Identifying Sources of Fecal Pollution in Water as Function of Sampling Frequency Under Low and High Stream Flow Conditions

TR Number



Journal Title

Journal ISSN

Volume Title


Virginia Tech


Sources of fecal pollution were evaluated as a function of sampling frequency with stream samples from Mill Creek, Montgomery County, VA. Samples were collected monthly for one year, plus weekly for four consecutive weeks during seasonal high flows (March), and seasonal low flows (September-October), plus daily for seven consecutive days within the weekly schedules. Thirty stream samples were collected from each of two sites (60 total) in Mill Creek, and 48 isolates of E. coli per sample (total of 2,880 stream isolates) were classified by source using antibiotic resistance analysis (ARA) and comparing the resulting patterns against a known-source E. coli library (1,158 isolates). The same process was performed with enterococci isolates against an enterococci library (1,182 isolates). The average rate of correct classification (ARCC) for the E. coli library with a three-way split (human, livestock, and wildlife) was 89.0%, and the ARCC of the species-specific E. coli library (cattle, deer, goose, human, misc. wildlife) was 88.9%. The ARCC of the enterococci library with a three-way split was 85.3%, and the ARCC of the species-specific enterococci library was 88.1%. The results did not justify the need for daily or weekly sampling, but indicated that monthly was adequate (quarterly and every-other-month were not). There was a seasonal effect as the human signature was highest during high flow while the livestock signature dominated during low flow. The results also indicated that sampling should be done over a period of time that includes both seasonal wettest and driest periods (at least 8 months).



bacterial source tracking, watersheds, enterococci, fecal coliforms, E.coli