To evaluate the effects of the Fukushima nuclear accident on the surrounding area, we studied the pale grass blue butterfly Zizeeria maha, the most common butterfly in Japan. We here review our important findings and their implications. We found forewing size reduction, growth retardation, high mortality rates, and high abnormality rates in the field and reared samples. The abnormality rates observed in September 2011 were higher than those observed in May 2011 in almost all localities, implying transgenerational accumulation of genetic damage. Some of the abnormal traits in the F1 generation were inherited by the F2 generation. In a particular cross, the F2 abnormality rate scored 57%. The forewing size reduction and high mortality and abnormality rates were reproduced in external and internal exposure experiments conducted in our laboratory using Okinawa larvae. We observed the possible real-time evolution of radiation resistance in the Fukushima butterflies, which, in retrospect, indicates that field sampling attempts at the very early stages of such accidents are required to understand the ecodynamics of polluted regions. We propose, as the postulates of pollutant-induced biological impacts, that the collection of phenotypic data from the field and their relevant reproduction in the laboratory should be the basis of experimental design to demonstrate the biological effects of environmental pollutants and to investigate the molecular mechanisms responsible for these effects.

The meltdown of the Fukushima Dai-ichi Nuclear Power Plant (NPP) caused the release of a massive quantity of artificial radionuclides into the environment. The accident was the second-largest nuclear accident in history. Several studies have measured radionuclide activity concentrations in environmental samples from the contaminated regions (Hashimoto et al. 2012; Koarashi et al. 2012; Hosoda et al. 2013) and simulated deposition and distribution of artificial radionuclides from the NPP (Garnier-Laplace et al. 2011; Kinoshita et al. 2011; Yasunari et al. 2011; Buesseler et al. 2012; Fisher et al. 2013). Other studies have demonstrated the accumulation of these radionuclides in the bodies of wild and domesticated organisms and in agricultural products (Kakiuchi et al. 2012; Sasaki et al. 2012; Tagami et al. 2012; Fukuda et al. 2013; Hayama et al. 2013; Mizuno and Kubo 2013; Nakanishi and Tanoi 2013). However, scientific reports on the biological impacts of this accident are still scarce.

Evaluations of animal abundance in the highly contaminated area conducted shortly after the accident indicated that the abundance of butterflies decreased, as did, to a lesser extent, the abundance of cicadas (Møller et al. 2012; Mousseau and Møller 2012a, 2012b; Møller et al. 2013), which suggests that the radionuclides released might have had a fatal effect on these insects. Consistent with this finding, phenotypic changes in organisms have been reported in a lycaenid butterfly known as the pale grass blue butterfly, Zizeeria maha (Hiyama et al. 2012a, 2013; Møller and Mousseau 2013). To the best of our knowledge, this is the only study to date that has conclusively demonstrated phenotypic changes in organisms that are highly likely to have been caused by the Fukushima nuclear accident. In the present paper, we describe this study on the pale grass blue butterfly, in accordance with our research philosophy. Readers are encouraged to refer to the original paper (Hiyama et al. 2012a), the communication paper (Hiyama et al. 2013), and the commentary paper (Møller and Mousseau 2013) for additional information.

Basic Research Philosophy

What Is Most Important about Fukushima Research?

The primary purpose of Fukushima research is to know exactly what has been happening to organisms living in the polluted area after the nuclear accident at the Fukushima Dai-ichi NPP that occurred on 11 March 2011. We emphasize the importance of phenotypic changes found in the field (or in their progeny in the laboratory) at the individual (i.e., organismal) level because phenotypes, including pathological ones, determine the fitness or well-being of an individual and thereby affect reproductive performance. This in turn directly affects subsequent generations. That is, phenotypic impacts are health effects that could threaten the existence of individuals and change the fate of future populations. The results at the individual level can then be used to infer effects at the population level and similarly at the cellular and molecular levels.

Without a doubt, field work to collect organisms and laboratory work to examine their phenotypic abnormalities are of primary importance in understanding the effects of the Fukushima accident on the surrounding area. Among the various types of field studies needed are efforts to detect phenotypic changes that might have been caused by the accident. However, the phenotypic changes observed could have been caused by many other environmental factors. What researchers can say from the field data is at most that a correlation (or a coincidence) exists between the detected abnormal phenotypes and the level of radionuclide contamination. A cause-and-effect relationship is difficult to prove by these field studies. Repeated field observations are of great importance, and additionally, experimental confirmation to reproduce the situation in the Fukushima area in the laboratory is required to support any supposition of a possible cause-and-effect relationship. The mechanisms of such a relationship can then be studied at the molecular level using a reductionist approach.

To be sure, it is entirely possible to perform molecular studies using model systems such as cell cultures or mouse systems, and such studies contribute to the body of knowledge on radiation biology, but their relevance to the Fukushima case should be demonstrated separately. It is not surprising to find that molecular reductionism without knowledge of the phenotypic phenomena to be tested has no relevance to what is really happening in the Fukushima area.

Prompt Field Work and a Controversy over the Original Paper

When we published our first paper on this subject in Scientific Reports (Hiyama et al. 2012a), it received immediate attention from the international media and readers from all over the world. Surprisingly, some established scientists who are not familiar with insect physiology and butterfly biology criticized our prompt field work as being “too fast,” meaning that “more careful examination is required” and thus that the findings were “not trustworthy.” Although we agree that more studies are required, our data demonstrated that prompt field work is one of the most crucial elements in this type of research if the changing ecodynamics in the Fukushima area are to be properly evaluated. A very important lesson from the Chernobyl accident is that the lack of field work in some years after the accident made it difficult for researchers to accurately evaluate the biological impacts of the accident, resulting in much confusion over this issue (Møller and Mousseau 2006).

Some of the controversy over our original study was described in the News section of Nature (Callaway 2013). After a fair introduction of our work, the article states: “Other scientists take issue with the reports of ecological harms from Fukushima. They say that Otaki’s research is flawed, because wing shape and other butterfly traits vary naturally with geography.” This criticism has no scientific basis, although such a possibility should have been discussed more carefully in the original paper. We have evidence that these butterfly traits do not vary naturally with geography (see Hiyama et al. 2013; Iwata et al. 2013; and future research papers of ours). We replied to various comments from readers in BMC Evolutionary Biology (Hiyama et al. 2013).

Science and Politics

One of our positions must be clarified here to avoid unnecessary and unproductive discussions. The Nature News article (Callaway 2013) may be wrongly interpreted to imply that we (scientists who study this topic seriously) are mad scientists who do not care about people living in the Fukushima area (Steen and Wayne 2013), partly because our data may “scare people” there. As Steen and Wayne (2013) indicate, we (the authors) have no intention of scaring people living in the Fukushima area. Many people living there wish to know the truth. We have not received any complaints from those who live in the Fukushima area. Rather, we have received thank-you messages and even offers of help and donations from them. This is largely because of an almost complete paucity of scientific data regarding what has been happening in organisms living in the contaminated area. It is this situation that most likely scares people more than anything. We propose that scientists should investigate the truth first and then, as human beings, discuss politics based on scientific facts. This should be performed in the interest of helping (not scaring) people living in the contaminated area.

Basic Research Design

The Pale Grass Blue Butterfly: The Indicator Species of Choice

We are interested in the formation of the color patterns of butterfly wings, and for that purpose, we have established a rearing system for the Japanese pale grass blue butterfly (Hiyama et al. 2010). We intend to use this butterfly as a lycaenid model organism for studies of developmental physiology and genetics. We are also interested in the evolution of this butterfly’s color patterns (Buckley et al. 2010; Otaki et al. 2010). We have also been considering using this butterfly as an environmental indicator (Hiyama et al. 2012b). At the time of the nuclear accident, we were studying this butterfly, and this is why we were able to study the effects of the accident on this species promptly.

There are many advantages to using the pale grass blue butterfly as a model organism (Hiyama et al. 2010, 2013). It is one of the smallest butterflies in Japan. Its forewing size, from the base to the apex, is a little more than 1.0cm. The flying ability of the pale grass blue butterfly is one of the lowest among butterflies, which means that this butterfly does not fly long distances. Its small size, low flying ability, and short life cycle (approximately 1 month) allow researchers to breed hundreds or thousands of individuals in the laboratory to perform rigorous statistical analyses. Furthermore, the overall color pattern of the pale grass blue butterfly, that is, distinguishable black dots against a light gray background, is very simple and amenable to rigorous analysis. The pale grass blue butterfly is likely the most successful, abundant, and common butterfly species in Japan, which means that it is easy to collect in the field. This species of butterfly adapts quickly to various environments. In fact, the pale grass blue butterfly shares living space with humans not only in rural areas but also in highly urbanized areas (Figure 1). It has recently been demonstrated by citizen science data that the pale grass blue butterfly is the most commonly found butterfly in Tokyo (Washitani et al. 2013). Biogeographically, it is found in most places throughout Japan, except Hokkaido (Shirôzu 2000). This popularity is reflected in its Japanese name, Yamato shijimi, meaning the Japanese lycaenid. The pale grass blue butterfly lives on or just above the surface of the ground throughout its life stages because its host plant, Oxalis corniculata (Katabami in Japanese), is small (its height is approximately 10cm). Thus, this butterfly is useful in monitoring ground surface environments and the effects of pollution. Interestingly, this butterfly is monophagous. That is, it eats only a single species of plant, which simplifies the field work, the rearing method, and the interpretation of field and experimental results (Hiyama et al. 2010).

Figure 1.

Urban habitat of the pale grass blue butterfly. Dozens of pale grass blue butterflies, Zizeeria maha, were found on garden flowers. Pictures were taken in Naha, Okinawa, on 28 August 2013. (A) Flower gardens along a sidewalk in the most urbanized area of Okinawa. (B) Flowers of a Helenium cultivar and a garden space that harbors the host plant Oxalis corniculata for the pale grass blue butterfly. This picture show a part of the sidewalk shown in (A). (C and D) An adult butterfly visiting a flower.

Figure 1.

Urban habitat of the pale grass blue butterfly. Dozens of pale grass blue butterflies, Zizeeria maha, were found on garden flowers. Pictures were taken in Naha, Okinawa, on 28 August 2013. (A) Flower gardens along a sidewalk in the most urbanized area of Okinawa. (B) Flowers of a Helenium cultivar and a garden space that harbors the host plant Oxalis corniculata for the pale grass blue butterfly. This picture show a part of the sidewalk shown in (A). (C and D) An adult butterfly visiting a flower.

On the other hand, there may be some disadvantages at present to using this butterfly species as a model organism. First, there is no established pure line available, and hence, genetic results from crossing experiments may be variable and may not show a simple Mendelian inheritance (Hiyama et al. 2010; Iwata et al. 2013). Our crossing experiments suggest that it is not easy to establish a pure line in this species (Hiyama et al. 2010; Iwata et al. 2013). Second, almost no molecular data are yet available for this species. However, this could be addressed by genome sequencing of this butterfly species in the future. Fortunately, a few butterfly genomes have already been sequenced (Zhan et al. 2011; The Heliconius Genome Consortium 2012), and the information obtained from these butterfly genomes may be of use in sequencing the genome of the pale grass blue butterfly.

Fukushima and Okinawa: Contrasting Locations in Japan

Our experiments were performed in our laboratory in Okinawa. Okinawa is 1700 km from the Fukushima Dai-ichi NPP. This long distance from the NPP and minimal contamination by the accident were advantages in this research. Minimal contamination is not a trivial concern because it is necessary to perform experiments in a contaminant-free space. This is especially relevant when we examine F1 and F2 effects for the purposes of eliminating acute physiological effects on these generations and focusing on heritable effects. Irradiation experiments must also be performed in a contaminant-free space. An additional advantage is that Okinawa does not have any NPP nearby, which is exceptional for locations in Japan and eliminates the possible effect of low-dose exposure from artificial radionuclides that may be released from NPPs to surrounding environments. Indeed, Okinawa is the only place in Japan where no electricity is supplied from a nearby NPP.

Research Framework

We explain below our research framework, according to Scientific Reports (Hiyama et al. 2012a). Our research consisted of 4 sets of experiments. First, field sampling was conducted in the Fukushima area in May 2011, 2 months after the accident. This early data collection effort was highly important because we wanted to collect the first-voltine adults, that is, the first-generation adults of that year. These adults also represented the first generation after the accident that was directly exposed to radionuclides from the NPP as larvae. Then, in the laboratory, we obtained the F1 and F2 generations from the field-collected adults (defined as the P generation) to examine the possible influences on these generations. Because these F1 and F2 generations were obtained in Okinawa, which is a contaminant-free environment, and because they were reared under optimal conditions in our laboratory, phenotypic abnormalities found in the F1 and F2 generations may be attributed to heritable defects of their parents’ germ-line cells.

Second, we conducted another round of field sampling in September 2011, 6 months after the accident, collecting butterflies from the same localities at which butterflies were previously collected in May 2011. In this second round of sampling, we caught the fourth or fifth generation of the year, which was also the fourth or fifth generation after the accident. We obtained the F1 generation of these butterflies in Okinawa. These May and September results can be compared to check for cumulative transgenerational effects.

Third, we performed an external exposure experiment using 137Cs as a radiation source. Although many different radionuclides were released from the Fukushima Dai-ichi NPP, 137Cs appears to have been one of the major radiation sources. For this experiment, we used the Okinawa larvae. It is important to emphasize that the Okinawa larvae were almost certainly not affected by radionuclides from the Fukushima Dai-ichi NPP.

Fourth and most important, we performed an internal exposure experiment. For this experiment, leaves of the host plant were collected from the Fukushima area and were given to the Okinawa larvae, which were almost certainly not affected by artificial radionuclides from the Fukushima Dai-ichi NPP. This simple experiment was the most informative one in our study. We can safely assume that the larvae living in the contaminated area consumed contaminated leaves, although some of the short-half-life nuclides would have been extinct by the time of our leaf collection.

It should be noted that our conditions for the internal exposure experiment did not completely match what the pale grass blue butterflies in the Fukushima area experienced. External irradiation was experienced simultaneously with internal exposure in the field. Moreover, internal exposure can be caused by inhalation of air in addition to consumption of leaves.

It is important to identify which of the numerous phenotypic traits of this species were measured. For the adult butterflies captured in the field, we necessarily focused on external morphological traits, based on the assumption that external morphology can be used as an indication of environmental impacts. However, the field-caught adults were all survivors of the accident and therefore do not provide information about other individuals of the species that did not survive to the adult stage. To compensate for this lack of information, we obtained the F1 and F2 generations from the field-caught adults in the laboratory, and we examined the mortality rates of larvae and pupae. Additionally, for the F1 and F2 generations, we examined developmental physiological traits such as eclosion and pupation time (i.e., the time in days from egg to eclosion or pupation). These physiological traits can be examined only by obtaining offspring in the laboratory.

Concise Review of the Results

Sampling from the Fukushima Area and Obtaining the F1 and F2 Generations

We first measured the forewing size of the field-caught adults. Forewing size has been recognized as a reasonable indicator of the body size of a butterfly in lepidopterology. Measuring the head-to-tail length is imprecise because the abdomen can stretch, shrink, and curve considerably. Wings are the largest construct in butterflies and have a fixed size for a given individual. We found that the forewings of the adult butterflies collected from Fukushima, Motomiya, and Koriyama were smaller than those from northern and southern localities (Figure 2). Note that Fukushima, Motomiya, and Koriyama are located close together. When the forewing size was plotted against the ground radiation dose at the time of the butterfly sampling, we obtained a reasonable inverse correlation: The higher the ground radiation dose was, the smaller the forewing size was. Note that in addition to the smallness of the forewings, the variation in forewing size (as indicated by the standard deviation and the standard error) for the butterflies collected in these 3 cities were larger than those collected in the other localities. This greater variation in forewing size and the smallness of the forewings suggest that the environmental factor that affects the size may not be a general one, such as the climate, but rather a factor that varies in the microenvironment, such as pollution. Dose measurements at a given collection site in the Fukushima area occasionally varied by more than one order of magnitude within a several-meter range, which may be reflected in the large variation in the forewing sizes of the butterflies caught at the 3 localities near the NPP.

Figure 2.

Collection localities and forewing size reduction observed in the first-voltine males in 2011 (the first generation after the accident). (A) Ten collection localities around the Fukushima Dai-ichi NPP (Shiroishi City, Fukushima City, Motomiya City, Koriyama City, Hirono Town, Iwaki City, Takahagi City, Mito City, Tsukuba City, and Tokyo). Many artificial radionuclides, represented by cesium, are believed to be distributed in the region northwest of the NPP, including Fukushima, Motomiya, and Koriyama. However, earlier deposition, represented by iodine, is likely to be distributed in a concentric fashion, affecting the southern region, including Hirono, Iwaki, and Takahagi. (B) Forewing size of the first-voltine males collected in May 2011. Error bars indicate the standard deviation. The numbers (n) indicate the number of individuals subjected to measurements. (C) Scatter plot between the forewing size and the ground radiation dose. Error bars indicate the standard error. A reasonably high Pearson correlation coefficient r and reasonably small P value were obtained.

Figure 2.

Collection localities and forewing size reduction observed in the first-voltine males in 2011 (the first generation after the accident). (A) Ten collection localities around the Fukushima Dai-ichi NPP (Shiroishi City, Fukushima City, Motomiya City, Koriyama City, Hirono Town, Iwaki City, Takahagi City, Mito City, Tsukuba City, and Tokyo). Many artificial radionuclides, represented by cesium, are believed to be distributed in the region northwest of the NPP, including Fukushima, Motomiya, and Koriyama. However, earlier deposition, represented by iodine, is likely to be distributed in a concentric fashion, affecting the southern region, including Hirono, Iwaki, and Takahagi. (B) Forewing size of the first-voltine males collected in May 2011. Error bars indicate the standard deviation. The numbers (n) indicate the number of individuals subjected to measurements. (C) Scatter plot between the forewing size and the ground radiation dose. Error bars indicate the standard error. A reasonably high Pearson correlation coefficient r and reasonably small P value were obtained.

It should also be noted that the temperature–size rule states that in low-temperature regions, animal body size is relatively large. Our observation of the Fukushima samples contradicts this rule: The Fukushima population of the pale grass blue butterfly was found to be smaller than its southern and northern populations. Nevertheless, many critics have argued that this size reduction is due to a temperature or latitude effect. In response to this criticism, samples from various localities throughout Japan were collected, and we are now analyzing them.

In addition to the forewing size, we examined morphological abnormalities in the field-collected samples and the laboratory-reared F1 samples. Morphological abnormalities were found in the legs, palpi, thoraxes, compound eyes, and wings (Figure 3). Some of these abnormalities were found to occur symmetrically on both the right and left sides (in approximately 10% of abnormal individuals), and many were found on either the right or the left side (in approximately 90% of abnormal individuals). Asymmetric abnormalities may imply that these individuals are genetic mosaics found primarily in the F1 generation (Iwata et al. 2013). Genetic mosaics are produced by single-strand DNA damage because the damage in sperm is repaired only after the first zygotic division (Grigliatti 1998; St Johnston 2002). However, asymmetric abnormalities were also found in the F2 generation in our crossing experiments, implying that differences in the expressivity of genes may also be involved. Even within a normal range of phenotypes, asymmetric ones are considered less fit (Møller and Swaddle 1997). Importantly, in all localities from which we collected butterflies except Mito, the September abnormality rates (in the fourth or fifth generation after the accident) were always higher than the May abnormality rates (in the first generation after the accident) in both the P and F1 generations (see Figure 9A, B in Hiyama et al. 2013), implying transgenerational accumulation of genetic damage after 4 or 5 generations.

Figure 3.

Symmetric and asymmetric morphological abnormalities. Symmetric and asymmetric ones are shown in (AM) and (NR), respectively. Arrows indicate abnormal structures. All abnormal specimens are of the F1 or F2 generation except (B) and (M), which are the field-collected individuals (defined as the P generation). All abnormal specimens were obtained from the samples caught in May or September 2011. Normal morphology was shown in (SV). (A) Dented compound eyes. Fukushima F1 (May). Scale bar indicates 500 μm. (B) Dented compound eyes. Shiroishi P (May). (C and D) Deformation of compound eyes. The left eye (C) is severely deformed, whereas the right eye (D) of the same individual is lightly affected. Hirono F1 (May). (E) Tumors in the thorax. Takahagi F1 (September). (F) Unexpanded small forewings. Hirono F1 (May). (GM) Symmetric color pattern changes. Abnormal sites are boxed and indicated by arrows. Scale bar indicates 1cm, which applies to (GR) and (V). (GI) Spot fusion. Hirono F1 (May). (J) Spot fusion. Iwaki F1 (May). (K) Spot dislocation and loss. Motomiya F1 (May). (L) Spot elongation and fusion. Hirono F1 (September). (M) Incomplete scale development. Fukushima P (September). (NR) Asymmetric color pattern changes. Abnormal sites are boxed and indicated by arrows. (N and O) Spot fusion. Fukushima F2 (May). (P) Spot elongation. Hirono F2 (May). (Q) Spot fusion. Iwaki F1 (May). (R) Extra spot. Iwaki F1 (May). (SV) Normal morphology of eyes (S, T; the F1 generation, Nikaho, Akita Prefecture), thorax (U; the F1 generation, Nikaho, Akita Prefecture), and wing color pattern (V; the F1 generation, Okinawa). Scale bars indicate 500 μm. Some of these specimens are also shown in Hiyama et al. (2012a, 2013).

Figure 3.

Symmetric and asymmetric morphological abnormalities. Symmetric and asymmetric ones are shown in (AM) and (NR), respectively. Arrows indicate abnormal structures. All abnormal specimens are of the F1 or F2 generation except (B) and (M), which are the field-collected individuals (defined as the P generation). All abnormal specimens were obtained from the samples caught in May or September 2011. Normal morphology was shown in (SV). (A) Dented compound eyes. Fukushima F1 (May). Scale bar indicates 500 μm. (B) Dented compound eyes. Shiroishi P (May). (C and D) Deformation of compound eyes. The left eye (C) is severely deformed, whereas the right eye (D) of the same individual is lightly affected. Hirono F1 (May). (E) Tumors in the thorax. Takahagi F1 (September). (F) Unexpanded small forewings. Hirono F1 (May). (GM) Symmetric color pattern changes. Abnormal sites are boxed and indicated by arrows. Scale bar indicates 1cm, which applies to (GR) and (V). (GI) Spot fusion. Hirono F1 (May). (J) Spot fusion. Iwaki F1 (May). (K) Spot dislocation and loss. Motomiya F1 (May). (L) Spot elongation and fusion. Hirono F1 (September). (M) Incomplete scale development. Fukushima P (September). (NR) Asymmetric color pattern changes. Abnormal sites are boxed and indicated by arrows. (N and O) Spot fusion. Fukushima F2 (May). (P) Spot elongation. Hirono F2 (May). (Q) Spot fusion. Iwaki F1 (May). (R) Extra spot. Iwaki F1 (May). (SV) Normal morphology of eyes (S, T; the F1 generation, Nikaho, Akita Prefecture), thorax (U; the F1 generation, Nikaho, Akita Prefecture), and wing color pattern (V; the F1 generation, Okinawa). Scale bars indicate 500 μm. Some of these specimens are also shown in Hiyama et al. (2012a, 2013).

Next, we examined eclosion time, which is the number of days from egg production to eclosion. Individuals from Tsukuba eclosed quickly, but others, such as those from Fukushima and Hirono, were slower to eclose (Figure 4A), indicating growth retardation. However, there were individual variations in eclosion time even within a single population, which may have arisen from genetic heterogeneity in a population or from differences in the microenvironment experienced by the parents (i.e., epigenetic maternal effect).

Figure 4.

Eclosion dynamics and abnormality rates. (A) Eclosion over time for the May 2011 F1 samples from different localities. (B) Average eclosion time versus distance from the NPP for the May 2011 F1 samples. Error bars represent standard error. (C) Abnormality rate of adult appendages versus distance from the NPP for the May 2011 F1 generation. (D) Abnormality rate versus ground radiation dose for the September 2011 P generation.

Figure 4.

Eclosion dynamics and abnormality rates. (A) Eclosion over time for the May 2011 F1 samples from different localities. (B) Average eclosion time versus distance from the NPP for the May 2011 F1 samples. Error bars represent standard error. (C) Abnormality rate of adult appendages versus distance from the NPP for the May 2011 F1 generation. (D) Abnormality rate versus ground radiation dose for the September 2011 P generation.

We observed a reasonable correlation between eclosion time and the distance from the NPP: The closer a locality was to the NPP, the longer the eclosion time was for individuals from that locality (Figure 4B). The abnormality rate of appendages in the F1 generation was also highly correlated with the distance from the NPP (Figure 4C). In contrast to these distance correlations, which were detected only in the F1 generation, the correlations with the ground radiation dose were detected only in the P generation (Figure 4D and Table 1). The possible reasons for these results are discussed later (see Further Discussion).

Table 1

Correlation analysis of important traits in the P and F1 generations from the May and September 2011 samples

Generation Traits Correlation Correlation 
May 2011 September 2011 
Forewing size Ground radiation dose None 
Abnormality rate (adults) None Ground radiation dose 
Abnormality rate (adult appendages) None Ground radiation dose 
F1 Forewing size None None 
F1 Abnormality rate (all developmental stages) None Distance from the NPP 
F1 Abnormality rate (adults) Distance from the NPP Distance from the NPP 
F1 Abnormality rate (adult appendages) Distance from the NPP None 
F1 Eclosion time Distance from the NPP Distance from the NPP 
Generation Traits Correlation Correlation 
May 2011 September 2011 
Forewing size Ground radiation dose None 
Abnormality rate (adults) None Ground radiation dose 
Abnormality rate (adult appendages) None Ground radiation dose 
F1 Forewing size None None 
F1 Abnormality rate (all developmental stages) None Distance from the NPP 
F1 Abnormality rate (adults) Distance from the NPP Distance from the NPP 
F1 Abnormality rate (adult appendages) Distance from the NPP None 
F1 Eclosion time Distance from the NPP Distance from the NPP 

Correlations are indicated in this table when Pearson correlation coefficient |r| > 0.7.

Inheritance of Abnormal Traits

Despite the fact that we used completely normal males and females to produce the F1 generation, F1 individuals exhibited high abnormality rates (see Figure 9B in Hiyama et al. 2013), suggesting germ-line damage in the P generation. We then examined whether the abnormalities found in the F1 generation could be inherited by the F2 generation. We chose abnormal F1 females from the collection localities (excluding Shiroishi, from which no abnormal F1 females were obtained), and we crossed them with normal F1 males from Tsukuba, the farthest city from the NPP among the sampled. A relatively small number of eggs was produced (see Supplementary Table 5 in Hiyama et al. 2012a), indicating a decrease in reproductive performance. In the F2 generation, abnormality rates of more than 40% were recorded in progeny from Takahagi, Iwaki, and Motomiya, although these results are not surprising because their female parents had abnormal traits. The abnormal individuals of the F2 generation were classified into 3 categories: identical to F1, homologous with F1, and not related to F1. In a particular case, “Iwaki1,” we observed an abnormality rate of 56.9% in the F2 generation (Figure 5; see Supplementary Table 6 in Hiyama et al. 2012a). The inheritance of this abnormal trait in approximately half of the progeny may suggest a simple Mendelian inheritance. Although only 67% of the individuals of the abnormal F2 generation had phenotypes identical or homologous to their F1 parents, this might have occurred because of the heterogeneity of their genetic background.

Figure 5.

Heritability of an F1 trait by the F2 generation. (A) A schematic diagram of genetic crosses. In this diagram, an abnormal female from the Iwaki F1 generation was crossed with a normal male from the Tsukuba F1 generation, producing the F2 generation. AR indicates the abnormality rate. (B) An F1 female from Iwaki used for a cross. This female has elongated spots. See Figure 3L for a similar (but not identical) phenotype. Scale bar indicates 1cm. (CD) The F2 generation individuals produced from the cross. They are phenotypically similar to the female parent.

Figure 5.

Heritability of an F1 trait by the F2 generation. (A) A schematic diagram of genetic crosses. In this diagram, an abnormal female from the Iwaki F1 generation was crossed with a normal male from the Tsukuba F1 generation, producing the F2 generation. AR indicates the abnormality rate. (B) An F1 female from Iwaki used for a cross. This female has elongated spots. See Figure 3L for a similar (but not identical) phenotype. Scale bar indicates 1cm. (CD) The F2 generation individuals produced from the cross. They are phenotypically similar to the female parent.

These results suggest that the F1 abnormalities were inherited by the F2 generation, but we admit that our experimental evidence for genetic inheritance is not conclusive. Similarly, we do not have sufficient evidence of heritable DNA damage. Obviously, more studies are necessary to examine this point.

In the F1 and F2 generations, we observed diverse color pattern aberrations: spot fusion, extra spots, spot elongation, spot dislocation, spot loss, incomplete scale development, and disorganized spot arrays (Figure 3). These abnormal individuals often had other morphological and behavioral abnormalities and reduced reproductive performance. These various and unpredictable changes are in sharp contrast to predictable color pattern changes induced by temperature shock (TS), which are not associated with other abnormalities (Otaki et al. 2010; Hiyama et al. 2012b). On the other hand, these various and unpredictable changes were similar to those found in a mutagenesis experiment (Iwata et al. 2013).

However, one of the abnormal phenotypes (spot elongation) obtained from Iwaki (Figure 5) warrants discussion. This phenotype was inherited from the F1 generation by the F2 generation. Although it is different from the TS-induced phenotypes obtained under our standard conditions and those observed in the field (Otaki et al. 2010, Hiyama et al. 2012b), this spot elongation bears some resemblance to a TS-induced phenotype. Individuals exhibiting this abnormal phenotype may be mutants for genes that are responsible for the TS-induced phenotypes.

External Exposure Experiment

We used the Okinawa larvae for the external exposure experiment. Using 137Cs as a radiation source, we used cumulative 22–55 mSv (180–280h) conditions (dose rate: 0.12–0.20 mSv/h) and 57–125 mSv (177–387h) conditions (dose rate: 0.32 mSv/h). We observed abnormalities in legs, antennae, palpi, eyes, and wings, including color pattern changes. Equally important, we observed forewing size reduction (see Figure 5b in Hiyama et al. 2012a) and a dose-dependent decrease in the survival rate (see Figure 5c in Hiyama et al. 2012a). Control groups (not exposed to radiation) that were genetically identical to the irradiated groups did not exhibit any change in survival rate under our standard conditions.

The level of exposure used was very high in comparison with that at the collection localities. In our measurements, the highest level of radiation activity measured at our collection localities, 3.09 μSv/h, was recorded at Motomiya in May 2011 (Hiyama et al. 2012a). However, it is certain that localities much closer to the NPP are contaminated at levels at least as high as the level of our irradiation experiments. For example, a survey on 25 July 2013 recorded a ground radiation dose of 320 μSv/h in Okuma in Fukushima Prefecture (Fujita et al. 2013). Levels measured earlier would have been even higher. It is worth mentioning that the purposes of this external exposure experiment were to examine whether abnormalities similar to those of field-caught individuals and their progeny would be observed under experimental conditions and to determine whether any dose-dependent mortality change would be observed. We also must consider that wild populations are more vulnerable to radiation than the populations used in the controlled experiment (Beresford and Copplestone 2011; Garnier-Laplace et al. 2013).

Internal Exposure Experiment

We collected leaves from 5 localities—Ube, Hirono, Fukushima, Iitate (the montane region), and Iitate (flatland)—for the internal exposure experiment. These leaves were given to the Okinawa larvae in the laboratory in Okinawa, a contaminant-free environment. The 137Cs and 134Cs activities in the collected leaves are shown in Supplementary Table 8 in Hiyama et al. (2012a). For example, the 137Cs activity of the leaves from Fukushima was 7900 times greater than that of the leaves from Ube, a noncontaminated city where a relatively high level of natural radiation has been detected.

We observed morphological abnormalities in antennae, palpi, eyes, and wings. Many larvae that ate leaves from Hirono, Fukushima, and Iitate died, at rates that depended on the cesium activity in the leaves (see Figure 5d in Hiyama et al. 2012a). The total abnormality rate (defined as the sum of the numbers of dead and abnormal individuals divided by the number of larvae reared) appeared to be saturated at the radiation level of Fukushima (Figure 6A). Among the abnormal individuals, the proportion of individuals with abnormalities in appendages seemed to be largely constant (approximately 40%), whereas the proportion of individuals with color pattern abnormalities increased with the radiation activity in the host plant leaves (Figure 6B). It is important to note that we observed a decrease in the forewing size of individuals given leaves from Fukushima and Iitate (flatland), compared with the forewing size of individuals given leaves from Ube (data not shown, but see Figure 5f in Hiyama et al. 2012a).

Figure 6.

Abnormality rates obtained from the internal exposure experiments. (A) Total abnormality rates. The total abnormality rate is defined as the sum of the numbers of dead and abnormal individuals divided by the number of larvae reared. (B) Categorization of abnormalities among the surviving adults.

Figure 6.

Abnormality rates obtained from the internal exposure experiments. (A) Total abnormality rates. The total abnormality rate is defined as the sum of the numbers of dead and abnormal individuals divided by the number of larvae reared. (B) Categorization of abnormalities among the surviving adults.

Possible Evolution of Radiation Resistance

We detected a possible evolution of radiation resistance, which was suggested by the change in the adult abnormality rate over time (data not shown). We have performed field surveys twice a year since 2011, and thus we have accumulated data on the abnormality rates of field-caught adult samples. The abnormality rate peaked in September 2011 and then decreased sharply. It appears that we are now observing real-time evolution of radiation resistance. This is still a tentative conclusion, and precise data will be published elsewhere in the future.

Here, we present survival curves for the F1 generation from the May 2011 and September 2011 samples (Figure 7A,B). The original numerical data for these survival curves have been published in Supplementary Table 3 in Hiyama et al. (2012a). In May 2011, the F1 individuals from Iwaki, Hirono, and Motomiya had lower survival rates than those from Fukushima, Mito, Tsukuba, and Takahagi, which had similar survival rates. The Shiroishi F1 individuals had the highest survival rate. In September 2011, the F1 individuals from Kobe, Takahagi, Fukushima, and Iwaki had similarly high survival rates. Because Kobe is far from the NPP, the survival rates of these 4 local populations can be considered to be within the normal range. Compared with the May 2011 data, the September 2011 data appeared to indicate an increase in the survival rate of the Iwaki population, not much change in the survival rate of the Motomiya population, and a decrease in the survival rate of the Hirono population. The decrease in the survival rate of the Hirono population may be due to Hirono being the closest of the localities we studied to the NPP. The Hirono population appeared to be affected more severely than the populations at the other localities by the early deposition of iodine and other short-half-life artificial radionuclides (Chino et al. 2011; Torii et al. 2013).

Figure 7.

Survival curves of the F1 generation. (A) The F1 generation of the May 2011 samples. (B) The F1 generation of the September 2011 samples.

Figure 7.

Survival curves of the F1 generation. (A) The F1 generation of the May 2011 samples. (B) The F1 generation of the September 2011 samples.

Further Discussion

Importance of Phenotypic or Physiological Changes

To summarize our results, we observed reduced and more variable forewing size in the field samples, and we observed growth retardation in the reared samples. Furthermore, we observed morphological abnormalities, increases in abnormality rates from May to September 2011, and inheritance of abnormal traits. In the external and internal irradiation experiments, we observed decreases in survival rate, reduced forewing size, and morphological abnormalities. We obtained similar results in the external and internal exposure experiments and in the field and laboratory studies.

It is important to stress that our research was guided by hypotheses based on field work at the organismal level. We started examining phenotypic changes immediately after the accident. Without these phenotypic data, most laboratory work would simply cause confusion because it cannot be focused on explaining what happens at the organismal level.

Correlation and Heritability

It is likely that cesium and iodine were not deposited together in the same way in the Fukushima accident. Early deposition of iodine (and other short-half-life radionuclides) occurred in a concentric fashion from the NPP, whereas late deposition of cesium occurred mainly toward the regions northwest of the NPP (Chino et al. 2011; Torii et al. 2013). Because the ground radiation dose at the time of our dose measurements primarily reflected cesium activity, the correlations of the observed ground radiation dose with the forewing size and abnormality rates in the P generation most likely indicate that changes were caused primarily by somatic cellular damage from long-term low-level exposure to cesium. This interpretation is consistent with the results of our external and internal exposure experiments, which produced forewing size reduction and abnormal phenotypes. In contrast, correlations were observed between the distance from the NPP and the eclosion time and the abnormality rates in the F1 generation only, which indicates that these changes were mainly caused by germ-line damage from relatively high-level exposure from the early deposition of iodine and other short-half-life nuclides. It appears that genetic damage caused by radioactive iodine immediately after the accident is inherited in the wild population of this butterfly, which was detected in the September F1 generation. Difficulty in detecting the possible iodine effect in the P generation is likely because only relatively healthy individuals were collectable in the field. Relatively mild abnormalities in the P generation may mostly be a result of the cesium effect on somatic cells.

Together with the fact that the abnormality rates were higher in September 2011 than in May 2011 in both the P and F1 generations for all localities surveyed (except the P generation from Mito), it is reasonable to conclude that germ-line genetic damage was mostly introduced immediately after the accident and was inherited and accumulated by the subsequent generations. The results of our crossing experiment are also consistent with the heritable nature of radiation-induced germ-line damage. Many of the morphological abnormalities observed in the P generation may have been somatic or physiological damage caused by cesium activity, although iodine effects may have contributed to the damage.

Internal Exposure as a Key Experiment

Our simple experimental design for internal exposure was considered to be the best way to reproduce what the pale grass blue butterflies in the Fukushima area experienced after the accident and to determine what happened to them. To be sure, it is possible to take a reductionist approach by feeding foods containing radioactive cesium to larvae. However, if the goal is to determine what happened in Fukushima after the accident, our experimental conditions may be better suited than the reductionist approach, considering that many different radioactive pollutants may be contained in the host plant leaves in the Fukushima area, that these pollutants together may cause the biological effects observed, and that the radionuclides released may be in the form of unique physical and chemical structures, such as microparticles (Adachi et al. 2013), that could cause additional biological effects, independent of radiation.

One could argue that the effects observed may have been caused not by cesium radioactivity but rather by the chemical toxicity of cesium itself. However, this is unlikely. The concentrations of radioactive cesium (137Cs and 134Cs) as pollutants were much lower than the natural concentrations of nonradioactive cesium (133Cs) in the environment. The mass of 100 Bq of 137Cs was only 31.2 pg, and 100 Bq of 134Cs is approximately 6.7% of 100 Bq of 137Cs (Shozugawa 2013). The toxicity of cesium is rather low, at least in mammals (Johnson et al. 1975; Pinsky et al. 1981). In a study of pupae of the butterfly Junonia orithya, injection of a cesium chloride solution (1.0M) did not cause any color pattern changes and did not seem to be toxic in comparison with lithium chloride, sodium chloride, and water (Dhungel and Otaki 2009; Otaki JM, unpublished data).

Low-Dose-Induced Abnormality and Mortality

Insects have been said to be relatively resistant to radiation. Many critics have argued that insects are much more resistant to radiation than our results suggest (Hiyama et al. 2013). However, our experimental conditions (i.e., irradiation dose, duration, timing, and so on) were very different from those of exposure experiments that have been performed before. Our study examines “long-term low-dose” (chronic) exposure as opposed to “short-term high-dose” (acute) exposure. Moreover, the pale grass blue butterfly is different from other species that have been used in previous exposure experiments. It is important to stress that insects are known to be highly diverse in their morphology, life history, and physiology. It is understandable that radiation sensitivity is fundamentally species dependent. Radiation sensitivity could also be population dependent within a species or even individual dependent within a population. Our results serve as a reminder that our knowledge of the radiation sensitivity of organisms is far from perfect.

We observed diverse color pattern changes and abnormalities in the Fukushima samples. Similar types of color pattern changes were observed in a mutagenesis study (Iwata et al. 2013), which supports the idea that at least some of the Fukushima abnormalities originated from DNA damage. In insects at least, DNA damage in sperm is repaired after the first zygotic division (Grigliatti 1998; St Johnston 2002). This means that one daughter cell receives a nondamaged strand of DNA and the other daughter cell receives a damaged strand of DNA, resulting in a mosaic mutant. If sperm cannot inherently repair any damage, the occurrence of mosaic mutants may be unavoidable in the areas around Fukushima where irradiation continues, even though insects can evolve more efficient DNA repair systems relatively quickly.

Changes in Abnormality and Survival Rates over Time

Although this is a tentative conclusion, it appears that we witnessed the real-time evolution of radiation resistance in the pale grass blue butterfly. The pale grass blue butterflies around Fukushima seem to have adapted to the contaminated environment in approximately 2 years. It is important to note that not all individuals died at the same age even under a given irradiation condition in the external and internal exposure experiments. Among those that survived, some were sick or weak, and some were very robust or strong. This variation provides the basis of the evolution of radiation resistance. In the field, this inherent variation at the population level may be enhanced by the introduction of random genetic mutations caused by irradiation, although most of the induced mutations may be harmful or functionally neutral.

Interestingly, it appears that even at the peak time of abnormalities (i.e., September 2011), survival rates were partly restored in all the local populations examined, except the Hirono population, compared with the survival rates of May 2011. The decline of the survival rate in the Hirono population may suggest that it was affected severely by the deposition of iodine (and other short-half-life nuclides) immediately after the nuclear accident.

Generalization of the Evaluation System

Here, we stress methodologically unique aspects of this study that may not be found in other studies. We monitored butterflies in several localities in the Fukushima area, beginning immediately after the accident. We evaluated the possible radiation effects in the F1 and F2 generations in experiments carried out in Okinawa, a contaminant-free environment. Furthermore, we reproduced the field and rearing results by external and internal experiments using the Okinawa larvae and the contaminated leaves.

Based on our experimental design described above, we propose “the postulates of pollutant-induced biological impacts” below, which use Koch’s postulates of infectious disease and their advanced version as a reference (Evans 1976) to prove that an environmental pollutant (or a group of pollutants) in question is a causal factor. Although we used the pale grass blue butterfly, other organisms can be tested similarly according to our postulates. The postulates can be applied to other nuclear accidents that will occur in the future and also to other artificial environmental pollution by industrial and agricultural chemicals.

  1. Spatial relationship: Organisms living around the sources of artificial pollutants suffer from higher incidence of death or abnormality than those living farther away. A spatial gradient of incidence of death or abnormality may exist.

  2. Temporal relationship: Organisms living around the sources of pollutants suffer from higher incidence of death or abnormality after the release of the pollutants than before the release of the pollutants. If the level of the pollutant changes temporally, a change in the incidence of death or abnormality may follow.

  3. Direct exposure: Pollutants should be detected from organisms suffering from death or abnormality or from their foods, collected from around the source of the pollutants.

  4. Phenotypic variability or spectrum: A variety or spectrum of abnormal phenotypes should be detected in organisms living around the source of artificial pollutants after the release of the pollutants into environment.

  5. Experimental reproduction (external exposure): When healthy organisms that have never been exposed to the pollutants are exposed to them externally, the external exposure should cause death or abnormality at rates similar to those observed in the field.

  6. Experimental reproduction (internal exposure): When the pollutants are introduced into healthy organisms that have never been exposed via foods that were collected from around the source of the pollutants, the internal exposure should cause death or abnormality at rates similar to those observed in the field.

Some comments are necessary here. With respect to postulate 4, a specific nonradioactive chemical pollutant may simply inhibit a specific molecular pathway, in which case a spectrum (but not a variety) of abnormal phenotypes may arise, depending on the sensitivity of the organisms. With respect to postulates 5 and 6, “healthy organisms” should not have already evolved resistance through past continuous exposure to the same or similar pollutants. In addition, with respect to postulates 5 and 6, because organisms are exposed to the pollutants both externally and internally in the field, a combination of external and internal exposures may replicate the field situation more faithfully. Further reductionist identification of harmful pollutants may be carried out by adding a substance of interest (e.g., radioactive iodine or cesium in the case of the Fukushima accident) to food collected from a contaminant-free environment and feeding this modified food to healthy organisms. However, in laboratory experiments, reproduction of physical and chemical structures of pollutants, such as microparticles of cesium (Adachi et al. 2013), may not be easy to accomplish.

An additional postulate, although it is one that we think is not required, may be that elimination of the pollutants should restore the health of the organisms. If DNA is damaged, elimination does not restore health completely. However, when the pollutants are eliminated, physiological stress associated with the pollutants is eliminated, which may result in significant improvement of the fitness of the individual. At the population level and through generations, health restoration will result in most cases. However, similar results may occur even in a continuously polluted environment because organisms can evolve to adapt to that environment, which somewhat complicates the issue.

In “the postulates of pollutant-induced biological impacts” listed above, we did not include any demonstration of genetic impact. Demonstration of genetic damage would require establishing its inheritance by examining the F1, F2, and preferably F3 generations. Such rearing experiments may be performed using field-caught samples and using irradiated samples as the parent generation. Furthermore, molecular identification of DNA damage may be pursued by genomic analysis.

Funding

Takahashi Industrial and Economic Research Foundation, Tokyo, Japan .

Acknowledgments

We thank TY Steen and ML Wayne for inviting us to make a presentation to the Society for Molecular Biology and Evolution and for giving us an opportunity to write this paper. We thank amateur Japanese lepidopterists who helped us to collect pale grass blue butterflies from many localities in Japan. We also thank many people who contributed to studies published in Hiyama et al. (2012a, 2013).

References

Adachi
K
Kajino
M
Zaizen
Y
Igarashi
Y
.
2013
.
Emission of spherical cesium-bearing particles from an early stage of the Fukushima nuclear accident
.
Sci Rep
 .
3
:
2554
.
Beresford
NA
Copplestone
D
.
2011
.
Effects of ionizing radiation on wildlife: what knowledge have we gained between the Chernobyl and Fukushima accidents?
Integr Environ Assess Manag
 .
7
:
371
373
.
Buckley
J
Bridle
JR
Pomiankowski
A
.
2010
.
Novel variation associated with species range expansion
.
BMC Evol Biol
 .
10
:
382
.
Buesseler
KO
Jayne
SR
Fisher
NS
Rypina
II
Baumann
H
Baumann
Z
Breier
CF
Douglass
EM
George
J
Macdonald
AM
et al.  
2012
.
Fukushima-derived radionuclides in the ocean and biota off Japan
.
Proc Natl Acad Sci USA
 .
109
:
5984
5988
.
Callaway
E
.
2013
.
Fukushima offers real-time ecolab
.
Nature
 .
499
:
265
266
.
Chino
M
Nakayama
H
Nagai
H
Terada
H
Katata
G
Yamazawa
H
.
2011
.
Preliminary estimation of release amount of 131I and 137Cs accidentally discharged from the Fukushima Daiichi nuclear power plant into the atmosphere
.
J Nucl Sci Technol
 .
48
:
1129
1134
.
Dhungel
B
Otaki
JM
.
2009
.
Local pharmacological effects of tungstate on the color-pattern determination of butterfly wings: a possible relationship between the eyespot and parafocal element
.
Zoolog Sci
 .
26
:
758
764
.
Evans
AS
.
1976
.
Causation and disease: the Henle-Koch postulates revisited
.
Yale J Biol Med
 .
49
:
175
195
.
Fisher
NS
Beaugelin-Seiller
K
Hinton
TG
Baumann
Z
Madigan
DJ
Garnier-Laplace
J
.
2013
.
Evaluation of radiation doses and associated risk from the Fukushima nuclear accident to marine biota and human consumers of seafood
.
Proc Natl Acad Sci USA
 .
110
:
10670
10675
.
Fujita
Y
Idokawa
K
Hirose
T
Hirokawa
R
.
2013
.
Radioactivity contamination map based on actual measurements: Okuma and Futaba in Fukushima Prefecture (in Japanese)
.
Days Japan
 .
10
:
10
17
.
Fukuda
T
Kino
Y
Abe
Y
Yamashiro
H
Kuwahara
Y
Nihei
H
Sano
Y
Irisawa
A
Shimura
T
Fukumoto
M
et al.  
2013
.
Distribution of artificial radionuclides in abandoned cattle in the evacuation zone of the Fukushima Daiichi Nuclear Power Plant
.
PLoS One
 .
8
:
e54321
.
Garnier-Laplace
J
Beaugelin-Seiller
K
Hinton
TG
.
2011
.
Fukushima wildlife dose reconstruction signals ecological consequences
.
Environ Sci Technol
 .
45
:
5077
5078
.
Garnier-Laplace
J
Geras’kin
S
Della-Vedova
C
Beaugelin-Seiller
K
Hinton
TG
Real
A
Oudalova
A
.
2013
.
Are radiosensitivity data derived from natural field conditions consistent with data from controlled exposures? A case study of Chernobyl wildlife chronically exposed to low dose rates
.
J Environ Radioact
 .
121
:
12
21
.
Grigliatti
TA
.
1998
.
Mutagenesis
. In:
Roberts
DB
, editor.
Drosophila: a practical approach
 .
2
nd ed.
Oxford
:
IRL Press
. p.
55
84
.
Hashimoto
S
Ugawa
S
Nanko
K
Shichi
K
.
2012
.
The total amounts of radioactively contaminated materials in forests in Fukushima, Japan
.
Sci Rep
 .
2
:
416
.
Hayama
S
Nakiri
S
Nakanishi
S
Ishii
N
Uno
T
Kato
T
Konno
F
Kawamoto
Y
Tsuchida
S
Ochiai
K
et al.  
2013
.
Concentration of radiocesium in the wild Japanese monkey (Macaca fuscata) over the first 15 months after the Fukushima Daiichi nuclear disaster
.
PLoS One
 .
8
:
e68530
.
Hiyama
A
Iwata
M
Otaki
JM
.
2010
.
Rearing the pale grass blue Zizeeria maha (Lepidoptera, Lycaenidae): toward the establishment of a lycaenid model system for butterfly physiology and genetics
.
Entomol Sci
 .
13
:
293
302
.
Hiyama
A
Nohara
C
Kinjo
S
Taira
W
Gima
S
Tanahara
A
Otaki
JM
.
2012a
.
The biological impacts of the Fukushima nuclear accident on the pale grass blue butterfly
.
Sci Rep
 .
2
:
570
.
Hiyama
A
Nohara
C
Taira
W
Kinjo
S
Iwata
M
Otaki
JM
.
2013
.
The Fukushima nuclear accident and the pale grass blue butterfly: evaluating biological effects of long-term low-dose exposures
.
BMC Evol Biol
 .
13
:
168
.
Hiyama
A
Taira
W
Otaki
JM
.
2012b
.
Color-pattern evolution in response to environmental stress in butterflies
.
Front Genet
 .
3
:
15
.
Hosoda
M
Tokonami
S
Tazoe
H
Sorimachi
A
Monzen
S
Osanai
M
Akata
N
Kakiuchi
H
Omori
Y
Ishikawa
T
et al.  
2013
.
Activity concentrations of environmental samples collected in Fukushima Prefecture immediately after the Fukushima nuclear accident
.
Sci Rep
 .
3
:
2283
.
Iwata
M
Hiyama
A
Otaki
JM
.
2013
.
System-dependent regulations of colour-pattern development: a mutagenesis study of the pale grass blue butterfly
.
Sci Rep
 .
3
:
2379
.
Johnson
GT
Lewis
TR
Wagner
WD
.
1975
.
Acute toxicity of cesium and rubidium compounds
.
Toxicol Appl Pharmacol
 .
32
:
239
245
.
Kakiuchi
H
Akata
N
Hasegawa
H
Ueda
S
Tokonami
S
Yamada
M
Hosoda
M
Sorimachi
A
Tazoe
H
Noda
K
et al.  
2012
.
Concentration of 3H in plants around Fukushima Dai-ichi Nuclear Power Station
.
Sci Rep
 .
2
:
947
.
Kinoshita
N
Sueki
K
Sasa
K
Kitagawa
J
Ikarashi
S
Nishimura
T
Wong
YS
Satou
Y
Handa
K
Takahashi
T
et al.  
2011
.
Assessment of individual radionuclide distributions from the Fukushima nuclear accident covering central-east Japan
.
Proc Natl Acad Sci USA
 .
108
:
19526
19529
.
Koarashi
J
Moriya
K
Atarashi-Andoh
M
Matsunaga
T
Fujita
H
Nagaoka
M
.
2012
.
Retention of potentially mobile radiocesium in forest surface soils affected by the Fukushima nuclear accident
.
Sci Rep
 .
2
:
1005
.
Mizuno
T
Kubo
H
.
2013
.
Overview of active cesium contamination of freshwater fish in Fukushima and Eastern Japan
.
Sci Rep
 .
3
:
1742
.
Møller
AP
Hagiwara
A
Matsui
S
Kasahara
S
Kawatsu
K
Nishiumi
I
Suzuki
H
Ueda
K
Mousseau
TA
.
2012
.
Abundance of birds in Fukushima as judged from Chernobyl
.
Environ Pollut
 .
164
:
36
39
.
Møller
AP
Mousseau
TA
.
2006
.
Biological consequences of Chernobyl: 20 years on
.
Trends Ecol Evol
 .
21
:
200
207
.
Møller
AP
Mousseau
TA
.
2013
.
Low-dose radiation, scientific scrutiny, and requirements for demonstrating effects
.
BMC Biol
 .
11
:
92
.
Møller
AP
Nishiumi
I
Suzuki
H
Ueda
K
Mousseau
TA
.
2013
.
Differences in effects of radiation on abundance of animals in Fukushima and Chernobyl
.
Ecol Indicat
 .
24
:
75
81
.
Møller
AP
Swaddle
JP
.
1997
.
Asymmetry, developmental stability, and evolution
 .
Oxford
:
Oxford University Press
.
Mousseau
TA
Møller
AP
.
2012a
.
Entomological studies in Chernobyl and Fukushima
.
Am Entomol
 .
58
:
148
150
.
Mousseau
TA
Møller
AP
.
2012b
.
Chernobyl and Fukushima: differences and similarities, a biological perspective
.
Trans Am Nucl Soc
 .
107
:
200
202
.
Nakanishi
TM
Tanoi
K
.
2013
.
Agricultural implications of the Fukushima nuclear accident
 .
Tokyo (Japan)
:
Springer Japan
.
Otaki
JM
Hiyama
A
Iwata
M
Kudo
T
.
2010
.
Phenotypic plasticity in the range-margin population of the lycaenid butterfly Zizeeria maha
.
BMC Evol Biol
 .
10
:
252
.
Pinsky
C
Bose
R
Taylor
JR
McKee
JSC
Lapointe
C
Birchall
J
.
1981
.
Cesium in mammals: acute toxicity, organ changes and tissue accumulation
.
J Environ Sci Health Part A Environ Sci Eng
 .
16
:
549
567
.
Sasaki
K
Hayashi
M
Narita
T
Motoyama
M
Oe
M
Ojima
K
Nakajima
I
Muroya
S
Chikuni
K
Aikawa
K
et al.  
2012
.
Radiocesium distribution in the tissues of Japanese Black beef heifers fed fallout-contaminated roughage due to the Fukushima Daiichi Nuclear Power Station accident
.
Biosci Biotechnol Biochem
 .
76
:
1596
1599
.
Shirôzu
T
.
2000
.
Standard of butterflies in Japan
 .
Tokyo (Japan)
:
Gakken
(in Japanese).
Shozugawa
K
.
2013
.
Advances in gamma-ray analysis (in Japanese)
.
Kagaku
 .
83
:
723
726
.
Steen
TY
Wayne
ML
.
2013
.
Fukushima: ‘ecolab’ branding insensitive
.
Nature
 .
500
:
400
.
St Johnston
D
.
2002
.
The art and design of genetic screens: Drosophila melanogaster
.
Nat Rev Genet
 .
3
:
176
188
.
Tagami
K
Uchida
S
Ishii
N
Kagiya
S
.
2012
.
Translocation of radiocesium from stems and leaves of plants and the effect on radiocesium concentrations in newly emerged plant tissues
.
J Environ Radioact
 .
111
:
65
69
.
The Heliconius Genome Consortium
.
2012
.
Butterfly genome reveals promiscuous exchange of mimicry adaptations among species
.
Nature
 .
487
:
94
98
.
Torii
T
Sugita
T
Okada
CE
Reed
MS
Blumenthal
DJ
.
2013
.
Enhanced analysis methods to derive the spatial distribution of 131I deposition on the ground by airborne surveys at an early stage after the Fukushima Daiichi nuclear power plant accident
.
Health Phys
 .
105
:
192
200
.
Washitani
I
Yoshioka
A
Suda
S
Yasukawa
M
Kitsuregawa
M
.
2013
.
The pale grass blue butterfly based on the Tokyo butterfly monitoring by citizens (in Japanese)
.
Kagaku
 .
83
:
961
966
.
Yasunari
TJ
Stohl
A
Hayano
RS
Burkhart
JF
Eckhardt
S
Yasunari
T
.
2011
.
Cesium-137 deposition and contamination of Japanese soils due to the Fukushima nuclear accident
.
Proc Natl Acad Sci USA
 .
108
:
19530
19534
.
Zhan
S
Merlin
C
Boore
JL
Reppert
SM
.
2011
.
The monarch butterfly genome yields insights into long-distance migration
.
Cell
 .
147
:
1171
1185
.

Author notes

Corresponding editor: Tomoko Steen