Distribution of radioactive cesium in soil and its uptake by herbaceous plants in temperate pastures with different management after the Fukushima Dai-Ichi Nuclear Power Station accident

Abstract

The accident at Fukushima Dai-Ichi Nuclear Power Station (NPS) extensively contaminated the agricultural land in the Tohoku region of Japan with radioactive cesium [sum of cesium-134 (¹³⁴Cs) and cesium-137 (¹³⁷Cs)]. We evaluated the status of radioactive cesium (Cs) contamination in soil and plants at the Field Science Center of Tohoku University, northern Miyagi prefecture, 150 km north of the NPS. In seven pastures with different management, we examined: (1) the distribution of radioactive Cs in soil, (2) the concentration of radioactive Cs in various herbaceous plant species and (3) the change in radioactive Cs content of plants as they matured. We collected samples of litter, root mat layer (root mat soil and plant roots), and subsurface soil (0–5 cm beneath the root mat) at two to three locations in each pasture in December 2011 and May 2012. The aboveground parts of herbaceous plants (four grasses, two legumes, and one forb species) were collected from May 9 to June 20, 2012, at 14-d intervals, from one to five fixed sampling locations in each pasture. The distribution of radioactive Cs in soil differed among pastures to some degree: a large proportion of radioactive Cs was distributed in the root mat layer. Pasture management greatly influenced the radioactive Cs content of herbaceous plants (p < 0.001); plant species had less influence. Radioactive Cs content was highest (> 3 kBq kg⁻¹ dry weight) on May 9 and significantly decreased with maturity (p < 0.001) for most of the pastures, whereas it remained low (0.04–0.18 kBq kg⁻¹ dry weight) throughout the measurement period in the pasture where composted cattle manure was applied. The soil-to-plant transfer factor was negatively correlated to pH(H₂O) (R² = 0.783, p < 0.001) and exchangeable K content (R² = 0.971, p < 0.001) of root mat soils, which suggests that surface application of composted cattle manure reduces plant uptake of radioactive Cs by increasing the exchangeable K content of the soil. The radioactive Cs content of plants decreased with plant maturity; its degree of decrease (May 9 to June 6) was smaller in legumes (80.6%) than grasses (55.5%) and the forb (58.6%). Radioactive Cs content decreased with plant maturity; also, the proportion remaining in the aboveground plant was higher in legumes (80.6%) than grasses (55.5%) and the forb (58.6%).

Key words:

• herbage
• pasture management history
• plant taxa
• radioactive cesium
• root mat soil

INTRODUCTION

The Great East Japan Earthquake on March 11, 2011, followed by the tsunami, triggered the accident at the Fukushima Dai-Ichi Nuclear Power Station (NPS). The radioactive fallout was dispersed by wind through several regions in eastern Japan, and was deposited during rainfall and snowfall after the accident (Chino et al. 2011; MEXT 2011; Katata et al. 2012; Terada et al. 2012). The radioactive fallout extensively polluted agricultural land, including permanent pastures and meadows, with radioactive cesium [the sum of cesium-134 (¹³⁴Cs) and cesium-137 (¹³⁷Cs)].

Extensive investigation of radioactive cesium (Cs) pollution of agricultural land after the Chernobyl disaster and atmospheric bomb testing showed that radioactive Cs fallout deposited on the soil surface remains in the litter layer or soil surface of non-cultivated land such as pasture (Riise et al. 1990; Varskog et al. 1994). Consequently, the radioactive Cs accumulated in the surface or litter layer is readily taken up by herbaceous plant roots that are distributed in these layers, resulting in relatively higher soil-to-plant transfer factors (TFs) in pastures than in tilled fields (IAEA 2010).

Plant uptake of radioactive Cs is affected by the physical and chemical factors of soil (Zhu and Smolders 2000; Ehlken and Kirchner 2002). For example, uptake of radioactive Cs by plants was increased by an increase in soil nitrogen (N) content and a decrease in potassium (K) content (Evans and Dekker 1969; Belli et al. 1995). Soil organic matter content also increases radioactive Cs uptake by plants (van Bergeijk et al. 1992; Absalom et al. 1996; Sanchez et al. 1999; Grytsyuk et al. 2006; Tulina et al. 2010), due to the lack of sufficient clay minerals to fix radioactive Cs (Rigol et al. 2002). These studies suggest that pasture management, such as fertilization and usage (e.g., herbivore grazing, cutting), affects soil-to-plant transfer of radioactive Cs. In addition, uptake of radioactive Cs is known to differ among plant species (e.g., Broadley and Willey 1997; Willey and Martin 1997). The radioactive Cs content of herbage decreased with maturity in ryegrass, whereas it increased in legumes (Paasikallio and Sormunen-Cristian 2002).

However, it is unclear whether the findings in Europe after the Chernobyl accident are relevant to the status in Japan after the NPS accident. In particular, the contamination status of pastures has not been documented as well as that of arable land (Tsuiki and Maeda 2012). The aim of this study was to measure radioactive Cs in the belowground and aboveground parts of plants from temperate pastures under different management history, and to examine whether (1) the radioactive Cs content of soils and plants differ with pasture management history, (2) the radioactive Cs content of plants differs among plant species or plant taxa and (3) the radioactive Cs content of plants decreases with maturity.

MATERIALS AND METHODS

Study area

The study was conducted in temperate pastures of the Field Science Center (FSC) at the Graduate School of Agricultural Science, Tohoku University, Japan (38°44’N, 140°15’E, 300 m elevation above sea level). The FSC is 150 km from the Tokyo Electric Power Company’s NPS (Fig. 1). The soil is classified as a Haplic, non-allophanic Andosol according to Soil Classification for Cultivated Soils in Japan (Classification Committee of Cultivated Soils 1996), or as an Alic Hapludands in Soil Taxonomy (Soil Survey Staff 1999).

Figure 1 Locations of the Fukushima Dai-Ichi Nuclear Power Plant and the Tohoku University Field Science Center (FSC) sampling site.

Seven pastures with different size and management history were studied (Table 1). Two pastures (5–2 and 14–2) were used as grazing paddocks for sheep and cattle, respectively, and the others were used as cutting meadows. In spring 2010, chemical fertilizer was applied to six pastures, and only composted cattle manure was applied to the surface of the seventh pasture (9–2). All pastures were abandoned and received no management (i.e., cultivation, fertilization, animal grazing, cutting, etc.) after the earthquake.

Table 1 The area, year established, management and sampling design for each pasture

Pasture number	Area (ha)	Established year	Usage	Chemical fertilizer^† N-P2O5-K2O (kg ha^−1 y^−1)	Composted cattle manure^†,‡(Mg ha^−1)	Targeted plant species and taxa^§	Number of locations for herbage sampling
5-2	3.0	1986	Grazing, sheep	106-53-53	–	Anthoxanthum odoratum L. (G)	4
						Trifolium repens L. (L)
9-1	2.1	2000	Cutting	213-79-79	4	Phalaris arundinacea L. (G)	3
						Rumex obtusifolius L. (F)
9-2	1.3	2010	Cutting	–	30	Dactylis glomerata L. (G)	3
						Trifolium pratense L. (L)
13-1	2.8	1992	Cutting	121-45-45	4	Dactylis glomerata L. (G)	4
						Phalaris arundinacea L. (G)
						Rumex obtusifolius L. (F)
14-2	3.6	1991	Grazing, cattle	106-53-53	–	Dactylis glomerata L. (G)	3
						Lolium perenne L. (G)	1
						Trifolium repens L. (L)	4
18-2	4.0	1996	Cutting	91-30-30	–	Dactylis glomerata L. (G)	5
						Rumex obtusifolius L. (F)
22	0.6	2000	Cutting	77-26-26	4	Phalaris arundinacea L. (G)	2

^† Soil amendment composition (nitrogen[N] - phosphate[P₂O₅] - potassium oxide[K₂O]) and the amount applied in 2010.

^‡ The manure contained 80-145-528 N-P₂O₅-K₂O (kg ha⁻¹).

^§ Plant taxa were classified as grasses (G), legumes (L) or forb (F).

Pasture measurement and collection of herbage

Two to five sampling sites (radius 10 m) were systematically allocated within each pasture at least 50 m apart from each other. The position of each site was recorded by a global positioning system (eTrex®, personal navigator®, GARMIN, Olathe, USA).

Four grasses (Dactylis glomerata L., Phalaris arundinacea L., Lolium perenne L., and Anthoxanthum odoratum L.), two legumes (Trifolium repens L. and T. pratense L.) and one forb weed (Rumex obtusifolius L.) were chosen as the target plant species (Table 1). These species were dominant in the individual pastures. Pasture air radiation dose rate, leaf mass height and ear height were measured and samples collected on May 9, May 23, June 6 and June 20, 2012; only D. glomerata was collected on June 20. Leaf mass and plant height were measured for each plant species. Ear height was measured for grass species when ear emergence was observed. The plants were cut at 3 cm above the soil surface, air-dried at 70°C for 72 h, and milled through a 2-mm screen to produce homogeneous samples. All samples were stored dry until analysis of radioactive Cs concentrations.

Air radiation dose rate of pastures

Air radiation dose rates (μSv h⁻¹) were measured with an environmental radiation monitor (γ survey meter TCS-172B, Hitachi-Aloka Medical, Ltd., Tokyo, Japan) at 1 m above the ground surface at each location. Air radiation was measured on the same day as pasture measurement, or the following day if it rained.

Soil collection and analysis

Soil samples were collected from two to three locations in each pasture in December 2011 and May 2012. After harvesting the aboveground part of the plants, a quadrat (28 × 28 cm) was set, litter on the soil surface was collected, and then soil samples were collected. In this study, root mat soil was defined as the surface soil layer that was tightly bound with plant roots and easily separated from the subsurface layer. The depth of root mat soil differed between sampling sites, and depended on the development of a tightly bound root layer. Subsurface soil, 0–5 cm depth immediately beneath of the root mat layer, was collected with a core sampler (5 cm height, 100 mL in volume). The root mat layer was air-dried and separated into roots and soil, which we designated “root” and “root mat soil”, respectively. Soil samples (root mat soil and subsurface soil) were sieved to 2 mm. Litter and roots were oven-dried at 70°C for 72 h and ground to pass through a 2-mm screen.

Radioactive Cs, pH(H₂O), and exchangeable K of root mat soil and subsurface soil samples were measured. Soil exchangeable K was extracted with a mixture of 0.05 M ammonium acetate and 0.0114 M strontium chloride (soil:extractant = 1:200) (Committee for Soil Environment Analysis 1997) and measured using an atomic absorption spectrophotometer (Hitachi Z-2300, Hitachi Co., Japan).

Measurement of radioactive Cs content of samples

The activity concentrations of ¹³⁷Cs and ¹³⁴Cs in dried soil and plant samples were determined with a gamma counter (WIZARD²® 2480, PerkinElmer, Waltham, USA) equipped with a sodium iodide (NaI) detector. We used either the measurement protocol provided by Perkin Elmer, Japan, or the protocol of Yin et al. (2012). Both protocols gave identical values. Each sample was loaded into a 20-mL plastic vial, weighed and measured for 600–1800 s. The average radioactive Cs concentration of duplicate samples was expressed as the total activity of ¹³⁴Cs and ¹³⁷Cs per unit of dry weight (dry wt.; Bq kg⁻¹ dry wt.). The standard deviation of each measurement with the gamma counter was less than 10%. All measured activities were corrected for radioactive decay to May 9, 2012, for estimates of soil-to-plant transfer. The measured activities of plant samples were corrected for each sampling date to compare the Cs concentrations in plants as they matured.

Data analysis

Radiocesium concentrations in soil and plant samples and soil properties for each pasture were summarized as geometric means. The total amount of radioactive Cs fallout (kBq m⁻²) was evaluated as the sum of the products of radioactive Cs contents (Bq kg⁻¹ dry wt.) and the amounts of material (kg dry wt. m⁻²) for each compartment (litter, plant roots, root mat soil and subsurface soil).

Plant species were classified into three taxa (i.e., grasses, legumes and a forb) before statistical analysis. A general linear model was used to analyze the factors affecting the radioactive Cs content of the plants. Radioactive Cs contents were logarithmically transformed to fit a normal distribution. The main effects were pasture, sampling date and plant taxa. The interactions included in this model were pasture × sampling date, pasture × plant taxa and sampling date × plant taxa.

The change in radioactive Cs content of the aboveground part of each plant species from each pasture was evaluated as the plants matured by calculating the proportion of radioactive Cs content on each sampling date to the content on May 9, 2012. The relative values were analyzed by a general linear model for the main effects of sampling date and plant taxa and the interaction between them. All the analyses were performed with IBM® SPSS® statistics v. 21 (IBM Corporation, New York, USA) (IBM Corporation 2012).

Pearson’s correlation analysis was performed to test whether the K content of root mat soil was related to that of subsurface soil. Regression analysis was performed to assess the relationship between soil-to-plant transfer factor of radioactive Cs (TF) and soil characteristics for each plant species. TF was defined as follows:

(1)

where all radioactive Cs contents of litter, root mat soil, roots from the root mat and subsurface soil are included in the denominator, and the concentration is expressed as a weighted average of these components. This is because the radioactive Cs in any of these components may be a source of plant uptake, and because we could neither differentiate living roots from dead root tissues nor remove all fine soil particles in the rhizoplane. Hence, the definition of TF in the present work is a little different from that widely used in the literature (IAEA 2010).

RESULTS

Distribution of radioactive Cs and chemical characteristics of the pasture soils

The pattern of dry-matter distribution between the compartments differed among pastures (Table 2). For example, in pasture 22, which was used as a cutting meadow, a large proportion of dry matter was in the root mat layer because of the well-developed root mat. The thickness of the root mat in pasture 22 was 4–5 cm, whereas in pastures 9–1, 9–2 and 13–1 it was only about 1 cm because of poorly developed root mats. Subsurface soil accounted for the highest proportion of belowground dry material (19.1–29.7 kg m⁻²), whereas the proportion in litter was the lowest (0.05–0.37 kg m⁻²).

Table 2 The dry matter distribution, mean pH(H₂O), exchangeable potassium (K) concentration and radioactive cesium (Cs) fallout and content of each compartment in the pastures

	Dry matter distribution (kg m^−2)				pH(H2O)		Exchangeable K concentration (cmol kg^−1 dry wt)		Total radioactive Cs fallout^† (kBq m^−2)	Radioactive Cs content^† (kBq kg^−1 dry wt)
Pasture number	Litter	Roots	Root mat soil	Subsurface soil	Root mat soil	Subsurface soil	Root mat soil	Subsurface soil		Litter	Roots	Root mat soil	Subsurface soil
5-2	0.05	1.69	2.07	18.95	5.1	4.9	1.39	0.57	32.4	2.12	5.15	4.24	0.46
9-1	0.09	0.72	1.14	19.03	5.1	5.1	0.93	0.24	29.7	9.38	9.66	6.68	0.43
9-2	0.37	0.71	1.92	27.50	6.5	6.2	4.27	2.62	42.2	2.55	6.06	6.85	0.50
13-1	0.14	0.98	1.31	20.79	4.7	4.6	0.79	0.24	26.8	9.58	9.42	4.91	0.25
14-2	0.11	2.29	4.68	25.94	4.9	5.0	0.41	0.09	28.7	11.87	5.74	1.90	0.06
18-2	0.19	0.45	3.33	26.73	4.8	4.9	1.06	0.27	33.2	10.01	8.14	4.47	0.18
22	0.07	5.72	7.16	29.56	4.7	4.7	0.24	0.13	29.4	4.44	4.00	0.54	0.01

^† Radioactivities were corrected for radioactive decay to May 9, 2012.

All values are geometric means.

Radioactive Cs contents tended to be higher in the litter (2.13–11.87 kBq kg⁻¹ dry wt.) and roots (4.00–9.66 kBq kg⁻¹ dry wt.) than in root mat soil (0.54–6.68 kBq kg⁻¹ dry wt.) and subsurface soil (0.01–0.50 kBq kg⁻¹ dry wt.) (Table 2). The distribution of radioactive Cs differed among pastures (Fig. 2). For example, 35.9–43.3% of radioactive Cs was in the subsurface soil from pastures 5–2, 9–1, and 9–2, whereas it was 1.1–7.1% in pastures 14–2 and 22. The relative distribution of radioactive Cs was low in the litter (0.3–5.6%). The soil fractions (i.e., root mat soil and subsurface soil) contained 18.4–84.5% of radioactive Cs per area, in which the root mat soil tended to accumulate more radioactive Cs (17.3–59.3%) than the subsurface soil (1.1–43.3%).

Figure 2 Relative distributions of radioactive cesium (Cs) in litter, plant roots, root mat soil and subsurface soil in the seven sampled pastures. Radioactivities were corrected for radioactive decay to May 9, 2012.

The total average radioactive Cs fallout was 31.3 kBq kg⁻¹, with a range of 26.8–42.2 kBq m⁻² (Table 2). These values are comparable with those predicted from the radioactive Cs distribution map produced with aerial monitoring data (MEXT 2011; JAEA 2013).

The soil pH(H₂O) was 4.6–6.5 and the exchangeable K content was 0.09–4.27 cmol kg⁻¹ dry wt. (Table 2). Both values were high in pasture 9–2 where composted cattle manure was applied. The exchangeable K contents in root mat soils were significantly correlated with those in subsurface soils (r = 0.987, p < 0.001).

Maturity of the target plant species

Leaf mass and plant height increased in all plant species over the sampling period in 2012 (Fig. 3). Ear emergence was observed from May 23 for A. odoratum and from June 6 for D. glomerata and P. arundinacea. Blooming was observed from June 6 for R. obtusifolius and T. repens.

Figure 3 Plant height (solid line and symbols) and ear height (dashed line and open symbols) of the seven herbaceous plants (Dactylis glomerata L., Phalaris arundinacea L., Anthoxanthum odoratum L., Lolium perenne L., Trifolium pratense L., Trifolium repens L. and Rumex obtusifolius L.). Vertical bars represent standard deviation (n = 4 for D. glomerata, n = 3 for P. arundinacea and R. obtusifolius, and n = 2 for A. odoratum and T. repens).

Air radiation dose rate of the pastures

Air radiation dose rate was almost constant throughout the sampling period. Relatively high values were recorded in pastures 13–1, 14–2 and 18–2 (0.158–0.178 μSv h⁻¹); the lowest value was in pasture 5–2 (0.140 μSv h⁻¹).

Radioactive Cs content in herbage

Radioactive Cs content in collected plants varied among the pastures (Fig. 4). The effects of pasture (p < 0.01) and sampling date (p < 0.001) were significant, but Cs contents did not differ significantly among plant taxa (p > 0.10). There were no significant interactions of pasture × sampling date, pasture × plant taxa and sampling date × plant taxa (Table 3).

Figure 4 Geometric mean of radioactive cesium (Cs) content of the aboveground part of the seven herbaceous plants (Dactylis glomerata L., Phalaris arundinacea L., Anthoxanthum odoratum L., Lolium perenne L., Trifolium pratense L., Trifolium repens L. and Rumex obtusifolius L.) in each pasture. The number in each graph is the pasture name. No vertical bars are shown because the geometrical standard deviation is small.

Table 3 Results for the general linear model used to analyze the factors affecting the radioactive cesium (Cs) content of the aboveground part of plants^†

Factors		Degree of freedom	Mean square	F value	Probability
Intercept	Model	1	1351.02	256.67	0.000
	Error	6	5.26
Pasture	Model	6	6.40	20.58	0.006
	Error	4	0.31
Sampling date	Model	3	0.79	11.69	0.000
	Error	24	0.07
Plant taxa	Model	2	0.28	0.89	0.481
	Error	4	0.31
Pasture × sampling date	Model	15	0.07	0.99	0.511
	Error	14	0.07
Pasture × plant taxa	Model	4	0.31	4.61	0.014
	Error	14	0.07
Sampling date × plant taxa	Model	4	0.07	1.04	0.422
	Error	14	0.07

^† Radioactive Cs contents were logarithmically transformed to fit normal distribution before the statistical analysis.

High radioactive Cs content was recorded in pastures 14–2 (2.85–3.98 kBq kg⁻¹) and 22 (4.27 kBq kg⁻¹) on May 9 (Fig. 4). Radioactive Cs contents were also high on May 9 in P. arundinacea in pasture 9–1 (3.51 Bq kg⁻¹). In contrast, radioactive Cs content was very low in plants collected from pasture 9–2 (0.04–0.18 kBq kg⁻¹) throughout the sampling period.

The radioactive Cs content decreased with maturity in all sampled plants (Fig. 4). The tendency was clearly observed in pastures 9–1, 14–2 and 22, where the radioactive Cs content of herbage was high on May 9.

The relative values of the radioactive Cs content significantly decreased with maturity of plants (p < 0.001), but the extent differed among plant taxa (p = 0.05) (Fig. 5). There was no significant interaction between maturity and plant taxa. Relative to the radiation in the grasses and the forb (R. obtusifolius) on May 9, the radiation by May 23 had decreased to 69.7–72.9% and by June 6 to 55.5–58.6%. In contrast, radiation in legumes on June 6 was still 80.6% of that on May 9.

Figure 5 Relative radioactive cesium (Cs) content of the aboveground part of grasses (Dactylis glomerata L., Phalaris arundinacea L., Anthoxanthum odoratum L. and Lolium perenne L.; n = 9), legumes (Trifolium pratense L. and Trifolium repens L.; n = 3) and the forb (Rumex obtusifolius L.; n = 3) with maturity. The values are presented as the percentage of radioactive Cs content on each sampling date relative to that on May 9.

Relationship between TF and soil characteristics

The soil-to-plant TF varied among pastures, but was almost constant among plant species within the same pasture (Table S1). The lowest TF value (0.11) was calculated for pasture 9–2, where composted cattle manure was applied. The highest value (6.23) was calculated for pasture 22, which was used as a cutting meadow. The variation of TF was well explained by both pH(H₂O) (R² = 0.783, p < 0.001) and exchangeable K content (R² = 0.971, p < 0.001) in root mat soil (Fig. 6).

Figure 6 Relationship of transfer factor (TF) to (A) pH(H₂O) and to (B) exchangeable potassium (K) content (cmol kg⁻¹) in root mat soil, for the seven herbaceous plants (Dactylis glomerata L., Phalaris arundinacea L., Anthoxanthum odoratum L., Lolium perenne L., Trifolium pratense L., Trifolium repens L. and Rumex obtusifolius L.). *** represents significance (p < 0.001). TF = radioactive cesium (Cs) content of the aboveground part of the plant (Bq kg⁻¹ dry weight)/radioactive Cs content of the soil (Bq kg⁻¹ dry weight), where all radioactive Cs contents of litter, root mat soil, roots from the root mat and subsurface soil are included in the denominator, and the concentration is expressed as a weighted average of these components.

DISCUSSION

The accident at NPS released a substantial amount of radioactive cesium, thereby contaminating soil over a wide area. The survey conducted by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan, estimated that the contamination levels broadly ranged from less than 10 kBq m⁻² to more than 3000 kBq m⁻² (MEXT 2011). For our study area, they estimated the contamination level to be 10–30 kBq m⁻² for both radionuclides of ¹³⁴Cs and ¹³⁷Cs (JAEA 2013), which is comparable to our measurements (29.4–42.2 kBq m⁻²).

At all locations in our study area, the root mat layer (plant roots and root mat soil) accumulated a large proportion of the deposited radioactive Cs. In the soil fractions, a higher proportion of radioactive Cs was in the root mat soil than in the subsurface soil (Fig. 2). Studies in northeastern Japan after the Fukushima accident showed high radioactive Cs contents in the top layer of untilled arable soils (Koarashi et al. 2012; Yamaguchi et al. 2012; Shiozawa 2013). Many studies conducted after the Chernobyl accident indicated that vertical migration of radioactive Cs was very slow, and most radioactive Cs was present in the upper soil layers (e.g., Almgren and Isaksson 2006). Furthermore, biorecycling of radioactive Cs found in forest ecosystems (Kruyts and Delvaux 2002) might also occur in pasture ecosystems. Radioactive Cs deposited onto a pasture surface may be taken up by roots developed in the root mat layer; then the radioactive Cs absorbed by plants may fall onto the pasture surface layer as dead tissues, unless the aboveground part of the plant is removed by harvesting or grazing.

The radioactive Cs content of the aboveground part of plants differed between pastures, despite the similarity in total radioactive Cs fallout. Many studies have shown that plant uptake of radioactive Cs is higher from soil rich in organic matter (van Bergeijk et al. 1992; Absalom et al. 1996; Sanchez et al. 1999; Grytsyuk et al. 2006); this is because organic matter decreases the affinity of clay for Cs, which increases Cs mobility (Dumat et al. 1997; Sanchez et al. 1999; Staunton et al. 2002). The soil used in our study was an Andosol, which is rich in organic matter (Nanzyo 2002). However, very low radioactive Cs content in herbaceous plants was observed in pasture 9–2, in which composted cattle manure was applied to the surface (Fig. 4). Soil-to-plant TF was negatively related to pH(H₂O) and exchangeable K content in soil in all the pastures (Fig. 6). It is well known that plant uptake of radioactive Cs is suppressed in soil with a high exchangeable K content; in fact, the ability of K to suppress radioactive Cs uptake is larger than the ability of organic matter to promote it (Shaw and Bell 1991; Belli et al. 1995; Sanchez et al. 1999; Rigol et al. 2002). This is because K⁺ can compete with Cs for absorption by plant roots (Robison and Stone 1992; Roca and Vallejo 1995; Zhu and Smolders 2000; Ciuffo et al. 2003). The high soil exchangeable K content in pasture 9–2 indicates that surface application of composted cattle manure greatly reduced plant radioactive Cs uptake by increasing exchangeable K content in the soil. It remains uncertain whether pH(H₂O) is directly related to radioactive Cs uptake by plants.

In our study, radioactive Cs content significantly decreased with maturity of the aboveground part of plants; in addition, the extent of decrease was less in legumes than in the grasses and the forb R. obtusifolius. Paasikallio and Sormunen-Cristian (2002) reported that radioactive Cs concentrations decreased in ryegrass and increased in clover and alfalfa with time during the growth period. They suggested that more ammonium (NH₄⁺) was available in legumes at the later stage of growth due to symbiotic N fixation, and that the NH₄⁺ availability may result in the relatively higher radioactive Cs uptake in legumes than grasses and other forbs. It is well known that NH₄⁺ in soil increases the radioactive Cs content of plants (Evans and Dekker 1969; Shaw and Bell 1991; Belli et al. 1995; Sanchez et al. 1999; Rigol et al. 2002), because NH₄⁺ effectively mobilizes Cs in soils (Jackson et al. 1965; Sanchez et al. 1999). However, it is uncertain whether N fixation in legumes is related to uptake of radioactive Cs. To clarify the temporal change in concentration of radioactive Cs in plants, it may be necessary to clarify the change in concentration of K and other nutrients.

In conclusion, pasture management history and plant maturity were major factors affecting radioactive Cs content in the aboveground part of plants. Surface application of composted cattle manure reduced radioactive Cs uptake by plants by increasing the exchangeable K content of the soil; however, increased exchangeable K content may reduce herbage quality by degrading the mineral balance (Kayser and Isselstein 2005). Plant maturity reduced the Cs content of plants, but late harvesting can decrease the nutritive value and digestibility of herbage. However, both application of composted cattle manure and late harvesting would be effective management strategies to produce forage with radioactive Cs contents that are below the contamination limit for animal feed (MAFF 2013).

Supplementary Material

The supplementary material for this article is available online from: http://dx.doi.org/10.1080/00380768.2014.954269.

ACKNOWLEDGMENTS

We thank the staff and students at the Laboratory of Land Ecology and the Laboratory of Environmental Crop Science, FSC, Tohoku University, for useful advice and support with field sampling and measurements. The present work was in part supported by Tohoku University and JSPS KAKENHI Grant No. 26511003.