Seasonal total coliform dynamics in a drinking water reservoir

Maintaining high-quality drinking water supply reservoirs is important for protecting public health. Despite extensive watershed protection efforts, reservoirs can still experience seasonal, elevated total coliform bacteria concentrations, indicator bacteria commonly used for regulations. This study aimed to understand associations between concentrations of total coliform and an array of water quality, soil, and meteorological parameters over 10 years to identify potential causes and correlations of elevated total coliform bacteria concentrations in a protected watershed and clear, oligotrophic waters. Leveraging long-term data, we performed extensive data analysis and a data-driven model to investigate these relationships in the Quabbin Reservoir (Massachusetts, USA). Data analysis and data-driven modeling results indicated that proxies of algae, organic matter, and dry conditions, as well as water temperature and dissolved oxygen, were most associated with increased total coliforms in the reservoir. Although indicator bacteria such as total coliform are frequently used for routine monitoring, our findings highlight that it was unlikely that their proliferation is indicating a likely elevated risk in the reservoir. The studied reservoir has pristine water quality with low variability and low fecal bacteria indicator levels with no sign of external contamination; therefore, the high concentrations of total coliform bacteria in the summer is likely an autochthonous process. Additionally, applying machine learning methods to leverage long-term routine data collected by monitoring agencies highlights opportunities to better understand how to maintain high-quality surface water in drinking water supply reservoirs through a rapidly changing climate.