International audience

Many tropical coastal ecosystems face human pressures related to tourism, land or sea use. We developed a practical procedure to involve stakeholders in an early stage of an ecological research project to map the Social-Ecological System (SES) in our case study Lac Bay, Bonaire island, as well as to identify and prioritize ecological research questions and nature management options in relation to a recent new threat: massive sargassum landings. In our procedure we used the Group Model Building methodology for identifying drivers, key variables and feedback loops in this SES. The underlying mechanisms of driving feedbacks were revealed and shared during these sessions. We identified and prioritized urgent ecological research questions for the conservation of seagrass beds and mangrove forests, and practical measures for nature management in Lac Bay. Both were used in follow-up scientific research and nature management plans, illustrating the applicability of our procedure for early science-stakeholder interaction.

International audience | Treatment Wetlands (TW) are employed in the treatment of wastewater, participating in a decentralised and sustainable urban water management. Their designs and usages vary greatly across the world, which limit the aggregation of comparable data for deeper understanding of the mechanisms ruling over water treatment. We developed a hybrid modelling methodology to combine mechanistic and empirical pre-sizing models, applied to a context of low amount of data and knowledge on Horizontal and Vertical Flow TW. We first collected data from the scientific literature and introduced a data quality validation step providing reliability weights associated with each observation. We secondly trained and compared machine-learning models on the assessed dataset. We finally tested two hybridizations with the tank-in-series model, and predicting an optimized surface with a uncertainty interval. This methodology is reproducible and we believe hybrid models can provide more accurate and reliable predictions, therefore facilitating implementations of TW in urban plannings.

Sargassum spp. strandings in the tropical Atlantic harm local ecosystems due to toxic sulfide levels. We conducted a mesocosm experiment to test the efficacy of iron(III) (hydr)oxides in (a) mitigating sulfide toxicity in mangroves resulting from Sargassum and (b) reducing potentially enhanced greenhouse gas emissions. Our results show that iron addition failed to prevent mangrove mortality caused by highly toxic sulfide concentrations, which reached up to 15,000 μmol l−1 in 14 days; timely removal may potentially prevent mangrove death. Sargassum-impacted mesocosms significantly increased methane, nitrous oxide, and carbon dioxide emissions, producing approximately 1 g CO2-equivalents m−2 h−1 during daylight hours, thereby shifting mangroves from sinks to sources of greenhouse gasses. However, iron addition decreased methane emissions by 62 % and nitrous oxide emissions by 57 %. This research reveals that Sargassum strandings have multiple adverse effects related to chemical and ecological dynamics in mangrove ecosystems, including greenhouse gas emissions.

Marine plastic pollution poses significant ecological, economic, and social challenges, necessitating innovative detection, management, and mitigation solutions. Spectral imaging and optical remote sensing have proven valuable tools in detecting and characterizing macroplastics in aquatic environments. Despite numerous studies focusing on bands of interest in the shortwave infrared spectrum, the high cost of sensors in this range makes it difficult to mass-produce them for long-term and large-scale applications. Therefore, we present the assessment and transfer of various machine learning models across four datasets to identify the key bands for detecting and classifying the most prevalent plastics in the marine environment within the visible and near-infrared (VNIR) range. Our study uses four different databases ranging from virgin plastics under laboratory conditions to weather plastics under field conditions. We used Sequential Feature Selection (SFS) and Random Forest (RF) models for the optimal band selection. The significance of homogeneous backgrounds for accurate detection is highlighted by a 97 % accuracy, and successful band transfers between datasets (87 %–91 %) suggest the feasibility of a sensor applicable across various scenarios. However, the model transfer requires further training for each specific dataset to achieve optimal accuracy. The results underscore the potential for broader application with continued refinement and expanded training datasets. Our findings provide valuable information for developing compelling and affordable detection sensors to address plastic pollution in coastal areas. This work paves the way towards enhancing the accuracy of marine litter detection and reduction globally, contributing to a sustainable future for our oceans.

The North Sea is one of the most industrialised marine regions globally. We integrated cetacean-dedicated aerial surveys (2015–2022) with environmental covariates and ship positions from the Automatic Identification System (AIS) to investigate the disturbance radius and duration on harbour porpoise distribution. This study is based on 81,511 km of line-transect survey effort, during which 6511 harbour porpoise groups (8597 individuals) were sighted. Several proxies for ship disturbance were compared, identifying those best explaining the observed distribution. Better model performance was achieved by integrating maritime traffic, with frequent traffic representing the most significant disturbance to harbour porpoise distribution. Porpoises avoided areas frequented distance over time improved model performance, while reasons for the lower performance of predicted ship sound levels remain unclear. This study demonstrates the short-term effects of maritime traffic on harbour porpoise distribution.

Catchment-scale plastic pollution assessments provide insights in its sources, sinks, and pathways. We present an approach to quantify macroplastic transport and density across the Odaw catchment, Ghana. We divided the catchment into the non-urban riverine, urban riverine, and urban tidal zones. Macroplastic transport and density on riverbanks and land were monitored at ten locations in December 2021. The urban riverine zone had the highest transport, and the urban tidal zone had the highest riverbank and land macroplastic density. Water sachets, soft fragments, and foam fragments were the most abundant items. Our approach aims to be transferable to other catchments globally.

Worldwide, coastal waters contain pollutants such as nutrients, plastics, and chemicals. Rivers export those pollutants, but their sources are not well studied. Our study aims to quantify river exports of nutrients, chemicals, and plastics to coastal waters by source and sub-basin worldwide. We developed a new MARINA-Multi model for 10,226 sub-basins. The global modelled river export to seas is approximately 40,000 kton of nitrogen, 1,800 kton of phosphorous, 45 kton of microplastics, 490 kton of macroplastics, 400 ton of triclosan and 220 ton of diclofenac. Around three-quarters of these pollutants are transported to the Atlantic and Pacific oceans. Diffuse sources contribute by 95–100 % to nitrogen (agriculture) and macroplastics (mismanaged waste) in seas. Point sources (sewage) contribute by 40–95 % to phosphorus and microplastics in seas. Almost 45 % of global sub-basin areas are multi-pollutant hotspots hosting 89 % of the global population. Our findings could support strategies for reducing multiple pollutants in seas.

Cryptic species are rarely considered in ecotoxicology, resulting in misleading outcomes when using a single morphospecies that encompasses multiple cryptic species. This oversight contributes to the lack of reproducibility in ecotoxicological experiments and promotes unreliable extrapolations. The important question of ecological differentiation and the sensitivity of cryptic species is rarely tackled, leaving a substantial knowledge gap regarding the vulnerability of individual cryptic species within species complexes. In times of agricultural intensification and the frequent use of pesticides, there is an urgent need for a better understanding of the vulnerability of species complexes and possible differences in adaptive processes. We used the cryptic species complex of the aquatic amphipod Gammarus roeselii, which comprises at least 13 genetic mtDNA lineages and spans from small-scale endemic lineages in Greece to a large-scale widely distributed lineage in central Europe. We exposed eleven populations belonging to four lineages to the neonicotinoid thiacloprid in an acute toxicity assay. We recorded various environmental variables in each habitat to assess the potential pre-exposure of the populations to contaminants. Our results showed that the populations differed up to 4-fold in their tolerances. The lineage identity had a rather minor influence, suggesting that the cryptic species complex G. roeselii does not differ significantly in tolerance to the neonicotinoid thiacloprid. However, the observed population differentiation implies that recent pre-exposure to thiacloprid (or similar substances) or general habitat contamination has triggered adaptive processes. Though, the extent to which these mechanisms are equally triggered in all lineages needs to be addressed in the future. Our study provides two key findings: Firstly, it shows that observed phylogenetic differences within the G. roeselii species complex did not reveal differences in thiacloprid tolerance. Second, it confirms that differentiation occurs at the population level, highlighting that susceptibility to toxicants is population-dependent. The population-specific differences were within the range of accepted intraspecific variability from a regulatory standpoint. From an evolutionary-ecological perspective, it remains intriguing to observe how persistent stresses will continue to influence tolerance and whether different populations are on distinct pathways of adaptation. Given that the potential selection process has only lasted a relatively short number of generations, it is crucial to monitor these populations in the future, as even brief exposure periods significantly impact evolutionary responses.