Graduation Year


Document Type




Degree Granting Department


Major Professor

Valerie J. Harwood, Ph.D.

Committee Member

Earl McCoy, Ph.D.

Committee Member

John Lisle, Ph.D.


Fecal coliforms, Indicator Organisms, Ribotyping, Antibiotic Resistance Analysis, Population Biology


Bacterial source tracking (BST) methods often involve the use of phenotypic or genotypic fingerprinting techniques to compare indicator bacteria such as Escherichia coli isolated from unknown sources against a library of fingerprints from indicator bacteria found in the feces of various known source animals. The predictive capability of a library is based in part on how well the library isolates reflect the true population diversity of indicator bacteria that can potentially impact a water body. The purpose of this study was to compare the behavior of E. coli population structures in the feces of humans, beef cattle and horses across different parameters. Ribotyping and antibiotic resistance analysis were used to "fingerprint", or subtype E. coli isolates. Significantly greater diversity was observed in the E. coli population of horses compared to the human or beef cattle sampled. Subtype sharing between individuals from all host categories was infrequent, therefore the majority of E. coli subtypes were sampled from a single individual. The dominant E. coli populations of nine individuals (three per host source category) were monitored over time, which demonstrated that E. coli subtypes within a host individual vary on a monthly time frame, and an increase in the frequency of subtype sharing was noted between individuals within the same source group over time. The E. coli population of a single human that had just finished antibiotic treatment was studied on a daily basis for one month. The loss of an E. coli subtype with high antibiotic resistance was observed over time, however there was a single dominant E. coli subtype that was present at every sampling event during the entire month. Geographic distinctiveness of E. coli populations was investigated by sampling four herds located in different geographical regions. We observed that E. coli populations are not geographically distinct, but are somewhat individual-specific, as most E. coli isolates had a subtype that was found in a single individual. This study defines factors that should be considered when constructing a successful BST library, and suggests that E. coli may not be the appropriate indicator organism for BST.