Quantifying Potential Sources of Microbial Contamination in Household Drinking Water Samples
Allevi, Richard Paul
MetadataShow full item record
In Virginia, over one million households rely on private water supplies (e.g. well, spring, cistern). Previous literature acknowledges bacterial contamination in private water supplies as a significant public health concern in the United States. The present study tested private wells and springs in 20 Virginia counties for total coliforms (TC) and E. coli (EC) along with a suite of chemical contaminants. Sample collection was organized by the Virginia Household Water Quality Program (VAHWQP), a Virginia Cooperative Extension effort managed by faculty in the Biological Systems Engineering Department. Microbial and chemical source tracking were used to identify possible sources of contamination. A logistic regression was employed to investigate potential correlations between TC contamination and chemical parameters (e.g. NO3-, turbidity) as well as homeowner provided survey data describing system characteristics and perceived water quality. TC and EC contamination were quantified via the Colilert (www.idexx.com) defined substrate method for most probable number (MPN) of EC and TC per 100 mL of water. Of the 538 samples collected, 41% (n=221) were positive for TC and 10% (n=53) for EC. Chemical parameters were not statistically predictive of microbial contamination. Well depth, water treatment, and farm location proximate to the water supply were factors in a regression model that predicted presence/absence of TC with 74% accuracy. Microbial and chemical source tracking techniques (Polymerase Chain Reaction (PCR) and fluorometry, respectively) identified 4 of 26 samples as likely contaminated with human wastewater. Application of these source-tracking analyses on a larger scale will prove useful in defining remediation strategies.
- Masters Theses