Assessment of Dairy Product Quality Utilizing Bacterial Enumeration and Metabolite Detection.

Abstract

A study was conducted to investigate the use of bacterial numbers and metabolites as estimators of the potential shelf-life of perishable dairy products, i.e., whole milk, skim milk and cottage cheese. The study was conducted in three phases. In the first phase, a method of impedance detection was performed at two incubation temperatures (18(DEGREES)C and 21(DEGREES)C), and its ability to estimate shelf-life of 100 commercially purchased milk samples compared to standard plate count (SPC), psychrotrophic bacteria count (PBC), modified psychrotrophic bacteria count (mPBC) and Mosely keeping quality test (MSPC). MSPC appeared to possess an adequate relationship to shelf-life (R = -0.839). Impedance detection time (IDT) at 21(DEGREES)C and 18(DEGREES)C proved to have the most significant relationships to shelf-life (R = 0.877 and 0.868, respectively). In the second and third phases of the study, samples of skim and whole milk, and cottage cheese were inoculated with Pseudomonas fluorescens P27 at levels of 0, 10('3) and 10('5) per ml or g. Samples were stored at 7(DEGREES)C. The milks were analyzed every 5 d and cottage cheeses every 7 d until determined to be organoleptically unacceptable. Each sample was analyzed for IDT at 21(DEGREES)C, PBC, proteolysis and endotoxin concentration. Each sample was subjected to preliminary incubation (PI) of none, 21(DEGREES)C 7 h, 21(DEGREES)C 14 h, 13(DEGREES)C 18 h or 18(DEGREES)C 18 h, then enumerated for SPC, mPBC and crystal violet TTC count (CVT). All bacterial enumerations were significantly related to potential shelf-life of pasteurized fluid milks and cottage cheese but were of little predictive value. Proteolysis was also significantly related to potential shelf-life of products studied but proved to be of more value for estimation of cottage cheese quality. Endotoxin concentration and IDT were significantly related to shelf-life and proved valuable for the formulation of predictive regression equations. Impedance detection resulted in the preferred prediction equation suitable for pasteurized fluid milks and cottage cheese.