Ecosystem effects of the Kyshtym accident

The 1957 'Kyshtym accident' at the Mayak Plutonium production site in the Southern Urals of Russia was the World's first major nuclear accident, leading to evacuation of 1000 km² of land and causing significant environmental damage. Sergey Fesenko, an internationally recognised authority on the Kyshtym and Chernobyl accidents, presents a very valuable summary of the fascinating ecosystem research conducted after the accident (Fesenko 2019). In the years after the explosion, most of the released radioactivity decayed away, but the remaining (approximately 5% of the total release) relatively long-lived Sr-90 (half-life 28.8 years) caused chronic long-term exposure of wildlife.

As documented by Fesenko (2019), this research represents some of the first real field studies of the effects of radiation on the natural environment. Doses to pine needles during the 'acute' phase of the accident reached 800 Gy and doses to mammals and birds exceeded 10 Gy in the most contaminated areas. Soviet researchers documented the lethal and sub-lethal effects of these huge doses on trees and small and large mammals in the affected areas. In a process repeated in the Chernobyl 'Red Forest' nearly 30 years later, they observed that extremely high beta-radiation doses to the leaves and buds of trees had lethal effects. Domestic animals exposed to external doses of 1.4–3.0 Gy and 4–54 Gy to the gastrointestinal tract (GIT) showed symptoms of acute radiation sickness and most died.

The studies of the effects of much lower chronic dose rates (from the early 1960s onwards) made a major contribution to interesting new concepts in radioecological and radiobiological research, including 'radioadaptation' of animals and plants to living in contaminated areas. They also included findings of long-term ecosystem effects. Apparent effects on soil invertebrate populations were seen at dose rates (5–12 mGy d⁻¹) which, though by normal standards are very high, are nevertheless significantly lower than might be expected from laboratory experiments on these organisms. Similarly, as late as the 1990s, major reproductive effects on birds were observed; though not given in Fesenko (2019), the ⁹⁰Sr contamination level is reported in (Sazykina and Kryshev 2006) as being extremely high (55 MBq m⁻²). Some of the findings summarised in the review are relevant to research apparently showing effects on organisms at Chernobyl at dose rates much lower than would be expected from laboratory radiobiological studies (Garnier-Laplace et al 2013), though the strength of evidence in these studies has been questioned (Smith 2019).

We have to acknowledge (as Fesenko does (2019)) the inevitable limitations of field research into the effects of radiation on wildlife. Studies of long-term (often subtle at low chronic dose rates) effects after Kyshtym, Chernobyl and, to a much lesser extent, Fukushima may be confounded by early ecosystem damage from extremely high dose rates due to short-lived radionuclides. At both Kyshtym and Chernobyl, initial damage of forest ecosystems was severe in the (most usually studied) hot-spots. Further confounding can occur from the difficulty in separating radiation effects from natural environmental variation. Simple contrasts between high dose rate areas and 'controls' are rarely sufficient to demonstrate effects as (unless there are very many true replicates) they can rarely capture this variation. It is all too easy for authors (and reviewers) to assume, uncritically, that observed correlations between some biological effect and dose rate is causal. But models apparently accounting for radiation effects together with potential confounders such as habitat quality and food availability rarely account for more than half of the observed variation in data. Co-correlations between radiation and unknown (or unstudied) variables can also confound results.

To some readers, debates about the ecosystem effects of radiation at sites which have suffered major nuclear accidents may seem somewhat academic. Dose rates at which significant affects are seen, even in some of the more surprising studies, are usually (but not always) orders of magnitude higher than those arising from regulated releases of radioactivity to the environment. However, the environmental impacts of nuclear accidents generate significant media, public and political interest and need thorough and ongoing scientific research to test effect hypotheses. We need to continue the work begun by the pioneering Soviet scientists at Kyshtym on the long-term ecosystem effects of chronic low dose rates. But, I think, with a more modern understanding of the difficulties of attributing causation in a complex natural environment.