This study evaluates for the first time the modifications in the microbial communities and sensory attributes caused by Self-Induced Anaerobiosis Fermentation (SIAF) compared to the Conventional processing of Coffea canephora var. Conilon. Microorganisms were identified through high-throughput sequencing of the 16S rRNA V3/V4 region for bacteria and the ITS region for fungi. Sensory attributes of roasted coffee were evaluated by Q-Graders. The relationship between microbial population, processing methods, and sensory attributes was investigated using principal component analysis. Before fermentation, 74 bacterial and 21 fungal species were identified in the natural coffee, whereas 44 bacterial and 15 fungal species were found in the pulped coffee. Torulaspora, Wickerhamomyces, and Meyerozyma exhibited more ITS region sequences, while Acetobacter, Enterobacter, and Lysinibacillus were predominant in the 16S region. In the natural coffee, Wickerhamomyces showed the highest relative abundance (45%) at 0 h. After 72 h, Meyerozyma (45%) and Torulaspora (75%) prevailed in Conventional processing and SIAF, respectively. In the pulped coffee, Torulaspora was the most abundant in the SIAF method, before (92%) and after (81%) fermentation, while Wickerhamomyces (39%) dominated after 72 h in the Conventional method. Enterobacteriaceae levels decreased, while Lactobacillaceae levels increased in SIAF natural coffee during the fermentation process. SIAF favored the presence of yeast and LAB while inhibiting mycotoxigenic fungi and Enterobacteriaceae. Torulaspora, Lactiplantibacillus, and Lactococcus showed the highest Pearson correlation coefficient with flavor (0.92), aftertaste (0.99), and bitterness/sweetness (0.89), respectively. Changes in coffee microbiota during SIAF improved sensory attributes, resulting in better-quality beverages.
Keywords: Torulaspora delbrueckii; Coffee quality; Conilon coffee; High-throughput sequencing; Sensory quality.
© 2025. The Author(s), under exclusive licence to Springer Nature B.V.