| This research focused on two microbial source tracking (MST) techniques: (1) typing F+RNA coliphages into one of four genetically distinct groups to distinguish human from animal fecal contamination and (2) antibiotic resistance analysis (ARA) of Escherichia coli to identify waters impacted by anthropogenic sources of pollution. Coliphages were isolated from impaired surface waters and proximal known fecal sources by single agar layer (SAL) and enrichment presence/absence (EP/A) methods. Confirmed F +RNA coliphages were genetically or serologically typed for MST. The EP/A technique detected coliphages infecting E. coli Famp in 38 (66%) of 58 surface water samples negative for F+ coliphages by the SAL method. Coliphages isolated by EP/A were found to be less representative of coliphage diversity within a sample. The majority of the 534 F+ RNA coliphage isolates from surface waters typed as group I. Group II and/or III viruses were identified from 14 surface water samples, most of which were downstream of wastewater treatment plant discharges, and therefore likely impacted by human-source fecal contamination. Therefore, F+ RNA coliphage typing appears to be useful for distinguishing human from animal fecal contamination.; F+RNA coliphage isolates from a South Carolina hog lagoon were found upon additional testing to belong to subgroup III. These isolates were found by sequencing analysis to be more closely related to MX1 than Q-beta. However, analysis of additional group III hog lagoon isolates from North Carolina found them to be more closely related to Q-beta. These results demonstrate that exceptions exist to the general associations between coliphage groups and their sources, and that F+RNA coliphage typing by serology or hybridization cannot be used for absolute distinction of pollution sources.; ARA of 631 E. coli isolates showed that 40 (6.3%) displayed resistance to at least one antibiotic and 35 (5.5%) were resistant to multiple antibiotics. A multiple regression model estimated that percent resistance can be predicted from a combination of landcover variables, including percent impervious surface and percent landscaped vegetation (adjusted r2 = 0.970). These results demonstrate that detection of antibiotic resistant bacteria can be useful for the identification of anthropogenically impacted waters in developing areas. |
