Review of radiation effects in non-human species in areas affected by the Kyshtym accident

The first visible radiation effects were observed in coniferous trees as early as spring 1958, when yellow–red dried needles and the suppression of the birch apical meristem were observed in the EURT areas close to the radioactivity release point [13].

In 1959–60, such effects were also observed over 10–30 km along the axis of the release [13, 18]. A substantial fraction of the buds did not burst and the rest were able to develop only short bundles of sprouts [19]. During the first two years after the accident, pine trees died completely within the area with the initial contamination density of ⁹⁰Sr at a level of 6.0–8.0 MBq m⁻², where doses to the needles were in a range of 20–40 Gy and doses to the buds ranged from 10–20 Gy [17–19].

The area of the deceased forest was about 50 km², whilst at an area of 80 km² (with initial ⁹⁰Sr contamination density ranging between 3.7–7.4 MBq m⁻²) the fraction of the deceased pine trees reached 95%. Severe damage of the pine tree crowns was observed in areas with initial ⁹⁰Sr contamination density at a level of 6.0–8.0 MBq m⁻², where doses to needles and buds were 5–10 Gy and 2–4 Gy, respectively. The radiation effects also included the delay or loss of buds, delayed sprouting development and subsequent dying of crowns [18].

Radiation effects in deciduous forests were observed in the contaminated parts of the EURT, where the dose to buds reached 100–200 Gy. The mortality of 50% of deciduous trees was observed at the site with area of around 3 km², with ⁹⁰Sr contamination density at about 140 MBq m⁻² [17]. Thirty percent of trees totally lost their crowns, and 70% of the understorey died at that contamination density. Mortality of deciduous trees of around 10% was observed at the area of 12 km² with initial ⁹⁰Sr contamination density of ⁹⁰Sr of 90–100 MBq m⁻². The doses to the crowns (bud meristem) at the site were about 100 Gy.

Thus, the absorbed doses of 100–150 Gy accumulated in the bud meristem induced the partial drying of trees and the death of birch trees was observed at the 'acute' dose of 200 Gy [13, 17].

Around 1% of the deciduous trees have lost crowns at the sites with the lower ⁹⁰Sr contamination density of 37–59 MBq m⁻². Such effects were observed within an area of 15 km² with doses of 40–60 Gy during the first year after the accident.

Young and weakened trees, which are normally below the main trees' tier, showed higher radiosensitivity compared to the trees comprising the main tier. In the area with doses of 40–60 Gy the mortality of weakened trees reached 30% [18, 19].

In 1961, both birch and pine trees in the affected areas demonstrated increased rates of chromosome aberrations, the presence of abnormal pollen, decreases in tree development, structural anomalies, and decreases in cellular mitotic index [20].

The rate of chromosome mutations in plants was constantly decreasing with time, for several reasons. First, the doses to the generative tissues were also constantly reducing, and second, because of the influence of DNA reparation processes and the natural selection of specimens, which adapted to existing environmental factors.

The results of studies with a provocative (additional) irradiation of plant seeds sampled in 1967 from sites with doses of 20–200 Gy during the first year after the accident suggested that the surviving plants have increased radioresistance compared to the unaffected populations [13, 17–19].

Shevchenko et al [20] reported data on the effects of extra γ-irradiation (100–150 Gy) of dormant buds of birch grown at the densities of 37–140 MBq (⁹⁰Sr) m⁻², and did not find any decrease in the mutation rate in the meristem of the shoot cells.

The assumption behind these data is that, after some period of increased mutation rate, the population of the affected organism can stabilise at some higher radioresistance level with lower mutation rate [20]. The duration of the period required to achieve such an adaptation depends on the mutation type. For chromosome aberrations, it can be achieved within a few years of exposure beginning, whilst for other mutations more than 30 years are required to reach such an equilibrium [20].

Another lesson learned from these studies was the conclusion that higher relative radiation damage is observed at lower dose. This phenomenon, called the 'genetic efficiency' of radiation, consisted of the increase in the yield of genetic effects per unit of absorbed dose with the reduction of the radiation dose. The chromosome aberration yield per 1.0 Gy in the cones of pine trees, measured for the cones which received 0.2 Gy, was higher by a factor of 2.7 than that measured for the cones exposed to a dose of 1.6 Gy [22].

Over 3–5 years after the accident, the affected forest ecosystems began to recover. However, if the recovery of birch forests through sprouting of surviving trees had a good chance of success, the restoration of pine forests through seed propagation from the adjacent territory occurred much later [17]. The most affected pine forest was replaced by either a birch forest or mixed pine-birch forests with dominating birch trees. The recovery of the affected forest failed to achieve the pre-accidental state in most contaminated sites.

The exposure of the generative tissues of plants can be different for the same radionuclides and contamination density, depending on plant architectonic and biomass density. In particular, plants may be classified into several groups, depending on the location of winter brood buds relative to the contaminated soil surface. Brood buds of Phanerophytes (trees and high shrub species) are located much above the ground, whilst brood buds of Chamaephytes (such as bilberry shrubs) winter in the snow layer near the top soil surface. Brood buds of Hemicryptophytes (such as reed grass) are located at the soil surface and buds of Cryptophytes (meadow grass, fescue, speedwell, etc) are in the soil layer. Therophytes represents species with winter seeds. Such peculiarities in the location of the brood buds can themselves result in very different radiation doses and biological effects [23, 24].

Major changes in the composition and productivity of the herbaceous species were observed in the spring–summer after the accident. Disappearance of Therophytes species and reduction of the total above-ground biomass by nearly 30% were observed at the sites with a ⁹⁰Sr contamination density of 37 MBq m⁻² or higher [13, 23].

Hemicryptophytes were the dominant species during the first years after the accident, although they received the highest exposure doses. Measured in terms of the leaves' projective cover, they amounted to nearly 70%, whilst Cryptophytes showed suppressed growth.

After 5–6 years after the accident, a considerable reduction in Hemicryptophytes species was observed. It is worth mentioned that such a reduction was anticipated based on damage to the reproductive organs and increase in the proportion of some Cryptophytes species. In the next 5–6 years, a slow recovery of the communities of herbaceous plant was observed, although it has failed to achieve pre-accidental states [24, 25].

Overall, these results showed that in terms of the plant composition measured by leaf area specific to different plant groups, the plant community did not recover at the sites where doses absorbed by plants were above 30–50 Gy, and this dose was documented as a threshold value for herbaceous plants' recovery. At the higher absorbed dose, a change in the structure of the plant community was observed as the proportion of Hemicryptophytes species was lowering, and that of Cryptophytic species was growing [23].

It should be emphasised that such a criterion for the assessment of radiation effects at the community level is rather sensitive and combines the effect of radiation with the effects of competition for ecological resources such as solar radiation and nutrients.

The frequencies of alterations in the enzyme electrophoretic mobility of 6.6% and 4.5% were observed for herbaceous plants (Centaurea scabiosa) on sites with the absorbed dose rates of 12 and 6 mGy d^–1 in 1982. At the uncontaminated site with similar environmental characteristics the enzyme electrophoretic mobility value was as low as 0.4%. Similar effects were observed for chlorophyll mutations [20, 21].

A statistically significant increase in the aberration rate in the first mitosis in seedlings was also found to be characteristic only for the sites with doses to plants (Crepis tectorum) of 100–180 mGy yr^–1 [21, 24]. At the same time, no statistically significant effects were found for plants exposed to radiation dose of 2–50 mGy yr^–1 [20, 21].

Later on, the rate of chromosome aberration was measured in the cells of the root meristem of seedlings from military grass (Crepis tectorum). The samples were taken from five sampling sites with ⁹⁰Sr contamination density ranging from 92–9200 kBq m⁻². The absorbed dose to the seedling meristem varied from 2–180 mGy yr⁻¹ and chromosome aberration rates were in a range 0.46%–0.78% for all sites except the most contaminated area with a ⁹⁰Sr contamination density of 9200 kBq m⁻², where a chromosome aberration rate of 1.45% was statistically different from the above values [21].

To assess the role of 'adaptive response' for herbaceous plants, the seeds of ten species of wild grass (narrow leaf vetch, forest speedwell, mountain clover, silverweed cinquefoil, etc) sampled at the contaminated sites were exposed to additional γ-irradiation with a dose of 150 Gy. The results demonstrated a significant increase in seed germination and a decrease in the rate of chromosome aberrations in seedlings compared to those measured in the non-exposed plants. It was shown that the radioresistance increased three–four-fold for relatively radiosensitive plants, and remained at the same level in more radioresistant species [20].

Other evidence of radiation adaptation was reported by Shevchenko et al [20], who carried out an experiment with a provocative γ-irradiation (20 Gy) of Centaurea scabiosa seeds sampled at the sites with contamination densities varying between 37–140 MBq (⁹⁰Sr) m⁻². It was shown that the rate of chromosome aberrations for the most exposed plants (19.7%) was lower than that of the less affected seeds (30.4%) sampled, and much lower than in the plants sampled at the uncontaminated site (42.9%). The differences were statistically significant. Similar results, demonstrating an increase in the proportion of plants sampled in highly contaminated areas that were substantially resistant to radiation, were observed in 1995 for another plant species, Crepis tectorum L. [21].

Investigations of the genetic effects of radiation on Centaurea scabiosa showed that the genetic efficiency of radiation under chronic exposure amounts on average to some 1 · 10⁻² mutations Gy⁻¹ per locus. This value increased 3.9 times with a dose reduction from 8.5 to 0.4 Gy, demonstrating the clear dependence of genetic efficiency on absorbed dose.

Soil invertebrates were one of the most exposed species in areas affected by the Kyshtym accident [26–28]. Investigations were performed on the site (five locations) with average absorbed doses in the top 1 cm layer of soil of 5–12 mGy d⁻¹ located in the birch forest. The doses to invertebrates were directly measured by thermoluminescent dosimeters [26]. The number of soil invertebrate species found on the contaminated site in 1969–71 was around two-fold lower than at the control sites located in similar uncontaminated birch forest [26, 27]. The highest reduction in the number of animals was reported for saprophages, polypodies and earthworms (figure 1). Each such group of soil invertebrates was represented at the contaminated site by a few specimens [28]. Substantial reduction of these species at the contaminated site can be explained by the relatively low mobility of saprophages, which during their life cycle were in long-term contact with the most contaminated soil [13].

Zoom In Zoom Out Reset image size

The total number of mites at the contaminated sites was not statistically different from that at the control sites. However, the number of mite species at the contaminated sites was around four-fold lower, illustrating the difference in radiosensitivity of different mite species to radioactive contamination [28].

Another study performed 30 years after the accident showed that even after such a long time, the community of invertebrates had not been restored to a state close to that found in the surrounding uncontaminated areas, and the numbers of the soil animals under study reached 15%–77% of those observed in similar but uncontaminated sites [27] (figure 1). These data also demonstrate dramatic changes in the structure of the soil invertebrates, and variations in the rates of the recovery of the invertebrate community.

The effects of radiation observed in different soil invertebrate groups differed substantially. Detritivores (earthworms, millipedes) and centipedes were most affected because they occupied the most contaminated forest soil horizons. These animals are sedentary and consumed very contaminated detritus. Centipedes mainly consume earthworms and millipedes, i.e. demonstrated high accumulation of radionuclides because of a trophic chain effect.

Overall, the trophic structure of the soil invertebrates tended toward some simplification, and biodiversity of the individual invertebrate species at the contaminated sites was lower than that observed at the unaffected sites. So, at the uncontaminated sites 19–20 species of crusade mites were found within the top soil horizon; at the highly contaminated sites, only 4–5. The number of species of crusade mites was similar in the deeper soil horizons: 9–10 species at the contaminated sites and 10–13 at the uncontaminated sites, reflecting the effects of high doses within the top soil horizon [28].

Another striking point relevant to the effects of radiation on the soil invertebrates is the extremely high impact and slow recovery of these animals compared to the other affected non-human species [27, 28]. This was not expected based on the data from laboratory experiments with acute γ-irradiation. The possible reasons for such effects are as follows:

• Firstly, the effects observed during studies performed in 1968–69 and 1987–88 reflect the consequences of the extremely high impact during the acute phase of the accident and the slow decrease of the chronic doses afterwards.
• Secondly, during a life cycle, soil invertebrates pass through several developmental stages, very different in terms of radiosensitivity, and require several years to reach a fully developed stage. The irradiation of animals at the radiosensitive stages (with quite low LD₅₀) could induce adverse effects observed later.
• Finally, soil invertebrates have very low mobility and there was a lack of inflow of non-exposed animals from surrounding, uncontaminated areas.

Domesticated animals, such as cattle and sheep, were highly exposed after the accident and the first radiation effects on these animals were observed shortly after the accident. The evacuation of the public and animals from the most affected area was conducted 12 days after the accident, and during that time the animals were grazing at the sites with a total contamination density of around 900–1100 MBq m⁻². They received external doses of 1.4–3.0 Gy and the GIT doses reached 4–24 Gy [13, 14]. These doses resulted in the mortality of the exposed cattle with symptoms specific to acute radiation sickness, including bleeding of mucous membranes and leucopenia. Cattle grazing outside of the radioactive trail, but still in proximity to it, received external doses of around 0.1 Gy and doses to the GIT of 1.0–2.0 Gy. These animals survived for at least six months, although some detrimental changes were found in the blood-producing system of the animals [14].

Similar effects were observed for sheep. Sheep grazing on the sites close to the axes of the release received external doses of 1.4–3 Gy and GIT doses of 8–54 Gy in the 12 days after the accident. As was mentioned earlier for cattle, the exposure resulted in symptoms of acute radiation sickness and death in most of the animals. No substantial radiation effects were observed in sheep that were grazing at the less contaminated sites (100–200 MBq m⁻² of the total activity). The calculated doses during the 12 days after the accident amounted to 0.1–0.2 Gy whilst the GIT doses were at the level of 2–4 Gy. Temporary changes in the blood-producing system of the animals recovered within a few months of the evacuation of the animals [13, 14].

The evaluation of the radiation effects in mouse-like rodents was the focus of research in the Kyshtym affected areas from the early 1960s [29–39]. Lethal effects in mouse-like rodents were found under doses of 10–20 Gy [13, 19]. In the areas with such exposure levels, considerable reduction of population and dramatic changes in the species structure of the mouse-like rodent community were observed [29, 32]. In particular, at the end of the spring of 1958, the population of short-tailed voles (Microtus agrestis) sampled in the area with an initial total contamination density of 70–130 MBq m⁻² consisted of only 2% older voles (born in early or mid-summer 1957), whilst the unaffected population included 50% young and 50% older mice [29]. During the first winter after the accident, around 60% of the short-tailed vole population died. At a similar site outside of the highly contaminated areas, the death of mice was lower than 18% [32]. Similar effects were noted for some other mouse-like species inhabiting the EURT, such as the backed mouse (Clethrionomys rutilus) [19].

Some detrimental morphological effects were found in the Microtus agrestis population inhabiting an area with initial contamination of 70–130 MBq m⁻². Upper teeth of abnormal size, never observed in non-exposed animals and preventing their normal feeding, were observed in 16% of the affected mouse population [19]. However, osteosarcoma incidents were not found in the mouse-like rodents, although ⁹⁰Sr concentrations in bones were extremely high [13].

Radiation effects were also observed for mice that received absorbed doses of between 1–10 Gy. In 1962, the reproductive period of wood mice exposed to a dose of 4.0 Gy per lifetime was shorter than that of unaffected animals. In particular, in August, there were 10.4% pregnant females of the control population, and only 0.8% in the population of the exposed mice. The number of embryos in exposed European wood mice (Apodemus sylvaticus) was smaller than that measured at the uncontaminated sites [19].

Statistically significant radiation effects were registered at both the organism and the population levels at the sites where the average annual dose to red bone marrow amounted to 4–8 Gy in 1962 and 0.6–1.0 Gy in 1981 [32, 39]. Substantial changes were reported in animal mass and size and in some animal tissues such as the spleen and liver. They were also noted for craniological performance, blood and other morphophysiological parameters. The spleen weights of bank voles trapped in the affected sites were 48%–60% of those measured in animals sampled in uncontaminated plots, whilst the erythrocytes, leucocytes and agranulocytes of affected bank voles were, respectively, at levels of 109%, 93% and 85% of the corresponding values measured for unaffected specimens [19]. The observed radiation effects correlated with the concentrations of ⁹⁰Sr in the skeleton [13].

Reported data indicated a reduction in the life span of the exposed animals, a decrease in their survival rate and an increase in their death rate during wintering. Reduction in the reproductive potential and increased (up to two-fold) incidence of haemo- and ecto-parasitism were also observed for animals exposed to a life span dose of 5.0 Gy [30, 32].

It is likely that the rodent population in the most contaminated areas was to some extent restored mainly through migration of animals from less contaminated areas. Thus, the populations of mouse-like rodents at the start of observations (1962) could be formed from animals which were very different in terms of received doses [13, 19].

Such effects were not observed for northern red-backed mice (Clethrionomys rutilus), which received doses ranging from 0.1–1 Gy. Physiological effects such as hypooxygenia, high breathing rate and high rectal temperature were found to be typical in the wood mouse populations inhabiting the sites with an annual dose of around 1.0 Gy [19].

The mouse-like rodents showed a high adaptation to the elevated radiation levels observed for the population living in the most contaminated part of the EURT for 20 generations and more [19, 32–39].

Additional exposure of Clethrionomys rutilus (the northern red-backed mouse) to the lethal external γ-radiation dose of 11 Gy (⁶⁰Co source) was performed in 1970–72. Death of the unaffected population came 12.6 ± 0.6 d after the lethal exposure, whilst death of the population sampled from the highly contaminated area came 26 d after the lethal exposure [34, 36–39].

Similar effects were observed for Apodemus sylvaticus (the European wood mouse), exposed in 1978 to the acute γ-radiation dose of 8.0 Gy. Although all animals taken from the uncontaminated area died within two weeks of exposure, a proportion of the animals trapped at the site with an initial contamination of 70–130 MBq m⁻² survived the whole observation period [37].

This serves as evidence of strong changes in the genotype of affected animal populations with inheritable properties because of long-term chronic exposure. Animals trapped in the affected area also had enhanced resistance to some other extreme factors, such as a sharp change of temperature [30]. This can be considered as a result of the increase of variability of population characteristics and, hence, the higher ability of animals to adapt to new environmental conditions, including not only elevated radiation but some other factors as well [30, 38, 39].

It was also reported that mouse-like rodents (bank voles and wood voles), sampled at the most contaminated sites 20–30 generations after the accident, demonstrated an increased rate of chromosome aberrations and aneuploid karyotypes in bone marrow cells. Further research performed with the same species has shown that this effect was inherited in the first, second and third generations of mice born from the exposed population. The presence of surviving animals after exposure to the lethal dose was typically observed in the progeny of the affected mice, whilst animals of the control population were quickly coming to their death. Furthermore, over time, the proportion of radioresistant animals increased, reaching 20% after 12–15 generations [38, 39].

An increased rate of chromosome aberrations and aneuploid karyotypes in bone marrow cells was reported for mouse-like rodents. However, karyotypes that could transfer hereditary alterations, and contribute to accumulating adverse genetic effects across generations, were not detected [20].

There is also some evidence that mice inhabiting more contaminated areas could more easily be trapped by predators. Thus, ⁹⁰Sr in the bones of about 80% of mice taken by buzzards (Buteo buteo) reached 35 kBq kg⁻¹. The proportion of mice with higher ⁹⁰Sr concentration in their bones, sampled by mechanical traps within that area, was only 20% and most of the mice had concentrations one order of magnitude lower. This demonstrates the presence of ecological mechanisms that serve for the natural elimination of the most affected animals with a lack of adaptation to avoiding predatory birds [19].

Absorbed doses of around 1.0 Gy to the GIT of large herbivores have led to a reduction of the animal population. At the sites with a contamination density of 37 MBq (⁹⁰Sr) m⁻², animals could receive an additional external dose of 2–3 Gy. In that case, early radiation effects and even the death of some animals can be anticipated. Some reduction in the number of moose and roe deer was observed in 1957–58 in areas where the GIT doses could reach 10–30 Gy. However, increased mortality of large animals was not documented, most likely because of the high migratory ability of large herbivores.

In general, birds are more radioresistant than mammals. Some estimates of table 2 show that rather high doses could be received by some wintering bird species (magpie, raven) inhabiting the highly contaminated areas in 1957–58. Although existing estimates show that in the spring of 1958, a ten-fold decrease in the number of wintering birds compared to pre-accident years was registered, no direct data on the increase of the bird mortality rate during the first year after the accident are available [13, 19].

Reproductive success of some birds (flycatchers) was studied in the 1990s based on placing artificial nests in the highly contaminated area [40, 41]. From 30 nests, only six were occupied with one baby bird, in spite of the fact that reproductive success was about 92% in the population occupying the uncontaminated site.

Comprehensive studies of radiation effects in amphibians started in the most contaminated part of the EURT only in 1992–95, i.e. with a high delay after the accident [42–45]. The sampling site for brown frogs (Rana arvalis) was selected near Lake Berdenish within the area with a ⁹⁰Sr contamination density of 56 MBq m⁻². The observations showed that the exposed population was characterised by reduced reproductive success, and 17% of the young frogs had visible morphological abnormalities. Additionally, the sizes of the eggs, as well as the sizes of adult specimens, were smaller than those observed for non-exposed frogs. The mortality of the frogs' embryos was also higher in the exposed population compared to the frogs sampled at the low contaminated site, and fore larvae developed from only 10% of eggs.

Some detrimental radiation effects were observed in viviparous lizards sampled in 1992–95 at the highly contaminated site of the EURT [43, 45]. Although no morphological changes were observed for adults, 26.6% of the studied embryos had clear abnormalities, including duplications of the head, tail, legs or fingers, and abnormal neck curvatures. As for the control population, such abnormalities were found in less than 2.1% of the population.

During the 'acute' period, the fish from the most affected lakes, Berdenish and Urus-Kul, received doses of 10–40 Gy (table 2). Such absorbed doses could induce a decrease in productivity of fish populations due to radioactive damage to the spawn. However, no critical biological effects, such as a substantial reduction in fertility or life span, were documented. The observations of fish populations in these lakes, started in 1960, did not show a substantial reduction in fish productivity that would suggest possible biological compensation for radioactive damage.

Fish, molluscs and crustaceans are most radiosensitive at the ontogeny stage corresponding to the early division of generative cells. Based on the information on the increase of the mutation rates at a dose rate higher than 1–2 cGy d^–1, it was assumed that the reproductive system and laid spawn are vulnerable in the fish inhabiting these EURT lakes [20].

At the same time, observations of the carp population of Lake Urus-Kul, exposed to radiation over 16 generations, did not show statistically significant differences in the rate of chromosome aberrations compared to a similar uncontaminated lake. Thus, for dose rates to the gonad of 5.0 mGy d^–1 (reported doses to spawn were 1–2 mGy d^–1) the yield of chromosome aberrations was measured to be as low as 2.5%. This value was a bit lower than that observed in the control site of 3.3% [20]. Similar effects were observed for the golden carp population sampled in Lake Berdenish [13, 19]. Overall, variabilities in 14 morphometric parameters in crucian carp from lakes Urus-Kul and Berdenish were found not to be statistically different from those in similar lakes located outside the EURT [13].

At the same time, some morphological anomalies were observed in 15% of the fish sampled in the Urus-Kul Lake. These effects mainly included anomalies of gonad structure (unpaired gonads, sterility) [46]. Some morphological anomalies were also observed in 1972–75 in 24% of the fish sampled in the Berdenish Lake, including abnormalities of gonads (17%) and sterility (25%) [47–49].

Based on this research, an average rate of mutations in the fish of the most affected lakes was estimated to be 10⁻³–10⁻² mutations per locus per generation, with that for uncontaminated lakes 10⁻⁶–10⁻⁵ [20]. However, the dose–response relationship for the fish was not clearly documented.

The research performed at the Urus-Kul Lake, where the doses to freshwater biota ranged in 1957–58 between 1.0–40.0 Gy during the first year after the accident, demonstrated a 1.5–2.0-fold decrease of biodiversity for the communities of the species related to all trophic levels, namely, for phytoplankton, zooplankton and benthic invertebrates [50]. The degree of the development of most species and their biomass per cubic metre in the affected lake was also much lower compared to those measured for similar lakes located outside the contaminated zone [50].