Mechanistic modelling of reprotoxic effects of pesticides in populations of the freshwater gastropod Lymnaea stagnalis

Alpar Barsi

Supervisors: Virginie Ducrot, Tjalling Jager and Laurent Lagadic


Overall aim of the project

Endocrine disruptors are exogenous substances or mixtures that alters the functioning of the endocrine system and consequently cause adverse health effects in an intact organism, or its progeny, or (sub)populations. This project aims at exploring how effects of endocrine disrupting chemicals (EDCs) on growth, sexual maturity and reproduction of the pond snail Lymnaea stagnalis affect its population dynamics using a mechanistic approach. Studying the mechanisms of reprotoxicity of these chemicals in snails will allow providing original knowledge and models that will enable risk assessors to predict environmental risk of EDCs on invertebrate populations.


Specific hypotheses to be tested/questions to be addressed

How can we understand and predict biological effects of EDCs?

Some EDCs induce complex responses (e.g. inverted U-shaped dose-response curve) in snails. This PhD aims at trying to understand why and to predict those responses from toxicity test data using models. In order to understand why complex responses occur, we will conduct toxicity tests allowing us to assess the dose-response curves of the snails exposed to:

–         natural steroid hormones (e.g. estradiol and testosterone) that are known to interact with the neuro-endocrine system in L. stagnalis and play a role in reproduction,

–         chemicals that have the potency to interact with the neuro-endocrine system of gastropods (e.g. estrogens and androgens mimetics).

If the response to those chemicals is similar to the response of to the natural hormones (i.e. positive control), then similar mechanisms of action could be responsible for toxic effects. Those mechanisms are already well known for natural hormones in L. stagnalis e.g. we know the hormonal pathways for endocrine control of growth and reproduction in these snails. We will use this information as a basis to formulate hypothesis about the mechanisms of action of the EDCs. We will then use models in order to capture those mechanisms in equations that allow predicting responses of snails to EDCs. Therefore, we need to use models that are based upon physiological assumptions. Dynamic Energy Budget (DEB) models have been developed in order to predict individual performances depending on environmental conditions (e.g. temperature, food availability, presence of toxicants). They are based upon a set of rules that specify the acquisition of energy in an organism and its allocation to the major physiological functions like growth, maturity and reproduction. Relative allocation to those physiological functions is hormonally controlled. Therefore, DEB models appear as a relevant mechanistic tool to investigate how hormonal changes (induced by chemicals like EDCs) could lead to altered growth, maturity and reproduction. We will use this model in order to test different scenarios for putative mechanisms of action of EDCs that explains the observed effects on individual life-cycle traits.


Are some adaptations of standard DEB model needed when predicting effects of EDCs on individuals of L. stagnalis?

The standard DEB model is a generic physiological model that can be applied to most animals. The standard DEB model for L. stagnalis is being developed by Elke Zimmer (CREAM DEB-1 project). Yet, this model applies only in uncontaminated conditions. Therefore, we will include equations that allow us to describe the toxicokinetics and toxicodynamics of the studied compounds. Firstly, we will use existing generic equations that apply to all toxicants (i.e. no particular mode of action is assumed). Secondly, we will test whether or not they apply to describe the effects of compounds that may act through endocrine disruption. If not, predicting the effects of those chemical might require some model adaptations. We will thus adapt the model based upon the current knowledge about physiology and endocrinology of L. stagnalis.


Are the effects of environmental concentrations of EDCs relevant for the population dynamics?

EDCs are present in the aquatic environment at low concentrations that are able to affect non-target species as L. stagnalis. We want to know if those low concentrations might generate deleterious effects on populations. We also want to understand how the effects we observe on individual propagate to the population level. To answer to these questions, we will investigate how changes in life-cycle traits (e.g. growth, sexual maturity, and reproduction) caused by EDCs affect populations of the pond snail. Therefore, DEB models will be coupled to population models.


Approach to be used

The overall approach that will be used is based upon the coupling of laboratory experiments and modelling.


Experimental approach:

EDCs are known to generate sublethal effects on development and reproduction in snails. Current standard acute toxicity tests might not be sensitive enough to detect sublethal effects of low doses of EDCs. Thus, a new approach in testing EDCs on snails is currently under validation by OECD for two snail species, including L. stagnalis. New tests should encompass the most sensitive endpoints and parts of the life-cycle that are the most responsive to EDCs. These test protocols will be used in this project to investigate effects of EDCs on pond snails.


When testing EDCs, complex dose-responses can occur. To define a window of sensitivity of snails to EDCs, toxicity range-finding tests will be performed. Snails will be exposed to four compounds which have been shown to induce reprotoxic effect in L. stagnalis: synthetic substitutes for natural estrogens and androgens (17α-ethinylestradiol (EE2) and 17α-methyltestosterone (MT), respectively) and estrogen and androgen mimics (4-n-nonylphenol (NP) and triphenyltin (TPT), respectively). Effects of these chemicals over the life-cycle will be explored in partial-life cycle (PLC) tests. PLC tests will be performed on juvenile and adult snails. Following endpoints will be monitored: survival, growth, time to first reproduction in juveniles, and oviposition rate and cumulated number of eggs in adults. Effects on progeny will be assessed through hatching success, time to hatch and size at birth. Results of the PLC tests will not only enable us to quantify effects of EDCs in an organism but also will be used as input data for calibration DEB models. Additional tests can be expected if DEB models require more data to estimate toxicological parameters.


Modelling approach:

In order to predict effects of low doses of EDCs on individuals, we will use DEB models. Existing DEB models will be used to estimate physiological (e.g. reproduction efficiency) and toxicological (e.g. no effect concentration) parameters that are relevant for population dynamics. If models are not able to predict effects adequately, they will be adapted. Furthermore, since the Euler-Lotka model has already been included in DEB model scripts (DEBtoxM), it will be used to extrapolate effects on population level. Although Euler-Lotka model does not take into account individual variability, its advantage is that the model is simple and easy to use.


An innovative approach of contrasting at least two population models (e.g. Euler-Lotka and Individual Based Models (IBMs) with different complexity is proposed by the CREAM project. It will allow us to understand how predictions change depending on models, which is important information for risk assessors. Therefore, next step will be modelling effects of EDCs on populations of L. stagnalis using IBMs. In IBMs, population dynamics arise from interactions of individuals with their environment and with each others. Such property of IBMs allows predicting effects of EDCs more realistically than the Euler-Lotka model does. DEB models will be integrated into IBMs as submodels. Combining DEB and IBMs will provide mechanistic modelling tools that are relevant for predicting effects of EDCs on populations of the pond snail.


Expected significance of the results

Integrating DEB into IBMs will provide original knowledge and mechanistic models for predicting reprotoxic effects of EDCs on populations of the pond snail. These tools will represent an innovative approach in modelling and enhance basis for further ecotoxicological research. Moreover, models will be developed in a coherent and transparent way within the TRACE (Transparent and Comprehensive Ecological Modelling Documentation) framework. It will help implementing the models for the regulatory risk assessment of chemicals.


Relation to other CREAM projects

Both DEB-1 and DEB-2 projects use DEB theory as a framework to investigate effects of toxicants on individuals of the pond snails. E. Zimmer is investigating how the effects of toxicants on the embryonic development of L. stagnalis may lead to changes in the population dynamics. My focus is on investigation of toxicants effects on juvenile and adult stages and how these effects propagate to population level. Planed visits to co-supervisor Tjalling Jager and E. Zimmer (both work at the Free University of Amsterdam; VU) will allow me to extend knowledge of DEB theory and gain experience with software tools for DEB modelling. In order to explain effects of EDCs in pond snails, data from experiments performed at the INRA will be analysed and discussed. E. Zimmer has already made an internship at the INRA where she performed experiments on pond snails with the goal to investigate reproduction investment The results are not only relevant for her project, but also for mine ( Findings from both projects will enable us to build models that allow to adequately predict effects of EDCs in pond snails.


Publication plan for the project

There will be five papers produced by the end of this PhD project.


Exposure of the pond snails to EDCs can lead to changes in their life-cycle traits. Effects of estrogen and androgen mimetic chemicals will be discussed in the first two manuscripts separately, based upon the results from PLC tests. It will be demonstrated how those compounds affect e.g. growth and time to first reproduction in juveniles, or growth and reproductive outputs in adults, and effects to their progeny (e.g. hatching success). In these manuscripts, modelling will not be included.

In the next manuscript (also presenting effects of estrogen and androgen mimics), it will be shown how effects of EDCs on individuals can be modelled within the DEB framework. The specific question will be stressed: is the standard DEB model able to explain a complex dose response curve of the pond snail to EDCs? If not, in order to describe the effects of EDCs, model modifications will be made and discussed. Finally, effects on individual will be upscaled to population level with the Euler-Lotka equation.


Following manuscript will show how the effects of EDCs on pond snails can be predicted at the population level. For this purpose, developed IBMs will be presented in the paper. Also, in order to predict effects of EDCs on populations, we will demonstrate how DEB models can be integrated into IBMs s. Effects of EDCs on life-cycle traits of the pond snail and their relevance for populations will be analyzed and discussed.


One paper is planned to be published in cooperation with E. Zimmer. Importance of effects of EDCs on embryogenesis and life-cycle traits for the populations of the pond snail will be investigated and discussed. In this study we will use modelling approach that combines both DEB models and IBMs.