Spatial and temporal variability of ribotyping results at a small watershed in South Carolina.

The utility of library-based ribotyping methods for a very small study area was evaluated through comparison of local results to libraries with differing spatial and temporal scales. Ribotyping of Escherichia coli isolates was used to evaluate sources of fecal pollution at a coastal golf course in Beaufort County, South Carolina. Thirty-five E. coli isolates were obtained from water samples from a detention pond for testing against several local and regional libraries of known-source isolate patterns. A library of 92 E. coli ribotype patterns was created from wildlife feces obtained on the site. Additional libraries were available for comparison, including a library from Morgan Island, a small, geographically isolated area (including a monkey colony), and a library from ongoing statewide assessments. Seventeen (49%) of the unknown E. coli isolates matched isolates from raccoon and deer scat from the local library. Two isolates (6%) were matched with monkey sources from Morgan Island, and 13 (37%) were matched to raccoon, deer, and cows from the statewide assessment. Evaluation of repeated ribotyping analyses at the study area revealed evidence of temporal variability of potential sources in the local library. Only one of the isolates from the second year of fecal samples successfully matched with a fecal isolate from the previous year. The results from this study suggest that source identification results were variable both spatially and temporally, and that local, temporally specific libraries are most appropriate for library-based MST studies in small watersheds. Results also suggest that it will be difficult to employ adequate sample sizes to satisfactorily address unknown pattern variability.