ABSTRACT
AIMS: Carbon source utilization profiles as a phenotypic fingerprinting methodology for determining sources of faecal pollution in water were evaluated. METHODS AND RESULTS: Three hundred and sixty-five Enterococcus isolates were collected from known faecal sources in four different geographical regions and were identified to species with the commercial Biolog system. Discriminant analysis (DA) was used to identify the substrate-containing wells that best classified the 365 isolates by source. By using 30 of the 95 wells for the analysis, the average rate of correct classification (ARCC) by source was 92.7% for a human vs non-human two-way classification when isolates from all regions were combined into one library. Corresponding ARCCs for other classification schemes were 81.9% for a four-way classification of human vs livestock vs wildlife vs domestic pets, and 85.7% for a three-way classification without human isolates. When three individual libraries were made based on classification of sources within Enterococcus species, the ARCC was 95.3% for the Ent. faecalis library, 95.8% for the Ent. gallinarum library and 94.7% for the Ent. mundtii library. Thirty Enterococcus isolates (unknown sources) were obtained from each of three stream sites where a specific source of pollution was apparent; 90.0% of the isolates from a human-suspected source were classified as human, 86.6% were classified as livestock from a livestock-suspected site, and 93.3% were classified as wildlife from a wildlife-suspected site. CONCLUSIONS: Phenotypic fingerprinting with carbon source utilization profiles provided levels of correct classification by sources from an Enterococcus library that were in the upper range of those reported in the literature. ARCCs for three Enterococcus species-specific libraries were very high and may be the best approach for further developing this concept and methodology. SIGNIFICANCE ANC IMPACT OF THE STUDY: The results, based on a modest Enterococcus library and a preliminary field validation test, demonstrated the potential for carbon source utilization profiles to be employed as a phenotypic method for determining sources of faecal pollution in water.
