FINE GRAINED LONGITUDINAL ANALYSIS OF COCOA BEAN FERMENTATION PROVIDES INSIGHTS INTO THE DYNAMICS OF MICROBIAL POPULATIONS

Abstract

Cocoa bean fermentation is an important microbial driven process where metabolites that affect chocolate quality and aroma are produced. Although considerable research has been devoted to the yeast and bacteria species involved in this process, less attention has been paid to the role of populations and strains, which hinders its selection, monitoring and use. Here we present a study that evaluates the microbial diversity associated to the tools and bean mass during spontaneous cocoa fermentation and in two distinct agro-ecological zones in Colombia. Using high-throughput sequencing of molecular markers for bacteria and yeast, we established the dynamics at the species-level (OTUs) and strains-level (oligotypes). Our results show that cocoa bean fermentation is catalyzed by a composite of strains within each OTU and not by one single strain. Eventhough we found only a few bacterial OTUs, one Enterobacter, three of Lactic Acid Bacteria and two of Acetic Acid Bacteria, these could be further split into 6, 23 and 19 oligotypes, respectively. Only two fungal OTUs were found. Comparison between fermentations suggest that local protocols generated a specific footprint in the dynamics of the microbial communities and that tools are reservoirs of some of those groups. The population analysis shows that the oligotypes that become most dominant are the same in the two locations, coupling co-abundance and dominance analysis we suggest that a combination of Enterobacter and Acetobacter oligotypes seem more optimal for the starter cultures. In conclusion, the results presented here show that exploring the fine level dynamics of microbial fermentation is necessary to understand the patterns of the dominance of specific populations and can be used as a valuable approach to select and monitor specific bacteria for the design of starter cultures in the food industry.\n\nIMPORTANCEIn Colombia, the lack of tools to validate and standardize fermentation protocols are one of the principal reasons why Colombian cocoa beans are not recognized as \"fine-flavor\" and widely marketed. Because of the large influence of the microbial fermentation in cocoa quality, the present study explores the microbial dynamics using high-throughput sequencing of molecular markers in two of the most important producer regions in Colombia. The results show that the identification of dominant and transitive strains can be used to select, design and monitor starter cultures and/or the effect of adjustments in the fermentation protocols.