Integrated biogeochemical, geographical and hydrological approaches to track sources of contaminants in mixed land-use watersheds – CHYPSTER

An innovative spatial analysis methodology, combining geomatics with surveys of land use and practices, was developed and implemented in two mesoscale watersheds to identify the origin of water at outlet points in relation to hydrological processes and land use, as well as any potentially associated contaminants. Surveys were conducted with stakeholders in the territory in order to identify practices and management methods at the watershed scale. At the same time, maps of land use and cover were produced from remote sensing data and various databases, in order to finely represent the activities and uses on the two watersheds studied: the Claduègne watershed (in Ardèche) and the Ratier watershed (sub-basin of the Yzeron, near Lyon). This methodology made it possible to identify and locate potential sources of contaminants for both watersheds and to select sampling sites in streams.

The CHYPSTER project was then based on the acquisition of a large set of data from analyses (chemical and microbiological) of water samples collected on these two watersheds over several seasons with, on the one hand, field campaigns on water samples from diffuse sources (collected from small streams downstream of sub-watersheds that are homogeneous in terms of land cover and land use, geology, and soil type), as well as point sources of anthropogenic discharge, such as storm water outfalls or wastewater treatment plant discharges; and, on the other hand, samples collected during rainfall/hydrological events at mixing sites (or outlets) in the streams.

This robust dataset made it possible to define source-specific biogeochemical fingerprints and to develop and validate a methodology for decomposing water flow at watershed outlets during hydrological events - that is, to estimate the contributions of the previously identified sources using a geochemical mixing model. Three spatial water quality mapping campaigns (so-called “low-cost”) were conducted to supplement these data, providing a snapshot of the water quality in the streams of both watersheds at a given point in time, using simple analytical methods (pH, conductivity, UV-visible spectra, and microbial DNA markers).

We then compared the estimates of the contributions from the sources to the outlets of the two watersheds obtained through hydrological modeling, with the results of the decompositions derived from biogeochemical mixing models for the selected hydrological events. Finally, the validated hydrological models were applied to targeted scenarios to estimate the contribution of various sources and, ultimately, to predict water quality in the river, based on different trajectories (climatic and land-use) that were developed in collaboration with local socioeconomic stakeholders.

An innovative research approach—combining geography, surveys, chemical analyses, microbiology, and hydrology—was successfully implemented in two mixed-use watersheds. Detailed land-use maps (both present and historical) are available for these two watersheds. This methodology made it possible to identify potential sources of contaminants.

For each watershed, water samples were collected from about ten “source” sites and at the outlets during six hydrological events. The samples were analyzed using both conventional and innovative indicators: major and trace dissolved elements, characteristics of dissolved organic matter, and microbial indicators. At the same time, more than 200 organic micropollutants from chemical families representative of the main sources of contamination were analyzed in the same samples. In addition, two low-cost spatial water quality campaigns on the Ratier River and one on the Claduègne River supplemented these data with rapid measurements.

These datasets constitute a robust and exploitable resource to address the project's objectives.

We successfully applied a geochemical mixing model to both watersheds to trace the origin of the water at the outlets during rainfall events and estimate the contributions from various sources. At the same time, we identified cocktails of organic and microbiological contaminants typical of each source category in both watersheds.

A module for spatial decomposition of river flow was developed and integrated into the J2000P distributed hydrological model for the Ratier River and into the new version of the model developed and calibrated for the Claduègne River. The contributions from the sources simulated by J2000P were compared with the results from mixing models for the targeted hydrological events. The results from the mixing model demonstrated contamination of the Ratier waters by wastewater. These results are supported by the presence of fecal markers and indicators of dissolved organic matter specific to wastewater. In the Claduègne, the mixing model showed a significant contribution from surface runoff during floods and helped to validate the hydrological model. This study enabled a critical assessment of the strengths and weaknesses of each model and helped to consolidate the conceptual understanding of the hydrological functioning of the two watersheds.

A landscape, hydrological, and climate projection was conducted. Past land-use patterns were traced. Several representations of the landscapes by 2050 were developed, resulting in three scenarios for each watershed, which were translated into land-use maps. Finally, hydrological models were applied to conduct future simulations for both watersheds.

The results of the CHYPSTER project facilitate the identification and quantification of the main sources of contaminants related to land use in mixed land-use (multiple sources) and mesoscale watersheds, in order to define and initiate relevant actions (e.g., management of storm water outfalls and retention basins for the urban part, buffer zones in agricultural areas) to reduce inputs of contaminants to the hydrosystems and associated ecological risks and impacts. The future projections of river water quality, based on land use and cover change scenarios, will provide keys for water resources and land-use management at the watershed scale to reduce the potential contamination.

This project also provides new insights into the characterization of chemical cocktails and microbial contamination in rivers.

This project will contribute to the development of innovative tools and markers for water quality monitoring, transferable to other research laboratories, including private ones.

With regard to developments in non-targeted molecular analysis of dissolved organic matter, our research continues to focus on characterizing the chemical pressure signatures of streams in small watersheds and developing strategies for non-targeted analysis of organic compounds (Fanny Mathiaux PhD thesis, 2025–2028, RiverLy).

A better understanding of the main chemical and microbiological contaminants found in watersheds, as well as their sources and the role of these various factors in their fate, will provide greater insight into their large-scale distribution and, ultimately, help limit the release of these contaminants into the natural environment and/or propose appropriate mitigation strategies.

Finally, the CHYPSTER project has enabled the launch of three new research projects by project partners or in collaboration with them:

- HOTSPOT (AERMC / ZABR; 2026–2028) — Observing, describing, and modeling the connectivity of surface and groundwater flows in a small watershed with a Mediterranean climate (upper Claduègne watershed). Coordinator: Céline Duwig (IGE); partners: PACTE, RiverLy.

- CŒUR (AERMC / OTHU; 2026–2028) - Develop a tool to estimate the contributions of various urban sources to water and suspended particulate matter flows at the outlet of an urbanized watershed. Coord. Matthieu Masson (RiverLy), partners: DEEP (INSA-Lyon), Edytem (University of Savoie Mont Blanc), Graie, SAGYRC, Lyon Metropolitan Area.

- AQUASUBURB, ANR-25-CE22-3861 (2026–2030) – Quantification of transit times and underground flow paths of peri-urban base flow. Coordinator: Jérémie Bonneau (INSA-DEEP), partner: RiverLy.

The CHYPSTER project was presented at the OZCAR seminar (March 2022) and the OTHU seminar (June 2022), which allowed to communicate the objectives and the first results of the project to a large scientific audience at the national level. The advances on hydrological and geochemical models have been presented at several national and international conferences (O. Grandjouan thesis).
The article Bouchali et al. (Microorganisms, 2023; doi.org/10.3390/microorganisms11040922) presents an exploratory study of the modalities of correlations analyses between concentrations of pharmaceutical compounds and bacterial taxonomic entities for collected samples within impacted watersheds.
The paper by Boukra et al. (STOTEN, 2023; dx.doi.org/10.1016/j.scitotenv.2023.162104) explores the relationship between land use and cover, and dissolved organic matter fingerprints obtained in river samples downstream of homogeneous sub-watersheds.

The presence of numerous chemical and microbiological contaminants in aquatic ecosystems, including toxic (e.g. endocrine disruptors) and/or persistent contaminants can lead to a significant degradation of the ecological status of natural waters and increase human and animal exposures to these contaminants through drinking water, crop irrigation or recreational uses; they may also trigger a rise of antimicrobial resistance (AMR) genes and bacterial pathogens in hydrosystems. Mesoscale watersheds (~10-1000 km²) with mixed land-use and heterogeneous geologies are particularly sensitive to such contaminations. In the near future, increasing surface artificialization as well as climate change could result in critical degradation of water quality in these watersheds. The identification of contaminant sources in mesoscale watersheds, including non-point sources, the estimation of their contributions to watershed outlets, and the prediction of water quality in a context of climate change and land-use trajectory are key issues for the development of action plans to reduce contamination of freshwater ecosystems and protect their ecological state.
This CHYPSTER project aims to define a new interdisciplinary approach to identify the sources of anthropogenic contaminants and predict future water quality trajectories based on hydrological processes, changes in land-use and anthropogenic activities, and climate change. This approach will be applied on two mesoscale watersheds (the Claduègne and Yzeron sites that belong to the national research infrastructure OZCAR) by combining diagnosis of land cover and uses of chemicals, biogeochemical fingerprinting, and distributed hydrological modelling. Methodological expertise in spatial analysis, combining geomatics and survey, will be mobilized in order to identify potential sources of contaminants and their location. Advanced passive sampling, and target and non-target analytical tools, both chemical and microbiological, will be deployed to obtain specific datasets on contaminant sources and allow building-up new biogeochemical fingerprints that will be combined with flow contributions estimated by the hydrological model for hydrological events. The use of the validated hydrological model on scenarios of climate change and territories trajectories evaluated with local socio-economic actors will allow to make predictions on the surface water chemical and microbiological qualities at the outlet of watersheds.
This project will bring new knowledge on the characterization of chemical cocktails and microbial contamination (AMR and pathogen–related genes, source-specific keystone taxa) of freshwater ecosystems. The results will facilitate the location, identification and quantification of the main sources of contaminants (including Contaminants of Emerging Concerns [CECs]) along with land-use in catchments of mixed land-use (multiple sources) and mesoscale, in order to define and then initiate relevant actions (e.g. management of storm water outfalls and retention basins for the urban part, buffer zones in agricultural areas) to reduce contaminant pressure onto hydrosystems and the related hazards and ecological impacts. The projections of future river hydrology and water quality according to land-use and land cover change scenarios will also give keys for water and soils management at the catchment scale, in order to reduce contaminations. This project will also contribute to the development of innovative tools and markers for the monitoring of water quality that will be transferable to other research laboratories including private ones. Interestingly, the use of passive samplers termed germcatchers will allow to infer the relationships between chemical pollutants and bacterial taxa developing as biofilms in river systems. These data will thus bring novel insights on resistomes emerging in the environment, and the associated AMR gene recruitments or bacterial taxa enrichments.