During the Fukushima Daiichi Nuclear Power Plant (F1NPP) accident that occurred in March 2011, radionuclides that were released into the atmosphere contaminated the surrounding environment1,2. Since the accident, much attention has been paid to the biological consequences of contamination by radionuclides. To detect the biological changes in the environment, various wild organisms, such as Japanese monkeys3, lycaenid butterflies4, and gall-forming aphids5, inhabiting the surrounding area have been investigated as possible indicator organisms. However, further studies using radiation-responsive indicator organisms help us to reach a consistent conclusion, whether radiological contamination from the F1NPP accident had a biological impact on the environment.
For the purpose of biomonitoring of the radiological contamination, nevertheless, coniferous plants have been demonstrated to be suitable indicator organisms because of their high radiosensitivity, which was revealed decades ago by field examination using gamma irradiation facilities6,7,8,9. Radiosensitive damages in conifers were reported after the Chernobyl nuclear accident in 1986, where two local coniferous species, Scots pine (Pinus sylvestris) and Norway spruce (Picea abies), showed distinct biological damage in the radioactively contaminated areas10,11,12. Under experimental and accidental exposure, morphological changes, particularly in branching of the main axis, were shown to be the most frequently observed radiosensitive responses of coniferous plants6,7,8,9,10,11,12.
Coniferous tree species are grown in the area highly radioactive contaminated by the F1NPP accident, where Japanese fir (Abies firma) is one of the most common naturally grown species. Different from other coniferous species, young-tree populations of Japanese fir are abundant, because this species has the characteristic ability to sprout even on the shaded forest floor. The short height of young trees enables the easy observation of morphological changes in the whole tree. In addition, the regular annual branching of Japanese fir trees enables determination of the year that any morphological changes occurred through a number of past years (Fig. 1).
Most of the naturally grown Japanese fir trees showed a typical monopodial branching pattern to form a trunk with one main axis (Fig. 3A), whereas some trees showed distinctive morphological defects on the main axis of the trunk (Fig. 3B,C). Independently of the growing site, these defects were characterized by irregular branching at the whorls of the main axis with a distinct deletion of the leader shoot that normally elongates vertically to form the main axis. The space of the deleted leader shoot was filled in by the remaining lateral branches that either extended upwards (Fig. 3B) or retained their horizontal position (Fig. 3C).
The overall frequency of the morphological defects of the main axis varied among observation sites, S1, S2, and S3, but it was significantly higher in each site compared to the control, S4 (chi-square test with df = 1, p = 2.1 × 10−58, 3.7 × 10−17, and 8.1 × 10−7, respectively, Bonferroni-corrected; Fig. 4). The frequency corresponded to the ambient dose rate at the observation sites that represented the local levels of radionuclide contamination (S1 > S2 > S3 > S4, Table 1). A high frequency of defects was observed in S1, where 125 out of 128 trees showed branching defects of the main axis.
Branching defects of the main axis were analyzed separately in each annual whorl (Fig. 5). Compared to the whorls of 2010, which had been generated before the F1NPP accident, the frequency of deleted leader shoots was significantly increased in the whorls after 2012 (sites S1 and S3), or those after 2013 (site S2). The frequency peaked in the whorls of 2013 and tended to decrease in the whorls of 2014 in every observation site. The variation patterns in the series of annual whorls were similar among the sites, whereas no annual variation was observed in the control site, S4. These results indicated that the deletion of leader shoots occurred most frequently in the whorls that elongated from terminal winter buds during the growing season of 2012–2013.
Despite the significant increase in the frequency of deleted leader shoots in annual whorls around 2013 in the observation sites S1–S3, the number of lateral branches that elongated from the same whorls did not show annual variation that corresponded to the deletion frequency of leader shoots (Fig. 6). The number of lateral branches was not different among annual whorls even in S1 (one-way ANOVA, p = 0.84), in which the frequency of leader shoot deletions varied most intensely compared to the other observation sites (Fig. 5). On the other hand, the number of lateral branches showed significant annual variation in S2, S3, and S4 (one-way ANOVA, p = 1.4 × 10−7, 6.3 × 10−3 and 1.5 × 10−8 for S2, S3, and S4, respectively); however, the annual variation patterns were independent from the frequency of leader shoot deletions. In addition, the variation in lateral branch number among the sites did not correspond to the frequency variation of deleted leader shoots. This indicated that the deletion of leader shoots occurred independently of the change in lateral branch number that elongated from the whorls.
Differences in the development of the leader shoots and lateral branches were also observed from a close inspection of the defected whorls. At each site, the deleted leader shoots left no marks among normal lateral branches (Fig. 7A). Similar structures were also observed in the winter buds of 2015 at the top of the main axis, where normal lateral buds with completely deleted apical buds were sometimes observed (Fig. 7B). These observations demonstrated that the deletion of leader shoots probably resulted from the deletion of apical buds at an early stage of their development, independently of the formation of lateral buds.