This work was aimed at isolating and identifying the microbiota present during the semi-dry method of coffee processing using polyphasic methods and to evaluate microbial diversity with PCR-DGGE. Samples of Coffea arabica L. were collected during different processing stages in southern Minas Gerais, Brazil. The bacterial and fungal isolates were phenotypically characterised and grouped according to the ARDRA technique, in which the 16-23S and ITS1-5.8S regions of the rDNA were sequenced for species identification. The bacterial counts varied from 3.7 to 7 log CFU g(-1). The yeast counts ranged from 3.4 to 6.9 log CFU g(-1), and the filamentous fungal population varied from 2 to 3.7 log CFU g(-1). Bacillus subtilis, Escherichia coli, Enterobacter agglomerans, Bacillus cereus and Klebsiella pneumoniae were the predominant bacteria detected during the processing of the coffee, and Pichia anomala, Torulaspora delbrueckii and Rhodotorula mucilaginosa were the dominant yeasts. All of the yeast and bacterial species detected by PCR-DGGE were isolated using culture-dependent methods, with the exception of one uncultivable bacterial species. Aspergillus was the most common genus among the filamentous fungal isolates. The use of polyphasic methods allowed a better characterization of the microbiota that is naturally present in semi-dry processed coffee.
Copyright © 2010 Elsevier Ltd. All rights reserved.