Comparison of population level and individual level endpoints to evaluate ecological risk of chemicals.

Ecological risk assessments (ERA) of chemicals are often based on mortality and reproduction of individuals. To protect populations, fixed safety factors are applied to the data. However, the relationship between individuals and populations cannot easily be described by predefined numbers. The use of population models may reduce uncertainty and, hence, the risk for erroneous assessments. However, introducing models also introduces additional complexity. Therefore, it is desirable to keep the models as simple as possible. The objective of the present study was to determine whether simple risk equations or matrix models can improve ERA compared to traditional endpoints. To examine this, complex models that included environmental stochasticity and density dependence were used to simulate population level risk based on dose-response data for five chemicals. The risk, measured as probability for pseudo extinction and recovery time, was then compared to risk estimates based on individual level data (acute and chronic), risk equations, and simple matrix models. The results showed that the simple matrix models reduced uncertainty by more than 88% and 76% compared to acute and chronic data, respectively. Also the simple risk equation reduced uncertainty considerably (80% and 61% compared to acute and chronic data, respectively).