Monitoring methodologies reflecting the long-term quality and contamination of surface waters are needed to obtain a representative picture of pollution and identify risk drivers. This study sets a baseline for characterizing chemical pollution in the Danube River using an innovative approach, combining continuous three-months use of passive sampling technology with comprehensive chemical (747 chemicals) and bioanalytical (seven in vitro bioassays) assessment during the Joint Danube Survey (JDS4). This is one of the world's largest investigative surface-water monitoring efforts in the longest river in the European Union, which water after riverbank filtration is broadly used for drinking water production. Two types of passive samplers, silicone rubber (SR) sheets for hydrophobic compounds and AttractSPETM HLB disks for hydrophilic compounds, were deployed at nine sites for approximately 100 days. The Danube River pollution was dominated by industrial compounds in SR samplers and by industrial compounds together with pharmaceuticals and personal care products in HLB samplers. Comparison of the Estimated Environmental Concentrations with Predicted No-Effect Concentrations revealed that at the studied sites, at least one (SR) and 4-7 (HLB) compound(s) exceeded the risk quotient of 1. We also detected AhR-mediated activity, oxidative stress response, peroxisome proliferator-activated receptor gamma-mediated activity, estrogenic, androgenic, and anti-androgenic activities using in vitro bioassays. A significant portion of the AhR-mediated and estrogenic activities could be explained by detected analytes at several sites, while for the other bioassays and other sites, much of the activity remained unexplained. The effect-based trigger values for estrogenic and anti-androgenic activities were exceeded at some sites. The identified drivers of mixture in vitro effects deserve further attention in ecotoxicological and environmental pollution research. This novel approach using long-term passive sampling provides a representative benchmark of pollution and effect potentials of chemical mixtures for future water quality monitoring of the Danube River and other large water bodies.

• Keywords
• Bioanalytical equivalent, Concentration addition, Iceberg modelling, Mixture effects, Passive sampling, Risk prioritization,
• MeSH
• Androgen Antagonists MeSH
• Water Pollutants, Chemical * toxicity   analysis   MeSH
• Ecotoxicology MeSH
• Estrone MeSH
• Water Quality * MeSH
• Environmental Monitoring methods   MeSH
• Rivers chemistry   MeSH
• Publication type
• Journal Article MeSH
• Research Support, Non-U.S. Gov't MeSH
• Names of Substances
• Androgen Antagonists MeSH
• Water Pollutants, Chemical * MeSH
• Estrone MeSH

A conventional evaluation methodology for drinking water pollution focuses on analysing hundreds of compounds, usually by liquid chromatography-tandem mass spectrometry. High-resolution mass spectrometry allows comprehensive evaluation of all detected signals (compounds) based on their elemental composition, intensity, and numbers. We combined target analysis of 192 emerging micropollutants with nontarget (NT) full-scan/MS/MS methods to describe the impact of treatment steps in detail and assess drinking water treatment efficiency without compound identification. The removal efficiency based on target analytes ranged from - 143 to 97%, depending on the treatment section, technologies, and season. The same effect calculated for all signals detected in raw water by the NT method ranged between 19 and 65%. Ozonation increased the removal of micropollutants from the raw water but simultaneously caused the formation of new compounds. Moreover, ozonation byproducts showed higher persistence than products formed during other types of treatment. We evaluated chlorinated and brominated organics detected by specific isotopic patterns within the developed workflow. These compounds indicated anthropogenic raw water pollution but also potential treatment byproducts. We could match some of these compounds with libraries available in the software. We can conclude that passive sampling combined with nontargeted analysis shows to be a promising approach for water treatment control, especially for long-term monitoring of changes in technology lines because passive sampling dramatically reduces the number of samples and provides time-weighted average information for 2 to 4 weeks.

• Keywords
• Drinking water treatment, Log2FoldChange, Nontarget analysis, Organic micropollutants, Removal efficiency, Treatment effect,
• MeSH
• Water Pollutants, Chemical * analysis   MeSH
• Water Purification * methods   MeSH
• Environmental Monitoring methods   MeSH
• Ozone * analysis   MeSH
• Drinking Water * analysis   MeSH
• Tandem Mass Spectrometry methods   MeSH
• Publication type
• Journal Article MeSH
• Names of Substances
• Water Pollutants, Chemical * MeSH
• Ozone * MeSH
• Drinking Water * MeSH