Chemosphere xxx (2013) xxx–xxx Contents lists available at ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere Sustainable agricultural use of natural water sources containing elevated radium activity Effi Tripler a,⇑, Gustavo Haquin b, Jean Koch b, Zehava Yehuda a,c, Uri Shani c a Southern Arava Research and Development, Hevel-Eilot 88820, Israel Radiation Safety Division, Soreq Nuclear Research Center, Yavne 81800, Israel c Department of Soil and Water Sciences, Faculty of Agricultural, Food and Environmental Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot 76100, Israel b h i g h l i g h t s  The environmental implications of using water containing Ra for irrigation were investigated.  Radium was found to accumulate in crops leaves following the evapotranspiration current.  Sorption of 226 Ra to soil particles hinders its matrix mobility. 226 Ra of 0.6–1.6 Bq L1.  Crops can be irrigated with the activity of a r t i c l e i n f o Article history: Received 30 June 2013 Received in revised form 11 November 2013 Accepted 12 November 2013 Available online xxxx Keywords: Radium Irrigation water Lysimeter study Radiological hazard a b s t r a c t Relatively elevated concentrations of naturally occurring radium isotopes (226Ra, 228Ra and 224Ra) are found in two main aquifers in the arid southern part of Israel, in activity concentrations frequently exceeding the limits set in the drinking water quality regulations. We aimed to explore the environmental implications of using water containing Ra for irrigation. Several crops (cucumbers, melons, radish, lettuce, alfalfa and wheat), grown in weighing lysimeters were irrigated at 3 levels of 226Ra activity concentration: Low Radium Water (LRW) < 0.04 Bq L1; High Radium Water (HRW) at 1.8 Bq L1 and (3) Radium Enriched Water (REW) at 50 times the concentration in HRW. The HYDRUS 1-D software package was used to simulate the long-term 226Ra distribution in a soil irrigated with HRW for 15 years. Radium uptake by plants was found to be controlled by its activity in the irrigation water and in the soil solution, the physical properties of the soil and the potential evapotranspiration. The 226Ra apeared to accumulate mainly in the leaves of crops following the evapotranspiration current, while its accumulation in the edible parts (fruits and roots) was minimal. The simulation of 15 years of crop irrigation by HYDERUS 1-D, showed a low Ra activity concentration in the soil solution of the root zone and a limited downward mobility. It was therefore concluded that the crops investigated in this study can be irrigated with the natural occurring activity concentration of 226Ra of 0.6–1.6 Bq L1. This should be accompanied by a continuous monitoring of radium in the edible parts of the crops. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction Relatively elevated concentrations of natural radium isotopes are found in groundwater in the southern part of Israel in the two main aquifers of the Negev Desert and the Arava Valley: the Nubian Sandstone aquifer (Kurnub Group) and the Lower Cretaceous aquifer (Judea Group). Radium is transferred from the host rock into the aquifer by geochemical processes and is commonly Abbreviations: Ra, Radium; BTC, breakthrough curve; ETp, potential evapotranspiration; MCL, Maximum Contaminant Level; TF, transfer factor; REW, Radium Enriched Water; HRW, High Radium Water; LRW, Low Radium Water. ⇑ Corresponding author. Tel.: +972 54 9799182. E-mail address: tripler@agri.huji.ac.il (E. Tripler). found in the groundwater as four isotopes: 226Ra (T1/2 = 1600 y) from the 238U decay series, 228Ra (T1/2 = 5.75 y), and 224Ra (T1/2 = 3.66 d) from the 232Th decay series and to a lesser exten 223 Ra (T1/2 = 11.435 d) from the 235U decay series (Vengosh et al., 2007). The water in some of the wells in the Southern Arava contains radium isotopes at concentrations of 1–2 Bq L1. These activity concentrations exceed the Maximum Contaminant Level (MCL) for drinking water in Israel set at 0.5 Bq L1 and 0.2 Bq L1 for 226Ra and 228Ra, respectively (Koch and Haquin, 2008). With regard to uptake by plants, most of the information is expressed as a transfer factor (TF), which is calculated as the ratio of the element concentration in plants to its concentration in soil. A comprehensive compilation of soil to plant transfer factors was 0045-6535/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 2 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx conducted by the International Atomic Energy Agency (IAEA, 2010), in which they are reported separately for temperate environments and for tropical and subtropical environments. Ra transfer factors are reported for up to 12 crop categories, for different plant tissues in part of the crop categories, and for up to 4 soil types. The largest amount of Ra data relates to temperate environments, far fewer data are available for tropical ecosystems and none are reported for subtropical environments (to which the Arava Valley belongs). The TF depends on soil characteristics, plant type, tissue type, climate conditions and the physico-chemical form of the radionuclide (Vandenhove et al., 2005; Vandenhove et al., 2009; IAEA, 2010). Field and laboratory measurements can be combined with mathematical models to yield better predictions of the long term distribution pattern of radionuclides in the soil. Particularly, in cases where environmental and health hazards can occur under conditions of regularly fertigation with radionuclides. Mechanism-base numerical models have become a frequently used tool for simulations and pre dictions related to water and ions in the vadoze zone (Van Genuchten and Šimůnek, 2004). Long-term (<200 years) simulation of 226Ra fate in agricultural ecosystems, where phosphogypsum fertilizer was applied for 27 years, was studied by Coelho et al. (2013), using the HP-1 multicomponent transport model. The software combines HYDRUS 1D (Šimůnek et al., 1998), a model simulating one-dimensional variably saturated water flow and ion transport in soils, with PHREEQC (Parkhurst and Appelo, 1999), a model that calculates geochemical reactions and ions distribution in soils. Elevated radium concentrations may be a dominant factor in the potential utilization of groundwater as drinking water, as well as for agricultural purposes. Natural occurring radionuclides, including radium, are present in foodstuffs at varying activity concentrations. In prone areas, where radium concentrations in the ground water are significantly higher than the limits in the drinking water standards, it is important to investigate the effect of using such water sources for agricultural purposes. Irrigation with water containing elevated radium concentrations may contaminate the soil and the radium may consequently find its way into the food chain. The overall objective of the current study was to develop criteria for the sustainable use of water containing high level of radium as irrigation water. Specifically, we aimed to explore chemical reactions of the added radium to soil particles and its movement in the soil, to investigate radium uptake and transport in plants and to develop a comprehensive soil-plant model for predicting Ra uptake in plants. 2. Materials and methods 2.1. Lysimeter studies To study the effect of drip irrigation on radium accumulation in the soil and plant uptake an experimental setup consisting of 9 weighing-drainage lysimeters was installed at Yotveta in the Arava Valley in southern Israel. The lysimeter system, illustrated in Fig. 1, allowed us to carry out controlled experiments in which water and Ra fluxes were measured with high accuracy, and hence, to compute water and ions balances. Each lysimeter consisted of a PVC container filled with soil, a bottom layer of highly conductive porous media (rockwool) in tight contact with the soil, and a drainage pipe filled with the same material extending downward from the lysimeter bottom. The height and the diameter of the growth containers were 0.85 m and 1.5 m, respectively, corresponding to a volume of 1.5 m3. The rockwool drainage extension prevented saturation at the lower soil boundary while permitting water to flow Fig. 1. A schematic representation of a single weighing-drainage lysimeter. out of the soil and be collected. The lysimeter system included automatic water and fertilizer preparation and delivery. Ion and water boundary conditions of the lysimeters are as in Tripler et al. (2012). Individual lysimeters were positioned on square weighing platforms with load cells situated in each corner. All lysimeters were filled with local soil, either an Arava loamysand soil, or an Arava sandy soil). The physical and mechanical parameters of the soils are described in Table 1. Both soils have average activity concentrations of 226Ra of 21 ± 3 Bq kg1 and 232 Th of 12 ± 2 Bq kg1. Several controlled experiments were carried out at three radium concentration levels of the irrigation water: (1) lysimeters 1–3 with Radium Enriched Water (REW) – 50–100 Bq L1; (2) lysimeters 4–6 with High Radium Water (HRW) at 1.8 Bq L1; and (3) lysimeters 7–9 with Low Radium Water (LRW) – <0.04 Bq L1. The purpose of the REW treatment was to simulate an accelerated process of accumulation over a long period of time, i.e. to study the effect of many years of irrigation with water containing high concentration of radium. Along the course of the entire study, the daily irrigation quantity was set so that the drainage is 1/4–1/3 of the total irrigation water amount. This criterion afforded optimized steady state conditions of water content and soil solution ions concentration. Each treatment was equipped by an irrigation valve and a water meter, which gave a signal pulse, for every 5  104 m3 of delivered water, to an attached automated controller. The daily water balance was calculated from: Z 0 L @h dZ ¼ IrðtÞ  DrðwÞ  ETðw; ET p Þ @t ð1Þ where h (m3 m3) is the water content, t is the time, z is a specific depth of interest, L is the total depth of the lysimeter, Ir is the irrigation water amount, Dr is the collected drainage water at the bottom of the lysimeter, ET is the evapotranspiration, whereas w and ETp represents the soil’s matric head and the potential evapotranspiration, respectively. The right hand side of Eq. (1) expresses the measured difference in the weight of the lysimeter between 00:30 and 23:30 of the same day. Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 3 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx Table 1 Physical characterization of Southern Arava soils. Soil parameter a Particle size distribution Sand Silt Clay Bulk densityb Organic matter content Sat. water contenta Residual water contenta Sat. hyd. conductivityc Air entry pressure headc Pore size distribution indexc a b c hs hr Ks wa b Arava loamy sand Arava fine sand Units 83.0 8.0 9.0 1.30 1.30 0.36 0.030 0.15 0.20 0.55 93.0 4.0 3.0 1.60 <0.50 0.30 0.005 0.21 0.05 0.39 % % % Mg m3 % m3 m3 m3 m3 m h1 M – Shani et al. (1987). Ben-Gal and Shani (2002). Shani et al. (2007). 226 Ra uptake was studied in various crops. Cucumbers, melons, lettuce and radish were grown on an Arava loamy-sand soil in lysimeters. Wheat grass, tomatoes and alfalfa were grown on an Arava sandy soil. Alfalfa was grown continuously for more than 6 months. The accumulated amount of 226Ra applied through irrigation, for each lysimeter, during all the experiments, is shown in Table 2. Plant materials sampled from each lysimeter were ovendried over night at 105 °C, ground and quantitatively ashed at 450 °C. Ash samples were placed in standard cylindrical sealed containers of 55 ml volume. The ash samples were then placed in standard cylindrical 55-mL containers, in which they were kept sealed during 21 d to achieve secular equilibrium of the radon with its decay products. They were then measured by low background high resolution gamma spectrometry during at least 24 h. The 226Ra activity concentration in the ash was determined using the gamma peaks count rate of 214Bi at 609.3, 1120.3 and 1764.5 keV and of 214Pb at 295.2 and 351.9 keV. The counting efficiency of the measuring geometry was calculated using MCNP simulations and a measured standard solution (QCY48 E&Z). The accuracy of the measurements is periodically demonstrated through the Mixed Analyte Performance Evaluation Program (MAPEP, 2013). Additionally, core samples of the same volume were also taken from the soil of the lysimeters and measured by high resolution gamma ray spectrometry (Haquin et al., 2008). The 226Ra minimum detectable activity concentrations for 24-h measurements are 0.05 and 0.17 Bq kg1 for vegetation and soil samples, respectively. Batch experiments and miscible-displacement study were carried in the laboratory, in order to determine the partitioned coefficient of 226Ra (Kd) and the breakthrough curve (BTC). The ambient temperature, throughout the experiments, was maintained at 25 ± 1 °C. 2.2. 226 Ra adsorption isotherms Increasing concentrations of 226Ra (2 Bq L1 to 2  105 Bq L1) were added to 30 g of oven-dried Arava sandy loam soil in a total volume of 100 ml. Solutions were adjusted to a pH of 7.5 using 0.1 M NaHCO3. The test tubes were shaken at 25 °C for 24 h or for 30 d. They were then centrifuged at 3000 rpm for 1/2 h and the supernatant was filtered. The equilibrium concentrations of Ra in the solution and in the solid phase were determined by direct gamma ray spectrometry (Haquin et al., 2008). The adsorption isotherms were constructed by plotting the sorbed 226Ra activity (Bq Kg1) against its activity in the equilibrium solution (Bq L1). 2.3. 226 Ra breakthrough curve Oven-dried Arava sandy loam soil at a weight of 1038.4 g was packed in a PVC column diameter and length of 5 and 20 cm, respectively). The bulk density was measured and the porosity was calculated accordingly, yielding values of 1.71 g cm3 and 0.35, respectively. Before the displacement experiment, the soil column was slowly saturated by feeding from the bottom a background solution of 0.1 M NaHCO3 buffered to pH 7.5. After an initial equilibration period with more than 10 pore volumes (PV), a continuous pulse of 226Ra solution having an activity concentration of 4545 Bq L1 was injected at the top of the soil column. The average pore-water velocity was 6.1 cm h1. Effluent samples were collected from the bottom of the column. The radon gas was purged from the samples with aged nitrogen and the radium was then measured using a Quantulus 1220 instrument (Perkin Elmer Ltd.) by alpha pulse height analysis for alpha/beta separation in the Liquid Scintillation Counting (LSC) technique. The BTC was then constructed by plotting relative concentration (effluent concentration Table 2 The accumulated 226Ra activity in the irrigation water for each lysimeter (1–9), in the different experiments (1–6). The confidence intervals were calculated from the total variance, which is the sum of the variance contributions of each water quantity applied at every irrigation event. Lysimeter 1 2 3 4 5 6 7 8 9 Accumulated 226 Ra activity in irrigation per experiment (Bq 11 103) 1 Cucumbers 2 Melons 3 Lettuce/Radish 4 Wheat grass 5 Tomatoes 52.80 ± 0.6176 52.80 ± 0.6176 52.80 ± 0.6176 0.86 ± 0.0102 0.86 ± 0.0102 0.86 ± 0.0102 0.02 ± 0.0002 0.02 ± 0.0002 0.02 ± 0.0002 55.30 ± 0.6411 55.30 ± 0.6479 87.30 ± 1.0208 2.11 ± 0.0241 2.11 ± 0.0241 2.11 ± 0.0241 0.05 ± 0.0005 0.05 ± 0.0005 0.05 ± 0.0005 55.32 ± 0.6461 56.12 ± 0.6464 88.12 ± 1.0308 2.92 ± 0.03418 2.92 ± 0.03418 2.92 ± 0.03418 0.06 ± 0.0007 0.06 ± 0.0007 0.06 ± 0.0007 55.24 ± 0.6569 55.24 ± 0.6569 60.65 ± 0.7094 60.65 ± 0.7094 4.66 ± 0.0545 4.66 ± 0.0545 10.07 ± 0.1178 10.07 ± 0.1178 0.10 ± 0.0011 0.10 ± 0.0011 0.22 ± 0.0025 0.22 ± 0.0025 6 Alfalfa 11.88 ± 0.1391 0.13 ± 0.0014 Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 4 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx divided by influent concentration) against dimensionless time represented by the PV. 2.4. 226 Ra transport modeling 3. Results and discussion The measured BTC was used to calculate transport parameters of 226Ra in the Arava soils, by means of a linear sorption non-equilibrium transport model (Šimůnek et al., 1998), expressed as: hR @C @2C @C @S ¼ hD 2  hv  qb @t @x @x @t ð2Þ @S ¼ a½K d C  S @t R¼1þ considered, since its half-life (1600 years) is much greater than the time period of the simulation. ð3Þ qb K d ð4Þ h where R is the retardation factor (dimensionless), C is the activity concentration of the solution (Bq cm3), D is the hydrodynamic dispersion coefficient (cm2 h1), x is the distance (cm), v is the average pore-water velocity (cm h1), qb is the bulk density (g cm3), S is the sorbed concentration (Bq Kg1) and a is a first-order rate constant describing the kinetics of the sorption process (h1). The rate of Ra sorption (@S/@t) was calculated assuming a linear isotherm (e.g. Eq. (3)). Data of the BTC and the initial and boundary conditions of the miscible displacement experiment were used as an input for STANMOD software (van Genuchten et al., 2012), in order to calculate 226 Ra transport parameters of D, R and a. STANMOD is a windows-based computer software package for evaluating solute transport in porous media using analytical solutions of the convection–dispersion solute transport equation. The patterns of Ra transport and distribution in the soil and in the drainage water were modeled by means of the HYDRUS 1-D software package (Šimůnek et al., 1998). This software simulates the one-dimensional water flow and solute transport involved in consecutive first-order decay reactions in variably saturated soils. It uses the Richards equation for simulating variably saturated flow and the advection–dispersion equations for solute transport introduced in Eq. (2). Spring and autumn growth of tomatoes was simulated for 15 years, under the southern Arava typical climatic conditions and irrigation water quality (1–2 Bq L1 226Ra) and with the fitted sorption and transport parameters. The soil hydraulic parameters, i.e. k(w) and w(h), were taken from Table 1. The irrigation water quantity and potential evaporation in the autumn amount to 610 and 530 mm, respectively. Accordingly, the timevariable boundary conditions in the spring were set to 720 and 625 mm, respectively. The soil profile was leached with 200 mm of water prior to every growing season. Although HYDRUS 1-D is capable to account for radioactive processes, 226Ra decay was not 226 Ra uptake in plants irrigated with water containing elevated concentrations of the radionuclide was measured for cucumbers, melons, lettuce, radish, wheat grass, tomatoes and alfalfa. Average activity concentrations in the leaves and fruits where relevant, are presented in Table 3. The 226Ra activity concentrations in the leaves of plants were up to 37 times higher than its concentrations in the fruits (Table 3). The range of 226Ra activity concentrations in edible parts of crops irrigated with HRW (2 Bq L1) was narrow, regardless of the soil type. The influence of the accumulated monthly evapotranspiration on the 226Ra translocation in alfalfa leaves is shown in Table 4. 226Ra uptake in alfalfa plants increased with increasing evapotranspiration. A paired-sample t-test was used to reject the hypothesis that 226Ra activity concentrations in leaves of alfalfa, irrigated from May to Nov. 2008 with LRW, are similar to those in leaves of plants irrigated with HRW (p = 0.112). A linear relationship was found between the accumulated 226Ra activity in the soil and the normalized (to ET) 226Ra activity concentration in leaves (Fig. 2), without dependence on the type of crop grown. Two linear regression curves were obtained, one for each soil type, indicating different levels of 226Ra uptake by crops. 226 Ra accumulation in the leaves of crops grown on the Arava loamy-sand soil is higher than in those of crops grown on the Arava sandy soil. The measured depth profile of the total 226Ra activity concentration for the lysimeters irrigated with REW is presented in Fig. 3. It is evident that 226Ra accumulation in the soil upper layer is greater in the Arava loamy-sand soil than in the sandy soil. Downward 226Ra transport is more pronounced in the sandy soil, due to its low clay content and high hydraulic conductivity. Our results indicate that 226Ra uptake by crops grown in Arava loamy-sand soil, characterized by a clay content of 9%, is higher than its uptake in the sandy soil (clay content of 3%). However, previous studies showed a negative correlation between CEC and 226 Ra activity in plants (Vandenhove and Van Hees, 2007), and between clay content and availability of 226Ra to plants (Blanco Rodríguez et al., 2008). This contradiction can be explained by the 1-D distribution of 226Ra in both soils as shown in Fig. 3. The total 226Ra activity concentration in the topsoil is higher in the Arava loamy-sand soil than in the sandy soil. It is noteworthy that the total 226Ra activity added to the loamy-sand soil and the sandy soil is 55 315 and 60 646 Bq, respectively (Table 2). Similarly, the total calculated radium in the profile in the corresponding soils was 2360 and 2225 Bq kg1 cm1. Therefore, a higher uptake intensity of 226Ra is expected for the loamy-sand soil, since under regulated drip irrigation regime, the greatest root density is likely to be found in the topsoil. Table 3 The 226Ra activity in fruits and leaves of irrigated crops grown in experiments 1–5, grown on Arava loamy-sand soil and Arava fine-sand soil. The numbers appear inside the square brackets represent one standard deviation around the mean. Lysimeter (226Ra concentration in water) a 226 Ra# concentration (Bq kg1) Cucumber (3a) Melon (3) Lettuce (3) Radish (3) Wheat grass (2) Tomatoes (2) Fruit Leaves Fruit Leaves Leaves Fruit Leaves Leaves Fruit Leaves 1–3 (50–100 Bq L1) 6.6 [1.52] 177 [43.68] 2.6 [0.65] 66 [17.28] 7.9 [2.07] 3.3 [0.86] 13.2 [3.19] 14.2 [3.34] 0.6 [0.15] 20.1 [5.12] 4–6 (2 Bq L1) <0.6 1.7 [0.41] <0.15 0.90 0.23] 1.3 [0.31] <0.16 0.9 [0.24] 0.6 0.14] 0.1 [0.02] 3.7 [0.98] 7–9 (<0.04 Bq L1) <0.6 1 [0.26] <0.12 0.30 [0.07] 1.1 [0.26] <0.12 <0.17 0.3 [0.07] <0.05 2.3 [0.58] Number of replicates or lysimeters. Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 5 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx Lysimeter 9 226Ra (Bq kg1) May Jun. Jul. Aug. Oct. Nov. 407 ± 44.21 304 ± 32.46 318 ± 38.69 449 ± 44.21 524 ± 60.62 280 ± 29.49 <0.85 0.45 ± 0.13 1.05 ± 0.22 1.40 ± 0.55 1.76 ± 0.45 1.07 ± 0.24 <1.0 0.55 ± 0.15 0.41 ± 0.17 0.63 ± 0.16 1.49 ± 0.31 0.73 ± 0.23 R 2 = 0.98 50000 40000 30000 20000 10000 S = 8.37 ⋅ C R 2 = 0.99 0 0.175 0 Exp. 1-3 3000 (Bq l-1) x Fig. 4. Sorption isotherms of 226Ra onto Arava loamy-sand soil. equilibrated with the soil for 1 or 30 d. 0.125 226 Radium was 0.100 Y = 1.67 ⋅10−6 X + 0.0087 * R 2 = 0.96; p < 0.0001 0.075 0.4 Measured Fitted 0.050 0.3 Y = 6.43 ⋅10−7 X − 0.00012 0.025 0 R 2 = 0.94; p < 0.0001 C/C * 2000 Ra activity concentration in equilibrium solution Linear regression exp. 1-3 Linear regression exp. 4-6 (Bq kg-1 mm-1) 1000 226 Exp. 4-6 0.150 S = 22.4 ⋅ C -1 Lysimeter 6 226Ra (Bq kg1) 1d 30 d 60000 (Bq xkg ) Accumulated ET (mm) Normalized 226Ra activity concentration in leaves Month 70000 Sorbed 226Ra activity concentration Table 4 Concentrations of 226Ra in leaves of alfalfa, irrigated with HRW and LRW (lysimeters 6 and 9, respectively), and cumulative evapotranspiration, during the growth season. The confidence intervals for the ET were calculated from the characteristic error of the water meter, and the intervals for the 226Ra content in the leaves were calculated from the Mixed Analyte Performance Evaluation Program (MAPEP, 2013). 0.000 0 20000 Accumulated 40000 226Ra 60000 80000 0.2 100000 0.1 activity in irrigation (Bq) Fig. 2. Normalized⁄ 226Ra activity in leaves as a function of accumulated applied to the soil by means of drip irrigation. 226 Ra 0.0 0 2 4 6 8 10 Pore volumes Total 226Ra in soil (Bq kg-1) 0 50 100 150 200 250 0 10 Depth (cm) 20 30 40 50 60 Exp 1-3 Exp 4-5 Fig. 3. Depth profile of 226Radium in two lysimeters irrigated with water enriched with 226Ra, measured at the end of two experiment sets. The loamy-sandy soil is represented by filled squared symbols, while the fine-sandy soil is symbolized with non-filled circles. The error bars represent one standard deviation around the mean. The temporal pattern of the sorption isotherms, obtained when equilibrating the soil with elevated 226Ra solutions, is illustrated in Fig. 4. A linear sorption model was assumed, since the range of 226 Ra in the equilibrium solution was in the magnitude of 1013– 10 M. The Kd value (24.2 l Kg1) obtained by equilibrating the 10 soil with 226Ra solution for 30 d is significantly higher than its value when equilibrated for only 1 d (8.3. 7 l Kg1), indicating that Ra Fig. 5. The 226Ra miscible displacement data and optimized simulation (filled and unfilled dotted curve, respectively), for 226Ra transport in Arava loamy-sand soil, at an input concentration (C0) of 4545 Bq L1. sorption to soil particles is time dependent. Tachi et al. (2001) reported that the Kd of 226Ra sorbed on bentonite and smectite is in the range of 10–103 l kg1 and its sorption mechanism is dominated by ion exchange phenomena. The clay content of the soil used in the current work (9%) is much lower than in the study of Tachi et al. and therefore a lower Kd is expected. The 226Ra breakthrough curve is shown in Fig. 5. Radium activity was detected in the effluent approximately after half a PV was collected. The relative activity of the effluent (C/C0) rose to 5% at 1 PV and from there on Ra activity increased at a constant rate. A relative activity of 32% was reached after 9.5 PV. Measured 226Ra was compared to predicted steady-state values calculated STANMOD software using the following equation: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u n u1 X RMSE ¼ t ðRaOi  RaPi Þ2 n i¼1 ð5Þ where RaOi and RaPi are the measured and predicted 226Ra activities, respectively, and n is the total number of measurements indexed by i. The calculated RMSE for the data given in Fig. 5 was 0.0166. Therefore, the fitted curve that was calculated with the STANMOD software, assuming linear sorption non-equilibrium miscible displacement, is in good agreement with the measured results. Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 6 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx Table 5 Parameters estimation, for Arava sandy soil, obtained by the application of the linear non-equilibrium model, to the 226Ra miscible displacement experiment, at an input concentration of 4545 Bq L1. Parameter Units Value Lower Upper D R Kd cm2 h1 – L kg1 h1 2.09 20.2 3.61 0.05 4.39 13.3 2.12 0.015 8.57 27.1 5 0.085 a 2.5 226 * Ra in soil solution (Bq l-1) A Root-zone Ra drainage Ra activity 2.0 1.5 C 1.0 B 0.5 0.0 0 2 4 6 8 10 12 14 Years Fig. 6. HYDRUS 1-D simulation of 226Ra activity concentration in the rhizosphere solution (Balck) and in the deep drainage at a depth of 100 cm (gray) as a function of consecutive days. Circled letters indicates a rise in the activity in the root-zone and in drainage water (A and B, respectively), and a temporal reductions in the root zone activity (C). The values of the 226Ra BTC parameters calculated with STANMOD are presented in Table 5. The calculated Kd is smaller than the value obtained from the two batch experiments. This can be explained by the short time period of the BTC measurements (approximately 6 h); compared with the two batch experiments. Evidently, the partition coefficient is a function of the contact time of Ra in the solution with the adsorbing matrix. The relative adsorption equilibrium time of radium to soils was previously found to be reached within hours to days; Wang et al. (1993) found that adsorption approaches equilibrium after an equilibration time of about 11 h. Similar findings were found by Laili et al. (2010) on organic matter. The fitted value of D is small, suggesting that both Ra hydrodynamic dispersion and diffusion are hindered by sorption to soil particles. The fitted values of D, R and a are also in good agreement with results obtained by Tsang et al. (2007), for Cd sorption and miscible displacement measurements: 2.7– 5.71 cm2 h1, 21–33 and 0.015–0.048 h1, respectively. Simulation of 15 years of tomatoes irrigated with HRW quality (1.8 Bq L1), under the Southern Arava climatic conditions is presented in Fig. 6. The average root zone 226Ra activity was weighted by the modeled 1-D root density pattern. The 226Ra activity concentration in the root zone reaches a steady-state level of 1.95 Bq L1 after 6 years, while its activity in the drainage water (below 100 cm) reaches a value of 1.24 Bq L1 after 15 years. Irrigation with saline water entails salt leaching, before each growing period. As a result, a temporal rise in Ra activity, both in the root-zone and in its deep-drainage (A and B, respectively), and counter-wisely, a short-term reduction after steady-state condition has been achieved (C), was calculated by HYDRUS 1-D. 4. Conclusions The effect of irrigation with water containing elevated 226Ra concentration on soil and crops has been experimentally studied, on well irrigated crops grown in lysimeters. The results revealed that the 226Ra is mainly distributed in the upper 10–20 cm of the soil. Radium is then transported into the crop leaves and fruits. It was found that Ra uptake in plants is mainly controlled by environmental conditions: soil solution activity, soil texture and potential evapotranspiration. 226Ra accumulates in the leaves following the evapotranspiration current independently of the crop type, while its accumulation in fruits and roots is minimal. For the sake of comparison, 226Ra activity concentration in the edible parts (excluding leaves), is well under the activity concentration of the a-emitting radionuclides recommended in the Codex Alimentarius for radionuclides of anthropogenic origin. The 226Ra mobility in a typical Arava loamy-sand soil is hindered by its sorption to soil particles and its sorption in this soil was found to be time-dependent Importantly, radium activity in the edible parts is a function of its soil solution activity. Simulation of 15 years of crop irrigation with relatively high 226 Ra activity concentration using a 1-D transport model predicts low Ra activity in the root zone, which may cause a minor, if not negligible, accumulation in edible tissues other than leaves. Irrigation method has a potentially great impact on Ra distribution in agricultural ecosystems. The crops in this study were irrigated with drippers, installed at each lysimeter. Therefore, it should be noted that Ra accumulation in the leaves can be higher, in case of sprinkler irrigation regime, since foliage uptake of Ra is likely to occur. Further research should be focused on the kinetics of Ra adsorption to soil particles, on Ra–Ca competitive adsorption, Ra sorption under high ionic strength and lysimeter miscible displacement studies under unsaturated water content conditions. Acknowledgement This project was funded by Grant No. 4500121077 from Israel Water Authority, Department of Water Quality. References Ben-Gal, A., Shani, U., 2002. A Highly conductive drainage extension to control the lower boundary condition of lysimeters. Plant Soil 239 (1), 9–17. Coelho, C.R.B., Batalha, M.S., Jacques, D., Barbosa, M.C., van Genuchten, M.Th., 2013. Using HYDRUS-HP1 to estimate 226Ra transport in soils following the use of phosphogypsum in agriculture. In: Proceedings of the 4th International Conference Hydrus Software Applications to Subsurface Flow and Contaminant Transport Problems, March 21–22. prague, czech republic. Haquin, G., Tripler, E., Koch, J., Gazit-Yaari, N., Yehuda, Z., Shani, U., 2008. 208 Sustainable use of natural water sources containing elevated radium activity. In: Proc. of the International Conference IRPA 12 Conf. Buenos Aires. IAEA, 2010. Handbook of parameters values for the prediction of radionuclide transfer in terrestrial and freshwater environments. Technical Reports Series No. 472. International Atomic Energy Agency, Vienna. Koch, J., Haquin, G., 2008. A new approach in dealing with the radiological quality of drinking water. In: Proc. of the 12th Int. Conf. of the International Radiation Protection Association (IRPA 12), Buenos Aires, 19–24 Oct. Laili, Z., Yasir, M.S., Omar, M., Ibrahim, M.Z., Philip, E., 2010. Influence of humic acids on radium adsorption by coir pith in aqueous solution. Sains Malays 39, 99–106. MAPEP, 2013, DOE, <http://www.id.energy.gov/resl/mapep/mapep.html>. Parkhurst, D.L., Appelo, C.A.J., 1999. User’s guide to PHREEQC (Version 2). A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations. Water Resources Investigations Report 99–4259, Denver CO. Blanco Rodríguez, P., Vera Tomé, F., Lozano, J.C., Pérez-Fernández, M.A., 2008. Influence of soil texture on the distribution and availability of 238U, 230Th, and 226 Ra in soils. J. Environ. Radioactive 99, 1247–1254. Shani, U., Hanks, R.J., Bresler, E., Oliveira, A.S., 1987. Field method for estimating hydraulic conductivity and matric potential-water content relations. Soil Sci. Soc. Am. J. 51, 298–302. Shani, U., Ben-Gal, A., Tripler, E., Dudley, L.M., 2007. Plant response to the soil environment: an analytical model integrating yield, water, soil type and salinity. Water Resour. Res.. http://dx.doi.org/10.1029/2006WR005313. Šimùnek, J., Šejna, M., van Genuchten, M.T., 1998. The HYDRUS-1D software package for simulating the one-dimensional movement of water, heat, and multiple solutes in variably-saturated media: Version 2.0. IGWMC-TPS-70. International Ground Water Modeling Center, Colorado School of Mines, Golden, CO. Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020 E. Tripler et al. / Chemosphere xxx (2013) xxx–xxx Tachi, Y., Shibutani, T., Sato, H., Yui, M., 2001. Experimental and modeling studies on sorption and diffusion of radium in bentonite. J. Contam. Hydrol. 47, 171–186. Tripler, E., Shani, U., Ben-Gal, A., Mualem, Y., 2012. Apparent steady state conditions in high resolution weighing-drainage lysimeters containing date palms grown under different salinities. Agric. Water Manage 107, 66–73. Tsang, D.C.W., Zhang, W., Lo, I.M.C., 2007. Modeling cadmium transport in soils using sequential extraction, batch, and miscible displacement experiments. Soil Sci. Soc. Am. J. 71 (2–3), 674–681. Van Genuchten, M.T., Simunek, J., 2004. Integrated modeling of vadose zone flow and transport processes. In: Feddes, R.A., de Rooj, G. (Eds.), Unsaturated Zone Modeling: Progress, Applications and Challenges. Springer, New York, pp. 37–69. van Genuchten, M.Th., Pontedeiro, E.M., Šimůnek, J., 2012. STANMOD: model use, calibration, and validation. Trans. ASABE 55, 1353–1366. Vandenhove, H., Van Hees, M., 2007. Predicting radium availability and uptake from soil properties. Chemosphere 69, 664–674. 7 Vandenhove, H., Eyckmans, T., Van Hees, M., 2005. Can barium and strontium be used as tracer for radium in soil–plant transfer studies? J. Environ. Radioactiv. 81 (2–3), 255–267. Vengosh, A., Peri, N., Haquin, G., Paytan, A., Pankratov, I., Elhanani, S., Karpas, Z., 2007. Mechanisms of radium mobilization for radium-rich groundwater from the Nubian Sandstone and carbonate aquifers in the Negev, Israel: implications for fossil groundwater resources in the Middle East. Presented at the American Geophysical Union 2007 Joint Assembly. Acapulco, Mexico, 22–25 May 2007. Paper H42A-03. Vandenhove, H., Olyslaegers, G., Sanzharova, N., Shubina, O., Reed, E., Shang, Z., Velasco, H., 2009. Proposal for new best estimates of the soil-to-plant transfer factor of U, Th, ra, Pb and Po. J. Environ. Radioactiv. 100, 721–732. Wang, R.S., Chau, A.S.Y., Liu, F., Cheng, H., Nar, P., Chen, X.M., Wu, Q.Y., 1993. Studies on the adsorption and migration of radium in natural minerals. J. Radioan. Nucl. Ch. Ar. 171, 347–364. Please cite this article in press as: Tripler, E., et al. Sustainable agricultural use of natural water sources containing elevated radium activity. Chemosphere (2013), http://dx.doi.org/10.1016/j.chemosphere.2013.11.020