Sie können Operatoren mit Ihrer Suchanfrage kombinieren, um diese noch präziser einzugrenzen. Klicken Sie auf den Suchoperator, um eine Erklärung seiner Funktionsweise anzuzeigen.
Findet Dokumente, in denen beide Begriffe in beliebiger Reihenfolge innerhalb von maximal n Worten zueinander stehen. Empfehlung: Wählen Sie zwischen 15 und 30 als maximale Wortanzahl (z.B. NEAR(hybrid, antrieb, 20)).
Findet Dokumente, in denen der Begriff in Wortvarianten vorkommt, wobei diese VOR, HINTER oder VOR und HINTER dem Suchbegriff anschließen können (z.B., leichtbau*, *leichtbau, *leichtbau*).
Dieser Artikel geht in die komplexe Welt der Diagnosewerkzeuge ein, die verwendet werden, um die Risiken von Chemikalien für Süßwasser-Ökosysteme zu bewerten. Zunächst werden die Herausforderungen bei der Messung und Bewertung der Auswirkungen chemischer Mischungen auf Wasserorganismen skizziert und die Notwendigkeit eines einheitlichen Bewertungsrahmens betont. Der Artikel untersucht verschiedene Diagnoseinstrumente, darunter Messgrößen für toxische Druckmessungen, Bioassays und ökologische Überwachungsindizes, wobei ihre Stärken und Grenzen hervorgehoben werden. Sie diskutiert auch die Integration dieser Instrumente in den Rahmen der ökotoxikologischen Risikobewertung und gibt Empfehlungen für weitere Verbesserungen. Eine Fallstudie an deutschen Bächen zeigt den Einsatz dieser Instrumente und zeigt, wie sie eingesetzt werden können, um die ökologischen Risiken realistischer Schadstoffgemische einzuschätzen. Der Artikel schließt mit der Beseitigung wichtiger Datenlücken und methodischer Fortschritte, die notwendig sind, um eine bessere Abstimmung zwischen den Ergebnissen verschiedener Diagnosewerkzeuge zu erreichen, und spricht sich für einen integrativen Rahmen zur Verbesserung der Zuverlässigkeit ökologischer Risikobewertungen aus.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
The risk assessment of chemicals relies on multiple tools to quantify the ecological responses of ecosystems to existing chemical pollution. These tools are broadly categorized into three major groups: toxic pressure assessments, bioassays, and ecological monitoring. Here, we examine the strengths and limitations of these approaches, their current level of implementation for freshwater ecosystems across Europe, and their ability to evaluate the impacts of chemicals under field conditions. Additionally, we analyze the correspondence between results obtained from these tools when applied to a monitoring dataset from German streams. Our evaluation showed that no single tool can perfectly characterize the environmental impacts of chemical mixtures. However, each provides distinct lines of evidence, enabling the identification of chemicals driving ecological risks and the biological endpoints most likely to be affected, with ecological monitoring tools having the potential to show long-term ecosystem impairment. Finally, we propose recommendations to better understand the discrepancies between the outcomes of different methods and explore their potential integration into a unified water quality evaluation framework.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
Aquatic ecosystems are exposed to multiple organic and inorganic chemicals, including pesticides, pharmaceuticals, metals, and industrial compounds (Busch et al., 2016; Wilkinson et al., 2022). These chemicals form complex mixtures that can affect aquatic organisms simultaneously and in sequence (Van Gils et al., 2020; Rorije et al. 2022). Such complexity highlights the importance of diagnostic (or retrospective) risk assessments, which aim to identify the impacts of chemicals that are in use or that have already been emitted into the environment, relying on advanced tools to characterize chemical exposure and effects, and supporting regulatory decisions on restricting, phasing out, or banning of substances (Vijver et al. 2017; Rico et al. 2021a). To date, our ability to quantitatively assess the risks of chemicals to aquatic ecosystems remains limited due to several challenges. These include difficulties in measuring every single substance that contributes to the environmental exposome (Scholz et al., 2022), the lack of toxicity data for many substances (Treu et al. 2024), the extrapolation of individual-level effects assessed under laboratory conditions to higher levels of biological organization (Schneeweiss et al., 2023), and accounting for multiple (non-)chemical stressors that collectively influence the structure and function of ecosystems (Rico et al., 2016; Van den Brink et al., 2018).
In Europe, the Water Framework Directive (WFD) incorporates multiple lines of evidence to assess the ecological status of aquatic ecosystems and to try to establish causality between measured sources of ecosystem impairment and ecological effects (EC, 2010), following methods that align to the TRIAD approach. The TRIAD approach, introduced in the 1980s by Long and Chapman (1985) for characterizing risks in contaminated sediments, integrates three lines of evidence: chemical (characterization of contaminant exposure and toxic pressure assessment), (eco)toxicological (use of in vivo bioassays), and ecological (biological indices measured under field conditions). This approach was later expanded to account for site-specific features and distinguish between chemical and non-chemical stressor effects (Chapman, 2000). Since then, the approach has been applied in several studies to identify chemicals and biological components affected in contaminated sites (Wolfram et al. 2012, Wolf et al., 2020; Martinez-Haro et al. 2022; Lee, Khim 2022).
Anzeige
The toolbox that supports the diagnostic risk assessment of chemicals in aquatic ecosystems, in line with the TRIAD approach, has progressed significantly over the past decades (Fig. 1). For instance, toxic pressure assessment metrics, which estimate the likelihood that contaminants will adversely affect biological communities by integrating exposure concentrations with toxicity data, enable the calculation of risk measures based on deterministic or probabilistic methods (De Zwart and Posthuma, 2005; Van den Brink et al., 2002). Traditionally, ecotoxicological assessments were based on in vivo animal tests that provide direct evidence of biological effects in sentinel organisms and biological systems. More recently these have been complemented by in vitro bioassays that range from simple high-throughput screening tools to sophisticated assays using omics techniques. In vitro bioassays allow the characterization of substances’ modes of action, and in combination with fractionation and chemical analysis, even molecular structural identifications and characterization of complex environmental exposures (Escher et al., 2020). Additionally, the repertoire of biological indices that provide quantitative metrics derived from the presence, abundance, and structural and functional parameters of biological communities has grown substantially (Ofogh et al., 2024).
Fig. 1
Toolbox for the diagnostic risk assessment of chemicals in freshwater ecosystems
Despite these advances, the complementary use of these tools for diagnostic risk assessment of contaminant mixtures remains limited (Brack et al., 2017; Munthe et al., 2017; Nowell et al. 2024). Integrating these tools into the ecotoxicological risk assessment framework presents opportunities to partially validate the outcomes obtained by other assessment methods, to elucidate toxicological mechanisms, and extrapolate effects to higher levels of biological organization. In light of this, the aims of this paper are: (1) to discuss the development and current implementation status of these tools in diagnostic risk assessment for chemicals in aquatic ecosystems; (2) to evaluate how and to what extent these tools have been employed to validate ecosystem-level effects; and (3) to provide recommendations for further integration of these tools into the ecological risk assessment framework. These discussions are based on the outcomes of a workshop organized by scientists from the NORMAN network (https://www.norman-network.net/), which focuses on improving the risk assessment of chemicals of emerging concern at the ecosystem level. Furthermore, the article presents an evaluation of pesticide mixtures’ effects in German streams, where a suite of diagnostic tools has been used as an example.
Progress in the development of diagnostic tools
Toxic pressure assessment tools
The measurement of chemical exposure concentrations under field conditions enables the characterization of theoretical toxic pressure or ecological risks posed by chemical mixtures, utilizing laboratory and/or semi-field toxicity data for individual substances. Laboratory toxicity data for standard test species have been applied to calculate Toxic Units (TUs), which scale measured environmental concentrations of chemicals to a reference toxicity value, such as the acute Effective Concentration for 50% of organisms (EC50) for Daphnia magna. Aggregated measures, assuming additivity, have also been developed, such as the summation of TUs for individual compounds in a mixture (Liess and von der Ohe 2005). Additionally, Predicted No Effect Concentrations (PNECs), derived using assessment factors, are employed to calculate risk quotients (RQs) for individual compounds (Ginebreda et al., 2010). One of the most extensive PNEC datasets is maintained by the NORMAN network, encompassing PNECs for 93,556 substances (accessed 9th of March 2025) derived from experimental data and QSAR models for various environmental compartments (https://www.norman-network.com/nds/ecotox/).
The RQ approach is widely recommended for identifying unacceptable ecological risks in regulatory frameworks such as the WFD (EC, 2013). Under the WFD, environmental quality standards are employed using both maximum exposure and annual average concentrations measured across various sites. This approach has been effective in risk prioritization for large chemical monitoring datasets due to its minimal data requirements (e.g., Wilkinson et al., 2022). However, the RQ approach falls short of estimating the actual magnitude of ecological effects on individual trophic levels when the calculated RQ of one or the sum of RQs for multiple chemicals exceeds one.
Anzeige
Laboratory toxicity data for multiple species can be used to construct Species Sensitivity Distributions (SSDs). These distributions allow for a probabilistic estimation of the potentially affected fraction (PAF) of species impacted by individual chemicals. When applied to multiple substances, this approach is referred to as the multi-substance potentially affected fraction (msPAF; De Zwart and Posthuma, 2005). The msPAF is derived by applying SSD parameters to all monitored compounds in a sample, using mixture models such as concentration addition or independent action (De Zwart and Posthuma, 2005). The approach relies on extensive toxicity datasets to estimate a theoretical SSD and the calculation of the proportion of species affected by a contaminant mixture (Posthuma et al., 2019). Similarly to the TU or RQ methods, a significant limitation of the msPAF approach is its inability to quantify the magnitude of impacts on populations and its omission of ecological processes, such as species interactions and indirect effects caused by changes in water quality or ecosystem functions. Despite these limitations, the msPAF approach has been used for diagnostic chemical risk assessments in several countries (Rämö et al., 2018; Posthuma et al., 2019; Rico et al., 2021b; Oginah et al., 2025).
Micro-/mesocosm studies have traditionally been used to derive No Observed Effect Concentrations (NOECs) from which PNEC values are extrapolated for use in high-tier prospective risk assessments (Fig. 1; EC, 2011); EFSA, (2013). The advantage of model ecosystem studies (i.e., micro- and mesocosms) lies in their greater ecological realism, as they account for both direct and indirect chemical effects on populations and community-level responses. Moreover, data from these studies can be used to inform predictive models for ecosystem-level effects. One example is the PERPEST model (Van den Brink et al., 2002), which utilizes a database of model ecosystem experiments to estimate the probability of slight or severe effects on various biological endpoints caused by chemical mixtures measured in the environment. This is achieved using machine learning tools such as case-based reasoning. Currently, the source dataset is populated with model ecosystem studies focused on pesticides, which means that, to date, the model has been used exclusively for risk assessments of these compounds (Van den Brink et al., 2006a; Rämö et al., 2018). However, the use of probabilistic approaches based on micro- and mesocosm experimental results, or similar high-tier datasets, shows significant promise for advancing diagnostic chemical risk assessments encompassing higher levels of biological organization (e.g., Mentzel et al., 2024).
Bioassays
A comprehensive battery of bioassays is available to evaluate the impact of chemical mixtures at the sub-individual and individual levels (Fig. 1; Schuijt et al., 2021). These bioassays, which include in vitro and in vivo techniques, enable the detection of chemical effects on morphological and physiological responses in model species and/or (sub)cellular systems (Barata et al., 2007; König et al., 2017). In vitro bioassays utilize cell cultures or subcellular systems derived from organisms such as mammals or fish, or modified bacteria. Reporter gene assays are particularly popular, as they amplify and visualize cellular processes such as the activation of estrogen receptors by estrogenic chemicals. When native nuclear receptors or transcription factors—activated at specific steps in cellular toxicity pathways—bind to chemicals, they trigger the transcription and translation of a reporter gene. This gene encodes easily measurable proteins, such as fluorescent proteins or enzymes (e.g., β-galactosidase or luciferase). The resulting measurable response (e.g., fluorescence intensity or enzyme activity) correlates with the degree of receptor binding, providing a direct proportional measure of the chemical’s (toxic) effect, which is usually expressed in dilution times needed to find a non-toxic effect (i.e., a non-significant measurable response of the endpoint or the measured enzymatic activity; Escher et al., 2021). Some of these bioassays have been used to establish effect-based trigger (EBT) values for surface water (Escher et al., 2018; Neale et al. 2023). Furthermore, these assays can be supplemented with sequential chemical analyses to confirm the specific chemicals responsible for the observed biological activity. Additionally, fractionation techniques, applied under the framework of Effect-Directed Analysis (EDA), can be used to select chemical compounds exhibiting specific activities from complex contaminant mixtures (Brack et al., 2016), helping in the risk prioritization of substances and on the determination of toxic mode of action of single substances or mixtures.
Active monitoring is conducted using sentinel organisms either caged within the study area (Liess and Schulz 1999) or tested under laboratory conditions using water samples collected from the field, which are commonly termed in vivo bioassays. As contaminants often occur at low concentrations without obvious effects, water samples used for in vivo testing are typically enriched, e.g., via solid-phase extraction, to cover a wider range of exposure concentration levels and amplify response detection (Neale et al., 2018). The assays measure a broad spectrum of response endpoints, encompassing both exposure biomarkers that identify pollutants through tissue bioaccumulation analysis or metabolite detection, and effect biomarkers that evaluate biological responses to pollution. Effect biomarkers include molecular and biochemical responses such as oxidative stress, acetylcholinesterase inhibition, and DNA damage, as well as cellular and histopathological alterations like lysosomal membrane stability, hepatic lesions, and micronucleus formation. Additionally, physiological and behavioral changes including reproductive output, swimming behavior, and heart rate are assessed, along with population-level indicators such as sex ratio. These assays frequently employ model organisms like zebrafish (Danio rerio) and water fleas (Daphnia magna) for laboratory testing and Gammarus sp., Dreissena sp., or sticklebacks for field monitoring, with results typically expressed as percentage effect relative to acute endpoints (for further details see Schuijt et al., 2021).
One of the primary strengths of both in vitro and in vivo techniques is their capacity to identify key toxicity mechanisms influenced by chemicals in environmental samples. When conducted with water extracts, these techniques primarily capture the effects of retained organic chemicals. Conversely, when water is directly dosed, they can reflect the combined impacts of organic contaminants, inorganic contaminants, and organic matter, as well as the influence of the sample’s physico-chemical properties (e.g., pH or dissolved oxygen levels), which may contribute to the mixture’s overall toxicity. These techniques are highly effective for detecting physiological damage at the cellular or individual level in sentinel organisms, providing valuable insights into cumulative stress responses. However, their ability to predict broader population- or ecosystem-level impacts remains limited.
Ecological monitoring tools
Ecological monitoring typically involves collecting biological samples to characterize aquatic ecosystems’ ecological status over time. This approach enables assessment of responses to both chemical and non-chemical stressors, which can be jointly monitored (Fig. 1). Biological samples are used to calculate integrated ecological indices that account for ecosystems disturbances at various trophic levels, e.g., primary producers, primary and secondary consumers, or predators. These indices are integral components of large-scale monitoring programs, serving to assess the ecological status of water bodies as part of the European WFD (Birk et al. 2012). Diverse metrics have been established in many of the EU member states, whose results have been compared in laborious intercalibration studies to determine reference conditions and ecological status classes for different water body types (Furse et al. 2009). These include, for example, the more general taxonomic indices for macroinvertebrates, such as diversity, richness or evenness, and the Saprobic Index (Rolauffs et al., 2004), the % of Ephemeroptera, Plecoptera, Trichoptera (%EPT) index (Weber, 1973) or the Biological Monitoring Working Party (BMWP) index (Hawkes 1997), which indicate general degradation compared to a reference situation. One of the main advantages of these methods is that they attain a high level of ecological realism, thus providing an integrated measure of community differences with undisturbed systems due to natural or anthropogenic perturbations. Moreover, such indices can be harmonized into Ecological Quality Ratios, which are used for European-level assessments of aquatic ecosystems (Solheim et al. 2025). However, the use of these ecological indices for the diagnostic risk assessment of potentially toxic chemicals is challenging, mainly due to their correlation with a wide range of non-chemical stressors, including organic matter pollution, nutrients, habitat modification or hydrological alterations (Schuwirth et al. 2015; Rico et al. 2016; Liess et al. 2021). This correlation has led to discussion on whether chemical-specific indices (i.e., indices that unequivocally determine the effects caused by chemicals with specific toxic mode of action) could be developed, with limited success so far (Schuwirth et al. 2015).
During the last decade, there has been a plea for the development of ecological indices based on biological traits, as these allow clearer links to be established between chemical effects and functional ecosystem responses, facilitating comparison of community responses across water bodies and (eco)regions (Menezes et al. 2010; Aazami et al. 2015). Several trait-based approaches exist, such as those for insecticides (Rico and Van den Brink, 2015) and salinity (Schäfer et al., 2011). Among these, the SPEAR index is notable for incorporating species’ relative sensitivity to pesticides and specific traits related to the probability of exposure and recolonize/recover from contaminant pulses (Liess and von der Ohe, 2005; Von der Ohe and Goedkoop 2013). The SPEAR index has been used in several research studies in Europe and other continents (see Hunt et al. 2017 and references therein) to evaluate the effects of pesticides, but the widespread inclusion of such trait-based indices into regulatory risk assessment has been limited so far.
Ecological monitoring research has also evolved to incorporate molecular tools, such as environmental DNA (eDNA) metabarcoding techniques, to estimate species diversity and assess chemical impacts on ecosystem functions (Zhang, 2019). A key advantage of these methods is their non-lethal nature for certain monitored organisms, minimizing habitat disruption while enabling assessments of diversity across a wide taxonomic spectrum, including rare species often overlooked by conventional sampling methods (Deiner et al., 2016). Over the past decade, there has been a growing effort to adapt indices used in the WFD to incorporate eDNA monitoring data for fish and other biological quality elements (Hering et al., 2018; Pont et al., 2021). However, the use of eDNA monitoring to assess chemical risks—particularly to establish connections between exposure and adverse effects across different levels of biodiversity—remains relatively limited (Schuijt et al., 2024). Moreover, linking chemical exposure to population- or community-level effects using eDNA data requires robust statistical correlations or the development of calculated indices.
The establishment of environmental specimen banks and biobanking programs (based on gene diversity studies) represent a valuable approach to collecting high-quality biological samples in a standardized manner through long-term ecological monitoring initiatives (Garmendia et al., 2015). Cryopreserved specimen banks facilitate spatial and temporal comparisons between historical and present samples with chemical datasets, providing insights into the historical evolution of pollutant effects or abrupt (genetic) diversity changes caused by the environmental release or the ban of certain contaminant classes. Ideally, these efforts require a thorough evaluation of the chemical fingerprint at the sites where the specimens and biobanking samples are collected or, at the very least, robust knowledge of nearby contamination sources. To date, specimen banks have been successfully implemented in countries like Germany to monitor changes in chemical bioaccumulation patterns across various aquatic organisms and pollutants (e.g., https://www.umweltprobenbank.de/en/). Meanwhile, biobanking has primarily been applied within conservation biology. The integration of data from both specimen banks and biobanking, alongside other risk assessment tools, holds significant promise for inferring long-term chemical effects on aquatic biodiversity and environmental health.
Performance evaluation of diagnostic tools
The tools outlined above have been utilized to identify chemicals of concern and ecosystem components particularly susceptible to damage (in a general sense) based on measured chemical exposure levels. Despite their inherent differences, these tools are expected to yield complementary results, highlighting both the chemicals and biological components at risk. Several studies have attempted to compare the outcomes produced by different diagnostic tools and to identify potential sources of discrepancies through evaluation exercises (hereafter referred to as evaluations or evaluation studies).
Such evaluations have often focused on comparing toxic pressure tools and ecological monitoring outcomes. For instance, field monitoring data has been employed to correlate TUs with various taxonomic indices, such as the BMWP, the Saprobic Index, and the %EPT. Significant negative correlations were observed, though the R2 values were generally modest, typically not exceeding 0.3 in large rivers like the Danube (e.g., Rico et al., 2016), and values of 0.4 (Liess et al., 2021) and 0.7 (Liess and von der Ohe 2005) in streams and small rivers. Moreover, the sum of pesticide TUs was found to correlate with species richness derived from rarefaction curves, suggesting that current pesticide mixtures may be responsible for up to a 42% decline in species richness (Beketov et al., 2013). Toxic pressure units have also been extensively assessed against trait-based indices, particularly the SPEARpesticides index, showing significant correlations across various water bodies (Beketov et al., 2013; Liess et al., 2021). Schäfer et al. (2013) evaluated the predictive capacity of four toxic pressure units (TUs based on Daphnia magna, TUs for the most sensitive species, TUs based on SSD HC5, and the msPAF approach) against the SPEARpesticides index using samples from five European countries. Their findings indicated that that either TUs based on HC5 values (in Australia), TUs based on the most sensitive species (Denmark, France and Germany), or the msPAF values (Spain) performed best, depending on the location.
PAF and msPAF calculations have also been compared with field-measured changes in species richness and other biological indices. Carafa et al. (2011) reported significant negative correlations between msPAF values and biotic indices for macroinvertebrates (BMWP) and diatoms (IPS) in Spanish rivers. However, other studies investigating similar relationships with fish richness demonstrated weaker correspondence, likely due to confounding factors such as river flow connectivity and predominant non-chemical stressors (Posthuma and de Zwart, 2006). Posthuma and de Zwart (2012) investigated the relationship between msPAF-EC50 values (calculated for 45 compounds using laboratory acute EC50 data) and macroinvertebrate abundance data obtained from ditches and streams in the Netherlands. Their study involved modeling population abundance declines for each taxon relative to reference conditions, which are often unavailable or influenced by multiple abiotic variables. Using Generalized Linear Models, they predicted population abundance under both chemical stress and non-chemical stress conditions. Their findings showed that the fraction of taxa exhibiting a 50% or greater abundance reduction corresponded well with msPAF-EC50 predictions for the most sensitive species. However, msPAF tended to underestimate population declines for less sensitive taxa, possibly due to species interactions. Additionally, some taxa exhibited population increases, which were not predicted by the msPAF approach, being identified as a major source of discrepancy between both methods. Posthuma et al. (2020) identified a strong correlation between msPAF-EC50 values (95th percentile) for 24 priority substances monitored in European rivers and their ecological status. They proposed msPAF-EC50 thresholds for assessing the ecological status of rivers, in line with the classification system established by the Water Framework Directive (WFD). More recently, Oginah et al. (2025) quantified toxic pressure levels at over 1,000 sites in the Netherlands using the msPAF approach and compared these values to macroinvertebrate abundance and species richness loss. Their findings revealed that macroinvertebrate abundance and richness generally decrease with increasing toxic pressure, with a nearly 1:1 relationship observed between msPAF-calculated values and the potentially disappeared fraction of species.
TUs and msPAF results have also been compared with outcomes from micro- and mesocosm experiments. TUs calculated with toxicity data for standard test species show strong correlations with observed effects in model ecosystem experiments involving pesticides (Brock et al., 2000a, 2000b). Furthermore, data from these experiments have been used to assess the protectiveness of the Hazard Concentration 5% (HC5) values derived from SSDs built with laboratory toxicity data. In most cases, these HC5 values were between 1.1 and 4 times lower than the lowest population NOEC obtained from micro-/mesocosm experiments (Maltby et al., 2005, 2009; Van den Brink et al., 2006b; Del Signore et al., 2016). Studies validating the SSD approach with model ecosystem experiments have generally focused on threshold (HC5) values rather than the entire SSD curve (Hose and Van den Brink, 2004). Rico et al. (2018) compared SSDs built with chronic laboratory toxicity data for the insecticide imidacloprid and a mixture of five neonicotinoid insecticides with the fraction of species significantly affected in a mesocosm experiment. They found a very good agreement between the predictions made with the PAF and msPAF values obtained from the SSDs and the observed effects in the mesocosm experiment at different toxic stress levels, supporting the use of such predictive tools for determining direct toxic effects in (semi-)natural species assemblages.
Models like the PERPEST model, which leverage data from micro- and mesocosm experiments, enable the calculation of chemical (or mixture) effects on structural and functional ecosystem parameters. However, comparisons between PERPEST predictions and other toxic pressure tools, such as TUs or msPAF indices, remain unexplored. Such comparisons would be highly valuable, as toxic pressure tools primarily focus on direct chemical effects, whereas the PERPEST model accounts for both direct and indirect effects. Moreover, these approaches rely on fundamentally different datasets—laboratory toxicity tests for toxic pressure tools versus outdoor mesocosm experiments for the PERPEST model. Consequently, any observed alignment between the two would significantly bolster confidence in diagnostic assessments and improve their reliability.
The integration of in vivo and in vitro bioassay monitoring data with ecological monitoring data or toxic pressure metrics remains a relatively unexplored field. In vitro assay results are often compared directly with EBT values, which are typically derived from environmental quality standards by incorporating relative potency estimates and mixture effects (Escher et al., 2018; Neale et al. 2023). Within this framework, in vitro bioassays are frequently regarded as analytical tools, used to identify complex chemical mixtures and link them to chemical analytical results through the application of mixture models. For certain endpoints, such as hormone-mimicking effects measured using estrogenicity assays, nearly 100% of the observed mixture effects can be attributed to detected estrogenic chemicals (Könemann et al., 2018). Similarly, photosynthesis inhibition is largely driven by known herbicides (Glauch and Escher, 2020). However, the explanatory power diminishes significantly for less specific toxicity pathways that are further downstream from molecular initiating events. Despite the detection of dozens of chemicals in surface water or environmental samples, the modeled and measured mixture effects of these detected chemicals often explain less than 1% of the observed effects for such endpoints (Neale et al., 2020). This apparent discrepancy may initially seem surprising but becomes more understandable when considering that a typical non-target chromatogram contains tens of thousands of peaks, many of which represent unknown chemicals likely contributing to the observed mixture effects (Escher et al., 2020). A recent study investigating the presence of 225 chemicals in Dutch surface waters found that in vitro bioassay results could be directly linked to detected chemicals in only 6% of cases. Furthermore, the detected chemicals accounted for just 1–17% of the observed effects, suggesting that the majority of the effects were likely caused by undetected chemicals (Boonstra et al., 2025).
Multiple diagnostic tools applied to German streams
This section demonstrates how various diagnostic tools can be utilized to assess the ecological risks posed by realistic contaminant mixtures. The analysis is based on a dataset that includes invertebrate monitoring data, in vitro bioassay results, and pesticide concentrations measured in German streams (obtained from Neale et al., 2020; Liess et al., 2021). The subsequent sub-sections provide a detailed description of the dataset, outline the diagnostic indicators calculated, and explain the methods employed to enable a comparative assessment of different diagnostic tools.
The dataset
The dataset encompasses 56 sampling sites randomly selected from those monitored by Liess et al. (2021). Sampling was conducted between April and June in 2018 and 2019, using event-driven sampling (i.e., automatic sampling triggered by runoff events), which showed a higher number of substances as compared to samples taken during fair-weather conditions (Neale et al., 2020). Sites with certain wastewater treatment plant (WWTP) influence were excluded, to primarily focus on effects caused by pesticides and exclude potential indirect effects caused by ammonia and other organic decomposition by-products. Pesticide and metabolite analysis involved 108 substances, detected using UHPLC-MS/MS with detection limits in the nanograms per liter (ng/L) range. Macroinvertebrate monitoring was conducted during June of 2018 or 2019 using the multi habitat sampling method. Macroinvertebrate samples were identified to the lowest taxonomic resolution possible. Detailed methodologies for pesticide and macroinvertebrate sampling and analysis can be found in Liess et al. (2021). Water samples were also used for in vitro bioassays targeting endpoints such as cytotoxicity, estrogen receptor activation (ERα), aryl hydrocarbon receptor activation (AhR), peroxisome proliferator-activated receptor activation (PPARγ), and oxidative stress response (AREc32). The methods used for these in vitro tests are described in detail in Neale et al. (2020).
Data analysis
The following diagnostic tools and indices were calculated from the dataset:
Toxic pressure assessment
The maximum pesticide concentrations measured at each sampling site were used by Liess et al. (2021) to calculate the log(TUmax) (hereafter TUmax) based on LC50 toxicity values for Daphnia magna or Chironomus sp. Furthermore, these maximum pesticide concentrations were used to calculate acute and chronic msPAF values for aquatic organisms, following the SSD parameters and calculation methods (assuming concentration addition) outlined by Posthuma et al. (2019), and implemented in the Key Factor Toxicity tool (STOWA, 2025). Additionally, the probability of observing significant effects on macroinvertebrates based on the pesticide concentrations at the sampling sites was calculated using the PERPEST model. Effects were defined as the sum of the probabilities of observing effect classes 2 and 3. These correspond to slight effects (partial reductions in population abundance on individual sampling days) and clear effects (severe reductions in population abundance over several consecutive sampling days), as documented in the micro- and mesocosm experiments included in the PERPEST database (Van den Brink et al., 2002). The PERPEST model calculations were performed for pesticides with a maximum hazard unit (the ratio of the maximum measured concentration to the hazardous concentration for 50% of organisms (HC50) obtained from the compound´s SSD) exceeding 0.002, while the rest were excluded from the PERPEST calculations as were assumed not to contribute to the calculated toxic pressure. The 0.002 threshold was derived by applying two extrapolation factors: A factor of 5 to extrapolate from HC50 to HC01 (based on the commonly reported SSD slope of 0.7 by Posthuma et al., 2019), and a factor of 100 to account for the additive toxic pressure from the approximate 100 chemicals included in the analysis.
In vitro assessment
The effects detected with four reporter gene assays were expressed in effect concentrations triggering 10% of maximum effect for the activation of the arylhydrocarbon receptor (AhR), the estrogen receptor (ERα), the peroxisome-activated receptor (PPARy) and the activation of the oxidative stress response 50% over the control ECIR1.5 (AREc32). The cytotoxicity IC10 (i.e., inhibitory concentration 10%) measured for each of the four in vitro bioassays (AhR, Erα, PPARy and AREc32) by Neale et al. (2020) in the water samples from each site was transformed into in vitro TUs (TU = 1/IC10). Subsequently, the average of the four in vitro TUs was calculated and used here as an integrated measure of in vitro cytotoxicity.
Ecological monitoring
The macroinvertebrate measurements (individuals/m2) were used here to calculate four common biological indices: total abundance, species richness, diversity (Shannon), and evenness, using the CANOCO software version 5 (ter Braak and Šmilauer 2012). Furthermore, SPEAR values were obtained from the calculations done by Liess et al. (2021).
Finally, a Pearson correlation analysis was done to assess the relationship between the toxic pressure tools employed here (TUmax, msPAF acute, msPAF chronic and the probability of finding effects as calculated with PERPEST), the in vitro cytotoxicity indicator (in vitro TUs) and the ecological monitoring indicators (abundance, richness, diversity, evenness and the SPEAR index). The Pearson correlation analysis and graphical interpretation was done with the SR plot software (Tang et al. 2023). The correspondence between the different tools and indicators was evaluated based on the significance of the Pearson correlation coefficients. Significant correlations were plotted using linear regression models built with Microsoft EXCEL (Microsoft Corporation 2024). The pesticide and invertebrate monitoring data used in this study, together with the in vitro test outcomes and the calculated indices are provided in the Supporting Information file.
Study outcomes
Figure 2 presents the Pearson correlations between the various toxic pressure metrics, in vitro bioassay results, and ecological monitoring indices. Significant correlations were detected between the TUmax and other indicators. The strongest was a negative relationship with the SPEARpesticides index (Pearson r = −0.40; p = 0.002; R² = 0.16;), indicating that higher TUmax values correspond to lower abundance of pollution-sensitive invertebrate species. TUmax also positively correlated with msPAF acute and chronic (Pearson r = 0.28 and 0.29; p = 0.04; R² = 0.08 for both), implying that elevated TUmax increases the percentage of potentially affected species calculated with SSDs. Interestingly, TUmax showed a negative correlation with in vitro TUs (Pearson r = −0.31; p = 0.02; R² = 0.10), suggesting that higher toxic pressure in standard macroinvertebrate tests aligns with greater stress in bioassays (note that in vitro TUs represent 1/IC10 values; Fig. 3). Among the in vitro results, ERα activation was notably high, alongside markers of untreated wastewater contamination. This points to inputs from human waste—likely via road runoff or discharges from small urban areas with limited sewage treatment—despite the streams being selected for agricultural influence. Other bioassay endpoints showed activation, though mixture effects generally remained below established effect-based thresholds (EBTs). TUmax was not significantly correlated with trait independent measures of the invertebrate community structure.
Fig. 2
Pearson correlations between toxic pressure (i.e., log(TUmax), msPAFacute, msPAFchronic, PERPEST probability of effects), in vitro toxicity (i.e., in vitro TUs, expressed as 1/IC10 values) and ecological monitoring indices (i.e., abundance, richness, Shannon diversity, evenness, SPEARpesticides) calculated for pesticides and macroinvertebrates monitored in German streams based on the data provided by Liess et al. (2021) and Neale et al. (2020). The circles show the size and sign of the Pearson correlation coefficients (blue = positive, red = negative). Significant correlations (p-value < 0.05) are marked with a star inside of the circle. The dataset used for the Pearson correlation analysis is provided in the Supporting Information file
Linear regression models showing the relationship between (A) the msPAFacute and the Shannon diversity index, (B) the msPAFacute and the Evenness index, (C) the logTUmax and the in vitro TU, and (D) the logTUmax and the SPEAR index
A significant positive correlation was observed between msPAF (acute and chronic) and the probability of macroinvertebrate effects predicted by the PERPEST model (Pearson r = 0.58 and 0.68, respectively; p < 0.001; R² = 0.34 and 0.46). This relationship may stem from the PERPEST model’s reliance on HC50 values derived from SSDs for the calculation of pesticide hazard units for querying its micro-/mesocosm database. Additionally, msPAF acute exhibited a significant negative correlation with species evenness (Pearson r = −0.28; p = 0.03; R² = 0.09;), suggesting that exceeding the 5% species protection threshold (msPAF acute) reduces the prevalence of rare and potentially sensitive taxa (Fig. 3).
The results of this case study demonstrate that, among the toxic pressure indicators evaluated, the TU approach performs best, as it effectively predicts both bioassay outcomes and biological monitoring indicators such as SPEARpesticides. However, the overall correlation coefficients were relatively low, suggesting that confounding factors influence each of these tools, limiting their ability to perfectly align. Additionally, none of the indicators showed a statistically significant correlation with classical invertebrate biodiversity metrics, such as the Shannon index and taxonomic richness. Nevertheless, while the correlations were not significant, the msPAF approach (acute and chronic) and the PERPEST model exhibited the strongest predictive capacity among the methods tested on such biodiversity indices (Figs. 2 and 3).
Uncertainties and recommendations to improve the diagnostic framework
The evaluation studies (discussed above) and the findings of the case study on German streams highlight significant data gaps and inconsistencies in the implementation and correlation between different diagnostic tools, as summarized in Fig. 4. This section outlines key data gaps and methodological advancements necessary for achieving better alignment between the results of different diagnostic tools.
Fig. 4
Standing questions pinpointing to the potential misalignments between the results obtained from various diagnostic tools used in chemical risk assessment
Discrepancies between toxic pressure assessments and biomonitoring outcomes often arise from inadequate toxicity data to represent the species present in the ecosystem under investigation. This limitation frequently stems from the reliance on a small set of standard test species for TU calculations or from the lack of toxicity data required to construct SSDs. Promising advancements include the use of machine learning algorithms that leverage chemical and biological predictors (e.g., taxonomy, physiology, or traits) to estimate the sensitivity of untested (non-standard test) species (Zubrod et al., 2024) and to expand species coverage in SSDs (Liang et al., 2024), thereby improving the robustness of toxic pressure assessments.
Another limitation of toxic pressure assessments is their focus on the measurable fraction of the exposome, excluding numerous unknown chemical compounds (Boonstra et al., 2025). For instance, a nationwide study in Germany assessed 464 pesticide compounds and found that additive mixture effects increased aquatic risk indicators (e.g., for invertebrates and algae) by 3.2 times under realistic worst-case scenarios compared to single-pesticide predictions (Weisner et al., 2021). These findings emphasize the importance of incorporating as many chemicals as possible in toxic pressure calculations and potentially investigating the need for a Mixture Assessment Factor (Backhaus, 2023), as done in prospective risk assessment, when chemical data are limited or unrepresentative of the full exposome. The Mixture Assessment Factor was designed for managing risks from unknown unintentional mixtures in prospective risk assessment, but an adapted version of it could be used when only a limited number of substances is measured in diagnostic risk assessments.
Toxic pressure assessments typically rely on concentrations from single grab water samples, while biomonitoring outcomes often reflect cumulative exposures over extended periods and across multiple exposure matrices, such as sediments and biota (Ijzerman et al., 2024). To address this, toxic pressure calculations should incorporate data from integrative multi-species, multi-compartment approaches (Miller et al., 2021; Manjarrés-López et al., 2025) using passive sampling devices (Shaw and Mueller, 2009). This could provide a more realistic estimation of long-term toxic pressure on aquatic populations and communities.
Furthermore, toxic pressure assessments predominantly focus on chemical stressors, yet aquatic communities are simultaneously exposed to a range of non-chemical stressors, including altered water flow, salinity, temperature extremes, or biological invasions (Birk et al., 2020). While the integration of non-chemical stressors into toxic pressure assessments remains limited, progress has been made. For example, metal toxicity data can be adjusted for salinity and pH gradients to calculate TUs, and temperature-dependent SSDs could be derived for some chemicals, either based on experimental data (Wang et al., 2019) or extrapolated toxicity data using temperature-dependent toxicokinetic-toxicodynamic models (Mangold-Döring et al. 2022).
Environmental toxicants, such as pesticides, impose strong selective pressures on various species. While the evolution of pesticide resistance in agricultural fields is well-documented, evidence of adaptation in non-target species exposed to these toxicants remains inconsistent (Bundschuh et al., 2023). Some studies suggest that species diversity plays a critical role in the development of toxicant tolerance by modulating the balance between intra- and interspecific competition (Becker and Liess, 2017). Multispecies models offer a promising complementary approach for quantifying the net effects of interspecific interactions on ecosystem responses to chemical stress, but initial efforts reveal their limited capacity to replicate community-level responses observed under (semi-)field conditions (Loerracher et al., 2023). Future research should focus on enhancing the ecological realism of model predictions by incorporating factors such as microevolutionary adaptation, internal and external population recovery pathways, and species’ adaptive behaviors, which collectively enhance community resilience to chemical pollution.
Bioassays vs biomonitoring
The limited alignment between bioassay results and ecological monitoring outcomes can be attributed to several factors. A primary challenge lies in extrapolating in vitro effects to in vivo outcomes, which stems from multiple issues. First, chemical bioavailability and toxicokinetics in in vitro bioassays often fail to account for delayed or cumulative effects, limiting their ability to predict certain apical responses (Yoon et al., 2012; Grech et al., 2019). Second, there is a scarcity of in vitro bioassays that adequately represent the diversity of aquatic organisms. Many currently available assays were developed for human toxicology and focus on conserved vertebrate receptors, such as those found in fish, while largely neglecting other taxa like invertebrates and primary producers. Third, the in vitro-to-in vivo extrapolation models are primarily designed for human toxicology, creating significant gaps when applied to aquatic organisms (Villeneuve et al., 2019; Brinkmann et al., 2016; Wang et al., 2022). To address these limitations, research efforts should focus on expanding the in vitro toxicity testing battery to include a broader range of receptors characteristic of freshwater ecosystems (beyond vertebrate taxa). Additionally, improved methods for extrapolating in vitro responses to in vivo apical effects are needed.
Moreover, our understanding of how in vitro results translate to population and community-level outcomes remains limited. The Adverse Outcome Pathway (AOP) framework provides a valuable approach for describing how chemical exposure and in vitro effects propagate across different levels of biological organization (Ankley et al., 2024). In the past decade, significant progress has been made in describing molecular initiating events and key events and linking them to ecologically relevant endpoints such as survival, growth, and reproduction (e.g. Baldwin et al., 2009). However, AOPs have predominantly focused on model organisms, limiting their applicability across a broader range of species. There is a pressing need to develop knowledgebases and computational tools to enhance inter-chemical and inter-taxa extrapolation and to support quantitative predictions of adverse effects at higher levels of biological organization (Kramer et al., 2011; Ankley et al., 2024).
Another complicating factor is the potential for correlated responses of in vitro results to other stressors present in environmental samples. Field studies have revealed significant natural variability in the enzymatic profiles of aquatic organisms (e.g., acetylcholinesterase, catalase, glutathione-S-transferase) driven by environmental factors such as temperature, salinity, and other physicochemical variables (Menezes et al., 2006; Pfeifer et al., 2005; Ippolito et al., 2017). Understanding this variability is essential for disentangling chemical-induced effects from those arising due to (natural) environmental influences.
Toxic pressure vs bioassays
Discrepancies between toxic pressure assessments and bioassay results may stem from the presence of undetected toxic chemicals in environmental samples, which are excluded from toxic pressure calculations due to analytical limitations. In such cases, employing EDA techniques to reduce sample complexity through fractionation can aid in identifying toxicologically relevant compounds (Escher et al., 2020). Furthermore, combining suspect screening and non-target screening approaches allows for the identification of compounds without the need for reference standards, thereby broadening the scope of chemical analysis (Hollender et al., 2023).
Another critical factor is the appropriate selection of species and toxicological endpoints for toxic pressure calculations. Bioassays produce results that are specific to certain species or taxonomic groups; therefore, toxic pressure assessments must align with these parameters to enable accurate comparisons. For instance, bioassay results derived from fish or mammalian cell lines may not correlate well with toxic units TUs or msPAF calculated from datasets that include aquatic invertebrates or plants. This misalignment likely contributed to the weak correlation observed in the German streams case study.
Even when taxonomic alignment is achieved, false negatives may still occur if the receptor specificity of the bioassay does not encompass the broader range of receptors present in the model species. As discussed earlier, addressing these gaps requires the integration of additional receptors through alternative bioassays, enhancing the robustness and reliability of toxic pressure evaluations.
Towards an integrative framework
Each group of diagnostic tools described above may be represented by different (groups of) indices, constituting multiple criteria to assess chemical impacts on ecosystems. The integration of these results to provide robust risk conclusions and to define management decisions has been debated under the multiple Lines of Evidence (LOEs) approach (US EPA, 1998; EFSA, 2021). This approach assumes that combining multiple independent lines reduces uncertainty and increases confidence in risk estimates, so that combinations of such lines of evidence are required to achieve a robust conclusion. The Weight-of-Evidence (WoE) approach has been dedicated to the integration of multiple LOEs based on qualitative methods, grounded on comparative tables and triangulation methods, or quantitative methods, using matrix-based approaches and Bayesian networks (Weed 2005; Linkov et al., 2009; Becker et al., 2022). Their final integration usually depends on multicriteria decisional analysis methods that require expert judgment for weighing the final diagnostic outcomes (e.g., Semenzin et al., 2008). Providing a unique method that accommodates all new diagnostic tools available for different sources of chemical ecosystem impairment and freshwater ecosystems is beyond the scope of this paper, but it seems that the individual uncertainties of some of the new diagnostic tools and indicators available would need to be reevaluated in this context. Figure 4 addresses this by proposing a qualitative framework to assess such uncertainties. Additionally, integrating emerging tools like AOPs, mixture toxicity models, and machine learning algorithms could refine uncertainty quantification and improve data synthesis for environmental decision-making (Kienzler et al., 2022; Bell et al., 2023).
Conclusions
Multiple diagnostic tools have been developed during the last decades to provide diverse lines of evidence for assessing ecological risks under field conditions. These tools are designed to address the intricate interactions and impacts of chemical stressors, offering a more nuanced perspective on ecosystem health. As highlighted in this paper, each method evaluates distinct aspects of ecosystems and can be seen as highly complementary, contributing to a comprehensive understanding of the (net) effects of chemical pollution on aquatic ecosystems. However, the limited number of studies that have attempted to compare the outcomes of these methods (including this one) reveal inconsistencies. These may arise from factors such as variations in the availability and quality of toxicity data used in toxic pressure assessments, unknown (or not assessed) biotic and abiotic stressors in the respective samples, disciplinary biases driven by differing study needs, or challenges in extrapolating findings across different levels of biological organization, among others. Understanding these limitations is critical for interpreting chemical risk assessment results and may facilitate the hierarchical and complementary application of these tools. The integration of these diagnostic tools with AOPs, effect models, and machine learning algorithms, that help reduce the uncertainties in extrapolations across ecosystem receptors, levels of biological organization and ecological scenarios is expected to enhance the reliability of ecological risk assessments in the future.
We thank the support provided by the NORMAN network for the organization of the Workshop on 'Improving the use of (semi-)field data for the risk assessment of chemicals´ in November 2022. AR thanks support from the Talented Researcher Support Programme - PlanGenT (CIDEGENT/2020/043) of the Generalitat Valenciana. S.J.M. was supported by The Research Council of Norway through the basic funding to NIVA (#342628). The conclusions expressed in this paper represent the expert judgement of the authors, but not necessarily the opinion of their affiliation.
Compliance with ethical standards
Conflict of interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The use of diagnostic tools to assess the risks of chemicals to freshwater ecosystems: towards a unified evaluation framework
Verfasst von
Andreu Rico
Udo Hommen
Beate I. Escher
Alina Koch
Anne Bado-Nilles
Belén González-Gaya
Enya Cody
Francisco Sylvester
Gabriele Treu
Gastón Alurralde
Henner Hollert
Iker Alvarez-Mora
S. Jannicke Moe
Joanne de Jonge
Kelsey Ng
Manu Soto
Matthias Liess
Melis Muz
Mirco Bundschuh
Naroa Lopez-Herguedas
Nicolas Pucheux
Nikiforos Alygizakis
Peter C. von der Ohe
Rémy Beaudouin
Saskia Finckh
Tobias Schulze
Yves Verhaegen
Paul J. van den Brink
Aazami J, Esmaili Sari A, Abdoli A, Sohrabi H, Van den Brink PJ (2015) Assessment of ecological quality of the Tajan river in Iran using a multimetric macroinvertebrate index and species traits. Environmental Manag 56:260–269.
Ankley GT, Berninger JP, Maloney EM, Olker JH, Schaupp CM, Villeneuve DL, LaLone CA (2024) Linking mechanistic effects of pharmaceuticals and personal care products to ecologically relevant outcomes: A decade of progress. Environmental Toxicol Chem 43(3):537–548.
Backhaus, T, 2023. Improving the regulatory assessment of combination effects: steps towards implementing the mixture assessment factor (MAF) in chemical regulation. Report PM 8/21 to the Swedish Chemicals Agency [online]
Baldwin DH, Spromberg JA, Collier TK, Scholz NL (2009) A fish of many scales: extrapolating sublethal pesticide exposures to the productivity of wild salmon populations. Ecolog Appl 19(8):2004–2015.
Barata C, Damasio J, López MA, Kuster M, De Alda ML, Barceló D, Riva MC, Raldúa D (2007) Combined use of biomarkers and in situ bioassays in Daphnia magna to monitor environmental hazards of pesticides in the field. Environmental Toxicol Chem: Int J 26(2):370–379.
Becker J, Liess M (2017) Species diversity hinders adaptation to toxicants. Environmental Sci Technol 51:10195–10202.
Becker RA, Ankley GT, Edwards SW, Kennedy SW, Linkov I, Meek B, Sachana M, Segner H, Van Der Burg B, Villeneuve DL, Watanabe H, Barton-Maclaren TS (2022) Improving the use of the weight of evidence approach in chemical assessments. Environmental Toxicol Chem 41(6):1355–1369.
Bell SM, Edwards SW, Ogese MO (2023) Integrating machine learning with weight of evidence for ecological risk prediction’. Environmental Sci Technol 57(12):4910–4920.
Birk S, Chapman D, Carvalho L, Spears BM, Andersen HE, Argillier C, Hering D (2020) Impacts of multiple stressors on freshwater biota across spatial scales and ecosystems. Nature Ecol Evolution 4(8):1060–1068.
Birk S, Bonne W, Borja A, Brucet S, Courrat A, Poikane S, Solimini A, Van De Bund W, Zampoukas N, Hering D (2012) Three hundred ways to assess Europe’s surface waters: an almost complete overview of biological methods to implement the Water Framework Directive. Ecological Indic 18:31–41.
Boonstra H, de Baat ML, van der Meer F, Besselink H, Roessink I, Kraak MHS (2025) Capturing temporal variation in aquatic ecotoxicological risks: Chemical-versus effect-based assessment. Science Total Environ 967:178797.
Brack W, Ait-Aissa S, Burgess RM, Busch W, Creusot N, Di Paolo C, Escher BI, Hewitt LM, Hilscherova K, Hollender J, Hollert H (2016) Effect-directed analysis supporting monitoring of aquatic environments—an in-depth overview. Science Total Environ 544:1073–1118.
Van den Brink PJ, Boxall AB, Maltby L, Brooks BW, Rudd MA, Backhaus T, Spurgeon D, Verougstraete V, Ajao C, Ankley GT, Apitz SE (2018) Toward sustainable environmental quality: Priority research questions for Europe. Environmental Toxicol Chem 37(9):2281–2295.
Brack W, Dulio V, Ågerstrand M, Allan I, Altenburger R, Brinkmann M, Bunke D, Burgess RM, Cousins I, Escher BI, Hernández FJ (2017) Towards the review of the European Union Water Framework Directive: recommendations for more efficient assessment and management of chemical contamination in European surface water resources. Science Total Environ 576:720–737.
Brinkmann M, Schlechtriem C, Reininghaus M, Eichbaum K, Buchinger S, Reifferscheid G, Hollert H, Preuss TG (2016) Cross-species extrapolation of uptake and disposition of neutral organic chemicals in fish using a multispecies physiologically-based toxicokinetic model framework. Environmental Sci Technol 50(4):1914–1923.
Brock, TCM, Lahr, J and Van den Brink, PJ (2000a) Ecological risks of pesticides in freshwater ecosystems; Part 1: herbicides (No. 88). Alterra.
Brock, TCM, Van Wijngaarden, RPA and Van Geest, GJ, (2000b) Ecological risks of pesticides in freshwater ecosystems; Part 2: Insecticides (No. 89). Alterra.
Bundschuh M, Mesquita-Joanes F, Rico A, Camacho A (2023) Understanding ecological complexity in a chemical stress context–a reflection on recolonisation, recovery and adaptation of aquatic populations and communities. Environmental Toxicol Chem 42(9):1857–1866.
Busch W, Schmidt S, Kühne R, Schulze T, Krauss M, Altenburger R (2016) Micropollutants in European rivers: A mode of action survey to support the development of effect-based tools for water monitoring. Environmental Toxicol Chem 35(8):1887–1899.
Carafa R, Faggiano L, Real M, Munn‚ A, Ginebreda A, Guasch H, Flo M, Tirapu L, von der Ohe PC (2011) Water toxicity assessment and spatial pollution patterns identification in a Mediterranean River Basin District. Tools for water management and risk analysis Sci. Total Environ 409(20):4269–4279
Chapman PM (2000) The Sediment Quality Triad: then, now and tomorrow. Environmental Sci Pollut Res 13:351–356.
De Zwart D, Posthuma L (2005) Complex mixture toxicity for single and multiple species: proposed methodologies. Environmental Toxicol Chem: Int J 24(10):2665–2676.
Deiner K, Fronhofer EA, Mächler E, Walser JC, Altermatt F (2016) Environmental DNA reveals that rivers are conveyer belts of biodiversity information. Nature Commun 7(1):12544.
Del Signore A, Hendriks AJ, Lenders HR, Leuven RS, Breure AM (2016) Development and application of the SSD approach in scientific case studies for ecological risk assessment. Environmental Toxicol Chem 35(9):2149–2161.
EC, European Commission, 2010. Directorate-General for Environment, Guidance on chemical monitoring of sediment and biota under the water framework directive. Guidance document N° 25, Publications Office. https://data.europa.eu/doi/10.2779/43586.
EC, European Commission, 2013. Directive 2013/39/EU of the European Parliament and of the Council of 12 August 2013 amending Directives 2000/60/EC and 2008/105/EC as regards priority substances in the field of water policy. Off. J. Eur. Union (24.8.2013, L226/1-17).
EFSA, European Food Safety Authority (2021) Guidance on the use of weight of evidence in scientific assessments’. EFSA J 19(7):6643. https://doi.org/10.2903/j.efsa.2021.6643 Available at.CrossRef
EFSA, European Food Safety Authority Panel on Plant Protection Products and their Residues (PPR) (2013) Guidance on tiered risk assessment for plant protection products for aquatic organisms in edge-of-field surface waters. EFSA J 11(7):3290.
Escher BI, Stapleton HM, Schymanski EL (2020) Tracking complex mixtures of chemicals in our changing environment. Science 367(6476):388–392.
Escher BI, Aїt-Aїssa S, Behnisch PA, Brack W, Brion F, Brouwer A, Buchinger S, Crawford SE, Du Pasquier D, Hamers T, Hettwer K (2018) Effect-based trigger values for in vitro and in vivo bioassays performed on surface water extracts supporting the environmental quality standards (EQS) of the European Water Framework Directive. Science Total Environ 628:748–765.
Furse, MT, Hering, D, Brabec, K, Buffagni, A, Sandin, L and Verdonschot, PF eds., 2009. The ecological status of European rivers: evaluation and intercalibration of assessment methods (Vol. 188). Springer Science & Business Media.
Ginebreda A, Muñoz I, de Alda ML, Brix R, López-Doval J, Barceló D (2010) Environmental risk assessment of pharmaceuticals in rivers: relationships between hazard indexes and aquatic macroinvertebrate diversity indexes in the Llobregat River (NE Spain). Environment Int 36:153–162.
Garmendia L, Izagirre U, Soto M, Lermen D, Koschorreck J (2015) Combining chemical and biological endpoints, a major challenge for twenty-first century’s environmental specimen banks. Environmental Sci Pollut Res 22:1631–1634.
Glauch L, Escher BI (2020) The combined algae test for the evaluation of mixture toxicity in environmental samples. Environmental Toxicol Chem 39:2496–2508.
Grech A, Tebby C, Brochot C, Lou Dorne F, Quignot N, Beaudouin R (2019) Generic physiologically-based toxicokinetic modelling for fish: Integration of environmental factors and species variability. Science Total Environ 651:516–531.
Hawkes HA (1997) Origin and development of the biological monitoring working party score system. Water Res 32:964–968.
Hering D, Borja A, Jones JI, Pont D, Boets P, Bouchez A, Bruce K, Drakare S, Hänfling B, Kahlert M, Leese F (2018) Implementation options for DNA-based identification into ecological status assessment under the European Water Framework Directive. Water Res 138:192–205.
Hollender J, Schymanski EL, Ahrens L et al. (2023) NORMAN guidance on suspect and non-target screening in environmental monitoring. Environmental Sci Eur 35: 75. https://doi.org/10.1186/s12302-023-00779-4.CrossRef
Hose GC, Van den Brink PJ (2004) Confirming the species-sensitivity distribution concept for endosulfan using laboratory, mesocosm, and field data. Archives Environ Contamination Toxicol 47:511–520.
Hunt L, Bonetto C, Marrochi N, Scalise A, Fanelli S, Liess M, Lydy MJ, Chiu MC, Resh VH (2017) Species at Risk (SPEAR) index indicates effects of insecticides on stream invertebrate communities in soy production regions of the Argentine Pampas. Science Total Environ 580:699–709.
Ijzerman MM, Raby M, Letwin NV, Black T, Kudla YM, Osborne RK et al. (2024) Pesticide presence in stream water, suspended sediment and biofilm is strongly linked to upstream catchment land use and crop type. Ecotoxicology Environ Saf 288:117382.
Ippolito A, Giacchini R, Parenti P, Vighi M (2017) Natural variability of enzymatic biomarkers in freshwater invertebrates. Environmental Sci Pollut Res 24:732–742.
Kienzler A, Bopp SK, van der Linden S, Berggren E, Worth A (2022) Weight of evidence for chemical mixtures: A review of current approaches and future directions. Critical Rev Toxicol 52(1):1–22. https://doi.org/10.1080/10408444.2022.2035313.CrossRef
Könemann S, Kase R, Simon E, Swart K, Buchinger S, Schlüssener M, Hollert H et al. (2018) Effect-based and chemical analytical methods to monitor estrogens under the European Water Framework Directive. Trends Anal Chem 102:225–235.
König M, Escher BI, Neale PA, Krauss M, Hilscherová K, Novák J, Teodorović I, Schulze T, Seidensticker S, Hashmi MAK, Ahlheim J (2017) Impact of untreated wastewater on a major European river evaluated with a combination of in vitro bioassays and chemical analysis. Environmental Pollut 220:1220–1230.
Kramer VJ, Etterson MA, Hecker M, Murphy CA, Roesijadi G, Spade DJ, Spromberg JA, Wang M, Ankley GT (2011) Adverse outcome pathways and ecological risk assessment: Bridging to population-level effects. Environmental Toxicol Chem 30(1):64–76.
Lee J, Khim JS (2022) Revisited a sediment quality triad approach in the Korean coastal waters: Past research, current status, and future directions. Environmental Pollut 292:118262.
Liang R, Sinclair TM, Craig PS, Maltby L (2024) Spatial variation in the sensitivity of freshwater macroinvertebrate assemblages to chemical stressors. Water Res 248:120854.
Liess M, Schulz R (1999) Linking insecticide contamination and population response in an agricultural stream Environmental Toxicol Chem 18(9):1948–1955. https://doi.org/10.1002/etc.5620180913CrossRef
Liess M, Ohe PCVD (2005) Analysing effects of pesticides on invertebrate communities in streams Environmental Toxicol Chem: Int J 24(4):954–965.
Liess M, Liebmann L, Vormeier P, Weisner O, Altenburger R, Borchardt D, Brack W, Chatzinotas A, Escher B, Foit K, Gunold R (2021) Pesticides are the dominant stressors for vulnerable insects in lowland streams. Water Res 201:117262.
Linkov I, Loney D, Cormier S, Satterstrom FK, Bridges T (2009) Weight-of-evidence evaluation in environmental assessment: review of qualitative and quantitative approaches. Science Total Environ 407:5199–5205.
Loerracher AK, Schmidt J, Ebke P, Schmolke A, Abi-Akar F, Galic N, Ashauer R (2023) Characterization of patterns and variability in the dynamics of outdoor aquatic mesocosms: exploring the capabilities and challenges in data supporting aquatic system models. Ecotoxicology 32(6):782–801.
Long ER, Chapman PM (1985) A sediment quality triad: measures of sediment contamination, toxicity and infaunal community composition in Puget Sound. Marine Pollut Bull 16(10):405–415.
Maltby L, Brock TC, Van den Brink PJ (2009) Fungicide risk assessment for aquatic ecosystems: importance of interspecific variation, toxic mode of action, and exposure regime. Environmental Sci Technol 43(19):7556–7563.
Maltby L, Blake N, Brock TC, Van den Brink PJ (2005) Insecticide species sensitivity distributions: importance of test species selection and relevance to aquatic ecosystems. Environmental Toxicol Chem: Int J 24(2):379–388.
Mangold-Döring A, Huang A, Van Nes EH, Focks A, Van den Brink PJ (2022) Explicit consideration of temperature improves predictions of toxicokinetic–toxicodynamic models for flupyradifurone and imidacloprid in Gammarus pulex. Environmental Sci Technol 56(22):15920–15929.
Manjarrés-López DP, Martínez-Megías C, Vitale D, Picó Y, Rico A, Pérez S (2025) Occurrence, bioaccumulation, and ecological risk assessment of contaminants of emerging concern using native and invasive species as biomonitors. Emerg Contami 11(3):100533.
Martinez-Haro M, Acevedo P, Pais-Costa AJ, Neto JM, Vieira LR, Ospina-Alvarez N, Taggart MA, Guilhermino L, Ribeiro R, Marques JC (2022) Ecotoxicological tools in support of the aims of the European Water Framework Directive: A step towards a more holistic ecosystem-based approach. Ecological Indic 145:109645.
Menezes S, Baird DJ, Soares AM (2010) Beyond taxonomy: a review of macroinvertebrate trait-based community descriptors as tools for freshwater biomonitoring. Journal Appl Ecol 47(4):711–719.
Menezes S, Soares AM, Guilhermino L, Peck MR (2006) Biomarker responses of the estuarine brown shrimp Crangon crangon L. to non-toxic stressors: temperature, salinity and handling stress effects. Journal Exp Mar Biol Ecol 335(1):114–122.
Mentzel S, Martínez-Megías C, Grung M, Rico A, Tollefsen KE, Van den Brink PJ, Moe SJ (2024) Using a bayesian network model to predict risk of pesticides on aquatic community endpoints in a rice field—A Southern European case study. Environmental Toxicol Chem 43(1):182–196.
Miller TH, Ng KT, Lamphiere A, Cameron TC, Bury NR, Barron LP (2021) Multicompartment and cross-species monitoring of contaminants of emerging concern in an estuarine habitat. Environmental Pollut 270:116300.
Munthe J, Brorström-Lundén E, Rahmberg M, Posthuma L, Altenburger R, Brack W, Bunke D, Engelen G, Gawlik BM, van Gils J, Herráez DL (2017) An expanded conceptual framework for solution-focused management of chemical pollution in European waters. Environmental Sci Eur 29:1–16.
Neale PA, Escher BI, de Baat ML, Enault J, Leusch FDL (2023) Effect-Based Trigger Values Are Essential for the Uptake of Effect-Based Methods in Water Safety Planning. Environmental Toxicol Chem 42:714–726.
Neale PA, Braun G, Brack W, Carmona E, Gunold R, König M, Krauss M, Liebmann L, Liess M, Link M, Schäfer RB (2020) Assessing the mixture effects in in vitro bioassays of chemicals occurring in small agricultural streams during rain events. Environmental Sci Technol 54(13):8280–8290.
Neale PA, Brack W, Ait-Aissa S, Busch W, Hollender J, Krauss M, Maillot-Maréchal E, Munz NA, Schlichting R, Schulze T, Vogler B, Escher B (2018) Solid-phase extraction as sample preparation of water samples for cell-based and other in vitro bioassays. Environmental Sci Process Impacts 20:493–504.
Nowell LH, Moran PW, Waite IR, Schmidt TS, Bradley PM, Mahler BJ, Van Metre PC (2024) Multiple lines of evidence point to pesticides as stressors affecting invertebrate communities in small streams in five United States regions. Science Total Environ 915:169634.
Ofogh ARE, Dorche EE, Birk S, Fathi P, Shahraki MZ, Bruder A (2024) Improving the performance of macroinvertebrate based multi-metric indices by incorporating functional traits and an index performance-driven approach. Science Total Environ 931:172850.
Oginah SA, Posthuma L, Slootweg J, Hauschild M, Fantke P (2025) Calibrating Predicted Mixture Toxic Pressure to Observed Biodiversity Loss in Aquatic Ecosystems. Global Change Biol 31(6):70305.
Pfeifer S, Schiedek D, Dippner JW (2005) Effect of temperature and salinity on acetylcholinesterase activity, a common pollution biomarker, in Mytilus sp. from the south-western Baltic Sea. Journal Exp Mar Biol Ecol 320(1):93–103.
Pont D, Valentini A, Rocle M, Maire A, Delaigue O, Jean P, Dejean T (2021) The future of fish-based ecological assessment of European rivers: from traditional EU Water Framework Directive compliant methods to eDNA metabarcoding-based approaches. Journal fish Biol 98(2):354–366.
Posthuma L, De Zwart D (2006) Predicted effects of toxicant mixtures are confirmed by changes in fish species assemblages in Ohio, USA, rivers. Environmental Toxicol Chem: Int J 25(4):1094–1105.
Posthuma L, de Zwart D (2012) Predicted mixture toxic pressure relates to observed fraction of benthic macrofauna species impacted by contaminant mixtures. Environmental Toxicol Chem 31(9):2175–2188.
Posthuma L, van Gils J, Zijp MC, van de Meent D, de Zwart D (2019) Species sensitivity distributions for use in environmental protection, assessment, and management of aquatic ecosystems for 12 386 chemicals. Environmental Toxicol Chem 38(4):905–917.
Posthuma L, Altenburger R, Backhaus T, Kortenkamp A, Müller C, Focks A, de Zwart D, Brack W (2019) Improved component-based methods for mixture risk assessment are key to characterise complex chemical pollution in surface waters. Environmental Sci Eur 31(1):1–11.
Posthuma L, Zijp MC, De Zwart D, Van de Meent D, Globevnik L, Koprivsek M, Focks A, Van Gils J, Birk S (2020) Chemical pollution imposes limitations to the ecological status of European surface waters. Scientific Rep. 10(1):14825.
Rämö RA, van den Brink PJ, Ruepert C, Castillo LE, Gunnarsson JS (2018) Environmental risk assessment of pesticides in the River Madre de Dios, Costa Rica using PERPEST, SSD, and msPAF models. Environmental Sci Pollut Res 25:13254–13269.
Rico A, Van den Brink PJ (2015) Evaluating aquatic invertebrate vulnerability to insecticides based on intrinsic sensitivity, biological traits, and toxic mode of action. Environmental Toxicol Chem 34(8):1907–1917.
Rico A, Van den Brink PJ, Leitner P, Graf W, Focks A (2016) Relative influence of chemical and non-chemical stressors on invertebrate communities: a case study in the Danube River. Science Total Environ 571:1370–1382.
Rico A, Dafouz R, Vighi M, Rodríguez-Gil JL, Daam MA (2021a) Use of postregistration monitoring data to evaluate the ecotoxicological risks of pesticides to surface waters: A case study with chlorpyrifos in the Iberian Peninsula. Environmental Toxicol Chem 40(2):500–512.
Rico A, Arenas-Sánchez A, Pasqualini J, García-Astillero A, Cherta L, Nozal L, Vighi M (2018) Effects of imidacloprid and a neonicotinoid mixture on aquatic invertebrate communities under Mediterranean conditions. Aquatic Toxicol 204:130–143.
Rico A, de Oliveira R, de Souza Nunes GS, Rizzi C, Villa S, López-Heras I, Vighi M, Waichman AV (2021b) Pharmaceuticals and other urban contaminants threaten Amazonian freshwater ecosystems. Environment Int 155:106702.
Rolauffs P, Stubauer I, Moog O, Zahrádková S, Brabec K (2004). Integration of the saprobic system into the European Union Water Framework Directive: case studies in Austria, Germany and Czech Republic. Integrated Assessment of Running Waters in Europe, pp. 285–298.
Rorije E, Wassenaar PNH, Slootweg J, van Leeuwen L, van Broekhuizen FA, Posthuma L (2022) Characterization of ecotoxicological risks from unintentional mixture exposures calculated from European freshwater monitoring data: forwarding prospective chemical risk management. Science Total Environ 822:153358.
Schäfer RB, Gerner N, Kefford BJ, Rasmussen JJ, Beketov MA, de Zwart D, Liess M, von der Ohe PC (2013) How to characterise chemical exposure to predict ecologic effects on aquatic communities?. Environmental Sci Technol 47(14):7996–8004.
Schäfer RB, Kefford BJ, Metzeling L, Liess M, Burgert S, Marchant R, Pettigrove V, Goonan P, Nugegoda D (2011) A trait database of stream invertebrates for the ecological risk assessment of single and combined effects of salinity and pesticides in South-East Australia. Science Total Environ 409:2055–2083.
Schneeweiss A, Juvigny-Khenafou NP, Osakpolor S, Scharmüller A, Scheu S, Schreiner VC, Ashauer R, Escher BI, Leese F, Schäfer RB (2023) Three perspectives on the prediction of chemical effects in ecosystems. Global Change Biol 29(1):21–40.
Scholz S, Nichols JW, Escher BI, Ankley GT, Altenburger R, Blackwell B, Brack W, Burkhard L, Collette TW, Doering JA, Ekman D (2022) The Eco-Exposome concept: Supporting an integrated assessment of mixtures of environmental chemicals. Environmental Toxicol Chem 41(1):30–45.
Schuijt LM, Peng FJ, van den Berg SJP, Dingemans MML, Van den Brink PJ (2021) (Eco)toxicological tests for assessing impacts of chemical stress to aquatic ecosystems: Facts, challenges, and future. Science Total Environ 795:148776.
Schuijt LM, van Smeden J, van Drimmelen CKE, Buijse LL, Wu D, Boerwinkel M-C, Belgers DJM, Matser AM, Roessink I, Heikamp-de Jong I, Beentjes KK, Trimbos KB, Smidt H, Van den Brink PJ (2024) Effects of antidepressant exposure on aquatic communities assessed by a combination of morphological identification, functional measurements, environmental DNA metabarcoding and bioassays. Chemosphere 349:140706.
Schuwirth N, Kattwinkel M, Stamm C (2015) How stressor specific are trait-based ecological indices for ecosystem management? Sci. Total Environ. 505:565–572
Semenzin E, Critto A, Carlon C, Marcomini A (2008) Development of a site-specific ecological risk assessment for contaminated sites: Part II. A multi-criteria based system for the selection of bioavailability assessment tools. Science Total Environ 406(1-2):385–395.
Shaw M, Mueller JF (2009) Time integrative passive sampling: how well do chemcatchers integrate fluctuating pollutant concentrations?. Environmental Sci Technol 43(5):1443–1448.
Solheim, AL, Thrane, JE, Mentzel, S, Moe, SJ, 2025. Harmonised biological indicators for rivers and lakes: Towards European assessment of temporal trends in ecological quality. Ecological Indicators, 171.
Tang D, Chen M, Huang X, Zhang G, Zeng L, Zhang G, Wu S, Wang Y (2023) SRplot: A free online platform for data visualization and graphing. PLoS One 9 18(11):e0294236.
Ter Braak, CJ and Šmilauer, P, 2012. Canoco reference manual and user’s guide: software for ordination, version 5.0.
Treu G, Schulze J, Galert W, Hassold E (2024) Regulatory and practical considerations on the implementation of a mixture allocation factor in REACH. Environmental Sci Eur 36(1):101.
US EPA, 1998. Environmental Protection Agency. Guidelines for ecological risk assessment. EPA/630/R-95/002F. Risk assessment forum, Washington, DC.
Van den Brink PJ, Brown CD, Dubus IG (2006a) Using the expert model PERPEST to translate measured and predicted pesticide exposure data into ecological risks. Ecological Model 191(1):106–117.
Van den Brink PJ, Blake N, Brock TC, Maltby L (2006b) Predictive value of species sensitivity distributions for effects of herbicides in freshwater ecosystems. Human Ecol Risk Assess 12(4):645–674.
Van den Brink PJ, Roelsma J, Van Nes EH, Scheffer M, Brock TC (2002) Perpest model, a case-based reasoning approach to predict ecological risks of pesticides. Environmental Toxicol Chem: Int J 21(11):2500–2506.
Van Gils J, Posthuma L, Cousins IT, Brack W, Altenburger R, Baveco H, Focks A, Greskowiak J, Kuhne R, Kutsarova S, Lindim C, Markus A, van de Meent D, Munthe J, Schueder R, Schuurmann G, Slobodnik J, de Zwart D, van Wezel A (2020) Computational material for analysis for thousands of chemicals of emerging concern in European waters. Journal Hazard Mater 397:122655.
Vijver MG, Hunting ER, Nederstigt TA, Tamis WL, van den Brink PJ, van Bodegom PM (2017) Postregistration monitoring of pesticides is urgently required to protect ecosystems. Environmental Toxicol Chem 36(4):860–865.
Villeneuve DL, Coady K, Escher BI, Mihaich E, Murphy C, Schlekat T, Garcia-Reyero N (2019) High throughput screening and environmental risk assessment – State of the science and emerging applications. Environmental Toxicol Chem 38:12–26.
von der Ohe PC, Goedkoop W (2013) Distinguishing the effects of habitat degradation and pesticide stress on benthic invertebrates using stressor-specific metrics Sci. Total Environ 444:480–490
Wang JQ, Nolte TM, Owen SF, Beaudouin R, Hendriks AJ, Ragas AMJ (2022) A generalized physiologically based kinetic model for fish for environmental risk assessment of pharmaceuticals. Environmental Sci Technol 56:6500–6510.
Wang Z, Lui GC, Burton GA, Leung KM (2019) Thermal extremes can intensify chemical toxicity to freshwater organisms and hence exacerbate their impact to the biological community Chemosphere 224:256–264
Weber CI Biological field and laboratory methods for measuring the quality of surface waters and effluents: Cincinnati. Ohio, U.S.: Environmental Protection Agency; 1973 [EPA-670/4-73-001].
Weed DL (2005) Weight of Evidence: A Review of Concept and Methods. Risk Anal 25(6):1545–1557.
Weisner O, Frische T, Liebmann L, Reemtsma T, Roß-Nickoll M, Schäfer RB, Schäffer A, Scholz-Starke B, Vormeier P, Knillmann S, Liess M (2021) Risk from pesticide mixtures – The gap between risk assessment and reality. Science Total Environ 796:149017. https://doi.org/10.1016/j.scitotenv.2021.149017.CrossRef
Wilkinson JL, Boxall AB, Kolpin DW, Leung KM, Lai RW, Galbán-Malagón C, Adell AD, Mondon J, Metian M, Marchant RA, Bouzas-Monroy A et al. (2022) Pharmaceutical pollution of the world’s rivers. Proceedings Natl Acad Sci 119(8):2113947119.
Wolf JF, Prosser RS, Champagne EJ, McCann KS (2020) Variation in response of laboratory-cultured freshwater macroinvertebrates to sediment from streams with differential exposure to agriculture. Water, Air, Soil Pollut 231:1–18.
Wolfram G, Höss S, Orendt C, Schmitt C, Adámek Z, Bandow N, De Deckere E (2012) Assessing the impact of chemical pollution on benthic invertebrates from three different European rivers using a weight-of-evidence approach. Science Total Environ 438:498–509.
Yoon M, Campbell JL, Andersen ME, Clewell HJ (2012) Quantitative in vitro to in vivo extrapolation of cell-based toxicity assay results. Critical Rev Toxicol 42(8):633–652.
Zhang X (2019) Environmental DNA shaping a new era of ecotoxicological research. Environmental Sci Technol 53(10):5605–5612.
Zubrod JP, Galic N, Vaugeois M, Dreier DA (2024) Bio-QSARs 2.0: Unlocking a new level of predictive power for machine learning-based ecotoxicity predictions by exploiting chemical and biological information. Environment Int 186:108607.
