We re-analyzed field data concerning potential effects of ionizing radiation on the abundance of mammals collected in the Chernobyl Exclusion Zone (CEZ) to interpret these findings from current knowledge of radiological dose–response relationships, here mammal response in terms of abundance. In line with recent work at Fukushima, and exploiting a census conducted in February 2009 in the CEZ, we reconstructed the radiological dose for 12 species of mammals observed at 161 sites. We used this new information rather than the measured ambient dose rate (from 0.0146 to 225 µGy h−1) to statistically analyze the variation in abundance for all observed species as established from tracks in the snow in previous field studies. All available knowledge related to relevant confounding factors was considered in this re-analysis. This more realistic approach led us to establish a correlation between changes in mammal abundance with both the time elapsed since the last snowfall and the dose rate to which they were exposed. This relationship was also observed when distinguishing prey from predators. The dose rates resulting from our re-analysis are in agreement with exposure levels reported in the literature as likely to induce physiological disorders in mammals that could explain the decrease in their abundance in the CEZ. Our results contribute to informing the Weight of Evidence approach to demonstrate effects on wildlife resulting from its field exposure to ionizing radiation.
As explained by Beresford et al.4 simplistic measurements of exposure, such as ambient dose rates or soil concentration activities, cannot encompass all the complexity of actual exposure of wildlife. Contributions of internal and external irradiation pathways to the total dose rates have both to be considered, and this balance depends on radionuclides (type and energy of emitted radiation) and on animal species (age, diet, habitat, use of the environment). Largely used for humans, the dose reconstruction process would also allow the accurate characterization of wildlife exposure required to interpret it in terms of effect. It should be acknowledged that this approach incorporates larger uncertainties than for humans as it deals with interspecific variation in addition to inter-individual differences. However, it is an unavoidable step for a correct analysis of the dose–response relationship.
The effect of exposure to ionizing radiation for wild mammal populations in the Chernobyl Exclusion zone was examined. Our re-analysis of exposure levels included measures of animal life history and past and contemporary radionuclide levels in order to estimate total doses for individual organisms in a manner not previously conducted. This novel approach led to new insights that suggest that likely doses to some animals were often much higher than those estimated using simple measures of ambient radiation levels with doses consistent with those expected to generate deleterious effects based on conventional radiological protection criteria for ecosystems. Our new analysis suggests that a tenfold increase of the mean dose absorbed by mammals over their generation time corresponds to a decrease in total abundance of about 60%. These findings tend to confirm the linkage established by the initial study, by revising upwards the dose rate responsible for the observed effects. They remain however fragile, due to the many associated uncertainties. Especially, despite our effort to consider as many as possible confounding factors, the difficult characterization of some did not allow their fully satisfactory interpretation.
The process of dose reconstruction has proven necessary and useful, as this revision led to estimated levels of exposure correlated to observed effects on mammals consistent with the available knowledge on the toxicity of ionizing radiation on these animals. In large part, this new analysis resolves some of the conflicts that were mentioned in the introduction. However, dose reconstruction alone is not sufficient to demonstrate a causal relationship between exposure and the observed effects. But, it is a step in the right direction and, when combined with the large spatial scale of the Chernobyl disaster and the ability to survey biological consequences at multiple locations across a regional landscape, one can at least have some confidence that the observations are realistic in the light of current knowledge of radiation effects. Our results should be seen as one line in a larger Weight of Evidence approach to demonstrate in situ the effects of exposure to ionizing radiation on wildlife.
More generally, our re-analysis indicates that existing datasets can be a valuable source of new knowledge regarding effects of exposure to ionizing radiation for wildlife in the field, as long as dose reconstruction is possible. This reconstruction process may be long and complex, and there are unavoidable assumptions required to fill each data gap that may introduce additional uncertainties. However, the current study indicates that it may be productive, constructive and useful to implement modern re-analyses to review initial findings that did not fit “conventional wisdom”.
The findings of this study highlight two general issues that apply to all ecological research. The first issue relates to the care that must be taken in designing and planning field studies. It is crucial to pay sufficient attention to and adequately characterize the environmental conditions under which the study is conducted and to report these additional data as completely as possible when publishing so that subsequent re-analysis might be conducted if needed. The second issue concerns the need for sensitivity and uncertainty analyses in dose reconstruction processes (although this applies to any ecological analysis). It is only by collecting appropriate and relevant qualitative and quantitative information that field observations may be properly interpreted.