The distribution of U, Th and 226Ra derived from the phosphate fertilizer industries on an estuarine system in Southwest Spain

The Distribution of U, Th and 226Ra Derived from the Phosphate Fertilizer Industries on an Estuarine System in Southwest Spain A. Martínez-Aguirre, M. García-León Facultad de Física, Universidad de Sevilla, Apartado 1065. 41080 Sevilla, Spain & M. Ivanovich AEA Environment and Energy, Harwell Laboratory, Oxon OXI 1 ORA. UK ABSTRACT This paper reports an extensive study of the presence of natural radioactivity around a phosphate fertilizer factory complex sítuated in an estuarine area of southwest Spain. The study has concluded that the wastes from such industries are the cause of the enhancement of natural radioactivity in the immediate environment. Thus, significantly high levels of U- and Th-isotopes and 226 Ra are detected in water and sediment samples collected in this area. These conclusions based on the enhanced isotopic concentrations are further supported by the measured U, Th and Ra isotopic activity ratios being quite different from any observed efsewhere in undisturbed estuaries. These isotope activity ratios appear to be ver y sensitive indicators of waste disposal practices in such environments. INTRODUCTION Recently, severa! studies have been carried out in the vicinity of phosphate fertilizer industries in arder to establish the significance of the radiological impact of their operations on the environment as well as on the human population (Strain, 1979; Van der Heidje et al., 1988; Koster et al., 1991; Martínez-Aguirre & García-León, 1991a). While typical rocks contain on average sorne 2·8 ppm of U (UNSCEAR, 1977), a phosphate rock can contain up to 50-300 ppm of U (Gascoyne, 1982). The European Community imports annually sorne 15-17 MT of phosphate rock (Baestslé, 1991 ). The imported material is processed into different products generating wastes such as phosphogypsum. Each year 6-7 MT of phosphogypsum is discharged directly into the sea, and, in addition, sorne 8-11 MT is deposited as so lid wastes in piles (Baestslé, 1991 ). These piles remain unprotected from rainwater, wind and other elements, and thus they could be connected hydraulically to surface or groundwaters as a potential source of contamination. The phosphogypsum is mainly composed of CaS04. This means that it also contains U, Th and Ra in large quantities as a result of co-precipitation with Ca during the process of fertilizer production. It is therefore possible that natural processes could lead to the redistribution of these natural radionuclides into the environment surrounding the waste disposal sites. Strain ( 1979) found a significant increase of Ra activities in ground waters collected around a fertilizer factory at North Carolina, USA. Sorne other workers (Van der Heidje et al., 1988; Koster et al., 1991) have also studied the presence of natural radioactivity, mainly 226Ra and 21 ºPo, in European environments affected by such industries. Also, Paul et al. (1980) investigated significantly high 226Ra concentrations in phosphogypsum at Kerala, India. Recently, Martínez-Aguirre and García-León (1991a) found anoma lously high U-isotope activities in water and sediment samples collected at the Odie! and Tinto rivers in southwest Spain (Figs 1 and 2). Such enhancements were attributed to an industrial complex surrounded by these rivers in which two phosphate fertilizer factories were located. One factory processes around 1-4 MT per year of phosphate rock from Morocco, while the other uses around 0·2 MT of rock from Morocco, Senegal and Togo. Both factories release their wastes directly to the Odie! river or are stored in uncovered piles (the area marked with stars in Fig. 2). U sing the data given by Guimond and Hardin (1989) for phosphate rock from Morocco, the first factory alone may produce around 2-2 TBq of 226Ra, 2-3 TBq of 238U and 1 -4 TBq of 232Th each year. Part of this activity can obviously be redistributed over the environment that surrounds the industrial complex. This paper represents a more complete study of possible redistribution of the radioactivity from the wastes generated by the above phosphate fertilizer factories. To this purpose isotopes of U and Th and 226Ra have been extracted and determined from different surface water and sediment Mediterranean sea Fig. l. Map of Spain showing the studied area in the Andalousian region. collected during 1989. The radionuclide concentrations in these samples and the corresponding activity ratios, which reflect the daughters mobility, and probably, also the origin of the radionuclides, will be used to investigate the distribution of radioactivity over this area of southwestern Spain. SAMPLING AND EXPERIMENTAL METHODS Samples of river water and sediment were collected along the Odiel and Tinto river basins (Fig. 2) during July 1989. Surface water samples of 7-1 O l were collected in plastic bottles during low tide, and pH and temperature (ºC) were measured at the same time. The suspended matter was separated as soon as possible using 0-45 µm pore size Nuclepore filters. It has been preserved for future analysis, since it could provide interesting additional information. Then, the filtrate was acidified to pH ""2 by concentrated HN03, to avoid the growth of microorganisms and to minimize water/wall interactions. At the same time, bottom sediments were collected. Samples of about 1 kg were collected and stored in plastic bags. Then, in the laboratory, sub-samples were dried and powdered before analysis. The extraction method used for U- and Th-isotope separate is given in ® Piles of phosphogypsum O 3km L____J__j Fig. 2. Map of Odiel and Tinto rivers showing the sampling stations in 1989. Ivanovich and Murray (1992) and Martínez-Aguirre (1991). Only a brief description is given here. After homogenization, a few grams of dry sediment, spiked with known quantities of 236U and 229Th standard solutions for chemical yield determinations, are digested with HN03, aqua regia, HC104 and HF. The residue left after the above treatment is dissolved in 7M HN03. The U and Th are co-precipitated with Fe3+ by adding concentrated ammonia to the solution. The U- and Th-isotopes are first separated by using anion exchange resins (Dowex AG 1 X 8 hydrochloric form) conditioned with 8M HCl. Then each fraction is further purified by the removal of Fe whose presence would affect the r.x spectral resolution. For this last purification step the resins in the separation columns are conditioned with 7M HN03. These purified solutions are dried and U and Th are electroplated on to stainless steel planchets using the procedure of Lally and Eakins (1978). The activities are determined by a-spectrometry using surface barrier Si a-detectors. In the case of water samples, sorne 7-101 of water are spiked with 236U and 229Th standard solutions and then boiled down to approximately 11. The U and Th content is then co-precipitated with Fe3 + by the addition of (NH)40H. The precipitate, after separation from supernate by contri fuging, is dissolved by 7M HN03. Finally, the same procedure, described above for digested solid samples, is followed. For the 226Ra analysis in sediment samples, the 222Rn emanation method was used. The samples are first digested with HN03, aqua regia, HC104 and HF and then the residue is dissolved with 8M HCl and the volume made up to 0-51. The solution is then transferred into a sealed glass bottle and the original crop of 222Rn expelled by bubbling ultrapure He gas through. After approximately three weeks secular equilibrium between a new crop of 222Rn and 226 Ra in solution is reached. The 222Rn is again expelled and collected in a ZnOz-lined quartz counting cell. The 222Rn activity is measured by a gas scintillation counter and from it the 226Ra concentration in the sample is determined (Ivanovich & Murray, 1992). RESULTS AND DISCUSSION As already stated, the industrial zone under study is situated on a tongue of land surrounded by the rivers Odiel and Tinto, with their confluence at its tip (Fig. 2). The two rivers are affected differently by the industrial activity. Consequently, the presentation and the discussion of the results will be carried out separately for each river. In the tables and figures the authors have denoted with 'O' and 'SO' the water and sediment samples, respectively, from the Odie! river. Similarly, for the Tinto river samples, prefix letter 'T' and 'ST' have been used. Prefix letters 'OT' and 'SOT' correspond to samples derived from the confluence of the two rivers which forms a common mouth to the Atlantic Ocean. The numbers given to each sample identify the sampling station (Fig. 2). In addition to the U and Th isotopic activities expressed in mBq/1 (for water samples) or mBq/g (for sediment samples), U and Th mass concentrations are also given in µg/1 or µg/g. The two are related by the following conversion formulae: [U] = 0-08 l lA23s [Th] = 0-2436A232 (1) (2) where [U] and [Th] are the mass concentrations and A238 and A232 are the activities of 238U and 232Th, respectively. Odiel river The sampling stations are distributed along the lower reaches of the river close to the industrial zone area (Fig. 2). The exception is station 1 which is situated far upstream of this area close to the river source. Water The U- and Th-isotope activities are given in Table 1 together with pH values. It is apparent that all radionuclides measured in the collected water TABLE 1 U- and Th-Isotope Activities (m/Bq/1) and pH of Water Samples Taken Along the Odiel River 238Th Sample pH 23su 234u 232Th 230Th 01 3-01 197·9 ± 6·8 316-3 ± 10-7 13-4 ± 0-6 49.7 ± 1-7 22-9 ± 0-9 02 7-40 34-2 ± 1-1 36-3±1-2 0-58 ± 0-30 7-02 ± 1-33 2-65 ± 0-78 03 7-20 33·0 ± 2-0 35-0 ± 2-1 1-81 ± 0-14 12-0 ± 0-5 NM NM 04 7-12 54·1 ± 3.4 58-7 ± 3.7 2-21 ± 0-24 14-3 ± 0-7 05 7-08 52·7 ± 3-6 54.3 ± 3.7 2-51 ± 0-25 18-9±1-3 7-52 ± 1-20 8-83 ± 2-89 2-38±0-15 06 6-80 131-3±11-3 136-2± 11-7 4.73 ± 0-67 34-2 ± 3.3 07 6-96 35-8 ± 0-9 39-2 ± 0-9 1-83±0-12 11-2±0-4 NM, Not measured. sample follow the same distribution pattern along the river with two well defined peaks. The isotopic data obtained for sample O 1 are anomalous relative to the rest of the data set. The very low pH value of 3, due to local geology conditions, accounts for the observed enhanced concentration listed. The head waters of river Odiel are chemically aggressive, resulting in effective leaching and transfer of U and Th from the solid to the aqueous phase of the system. The Th isotopes, mainly insoluble in natural waters, have concentrations at least one order of magnitude higher than in average surface waters (Martínez-Aguirre & García-León, 1992; Osmond & lvanovich, 1992), reflecting the higher mobility ofTh at very low pH values. Higher concentrations of U- and Th-isotopes also persist downstream between sampling stations 04 and 06, with a clear peak of activity at sampling station 06. A similar pattern was observed for U-isotopes in samples collected in 1988 (Martínez-Aguirre & García-León, 1991a). The observed pattern is indicative of the existence of a local source between stations 04 and 06. The obvious candidate for this local source of radioactivity has to be related to the phosphate manufacturing complex located between stations 04 and 05. The observed shift in the location of the peak of activity from stations 05 and 06 in the two sampling campaigns is a result of a lower tidal input and a higher activity input (higher waste output from the phosphate factories) to the river at the time of sampling. This situation was encountered during the 1988 sampling campaign when the activity peak was observed higher upstream at station 05 (Martínez Aguirre & García-León, 1991a). Table 2 presents the U and Th contents of the waters collected in the Odiel river, and the corresponding Th/U mass ratio together with the 234U /238U, 23ºTh/234U and 228 Th/232 Th activity ratios. In Fig. 3, U and Th concentrations along the river channel are shown. The Th/U ratios found are consistent with the well known fact of higher trend of Th to be associated with the solid phase in rivers. On the other hand, such results are higher than normally encountered in most river waters (Osmond & Ivanovich, 1992) sampled upstream from the estuary. These results, however, are quite similar to the mass ratios obtained for the estuarine waters of the Guadalquivir river (Martínez-Aguirre, 1991; Martínez Aguirre & García-León, 1992). The 234U/238U activity ratios reported in Table 2 are unusually low for arriver system. A weighted average for the data for stations 02 to 07 is 1-07 ± 0-02. This value is more typical of a primary mineral activity ratio from a fertilizer manufacturing process than the average disequilibrium observed in rivers (Scott, 1982). Sample 01 is clearly a different case. Its low pH promotes the dissolution TABLE 2 U and Th Concentrations (µg/1), Th/U Mass Ratio and Isotopic Activity Ratios in Water Samples Taken Along the Odie! River Sample [U] [Th] Th/U 214u¡21su210Th/214U21sTh/212Th 01 ¡5.9 ± 0·6 3·31±0·15 0·208 ± 0·012 1·60 ± 0,02 0· 157 ± 0·008 1·70 ± 0·07 02 2.75 ± 0·09 0· 145 ± 0·075 0,053 ± 0·027 1,06 ± 0,03 0•71 ± 0-21 4-51 ± 2-49 03 2-68 ± 0-16 0-447 ± 0-035 0-167 ± 0,017 1-07 ± 0-02 0-34 ± 0-03 04 4.39 ± 0-30 0-546 ± 0-059 0,134 ± 0-017 1-08±0-10 0-24 ± 0-02 os 4·28 ± 0-29 0-620 ± 0-061 0-151 ± 0-018 1-02 ± 0-09 0-37 ± 0-04 3-00 ± 0-25 06 10-7 ± 0-9 ¡ .¡7 ± 0-16 0-110 ± 0-018 1-04 ± 0-04 0-25 ± 0-03 1-87 ± 0-67 07 2-88 ± 0-07 0-450 ± 0-029 0-157±0,011 1-10 ± 0-02 0-29 ± 0-01 1-31±0-11 (a) ú) úJ 02 03 04 05 06 07 (b) -a- o \ J \�\ \� ··� 01 02 03 04 05 06 C1l OT1 Fig. 3. U concentration (a) and Th concentration (b) in µg/1 in water samples along the Odie! river. of U-isotope activities from the solid phase to the river system. In such conditions, the 234 U-isotope is preferentially mobilized, resulting in 234U enrichment in the aqueous phase relative to 238U (Gascoyne, 1982) and hence the higher234U /238U activity ratio observed in sample O l. The 23ºTh/234U activity ratios reported in Table 2 are anomalously high when compared with other river systems. Thus, typical activity ratios for river waters are around Ü· l or less (Plater et al., 1992; Martínez-Aguirre & García-León, 1992). In the present case the weighted average 23ºTh/234 U (a) T1 T2 T3 T4 Ts Ts (b) ID ú) o -·� j T1 "f.2 T3 T4 Ts Ts OT1 OT2 Fig. 4. U concentration (a) and Th concentration (b) in µg/1 in water samples along the Tinto river. The material discharged at stations 2 and 4 must also contain U. The fact that enhanced levels of U have not been observed is probably related to the differential mobilities of Th and U in aqueous systems. The U being much more mobile than Th, it will tend to redistribute more rapidly in the mass of water along the river channel, resulting in a smoother distribution pattern, compared to Th which will be incorporated to the solid particles very fast. The isotope activity ratios presented in Table 6 for stations 2 and 4 may help to support this proposition. TABLE 6 U and Th Concentrations (µg/1), Th/U Mass Ratios and Isotopic Activity Ratios in Water Samples Along the Tinto River 234u¡2Jsu 230Th/234U22sTh/232Th 230Th/232Th Sample [U} [Th} Th/U TI 9-26 ± 0-32 4-82 ± 0-23 0-521 ± 0-031 1-83 ± 0-03 0-25 ± 0-01 3-60 ± 0-12 2-67 ± 0-09 T2 3-20±0-15 11-8±3,6 3-69±1·13 1-07 ± 0-04 2-14 ±0-64 0-27 ± 0-05 1-90±0-21 T3 3-82 ± 0-18 0-193 ± 0-028 0-050 ± 0-006 1-09 ± 0-04 0-11±0-01 1-27 ± 0-22 7-2 ± 0-8 T4 3-87±0-19 5.33 ± 1-21 1-38 ± 0-32 1-11 ± 0-04 0-77±0-17 0-10 ± 0-03 1-90 ± 0-21 T5 3-42 ± 0-08 0-136 ± 0-024 0-040 ± 0-007 1-08 ± 0-02 0-085 ± 0-007 7.1 ±0·3 T6 3-68±0-12 0-155 ± 0-015 0·042 ± 0-004 1-08 ± 0-03 0-077 ± 0-004 1-09 ± 0-17 6-0 ± 0-6 OT1 3.97 ± 0-11 0-135 ± 0-016 0-034 ± 0-004 1-09 ± 0-02 0-060 ± 0-004 0-91 ± 0·18 5.9 ± 0-7 OT2 2-80 ± 0-16 0-021 ± 0-008 0-008 ± 0-003 1-11 ± 0-05 0-016 ± 0-002 7.4 ± 2-8 The weight average of the 234U /238U activity ratio of 1-09 ± 0-02 ( excluding sample Tl) is statistically identical to the weighted average obtained for the Odiel river water samples. Therefore, similar conclusions can be drawn on the origin of U-isotope signatures. However, the impact of the presence of the phosphogypsum piles on the Tinto river waters with respect to U-isotopes is not as high as that of the direct input of phosphate fertilizer process wastes on the Odiel river. In fact, it is very difficult to identify the sources of U activities from the data presented in Table 6. If anything, the only firm conclusion is that any Th-enrichment at stations 2 and 4 is rapidly removed from the aqueous column and presumably adsorbed onto the solid particles surface. In fact, all other Th/U mass ratios are ,.._, 10�2, in good agreement with published data for other river systems (Scott, 1982; Martínez-Aguirre, 1991). The 23ºTh/234U activity ratios are consistent with the above arguments, adding weight to the conclusion that discharge to the Tinto river from the phosphogypsum piles contains a great <leal more Th than U. The decrease in the 228Th/232Th activity ratio at stations 2 and 4 in the Tinto river reflects the relative immobility of 228Ra, a precursor of 228Tb in the decay series, in the phosphogypsum piles. Sediment The isotope radioactivity in bottom sediments in the Tinto river are given in Table 7. If compared to published U data for bottom sediments in other river systems (Scott, 1982; Descamps & Foulquier, 1988; Morón et al., 1988; Plater et al., 1992) it becomes clear that the river bottom sediments fire accumulating U and its daughters. The concentrations of 238U, 234U, 226Ra and 23ºTh are lower than those reported in Table 3 for the Odiel river but significantly higher than the published levels for undisturbed river systems, always below 60 mBq/g for all of them (Scott, 1982; Martínez Aguirre , 1991; Martínez-Aguirre & García-León, 1992; Plater et al., 1992). This is not the case of 232Tb and its daughter 228Tb, where the concentrations are similar to those found in the literature for unperturbed systems, usually below 57 mBq/g for 232Tb (Scott, 1982; Martínez-Aguirre, 1991; Martínez-Aguirre & García-León, 1992). However, the fact that all U-derived activities increase towards the confluence with the Odiel river may reflect a possible transfer of radioactive material from the Odiel river to the Tinto river. The ways through which this occurs should be investigated, but must be related mainly to the tidal cycle and suspended matter dynamics. The corresponding activity ratios for the Tinto river bottom sediments are presented in Table 8. Again, as in the case of the Odiel river, the weighted average 234U /238U activity ratio of 1-04 ± 0-03 suggests an Sample - STl ST2 ST3 ST4 ST5 ST6 SOT1 ND, Not detected. TABLE 7 U- and Th-Isotope and 226Ra Activities (mBq/g) in Sediment Samples Taken Along the Tinto River 23su231u226Ra232Th230Th 15,7 ± 1-3 16·8 ± 0,8 25·2 ± 0-9 11,5±3·0 20-3 ± 4-0 195·2 ± 4-4 198-5 ± 4.5 146-0 ± 2-7 30·8 ± ¡.o 341,7±11·6 74.5 ± 2-8 80-9 ± 3.0 33.2 ± 0-5 45·0 ± 4.0 66·4 ± 5-2 133.1 ± 3,8 137,8 ± 3.9 49,6±1·1 36-1 ± 3·6 88-6 ± 6-6 342·0 ± 7.5 363·8 ± 7.9 46·1 ± ¡.o 25,3 ± 2·0 231-0 ± 12·3 516·9 ± 15-9 531-2± 16·3 149-9 ± 0-8 47.5 ± 4-0 490,8 ± 23·3 299-0 ± 12-6 304-5 ± 12-8 50-3±1·7 53-7±4·1 279·5 ± ¡5.¡ 22sTh ND 34.3 ± 1-5 48,9 ± 4-8 36-3 ± 4,1 30-8 ± 2·3 41·0 ± 4·0 55.4 ± 4.0 TABLE 8 U and Th Concentrations (µg/g), Th/U Mass Ratio and Isotopic Activity Ratios in Sediment Samples Along the Tinto River Sample [U][ThjTh/U STI 1·26 ± 0·06 2·84 ± 0-74 2·24 ± 0.59 ST2 15·7 ± 0-5 7-6±0-3 0-485 ± 0·023 ST3 6-0 ±0·2 11-1 ± l·O 1·85±0·18 ST4 10-7 ± 0·3 3.9 ± 0-9 0·831 ± 0-087 ST5 27-5 ± 0·6 6-2 ± 0·5 0,226 ± 0,019 ST6 41-6±1·3 11,7 ± 1-0 0-282 ± 0·025 SOT1 24-0 ± l·O 13·2± l·O 0,550 ± 0-048 234U/23su 230Th/234U 1,07 ± 0-07 1-21±0·25 1.02 ± 0,02 1-72 ± 0·07 1·09 ± 0·05 0·82 ± 0-07 1·04 ± 0,03 0,64 ± 0·05 1-06 ± 0-02 0·64 ± 0·04 1·03 ± 0,02 0-92 ± 0·05 1-02 ± 0·03 0·92 ± 0·06 226 Ra/230Th 1-24 ± 0,25 0-43 ± 0·02 0·50 ± 0,04 0,56 ± 0,04 0,20 ± 0,01 0·31 ± 0·02 0·18±0,01 22sTh/232Th 230Th/232Th 1-77 ± 0-56 1·11 ± 0-05 11·1 ±0·5 1-09 ± 0-13 1-48 ± 0· 15 1-01 ± 0,14 2-46 ± 0·26 1-22±0·10 9.2 ± 0·6 0,86 ± 0-10 10·3 ± 0·8 1-04 ± 0-09 5·2 ± 0·3 artificial ongm of U-isotope signatures such as phosphogypsum piles situated along its right bank. The general trend in the isotopic ratios suggest a U-enrichment in the Tinto river bottom sediments. Observations made from the sediments in the Odiel river are also valid for the Tinto river case. CONCLUSIONS A careful sampling campaign carried out in 1989 for river water and bottom sediment samples along the Odiel and Tinto rivers (southwest Spain) has yielded comprehensive U, Th and 226Ra isotopic data. These data have been interpreted to establish the connection between the observed anomalously high concentrations of U-, Th- and 226Ra-isotopes in waters and sediments, and the operations of phosphate fertilizer factories located in a zone virtually surrounded by both rivers. The conclusion based on the isotopic data is that both liquid and solid wastes arising from the fertilizer production, and discharged from this industrial zone, are responsible for the observed radioactivity enhancement in the immediate environment of the fertilizer plants. ACKNOWLEDGEMENTS The authors thank the Spanish DGICYT (contract PB89-0621) and the AMA of Junta de Andalucía for the financial support of this work. They also wish to thank J. M. Abril, G. Manjón and J. Díaz for helping in the sampling, and are obliged to Miss Sandra Hasler of Isotope Geoscience Section, Harwell Laboratory for invaluable help with the analytical work. REFERENCES Baetslé, L. H. (1991) Study ofthe radionuclides contained in wastes produced by the phosphate industry and their impact on the environment. Commission of the European Communities, EUR 13262 EN, Nuclear Science and Technology. Borrego-Flores, J. & Pendón-Martín, J. C. (1988). Algunos ejemplos de influencia de los procesos antrópicos en el medio sedimentario: La Ría de Huelva. Henares, Rev. Geol., 2, 299-305 (in Spanish). Descamps, B. & Foulquier, L. (1988). Natural radioactivity in the principal constituents of French river ecosystems. Radiation Protection Dosimetry, 24(1), 143----47. Gascoyne, M. (1982). Geochemistry of the Actinides and their daughters. In Uranium Series Disequilibrium: Applications to Environmental Problems, ed. Ivanovich, M. & Harmon, R. S. Clarendon Press, Oxford, chap. 2. Gascoyne, M. (1992). Geochemistry of the Actinides and their daughters. In Uranium Series Disequilibrium: Applications to Earth, Marine and Environ mental Sciences, ed. Ivanovich, M. & Harmon, R. S. Clarendon Press, Oxford, chap. 2. Guimond, R. J. & Hardin, J. M. (1989). Radioactivity released from phosphate containing fertilizers and from gypsum. Radiat. Phys. Chem., 34, 309-15. Ivanovich, M. & Murray, A. (1992). Uranium Series Disequilibrium: Applications to Earth, Marine and Environmental Sciences, ed. Ivanovich, M. & Harmon, R. S. Clarendon Press, Oxford, chap. 5. Koster, H. W., Marwitz, P. A., Berger, G. W., van Weers, A. W., Hegel, P. & Nienwenhvize, J. (1991). 210Po, 210Pb, 226Ra in aquatic ecosystems and polders, anthropogenic sources, distribution and enhanced radiation doses in the Netherlands. In V Int. Symp. on the National Radiation Environment, Salzburg, Austria. Lally, A. E. & Eakins, J. D. (1978). Paper presented at Symp. Deter. Radionuclides Environ. Bio. Mats. CEGB, London. Martínez-Aguirre, A. (1991 ). Radioactividad natural en diversos compartimentos naturales de Andalucía. PhD thesis, Seville University, Spain (in Spanish). Martínez-Aguirre, A. & García-León, M. (1991a). Natural radioactivity enhance ment by human activities in rivers of southwest of Spain. J. Radioanal. Nucl. Chem., Letters, 155(2), 97-105. Martínez-Aguirre, A. & García-León, M. (1991b). U- and Ra-isotopes in the Guadalquivir river, south of Spain. In V lnternational Symposium on the Natural Radiation Environment, Salzburg, Austria. Martínez-Aguirre, A. & García-León, M. (1992). Natural radioactivity in the Guadalquivir river, south of Spain (submitted). Miyake, Y., Sugimura, Y. & Tsubota, H. (1964). Content ofU, Ra, and Th in river waters in Japan. In The Natural Radiation Environment, eds. Adams, J. & Lowder, W. M . . University of Chicago Press, Chicago, IL, pp. 219-25. Miyake, Y., Sugimura, Y. & Yasujima, T. (1972). Thorium isotope content in river water in Japan. Papers in Meteorology and Geophysics, 24(1), 67-73. Moore, W. S. (1967). Amazon and Mississippi river concentrations of uranium, thorium and radium isotopes. Earth Planet. Sci. Lett., 2, 231-4. Morón, M.C., Martínez-Aguirre, A. & García-León, M. (1988). Determination of levels ofU, Th and Po from different environmental samples. In International Conference on Environmental Radioactivity in the Mediterranean Area. SNE ENS, Barcelona, Spain, p. 111. Osmond, J. K. & Ivanovich, M. (1992). Uranium series mobilisation and surface hydrology. In Uranium Series Disequilibrium: Applications to Earth, Marine and Environmental Sciences, ed. Ivanovich, M. & Harmon, R. S. Clarendon Press, Oxford, chap. 8. Paul, A., Londhe, V. & Pillai, K. (1980). 228Ra and 226Ra levels in a river environment and its modification by human activities. In Natural Radiation Environment III, Vol. 2. US Department of Energy, pp. 1633-54. Plater, A. J., Ivanovich, M. & Dugdale, R.E. (1992). Uranium series disequili brium in river sediments and waters: the significance of anomalous activity ratios. Appl. Geochem., 7(2), 1-5. Santschi, P., Li, Y. H. & Bell, J. (1979). Natural radionuclides in the water of Narragansett Bay. Earth Planet. Sci. Lett., 45, 201-13. Scott, M. R. (1982). The chemistry of U- and Th-series nuclides in rivers. In Uranium Series Disequilibrium: Applications to Environmental Problems, ed. lvanovich, M. & Harmon, R. S. Clarendon Press, Oxford, chap. 8. Strain, C. D. (1979). An evaluation of radium-226 and radon-222 concentrations in ground and surface waters near phosphate mining and manufacturing facility. Paper presented at Annual Meeting of the Health Physics Society, Philadelphia, PA. UNSCEAR (1977). Ionizing Radiation and Biological Effects. United Nations Scient. Comm. on Effect of Atomic Radiation, New York, pp. 86-91. Van der Heidje, H., Klijn, P. & Passchier, W. (1988). Radiological impacts of the disposal of phosphogypsum. Radiation Protection Dosimetry 24(1), 419-23.