ABSTRACT
Lightly preserved seafood products, such as cold-smoked fish and fish gravlax, are traditionally consumed in Europe and are of considerable economic importance. This work aimed to compare three products that were obtained from the same batch of fish: cold-smoked salmon (CSS) stored under vacuum packaging (VP) or a modified atmosphere packaging (MAP) and VP salmon dill gravlax (SG). Classical microbiological analyses and 16S rRNA metabarcoding, biochemical analyses (trimethylamine, total volatile basic nitrogen (TVBN), biogenic amines, pH, volatile organic compounds (VOCs)) and sensory analyses (quantitative descriptive analysis) were performed on each product throughout their storage at a chilled temperature. The three products shared the same initial microbiota, which were mainly dominated by Photobacterium, Lactococcus and Lactobacillus genera. On day 28, the VP CSS ecosystem was mainly composed of Photobacterium and, to a lesser extent, Lactococcus and Lactobacillus genera, while Lactobacillus was dominant in the MAP CSS. The diversity was higher in the SG, which was mainly dominated by Enterobacteriaceae, Photobacterium, Lactobacillus and Lactococcus. Although the sensory spoilage was generally weak, gravlax was the most perishable product (slight increase in amine and acidic off-odors and flavors, fatty appearance, slight discoloration and drop in firmness), followed by the VP CSS, while the MAP CSS did not spoil. Spoilage was associated with an increase in the TVBN, biogenic amines and spoilage associated VOCs, such as decanal, nonanal, hexadecanal, benzaldehyde, benzeneacetaldehyde, ethanol, 3-methyl-1-butanol, 2,3-butanediol, 1-octen-3-ol, 2-butanone and 1-octen-3-one. This study showed that the processing and packaging conditions both had an effect on the microbial composition and the quality of the final product.
ABSTRACT
Cold-smoked salmon is a lightly preserved fish product in which a mixed microbial flora develops during storage and where the interactive behaviour of micro-organisms may contribute to their growth and spoilage activity. The aim of this study was to assess the effect of the bacterial interactions between the main species contaminating the cold-smoked salmon on bacterial growth, chemical and sensory changes, and spoilage. First, Carnobacterium piscicola, Photobacterium phosphoreum, Lactobacillus sakei, Vibrio sp., Brochothrix thermosphacta and Serratia liquefaciens-like were inoculated as pure cultures on sterile cold-smoked salmon. All bacterial species grew well; Vibrio sp. was the fastest and L. sakei strains developed very rapidly as well with a high maximum cell density on cold-smoked salmon blocks (up to 10(9) cfu g(-1) after 10 days at 8 degrees C). Based on sensory analysis, Vibrio sp. was identified as non-spoilage bacteria, C. piscicola as very lightly and B. thermosphacta as lightly spoiling. L. sakei and S. liquefaciens-like were found to be the most spoiling bacteria. Secondly, C. piscicola and L. sakei, two species frequently occurring in the lactic flora of the product, were inoculated together and each of them in mixed cultures with respectively P. phosphoreum, Vibrio sp., B. thermosphacta, and S. liquefaciens-like. The growth of L. sakei was shown to strongly inhibit most of the co-inoculated strains i.e. P. phosphoreum, B. thermosphacta, S. liquefaciens-like and, to a lesser extent, Vibrio sp. The growth of C. piscicola seemed to be enhanced with B. thermosphacta and to develop earlier with P. phosphoreum and Vibrio sp. Conversely, S. liquefaciens-like and P. phosphoreum were weakly inhibited by C. piscicola. The main observation resulting from the sensory evaluation was the delay in the appearance of the spoilage characteristics in the mixed cultures with L. sakei, in particular L. sakei/ S. liquefaciens-like. On the other hand, the spoilage activity of the non-spoiler strains Vibrio sp. or the moderate spoilage strains B. thermosphacta and C. piscicola was increased when they were associated together. It is concluded that the spoilage behaviour of micro-organisms in mixed culture is significantly different from pure culture and explain the difficulty to find robust quality indices for this product.
