There are fundamental differences in determining the risk to humans following exposure to radiation and the risks to other organisms.53 Human risk analyses largely focus on cancer risks to individuals. Dose-response relationships are sufficiently well known that risk factors (i. e. probability of lethality from cancer per unit of dose) are established. In contrast, ecological risk to wildlife is concerned generally with populations of plants and animals. For most organ­isms, cancer induction is not relevant and suitable endpoints include morbidity (functioning less well), reduced reproductive success, mortality and chromo­somal damage. The dose-response relationships for these endpoints are not established for many wildlife groups, and therefore there are well established and quantified risk factors that equate dose to the probability of an outcome.

The endpoints considered to be most relevant in determining risks to wildlife are increased mortality, increased morbidity and decreased reproductive out­put. Of the three, changes in reproduction are thought to be the most sensitive to radiological exposures and relevant for the protection of wildlife populations (populations rather than individuals being likely to be the object of protection for environmental assessments).54 Much more data are needed, however, before we can confidently predict population level impacts to wildlife as a function of radiological exposures.55 Data are particularly scarce for the chronic, low-level exposures for which most assessments will be used. The available data on the biological effects of ionising radiation on wildlife have been compiled from the literature into an online database called the FREDERICA radiation effects database (http://www. frederica-online. org).56