Rapport intermédiaire du projet Vigie-Lacs
L’objectif du projet est d’acquérir des connaissances scientifiques indispensables à la préservation des communautés de plantes aquatiques des lacs et étangs du littoral aquitain fortement menacées par les activités humaines et le changement climatique. Ce projet pluridisciplinaire fait appel à plusieurs domaines scientifiques complémentaires, telles que l’écologie des communautés, l’autécologie, la biogéochimie, les biostatistiques, la génétique ou encore l’hydrogéologie.
(pp. 24, 21/02/2026)
UR EABX, INRAE, ECLA, USMB [Université de Savoie] [Université de Chambéry], INRAE, OFB, BioGeCo, UB, INRAE, EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
Differentiating estuarine dissolved organic matter composition by unsupervised and supervised machine learning
Differentiating the composition of Dissolved Organic Matter (DOM) in estuaries is a major environmental concern, as the DOM characteristics are closely linked to biogeochemical and ecological considerations (e.g. water properties and trophic cycling). However, tracing the spatiotemporal variations of estuarine DOM is challenging due to multiple sources and complex transformation processes. Here, we investigate the dynamics of estuarine DOM by analyzing the optical properties of DOM through UV-Visible absorbance and fluorescence spectroscopy, while also capturing the variability of DOM using machine learning algorithms and explainable artificial intelligence. To this aim, we collected sub-surface water samples (n = 249) from a human-impacted estuary with intense industrialization and urbanization in France (Seine Estuary) across distinct land use characteristics in contrasting hydrological conditions. We then applied unsupervised and supervised machine learning techniques to analyze the optical properties of DOM, which were determined by UV-Visible absorbance and Excitation-Emission Matrix (EEM) fluorescence spectroscopy combined with parallel factor analysis (PAR-AFAC). Our results show that unsupervised machine learning (K-means clustering) captures the spatial variabilities of DOM, identifying three distinct estuarine zones based on pronounced spatial variations of several DOM optical parameters. Supervised machine learning (Light Gradient Boosted Machine, LightGBM) further validates the rationality of the defined zonation. Subsequently, explainable artificial intelligence based on SHapley Additive exPlanations (SHAP) analysis shows that DOM in each zone has specific characteristics. Our model indicates that DOM in the Seine Estuary is primarily influenced by high molecular weight materials and autochthonous contributions in the upper estuary (Zone I). The dominant contribution to DOM in the mid-estuary (Zone II) comes from autochthonous and aromatic material as well as transformation and (photo)degradation products. Lower estuary (Zone III) is mainly characterized by aromatic DOM (subject to photodegradation), low molecular weight compounds, autochthonous DOM, as well as transformation and (photo)degradation products. Overall, this study presents a workflow for differentiating the composition of DOM, tracing the variability and dynamics of DOM along the land-to-sea continuum, and elucidating the involved processes. The approach developed in the Seine Estuary has significant implications for environmental management and can be adapted to other land-sea continuums.
(Water Research. vol. 284, n° 0043-1354, pp. 123900, 21/02/2026)
METIS, EPHE, PSL, INSU - CNRS, SU, CNRS, EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
Semi-empirical forecast modelling of rip-current and shore-break wave hazards
Sandy beaches are highly attractive but also potentially dangerous environments for those entering the water as they can be exposed to physical hazards in the surf zone. The most severe and widespread natural bathing hazards on beaches are rip currents and shore-break waves, which form under different wave, tide, and morphological conditions. This paper introduces two new, simple semiempirical rip-current and shore-break wave hazard forecast models. These physics-informed models, which depend on a limited number of free parameters, can be used to compute the time evolution of the rip-current flow speed V and shore-break wave energy E sb . These models are applied to a high-energy mesotidal-macrotidal beach, La Lette Blanche, in southwest France, where intense rip-currents and shore-break wave hazards co-exist. Hourly lifeguardperceived hazards collected during patrolling hours (from 11:00 a.m. to 07:00 p.m. LT (UTC+2)) during July and August of 2022 are used to calibrate the two models. These data are also used to transform V and E sb into a five-level scale from 0 (no hazard) to 4 (hazard maximised). The model accurately predicts rip-current and shore-break wave hazard levels, including their modulation by tide elevation and incident wave conditions, opening new perspectives for forecasting multiple surf-zone hazards on sandy beaches. In addition, daily-mean hazard forecasts demonstrate even greater predictive skill, which is important for conveying straightforward messages to the general public and lifeguard managers. The approach presented here only requires a limited number of beach morphology metrics and allows for the prediction of surf-zone hazards on beaches where wave and tide forecasts are available.
(Natural Hazards and Earth System Sciences. vol. 25, n° 1561-8633, pp. 2379–2397, 21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS, UR ETTIS, INRAE, SMGBL, UR EABX, INRAE
Hourly irradiance data measured underwater on oyster tables during a one-year semi-controlled experiment
This dataset shows the hourly irradiance data measured underwater on each oyster table (control and ALAN conditions, at the oysters' level) throughout the year of the experiment as well as the temperature recorded by each irradiance sensor.
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
Hourly sound pressure magnitude data measured underwater on oyster tables during a one-year semi-controlled experiment
This dataset shows the hourly sound pressure magnitude data measured underwater on each oyster table (control and ALAN conditions, at the oyster's level) throughout the year of experiment for different frequencies (10 Hz, 20 Hz, 30 Hz, 40 Hz, 50 Hz, 60 Hz, 70 Hz, 80 Hz, 90 Hz, 100 Hz, 200 Hz, 300Hz, 400 Hz, 500 Hz, 600 Hz, and 700 Hz). The mean sound pressure magnitude is also presented for each condition. In the ALAN condition, data are missing from August 13, 2024, until the end of the experiment due to the failure of the hydrophone used.
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
Hourly physicochemical parameters measured underwater on oyster tables during a one-year semi-controlled experiment
This dataset shows the hourly physicochemical parameters measured underwater throughout the year of the experiment. The temperature and water depth were measured on each oyster table (control and ALAN conditions, at the oyster's level), while the turbidity was measured only on the ALAN table, and conductivity and salinity were measured only on the control table.
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
Valve behavior of the European flat oyster Ostrea edulis and associated underwater sound and temperature data, recorded during a 18-month in situ experiment in Helgoland, Germany
Once widespread across European coasts, the native flat oyster Ostrea edulis has now disappeared from most of its historical range and is officially recognized as threatened. As a key ecological engineer, this species supports biodiversity by filtering water, stabilizing sediments, and providing complex reef habitats. Understanding and evaluating its behavior and biological rhythms in a natural environment before reintroduction, and how it responds to natural geophysical cycles, is essential to support effective restoration strategies. However, current knowledge on O. edulis remains limited, with most studies focusing primarily on reproduction under aquaculture or laboratory conditions. To help fill this gap, we conducted a 18-month in situ study to assess the valve behavior of Ostrea edulis in the field. The experiment took place at the Margate site (54.19°, 7.88°) near the island of Helgoland (Germany) from the 11th of March 2023 to the 31st of August 2024. The experimental setup consisted of 16 oysters disposed on individual cages in a customized oyster basket placed on a lander, a metallic structure immersed at 10m depth. Their valve behavior was continuously measured during 18 months using a High-Frequency Non-Invasive (HFNI) valvometer biosensor (Tran et al. 2023; Le Moal et al. 2023 for further details). Briefly, a pair of lightweight electrodes (<100 mg) was glued on each half-shell of each oyster and was linked to the HFNI valvometer by a flexible wire, allowing undisturbed oyster valve movement. An electromagnetic field was generated between the electrodes, allowing the measurement of the distance between each oyster's valve in continuous mode. In addition to the oyster behavior, environmental parameters were continuously measured underwater by the HFNI valvometer biosensor during the experiment, such as temperature and sound pressure magnitude. This compilation of datasets gives an overview of environmental parameters and behavioral data collected during this experiment.
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS, AWI, AWI
Valve behavior of the European flat oyster Ostrea edulis and underwater temperature, recorded during a 18-month in situ experiment in Helgoland, Germany
This dataset shows the hourly valve behavioral data of the 16 oysters Ostrea edulis throughout the 18 months of the experiment (11th of May 2023 - 31st of August 2024) in Helgoland (Margate) as well as the temperature recorded by the HFNI valvometer. The oyster valve behavior is characterized by 3 parameters: the Valve Opening Amplitude (VOA, the percentage of the valve opening relative to maximum opening), the Valve Opening Duration (VOD, the percentage of time that an oyster spends with its valves open), and the VOA/VOD. The data are presented for each oyster and as a group average. Missing data corresponds to the death of the oyster number 7 or to the stop of recording due to electrical failure on the oyster's valvometer electrodes.
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS, AWI, AWI
Assessing the invasion risk of the cnidaria Blackfordia virginica Mayer, 1910: a threat to the Baltic Sea ecosystem?
The ecological role, bloom extent and long-term dynamics of jellyfishes are mostly overlooked due to sampling limitations, leading to the lack of continuous long-term datasets. A rise in frequency and magnitude of jellyfish invasion around the world is shedding new light on these organisms. In this study, we estimate the current and future distribution of the introduced jellyfish Blackfordia virginica in the Baltic Sea. We determine the combination of favorable levels of temperature and salinity for this species by analyzing presence/absence data from areas outside the Baltic Sea and project the distribution of suitable habitat in the Baltic Sea across different scenarios with variable climate forcing and eutrophication levels. Our results show that suitability increases with rising temperature and optimal salinity range from 13 to 20 for this species. In addition, a relatively large area of the Baltic Sea represents favorable abiotic conditions for B. virginica , enhancing the concerns on its potential range expansion. Spatial analysis illustrates that the coastal areas of the southern Baltic Sea are particularly at risk for the invasion of the species. The observation of the projection of habitat suitability across time highlights that future Baltic Sea environmental conditions increase suitability levels for B. virginica and suggest a potential expansion of its distribution in the future.
(Biological Invasions. vol. 27, n° 1387-3547, pp. 106, 21/02/2026)
BOREA, MNHN, IRD, SU, CNRS, UA, GEOMAR, CCMAR, UAlg, EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS
From fixed to transgressive dunes, the conditions and timing of the transition along the Aquitaine coast, France
Today most of the coastal dunes in temperate latitudes, especially in the northern hemisphere, are relatively stable. However, over the last decade, the Gironde coast, southwest France, has experienced substantial natural dune remobilization following a major marine erosion event. Annual, large-scale and highresolution, airborne LiDAR data and Satellite imagery (Sentinel-2) are combined to address the coastal dune morphological changes and establish relations with forcing and controlling factors (vegetation cover, geomorphological descriptors). Between 2014 and 2023, about 10 out of 85 km of the Gironde dunes have switched from fixed to transgressive state. The analysis showed that in the vast majority of the cases the dominant process involved was dune front cannibalism. However, there is considerable spatial and temporal variability along the coast, depending on the vegetation cover evolution, the amount of sediment remobilized and the morphological characteristics of the dunes (steepness of the front slope, width).
(21/02/2026)
EPOC, EPHE, PSL, UB, INSU - CNRS, CNRS