Genetic Microbial Source Tracking Support QMRA Modeling for a Riverine Wetland Drinking Water Resource.

Riverine wetlands are important natural habitats and contain valuable drinking water resources. The transport of human- and animal-associated fecal pathogens into the surface water bodies poses potential risks to water safety. The aim of this study was to develop a new integrative modeling approach supported by microbial source tracking (MST) markers for quantifying the transport pathways of two important reference pathogens, and , from external (allochthonous) and internal (autochthonous) fecal sources in riverine wetlands considering safe drinking water production. The probabilistic-deterministic model QMRAcatch (v 1.1 python backwater) was modified and extended to account for short-time variations in flow and microbial transport at hourly time steps. As input to the model, we determined the discharge rates, volumes and inundated areas of the backwater channel based on 2-D hydrodynamic flow simulations. To test if we considered all relevant fecal pollution sources and transport pathways, we validated QMRAcatch using measured concentrations of human, ruminant, pig and bird associated MST markers as well as in a Danube wetland area from 2010 to 2015. For the model validation, we obtained MST marker decay rates in water from the literature, adjusted them within confidence limits, and simulated the MST marker concentrations in the backwater channel, resulting in mean absolute errors of < 0.7 log particles/L (Kruskal-Wallis > 0.05). In the scenarios, we investigated (i) the impact of river discharges into the backwater channel (allochthonous sources), (ii) the resuspension of pathogens from animal fecal deposits in inundated areas, and (iii) the pathogen release from animal fecal deposits after rainfall (autochthonous sources). Autochthonous and allochthonous human and animal sources resulted in mean loads and concentrations of and (oo)cysts in the backwater channel of 3-13 × 10 particles/hour and 0.4-1.2 particles/L during floods and rainfall events, and in required pathogen treatment reductions to achieve safe drinking water of 5.0-6.2 log. The integrative modeling approach supports the sustainable and proactive drinking water safety management of alluvial backwater areas.