RESUMO
Many water quality monitoring programs quantify Escherichia coli, an indicator of fecal contamination and potential sewage pollution. However, interpretation of E. coli data can be complex due to abiotic factors that influence its growth and mortality. The goal of this study was to quantify the variability of E. coli in a river and assess the impact of that variability on water quality monitoring study design and sewage pollution source identification. Over 1900 samples were collected and analyzed from 2007 to 2017 in the Norwalk River in Connecticut. Sixty-six percent of the samples collected during weekly to monthly monitoring had E. coli concentrations below 200 CFU/100 mL, indicating that elevated bacteria concentrations were captured infrequently. Patterns observed during daily sampling indicated that the randomization of sampling days within a week may support the identification of pollution sources driven by human behavioral patterns. Spatial autocorrelation in bacteria concentrations was not observed between sites, indicating that the sample locations were not spaced sufficiently close together to be redundant for monitoring. On finer spatial scales however, detection of a known pollution source was found to be challenging at even short distances downstream, with less than 25% of the original source concentration detected at 10 m downstream and less than 10% by 1000 m downstream, suggesting that a high density of study sites may be needed to detect potential sources. These findings can be used to better understand the natural variability of this important indicator organism in freshwater systems, and inform more efficient and effective monitoring.