Ecological Interpretation of Molecular Microbial Community Composition Data: Opportunities and Pitfalls.

Patterns in the composition of microbial communities are detected through molecular methods that take advantage of gene sequence variation across microbial taxa.  Insights into mechanisms behind these patterns are often gained from application of statistical methods borrowed from other branches of ecology.  Using this approach, many factors that affect soil microbial community composition have been detected, including land use strategies, plant communities, and various abiotic conditions.  However, moving beyond the confirmatory nature of this work requires both enhancing the ecological basis of data interpretation and careful recognition of the limitations of molecular data.  First, the peculiarities of various molecular methods need to be recognized so that data are cautiously interpreted.  For example, although application of ecological diversity indices to molecular fingerprinting data (such as T-RFLP or DGGE data) seems straightforward, the limited sensitivity of these methods in detecting moderately rare taxa results in a lack of correlation between fingerprint diversity and the true diversity of the microbial community in a typical soil.  Second, incorporating additional ecological theory will help us understand the origin of variation not explained by traditional multivariate ordination analysis.  For example, new statistical methods are required that would simultaneously assess the effects of abiotic constraints and competitive interactions.  Finally, it is also necessary to assemble the various factors affecting soil microbial communities into a scale-dependent hierarchy, to avoid falling into a “stamp-collecting” approach of cataloging a bewildering array of isolated mechanisms without understanding how they interact to control microbial communities in natural systems.