Abstract
A meta-analysis of existing available Illumina 16S rRNA datasets from drinking water source, treatment and DWDS were collated to compare changes in abundance and diversity throughout. Samples from bulk water and biofilm were used to assess principles governing microbial community assembly and the value of amplicon sequencing to water utilities. Individual phyla relationships were explored to identify competitive or synergistic factors governing DWDS microbiomes. The relative importance of stochasticity in the assembly of the DWDS microbiome was considered to identify the significance of source and treatment in determining communities in DWDS. Treatment of water significantly reduces overall species abundance and richness, with chlorination of water providing the most impact to individual taxa relationships. The assembly of microbial communities in the bulk water of the source, primary treatment process and DWDS is governed by more stochastic processes, as is the DWDS biofilm. DWDS biofilm is significantly different to bulk water in terms of local contribution to beta diversity and in types of taxa present. Water immediately post chlorination has a more deterministic microbial assembly, highlighting the significance of this process in changing the microbiome although elevated levels of stochasticity in DWDS samples suggest that this may not be the case at customer taps. 16S rRNA sequencing is becoming more routine and may have several uses for water utilities including detection and risk assessment of emerging pathogens like Legionella, Bacteroides and Mycobacterium; assessing the risk of nitrification of DWDS; improved indicators of process performance and monitoring for significant changes in the microbial community to detect contamination. Combining this with other quantitative methods like flow cytometry will allow a greater depth of understanding of the DWDS microbiome.
Highlights
Meta-analysis of 27 16S rRNA studies from drinking water systems, comprising 1994 samples, identifying 4556 AVS from full length 16S rRNA gene sequences
We have demonstrated that DWDS microbiomes are more strongly affected by stochastic processes.
Chlorine has a stronger selective pressure on the assembly of the microbiome than filtration processes
Pathogens such as Firmicutes form co-exclusionary relationships with other phyla after the addition of chlorine.
Legionella abundance may be a good indicator of treatment performance for water utilities.
Changes in species abundance and richness may be useful in detecting contamination in DWDS for water utilities.
Competing Interest Statement
The authors have declared no competing interest.