Transfer of elements relevant to nuclear fuel cycle from soil to boreal plants and animals in experimental meso- and microcosms

Transfer of elements relevant to nuclear fuel cycle from soil to boreal plants and animals in experimental meso- and microcosms

Science of the Total Environment 539 (2016) 252–261 Contents lists available at ScienceDirect Science of the Total Environment journal homepage: www...

690KB Sizes 6 Downloads 25 Views

Science of the Total Environment 539 (2016) 252–261

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Transfer of elements relevant to nuclear fuel cycle from soil to boreal plants and animals in experimental meso- and microcosms Tiina S. Tuovinen a,⁎, Anne Kasurinen a, Elina Häikiö a, Arja Tervahauta b, Sari Makkonen a, Toini Holopainen a, Jukka Juutilainen a a b

Department of Environmental Science, University of Eastern Finland, P.O. Box 1627, FI-70211 Kuopio, Finland Department of Biology, University of Eastern Finland, P.O. Box FI-70211, Kuopio, Finland

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• We studied transfer of elements in boreal food chain using meso- and microcosms. • Elements related to nuclear fuel cycle and mining were examined. • Higher uptake at lower soil concentrations was observed for primary producers. • Snails took up elements mainly from food but for U also soil was an element source. • Non-linear transfer of essential elements was observed for herbivore and decomposer.

a r t i c l e

i n f o

Article history: Received 8 April 2015 Received in revised form 31 August 2015 Accepted 31 August 2015 Available online 9 September 2015 Editor: D. Barcelo Keywords: Element uptake Food chain Concentration ratio Mesocosm Microcosms Radioecology

⁎ Corresponding author. E-mail address: tiina.tuovinen@uef.fi (T.S. Tuovinen).

http://dx.doi.org/10.1016/j.scitotenv.2015.08.157 0048-9697/© 2015 Published by Elsevier B.V.

a b s t r a c t Uranium (U), cobalt (Co), molybdenum (Mo), nickel (Ni), lead (Pb), thorium (Th) and zinc (Zn) occur naturally in soil but their radioactive isotopes can also be released into the environment during the nuclear fuel cycle. The transfer of these elements was studied in three different trophic levels in experimental mesocosms containing downy birch (Betula pubescens), narrow buckler fern (Dryopteris carthusiana) and Scandinavian small-reed (Calamagrostis purpurea ssp. Phragmitoides) as producers, snails (Arianta arbostorum) as herbivores, and earthworms (Lumbricus terrestris) as decomposers. To determine more precisely whether the element uptake of snails is mainly via their food (birch leaves) or both via soil and food, a separate microcosm experiment was also performed. The element uptake of snails did not generally depend on the presence of soil, indicating that the main uptake route was food, except for U, where soil contact was important for uptake when soil U concentration was high. Transfer of elements from soil to plants was not linear, i.e. it was not correctly described by constant concentration ratios (CR) commonly applied in radioecological modeling. Similar nonlinear transfer was found for the invertebrate animals included in this study: elements other than U were taken up more efficiently when element concentration in soil or food was low. © 2015 Published by Elsevier B.V.

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

1. Introduction Nuclear fuel cycle includes mining, usage of nuclear fuel in reactors, storage of spent nuclear fuel and final disposal of spent nuclear fuel (IAEA, 2002) and radioecological modeling is needed for predicting possible adverse impacts on biota and human health due to radionuclides released to the environment during the nuclear fuel cycle. In radioecological models used for such purposes, transfer of elements into organisms is commonly described by concentration ratios (CRs). In case of soil-to-plant transfer, for instance, CR is plant element concentration divided by soil element concentration. The definition of CR (IAEA, 2010) includes the assumption of linearity, i.e., concentration in organisms is supposed to be linearly related to soil concentration which leads to constant CR value regardless of soil concentration. There is evidence that plants take up many elements more efficiently at low than at high soil concentrations, which is inconsistent with the linearity assumption (Simon and Ibrahim, 1987; Cook et al., 1994; Palm, 1994; McGee et al., 1996; Morton et al., 2002). It has been suggested that transfer is non-linear for essential, but not for nonessential elements (Timperley et al., 1970). However, non-linear transfer has been reported for many metals including essential elements such as calcium, magnesium, manganese (Vera Tome et al., 2003), zinc (Zn) and copper and for non essential elements such as lead (Pb), cadmium (Krauss et al., 2002) and mercury (Han et al., 2006), and for radionuclides of uranium (U), thorium (Th) and radium (Sheppard and Sheppard, 1985; Mortvedt, 1994). Deviation from the linearity assumption was also supported by our recent study on uptake of U, molybdenum (Mo), nickel (Ni), Pb and Zn in five forest plant species (blueberry, fern, may lily, rowan, spruce) growing at two boreal forest sites in Central Finland (Tuovinen et al., 2011). For all the elements and plant species studied, CRs were higher at low than at high soil concentrations (Tuovinen et al., 2011). In spite of these observations of non-linearity, use of linear CRs continues widely in radioecological modeling because they are easy to use and empirically determined values are readily available for many plant species. Different processes are involved in transfer of elements into plants and animals. Transfer of elements from soil to plant occurs mainly via root uptake (Denny, 2002; Kabata-Pendias, 2011), but transfer to animals occurs via food consumption or cutaneous uptake (IAEA, 2010). Little information is available about the linearity of radionuclide transfer from soil or food to animals in terrestrial food chains. Because of mobility of animals, differing dietary compositions and feeding behaviors of animals, and heterogeneous soil element concentrations (IAEA, 2010), it is difficult to define food and soil concentrations that would adequately represent actual animal exposure at natural forest sites. The aims of this study were to determine transfer from soil or food to animals in a terrestrial food chain consisting of boreal forest species, and to confirm our earlier observational results (Tuovinen et al., 2011) on uptake into plants in a controlled experimental system. Assessing the linearity of transfer (validity of the use of constant CRs) was a particular focus of the study. For this purpose, we constructed mesocosms containing three trophic levels: primary producers (birch, fern and grass), herbivorous species (snails), and decomposers (earthworms). In addition, a separate microcosm experiment was conducted to estimate the contribution of food (birch leaves) alone and soil + food together for the total accumulation of U and the other elements (Co, Mo, Ni, Pb, Th, Zn) in snail tissues. Copse snails (Arianta arbostorum) are known to be both herbivorous and detritivorous organisms that can be exposed to contaminated soils by both digestive and cutaneous routes. Snails are also regarded as key components of terrestrial food webs, and they may contribute to the transfer of pollutants from soil and plants to top predators (de Vaufleury et al., 2006). Earthworms are commonly tested in bioaccumulation studies, because they have a high exposure potential to chemicals in soil, and are important part of diet to a variety of wildlife species (DeForest et al., 2011).

253

This study assessed the uptake of seven trace metals that are of interest with respect to naturally occurring radioactive material (U, Th) or spent nuclear fuel and radioactive waste (U, Co, Mo, Ni, Pb, Zn). Radionuclides of U (234U, 235U, 238U) can be released during entire nuclear fuel cycle; 210Pb is a member of the 238U decay chain; 60Co, 93Mo, 59Ni and 63Ni are activation products of nuclear power production (STUK, 2001; Nykyri et al., 2008; Hjerpe et al., 2010), and Zn used to inhibit corrosion in pressurized water reactors produces 64Zn and 65Zn (Betova et al., 2011). Transfer of total element concentrations was used as a model of transfer of radionuclides, based on the assumption that uptake of stable and radioisotopes is equal (IAEA, 2010). 2. Materials and methods 2.1. Experimental soils In early June 2011, 0.5 m3 of forest top soil was collected from a uranium-occurrence located near Nilsiä, in Eastern Finland (63°04'N, 27°54'E; for more details see Roivainen et al., 2011a). Screening of the Nilsiä site in May 2011 (data not given here) showed that soil U concentrations were above background levels near the edge of a 100-m long excavation pit done in the ‘1960s, and therefore this site was selected for U soil collection. Preliminary element analyses performed in June 2011 confirmed screening observation (U level was about 39 mg kg−1; Table 1). Another 0.5 m3 of top soil was collected from a forest site near the Research Garden of the University of Eastern Finland, Kuopio (62°53′N, 27°37′E) in June 2011. In this soil uranium levels were lower (0.3–0.8 mg kg−1, Table 1), and therefore it served as control soil in the mesocosm experiment. Both soil collection sites were herb-rich forests, and at both sites the distinction between organic and mineral soil layers was unclear, so the soil was collected from top to the depth of about 20 cm. Both soil types were cleaned of debris and larger stones and then mixed properly before the initiation of the greenhouse experiment. One soil sample (mixed from nine subsamples) per soil type was also collected for element analysis to determine the soil element concentrations before the experiment, data is given in Table 1. 2.2. Mesocosm experiment in 2011–2012 The transfer of U and other elements from soil to plants and animals was studied in a mesocosm experiment in a greenhouse at the Research Garden of the University of Eastern Finland (Kuopio campus), from 5th of July 2011 to 7th of September 2012. In early July 2011, 25 L of uranium-rich soil (US) or control soil (CS) was placed in 30 L plastic containers (n = 9 per soil type). Two downy birch (Betula pubescens) genotypes (R1 and Ha02) were selected from the tree collections of the Department of Biology, University of Eastern Finland (Kopponen et al., 2001), and micropropagated during spring 2011. Two birch plantlets per genotype, shoot heights ranging from 10 to 15 cm, were planted in each mesocosm. Scandinavian small-reeds (Calamagrostis purpurea ssp. Phragmitoides, two plants per mesocosm) and narrow buckler ferns (Dryopteris carthusiana, one plant per mesocom) were also introduced into the mesocosms. The seeds of C. purpurea spp. phragmitoides were obtained from the Botanical Garden of the University of Oulu, and ferns (including rhizome, roots and above ground parts) were collected from a forest near the Research Garden. During the growing seasons 2011 and 2012, the mesocosms were kept at 21 °C/19 °C (day/night) temperature under 18 light:6 dark photoperiod in a greenhouse. Mesocosms were regularly watered with tap water. Small amounts of irrigation water leached out of mesocosms, and this water was not recycled back to the mesocosms. One week after the establishment of the plant material into the mesocosms (on 13th of July 2011), six adult earthworms (Lumbricus terrestris) were placed in each mesocosm. Furthermore, copse snails (A. arbostorum, six animals per mesocosm) were allowed to feed on plants in three US and three CS mesocosms for 40 days from 30th of

254

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

Table 1 The arithmetic mean (±SE) for element concentrations (mg kg−1) and for soil properties (pH and particle size distribution) in control soil (CS) and uranium-rich soil (US). Element concentrations were determined in spring 2011 (before experiment, n = 1) and in autumn 2011 (n = 2 per soil type) and in autumn 2012 (n = 3 per soil type). Soil properties were analyzed once at the end of the experimental season in 2011 (n = 2 per soil type for pH and n = 1 for other soil properties). P-values b0.05 are shown in bold font.

U Co Mo Ni Pb Th Zn

Before experiment

2011

CS

US

CS

US

CS

US

Soil type

Year

Soil type × year

0.74 5.97 0.38 12.1 5.56 3.38 47.4

38.6 3.53 3.99 11.3 4.93 2.03 24.1

0.55 ± 0.002 4.76 ± 0.11 0.45 ± 0.004 11.0 ± 0.32 5.41 ± 0.08 1.69 ± 0.03 43.3 ± 0.56

48.6 ± 1.63 3.42 ± 0.14 2.48 ± 0.39 10.9 ± 0.62 4.24 ± 0.09 1.41 ± 0.06 27.5 ± 1.00

0.51 ± 0.02 5.09 ± 0.11 0.41 ± 0.002 10.9 ± 0.10 5.91 ± 0.10 2.49 ± 0.10 47.6 ± 1.39

34.0 ± 1.50 3.72 ± 0.10 2.80 ± 0.58 11.3 ± 0.31 4.87 ± 0.17 1.97 ± 0.36 28.5 ± 1.99

b0.001 b0.001 0.001 0.684 b0.001 0.150 b0.001

0.001 0.035 0.735 0.671 0.006 0.032 0.158

0.001 0.932 0.665 0.409 0.640 0.639 0.363

CS

US

P-values from t-test

5.45 ± 0.017 3.29 10 37.8

4.39 ± 0.025 4.68 4.2 15.4

b0.001

pH Humus (%) Clay (%) Silt (%)

2012

August till 10th of October in 2011. During the snail feeding period, the mesocosms were covered with a mesh cloth supported by a frame, which prevented the animals to move outside the mesocosm, but allowed them to move freely inside the mesh-frames. The earthworms were acquired from a commercial supplier (Yorkshire-Worms, UK), while the snails were collected from the surroundings of the Research Garden in Kuopio. At the end of the growing season, on 10th of October 2011, soil, plant leaf and root and animal samples were harvested for chemical element analyses from six CS and six US mesocosms. Before element analyses, plant, soil and earthworm samples from three mesocosms per soil type were pooled, so that the number of replicates is two for soil, plant and earthworm samples for both US and CS mesocosms. Due to high animal mortality, snails from all six CS and six US mesocosms were pooled per soil type, resulting in only one snail sample per soil type. The remaining three CS and three US mesocosms over-wintered in a dark cold room, where temperature was lowered gradually from + 8 °C to + 1 °C, from November 2011 until the beginning of May 2012. After over-wintering period, the remaining mesocosms were moved back to greenhouse conditions, and on 2nd of July 2012, new sets of adult earthworms (10 adult L. terrestris per mesocosm) were added to ensure that there were enough living earthworms for the second sampling in autumn 2012. The plant, soil, and earthworm samples of the second year of mesocosm experiment were harvested on 7th of September 2012. In 2012, soil, birch, fern and earthworm samples from three US and three CS mesocosms were kept separately (n = 3 per soil type), but because of small amount of leaf and root material in small-reeds, grass samples of CS and US mesocosms were pooled so that n = 2 for grass leaf per soil type and n = 1 for grass root data per soil type. 2.3. Microcosm experiment 2012 A separate microcosm experiment was established in summer 2012 in order to study the relative importance of direct soil contact vs. ingestion of food in the transfer of elements into snails. Twelve small rectangular plastic boxes (3.1 L) were used as microcosms, and these microcosms were divided into the following treatments: 1) CS plus birch leaves produced in the CS mesocosms, 2) US plus birch leaves produced in the US mesocosms, 3) only birch leaves from the CS mesocosms and 4) only birch leaves from the US mesocosms; n = 3 per each microcosm treatment. There were six snails per microcosm, and only birch leaves were used as food, because the mesocosm experiment in 2011 showed that snails mainly fed on birch leaves. Leaves from both birch genotypes were used in equal amounts. Before placing CS, US or leaves in the boxes, 2 cm of sand (0.5 to 1.2 mm grain size) was added into all microcosms. In the soil + leaves containing microcosms, a

P-values from MANOVA

1-cm layer of CS or US was placed on top of the sand layer and leaves were placed on the top of the CS or US soil. The microcosms were regularly moistened with tap water. The snails were allowed to feed on birch leaves in each microcosm for two months (from 14th of June till 14th of August in 2012). Birch leaves were added into the microcosms when necessary, but partly consumed leaves were not removed from the microcosms during the experiment. 2.4. Chemical analyses All soil, plant and animal samples were oven-dried at + 70 °C for 4 days before the element analyses. Before oven-drying, all root samples were washed with distilled water, whereas earthworms and snails were allowed to void their gut by starvation for 4 to 5 days and then they were ground in liquid nitrogen. Dried soil samples were sieved to ≤2 mm fraction, and leaf and root samples were milled to fine powder. Soil particle size distribution (n = 1 per soil type) was determined according to the modified method of ICP forest (2006) (see Roivainen et al., 2011a), humus content (n = 1 per soil type) spectrofotometrically with Walkley–Black method (Walkley and Black, 1934), and also soil pH (soil:water v:v 1:5; n = 2 per soil type) was analyzed. Chemical element analyses were carried out in the laboratory of Labtium Ltd. in Kuopio, Finland. The laboratory is accredited according to FINAS T025 (EN ISO IEC 170265). Total concentrations of U, Co, Mo, Ni, Pb, Th, Ti and Zn, were determined by inductively coupled plasmamass spectroscopy (ICP-MS) (Perkin-Elmer Sciex Elan 6000) after nitric acid digestion in microwave oven (US-EPA standard 30,151). ICP-MS analyses were conducted following the standard SFS-EN ISO 17294-2. The detection limits were: 0.01 mg kg−1 for U; 0.02 mg kg−1 for Th; 1.0 mg kg−1 for Co, Mo and Zn; 0.05 mg kg−1 for Pb; 2.0 mg kg−1 for Ni and Ti (DW basis). For element concentrations below the detection limit, half the detection limit was used as a substitute for statistical analyses. Uranium concentrations in CS were below detection limit for seven birch leaf samples, for two small-reed leaf samples and for one snail sample in CS mesocosms. Thorium concentrations were below the detection limit for the majority of plant leaf samples, few birch root samples and for whole snail data. Plant titanium (Ti) concentration N 10 mg kg−1 was used as an indicator of soil contamination (Cary et al., 1986) as living organisms only take up small amount, less than 3 mg kg−1, of Ti (Nisbet and Shaw, 1994; Cook et al., 2007). 2.5. Statistical analyses The difference in pH values between CS and US (Table 1) was tested with independent-samples t-test, and since the requirement for equal variances was not met, t-test for unequal variances was used (the degrees of freedom were adjusted with Welch–Satterthwaite method).

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

255

The genotype effect on birch leaf CR data (U, Co, Mo, Ni, Pb and Zn; Th concentration was below detection limit) was first tested with multivariate ANOVA (MANOVA), in which fixed factors were birch genotype, soil type and experimental year. Since the differences between the two birch genotypes were statistically significant only for one element (Ni), mean element concentrations of the two genotypes were calculated and used in the further statistical testing. The mesocosm data for element concentrations in plants and earthworms were analyzed with MANOVA in which fixed factors were soil type and experimental year. The snail data from mesocosms could not be statistically tested as n = 1. Soil-to-plant CRs were analyzed with MANOVA in which fixed factors were soil type and plant species, and soil-to-earthworm CRs with MANOVA in which soil type and experimental year were fixed factors. Experimental year was included in the statistical models because omitting it might have affected the results concerning effects of soil type. However, effect of study year was not the main research question of interest in this study, and therefore it is not discussed further in the text. The microcosm data was analyzed with MANOVA and for snail element concentrations and food-to-snail CRs the fixed factors were soil type and microcosm type and for soil-to-snail CRs soil type alone was used as fixed factor. Element concentrations and CRs in plant and animal data are presented as geometric means (GMs) and geometric standard deviations (GSDs). GMs are commonly used in radioecology, as CR values and concentrations generally follow a log-normal distribution (Sheppard and Evenden, 1997). Before MANOVA tests, natural logarithm transformation for element concentration and CR data were performed. All statistical analyses were performed using SPSS 19.0 for Windows (SPSS Inc., an IBM Company).

significant soil type × plant species interaction for leaf U, Table 3). For other elements, only plant species or soil type main effects on CRs (Table 3) were observed. Leaf CRs of Co and Ni were higher in US than in CS for all plant species, ferns having the lowest CRs for Co and birches having the lowest CRs for Ni in both soil types. In contrast, CRs of Mo were higher in CS than US for all plants, and of all plant species studied small-reeds had the highest soil-to-leaf CRs for Mo in both soil types. Soil-to-leaf CRs of Pb were somewhat higher in US than CS for all plant species, birch having the lowest CRs in both soil types. Concentration ratios of Zn were also higher in US than CS, but now CRs for birch leaves were approximately five times higher than those for ferns and grass in both soil types. Concerning the soil-to-root CRs, differences between plant species were observed for all elements studied, but difference between the two soil types was statistically significant only for Co, Mo and Zn (Table 3). Because of possible soil contamination of root samples, soil-to-root CRs are less reliable than soil-to-leaf CRs, and therefore soil-to-root CRs are not discussed in detail. Leaf element concentrations showed no clear increase with increasing concentration in soil and, correspondingly, CR decreased with increasing soil concentration. This is illustrated for Mo and Co in Fig. 1, but a similar pattern was observed in all plant species for all elements except for Ni and U. Soil Ni concentrations were similar in US and in CS, making it impossible to study the dependence on soil concentration. The behavior of U differed from that of all other elements studied. Uranium concentrations in plant leaves increased with increasing soil U concentration (Fig. 1a), but the CR for U was still lower at high soil concentrations (Fig. 1b; significant soil type effect in Table 2), suggesting that the linearity assumption (i.e., constant CR) may not hold even for U.

3. Results

3.3. Element concentrations and concentration ratios for earthworms

3.1. Soil characteristics and element concentrations

Uranium concentration in earthworm tissue ranged from 0.22 mg kg−1 in CS to 14 mg kg−1 in US and concentrations of U and Mo were higher in earthworms grown in US than in CS (Table 4). Transfer of U into earthworms seemed to follow the linearity assumption: its concentration in earthworm increased with increasing soil U concentration (Fig. 2a, Table 4), and based on the Fig. 2b CR did not essentially change with soil U concentration, even though there was a statistically significant difference between the CRs of the two soil type (Table 4). For other elements (Co, Pb, Zn), however, concentration in earthworm did not essentially change with increasing soil concentration, and soilto-earthworm CRs were higher at low than at high soil concentrations (statistically significant for Pb, Table 4). Based on MANOVA, Mo was the only element having statistically significantly higher earthworm concentration at higher soil concentration, and still the CRs for Mo were statistically significantly (Table 4) and visibly higher at low soil concentration than at high soil concentration (Fig. 2c-d). The concentration of Th in earthworms was higher in US than in CS (Table 4), although Th concentrations were similar in the two soils (Table 1).

Both experimental soils (US and CS) can be classified as acidic soils, which is characteristic of Finnish forests. Uranium-rich soil was more acidic and based on the particle size distribution US contained less clay and silt than CS (Table 1). The concentration of U was significantly higher in US than in CS (Table 1). Of the other elements, Co, Pb and Zn concentrations were lower in US than in CS, whereas Mo concentrations were higher in US when compared to CS (Table 1). 3.2. Element concentrations and concentration ratios in plants The uranium, Co and Ni concentrations of birch, fern and small-reed leaves were higher in US than in CS, whereas Mo concentration of birch leaves was higher in CS than in US (Table 2). Lead and Zn concentrations in plant leaves did not differ between the two soil types. Uranium concentrations of plant roots were higher in US than in CS. In birch root, Co concentration was higher in US than in CS, whereas Pb concentration was higher in CS than in US. Plant Ti concentration was used as soil contamination indicator for plant parts. Titanium concentrations of nearly all the root samples exceeded the threshold N 10 mg kg−1, and the root Ti concentration as a percentage of soil Ti concentration varied from 1.5% to 81.6%. Therefore, the majority of root samples was considered to be contaminated by soil and is not emphasized in the discussion. Minor soil contamination (0.9%–12.6%) was noticed in a few fern and small-reed leaf samples, but not in birch leaves. The soil-to-leaf CRs differed statistically significantly between plant species and soil types for all elements except for Th (Table 3). Concentration ratios of U were higher in CS than in US for all plant leaves, and this difference between CS and US was most obvious for birch. Concentration ratios of U for birch leaves were lower than CRs for leaves of other studied plant species in both soil types. The plant species showing highest CR for U was fern in CS and small-reed in US (statistically

3.4. Element concentrations and concentration ratios for snails In the microcosm experiment, U concentrations were circa 7-fold higher in snails kept in microcosm with both uranium-rich soil and birch leaves grown in US than in those containing only birch leaves grown in US (Table 5), indicating that direct soil contact was the main source of U for snails. In addition to direct soil contact, food (birch leaves) was another source of U for snails, as snail U concentration was four times higher in snails fed with leaves from US (higher U concentration in food but no contact with soil) than in those fed with leaves from CS (lower U concentration in food) in Table 5. For uranium, food-to-snail CRs were higher in US than in CS microcosms but only when there was also uranium-rich soil present in the microcosms (significant soil type × microcosm type interaction, Table 5), and as pointed out above, the main source for snail U in

256

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

Table 2 Element concentrations (mg kg−1) in leaves and roots of birch, fern and small-reed in control (CS) and in uranium-rich soil (US). Values are geometric means (geometric standard deviation = GSD is given in brackets). Data for 2011 (n = 2 for all plant samples per soil type) and 2012 (n = 2–3 for leaf samples, n = 1–3 for root samples) are shown. P-values b0.05 are shown in bold font. Detection limit for U is 0.01 mg kg−1 and for Th 0.02 mg kg−1; concentrations below detection limit are marked as b.d. 2011

2012

P-values from MANOVA

CS

US

CS

US

Soil type

Year

Soil type × year

Birch leaf U Co Mo Ni Pb Th Zn

b.d. 0.98(1.14) 1.25(1.26) 2.50(1.45) 0.09(1.20) b.d. 339(1.07)

0.05(1.12) 3.11(1.06) 0.55(1.57) 7.32(1.36) 0.12(1.49) b.d. 362(1.11)

0.01(2.09) 0.89(1.13) 0.79(1.21) 1.86(1.22) 0.15(1.33) b.d. 204(1.12)

0.13(1.26) 2.24(1.23) 0.50(1.56) 7.48(1.29) 0.20(1.29) b.d. 214(1.27)

b0.001 b0.001 b0.023 b0.001 0.057 – 0.447

0.012 0.054 0.168 0.168 0.005 – 0.001

0.542 0.211 0.435 0.082 0.849 – 0.971

Fern leaf U Co Mo Ni Pb Th Zn

0.02(2.22) 0.71(1.12) 1.61(1.06) 5.84(1.19) 0.53(1.25) b.d. 36.2(1.05)

0.73(1.53) 1.54(1.34) 1.39(1.05) 11.62(1.33) 0.62(2.41) b.d. 38.7(1.03)

0.02(1.06) 0.31(1.25) 0.90(1.17) 1.60(1.65) 0.52(4.10) b.d. 37.9(1.44)

0.39(1.13) 0.70(1.18) 1.14(1.17) 6.49(1.24) 0.44(1.91) b.d. 59.0(1.23)

b0.001 0.001 0.601 0.003 0.995 – 0.159

0.195 0.001 0.004 0.005 0.776 – 0.191

0.290 0.886 0.067 0.159 0.804 – 0.276

Small-reed leaf U Co Mo Ni Pb Th Zn

0.04(2.00) 2.27(2.86) 2.52(1.03) 17.26(1.65) 0.47(1.91) 0.04(6.59) 41.8(1.04)

5.79(1.85) 9.25(1.63) 4.06(1.11) 55.89(1.13) 0.85(1.63) 0.21(2.49) 59.4(1.09)

b.d. 0.52(1.02) 1.58(1.30) 2.86(1.43) 0.18(1.01) b.d. 37.6(1.12)

0.29(2.40) 0.97(1.30) 1.43(1.12) 6.33(1.31) 0.55(2.81) b.d. 59.1(1.56)

b0.001 0.040 0.107 0.007 0.152 0.238 0.094

0.003 0.004 0.001 b0.001 0.232 – 0.786

0.362 0.338 0.033 0.432 0.626 – 0.806

Birch root U Co Mo Ni Pb Th Zn

0.09(2.14) 3.84(2.14) 1.28(1.99) 21.84(2.17) 1.12(1.81) 0.10(7.63) 48.7(1.03)

2.90(1.84) 13.40(3.93) 0.97(2.13) 18.52(3.16) 0.40(1.52) 0.04(3.37) 53.4(1.15)

0.22(2.33) 13.31(3.50) 0.74(1.56) 15.91(2.16) 1.96(1.83) 0.54(1.50) 43.9(1.12)

23.20(1.76) 14.48(2.56) 1.04(1.29) 21.34(1.66) 1.83(1.45) 0.55(1.38) 47.7(1.13)

b0.001 0.027 0.711 0.528 0.011 0.486 0.122

0.002 0.047 0.110 0.539 0.001 0.012 0.070

0.120 0.022 0.260 0.982 0.033 0.467 0.861

Fern root U Co Mo Ni Pb Th Zn

0.78(1.09) 39.85(1.30) 11.10(1.14) 257.83(1.23) 6.94(1.01) 1.95(1.29) 37.2(1.02)

18.53(1.83) 138.98(1.21) 6.72(1.68) 154.36(1.99) 5.10(2.45) 1.18(1.71) 37.6(1.07)

0.38(1.48) 36.02(1.73) 1.34(1.11) 55.61(1.90) 3.36(1.60) 0.91(2.81) 43.4(1.09)

47.49(1.41) 36.43(1.54) 2.09(1.18) 43.12(1.47) 4.16(1.42) 0.77(1.84) 44.5(1.23)

b0.001 0.061 0.862 0.298 0.889 0.510 0.853

0.656 0.039 b0.001 0.006 0.197 0.257 0.115

0.017 0.064 0.025 0.714 0.448 0.738 0.929

Small-reed root U Co Mo Ni Pb Th Zn

0.12(1.01) 3.99(1.10) 2.57(1.38) 29.19(1.89) 1.88(2.08) 0.19(1.71) 79.3(1.02)

7.22(1.77) 14.72(1.33) 4.52(1.17) 55.53(1.74) 0.94(1.92) 0.29(4.09) 99.3(1.19)

0.29 18.10 2.60 55.00 1.90 1.13 88.1

13.65 18.81 3.47 29.42 1.01 0.15 91.4

0.008 0.068 0.195 0.988 0.385 0.469 0.344

0.169 0.041 0.626 0.999 0.954 0.608 0.926

0.794 0.075 0.599 0.345 0.965 0.316 0.472

these containers was soil not the food. Thus, the plot of food-to-snail CR for U as a function of food U concentration shows that CR was higher at low than at high food U concentration (Fig. 3), even though the difference of food-to-snail CRs between soil types was not statistically significant (Table 5). This may indicate that, although snail U concentration somewhat increases with food U concentration, the linearity assumption (constant CR) may not be fully valid. However, soil-to-snail CR for U was similar in both soil types. For elements other than U, the difference in snail concentrations between the two soil types was small (statistically significant for Co and Ni) or not found (Table 5). The food-to-snail CRs differed statistically significantly between soil types (as well as between food element concentrations) for Co and Ni and the soil-to-snail CRs for Co, Mo and Zn (Table 5). When food-to-snail CR was plotted against food element concentration as in Fig. 3, lower CRs were always observed in snails

receiving higher concentration in food, as shown for Co. Differences in food Pb concentrations were small (not statistically significant), but the CR pattern for Pb was similar to that observed for the rest of the elements, with lower CR in snails given more Pb in food. For uranium, Co, Ni and Pb, soil-to-snail CRs were lower than food-to-snail CRs, but for Zn soil-to-snail CRs were higher than foodto-snail CRs (Table 5). Snail element concentrations and soil-to-snail CRs were similar in mesocosm and microcosm (Table 5), indicating that transfer of elements in the microcosms is similar to that occurring in the more realistic conditions in the mesocosms. Soil-to-snail CRs for Co, Mo and Zn were higher in lower soil concentrations (Table 5), which was consistent with the pattern observed in food-to-snail CR data. For all elements studied, soil-to-snail CRs were lower than soilto-earthworm CRs (Tables 4 and 5), because of higher element concentrations in earthworm tissue compared to snail tissue.

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

257

Table 3 Soil-to-plant concentration ratios (CRs) of seven elements for birch, fern and small-reed leaves and roots in control (CS) and in uranium-rich soil (US). Values are geometric means (geometric standard deviation = GSD is given inside brackets). Both 2011 and 2012 data were included (for birch and fern n = 5 per soil type, for small-reed n = 3–5 per soil type). CRs of Th for birch and fern leaf are not presented because concentrations were below the detection limit (0.02 mg kg−1). P-values b 0.05 are shown in bold font. CS

US

Birch

Fern

Small-reed

Birch

Leaf U Co Mo Ni Pb Th Zn

0.01 (5.05) 0.19 (1.75) 2.25 (1.38) 0.19 (1.87) 0.02 (3.58) – 5.46 (1.75)

0.04 (1.49) 0.09 (1.67) 2.70 (1.34) 0.25 (2.19) 0.09 (2.74) – 0.81 (1.26)

0.03 (3.42) 0.22 (2.93) 4.68 (1.30) 0.64 (2.98) 0.05 (2.01) 0.01 (4.34) 0.89 (1.12)

0.002 (2.15) 0.71 (1.32) 0.20 (1.90) 0.67 (1.30) 0.04 (1.42) – 9.44 (1.44)

Root U Co Mo Ni Pb Th Zn

0.30 (2.44) 1.64 (3.30) 2.19 (1.75) 1.65 (2.09) 0.27 (1.87) 0.09 (4.45) 1.00 (1.14)

0.96 (1.61) 7.57 (1.50) 7.42 (3.04) 9.38 (2.60) 0.79 (1.73) 0.58 (2.74) 0.89 (1.07)

0.30 (1.67) 1.37 (2.34) 5.97 (1.26) 3.29 (1.76) 0.34 (1.71) 0.18 (2.39) 1.85 (1.04)

0.26 (3.97) 3.90 (2.87) 0.39 (1.75) 1.81 (2.16) 0.22 (2.23) 0.08 (5.16) 1.78 (1.18)

P-values from MANOVA Fern

Small-reed

Soil type

0.01 (1.37) 0.27 (1.66) 0.48 (1.49) 0.74 (1.49) 0.11 (2.00) – 1.78 (1.35)

0.02 (4.91) 0.67 (3.76) 0.84 (2.06) 1.36 (3.45) 0.14 (2.31) 0.02 (6.61) 2.12 (1.48)

0.002 b0.001 b0.001 b0.001 0.012 0.158 b0.001

b0.001 0.002 b0.001 0.004 b0.001 – b0.001

0.046 0.854 0.096 0.645 0.331 – 0.398

0.83 (2.29) 17.30 (2.30) 1.29 (2.05) 6.45 (2.27) 0.98 (1.72) 0.54 (1.70) 1.48 (1.22)

0.21 (1.98) 4.54 (1.20) 1.62 (1.27) 4.08 (1.67) 0.22 (1.56) 0.15 (3.29) 3.47 (1.19)

0.567 0.012 b0.001 0.939 0.531 0.749 b0.001

0.010 0.001 b0.001 b0.001 b0.001 0.004 b0.001

0.969 0.922 0.768 0.680 0.553 0.989 0.718

4. Discussion 4.1. Element uptake from soil and plant element concentrations As the results for “mobile” (measured with an extraction method that provides an estimate of the fraction available to plants) and total concentrations in soil were essentially similar in our previous studies (Roivainen et al., 2011a, 2011b, 2012; Tuovinen et al., 2011), we measured only soil total concentration in this study. Apart from concentration in soil, several factors including solid/liquid partition coefficient (Kd), pH (Echevarria et al., 2001) fertilization (Ryfyikiri et al., 2006), speciation of radionuclides (Salbu et al., 2004), interactions of elements, soil organic matter and clay contents (Roivainen et al., 2011a, 2011b, 2012) and rhizosphere processes (Ehlken and Kirchner, 2002), may impact the availability of elements in reality. However, simplifications are necessary in radioecological modeling, and CRs based on total concentration are generally used. The measured Co, Mo, Ni, Pb, Th and Zn concentrations in both soil types were within the range of natural background element concentrations of top soil in Europe (Salminen, 2005). In the uranium-rich soil selected for this study, the U concentration was above the range of natural background U concentration, but within in the control soil (Salminen, 2005). Low soil pH has been related to high plant uptake of U and it is assumed to be highest at pH 5 (Ebbs et al., 1998). In our study, pH values in the two soil types were low and in the range of high uptake of U by plants (Ebbs et al., 1998). In general, plant element concentrations in experimental mesocosms were within the range observed in forest plants growing in their natural habitat (Roivainen et al., 2011a, 2011b; Tuovinen et al., 2011). Concentrations of U, Co, Mo, Ni, Pb and Th were as high in plants and as they were in soil, whereas Zn clearly accumulated into plants, showing higher concentration in plants than in soil. High Zn concentrations in plants are explained by the fact that Zn is easily taken up by plants, as it is essential for plants and very mobile in soils (Kabata-Pendias, 2011). Especially in soil with low pH (around 5), exchangeable form of Zn is dominant and plant uptake of Zn is primarily related to exchangeable Zn (Sims, 1986). Soil pH has been found to be the main factor affecting uptake of Zn by plants, while the other soil properties, such as particle size fraction, cation exchange capacity and organic matter content, appear to be less important (Sims, 1986). Element partitioning within the plant is typically described as rootto-shoot ratio. In the present study, concentrations of U, Co, Mo, Ni, Pb and Th were higher in roots compared to leaves, whereas Zn concentrations were higher in leaves than roots. Our results were in agreement

Plant species

Soil type × plant species

with the study conducted in boreal forest, where concentrations of U (Roivainen et al., 2011a), Co, Mo, Ni and Pb (Roivainen et al., 2011b) were higher in roots than in leaves. Kabata-Pendias (2011) categorized Mo, Ni and Zn as mobile in the plant and Co and Pb as bound in roots. In the present study, Mo and Ni were bound to roots as well. Also soil contamination (despite of careful washing of the roots) may have elevated root concentrations in the present study, as many root samples showed Ti concentrations exceeding the limit value of 10 mg kg−1 indicating contamination (Cary et al., 1986). Roivainen et al. (2011a) did not find significant differences in CRs for U between fern (D. carthusiana), May lily (Maianthemum bifolium), rowan (Sorbus aucuparia) and Norway spruce (Picea abies) grown in natural boreal forest. In the present study, CRs for U and the other studied elements differed between the plant species studied. The plant species studied by Roivainen et al. (2011a, 2011b) were partly different from those used in this study. Perhaps more importantly, uncontrolled variation in natural growth conditions may have made it more difficult to detect small interspecies differences compared to the controlled conditions of the present study. Some interspecies differences were reported for other elements (Co, Mo, Ni, Pb) by Roivainen et al. (2011b), but the differences were relatively small compared to intra-species variation, and use of values pooled from several species was recommended as generic CRs for plant categories such as understory plants and trees. The results of the present study support the view that, apart from plant species, other factors may strongly influence uptake into plants: soil type had in some cases more marked influence on CR than plant species (most obvious for Mo, see Table 3). The observed soil-to-plant CRs in this study were generally within the range reported for boreal plant species (Sheppard and Sheppard, 1991; Reinman et al., 2001; Avila, 2006; Roivainen et al., 2011a, 2011b), but variations of observed CRs are typically large (Ehlken and Kirchner, 2002; Higley and Bytwerk, 2007; IAEA, International Atomic Energy Agency, 2010). However, as the present and our earlier studies (Tuovinen et al., 2011 and Tuovinen et al., 2013) suggest that linear uptake (constant CR) may not be a valid approach for radioecological modeling, CR values are not further discussed in relation to values reported elsewhere. 4.2. Uptake of elements to snails and earthworms The concentration of U in snails was seven times higher when the U source was soil + leaves in comparison to snails that received U only via birch leaves. This shows the importance of direct soil contact for U transfer into snails. Element uptake into snails via soil ingestion and

258

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

Fig. 1. Birch leaf element concentrations (a, c, e) and CRs (b, d, f) plotted against soil element concentration for a non-essential element (U) and for essential elements (Mo, Co). Both year 2011 and year 2012 data are included for two different birch genotypes (R1 and Ha02) grown in control soil (CS) or in uranium-rich soil (US). Geometric means are shown separately for CS and US (n = 10 per soil type).

direct soil contact has been reported also earlier (Gomot et al., 1989). In contrast to U, concentrations of Co, Mo, Ni, Pb and Zn in snail tissues did not depend on the presence of soil in our microcosm experiment. This

difference between U and the other elements is most likely due to low uptake of U into birch, as U concentration was much higher in soil than in birch leaves. Besides food-to-snail CRs, soil-to-snail CRs were

Table 4 Earthworm element concentrations and soil-to-earthworm concentration ratios (CRs) in control (CS) and in uranium-rich soil (US). Values are geometric means (geometric standard deviation = GSD is given in brackets). Earthworm concentration data shown for 2011 (n = 2 per soil type) and 2012 (n = 3 per soil type). P-values b0.05 are shown in bold font. Concentration ratios were calculated using the combined data of both years (n = 5 per soil type), but study year was included as a factor in the MANOVA. Element concentrations in earthworm mg kg−1 2011

U Co Mo Ni Pb Th Zn

2012

CRs to earthworm P-values from MANOVA

2011 + 2012

P-values from MANOVA

CS

US

CS

US

Soil type

Year

Soil type × year

CS

US

Soil type

Year

Soil type × year

0.22 (1.09) 8.93 (1.07) 0.80 (1.10) 1.44 (1.39) 0.79 (1.29) 0.03 (3.96) 415 (1.16)

9.46 (1.06) 8.29 (1.21) 1.35 (1.14) 3.53 (1.06) 1.07 (1.03) 0.28 (2.43) 358 (1.02)

0.29 (1.23) 9.84 (1.08) 0.88 (1.06) 5.05 (2.01) 1.71 (1.19) 0.47 (1.36) 454 (1.07)

14.0 (1.12) 8.83 (1.26) 1.35 (1.11) 4.99 (1.18) 1.85 (1.13) 0.58 (1.25) 336 (1.31)

b0.001 0.417 b0.001 0.167 0.113 0.035 0.092

0.011 0.476 0.467 0.029 0.001 0.008 0.910

0.623 0.874 0.473 0.157 0.316 0.064 0.516

0.49 (1.26) 1.91 (1.05) 2.02 (1.13) 0.28 (2.41) 0.22 (1.51) 0.07 (4.43) 9.55 (1.08)

0.31 (1.54) 2.39 (1.20) 0.52 (1.35) 0.39 (1.24) 0.32 (1.28) 0.26 (1.71) 12.32 (1.32)

0.001 0.057 b0.001 0.191 0.009 0.018 0.130

0.001 0.963 0.770 0.036 0.002 0.021 0.755

0.080 0.791 0.401 0.150 0.251 0.072 0.767

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

259

Fig. 2. Earthworm element concentrations (a, c) and CRs (b, d) plotted against soil element concentration for a non-essential (U) and an essential element (Mo). Both year 2011 and year 2012 earthworm data in control soil (CS) and in uranium-rich soil (US) are included. Geometric means are shown separately for CS and US (n = 5 per soil type).

calculated, as soil-to-animal CRs are generally used in common radioecological modeling software, like in ERICA Tool (Brown et al., 2008). Transfer of elements into earthworms occurs both via ingestion of food and soil and via direct skin contact with soil. This double exposure

is obviously more important for animals living in soil than for animals living with less intensive contact with soil, leading to higher element concentrations in decomposers (Heikens et al., 2001). Consistently with this, all element concentrations were higher in earthworm tissues compared to snails.

Table 5 Snail element concentrations and food-to-snail and soil-to-snail concentration ratios (CRs) for control soil (CS) and in uranium-rich soil (US) in two different microcosm types and in mesocosms. Values are geometric means (geometric standard deviation = GSD is given in brackets). P-values b0.05 are shown in bold font. The number of samples analyzed was 3 for microcosms and 1 for mesocosms. The concentration of Th in snail tissue was below detection limit (0.02 mg kg−1). Food-to-snail CRs in mesocosms were calculated using concentrations in birch leaves, as these were found to be the main food for snails. Microcosm with soil + leaves

Microcosm with leaves

P-values from MANOVA

CS

CS

US

Soil type

Micro-cosm type

Soil type × micro-cosm type

CS

US

Element concentration mg kg U 0.01 (1.15) 0.57 (1.32) Co 1.36 (1.03) 1.78 (1.01) Mo 0.41 (1.12) 0.42 (1.04) Ni 2.23 (1.08) 3.23 (1.37) Pb 0.48 (1.43) 0.30 (1.64) Zn 152 (1.18) 202 (1.26)

0.02 (1.20) 1.37 (1.09) 0.56 (1.53) 2.04 (1.10) 0.29 (1.10) 155 (1.20)

0.08 (2.26) 1.58 (1.09) 0.51 (1.32) 2.32 (1.06) 0.22 (1.73) 190 (1.30)

b0.001 b0.001 0.808 0.035 0.154 0.081

0.020 0.172 0.117 0.065 0.133 0.871

0.002 0.127 0.754 0.255 0.677 0.728

0.005 2.06 0.76 2.85 0.27 169

0.35 2.47 0.65 3.29 0.13 252

Food-to-snail CRs U 1.13 (1.90) Co 1.51 (1.06) Mo 0.52 (1.12) Ni 1.19 (1.06) Pb 3.04 (1.65) Zn 0.74 (1.28)

5.65 (1.86) 0.85 (1.24) 0.80 (1.68) 0.55 (1.42) 1.64 (2.11) 0.98 (1.45)

1.78 (1.79) 1.53 (1.05) 0.71 (1.29) 1.08 (1.24) 1.84 (1.34) 0.76 (1.36)

0.81 (2.76) 0.76 (1.22) 1.00 (2.13) 0.39 (1.65) 1.22 (2.16) 0.92 (1.43)

0.365 b0.001 0.199 0.001 0.182 0.246

0.118 0.548 0.368 0.294 0.291 0.915

0.023 0.498 0.864 0.539 0.778 0.819

0.98 1.92 0.58 0.92 2.98 0.48

7.32 0.83 1.26 0.43 1.23 0.74

Soil-to-snail CRs U 0.02 (1.22) Co 0.27 (1.02) Mo 1.01 (1.11) Ni 0.21 (1.09) Pb 0.08 (1.40) Zn 3.19 (1.20)

0.02 (1.30) 0.48 (1.05) 0.15 (1.36) 0.29 (1.43) 0.06 (1.62) 7.14 (1.12)

0.009 0.42 1.69 0.25 0.05 3.86

0.007 0.69 0.23 0.29 0.03 8.85

US

Mesocosm

−1

0.394 0.009 b0.001 0.130 0.318 b0.001

260

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

Fig. 3. Snail element concentrations (a, c) and CRs (b, d) plotted against food (birch leaf) element concentration for a non-essential (U) and an essential element (Co). The snail data are from microcosms with birch leaves only grown either in control soil (CS) or in uranium-rich soil (US). Geometric means are shown separately for CS and US (n = 3 per soil type).

4.3. Linearity of element transfer to boreal plants, herbivores and decomposers In this experimental study, plant element concentrations did not increase linearly with increasing soil concentration and CRs were higher at low than at high soil concentrations. This finding is consistent with the nonlinear transfer characteristics previously observed in boreal forest ecosystem (Tuovinen et al., 2011). Although the results concerning uptake into plants confirm previous findings from studies using different study designs (Tuovinen et al., 2011; McGee et al., 1996), more novel and unexpected data were obtained concerning transfer into animals. Our previous study in aqueous ecosystem (Tuovinen et al., 2013) showed that transfer of 137Cs from water to non-piscivorous fish species was non-linear, and followed a similar non-linear function as the one that was used to describe uptake from soil to plants (Tuovinen et al., 2011). However, transfer from nonpiscivorous fish species to piscivorous species was found to be linear (Tuovinen et al., 2013). Based on these findings, we assumed that only the first step in the uptake into food chain (from water to phytoplankton or aquatic plants, or from soil water into plants) would be nonlinear, and transfer via ingestion would be linear. The present findings on snails however did not support that assumption. For most elements studied, transfer into snails and earthworms was non-linear, and higher CRs were observed when concentration in food (birch leaves) or in soil was low. Decrease in soil-to-earthworm (Lumbricus rubellus and Eisenia andrei) CRs with increasing soil concentrations was also reported for element concentrations of As, Cd, Cu, Pb and Zn (Vermeulen et al., 2009) and for radionuclides of Cs, Sr and Zn (Keum et al., 2013). The behavior of U differed from that of the other elements studied, as its concentration was higher in snails given food with high U concentration. However, the transfer did not appear to be fully linear even for U, because higher CR values were observed at low food or soil U concentrations. The differences in the uptake of U and the other elements might be related to the fact that Co, Mo, and Zn are essential elements, and it is reasonable to assume that their concentrations in animal tissues are relatively tightly regulated (Windish, 2002),

resulting in approximately constant tissue concentration independent of concentration in food. Biological regulation is commonly observed for essential elements, but it may occur also for some non-essential elements, as suggested by the findings of Neuhauser et al. (1995) for Pb concentration in earthworm tissues. The uptake of non-essential elements may be influenced by chemical similarities and interactions with essential elements and this might be relevant for Pb because of its metabolic similarities with Ca (Pereza et al., 1998), which is an essential element for plants. Our results for Pb also seemed to follow a pattern similar to that observed for the essential elements, with no food or soil dependent differences in Pb concentration and lower (not statistically significant for snails) Pb CRs in snails or earthworms receiving higher concentration of Pb in food or in soil. Overall, the results from the present study and the previous study on aquatic systems (Tuovinen et al., 2013) suggest that uptake of essential elements from food to animals is highly non-linear (possibly because of biological regulation aiming at physiologically optimal levels), but the uptake of non-essential elements (U, Cs) is linear or nearly linear. However, some non-essential elements (such as Pb) may behave like essential elements because of chemical similarities with essential elements.

5. Conclusion Transfer of elements between soil and three different trophic levels, (primary producers, herbivores and decomposers) was studied experimentally from the perspective of radioecological modeling. For most of the elements studied, the results confirmed earlier observational findings suggesting that the linearity assumption generally used in radioecological modeling (constant concentration ratio) is not valid for transfer from soil to plants. The study produced novel data indicating that also transfer into animals is non-linear for many elements relevant to the nuclear fuel cycle. Similarly with uptake into plants, transfer into animals was more efficient (concentration ratio was higher) when concentration in soil or food was low. In contrast to the other elements

T.S. Tuovinen et al. / Science of the Total Environment 539 (2016) 252–261

studied, the uptake of U was nearly linear indicating that different modeling approaches may be required for different elements. Acknowledgements This study was funded by a grant from the Finnish Research Program on Nuclear Waste Management (KYT-2014 program). We thank Toivo Kuronen and Marjatta Puurunen for help in establishing and maintenance of the experiments at the Research Garden of UEF in Kuopio campus. References Avila, R., 2006. Model of the long-term transfer of radionuclides in forests. Technical Report TR-06-08. Swedish nuclear fuel and waste management Co., Stockholm. Betova, I., Bojinov, M., Kinnunen, P., Saario, T., 2011. Zn injection in pressurized water reactors — laboratory tests, field experience and modeling. Research Report No. VTT-R05511-11. Technical Research Center of Finland, Finland. Brown, J.E., Alfonso, B., Avila, R., Baresford, N.A., Copplestone, D., Pröhl, D., Ulanovsky, A., 2008. The ERICA tool. J. Environ. Radioact. 99 (9), 1371–1383. Cary, E.E., Grunes, D.L., Bohman, V.R., Sanchirico, C.A., 1986. Titanium determination for correction of plant sample contamination by soil. Agron. J. 78, 933–936. Cook, L.L., Inouye, R.S., McGonigle, T.P., White, G.J., 2007. The distribution of stable cesium in soils and plants of the eastern Snake River Plain in southern Idaho. J. Arid Environ. 69, 40–64. Cook, C.M., Sgardelis, S.P., Pantis, J.D., Lanaras, T., 1994. Concentrations of Pb, Zn and Cu in Taraxacum ssp. in relation to urban pollution. Bull. Environ. Contam. Toxicol. 53, 204–210. de Vaufleury, A., Coeurdassier, M., Pandard, P., Scheifler, R., Lovy, C., Crini, N., Badot, P.-M., 2006. How terrestrial snails can be used in risk assessment of soils. Environ. Toxicol. Chem. 25, 797–806. DeForest, D.K., Schlekat, C.E., Brix, K.V., Fairbrother, A., 2011. Secondary poisoning risk assessment of terrestrial birds and mammals exposed to nickel. Integr. Environ. Assess. Manag. 8, 107–119. Denny, H., 2002. Plant mineral nutrition. In: Ridge, I. (Ed.), Plants. Oxford University Press, New York., pp. 167–220. Ebbs, S., Brady, D.J., Kochian, L.V., 1998. Role of uranium speciation in the uptake and translocation of uranium by plants. J. Exp. Bot. 49, 1183–1190. Echevarria, G., Sheppard, M.I., Morel, J., 2001. Effects of pH on the sorption of U in soils. J. Environ. Radioact. 53, 257–264. Ehlken, S., Kirchner, G., 2002. Environmental processes affecting plant root uptake of radioactive trace elements and variability of transfer factor data: a review. J. Environ. Radioact. 58, 97–112. Geochemical Atlas of Europe. In: Salminen, R. (Ed.), Part 1: Background Information, Methodology and Maps/FOREGS. Geological Survey of Finland, Espoo. Gomot, A., Gomot, L., Boukraa, S., Bruckert, S., 1989. Infuence of soil on the growth of the land snail Helix aspersa. An experimental study of the absorption route for the stimulating factors. J. Molluscan Stud. 55, 1–7. Han, F.X., Su, Y., Monts, D.L., Waggoner, C.A., Plodinec, M.J., 2006. Binding, distribution and plant uptake of mercury in soil from Oak Ridge, Tennessee, USA. Sci. Total Environ. 368, 753–768. Heikens, A., Peijnenburg, W.J.G.M., Hendriks, A.J., 2001. Bioaccumulation of heavy metals in terrestrial invertebrates. Environ. Pollut. 113 (3), 385–393. Higley, K.A., Bytwerk, D.P., 2007. Generic approach to transfer. J. Environ. Radioact. 98, 4–23. Hjerpe, T., Ikonen, A.T.K., Broed, R., 2010. Biosphere Assessment Report 2009. Report 2010-3. Posiva Oy, Olkiluoto. IAEA, International Atomic Energy Agency, 2002. Heavy water reactors: status projected development. Technical Reports Series No. 407. International Atomic Energy Agency (STI/DOC/010/407 (ISBN: 9201115024). PHWR data is taken from Chapter 6, HWR fuel cycles). IAEA, International Atomic Energy Agency, 2010. Handbook of parameter values for the prediction of radionuclide transfer in terrestrial and fresh water environments. Technical reports series no. 472International Atomic Energy Agency, Vienna. Kabata-Pendias, A., 2011. Trace Elements in Soils and Plants. Fourth edition. CRC Press, Boca Raton FL (520 pp.). Keum, D.K., Jun, I., Lim, K.M., Choi, Y.H., Howard, B.J., 2013. Time-dependent transfer of 137 Cs, 85Sr and 65Zn to earthworms in highly contaminated soils. J. Environ. Radioact. 126, 427–433. Kopponen, P., Utriainen, M., Lukkari, K., Suntioinen, S., Kärenlampi, L., Kärenlampi, S., 2001. Clonal differences in copper and zinc tolerance of birch in metalsupplemented soils. Environ. Pollut. 112, 89–97.

261

Krauss, M., Wilcke, V., Kobza, J., Zech, W., 2002. Predicting heavy metal transfer from soil to plant: potential use of Freundlich-type function. J. Plant Nutr. Soil Sci. 165, 3–8. McGee, E.J., Johanson, K.J., Keatinge, M.J., Synnott, H.J., Colgan, P.A., 1996. An evaluation of ratio systems in radioecological studies. Health Phys. 70, 215–221. Morton, L.S., Evans, C.V., Estes, G.O., 2002. Natural uranium and thorium distributions in podzolized soils and native blueberry. J. Environ. Qual. 31, 155–162. Mortvedt, J.J., 1994. Plant and soil relationships of uranium and thorium decay series radionuclides — a review. J. Environ. Qual. 23, 643–650. Neuhauser, E.F., Cukic, Z.V., Malecki, M.R., Loeher, L.C., Durkin, P.R., 1995. Bioconcentration and biokinetics of heavy metals in the earthworms. Environ. Pollut. 89 (3), 293–301. Nisbet, A.F., Shaw, S., 1994. Summary of a five-year lysimeter study on the time dependent transfer of 137Cs, 90Sr, 239,240Pu, 241Am to crops from three contrasting soil types. 2. Distribution between plant different plant parts. J. Environ. Radioact. 23, 171–187. Nykyri, M., Nordman, H., Marcos, N., Löfman, J., Poteri, A., Hautojärvi, A., 2008. Radionuclide release and transport — RNT-2008. Report 2008-06. Posiva Oy, Olkiluoto. Palm, V., 1994. A model for sorption, flux and plant uptake of cadmium in soil profile: model structure and sensitivity analysis. Water Air Soil Pollut. 77, 169–190. Pereza MA, Ayala-Fierro F, Barber DS, Caserez E, Rael LT. Effect of micronutrients on metal toxicity. In: Reviews in Environmental Health (1998): Toxicological Defence Mechanics. Edit: Hook GER, Lucier GW. pp.206. Reinman, C., Koller, F., Frengstad, B., Kashulina, G., Niskavaara, H., Englmaier, P., 2001. Comparison of the elemnt composition in several plant species and their substrate from a 1500000-km2 area in Northern Europa. Sci. Total Environ. 278, 87–112. Roivainen, P., Makkonen, S., Holopainen, T., Juutilainen, J., 2011a. Soil-to-plant transfer of uranium and its distribution between plant parts in four boreal species. Boreal Environ. Res. 16, 158–166. Roivainen, P., Makkonen, S., Holopainen, T., Juutilainen, J., 2011b. Transfer of elements relevant to radioactive waste from soil to five boreal plant species. Chemosphere 83, 385–390. Roivainen, P., Makkonen, S., Holopainen, T., Juutilainen, J., 2012. Element interactions and soil properties affecting the soil-to-plant transfer of six elements relevant to radioactive waste in boreal forest. Radiat. Environ. Biophys. 51, 69–78. Ryfyikiri, G., Wannijn, J., Wang, L., Thiry, Y., 2006. Effects of phosphorous fertilization on the availability and uptake of uranium and nutrients by plants grown on soil derived from uranium mining debris. Environ. Pollut. 141, 420–427. Salbu, B., Lind, O.C., Skipperud, L., 2004. Radionuclide speciation and its relevance in environmental impact assessments. J. Environ. Radioact. 74 (1-3), 233–242. Sheppard, S.C., Evenden, W.G., 1997. Variation in transfer factors for stochastic models: soil-to-plant transfer. Health Phys. 72 (5), 727–733. Sheppard, M.I., Sheppard, S.C., 1985. The plant concentration ratio concept as applied to natural U. Health Phys. 48, 494–500. Sheppard, S.C., Sheppard, M.I., 1991. Lead in boreal soils and food plants. Water Air Soil Pollut. 57-58, 79–91. Simon, S.L., Ibrahim, S.A., 1987. The plant/soil concentration ratio for calcium, radium, lead and polonium: evidence for non-linearity with references to substrate concentration. J. Environ. Radioact. 5, 123–142. Sims, T., 1986. Soil pH effects on the distribution and plant availability of manganese, copper and zinc. Soil Sci. Soc. Am. J. 50 (2), 367–373. STUK (Radiation and Nuclear Safety Authority), 2001. Long-term Safety of Disposal of Spent Nuclear Fuel Guide YVL 8.4 STUK, Helsinki. Timperley, M.H., Brooks, R.R., Peterson, P.J., 1970. The significance of essential and nonessential trace elements in plants in relation to biogeochemical prospecting. J. Appl. Ecol. 7, 429–439. Tuovinen, T.S., Roivainen, P., Makkonen, S., Kolehmainen, M., Holopainen, T., Juutilainen, J., 2011. Soil-to-plant transfer of elements is not linear: results for five elements relevant to radioactive waste in five boreal forest species. Sci. Total Environ. 410, 191–197. Tuovinen, T.S., Saengkul, C., Ylipieti, J., Solatie, D., Juutilainen, J., 2013. Transfer of 137Cs from water to fish is not linear in two northern lakes. Hydrobiologia 700, 131–139. Vera Tome, F., Blanco Rodríguez, M.P., Lozano, J.C., 2003. Soil-to-plant transfer factors for natural radionuclides and stable element in a Mediterranean area. J. Environ. Radioact. 65, 161–175. Vermeulen, F., Van den Brink, N.W., D'Havé, H., Mubiana, V.K., Blust, R., Bervoets, L., De Coen, W., 2009. Habitat type-based bioaccumulation and risk assessment of metal and As contamination in earthworms, beetles and woodlice. Environ. Pollut. 157, 3098–3105. Walkley, A., Black, I.A., 1934. An examination of the Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci. 37, 29–38. Windish, W., 2002. Interaction of chemical species with biological regulation of the metabolism of essential trace elements. Anal. Bioanal. Chem. 372 (3), 421–425.