Résumé
Trait-based approaches are increasingly gaining importance in community ecology, as a way of finding general rules for the mechanisms driving changes in community structure and function under the influence of perturbations. Frameworks for lifehistory strategies have been successfully applied to describe changes in plant and animal communities upon disturbance. To evaluate their applicability to complex bacterial communities, we operated replicated wastewater treatment bioreactors for 35 days and subjected them to eight different disturbance frequencies of a toxic pollutant (3-chloroaniline), starting with a mixed inoculum from a full-scale treatment plant. Relevant ecosystem functions were tracked and microbial communities assessed through metagenomics and 16S rRNA gene sequencing. Combining a series of ordination, statistical and network analysis methods, we associated different life-history strategies with microbial communities across the disturbance range. These strategies were evaluated using tradeoffs in community function and genotypic potential, and changes in bacterial genus composition. We further compared our findings with other ecological studies and adopted a semi-quantitative competitors, stress-tolerants, ruderals (CSR) classification. The framework reduces complex data sets of microbial traits, functions and taxa into ecologically meaningful components to help understand the system response to disturbance and hence represents a promising tool for managing microbial communities.