A lab-scale model system for cocoa bean fermentation

Abstract

Lab-scale systems modelling the spontaneous cocoa bean fermentation process are useful tools to research the influence of process parameters on the fermentation and the final bean quality. In this study in Honduras, a 1-kg lab-scale fermentation (LS-F) was compared to a 300-kg on-farm fermentation (OF-F) in a multiphasic approach, analysing microbial counts, microbial species diversity, physico-chemical parameters, and final dried bean quality. Yeast and total aerobic counts of up to 8 log CFU/g during the LS-F were comparable to the OF-F, while counts for lactic acid bacteria and acetic acid bacteria were up to 3 log CFU/g lower during the LS-F than during the OF-F. While species of the genera Hansenia, Saccharomyces, and Acetobacter dominated most of the fermentation processes, the genera dominating the drying phases were Pichia, Trichosporon, Pediococcus, and Acetobacter. Dried beans resulting from the LS-F, compared to the OF-F, were similar in contents of acetic acid, 6 times lower in lactic acid, up to 4 times higher in residual sugars, and 3-12 times higher in polyphenols. Dried beans processed at LS showed a similar flavour profile in terms of astringency, bitterness, acidity, and brown, fine, and cocoa flavours, but 2 units higher off-flavours than OF processed beans. With 81%, the share of well-fermented beans from the LS-F complied with industrial standards, whereas 7% over-fermented beans were above the threshold. Conclusively, the 5-day model fermentation and subsequent drying successfully mimicked the on-farm process, providing a high-throughput method to screen microbial strains to be used as starter cultures.

Influence of two anti-fungal Lactobacillus fermentum-Saccharomyces cerevisiae co-cultures on cocoa bean fermentation and final bean quality

“…A possible reason could be that thermotolerant yeast species developed at high temperatures at the end of fermentations during IIIF, e.g. Pichia kudriavzevii and Trichosporon asahii, which we had isolated previously from drying cocoa beans in Honduras [4]. The higher temperatures during IIIF caused a similar effect on LAB and total aerobic mesophilic germs with higher counts on dried beans after IIIF than after IF and IIF.…”

“…The content of pulp was determined as single measurements using the method previously described [4] and expressed as a percentage of the pulp weight compared to the whole bean including the adherent pulp. To determine pulp pH, 20 g of pulp-bean mass was manually mixed with 9 times its own weight of dH 2 O for 1 min and pH was measured in the pulp-water mix using a portable pH-meter (VWR International, Pennsylvania, USA).…”

“…Similarly, the cotyledon pH was determined in a cotyledon-water mix produced by homogenizing 20 g of cotyledons after removing pulp and testa manually with 9 times the weight of dH 2 O for 2 min using a hand blender (Bamix AG, Mettlen, Switzerland). The temperature of the fermentation mass was recorded every 15 min with nine temperature sensors (DS18B20, Maxim Integrated, California, US) that were placed in different spots in the fermentation box as described [4].…”

“…The quality of dried beans resulting from each fermentation was assessed by a cut test with 300 dried cocoa beans and by tasting cocoa liquors as described [4]. The cocoa liquors were prepared and evaluated sensorially by a trained panel of seven at Chocolats Halba (Wallisellen, Switzerland) in August 2017.…”

“…Despite its spontaneous nature, the fermentation of cocoa beans is typically defined by a succession of yeasts, lactic acid bacteria (LAB), and acetic acid bacteria (AAB), as observed for fermentations in different parts of the world [ 4 – 7 ]. The activities of these microorganisms in the pulp result in the production of acids, ethanol, and heat, which diffuse into the bean leading to death of the seed embryo and structural and biochemical changes in the cotyledons [ 3 ].…”