Current environmental risk assessment (ERA) models are pragmatic tools for estimating the effects of chemicals in the environment. Consequently, the current approaches are often over-simplistic in terms of considering how ecosystems may respond to chemical exposure and as a result they may over or under predict fate and effects in the environment. Indeed, one of the major questions posed by risk assessors is ”can classical ecotoxicological studies for use in chemical risk assessment be used to predict relevant ecological endpoints such as population effects?”
The current approach in risk assessment uses simplistic assessment factors on LC50s or NOECs from standard studies, or at best species sensitivity distributions, to provide an assessment of effects in the field. An implicit assumption of such assessments is that the approach in some way integrates any effects of the chemical under consideration with other influences on a population's response to chemical exposure arising from the complexity of ecological interactions within natural communities. However, the data taken from conventional ecotoxicity tests do not do this. Despite the provision of significant amounts of data from chronic studies (notably reproduction tests) little of this information is used to extrapolate to potential population related effects that may be encountered in the environment. Existing data on survival, fecundity and growth can be integrated into simple population level models. These can be further refined using factors such as predator-prey relationships, density dependant growth rates and recruitment.
Review and examine the fitness for purpose (for use in risk assessment) of existing population models described in the literature (including the use of recruitment models used in fisheries population modelling). Consider the different input variables that are required and collate these, ranking them in terms of availability and usefulness, e.g. can they be extracted from standard ecotoxicity tests? The review will attempt to provide information relevant to the following:
How much complexity is required to deliver ecological relevance? What are the likely consequences for the conclusions from risk assessment? Whether such models can be used to include or predict effects on keystone species and the resulting impact on the food web.
The key deliverable is a report that addresses the suitability of such models for inclusion in ERA, food web interaction analysis and makes recommendations for further research; this should be achieved by;
- Summarising existing models, and ranking their order of suitability for use in ERA.
- Listing further data that could be added to standard toxicity tests in order to improve their suitability in population modelling. This should be based on sensitivity/elasticity analyses to identify the key drivers of the population level dynamics.
- Summarising case studies linked to each model that explores their validity as a predictor for population dynamics for use in ERA.
It is expected that the findings will developed into a peer reviewed publication, following presentation at a suitable scientific conference. Short interim reports on progress are required at 3 to 6-monthly intervals.