Use of numerical profiles for studying bacterial diversity

Abstract

A total of 308 bacteria were isolated from oil-storage tanks. Of these 20% were unidentifiable, even at the generic level. A numerical scoring method differentiated between the isolates and was used to estimate the diversity of the bacterial communities in the tanks over a period of 11 months. Although the scoring method suggested a higher diversity than did conventional identification, there was some consistency in the results produced by the two approaches. It is suggested that a scoring method based on only nine tests could be useful for estimating and comparing bacterial diversity in other habitats.