Long-term mobility of fallout 90Sr in ploughed soil, and 90Sr uptake by wheat grain

Science of the Total Environment 372 (2007) 595 – 604 www.elsevier.com/locate/scitotenv

Long-term mobility of fallout 90 Sr in ploughed soil, and uptake by wheat grain

90

Sr

N. Yamaguchi a,⁎, K. Seki b , M. Komamura a , K. Kurishima c a

Soil Environment Division, National Institute for Agro-environmental Sciences, 3-1-3, Kan-non-dai, Tsukuba, Ibaraki, 305-8604, Japan b Faculty of Agriculture, The University of Tokyo, 1-1-1, Yayoi, Bunkyo, Tokyo, 113-8657, Japan c WDB Co., Ltd., 1-6-1 Takezono, Tsukuba, Ibaraki, 305-0032, Japan Received 18 May 2006; received in revised form 5 October 2006; accepted 10 October 2006 Available online 21 November 2006

Abstract In this study, we evaluated the mobility of 90Sr in ploughed upland soil, which affects the residual amount in the soil and plant uptake on the basis of long-term monitoring data. Paired samples of soil and wheat grain were taken annually from 1961 to 1995 from 8 agricultural fields in Japan, and the concentrations of exchangeable 90Sr in soil and total 90Sr in wheat grain were determined. The concentration of exchangeable 90Sr in ploughed soil decreased exponentially with time. The environmental factor responsible for the decrease of exchangeable 90Sr in the ploughed layer, λe, was determined from the monitoring data of exchangeable 90Sr in the ploughed soil and the amount of fallout-derived deposition. The λe was larger from 1970 to 1980 than it was from 1980 to 1995, suggesting that an easily removable fraction of 90Sr in soil was preferentially lost from ploughed soil. Among various soil properties that we investigated, the main factor controlling the long-term mobility of 90Sr from ploughed upland soil and 90Sr uptake by wheat grain was the cation-exchange capacity (CEC) of soil. Our experimental results indicate that the entrapment of 90Sr on a cation-exchange site retards the downward migration and wheat uptake of 90Sr from ploughed soil. The empirical parameters that we obtained based on the long-term observation of a wheat-cultivated upland field in Japan could be used as reference data in order to roughly estimate the mobility of 90Sr in ploughed soil and soil-borne 90Sr transfer to wheat grain in the humid Japanese climate. © 2006 Elsevier B.V. All rights reserved. Keywords: Radiostrontium; Ploughed soil; Mobility; Decreasing trend; Wheat; Concentration ratio

1. Introduction Though more than 25 years have elapsed since the latest atmospheric nuclear test in 1980, long-lived artificial radionuclides, such as 137Cs and 90Sr, remain in the soil, and those trace amounts are still being absorbed by agricultural products (Komamura et al., 2005; Tsukada et al., 2005), resulting in a potential pathway of ⁎ Corresponding author. Tel./fax: +81 29 838 8433. E-mail address: [email protected] (N. Yamaguchi). 0048-9697/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2006.10.011

radiation to humans. The major source of 137 Cs and 90Sr in soils in Japan is the global fallout derived from the testing of nuclear weapons. Once deposited on soil surface, the fallout-derived radionuclides tend to migrate down to subsurface layers. The migration characteristics of radionuclides in soil have been shown to vary depending on the soil properties, climatic conditions, land use, and management practices (Baes and Sharp, 1983; Fernandez et al., 2006; Ivanov et al., 1997). A notable characteristic of the Japanese climate is high annual precipitation, which enhances the downward

596

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

migration of 90Sr in soil. On the other hand, the higher organic matter contents and cation exchange capacity of Andosols, the volcanic ash soils typically utilized in agricultural fields, could retard the downward migration of 90 Sr. Strontium is mostly present as an easily extractable form in soil, either complex with organic or bound to the exchange sites on clay and organic matter. The vertical distribution of radionuclides has been extensively investigated in order to predict the fate of radionuclides deposited on soil by using soils taken from the Chernobyl exclusion zone (Ivanov et al., 1997; Lujaniene et al., 2002) or by adding various concentrations of an experimental tracer to soil (Cline and Rickard, 1972; Forsberg and Strandmark, 2001; Fernandez et al., 2006). In addition to the vertical distribution, the long-term decrease in radionuclides from the surface soil would provide critical information leading to a better understanding of the fate of radionuclides in soils (Fresquez et al., 1998). The mobility of 90Sr in soil also controls the 90Sr uptake by plants, which affects the radiation dose to humans. The objective of this study was to find major factors among soil properties and climate affecting the mobility of 90Sr in ploughed soil based on long-term monitoring survey data of fallout-derived 90 Sr concentrations in wheat-field soils in Japan. The

relationships between 90Sr uptake by wheat grain and soil properties are also discussed. 2. Materials and methods 2.1. Sample collection A set of ploughed-layer soil and wheat grain was collected annually from each of eight upland fields at national or prefectural agricultural experimental stations in Japan (Fig. 1). The sampling periods, depths of ploughed layers, and relevant soil properties are shown in Table 1. Two kg of ploughed soil was collected from one or various sites in each field when wheat was harvested. The sampling spot was randomly selected in the same field every year. The depth of the ploughed layer was measured every year, and changes in the ploughed-layer depth were negligibly small over the 35-year sampling period. 2.2. Sample preparation and analysis Soils were air-dried, and large aggregates formed after air-drying were gently crushed with a mortar and then passed through a 2-mm sieve. Exchangeable fractions of 90Sr were extracted from 200 g of soil with 400 ml of 1 mol L− 1 ammonium acetate at pH 7 for

Fig. 1. Sampling sites.

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

597

Table 1 Physico-chemical characteristics of soils Sampling Site

Sapporo

Sampling period

1962–1996 1966–1995 1961–1995 1961–1995 1963–1995 1980–1994 1973–1994 1967–1995

Soil classification a

Hydric Andosols

Haplic Gleysols

Hydric Andosols

Haplic Fluvisols

Haplic Andosols

Tephric Fluvisol

Haplic Fluvisols

Eutlic Gleysols

15

15

15

15

15

15

15

15

LiC 0.70 5.4 76 16.4 0.9 1.7 0.3 48.0

HC 0.81 5.9 52 24.7 0.6 3 0.3 40.0

CL 0.63 5.9 96 17.7 0.6 1.8 0.2 45.0

LiC 1.51 6.1 14 13.9 4.0 1.1 0.1 23.8

LiC 0.65 6.1 49 18.3 3.2 2.1 0.2 39.4

CL 0.83 5.0 11 7.0 0.6 1.4 0.2 16.3

SCL 0.65 6.2 14 11 1.3 1.7 0.1 14.6

CL 1.13 6.3 19 8.3 0.2 0.7 0.2 13.2

18

3.7

19

16

25

12

30

30

18

6.8

28

17

22

37

34

23

Depth of ploughed layer Soil texture b Bulk density pH(H2O) Total carbon Exchangeable Ca Exchangeable Mg Exchangeable K Exchangeable Na Cation exchange capacity

Particle size distribution wt.%

Mean annual precipitation

cm kg L− 1

Nagaoka

Morioka

Iwanuma

Mito

Kumagaya

Futaba

San-yo-cho

g kg− 1 cmolc kg− 1 cmolc kg− 1 cmolc kg− 1 cmolc kg− 1 cmolc kg− 1 2.0– 0.2 mm 0.2– 0.02 mm 0.02– 0.002 mm b2 μm

31

38

33

29

28

31

18

24

33

52

20

39

26

21

19

23

mm

1111

2304

1242

1275

1347

1284

1115

1198

(LiC: light clay; CL: clay loam; HC: heavy clay; SCL: sandy clay loam). a Soil classification is based on world reference base for soil resources. b Determined using the soil texture triangle of the International Society of Soil Science.

12 h. One hundred mg of Sr as a carrier was added to the extract, and the extract was evaporated to dryness. The residue was decomposed by concentrated HNO3 and HCl and filtered by a quantitative filter paper (Advantec No.6; size of precipitate retained: 3 μm), and the filtrate was evaporated to dryness. Strontium was precipitated as SrCO3 by the addition of ammonium carbonate after Ca removal by treatment with fumic HNO3. Radioactive impurities were eliminated by scavenging on BaCrO4 and Fe hydroxide. To calculate the recovery rate, the precipitate of SrCO3 was weighed after being dried in an oven at 110 °C. After 2 weeks, the SrCO3 was dissolved in HCl, Y was collected on a filter by scavenging on Fe hydroxide, and β rays emitted from 90Y were counted by using a 4π or 2π gas-flow low-background counter (Aloka LBC-480, SC-511). For selected soil samples collected in 1970 and 1980, total 90Sr concentrations were also determined. For the analysis of the total 90Sr concentration in soil, 100 g of soil was heated to 450 °C and extracted by 1 L of 12 mol L− 1 HCl with heating at 80 °C for 3 h after the addition of 50 mg of Sr as a carrier. The extract was filtered, and 250 mg of Ca was added to the filtrate. The resulting carbonate precipitate, obtained at pH 10, was dissolved in HNO3, and Sr and Ca oxalates were precipitated at pH 4.0 to 4.2 by the

addition of oxalic acid. The obtained oxalate precipitate was dissolved in HCl, and then Ca and radiochemical impurities were separated from Sr by using a cationexchange resin (Dowex 50WX8). Beta rays emitted from 90Y were counted as mentioned above. Wheat grain (1 kg) was ashed in an electric furnace, the temperature of which was gradually raised to 450 °C over a period of 22.5 h, maintained at 450 °C for 6 h, increased to 500 °C over 30 min, and maintained at 500 °C for 11 h. The gray–black ash samples were pulverized by a mortar. For wheat samples collected from 1985 to 1995, 90Sr was extracted from the ashed and powdered samples with hydrochloric acid. β rays emitted from 90Y were counted after separating 137Cs with the precipitate of ammonium phosphomolybdate– chloroplatinic acid followed by the radio-chemical separation of 90Sr as described above. The analyses of 90 Sr in soil and wheat were performed according to the official standard procedure for 90 Sr determination proposed by the Science and Technology Agency in Japan (1983); the procedure has also been described in detail elsewhere (Komamura et al., 2005; Tsukada et al., 2005). The accuracy of measurement was checked annually using standard soil and wheat samples prepared by the addition of a known amount of 90Sr.

598

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

2.3. Decrease in exchangeable

90

Sr from ploughed soil

The concentration of exchangeable 90 Sr in the ploughed soil changes according to the following differential equation: dCðtÞ ¼ DðtÞ−ka CðtÞ dt

ð1Þ

ka ¼ ke þ k p

ð2Þ

ke ¼ kl þ kpl þ kage ;

ð3Þ

where C(t) is the exchangeable 90Sr concentration in the ploughed layer of soil (Bq kg− 1), D(t) is 90Sr added to the soil through radioactive fallout (Bq kg− 1 year− 1), and λa is the observed decay constant of 90Sr (year− 1). Here, we defined λa, the observed decay constant, as the sum of the environmental decay constant (λe) and the physical decay constant (λp) because the decrease in exchangeable 90Sr in the ploughed soil is caused not only by physical decay but also by other environmental factors, such as downward migration from the ploughed layer to the subsurface layer (leaching), plant uptake in the ploughed layer, and transformation into nonexchangeable 90Sr in the ploughed layer (aging). By assuming that the rate of decrease of exchangeable 90Sr due to these three factors is proportional to the concentration of exchangeable 90Sr in the ploughed layer, the first-order rate constants of leaching, plant uptake, and aging can be defined as λl, λpl, and λage, respectively, and the environmental decay constant can be expressed as the sum of these constants (Eq. (3)). In Eqs. (1) (2) and (3), the depth of the ploughed layer should be constant throughout the observation period. In deriving these equations, we made two assumptions. The first assumption is that the ploughed layer is homogeneously mixed every year and the concentration of exchangeable 90Sr does not change with the depth of the ploughed layer. The second assumption is that all of the fallout 90Sr deposited to the soil surface becomes exchangeable in soil. Although the second assumption is not necessarily clear in some cases, we chose to simplify the problem by making the assumption, rather than increasing the number of uncertain parameters in the model, because at this stage we cannot accurately estimate how much of the deposited 90Sr becomes non-exchangeable. If an assumption similar to that made by Takahashi et al. (2000) is used, D(t) can be calculated from the total amount of deposited 90 Sr (Bq m− 2 year − 1 )

obtained from the database of the Japan Chemical Analysis Research Center (2006), the depth of the ploughed layer (d cm), and the bulk density of soils (b kg L− 1) by DðtÞ ¼

If ðtÞ : 10db

ð4Þ

D(t) obtained by Eq. (4) can be substituted into Eq. (1) for the calculation of λa. In deriving Eq. (4), it was assumed that all additional 90Sr from atmospheric fallout reached the soil surface. For a more precise analysis, this assumption needs to be verified because, in practice, some fractions of atmospheric fallout will be trapped by wheat or a rotation crop and will not reach the soil surface. The general solution to Eq. (1) can be expressed as Z CðtÞ ¼ expð−ka tÞfC þ

DðtÞexpðka tÞdtg;

ð5Þ

where C is the constant of integration. In order to determine the value of λa from observed data, this equation was solved for two cases: when D(t) N 0 and when D(t) = 0. As shown in Results and discussion, the λa calculation was performed during two periods: the former period, from 1970 to 1980, where D(t) N 0, and the latter period, after 1980, where D(t) = 0. Case 1. D(t) N 0 Since the data of D(t) is given as the annual average value, we assume here that in the one-year period of ti ≤ t b ti + 1 (ti = 0,1,…), D(t) is constant and thus can be written as D(t) = D(ti). In this one-year period, Eq. (5) becomes 

 Dðti Þ CðtÞ ¼ Cðti Þ− expf−ka ðt−ti Þg ka Dðti Þ ðti VtVti þ 1Þ; þ ka

ð6Þ

therefore, substituting t = ti + 1 into this equation yields Cðti þ 1Þ ¼ Cðti Þe−ka þ Dðti Þ

1−e−ka : ka

ð7Þ

During the observation period, D(ti) (ti = 0,1,2, …) is given as the annual deposition; if we set C0 = C(0) and λa, the values of C(ti) (ti = 1,2, …) can be calculated from Eq. (7). Thus, by minimizing the root mean square error of ln(C(ti)) by non-linear regression, the optimized value of λa (and C0) can be determined.

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

599

Table 2 Total and exchangeable 90Sr concentrations in soil and ratios of exchangeable to total 90Sr Total

90

Sr

1970

Exchangeable 1980

90

1970

Ra

Sr 1980

1970

1980

4.0 ± 0.12 8.1 ± 0.17 3.1 ± 0.10

0.62 0.86 0.67 0.80 0.57

0.87 0.68 0.36

Bq kg− 1 Sapporo Nagaoka Morioka Iwanuma Mito Kumagaya Futaba San-yo-cho a

R¼

9.3 ± 0.30 22.7 ± 0.62 12.9 ± 0.18 8.4 ± 0.24 6.1 ± 0.07

4.6 ± 0.29 12.0 ± 0.24 8.7 ± 0.17

5.8 ± 0.29 19.6 ± 0.18 8.7 ± 0.54 6.7 ± 0.25 3.5 ± 0.26

2.9 ± 0.11 2.4 ± 0.16

3.3 ± 0.13 2.3 ± 0.14

2.7 ± 0.18 2.8 ± 0.23

2.9 ± 0.20 2.2 ± 0.08

1.02 0.92

0.81 1.22

Exchangeable 90 SrðBq kg −1 Þ : Total 90 SrðBq kg −1 Þ

Case 2. D(t) = 0, In this case, Eq. (5) becomes ð8Þ

CðtÞ ¼ C0 expð−ka tÞ;

where C0 = C(0). λa can, therefore, be calculated from the slope of the linear regression of the logarithmic plot of C(t). The observed and environmental residence halftimes (RHTa and RHTe) of 90Sr in ploughed soil are calculated by RHTa ¼

ln2 and ka

RHTe ¼

ln2 : ke

ð9Þ

ð10Þ

Once λa is calculated, λe, RHTa, and RHTe are calculated from Eqs. (2), (9) and (10), respectively. 3. Results and discussion 3.1. Decreasing trends of exchangeable ploughed soil

90

Sr from

Table 2 shows the total and exchangeable 90Sr concentrations of soils collected in 1970 and 1980 and ratios of exchangeable to total 90Sr. Most 90Sr in soil was in exchangeable form except for that in Andosols collected in 1980 from Morioka, where the ratios of exchangeable to total 90Sr was 0.36. The lower ratio of exchangeable 90Sr in Morioka was likely related to properties of the sorption site for 90Sr in soil. The cationexchange sites of allophone, imogolite, and humus, rich in Andosols, exhibit a higher affinity for Ca2+ over NH4+

(Okamura and Wada, 1984). In addition, Tsumura et al. (1984) showed that 83% of 90Sr spiked to humic acid extracted from Andosols was extracted by neither water nor 1 M CH3COONH4. No clear tendency was found if the ratios of exchangeable to total 90Sr were decreased or increased in 10 years from 1970 to 1980. Examination of the fate of the non-exchangeable 90Sr would be also important, but in this paper, only the long-term mobility of the exchangeable 90Sr is analyzed. Fig. 2 shows the exchangeable 90Sr concentrations in ploughed soil (C(t)) from 1961 to 1995. The concentrations of 90Sr in soil were related to the cumulative amounts of 90Sr that had been deposited on the soil surface as fallout due to atmospheric nuclear tests. The effects of the Chernobyl reactor accident in 1986 on the increase in 90Sr concentrations in soil were not very remarkable in Japan (Komamura et al., 2005). Regional differences in the amounts of fallout deposition were attributed to metrological and geomorphologic differences, and the fallout deposition on the Japan Sea side tended to be higher than that on the Pacific Ocean side (Aoyama et al., 1996). Among the observation plots that we investigated, the highest concentration of 90Sr in soil was observed in Nagaoka on the Japan Sea side. Once 90Sr is deposited on the soil surface, the behavior of 90Sr in soil should be mainly controlled by the soil properties and precipitation amounts. Although annual fluctuations were found, the concentration of 90 Sr in ploughed soil decreased after 1965 in most of the observation plots, suggesting that the leaching rates of 90 Sr from ploughed soil exceeded the accumulation rates through fallout deposition after 1965. Assuming that all 90Sr in fallout reached the soil surface and that the ploughed layer was homogeneously mixed every year, the contribution of the annual addition of fallout-derived 90Sr (D(t)) to ploughed soil (C(t))

600

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

Fig. 2. Changes in exchangeable 90Sr concentration in soil from 1961 tο 1995.

accounted for, at most, 3 to 30% in 1965 and 1 to 5% from 1970 to 1980. After 1980, the contribution of falloutderived 90Sr to soil was less than 1% because no atmospheric nuclear test resulting in 90Sr deposition on Japanese soil took place after 1980. Because the contribution of fallout-derived 90Sr on the changes in 90 Sr concentrations in ploughed soil was significant before 1970, we used the monitoring data after 1970 to calculate λa, λe, RHTa, and RHTe. Table 3 shows λa and λe as well as RHTa and RHTe. We divided the entire observation period from 1970 to 1995 into two stages: 1970 to 1980 (rapid stage) and 1980 to 1995 (slow stage) because of the different effects of 90 Sr deposited on soil through fallout. The rapid and slow components are denoted by the subscripts r and s, respectively. The observed decay constants, λar and λas, were larger than the physical decay constant (0.0241 year− 1), and RHTar and RHTas were shorter than the radioactive half-time, 28.8 year, as well. This finding indicates that, besides physical decay, the environmental factor was important for the decrease in the 90Sr concentration in soil. The rapid (RHTer) and slow (RHTes) components of the environmental residence halftime of 90Sr in the ploughed layer were in the ranges of 3.52 to 16.7 years and 9.93 to 32.1 years, respectively. The environmental residence half-time of 90Sr in a wheat field was longer than that of 90Sr in a Japanese paddy field (Takahashi et al., 2000). For example, the environmental residence half-time of the paddy field in Mito was 6.5 years (1987–1995; depth of ploughed layer: 13 cm),

whereas that of the wheat field in this study was 32.1 year (1980–1995; depth of ploughed layer: 15 cm; Table 1). Despite the similarities in soil and climate, 90Sr leached to depths of 80 cm in the paddy field and 45 cm in the upland field in Mito (Yamaguchi et al., unpublished results). The residence half-time for the upland field was much longer Table 3 Decay factors and residence half-time 1970–1980

Sapporo Nagaoka Morioka Iwanuma Mito Kumagaya a Futaba San-yo-cho

λar

λer

RHTar

RHTer

0.0656 0.0852 0.0963 0.0707 0.1102

0.0415 0.0611 0.0722 0.0467 0.0861

10.6 8.14 7.20 9.80 6.29

16.7 11.3 9.60 14.9 8.05

0.221 0.133

0.197 0.109

3.14 5.20

3.52 6.35

1980–1995

Sapporo Nagaoka Morioka Iwanuma Mito Kumagaya Futaba San-yo-cho

λas

λes

RHTas

RHTes

0.0467 0.0555 0.0513 0.0576 0.0457 0.0861 0.0858 0.0939

0.0226 0.0314 0.0272 0.0335 0.0216 0.0620 0.0617 0.0698

14.9 12.5 13.5 12.0 15.2 8.05 8.08 7.38

30.7 22.1 25.5 20.7 32.1 11.2 11.2 9.93

The subscript r denotes rapid component; s denotes slow component. a No monitoring data in Kumagaya before 1980.

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

than that for paddy field; therefore, such a small difference in the depth of the ploughed layer alone cannot explain the difference in RHT. In the paddy field, water-flooding management during rice cultivation was likely to have accelerated the downward migration of 90 Sr from ploughed soil. Although the downward migration of 90 Sr from ploughed soil was smaller in the upland wheat field than in the submerged paddy field, it is still an important factor in the decrease in 90Sr and is discussed in more detail below. As shown in Table 3, the rapid component of the decay factor is indeed more rapid than the slow component; i.e., λer is larger than λes. This finding agrees with a report by Takahashi et al. (2000), in which the λe for 90Sr in the paddy field was composed of rapid and slow components. The rapid component was predominant in the early stage, and the slow component was predominant in the later stage because most of the easily removable fractions of 90 Sr were lost in the early stage. In deriving Eqs. (7) and (8), we have made several assumptions, the validity of which affect the reliability of the results. The assumptions for D(t), that all 90Sr deposited as fallout reaches the soil surface without being trapped by plants and becomes exchangeable in the soil, seem unrealistic. Therefore, we consider the calculation of the slow component more reliable than that of the rapid component, where the value of D(t) is not negligible. The slow component, λes, is discussed in detail later in this paper. 3.2. Possible causes of environmental decreases of 90Sr in ploughed soil The λes in the bottom section of Table 3 indicates that the decreasing trends of 90Sr from ploughed layers differed from region to region. Possible environmental factors controlling the decreasing trends of exchangeable 90 Sr in ploughed soil are the following: (1) the downward migration of 90Sr from the ploughed layer to the subsurface soil with water (λl); (2) the uptake of 90Sr by plants (λpl); and (3) the decrease in the ratio of exchangeable to total 90Sr due to aging (λage) (Rigol et al., 1999). The effects of these three environmental factors on the decrease in exchangeable 90Sr from ploughed soil varied depending on the soil properties and climate, which differed from region to region. Among these factors, the first one, downward migration, should have the greatest effect on the decrease in 90Sr from the ploughed soil layer. In addition, the uptake of 90Sr by the wheat plant body should also have an important role in reducing 90Sr from ploughed soil. For example, in 1985, exchangeable 90Sr in soil in Mito was 1.3 Bq kg− 1, and the slow component of the environmental decay constant in

601

Mito was 0.02160. According to Eqs. (1) and (2), the amount of decrease of 90Sr from ploughed soil due to environmental factors in one year was estimated to be 0.028 Bq kg− 1 (λe ×C(t) = 0.02160 × 1.3). In that year, the wheat crop yield and the 90Sr concentration in the wheat grain were 3 t ha− 1 and 178 mBq kg− 1, respectively, in Mito. The calcium content in the wheat grain amounted to 7% of that in the entire plant (Tsukada et al., 1997). Since the distributions of Sr and Ca in the wheat were quite similar (White, 2001), the uptake of 90Sr by the entire wheat plant was estimated to be 0.0078 Bq kg− 1 soil, assuming that 7% of 90Sr was distributed in the wheat grain and the remainder was in the non-edible part. The uptake of 90Sr by the entire wheat plant accounted for 28% of the decreased amounts of 90Sr from ploughed soil due to an environmentally controlled factor in 1985. The contributions of wheat uptake to the decrease in 90Sr from ploughed soil differed according to crop yield, year, and region of harvest; the average was 30% in Mito. As there is 90 Sr uptake by some rotation crops other than wheat, the average λpl/λe ratio was more than 30% (λpl N 0.3λe) in Mito. Regarding the effect of aging, in this study, we did not determine the total 90Sr concentrations in ploughed soil every year; as a result, we could not find any evidence that exchangeable 90Sr decreased due to aging. Rigol et al. (1999) revealed a decrease in the exchangeable 90Sr fraction after 8 months or more, whereas Forsberg and Strandmark (2001) showed that, even after extended aging, 90Sr remained labile. According to Forsberg and Strandmark (2001), it is reasonable to assume that 5 to 10 years after deposition, a steady state in terms of plant availability is reached if the fallout is in a soluble form. At least for λes, the aging effect could be negligible. The accurate determination of each environmental factor, λl, λpl, and λage, from our data was not possible, but we were able to infer that λl N λpl N λage; i.e., the effect of downward leaching is the most important factor, the effect of plant uptake is of secondary importance, and the effect of aging might be least important. 3.3. Relationships of λes to soil properties and precipitation Table 4 shows the correlation coefficients among λes and relevant soil properties and precipitation. There were significant relationships between λes and cation exchange capacity (ρ = 0.833, p b 0.05, two-sided), indicating that the entrapment of 90Sr on the cationexchange site retarded the downward migration and plant uptake of 90Sr from ploughed soil. The difference in λes was not explained by the organic carbon content or clay content of soils alone but by a combination of

602

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

Table 4 Correlation coefficients between environmental decay factors of 90Sr from ploughed soil (λes) or mean concentration ratios of 90Sr in wheat grain to exchangeable 90Sr in soil from 1985 to 1995 (CRex) and the soil properties/annual precipitation amounts Soil properties

Bulk density pH(H2O) Total carbon Exchangeable Ca Exchangeable Mg Exchangeable K Exchangeable Na Exchangeable Ca + Mg Exchangeable Ca + Mg + Na + K CEC Base saturation Coarse sand (2.0–0.2 mm) Fine sand (0.2–0.02 mm) Silt (0.02–0.002 mm) Clay (b0.002 mm) Mean annual precipitation

Spearman's rank correlation coefficients λes

CRex

0.587 0.301 − 0.683 − 0.809⁎ − 0.464 − 0.719 − 0.386 − 0.738⁎ − 0.810⁎ − 0.833⁎ 0.429 0.120 0.429 − 0.347 − 0.310 − 0.143

−0.060 0.554 −0.587 −0.619 −0.171 −0.371 −0.617 −0.571 −0.619 −0.857⁎⁎ 0.405 0.659 0.761⁎ −0.731⁎ −0.762⁎ −0.143

⁎⁎p b 0.01, ⁎p b 0.05 (two-sided).

organic carbon and clay content, [λes Bq kg− 1 year− 1] = − 0.00068 [Clay %] − 0.00043 [OM g kg− 1] + 0.078, R2 = 0.74. This was related to the fact that CEC depends on the total carbon and clay contents. A plausible site for 90 Sr entrapment is soil organic matter and clay minerals, which provide a negative charge in the pH ranges of the soils observed. The exchangeable Ca and the sum of exchangeable Ca and Mg were also significantly related to λes. It is noteworthy that the amounts of exchangeable cations varied depending on the land management, and it is inadequate to evaluate the relationships with a long-term reduction of 90Sr. Nonetheless, the dependence of λes on exchangeable Ca and the sum of exchangeable Ca and Mg are reasonable because of the chemical similarity of Sr to Ca. The mobility of 90Sr in soil and its absorption by plants have often been related to the ratio of 90Sr to Ca, i.e., Sr unit (Squire, 1966; Sysoeva et al., 2005). In general, the amount of leaching depends on the climate, soil properties, and land management (Baes and Sharp, 1983). No clear relationships were found (ρ = − 0.143) between λes and the mean annual precipitation at our observation plots (Table 4), however, probably because there were no clear differences in the precipitation amounts at the observation plots, except at Nagaoka (Table 1). As the climate condition, precipitation, did not contribute to the different λes values, the soil properties were the primary factor

affecting the different λes values in this study. Among the observation plots that we investigated, different CECs were most responsible for the different λes values, indicating that CEC would be a good measuring tool for the estimation of the decreasing trend of 90Sr from ploughed soil, perhaps because the rates of leaching and plant uptake depend on CEC. 3.4. Relationships of 90Sr uptake by wheat grain to soil properties and precipitation The environmental decay factor, λe, is mainly composed of leaching and plant uptake, as stated above. In considering an internal radiation dose to humans through radioactivity in food, it is critical to estimate how much of a fraction of 90Sr in ploughed soil is absorbed by agricultural products. In addition to the amount of 90Sr in soil, the availability of 90Sr in soil should determine the 90Sr concentrations in plants. Two absorption pathways are considered for 90Sr uptake by wheat grain: a direct absorption pathway through 90Sr directly deposited on the plant body, and an indirect pathway via the absorption of soil-borne 90Sr through the roots. Before 1980, when radioactive fallout deposition was significant, the direct absorption pathway predominantly contributed to 90Sr uptake by wheat grain (Komamura et al., 2002). In response to the decrease in atmospheric fallout deposition, the contribution of soil-borne 90Sr to wheat uptake became the major absorption pathway. According to Komamura et al. (2002), the contribution of indirect absorption in Japan became nearly 100% after 1985. The Chernobyl accident in 1986 caused a significant increase in the 137 Cs concentration in wheat grain harvested in Japan, whereas the increase in the 90Sr concentration was less important (Komamura et al., 2002). The ratio of the 90Sr concentration in wheat grain to the exchangeable 90Sr concentration in soil was defined as CRex in this study, and the CRex was in the range of 0.05 and 0.2, depending on the sampling site (Table 5). The mean CRex between 1985 and 1995 was significantly related to CEC (ρ = − 0.857, p b 0.01, two-sided, Table 4). The data collected in 1986 were not used for calculations to avoid a possible influence of direct absorption of 90Sr due to the Chernobyl accident. It has been shown in numerous experiments that soil properties such as CEC, organic matter content, and clay content are important to determine the uptake of radionuclides by plants (e.g., van Bergeijk et al., 1992; Roca et al., 1997; Sysoeva et al., 2005). We showed that both λes and CRex had significant linear relations with CEC (Table 4).

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

603

Table 5 Concentration ratios of 90Sr in wheat grain to exchangeable 90Sr in soil from 1985 to 1995 (CRex) Year

Sapporo

Nagaoka

Morioka

Iwanuma

Mito

Kumagaya

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Average Standard deviation

0.0448 0.0530 0.0464 0.0608 0.0556 0.0691 0.0333 0.0596 0.0652 0.0436

0.0510 0.0631 0.0431 0.0542 0.0435 0.0516 0.0376 0.0429 0.0641 0.0547 0.0975 0.0540 0.0171

0.1445 0.0634 0.0648 0.0450 0.0485 0.134 0.0317 0.0381 0.0585 0.0198 0.0487 0.0633 0.0420

0.0524 0.0594 0.0494 0.0384

0.0989 0.0854 0.0759 0.0871 0.0536 0.0788 0.0556 0.0478 0.0954 0.0900 0.0761 0.0759 0.0181

0.220 0.251 0.071 0.118 0.122 0.147 0.120

0.0532 0.0118

As indicated above, CEC would be a good measure for the estimation of how much exchangeable 90Sr would be lost from soil due to leaching down to subsurface soil and absorption by plants based on the long-term monitoring data of fallout 90Sr in ploughed soil and wheat grain. This result had been expected based on the results of studies under controlled laboratory conditions, such as adsorption experiments (Khasawneh et al., 1968; Solecki, 2005) and pot experiments using an experimental tracer (Roca et al., 1997; Sysoeva et al., 2005; van Bergeijk et al., 1992). In real field conditions, there are numerous factors controlling 90Sr mobility in soil and its uptake by plants. Despite the numerous factors that could affect the mobility of elements in the soil environment of real field soil, this study revealed that the CEC of soil is the most important factor controlling 90Sr mobility and availability in soil. Acknowledgements This work was supported by a fund for radioactivity surveys in Japan from the Ministry of Education, Culture, Sports, Science, and Technology. We would like to express our gratitude to Messrs. N. Kihou, H. Fujiwara, and Mses. T. Hirose, S. Suzuki, and H. Morita for their assistance with sample preparation; to Dr. H. Tsukada of the Institute for Environmental Sciences for introducing us to the reference data on 90Sr distribution in the wheat plant; and to the staff members of the following institutions for providing us with samples for analysis over many years: Miyagi Prefecture Furukawa Agricultural Experiment Station, Ibaraki Agricultural Center, Saitama Prefectural Agriculture and Forestry Research Center, Yamanashi Prefectural Agricultural Research Center, Okayama Prefectural General Agriculture Center, and the incorporated administrative

0.0290 0.126 0.0556 0.0585 0.0345

Futaba 0.452 0.189

0.230 0.0838

0.135 0.133 0.124 0.166 0.177

0.139 0.0583

0.154 0.0266

San-yo-cho 0.0700 0.200 0.198 0.190 0.173 0.117 0.059 0.141 0.156 0.197 0.165 0.147 0.0501

agencies of the National Agricultural Research Center for the Hokkaido Region, National Agricultural Research Center Hokuriku Research Center, and National Agricultural Research Center for the Tohoku Region. References Aoyama M, Hirose K, Igarashi Y, Miyao T. Geochemical studies and analytical methods of anthropogenic radionuclides in fallout samples. Tech Rep Meteorol Res Inst 1996;36:1-34 [in Japanese]. Baes III CF, Sharp RD. A proposal for estimation of soil leaching and leaching constants for use in assessment models. J Environ Qual 1983;12:17–28. Cline JF, Rickard WH. Radioactive strontium and cesium in cultivated and abandoned field plots. Health Phys 1972;23:317–24. Fernandez JM, Piault E, Macouillard D, Juncos C. Forty years of 90Sr in situ migration: importance of soil characterization in modeling transport phenomena. J Environ Radioact 2006;87:209–26. Forsberg S, Strandmark M. Migration and chemical availability of 137 Cs and 90Sr in Swedish long-term experimental pastures. Water Air Soil Pollut 2001;127:157–71. Fresquez PR, Armstrong DR, Mullen MA. Radionuclides in soils collected from within and around Los Alamos national laboratory: 1974–1996. J Environ Sci Health 1998;A33(2):263–78. Ivanov YA, Lewyckyj N, Levchuk SE, Prister BS, Firsakova SK, Arkhipov NP, et al. Migration of Cs-137 and Sr-90 from Chernobyl fallout in Ukrainian, Belarussian, and Russian soils. J Environ Radioact 1997;35(1):1-21. Japan Chemical Analysis Research Center. Environmental radioactivity database. http://search.kankyo-hoshano.go.jp/top.jsp2006 May. Khasawneh FE, Juo ASR, Barber SA. Soil properties influencing differential Ca to Sr adsorption. Soil Sci Soc Am Proc 1968;32:209–11. Komamura M, Tsumura A, Kihou N, Kodaira K. 90Sr and 137Cs contamination of wheat produced in Japan. Survey and analysis during the years 1959 through 1995 including the Chernobyl accident. Radioisotopes 2002;51:345–63 [in Japanese with English summary]. Komamura M, Tsumura A, Yamaguchi N, Kihou N, Kodaira K. Monitoring 90Sr and 137Cs in rice, wheat, and soil in Japan from 1959 to 2000. Miscellaneous Publications of the National Institute for Agro-Environmental Sciences, vol. 28.; 2005. p. 1-56.

604

N. Yamaguchi et al. / Science of the Total Environment 372 (2007) 595–604

Lujaniene G, Plukis A, Kimtys E, Remeikis V, Jankunaite D, Ogorodnikov BI. Study of Cs-137, Sr-90, Pu-239, Pu-240, Pu238, and Am-241 behavior in the Chernobyl soil. J Radioanal Nucl Chem 2002;251(1):59–68. Okamura Y, Wada K. Ammonium–calcium exchange equilibria in soils and weathered pumices that differ in cation-exchange materials. J Soil Sci 1984;35:387–96. Rigol A, Roig M, Vidal M, Rauret G. Sequential extractions for the study of radiocesium and strontium dynamics in mineral and organic soils from Western Europe and Chenobyl areas. Environ Sci Technol 1999;33:887–95. Roca MC, Vallejo VR, Roig M, Tent J, Vidal M, Rauret G. Prediction of cesium-134 and strontium-85 crop uptake based on soil properties. J Environ Qual 1997;26:1354–62. Science and Technology Agency, Japan. Analytical method of radioactive strontium. Chiba: Japan Chemical Analysis Center; 1983. p. 1-50. [in Japanese]. Solecki J. Investigation of 85Sr adsorption on selected soils of different horizons. J Environ Radioact 2005;82:303–20. Squire HM. Long-term studies of strontium-90 in soils and pastures. Radiat Bot 1966;6:49–67. Sysoeva AA, Konopleva IV, Sanzharova NI. Bioavailability of radiostrontium in soil: experimental study and modeling. J Environ Radioact 2005;81:269–82.

Takahashi T, Komamura M, Uchida S. Application of 2-component model to the removal of 90Sr in paddy-field soil. Jpn J Health Phys 2000;35:359–64 [in Japanese with English abstract]. Tsukada H, Hasegawa H, Hisamatsu S, Sakurai N. Concentrations of elements in cereal plant components and soil-to-grain transfer factors. Annual report of the Institute for Environmental Sciences; 1997. p. 14–6. Tsukada H, Takeda A, Takahashi T, Hasegawa H, Hisamatsu S, Inaba J. Uptake and distribution of 90Sr and stable Sr in rice plants. J Environ Radioact 2005;81:221–31. Tsumura A, Komamura M, Kobayashi H. Behavior of radioactive Sr and Cs in soils and soil-plant systems. Bull Natl Inst Agric Sci Ser B 1984;36:57-113 [in Japanese with English abstract]. van Bergeijk KE, Noordijk H, Lembrechts J, Frissel MJ. Influence of pH, soil type, and soil organic matter content on soil-to-plant transfer of radiocesium and-strontium as analyzed by a nonparametric method. J Environ Radioact 1992;15:265–76. White JP. The pathways of calcium movement to the xylem. J Exp Bot 2001;52:891–9.