Molecular indicators used in the development of predictive models for microbial source tracking.

Area of Science:

• Environmental microbiology
• Molecular biology
• Water quality assessment

Background:

• Identifying fecal pollution sources in water bodies is crucial for public health.
• Existing chemical, microbial, and eukaryotic indicators have limitations in pinpointing pollution origins.
• Molecular methods offer a promising complement to traditional water analysis for microbial source tracking (MST).

Purpose of the Study:

• To evaluate the feasibility of various molecular indicators for MST.
• To develop and assess predictive models for identifying fecal pollution sources.
• To determine the optimal combination of molecular markers for accurate source identification.

Main Methods:

• Compared host-specific Bacteroidetes markers (HF134, HF183, CF128, CF193), Bifidobacterium adolescentis (ADO), Bifidobacterium dentium (DEN), Enterococcus faecium esp gene, and mitochondrial DNA (Humito, Bomito, Pomito).
• Developed MST predictive models using combinations of these molecular markers.
• Evaluated models based on the lowest number of markers yielding the highest source identification rate.

Main Results:

• No single molecular marker achieved 100% source identification.
• A combination of three markers (ADO, Bomito, Pomito) correctly identified 75.7% of sources (human, swine, bovine, poultry).
• Discriminating human from nonhuman sources was achieved with 84.6% accuracy using two markers (ADO, Pomito); a five-marker model (HF134, ADO, DEN, Bomito, Pomito) reached 90.1% accuracy.

Conclusions:

• Combined use of molecular markers significantly enhances sensitivity and specificity for MST.
• The optimal predictive model for distinguishing human from nonhuman fecal sources utilized five specific molecular markers.
• Multi-marker approaches are essential for robust and accurate determination of fecal pollution origins in water.