Abstract

Food industry is increasingly concerned in developing and applying rapid and nondestructive methods to offer safer and high quality foods to consumers. During the last years, Fourier transform near-infrared (FT-NIR) has been widely used to determine food quality based on spectrum. Likewise, FT-NIR has been proposed as an innovative and promising nondestructive rapid method capable to detect and identify microorganisms in foods; however, little progress has been made to date in this field. This study is a new attempt to apply FT-NIR technology to identify and quantify bacteria species in water-based systems in order to simulate water-based food matrices. For that, three different lactic acid bacteria—Lactobacillus plantarum, Leuconostoc mesenteroides, and Lactobacillus sakei—associated with spoilage in ready-to-eat meat, were analyzed by reflectance–transmitance FT-NIR in the spectral range 1,100–2,500 nm. Principal component analysis (PCA), and partial least squares (PLS) were applied to obtain prediction models. PCA and PLS showed a clear discrimination between the tested bacteria species whereas PLS method could succesfully quantify the concentration levels (3–9 log cfu/mL) and also distinguish between spoilage (7–9 log cfu/mL) and nonspoilage concentration levels (3–6 log cfu/mL). Results suggest that FT-NIR could be used efficiently to detect and quantify microorgasnisms in water-based food matrices. However, this study is an initial approach and therefore, it will be necessary to further research in order to really carry out its application to more complex food matrices and other microorganisms (i.e., food-borne pathogens).