Science of the Total Environment 482–483 (2014) 184–192
Contents lists available at ScienceDirect
Science of the Total Environment
j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv
Radioactive contamination of fishes in lake and streams impacted by theFukushima nuclear power plant accident
Mayumi Yoshimura a,⁎, Tetsuya Yokoduka b
a Kansai Research Center, Forestry and Forest Products Research Institute, Nagaikyuutaro 68, Momoyama, Fushimi, Kyoto 612-0855, Japanb Tochigi Prefectural Fisheries Experimental Station, Sarado 2599, Ohtawara, Tochigi 324-0404, Japan
H I G H L I G H T
• Concentration of 137Cs in brown trout was higher than in rainbow trout.• 137Cs concentration of brown trout in a lake was higher than in a stream.• 137Cs concentration of stream charr was higher in region with higher aerial activity.• Concentration of 137Cs in stream charr was higher in older fish.• Difference of contamination among fishes was due to difference in diet and habitat.
⁎ Corresponding author. Tel.: +81 75 611 1201; fax: +E-mail address: [email protected] (M. Yoshim
http://dx.doi.org/10.1016/j.scitotenv.2014.02.1180048-9697/© 2014 Elsevier B.V. All rights reserved.
a b s t r a c t
a r t i c l e i n f oArticle history:Received 27 November 2013Received in revised form 19 February 2014Accepted 25 February 2014Available online 18 March 2014
Keywords:Brown troutCharrLakeRadioactive cesiumStomach contentsStream
The Fukushima Daiichi Nuclear Power Plant (FDNPP) accident in March 2011 emitted radioactivesubstances into the environment, contaminating a wide array of organisms including fishes. We foundhigher concentrations of radioactive cesium (137Cs) in brown trout (Salmo trutta) than in rainbowtrout (Oncorhynchus nerka), and 137Cs concentrations in brown trout were higher in a lake than in astream. Our analyses indicated that these differences were primarily due to differences in diet, but thathabitat also had an effect. Radiocesium concentrations (137Cs) in stream charr (Salvelinus leucomaenis)were higher in regions with more concentrated aerial activity and in older fish. These results were alsoattributed to dietary and habitat differences. Preserving uncontaminated areas by remediating soilsand releasing uncontaminated fish would help restore this popular fishing area but would require a sig-nificant effort, followed by a waiting period to allow activity concentrations to fall below the thresholdlimits for consumption.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
A massive earthquake occurred in eastern Japan on 11 March 2011,causing a tsunami that washed over the Fukushima Daiichi NuclearPower Plant (FDNPP) on the east coast of Japan. Damage to the coolingsystem of the FDNPP resulted in several explosions, causing leakage of ra-dioactive substances. The FDNPP accident released 1.6 × 1017 Bq ofiodine-131 (131I), 1.8 × 1016 Bq of cesium-134 (134Cs) and 1.5 × 1016 Bqof cesium-137 (137Cs) into the surrounding environment (Oharaet al., 2011). Most of these radionuclides, including 131I, 134Cs and137Cs, were unevenly deposited over large areas of land in eastern
81 75 611 1207.ura).
Japan, reaching sites hundreds of kilometers from the FDNPP. Thedense radioactive plume spewing from the FDNPP moved north-westward, descending to ground level with precipitation and heavilycontaminating large tracts of the landscape. A smaller plume driftedto the south-west and contaminated areas such as the Oku Nikko andAshio regions in Tochigi Prefecture (Kinoshita et al., 2011; MEXT,2011). Atmospheric dose rates 0.5 m above the ground exceeded20 μSv/h in some hot spots N160 km from the FDNPP in May 2011(Tochigi, 2011).
The first phase of investigation revealed that a large portion of thedeposited 134Cs and 137Cs was trapped in the forest canopies and thesoil litter layer (Hashimoto et al., 2012). Radiocesium is easily adsorbedonto clay minerals and soil organic matter (Kruyts and Delvaux, 2002)and can be transported to streams and rivers with eroded soils(Fukuyama et al., 2005; Kato et al., 2010; Wakiyama et al., 2010).Dissolved 134Cs and 137Cs in running waters that are not adsorbed tosoil, are readily taken up by microbes, algae, plankton and plants. This
185M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
pathway of 134Cs and 137Cs transport eventually leads to uptake by fishesat higher trophic levels.
Radioactive contamination of fish should be prevented becausefishes may be taken by anglers and consumed as food. A safetythreshold of 100 Bq/kg of radioactive Cs was introduced in April2012, but activity concentrations greater than this have been detect-ed in fishes hundreds of kilometers distant from the FDNPP. There isclearly a pressing imperative to reduce radioactive Cs contaminationof food.
The Chernobyl accident released more than five million TeraBecquerel of radionuclides. Much radionuclides from the Chernobylaccident spread to Finland, Sweden and Norway, 2000 km to thenorthwest. Fish have been monitored for radioactivity ØvreHeimdalsvatn, a Norwegian subalpine lake (Brittain et al., 1991;Brittain and Gjerseth, 2010). Activity concentrations of 137Cs inbrown trout reached 8400 Bq/kg in 1987 and declined to 200–300 Bq/kg in 2008, but the contamination level has recently beenconstant and approached an asymptotic decline (Brittain andGjerseth, 2010). The broad effects of the Chernobyl accident havebeen examined and management measures have been designed forlarge geographical areas (Report of the Chernobyl Forum ExpertGroup ‘Environment’, 2006; Yablokov et al., 2009). However, theecological consequences of radioactive contamination in fishes arepoorly understood.
To precisely describe the mechanisms of diffusion and export of137Cs deposited in freshwater fish, we measured concentrations of137Cs in the muscle tissue and stomachs of brown trout (Salmo trutta),rainbow trout (Oncorhynchus mykiss) and kokanee (Oncorhynchusnerka) from a lake, that of brown trout (Salmo trutta) from a streamand that of charr (Salvelinus leucomaenis) from four streams. Here, wereport the results of preliminary investigations 21 months after theFDNPP accident.
2. Materials and methods
2.1. Study site
The study was conducted in the Nikko area (Lake Chuzenji, OkuNikko and Ashio regions) located approximately 160 km south-west of the FDNPP. According to an aircraft radioactivity surveyreported by MEXT (2012), the air dose rate in this area was0.1–0.25 μSv/h on 31 May 2012. The study area is mostly forested;the dominant tree species are broadleaf and deciduous. Otherareas are forested with Japanese cedar and cypress plantationsused for timber production. The field survey was conducted inLake Chuzenji and two headwater tributaries (Toyamasawa andYanagisawa streams) of the Kinu River in the Oku Nikko regionand in two headwater tributaries (Kuzosawa and Asosawastreams) of the Watarase River in the Ashio region, which formthe upper drainage component of the Tone River system, Honshu,Japan (Fig. 1).
2.2. Sampling
Fishes were captured in Lake Chuzenji and in the four streams inNovember and December, 2012 using fishing rod in the lake andbattery-powered backpack electrofishing (Smith-Root Inc. LR-24)units operated at 300-V pulse-DC in the stream. In central LakeChuzenji, three species of fishes, brown trout (Salmo trutta),rainbow trout (Oncorhynchus mykiss) and kokanee (salmon)(Oncorhynchus nerka) were sampled. In Toyamasawa stream, twospecies of fish, brown trout (Salmo trutta) and charr (Salvelinusleucomaenis) were sampled at site B (200 m stream segments). Inother three streams, only charr (Salvelinus leucomaenis) were sam-pled at sites E, G, and I (200 m stream segments). After capture, werecorded the body size (fork length) of each fish and sampled
muscle tissue for the analysis of radioactive concentration. Wethen collected and froze the stomachs for the analysis ofradiocesium concentrations in stomach tissue and for identificationof stomach contents. The age of charr specimens was determinedfrom sagittal otoliths using the surface reading method. The ageof the other three fishes was not determined; folk length wasused as a substitute metric.
In the four streams, air dose rates at 1-m above ground weremeasured with a γ survey meter adjacent to the stream (NaIscintillation counter; ALOKA TCS-172). Electrical conductivities(EC) of the streams were measured with a portable compacttwin conductivity meter (B-173; Horiba); pH was measuredwith a portable compact twin pH meter (B-212; Horiba). Wettedstream widths (SW) were measured with a measuring tape;stream velocities were measured with a portable meter (V-303,VC-301, KENEK). All of these environmental parameters weremeasured in stream riffles at each site along Toyamasawa stream(sites A, B, and C), Yanagisawa stream (sites D, E, and F),Kuzosawa stream (sites G and H) and Asosawa stream (site I)in December, 2012.
2.3. Radiocesium analysis and identification
Samples of fish stomach andmuscle tissuewere directly packed into100-ml polystyrene containers (U-8). The radioactive levels of 137Cs(662 keV) were measured with an HPGe coaxial detector system(GEM40P4-76, GEM20-70, Seiko EG&G, Tokyo, Japan; GC4020, Canberra,Japan) for 36,000 s–63,000 s depending on the sample weight.Gamma-ray peaks of 622 keV were used to determine 137Cs. Themeasurement system was calibrated using a standard gamma-raysource (MX033U8PP, Japan Radioisotope Association, Tokyo,Japan), and a standard soil sample (IAEA-444) was used to check ac-curacy. Fine adjustments to the measurements were made to corre-spond to the radiocesium concentration value for 1 December2012. The radioactivity of lake and stream water was not measuredbecause 137Cs concentrations in several lakes and streams were re-ported to be below the detection level (1 Bq/l) (MOE 2013, TochigiPrefecture, 2013).
The inner contents of the frozen stomach for the analysis ofstomach contents were removed and identified under a 50× mi-croscope (SMZ-U; Nikon). We identified specimens to the familylevel or higher following Merritt and Cummins (1996) and Kawai(1985).
2.4. Statistical analysis
Kruskal–Wallis tests or Mann–Whitney U-tests were used to com-pare the 137Cs concentrations. The Kendall test was used to clarify therelationship between folk length and 137Cs concentration. We conduct-ed principal component analysis (PCA) on the presence or absence ofeach fish stomach content for three species in Lake Chuzenji, forbrown trout in Lake Chuzenji and Toyamasawa stream, and for charrin the four streams. We examined the PCA axes among the three fishspecies and among the four streams using a Kruskal–Wallis multiplecomparison test and between lake and stream habitats using a Mann–Whitney U-test.
It would be useful to compare samples of the same species or thesame age at all sampling locations. However, the fishes sampled inthis study differed between the stream and lake and it was notpractical to collect specimens of equal size. Statistical analysis wasperformed using SYSTAT version 10 (SPSS Inc. 2000). Radiocesiumconcentrations that were below the detection level because of insuf-ficient weight for radiocesium analysis were excluded from thestatistical analysis.
Fig. 1. Study site of Lake Chuzenji and four streams in the Oku Nikko and Ashio areas, Tochigi Prefecture, Japan.● A–I: stream water sampling sites, □ B, E, G, I: fish sampling sites.
186 M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
3. Results
Aerial radioactivity was higher in the Ashio region than in theOku Nikko region (z = 2.33, n = 9, P b 0.05, Mann–Whitney U-test;Table 1), as was radiocesium deposition. These two regions did not dif-fer in air orwater temperature, pH, streamwidth or stream velocity, butEC differed between the regions (Table 1; Yoshimura andAkama, 2014).
137Cs concentrations in fish stomachs were highest in browntrout (Salmo trutta), lowest in rainbow trout (Oncorhynchusmykiss) and intermediate in kokanee (salmon) (Oncorhynchusnerka) (H = 11.1, n = 14, P b 0.005, Kruskal–Wallis test; Fig. 2).137Cs concentrations in fish muscle showed the same patternamong species (H = 590.5, n = 69, P b 0.0001, Kruskal–Wallis
Table 1The dose rate at 1 m above the ground, air temperature, water temperature, pH, electric conduc
Oku Nikko
Toyamasawa Yanagisaw
A B C D
Dose rate (μSv/h) 0.08 0.10 0.12 0.09Air temperature (°C) −0.8 1.7 0.7 0.1Water temperature (°C) 7.2 6.8 7.3 8.1pH 6.9 7.0 7.1 6.7Electric conductivity (μs/cm) 61 50 49 53Stream velocity (cm/s) 30 63 32 48Stream widths (m) 7.6 5.7 4.8 17.1Altitude (m) 1270 1300 1450 1270
*: P b 0.05.
test; Fig. 2). Radioactive Cs concentrations in stomach and muscletissues did not differ significantly among the three species. Therewas no relationship between 137Cs concentrations in fish muscleand fork length in three species (brown trout: τ = 0.27, n = 23,n.s.; kokanee: τ = 0.02, n = 24, n.s.; rainbow trout: τ = 0.17, n = 22,n.s.; Kendall test).
Stomach contents varied significantly among the three speciesof fish (Appendix A). Brown trout stomachs primarily containedfish and unknown digested foods. Kokanee primarily contained un-known foods and rainbow trout stomachs contained a wide varietyof fish and invertebrates. The first PCA axis explained 52% of thevariation in the stomach contents (mainly the number of speciesconsumed) but was not correlated with muscle concentration of
tivity, stream velocity and streamwidth at 9 sites onDecember 2012with statistical result.
Ashio
a Kuzosawa Asosawa U-test
E F G H I z
0.10 0.11 0.26 0.21 0.27 2.33*6.4 6.7 1.3 1.7 −1.9 0.654.9 2.6 4.0 4.2 1.6 1.816.4 6.9 6.5 6.3 6.8 1.56
33 51 73 59 64 2.07*35 58 70 68 57 1.815.8 8.1 3.0 3.5 5.8 2.69
1320 1380 810 950 790 2.33*
0
50
100
150
200
0 100 200 300 400 500 600
137 C
s (B
q/k
g)
Fork length (mm)Brown trout (stomach) Kokanee (stomach) Rainbow trout (stomach)
b
0
50
100
150
200
0 200 400 600 800
137 C
s (B
q/k
g)
Fork length (mm)
Brown trout Kokanee Rainbow trout
a
Fig. 2. Relationship between body length and 137Cs concentration of three species in LakeChuzenji. (a) Muscle; (b) stomach.
0
50
100
150
200
0 100 200 300 400 500 600
Fork length (mm)Chuzenji Lake Toyamasawa
a
0
50
100
150
200
0 100 200 300 400 500 600
Fork length (mm)Chuzenji Lake (stomach) Toyamasawa (stomach)
b
137 C
s (B
q/k
g)
137 C
s (B
q/k
g)
Fig. 4. Relationship between body length and 137Cs concentration in brown trout fromLake Chuzenji and Toyamasawa stream. (a) Muscle; (b) stomach.
0.8
Lake
187M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
137Cs (τ = 0.37, n = 15, n.s., Kendall test). The second PCA axis ex-plained 13% of the stomach content data (the lower part of the di-agram shows higher fish contents and the upper part of diagramshows higher insects contents) and was correlated with muscleconcentrations of 137Cs (τ = −0.56, n = 15, P b 0.005, Kendalltest). The PCA also showed that stomach contents of the three specieswere statistically different based on the first axis (H = 9.1, n = 15,
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
0 0.2 0.4 0.6
Axis 1 (52%)
Axi
s 2
(13%
)
0.8 1 1.2
Brown troutKokaneeRainbow trout
Fig. 3. Scatter diagram of the principal components analysis (PCA) of stomach contents ofthree fishes from Lake Chuzenji. Bubble size indicates the concentrations of 137Cs inmuscle tissue of each fish.
P b 0.05, Kruskal–Wallis test; Fig. 3) and the second axis (H = 6.2,n = 15, P b 0.05, Kruskal–Wallis test; Fig. 3). No significant correla-tion was observed between the PCA axes and 137Cs concentrationin any of the three fish species (Kendall test; Fig. 3).
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 0.2 0.4 0.6 0.8 1
Stream
Axi
s 2
(20%
)
Axis 1 (33%)
Fig. 5. Scatter diagram of stomach contents in the principal components analysis (PCA) ofeach brown trout in Lake Chuzenji and Toyamasawa stream. Bubble size indicates the con-centration of 137Cs in the muscle tissue of each fish.
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1
ToyamasawaYanagisawaKuzosawaasosawa
Axis 1 (30%)
Axi
s 1
(9%
)
Fig. 7. Scatter diagram of the principal components analysis (PCA) for stomach contents ofeach charr from the four streams. Bubble size indicates the concentrations of 137Cs inmuscle tissue of each fish.
188 M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
Concentrations of 137Cs in the stomach contents of browntrout were higher in Lake Chuzenji than in the streams (z =2.7, n = 11, P b 0.01, Mann–Whitney U-test; Fig. 4), and so were137Cs concentrations in muscle tissue (z = 6.0, n = 50, P b 0.0001,Mann–WhitneyU-test; Fig. 4). No significant difference in 137Cs concen-tration was observed between stomach and muscle tissues. Therewas significant relationship between 137Cs concentrations in fishmuscle and fork length in brown trout from streams (τ = 0.51, n =29, P b 0.001; Kendall test).
Stomach contents of brown trout differed between Lake Chuzenjiand Toyamasawa stream (Appendix A). In streams, the brown troutstomach contents were mainly insects. The first PCA axis explained33% of the variation in stomach content (mainly the number of speciesconsumed) and was not correlated with 137Cs concentrations in muscletissue (τ = 0.01, n = 14, n.s., Kendall test). The second PCA axis ex-plained 20% of the variation in stomach content data (the lower partof the diagram shows higher fish content and the upper part of thediagram shows higher insect content) and was correlated with137Cs concentrations in muscle tissue (τ = −0.40, n = 14, P b 0.05,Kendall test). Stomach contents were statistically different betweenlake and stream fish based on PCA axis 2 (z = 3.0, n = 14, P b 0.005,Mann–Whitney U-test; Fig. 5), but not on axis 1 (z=0.2, n=14, n.s.,Mann–Whitney U-test; Fig. 5). No significant correlation was detect-ed between these two axes and 137Cs concentrations of fish in eachhabitat (Kendall test; Fig. 5).
Concentrations of 137Cs in the stomach tissue of charr did not dif-fer among the four streams (H = 4.6, n = 18, n.s., Kruskal–Wallistest), but muscle activity concentrations did differ among thestreams (H = 16.2, n = 88, P b 0.001, Kruskal–Wallis test). Olderfish had higher muscle concentrations of 137Cs (Toyamasawa: H =7.8, n=35, P b 0.052; Yanagisawa:H=8.6, n=19, P b 0.05; Kuzosawa:H = 22.5, n = 33, P b 0.0001, Kruskal–Wallis test; Fig. 6), but no differ-ence was observed in the concentrations of 137Cs in stomach tissuesamong age.
The stomach contents of charr consisted primary of insections in allfour streams (Appendix B). Axis 1 of the PCA explained 30% of the var-iation in stomach contents (the right side of the diagram shows higheraquatic insect content and the left side of diagram shows higher terres-trial insects content) andwas correlatedwith concentrations of 137Cs inmuscle tissue (τ = 0.31, n = 26, P b 0.05, Kendall test). Axis 2 of thePCA explained 9% of the variation in stomach contents (the lower partof the diagram shows higher Bryophyta content) and was also correlat-ed with 137Cs levels in muscle (τ = −0.30, n = 26, P b 0.05, Kendalltest). Stomach contents among the four streams differed statisticallyalong PCA axis 1 (H = 22.2, n = 40, P b 0.001, Kruskal–Wallis test;Fig. 7), especially between fish in Toyamasawa and the other streams.Stomach contents also differed among the four streams based on PCAaxis 2 (H = 22.9, n = 40, P b 0.001, Kruskal–Wallis test; Fig. 7), espe-cially between fish in the Ashio region and the Oku Nikko region. No
0
20
40
60
80
100
137 C
s (B
q/k
g)
137 C
s (B
q/k
g)
Fish age0+ 1+ 2+ 3+
a
0
20
40
60
80
100
Fis0+ 1
b
Fig. 6. Relationship between fish age and concentration of 137Cs in stream fish. (a) Toyamasawmuscle and ♦: concentration of 137Cs in charr stomach tissue.
significant correlation was observed between these two axes and137Cs concentrations of fish in each of the four streams, except the rela-tionship between PCA axis 2 and 137Cs concentration inmuscle tissue offish from the Kuzosawa stream (τ=−0.73, n= 11, P b 0.005, Kendalltest; Fig. 7).
4. Discussion
Fishes were contaminated with radioactive substances even wheredose rates in the atmosphere were relatively lower (160 km distancefrom the FDNPP). In Lake Chuzenji, concentrations of 137Cs in muscletissues and stomachwere highest in S. trutta, lowest inO. mykiss and in-termediate in O. nerka. The lack of difference between stomach andmuscle concentrations of 137Cs indicates that the absence of a highlevel of contamination in the stomach contents and suggests that andradioactive materials are transferred from food items to muscle tissues.Research after the Chernobyl accident also showed that radioactive con-taminationwas caused by diet (Forseth et al., 1991; Ugedal et al., 1995),and dietary differences were reflected in the decay rates and ecologicalhalf-lives (Hessen et al., 2002).
The primary food of brown trout in Lake Chuzenji was goby,and the 137Cs concentration of goby in Lake Chuzenji averaged57.4 Bq/kg (n = 20, 1 December 2012, unpublished data). There wasno diet in the stomach of kokanee and they usually consume zoo-plankton. The 137Cs concentration of zooplankton in Lake Chuzenjiis 8.3 Bq/kg (n = 1, 1 December 2012, unpublished data). Thus,
137 C
s (B
q/k
g)
0
20
40
60
80
100
Fish age0+ 1+ 2+ 3+
c
h age+ 2+ 3+
a stream, (b) Yanagisawa stream, (c) Kuzosawa stream.○: Concentration of 137Cs in charr
189M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
higher 137Cs concentrations in brown trout are likely to be acquiredmainly from their diet. Some rainbow trout had 137Cs concentra-tions that were below the detection limit; these fishes may havebeen released as adults. Brown trout and kokanee are releasedinto Lake Chuzenji only as fry, but rainbow trout are released asboth fry and adult fish. Because fish released as adults spend onlya short period in the contaminated lake, their muscle tissue is ex-pected to be less contaminated.
The mean 137Cs concentrations of three families of aquaticstream insects (Perlidae, Perlodidae and Stenopsychidae) were67.9 Bq/kg (max: 271 Bq/kg, min: 0 Bq/kg; Yoshimura andAkama, 2014). Thus, 137Cs concentrations in aquatic insects in lakeare higher than those of zooplankton, but with a large variance. Activityconcentrations of more than 50 Bq/kg in some lake rainbow trout,which feed primarily on insects and are released as fry, indicate thatmuscle contamination is a result of diet. However, concentrations of137Cs did not differ greatly between goby in the lake and aquatic insectsin the streams. The difference in 137Cs concentrations between thesetwo fish species may instead be a result of habitat differences. Browntrout typically reside from benthic substrate to middle strata of thelake interior, whereas rainbow trout reside from middle to upperdepths close to the shoreline. Radioactive substances tend to condensemore in deeper area of the lake where less water flow occurs. Higher137Cs concentrations in brown trout may arise from these differencesin habitat preference.
Stomach contents of brown trout in Lake Chuzenji differed fromthose in Toyamasawa stream, and radiocesium concentrations ofstomach and muscle tissue also differed between the habitats, sug-gesting that the differences in contamination are due to habitat dif-ferences. However, 137Cs concentrations of goby consumed bybrown trout in the lake and aquatic insects consumed by browntrout in the stream did not differ. The primary environmental dif-ference between the lake and stream is the rate of water flow. Inlakes, particulate organic matter from the surrounding forest,which may be contaminated with 137Cs, settles to the bottom andreleases 137Cs by the decomposition of particle organic materialsgathered. The released 137Cs accumulates at the bottom of thelake and is not easily washed away. The lack of flow in deep por-tions of the lake may also result in accumulation of silt contaminat-ed with 137Cs. Sand substrates had higher 137Cs concentrations inLake Chuzenji (81.28 Bq/kg, n = 11, unpublished data) than inany of the four streams (26.59 Bq/kg; Yoshimura and Akama,2014). The 137Cs value of sand in the lake seems to be higher thanthat in the stream (Ministry of the Environment, 2011, 2013).Aquatic insects that inhabit stream pools have higher value of137Cs than insects in riffles (Yoshimura and Akama, 2014); waterflow rate may thus be a key factor determining the activity concen-trations in fishes. Retention time would also be related to contam-ination level and would lead to variability in radioactive cesiumconcentration in fish from different lentic environments (Håkanson,1992). Free 137Cs in stagnant water in lakes accumulates particulateorganic matter associated with the substrate. Thus, fishes livingnear the bottoms of lakes would intake highly 137Cs contaminatedmaterials into their bodies from dietary items settling to thebottom.
Brown trout in Norwegian lake contaminated by radioactivefallout from the Chernobyl accident retain substantial radioactivitylevels more than 20 years after the accident, and these levels donot appear to be declining (Brittain and Gjerseth, 2010), probablybecause of continuing input of radioactive substances from thecatchment and remobilization of sediments in the lake. Somemechanisms of 137Cs contamination of fishes remain unknown. Re-moval of radioactive substances is necessary to alleviate this prob-lem. At present, only catch-and-release fishing is permitted in LakeChuzenji. Although restoring the lake environment would take de-cades, areas in which fishing and fish consumption can be safely
practiced must be reclaimed. Full-grown rainbow trout released tothe lake had lower concentrations of 137Cs. Decontamination of soilsubstrates and release of non-contaminated fish in designatedareas would be the most rapid way to improve fish quality andenable tourism to resume. However, continual removal of highlycontaminated adult fish such as brown trout and release of uncon-taminated fry may constitute a more appropriate long-termapproach.
Among the four streams, radioactive Cs values in the muscle tis-sues of charr were elevated where aerial dose rates were higher. Al-though no differences in 137Cs values were observed in stomachamong the four streams, fishes in the Ashio region consumedmore Chlorophyta and Heterokontophyta. The 137Cs value of Chlor-ophyta in the Ashio region was 170 Bq/kg (n = 1, 1 December2012, unpublished data). The higher 137Cs value of fish in theAshio region might be a result of consumption of these algae in ad-dition to higher contamination from fallout. The higher 137Cs con-centrations in older fish might also be caused by continuousintake of contaminated food. However, a large degree of variationin 137Cs was detected in equal-aged fishes in a given stream. Thehabitat range of charr is generally larger than that of aquatic insectsand always includes a small riffle and a pool. Concentrations of137Cs in aquatic insects differed between pools and riffles evenwithin the same families. Habitat selection (proportion of timespent in pools and riffles) by individual fish may cause dietary var-iation in 137Cs, leading to variability in 137Cs levels among fish inthe same sampling area.
The PCA axes for fish diet were correlated with 137Cs concen-trations of muscle tissue in all three species of fish in the lake,in brown trout in the lake and streams, and in charr in all fourstreams, suggesting that dietary differences are primarily re-sponsible for the differences in radioactive Cs concentrationsamong these fish. However, the PCA axis for diet was not corre-lated with 137Cs concentrations in muscle tissue across species,habitat or streams, which suggests that the diet menu of eachindividual is not always almost the same and there is substan-tial variation of 137Cs concentration in individual diets. Fritschet al. (2008) reported that 137Cs in soil is transferred to plants,earthworms and snails, but that transfer rates of radionuclidesthrough the food web depend on the concentration of radio-active substances, exposure period and consumption rates.Habitat variability among individual fish would also affect137Cs transfer.
The areas where this study was conducted are extremely pop-ular fishing locations. However, only catch-and-release fishing iscurrently permitted because of the elevated radioactivity levels.As long as the diet and sand substratum are contaminated, levelsof radioactive Cs in fishes are expected to remain high, and con-tinual monitoring and management of radioactivity will benecessary.
Conflict of interest
The authors declare no conflicts of interests.
Acknowledgements
We thank Dr. Makino in the Forestry and Forest Products ResearchInstitute, Dr. Yamamoto in the National Research Institute of aquacul-ture and Dr. Yoshida in the fisheries experimental station of TochigiPrefecture for their assistance with sampling survey. Capture of fisheswas done with the permission from Tochigi Prefecture. This work wassupported in part by a research grant from the Council of Science andTechnology Policy in 2012.
Appendix A. Stomach contents of three species of fishes in Lake Chuzenji and of brown trout in the Toyamasawa stream
Lake Chuzenji Toyama sawa
Division Order Family Brown trout (5) Kokanee (5) Rainbow trout (5) Brown trout (9)
Nematomorpha Gordea Gordiidae ***Arthropoda Araneae Undetermined **
Lithobiomorpha Ethopolidae **Thysanura Lepismatidae * *Ephemeroptera Baetidae **
Heptageniidae **Ephemerellidae *Undetermined **
Plecoptera Nemouridae ***Chloroperlidae * *Perlodidae * **
Orthoptera Acrididae ***Hemiptera Membracidae ** *
Psyllidae ***Aphididae **Tettigometridae **Reduviidae **Lygaeidae **Pentatomidae *** ***Acanthosomatidae ** **Undetermined ** ***
Trichoptera Arctopsychidae **Philopotamidae *Stenopsychidae *Rhyacophilidae * **Brachycentridae **Limnephilidae ***** **Undetermined ** ***
Diptera Tipulidae **Chronomidae *Bibionidae ***Drosophilidae *Undetermined *** **
Coleoptera Staphylinidae **Undetermined ** **
Hymenoptera Chalcidoidea *Formicidae ** **Undetermined **
Insect (fragments) *** ****Insect egg **
Chordata Perciformes Gobiidae **** ***Ichthyic digest **** ***Ichthyic eggs **Ichthyic bone *Bird feather **
Unknown egg *Chlorophyta Ulotrichales Ulotrichaceae *Bryophyta Bryidae ****
Hepaticopsida **Magnoliophyta Hydrocharitales Hydrocharitaceae ***
Sapindales Aceraceae ***Plant (fragment) * ** **** **Sand *** ***Unknown digest **** **** ***** *****
Weight of each stomach content, *: b0.01, **: b0.1, ***: b1.0, ****: b10.0, *****: b100.0 (g).
Appendix B. Stomach contents of charr in four streams
Division Order Family Toyamasawa (14)
Yanagisawa (10)
Kuzo sawa(11)
Asosawa (5)
Nematoda ** *** ** *Arthropoda Opiliones Phalangiidae **
Acari Spercontidae *Araneae Thomisidae **
Undetermined *Isopoda Ligiidae **Collembola *Ephemeroptera Siphlonuridae ** * **
Baetidae ** ** *Heptageniidae ** *
190 M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
(continued)
Division Order Family Toyamasawa (14)
Yanagisawa (10)
Kuzo sawa(11)
Asosawa (5)
Ephemeridae ** *Ephemerellidae ** *** *** **Undetermined *
Plecoptera Scopuridae ***Capniidae * * *Nemouridae *** ** * *Taeniopterygidae * * *Chloroperlidae * * **Perlidae ** *** *Perlodidae ** *** *** ***
Orthoptera Acrididae **Hemiptera Membracidae **
Aphrophoridae **Psyllidae ** *Aphididae ** * *Lygaeidae **Pentatomidae *Acanthosomatidae ***Undetermined ** *
Neuroptera Osmylidae **Undetermined **
Trichoptera Arctopsychidae **Hydropsychidae ** **Philopotamidae *Stenopsychidae *** *** ***Glossosomatidae **** *** *** ***Rhyacophilidae ** ** **Limnephilidae *** ** ** **Brachycentridae *Goeridae *** *Lepidostomatidae * ** *** ***Uenoidae * **Undetermined ** *
Lepidoptera ***Diptera Tipulidae ** * **
Blephariceridae **Chronomidae ** ** **** ***Simuliidae * *Empididae * *Bibionidae ***Undetermined * ** **
Coleoptera Carabidae **Staphylinidae **Curculionidae **Chrysomelidae ***Undetermined ** * *
Hymenoptera Cynipoidea * *Chalcidoidea *Braconidae *Ichneumonidae *** *Formicidae *
Insect (fragments) **** ** *** ***Chordata Salmoniformes Salmonidae ****
Ichthyic bone **Unknown egg *Chlorophyta Ulotrichales Ulotrichaceae * *Heterokontophyta Centrales Melosiraceae *
Pennales Diatomaceae * *Bryophyta Bryidae **Plant (fragment) *** ** *** **Sand *** *** **** ***Unknown digest **** **** ***** ***
Weight of each stomach content, *: b0.01, **: b0.1, ***: b1.0, ****: b10.0, *****: b100.0.
Appendix B (continued)
191M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
References
Brittain JE, Gjerseth JE. Long-term trends and variation in 137Cs activity concentrations inbrown trout (Salmo trutta) from Øvre Heimdalsvatn, a Norwegian subalpine lake.Hydrobiologia 2010;642:107–13.
Brittain JE, Storruste A, Larsen E. Radiocesium in brown trout (Salmo-trutta) from a sub-alpine lake ecosystem after the Chernobyl reactor accident. J Environ Radioact1991;14:181–91.
Forseth T, Ugedal O, Jonsson B, Langeland A, Njastad O. Radiocaesium turnover inArctic charr (Salvelinus alpinus) and brown trout (Salmo trutta) in a Norwegianlake. J Appl Ecol 1991;28:1053–67.
Fritsch C, Scheifler R, Beaugelin-Seiller K, Hubert P, Coeurdassier M, Vaufleury AD, et al.Biotic interactions modify the transfer of Cesium-137 in a soil–earthworm–plant–snail food web. Environ Toxicol Chem 2008;27:1698–707.
Fukuyama T, Takenaka C, Onda Y. 137Cs loss via soil erosion from amountainous headwa-ter catchment in central Japan. Sci Total Environ 2005;350:238–47.
Håkanson L. Radioactive cesium in fish in Swedish lakes 1986–1988—general pat-tern related to fallout and lake characteristics. J Environ Radioact 1992;15:207–29.
Hashimoto S, Ugawa S, Nanko K, Shichi K. The total amounts of radioactively contam-inated materials in forests in Fukushima, Japan. Sci Rep 2012;2. [article number416].
192 M. Yoshimura, T. Yokoduka / Science of the Total Environment 482–483 (2014) 184–192
Hessen DO, Skurdal J, Hegge O, Hesthagen T. Radiocesium decay in populations of browntrout and Arctic char in the alpine Atna area, south-eastern Norway. Hydrobiologia2002;489:55–62.
Kato H, Onda Y, Tanaka Y. Using 137Cs and 210Pbex measurements to estimate soil redis-tribution rates on semi-arid grassland in Mongolia. Geophys J Roy Astron Soc 2010;114:508–19.
Kawai TIn: Kawai T, editor. An illustrated book of aquatic insects of Japan. Tokyo: TokaiUniversity Press; 1985.
Kinoshita N, Sueki K, Sasa K, Kitagawa J, Ikarashi S, Nishimura T, et al. Assessment of indi-vidual radionuclide distributions from the Fukushima nuclear accident coveringcentral-east Japan. Proc Natl Acad Sci 2011;108:19526–9.
Kruyts N, Delvaux B. Soil organic horizons as a major source for radiocesium biorecyclingin forest ecosystems. J Environ Radioact 2002;58:175–90.
Merritt RW, Cummins KW. An Introduction to the aquatic insects of North America. 3rded. Dubuque, IA: Kendall/Hunt; 1996.
Ministry of Education, Culture, Sports, Science, Technology (MEXT). Database on the re-search of radioactive substances distribution. http://radioactivity.nsr.go.jp/ja/con-tents/5000/4930/24/1305819_0727.pdf, 2011. [10 January 2014].
Ministry of Education, Culture, Sports, Science, Technology (MEXT). Database on the re-search of radioactive substances distribution. http://radb.jaea.go.jp/mapdb/portals/101002/?prefectures=09&category_b=b1201&year=2012, 2012. [10 January 2014].
Ministry of the Environment. Database on the environmental monitoring research of ra-dioactive substances in East Japan. http://www.env.go.jp/jishin/monitoring/result_pw111216-2.pdf, 2011. [10 January 2014].
Ministry of the Environment. Database on the environmental monitoring research of ra-dioactive substances in East Japan. http://www.env.go.jp/jishin/monitoring/result_pw131108-1.pdf, 2013. [10 January 2014].
Ohara T, Morino Y, Tanaka A. Atmospheric behavior of radioactive materials fromFukushima Daiichi Nuclear Power Plant. J Natl Inst Public Health 2011;60:292–9.
Report of the Chernobyl Forum Expert Group ‘Environment’. Environmental con-sequences of the Chernobyl accident and their remediation: twenty years ofexperience. Vienna, Austria: Radiological Assessment Reports Series, IAEA;2006.
Tochigi prefecture. Data base on the dose rate in the air in Tochigi Prefecture fromMay 13to May 31. http://www.pref.tochigi.lg.jp/kinkyu/documents/20110601_1400_50cm.pdf, 2011. [10 January 2014].
Tochigi prefecture. Data base on the activity concentrations in stream and lake sand inTochigi Prefecture. http://www.pref.tochigi.lg.jp/d03/kankyosho_houshaseibusshitu.html, 2013. [10 January 2014].
Ugedal O, Forseth T, Jonsson B, Njastad O. Sources of variation in radiocaesium levels be-tween individual fish from a Chernobyl contaminated Norwegian lake. J Appl Ecol1995;32:352–61.
Wakiyama Y, Onda Y, Mizugaki S, Asai H, Hiramatsu S. Soil erosion rates on forestedmountain hillslopes estimated using 137Cs and 210Pbex. Geoderma 2010;159:39–52.
Yablokov AV, Nesterenko VB, Nesterenko AV. Chernobyl, consequences of the catastrophefor people and the environment. Ann N Y Acad Sci 2009;1181.
Yoshimura M, Akama A. Radioactive contamination of aquatic insects in a stream im-pacted by the Fukushima nuclear power plant accident. Hydrobiologia 2014;722:19–30.
