Application of fast expectation-maximization microbial source tracking to discern fecal contamination in rivers exposed to low fecal inputs.

Abstract

Community-based microbial source tracking (MST) can be used to determine fecal contamination from multiple sources in the aquatic environment. However, there is little scientific information on its application potential in water environmental management. Here, we compared SourceTracker and Fast Expectation-maximization Microbial Source Tracking (FEAST) performances on environmental water bodies exposed to low fecal pollution and evaluated treatment effects of fecal pollution in the watershed utilizing community-based MST. Our results showed that FEAST overall outperformed SourceTracker in sensitivity and stability, and was able to discern multi-source fecal contamination (mainly chicken feces) in ambient water bodies exposed to low fecal inputs. Consistent with our previous PCR/qPCR-based MST assays, FEAST analysis indicates that fecal pollution has been significantly mitigated through comprehensive environmental treatment by the local government. This study suggests that FEAST can be a powerful tool for accurately evaluating the contribution of multi-source fecal contamination in environmental water, facilitating environmental management.