Measuring and predicting the effects of time variable exposure of pesticides on populations of green algae : combination of flow through studies and ecological modelling as an innovative tool for refined risk assessments

The EU risk assessment of pesticides for aquatic organisms relates measured effect data to exposure values obtained from model predictions. The FOCUS exposure models provide Predicted Environmental Concentrations in surface waters. Calculations use realistic worst-case scenarios, based on representative agricultural conditions. Pesticide input via spray drift, runoff or drain flow results in exposure patterns in a time-variable scale. Standard toxicity tests and effect data are based on the assumption of a single exposure event into a static water body. Relating the results of these standard tests to FOCUS time-variable exposure patterns is only possible by gross simplifications and the use of overly conservative worst-case assumptions. The topic on how to relate pulsed exposure patterns to simple exposure assumptions is discussed since the inclusion of the FOCUS scenarios in the EU risk assessment of pesticides. Population-level effects and time-to-recovery after time-variable exposure need to be considered for more realistic, but still conservative risk assessments. One way forward is the use of population models within the higher-tier risk assessment, in close combination with ecotoxicological experiments. Out of the multitude of interactions of pesticides and aquatic non-target organisms, the example of herbicides and algae was chosen for this work. This work aimed to develop and test a new approach to assess the effects of time-variable exposure on algae. This approach comprised the close combination of ecotoxicological experiments and population modelling. It involved the development of an algae model including all necessary steps for verification and validation as well the performance of laboratory experiments. A mechanistic compartment model, SAM-X (Simple Algae Model EXtended), was developed. This model describes population dynamics of the four different algae species under static as well as under continuous culture conditions. The experimental approach of this work was the development and application of a flow-through system for the continuous culture of algae. Thereby, the exposure of algae populations to a pesticide could be investigated in a time-variable scale. This setup was based on the Chemostat principle, which allowed a continuous operation in a stable manner without the need of external influence. The results of this work confirm the proposals available in literature to use flow-through systems as appropriate higher-tier tools to assess effects of complex toxicant exposure to algae. The algae model proved its capability to simulate effects on population growth and times to recovery after time-variable exposure of toxicants. Moreover, the results support the use of the algae model as a predictive tool to assess effects of pesticides on population level. Thus, an implementation of the model as an inherent part of the pesticide risk assessment would be an asset. The usefulness of the combined approach was confirmed by the results, as not only the experiments offered detailed information on population-level effects after pesticide exposure, also the model application showed practical advantages: It can be applied for designing and improving the flow-through study as a first step, to describe the results as a second step, and thirdly, to extrapolate to different exposure scenarios to predict effects on the algae populations. In summary, this work demonstrated an excellent option to represent pulsed exposure of a pesticide and corresponding effects on algae a bit closer to reality. The conclusive results of the combined approach confirm the beneficial use of population models as supporting tools in higher-tier risk assessment of aquatic toxicants