Real-time genomic pathogen, resistance, and host detection in wetland ecosystems from passive water sampling

Wetlands are critical ecosystems where anthropogenic pressures can create public health risks. Adopting a One Health framework, we developed and deployed an integrated surveillance strategy across twelve European wetlands with varying human impact. Our approach combines non-invasive passive water sampling with a multi-omic nanopore sequencing workflow to concurrently profile vertebrate communities (12S eDNA), the water microbiome, and antimicrobial resistance (AMR) genes. The framework successfully identified sites under heavy anthropogenic influence, particularly animal farming, as locations with high relative abundances of potential pathogens and AMR genes. We established a strong source-pathogen association at a French duck farm, where a dominant eDNA signal from farmed ducks (Anas platyrhynchos) coincided with high relative abundances of the pathogen Aeromonas veronii. Crucially, leveraging long-read data for metagenomic assembly allowed us to confirm this association with high confidence by locating multiple ß-lactamase genes directly on assembled A. veronii contigs. In addition to these contemporary threats, our approach also uncovered a legacy of historical pollution, identifying arsenic resistance genes (arsB, arsC) on contigs from Acidiphilium multivorum, a bacterium indicative of metal-contaminated sites. By genetically linking specific AMR determinants to potential pathogens and their likely agricultural sources, this work demonstrates a powerful framework for prioritizing sites for targeted environmental monitoring. This integrated approach moves beyond simple hazard detection to provide valuable genomic context essential for hazard identification and mitigation strategies under a One Health paradigm.