The cutting type of vegetables influences the spontaneous fermentation rate

International audience

إنجليزي. Fermented vegetables are mainly produced by the spontaneous fermentation of raw vegetables that are roughly or thinly cut, salted and incubated in an oxygen-free environment. Despite the variety of cutting types and their potential role in the rate of solute diffusion from vegetable tissue, and hence the fermentation rate, the effect of this factor has been little studied. Our aim was to investigate how cutting and small variations in salt concentrations impact the microbial and biochemical changes that occur during the spontaneous fermentation of vegetables. A 2  3 experimental design was set up with vegetable type (carrot/cabbage), cutting type (thin/rough), and salt concentration (0.8%/1%) as the different factors. The vegetables were pressed down in 500 mL-jars and then filled with brine, and two independent jars used at four stages to characterise microbial dynamics and biochemical changes by combining culturomics, 16S rRNA V5-V7 and gyrB metataxonomics, and targeted metabolomics. Culturomic and metataxonomic results revealed similar successions of the main bacterial groups in both vegetables, with Enterobacteriaceae (8 vs 7 log colony-forming units(CFU)/g) quickly replacing the initial microbiota, further replaced within a few days by lactic acid bacteria (9 vs 8 logCFU/g), mainly represented by Leuconostoc sp. The pH fell to 3.8 within 40 h in carrot and about two weeks in cabbage. Mannitol, lactic acid and acetic acid were the main metabolites produced in both vegetables. Viable Enterobacteriaceae were no longer detected after two weeks of fermentation, except in some roughly-cut cabbage samples. No pathogenic bacteria were found. Taxonomic profiles varied depending on the marker used, e.g. Leuconostoc was only detected with gyrB and vice-versa for Clostridium. The gyrB marker enabled markedly better resolution at the species level (for 97% of ASV vs only 20% for the 16S marker). Significant effects of the cutting type, and, to a limited extent, of the NaCl concentration, were observed. Thinly-cut vegetables generally displayed more rapid fermentation compared to roughly-cut vegetables, together with higher titratable acidity, e.g. 0.8% vs 0.3%, respectively, in grated and sliced carrot after 64 h incubation. In line with this, acids were produced more rapidly and levels of viable enterobacteria fell more quickly in thinly-cut vegetables, and particularly cabbage, where the surface area generated by cutting was ~20-fold greater in shredded cabbage than in leaf cabbage. Some leaf cabbage samples displayed atypical fermentations, with particular taxa and atypical metabolite profiles producing high levels of ethanol. These general trends were modulated by quantitative and qualitative differences between replicate jars.This study therefore confirms the highly diverse microbiota of spontaneously fermented vegetables and the tight competition between Enterobacteriaceae and lactic acid bacteria regarding their colonisation. For the first time it documents the effects of cutting