Geochemical Journal, Vol. 49, pp. e9 to e14, 2015
doi:10.2343/geochemj.2.0381
EXPRESS LETTER
Distribution and supply mechanisms of high uranium concentration
in the rivers of Okinawa Island, Japan
AKIHITO MOCHIZUKI,* KO HOSODA and MASAHITO SUGIYAMA
Graduate School of Human and Environmental Studies, Kyoto University, Yoshida-nihonmatsu-cho, Sakyo, Kyoto 606-8501, Japan
(Received March 27, 2015; Accepted June 18, 2015; Online published July 13, 2015)
Uranium concentrations in the rivers of Okinawa Island, Japan were determined in this study. Most rivers in the
southwestern area of the island showed high U concentrations, the highest of which was 87 and 11 times the average U
levels in Japanese rivers and worldwide, respectively. The U concentration in two rivers exceeded the Japanese guideline
for public waters. Thermodynamic calculations revealed that the predominant U species in the rivers was Ca2UO 2(CO3) 30
(aq) because of high concentrations of calcium and carbonate ions. However, the U concentrations were not explained by
either the congruent dissolution of limestone or the input of seawater and/or sea salt aerosols. Therefore, selective dissolution of U from limestone as well as from other rocks and soils via the formation of calcium-uranyl-carbonate complexes
may be a possible mechanism for the high riverine U concentrations.
Keywords: uranium, river, Okinawa Island, guideline value, supply mechanism
ceeded either Japanese or WHO guidelines. The average
concentration was 0.17 nmol L–1, which was lower than
the average of 1.3 nmol L–1 across major rivers worldwide (Palmer and Edmond, 1993). However, the U concentrations in the Kokuba and Hija Rivers, located in the
southwestern area of Okinawa Island, were 6.30 nmol
L–1 and 4.20 nmol L–1, which were 37 and 25 times the
average in Japanese rivers and the highest and secondhighest values in Japan, respectively. These U concentrations were comparable to the range of 4.2–7.1 nmol L–1
in the Brahmaputra River (Palmer and Edmond, 1993;
Chabaux et al., 2001), which is a major river with known
high U concentration (Palmer and Edmond, 1993).
In this study, we investigated the distribution of dissolved U in 18 rivers throughout Okinawa Island. Most
of the rivers in the southern area of the island showed
high U concentrations, two of which exceeded the Japanese guideline for U levels in public waters. In addition,
we determined the U concentration in spring water and
rocks near the rivers, and we discussed the mechanisms
that may produce high riverine U concentrations.
INTRODUCTION
Uranium has the potential to threaten human health
because of its chemical and, to a lesser extent, radioactive toxicity (e.g., Zaroma et al., 1998). In 2004, the World
Health Organization (WHO) set a guideline value for U
in drinking water of 63 nmol L–1 (15 µg L –1), which was
revised to 126 nmol L–1 (30 µg L–1) in 2011 (WHO, 2011).
Several countries also monitor U levels and have established domestic guideline values for U in drinking and
public waters; for example, 126 nmol L–1 (30 µg L –1) is
the guideline set by the United States Environmental Protection Agency (2000) and 8.4 nmol L –1 (2.0 µg L–1) is
the guideline set by the Japanese Ministry of the Environment (2004). If U concentrations in drinking or public waters exceed guideline values, the source and transport mechanisms of U, as well as the impacts on human
health, should be thoroughly investigated. In particular,
it is essential to discriminate whether high U concentrations are the result of natural processes or anthropogenic
contamination.
U concentrations were previously determined in 194
rivers throughout Japan (Mochizuki and Sugiyama, 2012),
ranging from 0.002 nmol L–1 to 6.30 nmol L–1; none ex-
MATERIALS AND METHODS
Study site
Okinawa Island (26°30′ N, 128°00′ E) is located approximately 650 km southwest from the four main islands
of Japan (Fig. 1a). The geology of Okinawa Island can be
classified into three areas based on geological structure:
*Corresponding author (e-mail: [email protected])
Copyright © 2015 by The Geochemical Society of Japan.
e9
Fig. 1. (a) Sampling locations on Okinawa Island (b) Geological features of Okinawa Island reported by Konishi (1965).
the Kunigami, Motobu, and Shimajiri belts (Fig. 1b;
Konishi, 1965). The Kunigami belt covers the northeastern and central areas of the island and is composed of
Mesozoic sedimentary and metamorphic rocks such as
sandstone, schist, and phyllite. The Motobu belt forms
Motobu Peninsula in the northwest area of the island and
is composed of Paleozoic sedimentary rocks such as limestone, sandstone, and chert. The Shimajiri belt lies in the
southern area of the island and is composed of two layers; its lower part is composed of Neogene marl (calcare10
A. Mochizuki et al.
eous mudstone) and mudstone, and its upper part is Quaternary limestone.
Since water chemistry is strongly influenced by the
geological features in these areas, Tohyama (1981) classified the rivers on Okinawa Island into three areas, similar to the geological classification of Konishi (1965). In
this study, we adopted the classification of Tohyama
(1981) and referred to the three areas as the Kunigami,
Motobu, and Shimajiri areas.
Sample collection
We conducted river water sampling on Okinawa Island in February 2012, March–April 2013, April 2014,
and June 2014. We collected either two or three samples
from each river in February 2012 and one sample from
each river during the other periods. We also collected
limestone and marl samples from outcrops in April and
June 2014 and spring water samples in June 2014. The
water and rock sampling stations are shown in Fig. 1a.
Sampling stations were numbered in ascending order with
downstream distance.
The Genka, Noha, Benoki, Aha, Fukuchi, Teima,
Kushi-okawa, Maja, and Okukubi Rivers are located in
the Kunigami area; the Yabu, Oi, and Manna Rivers are
located in the Motobu area; and the Tengan, Hija,
Futenma, Asato, Kokuba, and Mukue Rivers are located
in the Shimajiri area (Fig. 1b). All of the rock samples
and spring water samples were collected in the Shimajiri
area.
At each water sampling station, we measured water
temperature (WT) and electrical conductivity using a
portable electrical conductivity analyzer (CM-21P, Toa
DKK), dissolved oxygen concentration (DO) using a portable DO analyzer (DO-24P, Toa DKK), and pH via the
colorimetric method. Each water sample was immediately
filtered using a 0.45-µm polyvinylidene fluoride Millex®
filter (Millipore) attached to a polypropylene syringe. The
filter was washed with the water sample before filtration
and we confirmed that this filtration process did not contaminate samples. Two aliquots of filtrate were stored in
low-density polyethylene bottles. One was acidified to
pH 2 by adding ultrapure nitric acid (60%, Kanto Kagaku)
for major cation and trace element analysis, whereas the
other was stored without acid addition for major anion
analysis.
Analytical methods
The concentration of dissolved U was determined via
inductively coupled plasma mass spectrometry (ICP-MS)
(Element 2TM, Thermo Fisher Scientific) by using Bi as
an internal standard element (Mochizuki and Sugiyama,
2012). Ten mL of either the sample or a standard solution
was added into a polyethylene centrifuge tube spiked with
100 µL of bismuth stock solution (1 mg L–1) and 100 µL
of ultrapure nitric acid (60%).
The concentrations of Na and K were determined via
atomic emission spectrometry (Solaar iCE 3300, Thermo
Fisher Scientific). The concentrations of Mg, Ca, Sr, and
Ba were determined via inductively coupled plasma
atomic emission spectrometry (ICP-AES) (Optima
5300DV, PerkinElmer). The concentrations of Cl and SO4
were determined via ion chromatography (L-7000 series,
Hitachi and TSKgel IC-Anion-PW column, Tosoh). Alkalinity was determined by acid titration (0.05 mol L–1
Table 1. Dissolved U concentration (nmol L–1) in the rivers of
Okinawa Island. Bold-faced values exceed the Japanese guideline value of 8.4 nmol L–1 (2.0 µg L–1) for U in public waters
(Japanese Ministry of the Environment, 2004)
Sampling periods
February
2012
Kunigami area
Genka-10
Genka-20
Noha
Benoki
Aha
Fukuchi
Teima
Kushi-okawa
Maja
Okukubi-10
Okukubi-20
Motobu area
Yabu-10
Yabu-20
Oi
Manna
Shimajiri area
Tengan
Hija-10
Hija-20
Hija-30
Futenma
Asato
Kokuba-10
Kokuba-20
Kokuba-30
Mukue
0.008
0.011
March−April
2013
April
2014
June
2014
0.030
0.135
0.078
0.024
0.050
0.038
November
2010*
0.009
0.091
0.064
0.407
0.346
0.885
1.29
0.584
1.20
0.796
0.416
1.00
0.749
2.88
3.65
3.62
9.88
4.53
4.87
4.20
2.98
7.80
5.85
5.13
3.63
4.88
3.62
10.2
1.44
6.74
5.04
6.20
6.30
14.8
9.78
12.7
*Mochizuki and Sugiyama (2012).
of H2SO4) to pH 4.8. The concentration of PO4 was determined colorimetrically with an air-segmented continuous-flow analysis system (Auto-Analyzer II, BL Tech)
by using the ascorbic acid-molybdenum blue method.
Limestone and marl were decomposed according to
the method of Sugiyama (1996), with some modifications.
The rock samples were coarsely crushed by a stainless
steel hammer, powdered by a tungsten carbide mortar and
pestle, and dried at 105°C overnight. Approximately 100
mg of the limestone powder or 10 mg of the marl powder
was weighed in a 30 mL Teflon® vessel and mixed with
0.3 mL of 60% perchloric acid, 1.0 mL of 60% nitric acid,
and 1.0 mL of 48% hydrofluoric acid. The mixture was
heated at 120°C for 2 h, 150°C for 2 h, and 170°C for 8 h
and then dried completely at 170°C. The residue was
mixed with 0.5 mL of 60% perchloric acid and 1.0 mL of
60% nitric acid and was heated again to dryness at 170°C.
The residue was then dissolved in 6 mL of 0.5 mol L–1
nitric acid. The U and Ca concentrations in the solution
were determined via ICP-MS and ICP-AES, respectively.
Uranium in rivers of Okinawa, Japan
e11
Table 2. Concentrations of major elements and U in spring waters located in the Shimajiri area
Springs
Takaraguchi
Samukawa
Shiohira
Kadeshi
Yoza
Shimajiri Rivers average
pH
Na
(mmol L−1)
K
(mmol L−1)
Ca
(mmol L−1)
Mg
(mmol L−1)
Cl
(mmol L−1)
Alkalinity
(meq L−1)
SO4
(mmol L−1)
U
(nmol L−1)
7.2
7.2
7.4
7.1
7.3
7.71
1.11
1.43
1.22
1.07
1.39
2.62
0.230
0.122
0.070
0.114
0.452
0.199
1.95
2.48
2.07
2.44
3.25
2.27
0.212
0.241
0.239
0.300
0.497
0.777
0.872
1.63
1.14
1.02
1.30
1.65
3.89
5.15
3.61
3.85
5.47
5.11
0.269
0.522
0.367
0.625
0.645
0.894
1.14
1.86
1.05
1.02
1.73
5.56
Fig. 2. Piper diagram for river water samples collected in
Okinawa Island during March–April 2013. The percentage of
HCO 3 was calculated by assuming that the alkalinity was the
same as the equivalent concentration of HCO3.
Fig. 3. U and Ca concentration in the rivers of the Motobu and
Shimajiri areas. Solid, dotted, and dot-dashed lines represent
the U/Ca mole ratios for the limestone in the Shimajiri area
(0.290 nmol/mmol), typical limestone (0.84 nmol/mmol; Palmer
and Edmond, 1993), and marl in the Shimajiri area (10.5 nmol/
mmol), respectively.
RESULTS AND DISCUSSION
U concentration in Okinawa Island rivers
The dissolved U concentration in the rivers of Okinawa
Island is shown in Table 1. Concentrations in the
Kunigami, Motobu, and Shimajiri areas were 0.008–
0.584, 0.416–1.29, and 0.749–14.8 nmol L–1, respectively.
In addition to the Kokuba and Hija Rivers where high U
concentrations were observed in a previous study
(Mochizuki and Sugiyama, 2012; Table 1), some rivers
in the Shimajiri area (Futenma, Asato, and Mukue Rivers) showed high U concentrations greater than the range
of 0.002–2.05 nmol L–1 for all other Japanese rivers (see
Mochizuki and Sugiyama, 2012 and references therein).
Moreover, the concentrations exceeded the average of 1.3
nmol L –1 across all major world rivers (Palmer and
Edmond, 1993) and were comparable to a range of 4.2–
7.1 nmol L–1 for the Brahmaputra River and 8.4–16.7 nmol
L –1 for the Ganges River (Palmer and Edmond, 1993;
Chabaux et al., 2001).
e12
A. Mochizuki et al.
Influence of seawater on riverine U concentration
The U concentration in all samples of the Mukue River
and in two samples of the Kokuba River collected at
Kokuba-10 station (boldface in Table 1) exceeded the
Japanese guideline of 8.4 nmol L–1 (2.0 µg L–1) for U in
public waters (Japanese Ministry of the Environment,
2004). The Japanese Ministry of the Environment (2009)
had also previously reported U concentrations higher than
the guideline value in a few Japanese rivers. Most were
attributed to the mixing of river water and seawater near
the estuary because the typical U concentration in
seawater can reach 14 nmol L–1 (Nozaki, 2001). In addition, Tohyama (1981) suggested that the water chemistry
of rivers on Okinawa Island is strongly influenced by the
input of sea salt aerosols because the island is very small.
Therefore, U inputs via the mixing of seawater and via
sea salt aerosols are possible sources of riverine U in the
Shimajiri area.
Assuming that all of the dissolved Cl in rivers is derived from seawater and that the U/Cl ratio in seawater
does not change during the mixing of seawater with river
water or the transport of sea salt aerosols, the riverine U
concentration derived from seawater can be estimated by:
[U]* = [U/Cl]Sea × [Cl]
(1)
where [U]* is the riverine U concentration derived from
seawater, [U/Cl]Sea is the U/Cl mole ratio in seawater (2.5
× 10–8; Nozaki, 2001), and [Cl] is the dissolved Cl concentration in river water (Supplementary Table S1). The
[U]* in rivers of the Shimajiri area was 0.02–0.08 nmol
L–1 and accounted for only 0.2–1.8% of the riverine U
concentration. Therefore, the high U concentration in
these rivers was not attributed to the seawater inputs. This
finding was contrary to those in other Japanese rivers in
which the U concentration was higher than the domestic
guideline.
The contribution of seawater to riverine U concentration in the Motobu area was also considered to be minor
because the calculated [U]* accounted for only 2.0–4.9%
of the riverine U concentration. Conversely, [U]* accounted for 16–144% of the U concentration in most of
the rivers in the Kunigami area, which suggested that
seawater inputs may influence the riverine U concentrations in this area to some extent.
Influence of limestone on riverine U concentration
A Piper diagram of river waters on Okinawa Island is
shown in Fig. 2. Sodium and Cl were dominant in most
rivers of the Kunigami area, whereas Ca and HCO3 (measured as alkalinity) were dominant in rivers of the Motobu
and Shimajiri areas. These results suggest that the river
water chemistry is influenced by sea salt aerosols in the
Kunigami area, by the dissolution of limestone in the
Motobu area, and by the dissolution of both limestone
and marl in the Shimajiri area.
High riverine U concentrations in the Motobu and
Shimajiri areas (Table 1) suggest that limestone is a major source of U. Figure 3 shows U and Ca concentrations
in the rivers of the Motobu and Shimajiri areas. The average U/Ca mole ratio in the limestone collected in the
Shimajiri area was 0.290 nmol/mmol and the typical mole
ratio for limestone is 0.84 nmol/mmol (Palmer and
Edmond, 1993). Most of the U/Ca mole ratio values from
the Motobu area were between the two values, suggesting that the riverine U concentration in the area can be
attributed to the congruent dissolution of limestone in
which the release of elements from solid to solution occurs with the same mole proportions. Conversely, most
of the data in the Shimajiri area showed higher U/Ca ratios than those expected from the congruent dissolution
of limestone and cannot be explained by that process
alone. Results similar to that of rivers in the Shimajiri
area have also been reported for the Brahmaputra, Ganges, and Indus Rivers, and their high U concentrations
have been attributed to the weathering of rocks such as
black shale and granite, which have much higher U concentrations than limestone (Palmer and Edmond, 1993;
Singh et al., 2003). However, such uraniferous rocks have
not been reported in the Shimajiri area.
Possible mechanism for high U concentration
As discussed above, the input of seawater and the congruent dissolution of limestone are likely not major supply mechanisms for high U concentrations in the rivers
of the Shimajiri area. Although the major source of U in
these rivers is not known at present, we propose some
possible mechanisms here.
Tohyama (1981) suggested that the dissolution of marl
also influences riverine water chemistry in the Shimajiri
area. The U/Ca ratio in a marl sample collected in the
Shimajiri area for this study was 10.5 nmol/mmol (Fig.
3), higher than the ratio in the rivers. However, the U
concentration in spring water in the Shimajiri area, where
the weathering of marl and mudstone influences water
chemistry (Agata et al., 2001), was significantly lower
than the average U concentration in the rivers, although
the concentrations of major elements were similar in the
springs and rivers (Table 2). The contribution of the dissolution/weathering of marl and mudstone to riverine
water chemistry, including to the U concentration in the
Shimajiri area, should be investigated in future studies.
It is well known that uranyl ions (UO22+) form soluble complexes with carbonate ions to become
UO2(CO3)22– and UO2(CO3)34– (Langmuir, 1978), as well
as the ternary complexes of alkaline earth metals-uranylcarbonate MUO2(CO3)32– and M 2UO2(CO3)30 (aq), where
M is Mg, Ca, Sr, or Ba (Dong and Brooks, 2006). We
estimated the chemical species of U using MINEQL+ and
its database (Schecher and McAvoy, 2006), and the formation constants of the alkaline earth metal-uranyl-carbonate complexes (Dong and Brooks, 2006).
Ca2UO 2(CO3)30 (aq) was the predominant species of dissolved U in the rivers of the Motobu and Shimajiri areas,
whereas the predominant species in most rivers of the
Kunigami area was UO 2 (HPO 4 ) 2 2–. The rivers in the
Shimajiri area showed higher alkalinity and Ca concentration than those in the Motobu and the Kunigami areas
(Table S1). The saturation index (log(Q/Ksp), where Q is
the ion product and Ksp is the solubility product) for calcite was higher in the Shimajiri area (–0.41 to 1.30, average of 0.67) than in the Motobu area (–0.16 to 0.41, average of 0.18). Therefore, the formation of the calciumuranyl-carbonate complex may selectively increase the
dissolution of U from rocks and soils and enhance the
Uranium in rivers of Okinawa, Japan
e13
mobility of U in these rivers. Such selective dissolution
of U may occur at the interface between river water and
sediments or between surface runoff and the rocks and
soils. In addition, the desorption of uranyl species from
particulate matter and sediment in a high pH environment
(around 8.0; Table S1) may also produce high riverine U
concentrations in the area.
Phosphate fertilizers used in paddy fields in the
Shimajiri area (Shigematsu et al., 2007) may be another
source of riverine U. Phosphate fertilizer contains up to
840 nmol g–1 of U and is regarded as a major contamination source of U in natural waters (e.g., Zielinski et al.,
2000). It is well known that most U derived from phosphate fertilizer is retained in agricultural soils in Japan
(Tagami and Uchida, 2006). However, U may be released
from agricultural soils in the Shimajiri area if calciumuranyl-carbonate complexes are formed because of high
alkalinity and Ca concentration in surface waters. Additional research should be conducted on the source and
supply mechanisms that lead to high U concentrations in
the rivers of the Shimajiri area.
Acknowledgments—We thank Prof. Y. Sohrin of Kyoto University for providing the facilities for the ICP-MS analysis. We
also thank associate editor K. Yamamoto and two anonymous
reviewers for their constructive comments.
REFERENCES
Agata, S., Satake, H. and Tokuyama, A. (2001) Chemical characteristics and isotopic compositions of spring and river
waters in Okinawa Island. Chikyu Kagaku (Geochemistry)
35, 27–41 (in Japanese with English abstract).
Chabaux, F., Riotte, J., Clauer, N. and France-Lanord, C. (2001)
Isotopic tracing of the dissolved U fluxes of Himalayan rivers: Implications for present and past U budgets of the Ganges-Brahmaputra system. Geochim. Cosmochim. Acta 65,
3201–3217.
Dong, W. and Brooks, S. C. (2006) Determination of the formation constants of ternary complexes of uranyl and carbonate with alkaline earth metals (Mg2+, Ca 2+, Sr2+, and
Ba 2+) using anion exchange method. Environ. Sci. Technol.
40, 4689–4695.
Japanese Ministry of the Environment (2004) Environmental
quality standards for water pollution. Available at http://
www.env.go.jp/en/water/wq/wp.pdf
Japanese Ministry of the Environment (2009) Review of the
environmental quality standards for water pollution on the
protection of human health (second report). Available at
https://www.env.go.jp/council/09water/y090-21/mat03.pdf
(in Japanese).
Konishi, K. (1965) Geotectonic framework of the Ryukyu Islands (Nansei-shoto). Jour. Geol. Soc. Japan 71, 437–457
(in Japanese with English abstract).
e14
A. Mochizuki et al.
Langmuir, D. (1978) Uranium solution-mineral equilibria at low
temperatures with applications to sedimentary ore deposits. Geochim. Cosmochim. Acta 42, 547–569.
Mochizuki, A. and Sugiyama, M. (2012) Natural background
concentration of dissolved uranium in Japanese rivers. Jpn.
J. Limnol. 73, 89–107 (in Japanese with English abstract).
Nozaki, Y. (2001) Elemental distribution. Encyclopedia of
Ocean Sciences (Steele, J. H., Thorpe, S. A. and Turekian,
K. K., eds.), 840–845, Academic Press.
Palmer, M. R. and Edmond, J. M. (1993) Uranium in river water. Geochim. Cosmochim. Acta 57, 4947–4955.
Schecher, W. D. and McAvoy, D. C. (2006) MINEQL+: A
chemical equilibrium modeling system (ver. 4.6). Environmental Research Software.
Shigematsu, T., Yoshida, A., Yamashita, N. and Tanaka, H.
(2007) The effect of reclaimed wastewater on the soil and
water environment under the dry-field irrigation. Environ.
Eng. Res. 44, 39–47 (in Japanese with English abstract).
Singh, S. K., Dalai, T. K. and Krishnaswami, S. (2003) 238 U
series isotopes and 232Th in carbonates and black shales from
the Lesser Himalaya: implications to dissolved uranium
abundances in Ganga-Indus source waters. J. Environ.
Radioactiv. 67, 69–90.
Sugiyama, M. (1996) Simultaneous multi-element analysis of
aquatic suspended particulate matter. Bunseki Kagaku 45,
667–675 (in Japanese with English abstract).
Tagami, K. and Uchida, S. (2006) Use of a natural U/Th concentration ratio for estimation of anthropogenic uranium
concentration in Japanese agricultural soils due to application of phosphatic fertilizers. Radioisotopes 55, 71–78 (in
Japanese with English abstract).
Tohyama, E. (1981) The water qualities of the streams on
Okinawa Island and their origins II: The geochemical study
of the stream waters in Shimajiri area. Kogyo Yosui 277, 7–
14 (in Japanese).
United States Environmental Protection Agency (2000) Technical fact sheet: Final rule for (non-radon) radionuclides in
drinking water. United States Environmental Protection
Agency.
WHO (World Health Organization) (2011) Uranium in drinking-water: Background document for development of WHO
guidelines for drinking-water quality. WHO.
Zaroma, M. L., Tracy, B. L., Zielinski, J. M., Meyerhof, D. P.
and Moss, M. A. (1998) Chronic ingestion of uranium in
drinking water: A study of kidney bioeffects in humans.
Toxicol. Sci. 43, 68–77.
Zielinski, R. A., Simmons, K. R. and Orem, W. H. (2000) Use
of 234U and 238U isotopes to identify fertilizer-derived uranium in the Florida Everglades. Appl. Geochem. 15, 369–
383.
S UPPLEMENTARY MATERIALS
URL (http://www.terrapub.co.jp/journals/GJ/archives/
data/49/MS381.pdf)
Table S1