Terminal-restriction fragment length polymorphism (T-RFLP) analysis is widely used in microbial ecology studies. In the present study, T-RFLP analysis of PCR products digested by five restriction enzymes (AluI, HaeIII, MspI, Sau3AI and TaqI) was applied for 20 samples from three contrasting coastal environments to assess the biases associated with the choice of enzyme digestion and T-RF analysis. The five enzyme digestions produced highly variable species richness (in terms of number of T-RFs). Analysis of peak areas with a threshold of 0.5% of the total peak area, which recovered 92-96% of the total peak area, revealed different diversity indexes from the five enzyme digestions. Multidimensional scaling, based on matrices that were generated by scoring peak presence/absence and area, revealed similar bacterial community structure patterns among the 20 samples, regardless of the choice of restriction enzymes. Our results strongly argue that the choice of different digestion enzymes in the T-RFLP technique generated valid and consistent bacterial community structures but highly variable species richness and diversity indices. The biases associated with the choice of digestion enzymes needs to be evaluated carefully or at least to be addressed when using T-RFLP analysis. © 2008 Federation of European Microbiological Societies. Published by Blackwell Publishing Ltd. All rights reserved.

link_to_OA_fulltext