Identification of sources of Escherichia coli in South Carolina estuaries using antibiotic resistance analysis

Fecal pollution from nonhuman (pets, livestock or wildlife) and human sources is often one of the major factors associated with urbanization that contribute to the degradation of water quality. Methods to differentiate animal from human sources of fecal coliform contamination could assist resource managers in developing strategies to protect shellfish harvesting areas and recreational waters. In this study, surface water samples were collected from both a developed and an undeveloped watershed in coastal South Carolina. Influent and effluent samples from several wastewater treatment plants (WWTPs) in the same area were also collected. Most Probable Numbers (MPNs) of fecal coliforms were determined for all samples. Escherichia coli isolates were analyzed for antibiotic resistance (AR) to 10 antibiotics. Then, AR indices (no. of resistant/total no. of antibiotics tested), were calculated for each isolate and site. Results indicated that MPNs from the WWTP samples were significantly higher than those from the developed watershed which were significantly higher than those from the undeveloped watershed (p<0.0001). The AR analyses suggested that there was a trend toward increased antibiotic resistance in samples for the urbanized Broad Creek (BC) watershed. In the Okatee River (OR), E. coli isolates from three sites (20%) showed resistance to a single antibiotic (penicillin) but in BC, isolates from seven sites (47%) were resistant to multiple antibiotics, and the predominant resistance pattern was chlortetracyclineoxytetracycline-tetracycline. Raw sewage isolates from most WWTPs contained E. coli that exhibited resistance to multiple antibiotics. Cluster analysis indicated that all resistant OR sites had antibiotic resistant isolates that matched AR patterns found in isolates from WWTPs. Similarly, six of the seven sites in BC had AR patterns that matched with resistance patterns in WWTPs. These results suggest that AR testing may be a useful tool for differentiating E. coli from human and wildlife sources. Further testing of bacterial isolates from known animal sources is necessary to better assess the utility of this approach. Published by Elsevier B.V.