Climate change impacts, including rising temperatures and altered precipitation patterns, threaten drinking water quality. Drinking water distribution systems (DWDS) are particularly vulnerable to temperature fluctuations, impacting microbial growth and community composition. Higher temperatures can stimulate growth of certain species, including potential pathogens like *Legionella*. Changes in raw water sources also significantly affect the drinking water microbiome, especially initially after a switch. Climate change is predicted to increase water temperature and nutrient concentrations in both raw water sources and DWDS, potentially leading to issues with taste, color, and odor. Microbial growth in DWDS occurs in four phases: bulk water, suspended solids, pipe wall biofilm, and loose deposits. Biofilms provide a protected environment for bacteria, influenced by factors like nutrient availability and surface characteristics. The interaction between biofilm and bulk water communities is dynamic, with biofilm formation initially from bulk water colonizers, later influenced by a seed bank from bulk water taxa. However, biofilms, while representing the majority of biomass, are a crucial factor in drinking water quality deviations. Previous pilot studies have yielded conflicting results on the long-term effects of temperature increases on biofilm communities in chlorinated and non-chlorinated systems. This study aims to determine how climate change-induced variability in water source quality and increasing temperatures impact water biostability during distribution using a pilot-scale DWDS.

Several studies have examined the effects of temperature and water source changes on drinking water microbiomes. Some have shown increased heterotrophic plate counts (HPC) and bacterial cell concentrations at higher temperatures, while others found inconsistent correlations between temperature and bacterial growth, suggesting that other factors such as water quality (disinfectant residual, nutrients) and hydraulics also play a role. Variations in source water quality, particularly differences in organic carbon content, have been shown to significantly influence microbial communities. The impact of switching water sources on biofilm communities has been noted in some studies, but with varying timeframes for community restoration to a stable state. Existing research utilizing drinking water distribution pilots to simulate full-scale networks has shown mixed results regarding the effects of temperature on biofilm and bulk microbial communities, partly due to differences in experimental duration and chlorination status.

A pilot-scale DWDS with three identical loops (100 m each) operating at 16 °C, 20 °C, and 24 °C was used. The system was fed with alternating treated groundwater and surface water. Bulk water quality (cell densities, community composition, physicochemical parameters) was monitored for 137 days using online flow cytometry (FC) and 16S rRNA gene-based amplicon sequencing. Biofilm samples were collected using a coupon system at various time points. Flow cytometry was used to assess bulk and biofilm cell densities and phenotypic traits. 16S rRNA gene sequencing was employed for genotypic community analysis. Physicochemical parameters (conductivity, pH, pressure, flow, nitrogen forms, orthophosphate, iron, NPOC) were also measured. Growth curves were fitted to FC data to calculate growth rates and carrying capacities. Statistical analysis (PERMANOVA, Kruskal–Wallis test, Mann-Whitney test) was performed to determine significant differences.

Temperature significantly affected bulk bacterial cell densities, with higher temperatures leading to higher densities. Water source significantly influenced bulk cell densities and community composition. Growth rates and carrying capacities were higher at elevated temperatures with treated groundwater. Biofilm cell densities were not significantly affected by temperature. A mature biofilm developed from day 70 onwards, exhibiting a stable core microbiome resilient to temperature and water source changes. Analysis of cytometric data revealed significant differences due to both temperature and water source on microbial phenotypic traits. The bulk water community was primarily composed of Proteobacteria (Comamonadaceae, Sphingomonadaceae), with the source water significantly influencing composition. A core biofilm microbiome was identified, including Rhodocyclaceae, Xanthobacteraceae, Sphingomonadaceae, and Comamonadaceae, and was not affected by temperature increases or bulk water changes. Non-metric multidimensional scaling (NMDS) analysis showed higher similarity across mature biofilm samples compared to bulk water samples.

The findings demonstrate the importance of both temperature and water source quality in shaping the drinking water microbiome. Elevated temperatures in combination with treated groundwater promoted increased bacterial growth and carrying capacity in the bulk water, but did not influence biofilm development or community composition. Water source variations influenced both bulk and biofilm cell densities, particularly at higher temperatures, but the impact on biofilm community composition was limited, especially after the establishment of a mature biofilm. The stable core biofilm microbiome suggests a high resilience to environmental changes. The significant effect of the water source on the bulk community composition emphasizes the primary influence of the initial water quality on the overall microbial dynamics in the DWDS.

Increasing temperatures primarily impacted bulk water microbial densities, not biofilm development. Treated groundwater at elevated temperatures resulted in increased bulk water growth rates and carrying capacities. Water source significantly affected both bulk and biofilm communities, especially at higher temperatures, however the effect on biofilm composition was limited in mature biofilms. A core biofilm microbiome was identified and was resistant to changes. Future research could focus on extending the study duration to further investigate long-term effects and expanding the range of water sources and temperatures.

The pilot-scale system, while useful for controlled studies, may not perfectly reflect the complexity of real-world DWDS. The relatively short residence time (7 days) in the pilot may not fully represent the conditions in some full-scale systems. The alternation of water sources in the pilot system created a complex dynamic, and further research isolating the effects of individual factors is needed. The specific microbial analysis (16S rRNA) may have missed some aspects of community functionality.