Emerging investigators series: microbial communities in full-scale drinking water distribution systems - a meta-analysis

In this study, we co-analyze all available 16S rRNA gene sequencing studies from bulk drinking water samples in full-scale drinking water distribution systems. Consistent with expectations, we find that Proteobacteria, particularly Alpha-and Betaproteobacteria, dominate drinking water bacterial communities irrespective of origin of study and presence/absence of or disinfectant residual type. Microbial communities in disinfectant residual free systems are more diverse than in those that maintain a disinfectant residual. Further, we find positive associations between mean relative abundance and occurrence of bacteria within a disinfectant category group. The relative abundance and occurrence of key bacterial genera (e.g. Legionella, Mycobacterium, Pseudomonas) is influenced by the presence/absence of a disinfectant residual and the type of disinfectant residual used. Similarly, we find widespread distribution of bacterial genera that are of interest from both an ecological and process perspectives (e.g. nitrification, predation). By estimating the contribution of potential contaminating genera to published drinking water datasets, we recommend that routine sequencing of negative controls be included in drinking water studies. Finally, we test the utility of predicting the metabolic potential of drinking water communities using 16S rRNA gene data and recommend against this practice. Though data heterogeneity across available datasets is a major confounding factor in our meta-analysis, we recommend that efforts to standardize sample processing protocols to address it may not be optimal for the drinking water microbial ecology field at this juncture. Rather, we recommend standardizing data and meta-data reporting, starting with making all sequencing data publicly available, and sample sharing as means of supporting future efforts for comparative analyses across drinking water systems/studies.