Scientific Opinion on the state of the art of Toxicokinetic/Toxicodynamic (TKTD) effect models for regulatory risk assessment of pesticides for aquatic organisms

Towards a spatiotemporally explicit toxicokinetic-toxicodynamic model for earthworm toxicity

The aim of the environmental risk assessment of chemicals is the prevention of unacceptable adverse effects on the environment. Therefore, the risk assessment for in-soil organisms, such as earthworms, is based on two key elements: the exposure assessment and the effect assessment. In the current risk assessment scheme, these two elements are not linked. While for the exposure assessment, advanced exposure models can take the spatial and temporal scale of substances into account, the effect assessment in the lower tiers considers only a limited temporal and spatial variability. However, for soil organisms, such as earthworms, those scales play a significant role as species move through the soil in response to environmental factors. To overcome this gap, we propose a conceptual integration of pesticide exposure, ecology, and toxicological effects on earthworms using a modular modeling approach. An essential part of this modular approach is the environment module, which utilizes exposure models to provide spatially and temporally explicit information on environmental variables (e.g., temperature, moisture, organic matter content) and chemical concentrations. The behavior module uses this information and simulates the feeding and movement of different earthworm species using a trait-based approach. The resulting exposure can be processed by a toxicokinetic-toxicodynamic (TKTD) module. TKTD models are particularly suitable to make effect predictions for time-variable exposure situations as they include the processes of uptake, elimination, internal distribution, and biotransformation of chemicals and link the internal concentration to an effect at the organism level. The population module incorporates existing population models of different earthworm species. The modular approach is illustrated using a case study with an insecticide. Our results emphasize that using a modular model approach will facilitate the integration of exposure and effects and thus enhance the risk assessment of soil organisms.

Modeling Sublethal Effects of Chemicals: Application of a Simplified Dynamic Energy Budget Model to Standard Ecotoxicity Data

To assess ecological risks from chemical exposure, we need tools to extrapolate from the sublethal effects observed in the laboratory under constant exposure to realistic time-varying exposures. Dynamic energy budget (DEB) theory offers a mechanistic modeling approach to describe the entire life history of a single organism and the effects of toxicant exposure. We use a simplified model, which can be wholly calibrated from standard chronic bioassay data. Case studies on standard test organisms (Americamysis bahia and Pimephales promelas) are presented to demonstrate the calibration procedure, and for the second case, data are available to pseudovalidate model performance. We use these results to highlight gaps and shortcomings in the current state of the science, and we discuss how these can be overcome to maximize the potential of DEB theory in ecological risk assessment.

Linking Morphology, Toxicokinetic, and Toxicodynamic Traits of Aquatic Invertebrates to Pyrethroid Sensitivity

Pyrethroid insecticides are known to be highly toxic to most aquatic nontarget organisms, but little is known about the mechanisms causing some species to be highly sensitive while others are hardly affected by the pyrethroids. The aim of the present study was to measure the sensitivity (EC₅₀-values) of 10 aquatic invertebrates toward a 24 h pulse of the pyrethroid cypermethrin and subsequently test if the difference in sensitivity could be explained by measured morphological and physiological traits and modeled toxicokinetic (TK) and toxicodynamic (TD) parameters. Large differences were observed for the measured uptake and elimination kinetics, with bioconcentration factors (BCFs) ranging from 53 to 2337 at the end of the exposure. Similarly, large differences were observed for the TDs, and EC₅₀-values after 168 h varied 120-fold. Modeling the whole organism cypermethrin concentrations indicated compartmentation into a sorbed fraction and two internal fractions: a bioavailable and non-bioavailable internal fraction. Strong correlations between surface/volume area and the TK parameters (sorption and uptake rate constants and the resulting BCF) were found, but none of the TK parameters correlated with sensitivity. The only parameter consistently correlating with sensitivity across all species was the killing rate constant of the GUTS-RED-SD model (the reduced general unified threshold models of survival assuming stochastic death), indicating that sensitivity toward cypermethrin is more related to the TD parameters than to TK parameters.

Time-Variable Exposure Experiments in Conjunction with Higher Tier Population and Effect Modeling to Assess the Risk of Chlorotoluron to Green Algae

An algae population model was applied to describe measured effects of pulsed exposure to chlorotoluron on populations of Pseudokirchneriella subcapitata in 2 laboratory flow-through chemostat tests with different exposure regimes. Both tests enabled evaluation of adverse effects on algae during the exposure and population recovery afterward. Impacts on population densities after chlorotoluron exposure were directly visible as biomass loss in the chemostats. Recovery was observed after each exposure peak. The test results indicate that P. subcapitata is unlikely to show an increased sensitivity to chlorotoluron after pulsed exposure. No altered response or adaptation of the algae to chlorotoluron was observed, with the exception of the last high peak in flow-through test 2. Therefore, an adaptation to the test substance cannot be excluded after long-term exposure. However, recovery to the steady-state level after this peak indicates that the growth rate (fitness) was not significantly reduced in the population with higher tolerance. No differences in chlorotoluron impact on the populations over time in terms of growth were detected. Model predictions agreed well with the measured data. The tests and modeling results validate the model to simulate population dynamics of P. subcapitata after pulsed exposure to chlorotoluron. Model predictions and extrapolations with different exposure patterns are considered reliable for chlorotoluron. The good reproducibility of the population behavior in the test systems supports this conclusion. An example modeled extrapolation of the experimental results to other (untested) exposure scenarios shows a potential approach to using the validated model as a supportive tool in risk assessment. Environ Toxicol Chem 2019;38:2520-2534. © 2019 SETAC.

Recommendations to address uncertainties in environmental risk assessment using toxicokinetic-toxicodynamic models

Providing reliable environmental quality standards (EQSs) is a challenging issue in environmental risk assessment (ERA). These EQSs are derived from toxicity endpoints estimated from dose-response models to identify and characterize the environmental hazard of chemical compounds released by human activities. These toxicity endpoints include the classical x% effect/lethal concentrations at a specific time t (EC/LC(x, t)) and the new multiplication factors applied to environmental exposure profiles leading to x% effect reduction at a specific time t (MF(x, t), or denoted LP(x, t) by the EFSA). However, classical dose-response models used to estimate toxicity endpoints have some weaknesses, such as their dependency on observation time points, which are likely to differ between species (e.g., experiment duration). Furthermore, real-world exposure profiles are rarely constant over time, which makes the use of classical dose-response models difficult and may prevent the derivation of MF(x, t). When dealing with survival or immobility toxicity test data, these issues can be overcome with the use of the general unified threshold model of survival (GUTS), a toxicokinetic-toxicodynamic (TKTD) model that provides an explicit framework to analyse both time- and concentration-dependent data sets as well as obtain a mechanistic derivation of EC/LC(x, t) and MF(x, t) regardless of x and at any time t of interest. In ERA, the assessment of a risk is inherently built upon probability distributions, such that the next critical step is to characterize the uncertainties of toxicity endpoints and, consequently, those of EQSs. With this perspective, we investigated the use of a Bayesian framework to obtain the uncertainties from the calibration process and to propagate them to model predictions, including LC(x, t) and MF(x, t) derivations. We also explored the mathematical properties of LC(x, t) and MF(x, t) as well as the impact of different experimental designs to provide some recommendations for a robust derivation of toxicity endpoints leading to reliable EQSs: avoid computing LC(x, t) and MF(x, t) for extreme x values (0 or 100%), where uncertainty is maximal; compute MF(x, t) after a long period of time to take depuration time into account and test survival under pulses with different periods of time between them.

The impact of humic acid on toxicity of individual herbicides and their mixtures to aquatic macrophytes

This study investigates the impact of humic acid (HA) on the toxicity of selected herbicides and their binary mixtures to aquatic plants. The focus was on two auxin simulators (2,4-D and dicamba) and two photosynthetic inhibitors (atrazine and isoproturon). The results suggested that the addition of HA to the standard synthetic medium does not affect Lemna minor growth nor the toxicity of atrazine, but increases the toxicity of 2,4-D and the binary mixture of atrazine and 2,4-D. The addition of HA to the standard synthetic medium reversibly decreased the growth (biomass) of Myriophyllum aquaticum and enhanced the toxicity of individually tested herbicides (isoproturon and dicamba) as well as their binary mixture. The results showed delayed toxic effects of auxin simulators, especially 2,4-D in the Lemna test. The recovery after the exposure to individual photosystem II inhibitors (atrazine and isoproturon) is fast in both plant species, regardless of the presence of HA. In the case of selected mixtures (atrazine + 2,4-D and isoproturon + dicamba), recovery of both plant species was noted, while the efficiency depended on the herbicide concentration in the mixture rather than the presence or absence of HA.

Predictive models in ecotoxicology: Bridging the gap between scientific progress and regulatory applicability-Remarks and research needs

This paper concludes a special series of 7 articles (4 on toxicokinetic-toxicodynamic [TK-TD] models and 3 on quantitative structure-activity relationship [QSAR] models) published in previous issues of Integrated Environmental Assessment and Management (IEAM). The present paper summarizes the special series articles and highlights their contribution to the topic of increasing the regulatory applicability of effect models. For both TK-TD and QSAR approaches, we then describe the main research needs. The use of TK-TD models for describing sublethal effects must be better developed, particularly through the improvement of the dynamic energy budget (DEBtox) approach. The potential of TK-TD models for moving from lower (molecular) to higher (population) hierarchical levels is highlighted as a promising research line. Some relevant issues to improve the acceptance of QSAR models at the regulatory level are also described, such as increased transparency of the performance assessment and of the modeling algorithms, model documentation, relevance of the chosen target for regulatory needs, and improved mechanistic interpretability. Integr Environ Assess Manag 2019;00:000-000. © 2019 SETAC.

Modelling effects of time-variable exposure to the pyrethroid beta-cyfluthrin on rainbow trout early life stages

BACKGROUND

Available literature and regulatory studies show that the severity of effects of beta-cyfluthrin (a synthetic pyrethroid) on fish is influenced by the magnitude and duration of exposure. To investigate how the exposure pattern to beta-cyfluthrin (constant vs peak) may influence the response of the fish, we used a mechanistic effect model to predict the survival and growth of the rainbow trout over its early life stages (i.e. egg, alevin and swim-up fry). We parameterized a toxicokinetic-toxicodynamic (TKTD) module in combination with a dynamic energy budget model enabling us to describe uptake and elimination, as well as to predict the threshold concentration for survival and sublethal effects (feeding behaviour and growth). This effect model was calibrated using data from an early life stage experiment where trout was exposed to a constant concentration of cyfluthrin. The model was validated by comparing model predictions to independent data from a pulsed-exposure study with early life stages of rainbow trout.

RESULTS

The co-occurrence of effects on behaviour and growth raised the possibility that these were interrelated, i.e. impairment of feeding behaviour may have led to reduced food intake and slower growth. We, therefore, included 'effect on feeding' as mode of action in the TKTD module. At higher concentrations, the constant exposure led to death. The model was able to adequately capture this effect pattern in the calibration. The model was able to adequately predict the response of fish eggs, alevins and swim-up fry, from both the qualitative (response pattern) and quantitative points of view.

CONCLUSIONS

Since the model was successfully validated, it can be used to predict survival and growth of early life stages under various realistic time-variable exposure profiles (e.g. profiles from FOCUS surface water modelling) of beta-cyfluthrin.