HA10 Environmental Toxicology, Environmental Assessment
D135-136
8:00 AM - 12:00 PM, Thursday
() Comparison of three methods for fecal bacterial source tracking.
Farmer, J1, Bailey, F2, Ejiofor, A1, Johnson, T1, 1 Tennessee State University, Nashville, TN, USA2 Middle Tennessee State University, Murfreesboro, TN, USA
ABSTRACT- Bacterial source tracking is a valuable tool for remediation of non-point source fecal bacterial contamination. The purpose of this study was to compare three methods being used for bacterial source tracking; antibiotic resistance patterns (ARP), carbon utilization profiles (CUP), and DNA fingerprinting from pulsed-field gel electrophoresis (PFGE) of Escherichia coli. E. coli were isolated from cows, septic systems, and sewage influent and were "fingerprinted" by the above three methods. The data were then entered into a library for use in discriminant analysis. ARP were determined for each isolate by determining growth on tryptic soy agar plates containing various concentrations of 9 antibiotics. CUP were determined and bacteria were identified using the Biolog® bacterial identification system. PFGE was performed on restriction digests (XbaI restriction enzyme) of bacterial DNA extracts. Environmental bacteria from two sites on the Duck River, central Tennessee, were collected by the filter-membrane method and fingerprinted using the above three methods. Discriminant analysis was used to classify environmental bacteria according to host source. Results from discriminant analysis indicated that classification probabilities were higher for CUP (94%) than ARP and PFGE (71% and 70% respectively). Environmental E. coli from site 1 were classified using CUP as 53% sewage, 34% cow, and 13% septic. For the same bacterial isolates, ARP indicated 57% sewage and 43% cows, while PFGE data identified the bacteria as 46% sewage, 39% cow, and 15% septic. E. coli from site 2 were classified by CUP as 63% sewage and 37% cow. ARP indicated 50% sewage and cows and PFGE data grouped the bacteria as 40% sewage, 38% cow, and 22% septic. All three methods appear to have promise for fecal bacteria source tracking, but the higher classification probability for CUP indicates that it is the most reliable of the three methods.
Key words: antibiotic resistance patterns, bacterial source tracking, carbon utilization patterns, pulsed-field gel electrophoresis