Transfer of natural radionuclides from soils to plants in a marsh enhanced by the operation of non-nuclear industries

Transfer of Natural Radionuclides from Soils to Plants in a Marsh Enhanced by the Operation of Non-Nuclear Industries A. Martínez-Aguirre/ l. García-Orellanab & M. García-Leónb ªDep. Física Aplicada, EUITA, Universidad de Sevilla, Carretera Utrera, km 1, 41080 Sevilla, Spaín bFacultad de Física, Universidad de Sevilla, Apdo 1065, 41080 Sevilla, Spain ABSTRACT Two sampling campaigns were performed in 1993 at the marsh area (Odie/ marsh) located in southwestern Spain, in the city of Huelva. Spartina densi flora and substrate soil ( 5 cm deep) samples were collected in severa/ loca tions across the area in each campaign. Activity concentrations of 210Po, U and Th-isotopes were determined in the plant and the substrate samples. The production of phosphoric acid from phosphate mineral in the vicinity clearly enhances the concentrations of these radionuclides in certain areas of the marsh. Moreover, concentrations in plants are affected by the concentration of the same element in its substrate. Indeed, high concentration levels in plants are coincident with high concentration in soils. However, concentration ratios ( CR), defined as the ratio between the concentration of an element in the plant and of that in its substrate, are higher when substrate concentrations are low, whereas low CR values are found in areas where substrate concen trations are high. Moreover, both variables (CR and soil concentration) seem to be non-linear/y related, at least, in the case of radionuclides from the 238U decay chain. INTRODUCTION Environmental assessments use food-chain models to determine the <lose to man from radionuclides released to the biosphere. Traditional food chain models require a plant/substrate transfer coefficient, referred to as a concentration ratio (CR). The concentration ratio or transfer coefficient concept, is generally accepted and widely used in environmental transport models and describes the amounts of nuclide expected to enter a plant from its substrate. The CR approach requires us to assume that the plant and substrate concentrations are linearly related and that the line defining the relation ship goes through the origin. The variability of CR has previously been addressed, but few have specifically addressed the linearity assumption. This assumption implies that, under comparable conditions, the CR for an element is constant, regardless of substrate concentration (Sheppard & Sheppard, 1985). However, plant uptake of elements, both essential and non-essential for growth, differs and may follow a non-linear relationship. Cannon (1952) has stated that U may be a micronutrient for higher plants, which means that CR was not linearly related to the substrate concentration. Mordberg et al. (1976) obtained CR values by fitting linear and hyperbolic functions to data obtained using different soil concentrations of isotopes in the U-Ra series. They showed that the CR was constant over their observed soil concentrations. Cannon (1952) also suggested a linear relationship existing for U between plant and soil concentrations. However, Morishima et al. (1976) reported a non-linear relationship for U between vegetable and soil concentrations. Lopatkina et al. (1970) have suggested that the linearity assumption may hold only for primitive plant forms. Sheppard and Sheppard ( 1985) studied the linearity assumption for U in plants and substrate at two ore bodies on the Precambrian Shield. They found that the relationship between CR and substrate concentration observed in native plants tends to be hyperbolic, with a tendency for CR values to be larger at lower substrate concen trations. It appears, therefore, that U may behave as an essential element at low substrate concentrations. Many data suggest that plant concentrations are unrelated to soil concentrations, and when a line is fitted to the data, the intercept is usually greater than zero (Simon & lbrahim, 1987). Log-normal distributions of concentration ratios were reported by Sheppard and Evenden (1988). They proposed that the linearity assumption, although contrary to what is usually expected for nutrients (essential) elements, may be valid for limited ranges of soil concentrations or for elements not physiologically regulated by plants. Many experiments designed to measure CR values have used only a few different soil concentrations and thus, cannot statistically assess the linearity assumption. Simon and Ibrahim (1987) proposed that a curvi linear relationship reflecting a saturation-type mechanism in plants was more appropriate. A hypothetical relationship for the CR for essential and non-essential elements over a range of substrate concentrations is shown in Fig. 1. This relationship means that plants readily take up elements essential for plant growth when substrate concentrations are low, whereas plant uptake of non-essential elements is generally constant in this substrate concentration range (Timperley et al., 1970; Mengel & Kirkby, 1979). At high substrate concentrations, plant uptake of essential and non-essential elements can either be constant (no toxicity) or can decrease, leading to toxicity or death. Toxicity begins at a toxicity threshold, the substrate concentration at which plant growth, and hence uptake, begin to be impaired. Lethal substrate concentrations of an essential or non-essential element may not exist, and in that case, the CR may be constant for severa} orders of magnitude. The CR concept is CR = Cp/Cs = constant, and it is probably useful for non-essential elements at non-toxic substrate levels. The lack of correla tion between plant and substrate concentrations for non-essential elements could be explained if the element is mimicking an essential element (Sheppard & Sheppard, 1985). This fact would result in a non-linear rela tionship between the CR and the substrate concentration at low substrate concentration. This relationship could be modelled as: Deficient To!erant Non Essentiol- Soil Concentrotion Toxic ... Toxicity Threstíold Non-toxic /Lethal (1) Fig. l. Hypothetical relationship of CR vs soil concentration for both essential and non essential elements through the deficient to toxic soil-concentration range (Sheppard & Sheppard, 1985). where Cp is the concentration in the plant and Cs is the concentration in the substrate. This results in: Cp a CR =Cs = Cs +b. (2) If the first relationship does not have a zero intercept (a i-O), this would mean that CR (Cp/Cs) is dependent on the substrate concentration. Tracy et al. (1983) studied the uptake of 226Ra, 21°I>b and U from soils to garden produce. They observed a clear trend of increasing concentrations in plants with increasing concentration in soil substrate. In order to express this trend quantitatively, a linear regression analysis was carried out for the logarithms of the mean vegetable concentration vs the logarithms ofthe soil concentrations. This involves fitting a function of the form: or CR = acb-i s ' where a and b are parameters to be determined. (3) (4) A value of b equal to 1.00 would imply a linear relationship (CR = Cp/ Cs = cte). They found that their results were not inconsistent with the line arity assumption, at least for concentrations up to about 37 Bq g-1 in soil. The use of concentration ratios to predict plant concentrations from low substrate concentration for environmental assessment purposes should be investigated further for radionuclides known to mimic essential elements and should be measured under natural conditions (Sheppard & Sheppard, 1985). This paper reveals a study of the distribution and transfer of 210Po-, U and Th-isotopes from soils to Spartina densiflora (a wild plant species typical of marsh areas in Europe) collected from a saline marsh located in southwest Spain. The presence, in the vicinity, of a fertilizer complex which uses mineral with high U concentration could allow for a wide soil concentration range in the studied area (Martínez-Aguirre & García León, 1997). Indeed, the main aim of the work is to contribute a bit more in the study about how natural radionuclides are being incorporated in the plant from its substrate concentration. Thus, a study of the CR relation ship with the concentration in the soil substrate will follow. Although this species is not directly involved in the human food chain, information on the concentration levels and transfer of radionuclides from contaminated soils will provide important data on the transfer mechanisms in the case of those species more directly involved in the human diet. SAMPLES AND RADIOANAL YTICAL PROCEDURES Several stations were chosen across the Odiel marsh area of Huelva for sample collection. This area is clearly aff ected by the wastes released by two fertilizer factories located in the vicinity (see Fig. 2). The area is constituted by four small islands separated by four channels, which are affected, due to tidal movements, by the inflow of the Odiel riverwaters (Martínez-Aguirre & García-León, 1997). Samples (plant and substrate) were collected, during the low tide, at the border of the channels, in areas covered with water during the high tide. Atlant:c Ocean Huelva Town Fig. 2. Location of the sampling stations in the Odie! marsh area. The locations of the fertilizer complex and the phosphogypsum piles are also given. About 1 kg of soil sample was collected from each station and stored in plastic bags. At the laboratory, they were dried and powdered before the radionuclide analysis. The bulk dry density (g cm -3) and percentage of organic matter were determined in each soil sample. Spartina densiflora samples were collected at the same stations and at the same time as soil samples. The roots of the plants were separated in order to study only the overground part of the plants. Once separated, the samples were washed with water to remove any trace of solid particles. Finally, they were air dried and powdered for analysis. Aliquots of each sample (soil and/or plant) were used for Po- and U and Th-isotopes determinations, separately. They were digested with concentrated HN03 and aqua regia and the final residue dissolved in 8M HN03. Before the digestion, the samples were spiked with known amounts of 208Po activities for 210Po determination, and of 232U and 229Th activities for U and Th isotopes determinations, separately. Polonium-210 was separated from actinides by a solvent extraction method which used TBP (tributylphosphate) as the extractant phase. Polonium remained in the nitric phase, whereas actinides moved to the organic (TBP) phase (Martínez-Aguirre, 1991). The final Po solution was evaporated and self deposited on to 2.2 cm diameter sil ver planchets. Polonium activities were determined by r.(-spectrometry with ion-implanted detectors. For U- and Th-isotope determinations, an ion-exchange method was used. After redissolution of the residue with 8M HN03, precipitation of iron hydroxides was carried out by adding concentrated ammonium. Uranium- and Th-isotopes were separated and purified in a DOWEX AG1-X8 (chloride form) resin (Martínez-Aguirre, 1991). Final solutions were electroplated on to 2.2-cm diameter stainless steel planchets, sepa rately. Uranium- and Th-isotopes activities were also determined by r.( spectrometry with ion-implanted detectors. RESULTS AND DISCUSSION In Tables 1 and 2, data on 210Po-, U- and Th-isotopes in soils and Spartina densiflora for each sampling station are given, respectively. Table 1 also includes the percentage of organic matter in each soil sample. Moreover, in Table 3, concentration ratios (CR), defined as the concentration of an element in the plant (mBq g-1 dry weight) divided by the concentration of the same element in the substrate (mBq g-1 dry weight), are also given. The next sections discuss separately, data on soil and plant concentrations and, finally, the CR results. Soil substrate In general, by considering the levels of all radionuclides from the 238U decay chain, a reasonably wide range of activity concentration in the soil samples, is found. There are samples with low concentrations of 238U and descendants, which can be considered as clearly non-enhanced by the fertilizer industries. In general, concentration levels in these samples are below 70 mBq g-1, levels similar to those typical in undisturbed soils (ranging from 10 to 50mBqg-1, UNSCEAR, 1988). These samples were collected in stations further away from the fertilizer industries, in the west part of the Saltés island along the Punta Umbria river. Also, very low concentration levels were found in two stations at the east side of the Bacuta island opposite the fertilizer industries. This fact must be related to the composition of the sample. Indeed, values of density and organic matter (see Table 1) in both samples reflect that high amounts of sandy particles, with low activity concentrations, are contained in such samples. Sorne other samples reveal intermediate concentration levels (clearly higher than background, but not very high) for all radionuclides from the 238U decay chain, ranging from about 100 to 300 mBq g-1. These samples are located along the east part ofthe Enmedio island and the west part of the Bacuta island, washed by the Mojarrera channel. Finally, there are samples with much higher concentration levels than the rest of the samples. The levels are, in general, above 400 mBq g-1 for all radionuclides from the 238U decay chain. These samples are located in the east part of the Saltés island and the south of the Northern marsh in the Odiel river, all of them along the Odiel river upstream and downstream from the fertilizer industries. The difference found in the samples collected at the east side of the Saltés island must, again, be related to the composition of the samples. Thus, it can be seen that those samples with low concentrations (15 and 17) have density and organic matter values very different from the other samples (16 and 18) that have quite high concentrations of radionuclides. In general, excepting the case of samples with different composition, the differences in activity concentration between samples seem to depend on the location of the sampling sites, which indicates the influence of the inflow of Odiel riverwaters over each area. Thus, the bag of contaminants moves upstream and downstream of the Odiel river. Part of it is clearly moving to the Odiel marsh through the north of the Bacuta island that later flows into the Mojarrera channel. We can also compare our results with other data obtained in soils from the area of Huelva, but in the Tinto marsh area (see Fig. 2) and around the phosphogypsum piles (Bolivar et al., 1995). Levels below 40 and 60mBqg-1 for 210Po and 238U, respectively, in stations located in the east TABLE 1 Activity Concentrations (mBqg-1) of sorne Natural Radionuclides from the U and Th Decay Chains in Soil Samples from the Odie! Marsh Collected in February (F) and November (N) 1993 Code O.M.(%) 21op0 238U 234u 232Th 230Th 22sTh Fl 165 ± 7 189 ± 8 NI 10.8 199 ± 10 181 ± 8 197 ± 8 48.1 ± 2.5 186 ± 9 73.0 ± 7.0 N2 11.5 778 ± 41 698 ± 35 747 ± 37 88.1 ± 6.8 922 ± 52 160 ± 15 N3 14.8 643 ± 28 728 ± 32 751 ± 33 103 ± 5 898 ± 42 149 ± 13 N4 14.6 580 ± 26 540 ± 21 594 ± 23 61.1 ± 4.0 533 ± 26 144 ± 13 N13 15.4 222 ± 11 196± 7 217 ± 8 29.8 ± 1.8 112 ± 9 59.3 ± 6.3 F5 20.9 290 ± 31 360 ± 22 364 ± 22 N5 15.2 213 ± 10 232±9 248 ± 10 41.2 ± 3.0 208 ± 11 48.0 ± 3.3 N6 11.4 141 ± 7 147 ± 6 161 ± 6 31.0 ± 1.8 104 ± 5 14.0 ± 3.7 N7 13.1 130± 7 158 ± 6 179 ± 7 30.3 ± 2.2 125 ± 7 36.2 ± 2.5 N8 9.2 64.6 ± 4.0 64.8 ± 3.2 71.9 ± 3.5 26.6 ± 1.9 41.0 ± 2.6 30.9 ± 2.6 N9 9.7 188 ± 10 127 ± 5 142± 5 24.4 ± 1.6 101 ± 5 35.0 ± 4.0 FIO 12.9 25.1 ± 1.7 95.1 ± 4.9 116±6 NlO 12.7 27.9 ± 2.0 106±4 122± 5 23.6 ± 1.6 28.4 ± 1.9 22.5 ± 1.6 N23 14.5 210 ± 11 245±8 281 ± 9 41.4 ± 3.1 216 ± 13 62.1±5.4 N20 9.8 233 ± 11 258 ± 9 Nl8 17.3 565 ± 23 601 ± 21 Fl7 9.4 108±6 142 ± 10 N17 4.8 41.0 ± 3.2 17.6 ± 0.8 Fl6 13.7 367 ± 34 434 ± 16 Nl6 15.5 441 ± 20 511 ± 18 Nl5 4.6 131 ± 7 94.1 ± 3.7 Fl2 15.6 163 ± 9 147± 6 Nl2 9.3 44.2 ± 2.5 44.8 ± 1.6 Fll 13.3 124 ± 8 124 ± 6 Nll 9.2 194±9 75.4 ± 2.6 Nl9 5.7 54.4 ± 3.7 30.1±1.2 Fl4 12.5 40.5 ± 2.3 36.9 ± 1.6 N14 9.4 54.7 ± 3.9 37.8 ± 1.6 N21 1.1 16.2 ± 2.5 17.9 ± 0.8 N22 2.3 37.2 ± 2.5 32.4 ± 1.3 283 ± 10 44.l ± 3.6 630 ± 21 56.4 ± 3.8 163 ± 12 18.8 ± 1.0 26.5 ± 2.5 498 ± 23 523 ± 19 68.l ± 4.5 101 ±4 17.1±1.2 158±6 48.4 ± 1.8 26.7 ± 1.6 138 ±7 87.4 ± 3.0 22.0 ± 1.4 34.7 ± 1.4 15.4 ± 1.0 43.4 ± 1.8 43.7 ± 1.8 22.4 ± 1.6 19.3 ±0.9 4.9 ± 0.3 36.7 ± 1.4 11.8±0.7 256 ± 16 443 ± 22 30.0 ± 2.7 554 ± 23 87.4 ± 4.7 28.l ± 1.7 71.0 ± 3.7 31.7 ± 1.9 42.8 ± 2.7 8.8 ± 0.5 16.0 ± 0.9 78.5 ± 6.4 120 ± 12 32.8 ± 2.9 100 ± 12 46.3 ±4.5 25.0 ± 1.7 28.9 ±4.l 16.4 ± 1.3 22.5 ± 0.4 4.3 ± 0.9 13.3 ± 1.1 232Th/23su = 2_57e3su)-o.4ss and 232Tb/230Tb = 3.14(23ºTh)-o.sos, with regression coefficients of -0.8623 and -0.9267, respectively. (6) (7) These relationships may reflect that these activity ratios remain constant only for concentrations of 238U and 230Tb above sorne 200 mBq g-1• F or concentrations below this value, contamination of radionuclides from the 232Tb decay chain seems to be undistinguished. Spartina densiflora The same radionuclides were analysed in samples of Spartina densiflora, whose results are presented in Table 2. The activities range, in general, from 2 to 50 mBq g-1 for 210Po and U isotopes, with the activity concen trations of 21ºPo being higher than those of U-isotopes in most samples. There is a difference of one order of magnitude between the highest and the lowest concentrations. As in the case of soil substrates, the highest concentrations in Spartina densiflora were found at the Northern marsh and at the east side of the Saltés island. Thus, in general, the highest radioactive concentrations in plants are found in those collected in areas with the highest radioactive concentration in the soil substrate, and the lowest in those with the lowest concentrations in the substrate. Concen trations of 230Tb are generally lower than those of 210Po and U-isotopes. However, the same pattern of distribution is observed. Concentrations of radionuclides from the 232Tb decay chain are clearly much lower than those of radionuclides from the 238U decay chain (ranging from 0.26 to 4.14 for both Th isotopes). In general, the concentration of 228Tb is higher than that of 232Tb. This fact should be related with the higher mobility of 228Ra. However, it is also true that the radioactive concentration of both isotopes in plants collected in the Northern marsh and in the east side of the Saltés island are higher than in those samples collected in other stations. It seems interesting to study the relationship between these radio nuclides. In Fig. 5(a), radioactive concentrations of both U isotopes are plotted. It can be clearly observed that both isotopes are quite well linearly related (r = 0.996), as was found in the case of the soil substrate samples. Moreover, again the slope of the line (1.073 ± 0.018), for a y-axis intercept zero (0.174 ± 0.262), would give the 234U /238U activity ratio of the plan t. Thus, there is again a slight excess of the daughter isotope. The relation- so ,o JO • ,. .,A , . ' ,'o .§. 20 ::) �-· ,o .-· ,,/ ,,- ,o ·1d J .§., ' o L:C, . , • . o.o 0.5 1.0 20 ""u (mBqlgl .,. , ... ... .. . . 30 2.0 1.5 232Th(mBqlg) 'º . 1.!t 50 ,• 'º fº �B.. 20 10 b ,' . �,. ... . . . . ,' ,, ,. ...... :• ... , ....... . ,' ,' ,. 10 20 "ªu (mBqlg) )O 40 JO e 25 20 i" � f10 ....... -- _ ... - _ ... , '..... . ...�, .. -:.. o+----,----,-----r---..----,--� o ,o 15 20 ,s "'u (mBqlg) )•,.o '·' -2.CJ � .-,H .,...,.,.,, ... .t:: f o J.0 .. • ...... --. _,... . _,, • .§.. , 5 ¡::. 0,5 ........ • . . . o N ,.oj, .r'; . , . 1 o.o 1� 1!) L-:_-, __ ;;;----;i;----io---215 ""Th (mBQlgl 'º 25 10 1S ,,.U (mBqlg) " Fig. 5. Relationship between concentrations of radionuclides from the 238U decay chain in Spartina densiflura samples (a) 234U vs. 238U, (b)210po YS. 238U, (e) 230Tb vs. 238u, (d) 228Th vs. 232Th, (e) 232Th vs. 238u and (f) 232Th vs. 230Th. ship between the concentrations of 210Po and that of 238U [see Fig. 5(b)], is not clear. However, there is still a clear trend of increasing 210Po concen trations when concentrations of 238U increase (r = 0.8515). Comparing the U concentration in the plant with the y-intercept of this line (5.59 ± 1.45), the latter must be considered clearly different to zero. This y-intercept reflects the variation in the 210Po/238U activíty ratio with the 238U concentratíon in the plant. Thus, for a low U concentratíon (back ground levels), there is an excess of 210Po compared to 238U. This excess decreases to secular equílíbríum for high 238U concentratíons. In fact, for plants with enhanced concentrations, both radionuclídes are practically in secular equilibrium. Agaín, a linear relationship (r = 0.8223) can be found between the 23ºTh and 238U concentration in the plants [see Fig. 5(c)]. In this case, the value close to zero of the y-intercept (0.368 ± 1.39) gives a 23ºTh/238U activity ratio quite constant across all the area (0.638 ± 0.110). This value of the activity ratio reflects a clear deficity of 230Th in the plants, opposite to that found in soils. This diff erence between activity ratios in soils and plants can be explained by the lower mobility of Th compared to U. Both Th isotopes from the same decay chain (232Th and 228Th) are also linearly related (r = 0.8950), as can be observed in Fig. 5(d). The value of the y-intercept, gives practically secular equilibrium between isotopes in all samples. More ínteresting seems to be the relationship of the 232Th concentra tions with the 238U and 230Th concentrations [see Fig. 5(e) and 5(f)]. The concentration of 232Th in the plant clearly increases as the concentration of U and/or 23ºTh increases, following linear relationships (regression coefficients of O. 7220 and 0.9248, respectively). As they are isotopes of the same element, the relationship between both Th isotopes is much better. In both cases, the y-axis intercept is nearly zero, given 232Th/238U and 232Th/23ºTh activity ratios of 0.087 ± 0.021 and 0.143 ± 0.015, respec tively. Concentration ratios We were also ínterested in the study of the biological uptake and trans fer of these radionuclides from the soil substrate to the plant. Thus, we have calculated the experimental values of the concentration ratios for each radionuclide in each station. The results are presented in Table 3. For 210Po, concentration ratios ranging from 0.026 ± 0.002 to 0.948 ± 0.082, with a mean value of 0.172 ± 0.204, were found. The standard deviation of this mean value reflects the high dispersion of the data. In fact, differences above one order of magnitude are found. Moreover, it is interesting to notice that different values of this factor depend on the activity concentration of the soil substrate. Thus, high concentration ratios appear in stations with low activity concentrations in the substrate, and low in more contaminated stations. The same occurs in the case of all radionuclides from the 238U decay chain. For both U isotopes (238U and 234U), the experimental concentration ratios range from 0.022 ± 0.002 to 0.269 ± 0.03 l for 238U and from 0.024 ± 0.002 to 0.272 ± 0.032 for 234U, with mean values of 0.083 ± 0.061 and 0.082 ± 0.060, respectively. Again, in both cases, the concentration ratio (CR) depends on the activity concentration of the elements in the soil substrate. The same is found in the case of the concentration ratio for 23ºTh. It ranges from 0.010 ± 0.001 to 0.202 ± 0.022, with a mean value of 0.065 ± 0.057. Thus, it seems that the behaviour of the concentration ratio, which means the uptake by plants of radionuclides from its substrate, depends on the activity concentration of the element in the soil substrate. In the case of both Th isotopes from the 232Th decay chain, the concentration ratio ranges from 0.014 ± 0.004 to 0.065 ± 0.008 for 232Th and from 0.011 ± 0.001 to 0.084 ± 0.022 for 228Tb, with mean values 0.036 ± 0.018 and 0.042 ± 0.025, respectively. The dispersion of the data seems to be not as high as that for radionuclides from the 238U decay chain. However, we must consider that, in this case, the range of Th concentration in the soil substrate is not as wide as in the case of U and daughters. In order to study the behaviour of the concentration ratios for these radionuclides, concentration ratios versus the activity concentration at the radionuclide in the soil substrate have been plotted in Fig. 6 for 238U and daughters and in Fig. 7 for 232Th and 228Tb. The CR values clearly decrease when the activity concentration in the substrate soil increases. In fact, the data can be modelled by fitting functions of the form: CR = aC� (8) Thus, if b = O, this would mean that the concentration ratio is constant, regardless of the activity concentration in the soil substrate. Functions of this type have been fitted for 210Po, 238U, 234U and 230Tb, and these results are presented in Table 4. Thus, for all radionuclides from the 238U, the b value of the hyperbolic function is clearly different from zero, -0.750 ± 0.097, -0.445 ± 0.081, -0.438 ± 0.082 and -0.573 ± 0.098 for 210Po, 234U, 234U and 23ºTh, respectively. However, the linearity assump tion may fit better in the case of U isotopes. Thus, it seems that all these radionuclides follow the behaviour of an essential element for the plant TABLE4 Concentration Ratio (CR) Fitted Functions for Radionuclides from the 238U Decay Chain Radionuclide 210p0 23su 234u 230Th CR =aCt CR = 4.06 c;º·750 CR = 0.59 c;º·445 CR = 0.588 c,0·438 CR = 0.717 c,0573 l.96 0.239 0.247 0.349 0.097 0.081 0.082 0.098 r -0.8296 -0.7137 -0.7030 -0.8256 C, means the concentration of the radionuclide in the soil substrate. growth. Moreover, if activity concentrations in the soil substrate above 200mBqg-1 are only considered, the CR value would be apparently constant. Thus, considering only those stations with high activity concen trations in the soil substrate, CR values of 0.050 ± 0.023, 0.046 ± 0.020 and 0.023 ± 0.011 are obtained for 210Po, U-ísotopes and 23ºTh, respec tively. In this case, if the CR value is used to predict plant concentrations from substrate concentration, it will underpredict plant concentration when the substrate concentration is less than ,..._, 200 mBq g-1. Sheppard and Sheppard (1985) found that the CR for natural U in native plants at the " 'T . :T • º·' , .. : I " � º·'i o.,�,- .,_ .. �--.-------� --. ------�---+- --·--+-- 0.30 0.25 0,20 0::: 0,15 u 0,10 0.05 0.00 200 T 1 T t I 1! ·\.} I _{_ .! I ·. I Pi. ·I. ... 200 <-00 800 800 C8 (mBq/g) 234u I j _t ,oo 600 C. (mBqlgJ 800 O.JO :T , 1 b º·" .T 0.20 +-,-! .. 0:::: 0,15 J".1 ü ' � 0.25 '. º·'° ·r ,! 0.15 T 5 ' º·" l. 200 o.o, 'i:_ . I r .. I1í ··--]I; J[ 0.00 ¡ O 200 I J: I -- --- --- ---_]l_ - ,oo 600 800 C.(mBq/g) ·�-- -- -· -- -- ---------•.--- ' '°º ' 600 C.(mBq/g) ' --, 800 1000 Fig. 6. Concentration ratios vs. the activity concentration of the element in the soil substrate for (a) 210Po, (b) 238U, (e) 234U and (d) 23ºTh. Precambrian Shield appeared to be dependent upon substrate concentra tion below ""20 µgg-1. They assumed the use of the CR concept to predict U concentration in plants to be valid when substrate concentratíons are higher than about 20 µgg-1• In our case, this fact seems to be true for ali radionuclides from the 238U decay chain. The case of both Th isotopes from the 232Th decay chain is not very clear. The CR values (see Fig. 7) seem to be more or less constant compared with those of radionuclides from the 238U decay chain, with mean values of 0.036 ± 0.017 and 0,16 232Th • 0,14 0,12 0,10 a::: 0.08 III 0,06 T I I 0,04 1 t ]: I l: 0,02 ]: :q I I º·ºº' o20 40 60 80 100 120 C. (mBq/g) b 0,10 • 1 ! + I 1 0,05 -'-- I II lI ! I I I I 0,00 +-���--.-�-.---..-----,--���--.�-.---..---.-�-, o 20 40 60 80 100 1W 1� 100 100 C.(mBq/g) Fig. 7. Concentration ratios vs. the activity concentration of the element in the soil substrate for (a) 232Th and (b) 228Th. 0.042 ± 0.025 for 232Th and 228Th, respectively . However, the range of activity concentration of these radionuclides in the soil substrate, should be considered carefully. Thus, it seems that, when studying the uptake of radionuclides by plants from its substrate concentration, it is necessary to consider a reasonably wide range of activity concentration in the soil substrate. Otherwise, nothing can be clearly predicted. lt seems that at least those radionuclides from the 238U decay chain behave as essential elements for the growth of the plant. However, whether these elements are essential or are mimicking essential elements cannot be determined within this study. CONCLUSIONS Concentrations of 210 Po, U- and Th-isotopes in Spartina densiflora and soil substrate samples collected from a saline wet marsh have been studied. The activity concentrations show clearly the influence that the radio activity contained in the phosphate rock used for the phosphoric acid production has over the concentration of natural radionuclides in certain zones of the marsh, affecting the plants and the soils. Indeed, the stations more influenced by the Odiel waters present the highest concentrations in soils and plants. Concentration ratios for all radionuclides in the plant were determined. The concentration ratio of an element seems to be dependent on the activity concentration of the same element in its soil substrate. Thus, high CRs appear in plants growing in soils with a low concentration of radio nuclides, whereas low CRs appear in plants growing in soils with high concentrations. Moreover, the relationship between the CR and the activity concentration of the soil substrate seems to be hyperbolic, at least in the case of radionuclides from the 238U decay chain. As a consequence, if the CR value is used to predict plant concentrations from substrate concentration, it will underpredict plant concentration when the substrate concentration is less than "'200mBq g-1. However, in the case of 232Tb and 228Tb, the narrow range of activity concentration of these radio nuclides in the soil substrate, could be the reason for the apparent constancy of the CR with substrate concentration. ACKNOWLEDGEMENT This work was partially supported under EC contract FI 3P-CT920035. REFERENCES Bolívar, J. P., García-Tenorio, R. & García-León, M. (1995). Enhancement of natural radioactivity in soils and salt-marshes surrounding a non-nuclear industrial complex. Sci. Total Environ., 173/174. Cannon, M. L. (1952). The effect of uranium vanadium deposits on the vegeta tion of the Colorado Platea u. Am. J. Sci., 250, 735-70. Lopatkina, A. P., Komarov, V. S., Sergeyev, A. N. & Andreyev, A. G. (1970). On concentration of uranium by living and dead peat-forming plants. Geochem. Int., 7, 277-82. Martínez-Aguirre, A. (1991). Radioactividad natural en diversos compartimentos naturales de Andalucía. Ph.D thesis, University of Seville, Seville, Spain (in Spanish). Martínez-Aguirre, A. & García-León, M. (1997). Radioactive impact of the phosphate ore processing in a wet marshland in southwestern Spain. J. Environ. Radioactivity, 34(1), 45-57. Mengel, K. & Kirkby, E. A. (1979). Principies of Plant Nutrition (2nd Edn). Internatíonal Potash Institute, Bern, Switzerland. Mordberg, E. L., Aleksandruk, V. M., Kovygin, G. F., Shevchenkov, l. l., Blyumshtein, V.M. & Yushkevish, G. F. (1976). Translocation of isotopes of the uranium-radium series into the grain of certain agricultura! crops. J. Gigiena J. Sanitariia, 2, 58-61. Morishima, H., Koga, T., Kawai, H., Honda, Y. & Kaysurayana, K. (1976). Experimental studies on the movement of uranium in the soil to vegetables. Radioisotopes, 25, 773. Sheppard, S. C. & Evenden, W. G. (1988). The assumption of linearity in soil and plant concentration ratio: an experimental evaluation. J. Environ. Radio activity, 7, 221-47. Sheppard, M. l. & Sheppard, S. C. (1985). The plant concentration ratio concept as applied to natural U. Hlth Phys., 48, 494-500. Simon, S. L. & lbrahim, S. A. (1987). The plant/soil concentration ratio for calcium, radium, lead and polonium: evidence for non-linearity with refer ence to substrate concentration. J. Environ. Radioactivity, 5, 123-42. Timperley, M. H., Brooks, R. R. & Peterson, P. l. (1970). The significance of essential and non-essential trace elements in plants in relation to biogeo chemical prospecting. J. Appl. Eco/., 7, 429-39. Tracy, B. L., Prantl, F. A. & Quinn, J. M. (1983). Transfer of 226Ra, 210Pb and U from soil to garden produce: assessment of risk. Hlth Phys., 44, 469-477. UNSCEAR (1988) Sources, Effects and Risks of Jonizing Radiation. Annex A. Exposure from Natural Sources of Radiation. United Nations Scientific Committee on the Effects of Atomic Radiation, United Nations, New York.