In 2019 only evaluated ecological models will be accepted in applications
In recent guidance documents such as the EFSA Aquatic Guidance Document and EFSA guidance Birds & Mammals, the possibility of ecological modelling is included as an option for risk refinement. Applicants are increasingly taking advantage of this option. Often these are models that are developed by industry. When the Ctgb receives an application, these models have not yet been evaluated. This involves an evaluation that often has to be repeated for every situation and for every product. Until now the Ctgb has coordinated the evaluation of an ecological model by partially outsourcing this evaluation to an independent, qualified organisation. This procedure is going to change.
As of 1 January 2019, the Ctgb will accept applications with ecological models only if these models have previously been evaluated. Applicants are free to choose an independent and qualified organisation to conduct these evaluations. Upon receipt of the dossier, the evaluation of ecological population models will be reviewed according to the following criteria:
- The evaluation must be carried out by an independent and qualified organisation.
- Model documentation must be as described in the EFSA opinion (e.g. TRACE documentation).
- Evaluation report:
The evaluation report should follow the recommendations from the EFSA Scientific Opinion on Good Modelling Practice (2014). For example, a typical entry could consist of the following components
(This list is not exhaustive and more information can be requested on a case-by-case basis. Applicants can contact the Ctgb for more information about their dossier or an RFM):
- Evaluation of all literature data used in the model (ecological and toxicity data): This step checks which model parameters and assumptions are derived from the literature, how they were derived and how they are used in the model.
- Specification of the questions to be answered by the model and evaluation of the conceptual model.
- Evaluation of the conservatism of the environmental scenario, i.e. whether the landscape and exposure scenarios are representative and conservative for the requested use in the GAP.
- Model implementation and verification: evaluating the model code (functions, routines, logic of sequence and debugging if necessary) and whether the output of the model makes sense.
- Sensitivity analyses: analysis of which parameters are sensitive to model output (globally and/or locally).
- Model validation: comparison of model simulations with independent data sets and/or analysis of the output by means of model simulation patterns (pattern-oriented modelling).
- Uncertainty analyses: Sources and description of model uncertainties propagated to the model output.
- Executable model software with user manual and full source code.
These points are also included in the priority list for the intake. If these points are sufficiently addressed, the ecological population model can be part of the dossier.
For questions or discussion about the evaluation, an applicant can apply for an RFM.
Once a model has been evaluated, no further evaluation is needed for the adaptation of the model to new species. An evaluation of the input of the model is part of the assessment and is carried out internally.