Abstract
Background After the Fukushima Daiichi Nuclear Power Plant accident, various surveys were performed to measure the extent of radioactive material contamination in the marine sediments, surface waters, plankton and fish. Six months after the event, relatively high radiocesium contamination could still be detected in the fish and sediment of the coastal demersal environment. To determine the distribution of radioactive material in the demersal ecosystem adjacent to the event, we sampled and analyzed the dominant macroorganisms attached to the seafloor as well as environmental material, including biofilms, from the coastal demersal environment around Hisanohama Port, which is less than 30 km south of the accident site.
Results Our results showed that variable ratios of radiosilver/radiocesium occurred in the macrobenthos, macroalgae and sediment samples. However, the biofilm samples displayed high radiocesium contamination but did not show radiosilver contamination.
Conclusions These findings suggest that several different entry paths are available for radioactive elements to access the biological components of the ecosystem, and these paths may explain the disproportional and patchy distribution of radioactive elements among marine ecosystems.
Background
Following the Great East Japan Earthquake and catastrophic tsunami in March of 2011, the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident led to the release of large amounts of radioactive materials into the atmosphere and marine environment.
This environmental input of radioactive elements from the FDNPP introduced into the marine waters a variety of radioisotopes, especially radiocesium isotopes. Public monitoring showed that during the several months after the first explosion at the FDNPP, a rapid decrease in radioisotopes was observed in the nearby seawater. A joint official survey team from the Tokyo Power Electric Company (TEPCO), Fukushima prefectural government and the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) reported that the 137Cs contamination level in the seawater quickly decreased from over 10 k Bq/liter to less than 100 Bq/liter within a 2-month period, whereas the 137Cs found in the ocean sediments was continuously observed over this period at levels between 100 Bq/dry kg and 1000 Bq/dry kg until at least September 2012 [1]. This disproportional distribution of radioisotopes was also observed in marine fish by Wada et al. Wada reported that the retention time of radiocesium appears to be remarkably longer in the larvae of demersal fish, e.g., Sebastes cheni, than in surface fish, e.g., Eugraulis japonica [2]. Similar disproportional distributions among radionuclides was also reported in a terrestrial ecosystem. Nakanishi et al. reported 110mAg accumulation in the orb-web spider (Nephila clavata) and other arthropods [3]. Decay-corrected data from Nakanishi’s work showed that the radioactivity ratio between 110mAg and 137Cs (110mAg/137Cs) in specific samples of arthropods was over 1.0, whereas the same ratio (110mAg/137Cs) in the soil was 0.0014-0.0023, which indicates that these organisms may actively accumulate radiosilver more than radiocesium. This finding is inconsistent with the majority of the organisms from land ecosystems, which accumulate more radiocesium than radiosilver.
These disproportional distributions may occur via partitioning in complex ecosystems and food webs through processes that include biological concentration. To better understand the fate of radioisotopes in nearshore marine environments, the present study sampled water, sediment, biofilms, macrobenthos and macroalgae as basic environmental materials and entry points of organic/inorganic materials into the ecosystem. In the present study, samples were collected using a self-contained underwater breathing apparatus (SCUBA) at a depth of 5-10 m along the nearshore area around Hisanohama Port. Sampling was essentially performed weekly from the end of November 2011 to the beginning of February 2012 at 6 different points. Our results showed that macroalgae and macrobenthos accumulate and retain both radiocesium and radiosilver. The observed ratios of radiosilver/radiocesium are variable among the organisms, and they may be dependent on the specific usage and behavior of the organisms with respect to metals, such as specific cofactors, oxygen acceptors, metal uptake and other metabolic features. Although sediment also retains both radiocesium and radiosilver, biofilms contain only radiocesium. The microbial community structure is likely related to the radiocesium accumulation and retention in microbes.
Materials and Methods
Sampling
Sampling was performed from 11 November 2011 to 4 February 2012. The sampling site was the ocean along the coastline of Hisanohama, Iwaki City, Fukushima, Japan. We collected samples 10 times within the established time period at 11 Nov (week 1), 20 Nov (week 2), 26 Nov (week 3), 3 Dec (week 4), 10 Dec (week 5), 17 Dec (week 6), 14 Jan (week 7), 21 Jan (week 8), 29 Jan (week 9) and 4 Feb (week 10). Sampling was performed at 6 different areas near Hisanohama Port: Area 1 (37.153762N 141.006167E, weeks 1 and 5), Area 2 (approximately 37.149863N 141.001017E, weeks 2, 8 and 9), Area 3 (approximately 37.146579N 141.002390E, week 4), Area 3’ (approximately 37.147503N 141.008570E, week 10), Area 4 (approximately 37.133887N 141.002390E, weeks 3 and 7) and Area 4’ (approximately 37.142953N 141.003592E, week 6). These sampling sites are located 30 km south of the Fukushima Nuclear Power Plant #1. Sampling was performed by diving with a SCUBA unit. Divers sampled the bottom water and surface water (5 L) and collected bottom stones to obtain biofilms and locally observed macrobenthos and macroalgae.
Sediment core samples were collected at Area 2 using a piston corer prototype. The “core sampler” was composed of a 4.5 cm diameter/28 cm height cylinder and inscribed piston. Core samples were collected during a separate dive in weeks 6 (core sample 1) and 7 (core samples 2 – 4).
Samplings were performed with permission from Fukushima Prefecture and the fishery association of Iwaki City.
Sample preparation
The seawater samples were immediately filtered with two 24-mm-wide 0.22 µm Durapore filters (Millipore, MA, USA) sample until they clogged. All samples were frozen at −20°C in a field freezer prior to transportation back to laboratory.
In the laboratory, the samples were dried at 105°C overnight except for the filters, water and sediment core samples. The dried samples were crushed by a food mill and settled into a U8 container, and the weight and height were measured in the container. In the case of biofilms, half of the samples was dried as mentioned above and the other half was captured on 0.2 µm pore size membrane filters and stored at −20°C for ribosomal RNA gene analysis. The water samples were dried by boiling. The remaining salts were settled into a U8 container, and the weight and height were measured in the container.
Sediment core samples for radioactivity measurements were stored at −80°C. The cores were sliced by a clean saw at 5 cm intervals from the surface side of the sediment to the bottom side. Each piece of the core sample was freeze dried and then settled into a U8 container, and the weight and height measured in the container.
Measurement of radioactivity
The gamma ray activity of the samples in the containers was measured by a Germanium (Ge) detector. The detector setup is identical to those described in [4]. The energy resolution of 2 keV full width at half maximum (FWHM) was achieved at 1332.5 keV gamma rays. The sample location and the detector were surrounded with 15 cm lead blocks to reduce the background radiation in the gamma ray spectroscopy. The detection efficiency of the gamma rays was calibrated with an accuracy of 2 - 10% using a multiple gamma ray standard source, which ranged from 88 keV to 1836 keV. A calibrated 134gCs source was also used to correct for the coincidence summing for radioactivity determinations of 134gCs.
Every sample was contained in a U8 plastic container. The U8 container is a standard container in Japan that is used for measurements of absolute radioactivity with high accuracy by a Ge detector. The geometry of the container is cylindrical, with a diameter of 47 mm and height of 60 mm. The volume is approximately 100 mL. The efficiency calibration source is available within the container with a uniform density in the container.
Every sample was measured by the Ge detector for at least 8 hours to obtain sufficient statistics for each gamma ray peak. The relative geometry between the sample and the Ge detector was carefully reproduced by a guide apparatus at the sample location.
Gamma rays from 134,137Cs and 110mAg were observed for the majority of the samples. The radioactivity levels for every isotope were determined from the yield of the gamma rays corrected by the detection efficiency of the Ge detector and a geometrical acceptance between the sample and the detector. The geometry difference in the height of the sample from the calibration source was simulated by several electromagnetic simulation codes, including Geant4 [5,6] and EGS5 [7].
The radioactivity of the marine animals, macroalgae and biofilm samples were normalized by weight, and the radioactivity of the water samples and sediment core samples were normalized by volume.
Element analysis
To quantitatively analyze the content of 37 elements (7Li, 9Be, 23Na, 24Mg, 27Al, 39K, 43Ca, 45Sc, 47Ti, 51V, 52Cr, 55Mn, 57Fe, 59Co, 60Ni, 63Cu, 66Zn, 85Rb, 88Sr, 98Mo, 111Cd, 133Cs, 139La, 140Ce, 141Pr, 146Nd, 152Sm, 153Eu, 158Gd, 159Tb, 164Dy, 165Ho, 166Er, 169Tm, 172Yb, 175Lu, and 202Hg) in the samples (i.e., surface salts, surface microbes, biofilms), the samples were predigested with 5 mL of concentrated HNO3 for 1 h at room temperature. Next, the organic components were completely decomposed by wet-ashing using a microwave sample preparation system (Multi-Wave-3000, Perkin Elmer, MA USA) [8]. The digested samples were brought up to a volume of 50 mL with MilliQ water (MQW) and filtered through 5B filter paper (Advantec, Tokyo, Japan). The concentrations of the mineral elements were determined by Inductively Coupled Plasma Mass Spectrometry (ICP-MS, NexION300, Perkin Elmer). For the ICP-MS analysis, a portion of the filtrated samples was diluted appropriately with 0.01 mol L-1 HCl [8] The mineral nutrient concentration in the samples was calculated as a unit (mg g-1 in dry weight).
The obtained ICP-MS data were normalized by the root-sum-of-squares (RSS) levels among each element. Count data of radioactive cesium (134Cs and 137Cs) were treated with the same calculation for normalization and then merged with the normalized ICP-MS dataset (elements table).
ssrDNA analysis
Genomic DNA was prepared from stored biofilm samples by bead-beating with a phenol: chloroform: isoamyl alcohol (PCI) extraction. Bead-beating was performed at 3000 rpm for 5 min (TOMY, Tokyo Japan), and then nucleic acids were recovered by ethanol precipitation.
PCR amplification was performed with the extracted genomic DNA with two primer sets for 16S ssrDNA, 515F/806R [9, 10], and 18S ssrDNA, TAReuk454FWD1/TAReukREV3 [11]. The primers were modified for Illumina sequencing with a multi-index kit for the Nextera XT sequencing library construction kit (Illumina, CA, USA). We added a “read1” sequence (5’-TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG-3’) to the forward primers and a “read1” sequence (5’- GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA G-3’) to the reverse primers.
PCR was performed with a standard reaction of ExTaq DNA polymerase according to the manufacturer’s instructions (Takara Bio, Kyoto Japan) with 25 pmol of each primer/50 µl reaction. The reactions were performed as a two-step PCR (10 cycles of 94°C for 30 s, 55°C for 45 s and 72°C for 60 s followed by 20 cycles of 94°C for 30 s and 72°C for 60 s) for 16S and 3-step PCR (30 cycles of 94°C for 30 s, 55°C for 45 s and 72°C for 60 s) for 18S. The amplified products were purified by 1% agarose gel electrophoresis. The obtained amplified products were applied to the NexteraXT (Illumina CA USA) sequencing library construction process without the “tagmentation” process.
A multi-index sequencing library was applied to the MiSeq sequencing kit v1 and read by Illumina MiSeq according to the instructions in the paired-end sequencing mode. Data were treated by MiSeq reporter software, and the reads were obtained from individual samples. After the extraction of reads, reads from weeks 4, 5 and 8 for the 18S ssrDNA were insufficient for analysis. Therefore, the following analysis was performed using reads from weeks 1, 2, 3, 6, 7, 9 and 10.
The obtained read data were mapped onto the ribosomal RNA sequence library SILVA release 108 by the program package QIIME [12] using “Prefix-suffix OTU picking”, and the absolute abundance OTU table at the family level was constructed. Count zero data were manually removed from the table, and the relative abundance values were calculated. 16S data and 18S data were treated independently and then merged into a single OTU table with the relative abundance values.
Sequence data have been deposited in the DDBJ sequence read archive with accession number DRA004367.
Statistical analysis
For the statistical analysis, we used the program package R [13]. First, we drew a heatmap chart with the hierarchical cluster analysis (HCA). The heatmap.3 function in the GMD package, which is a package for non-parametric distance measurements between two discrete frequency distributions [14], was used to draw the heatmap with the HCA. Independent heatmaps were drawn from each element table and the OTU table with a two-axis HCA. A scaling option was used for the OTU table treatment according to the row direction but was not used for the elements table treatment.
A principal components analysis (PCA) was performed with the merged dataset, including both the out table and elements table, using the prcomp function in R. Score plots were drawn with principal component 1 (PC1) (x-axis) and PC2 (y-axis). The loading plots for quadrants II and III were manually drawn.
Spearman’s rank correlation coefficient between the elements dataset and the OTUs dataset was calculated with the cor function in R. A correlation matrix was used to draw the heatmap chart with the HCA using the heatmap.3 function of R.
Results and Discussion
Heterogeneous distribution of artificial radioisotopes in the coastal marine environment around Hisanohama
According to our gamma ray energy spectra, we found significant peaks corresponding to 134Cs and 137Cs in addition to natural radioisotope signals, such as 40K (data not shown), for all samples with the exception of the seawater samples. We also detected significant 110mAg signals from the macrobenthos, macroalgae and sediment samples. Significant signals of the radioisotopes 134Cs, 137Cs and 110mAg were not detected from the control samples taken from Nishi-kawana inside Tokyo Bay, which is approximately 280 kilometers from the FDNPP. Although 134Cs (half-life of 2.0652 years) and 137Cs (half-life of 30.1 years) from prior nuclear weapons tests in the Pacific Ocean and the Chernobyl disaster might have persisted in the seabed around Japan, our analysis of samples collected near Nishi-kawana did not detect these signals. Therefore, we concluded that the radioisotopes detected in the samples in this study originated from the FDNPP disaster rather than the environmental background or past nuclear weapon tests.
As presented in Table 1, among the results from the radioactivity measurements of a variety of samples, we noticed a heterogeneous distribution between radiocesium (134Cs and 137Cs) and radiosilver (110mAg). The retention of radiocesium was observed in all sediments, macrobenthos, macroalgae and biofilm samples, whereas non-detectable signals of these isotopes were observed in the surficial and bottom water. This result is surprising because the 134Cs radioactivity detected in the Pacific Ocean near the accident site was 3900 Bq/m3 (3.9 Bq/liter) in June 2011, which is one of the highest levels ever recorded [15]. Although our results do not directly indicate that marine organisms and sediments are able to accumulate radioisotopes at the concentrations found in this environment, we must consider that they are able to retain radioisotopes for at least up to 9 months, which represents the time lapse between the disaster and sample collection.
Sediment core samples were obtained from the sampling point “Area 2” by the core sampler. The analysis of these sediment core samples indicated the presence of both radiocesium and radiosilver. The measured radioactivities are shown in Additional file 1: Supplementary Table 1. To better understand the radioactivity in the same volume of space, we employ the unit Bq/liter to express the radioactivity of the core samples because the core samples contain different materials along their depth profile.
In Figure 1, the distribution of radioactivity along the sediment column is shown. Both radiocesium and radiosilver are distributed similarly in the surficial part of the core samples, which contains a large amount of radioactivity. The calculated regression line for the data showed that both radiocesium and radiosilver occur at depths above approximately 155 mm. The similar distribution of radiocesium and radiosilver suggest that both radioisotopes start to precipitate at almost the same time and they share a similar precipitation pattern. The ratio of radioactivity between 110mAg and 137Cs is 0.0046 on average, which is based on the detected radiosilver of each 5 cm slice of the core samples (Fig. 1b).
In the macrobenthos and macroalgae samples, we also detected radiosilver and radiocesium. However, the ratio between 110mAg and 137Cs varied among the samples.
As shown in Figure 2a and Additional file 2: Supplementary Table 2, macroalgae contained a relatively low ratio of radiosilver relative to radiocesium. The ratio between 110mAg and 137Cs averaged 0.06 to 0.07 among brown, red and calcareous algae (Fig. 2a); nevertheless, it varied among individuals (Fig. 2b). Except for green algae, the ratio of 110mAg/137Cs of macroalgae was approximately 10-fold larger than that of the surrounding sediments. Green algae contained almost no 110mAg.
As shown in Figures 3a and 3b and Additional file 3: Supplementary Table3, the macrobenthos samples showed a variable ratio between 110mAg and 137Cs.
Our results indicated that sea snails (Buccinidae) and oysters (Crassostrea gigas) consistently presented a high ratio of 110mAg/134Cs, which was also observed for the sponge Cliona chilensis. Samples of the Ascidian Halocynthia roretzi showed a relatively high ratio of 110mAg/134Cs (Fig. 3a).
Certain mollusks, such as sea snails and oysters, are known to use hemocyanin as an oxygen carrier [16]. A study on the relatively high ratio of 110mAg/137Cs in the spider, Atypus karschi indicated that the reaction center Cu in hemocyanin can be replaced by Ag, or 110mAg in this case [3]. Similarly, the sampled mollusks may have actively incorporated 110mAg from environmental water and taken up the radioisotope into the blood system. To corroborate this hypothesis, our results showed a higher ratio of radiosilver compared with that of radiocesium detected in the body (134Cs = 21.7923, 137Cs = 37.3885, 110mAg = 134.6388 (Bq/kg dry biomass)) than in the shell (134Cs = 12.6245, 137Cs = 18.8071, 110mAg = 17.7947 (Bq/kg dry biomass)).
A similar mechanism for accumulation and retention of the relatively high observed amount of radiosilver in Halocynthia roretzi is expected. Ascidians accumulate vanadium into their blood cells [17]. Although the function of vanadium accumulation is not yet fully understood and the mechanism of metal uptake in ascidians is still under investigation, these organisms may accumulate other metals using vanadium binding protein(s). Our samples of ascidians are from Halocynthia roretzi, and a high specificity for vanadium has not been reported in this species [18]. Our hypothesis is that a metal binding factor within the Halocynthia roretzi blood system with a low specificity for vanadium is able to promote 110mAg accumulation. Although our hypothesis is based on indirect evidence, it is partially supported by the observation that the Halocynthia roretzi body (including the blood system) showed a larger amount of radiosilver than was found in the shell, which can contain radiosilver because of attached organisms on the surface of the shell. We investigated three individuals of Halocynthia roretzi, and the body measurement results are 134Cs = 645,0563, 927.5638, 944.2312; 137Cs = 997.5926, 1437.1948, 1588.1469; 110mAg = 295.6794, 255.9421, 213.5846 (Bq/kg dry biomass) and the shell measurement results are 134Cs = 18.6579, 51.1889, 28.2958; 137Cs = 30.3881, 80.5451, 43.8699; 110mAg = not detected, 3.3797, not detected (Bq/kg dry biomass).
The sea urchin (Anthocidaris crassispina) samples and the majority of the sponge samples (Cliona chilensis) presented ratios of radiosilver/radiocesium that were lower relative to that of the other macrobenthos (Fig. 3a, b). Nevertheless, the feeding pattern of sea urchins and sea snails are similar because they are both deposit feeders, and the feeding patterns of sponges, oysters and Ascidians are also similar because they are all filter feeders. Interestingly, 4 of the 16 specimens of sponge presented remarkably high radiosilver/radiocesium ratios of approximately 0.5 (Fig. 3b, arrowheads). Variation were not observed in the radiosilver/radiocesium ratio corresponding to the three color variations of the samples of Cliona chilensis. A previous report showed that certain sponge species exhibit a strong ability to incorporate a variety of metal elements, including silver [19]. Indeed, sponges are known as "a sentinel organism" that can be used as a marker to monitor heavy metal pollution in water because they present a high capacity for specific metal accumulation [20]. According to Genta-Jouve et al., Acanthella acuta, which is a species of commonly found sponge, can incorporate 110mAg from the surrounding water and retain the element even after moving to clean water that does not contain 110mAg [19]; however, among the six sponge species analyzed in that study, this feature was only observed in Acanthella acuta. In our study, four specimens of sponge incorporated relatively high ratios of radiosilver relative to radiocesium (Fig. 3b arrowhead); therefore, we regarded these sponges as a “silver accumulator” species.
Thus, macroalgae and macrobenthic species present varying capacities to incorporate and retain radioisotopes based on their biological features. In addition, the radiosilver/radiocesium ratios of these organisms are higher than those in the sediment. Although we detected both radiocesium and radiosilver from almost all biological samples in this research, an interesting exception was observed because radiosilver was not detected in any of the biofilm samples, including biofilms collected from Area 2 (weeks 2, 8 and 9) (Table 2). This result suggested that microbes that compose biofilms specifically accumulate and retain radiocesium but not radiosilver, even when the surface of nearby sediments contains both radioisotopes.
“Cesium-philic” biofilm
Microbial communities that grow on surfaces of solid materials in the presence of water are known as biofilms. Our observations suggest that the biofilm community specifically accumulates radiocesium. Therefore, we analyzed the relationship between the microbial consortia within the biofilms and their elemental composition.
Quantified element data were combined with radioactivity data for radiocesium and utilized for the statistical analysis. Ribosomal RNA genes were sequenced and used to construct a dataset for the statistical analysis.
A HCA from the dataset of elements and radiocesium (Fig. 4) in which an accumulation of radioactive cesium, Cr, Ni, Sc, Rb, Li and cesium (Cs cluster) was performed for the week 1 sample. Another remarkable profile from the elemental analysis is observed in week 3, where high relative amounts of lanthanide elements were measured (Lanth cluster).
A PCA score plot (Fig. 5a) of the merged dataset between the elements/radiocesium and the microbial consortia showed that the week 1 sample was placed in the PC1 negative – PC2 positive direction and the week 3 sample was placed in the PC1 negative – PC2 negative direction. As mentioned above, the week 1 sample contained high relative amounts of the Cs cluster signals, and the week 3 sample contained high relative amounts of the Lanth cluster signals. Consistent with those findings, the PCA loading plot showed that the Cs cluster was placed in the PC1 negative – PC2 positive direction (Fig. 5b, Additional file 4: Supplementary Table 4), and the lanthanide cluster was placed in the PC1 negative – PC2 negative direction (Fig. 5b, Additional file 5: Supplementary Table 5). The results of the HCA and PCA analyses suggested that a “cesium-philic” biofilm occurred at the sampling point in week 1 and a “lanthanide-philic” biofilm occurred at the sampling point in week 3.
The PCA results from both datasets of elements and microbes showed candidate microbes in the “cesium-philic” and “lanthanide-philic” biofilms on the loading plots because those plots were placed in same loading direction, e.g., PC1 negative – PC2 positive (Cs cluster) or PC1 negative – PC2 negative (Lanth cluster) (Fig. 5c, Additional file 4 and 5: Supplementary Table 4 and 5). Indeed, the HCA heat map of the microbial consortia showed that the PC1 negative – PC2 negative-directed microbes mainly appeared in the week 1 column (Fig. 6 blue dots). The same tendency was observed with the Lanth cluster microbes (Fig. 6 green dots). These results suggest that these two OTU sets are positively correlated with the Lanth cluster (week 3 sample) and Cs cluster (week 1 sample). The correlation heat map chart between the element dataset and the OTU dataset supports this result
The heatmap chart has been sorted by the HCA for both axes (OTU axis and sample axis). Green and blue dots indicate contributing OTUs for the PC1 minus and PC2 plus direction and the PC1/PC2 minus direction, respectively. The relative intensity value has been scaled and centered in the row direction.
The radiocesium and Cs cluster elements are positively correlated to other multicellular eukaryotes, such as Hexacorallia and Oligochaete. For bacteria, high relative intensities of Bacteroidetes, Rhodobacteraceae and Gamma-proteobacteria were observed (Additoinal file 6: Supplementary Table 6 and Fig. 6). Although a primer set that did not target Archaea was used, Archeal sequences were obtained in this work by sequencing amplicons generated by the 16S ribosomal RNA primers set. An uncultured Crenarcheota in marine group I was observed to be one of the dominant OTUs from the biofilm sample of week 1 (Additional file 6: Supplementary Table 6).
The function of these microorganisms is poorly understood. However, Alpha- and Gamma-proteobacteria have been obtained from biofilms on the surface of nuclear fuel pools [21]. According to Sarro et al. [21], biofilms obtained from nuclear fuel pools accumulate 60Co. Cobalt accumulation in the week 1 sample was suggested by the PCA (Fig. 5b and Additional file 4: Supplementary Table 4) despite a lack of 60Co signal indicated by gamma-ray spectrography. The microbial community retrieved from the nuclear fuel pool was described as capable of accumulating both 60Co and 137Cs as reported by Tisakova et al. [22]. Our results suggest that the biofilm sampled in week 1 presents similar characteristics to the microbes described by Sarro et al. and Tisakova et al.
Conclusions
In this study, we found heterogeneous distributions of radiocesium and radiosilver among a variety of sample types from the coastal environment near the FDNPP disaster. Specific organisms that use hemocyanin, such as shell fish and Ascidians, accumulate and retain 110mAg, and similar observations have been reported in terrestrial spiders. We also found radiosilver in the sediment samples, suggesting that particular materials can be sources of radiosilver for these organisms, even if particles filtered from the water have low levels of radioactivity in our samples.
However, the biofilm samples did not display detectable signals of radiosilver, although the amount of radiocesium in these samples was higher than that found in other marine organisms. Biofilms may still act as an entry point for radiocesium to enter into the ecosystem. In this paper, we could not assess the functional mechanisms by which biofilms accumulate cesium specifically; thus, future research that includes cultivation experiments and meta-transcriptome analyses should be performed to further investigate the microbial communities described in this work and provide insights into this complex topic.
Declarations
Author’s contributions
SM, HO and JS designed research. SM, HO, KK, HY and JS performed sampling. YK and MI performed ICP-MS analysis and KS and JK partially performed sample preparation for ICP-MS and advice for statistical analysis. Sample preparation and molecular biological experiment was performed by SM and KK. Radioactivity measurement has been done by HO. SCUBA diving safety control and sampling arrangement has been done by HY and JS.
Competing interests
The authors declare no competing financial interests.
Availability of data and materials
The datasets used and/or analysed during the current study except for ribosomal RNA sequences data and datasets that showed in manuscript or supplementary tables are available from the corresponding author on reasonable request.
Ribosomal RNA sataset are available in the DDBJ sequence read archive with accession number DRA004367.
Electronic supplementary material
Additional file 1: Supplementary Table 1: Radio activity of core samples. (XLSX 44 KB)
Additional file 2: Supplementary Table 2: Radio activity of macro algae samples. (XLSX 47 KB)
Additional file 3: Supplementary Table 3: Radio activity of macro benthos samples. (XLSX 46 KB)
Additional file 4: Supplementary Table 4: Loading plot on PC1 minus - PC2 plus. (XLSX 52 KB)
Additional file 5: Supplementary Table 5: Loading plot results of PC1 minus - PC2 minus. (XLSX 51 KB)
Additional file 6: Supplementary Table 6: Cs cluster taxon in Week1 sample. (XLSX 30 KB)
Funding
This work was supported by RIKEN Incentive Research Project (FY2011) “Emergency monitoring for possible cycling of radio active compounds via oceanic food web” (SM, HO) and biofilm section was supported by MEXT Grant-in-Aid for Scientific Research on Innovative Areas Grant Number 23117003 (SM, KK).
Acknowledgements
Sampling was permitted by the Iwaki City Fishery Association and performed via the fishing vessel Shyo-ei-maru and crew, which belong to the abalone section of the Iwaki Fishery Association. The diving safety of SM was overseen by Toshiyuki Suzuki. The authors wish to acknowledge the profound academic comments and proofreading provided by R. Craig Everroad and Diogo M. O. Ogawa.