Research of quality indices for cold-smoked salmon using a stepwise multiple regression of microbiological counts and physico-chemical parameters.

AIMS The aim of the study was to assess the relationships between the remaining shelf-life (RSL) of cold-smoked salmon and various microbiological and physico-chemical parameters, using a multivariate data analysis in the form of stepwise forward multiple regression. METHODS AND RESULTS Thirteen batches of French cold-smoked salmon were analysed weekly during vacuum-packed storage at 5 degrees C for their lipid, water, salt, phenol, pH, total volatile basic nitrogen (TVBN) and trimethylamine contents, total psychrotrophic count, lactic acid bacteria, lactobacilli, B. thermosphacta, Enterobacteriaceae and yeast counts. At the sensory rejection time, the flora was dominated by lactobacilli, lactobacilli/Enterobacteriaceae or Carnobacteria/B. thermosphacta. Shelf-life was very variable (1->6 weeks) and was related to the initial Enterobacteriaceae load (P < 0.05), depending on hygienic conditions in the smokehouse. High correlations existed between the RSL and lactobacilli count (P < 0.01), yeast count (P < 0.05) and TVBN concentration (P < 0.01). A polynomial fitting the RSL as a function of those three factors was proposed (R(2) = 0.80). Assuming that lactobacilli count could not exceed 109 cfu g-1, a minimum of 36 mg-N 100 g-1 was necessary for a product to be rejected, with a yeast count of 104 cfu g-1. CONCLUSION Estimation of cold-smoked salmon quality is possible by measuring three parameters: lactobacilli and yeast counts and TVBN concentration. SIGNIFICANCE AND IMPACT OF THE STUDY The technical content is important for the smoked salmon industry and for development of quality standards for cold-smoked salmon.