Valorization of Fruit By-Products Through Lactic Acid Fermentation for Innovative Beverage Formulation: Microbiological and Physiochemical Effects

Abstract

The increase in food production is accompanied by an increase in waste, particularly agricultural by-products from cultivation and processing. These residues are referred to as agricultural by-products. To address this issue, biotechnological processes can be used to create new applications for these by-products. This study explored the use of LAB strains (Lactiplantibacillus plantarum, Streptococcus thermophilus, Lactobacillus delbrueckii subsp. bulgaricus, and Limosilactobacillus fermentum) on by-products such as white grape pomace, cocoa bean shells, apple pomace, and defatted roasted hazelnut to develop yoghurt-style fruit beverages. Microbial load and pH changes were monitored during a 24 h fermentation and 14-day shelf life at 5 °C. Concentrations of sugars, organic acids, and volatile organic compounds were also analyzed using HPLC and GC-qMS. The results showed that optimizing the matrix led to significant bacterial growth, with viable microbes remaining under refrigeration. In particular, the strain of L. plantarum tested on the cocoa bean shell yielded the most promising results. After 24 h of fermentation, the strain reached a charge of 9.3 Log CFU/mL, acidifying the substrate to 3.9 and producing 19.00 g/100 g of lactic acid. Aromatic compounds were produced in all trials, without off-flavours, and characteristic fermented food flavours developed. Additionally, secondary metabolites produced by lactic acid bacteria may enhance the health benefits of these beverages.

Keywords: bioactive compounds; functional food; gas chromatography; lactic acid bacteria; volatile compounds.

Figure 1

Evolution of the bacteria count (Log CFU/mL; left) and pH (right) of the LAB strains in Moscato grape pomace (MP), cocoa bean shells (CBSs), apple pomace (AP) and de-fatted roasted hazelnut (DH) during 24 h of incubation at 37 °C and 14 days at 5 °C. Bars represent statistical variances in triplicate measurements of duplicate fermentations (Dunn’s test, Bonferroni, p > 0.025).

Figure 2

PCoA on data dispersion produced on the multivariate distance matrix from sugars and organic acids detected by HPLC analysis. For each fruit by-product, the results obtained for each strain are reported. For each matrix, the initial condition (T0) was compared to the fermented beverage after 24 h.

Figure 3

PCoA on data dispersion produced on the multivariate distance matrix from VOCs. For each fruit by-product, the results obtained for each strain are reported. For each matrix, the initial condition (T0) is compared to the fermented beverage after 24 h.

Figure 4

Non-metric multidimensional scaling (NMDS) on data dispersion on VOCs. For each fruit by-product, the results obtained for each strain are reported. The main volatile compounds produced during each fermentation are highlighted.