Last update on 9 August 2012
In the continental environment, the mobility of uranium is all the more important as the conditions of the environment are oxidizing and acid. Uranium remains then as uranyl ion U(VI), dominant with acid pH, and may form complex molecules with hydroxides, carbonates or chloride ions or enrich the colloidal fraction. Adsorption on minerals or biological surfaces may delay transport as well as its reduction to U(IV).
The order of magnitude of radioecological parameters describing transfer in agricultural terrestrial ecosystems is known.
In the case of aquatic ecosystems, acute chemical toxicity has been assessed for various trophic levels while the consequences of chronic exposure to uranium at mining sites or following military use of depleted metal, are very poorly known, especially transfer by ingestion for animal organisms. The kinetic aspect of transfers is not known.
Radiotoxicity is characterised by a dominance of internal exposure related to internalisation of an alpha emitter radionuclide.
Uranium, a chemical element with atomic number of 92 is part of the actinide family. A pyrophoric grey metal that is very dense in its pure state, uranium is always found combined with other elements, especially oxygen. It has four possible valences (+III to +VI), with valences 4 and 6 the most common in ore. The conditions for passing from valence 4 to valence 6 depend on the reduction potential of the environment; they resemble the conditions for passing from ferrous iron to ferric iron. Hexavalent uranium is much more soluble than tetravalent uranium; it forms complexes, with the most common being uranylcarbonates and uranylsulfates.
Natural uranium is composed of three main isotopes (234U, 235U and 238U), all of which are radioactive. The two most abundant isotopes on earth are 238U and 235U, which have existed since the planet’s creation. The 234U isotope is produced by alpha disintegration of 238U and represents only a minute share of all uranium. On the other hand, it is more radioactive than the other isotopes and makes up approximately half of all radioactivity from natural uranium. The 235U isotope is the only natural fissile isotope. The enriched, depleted and reprocessed forms of uranium correspond to variable percentages of these isotopes. Other radioactive isotopes of uranium exist, but they are produced artificially (see table below).
Natural uranium such as it is extracted from ore contains by weight 99.275% of the isotope 238, 0.719% of isotope 235 and 0.0057% of isotope 234. Thus for 1 g of natural uranium and without taking into account radioactivity of decay products, the chemical element is distributed as follows:
- 0.99275 g of 238U, or approximately 12,311 Bq;
- 0.00719 g of 235U, or approximately 576 Bq;
- 0.000057 g of 234U, or approximately 12,880 Bq.
(Nucleonica; CE, 2009)
The origin of this radioelement is exclusively natural with redistribution related to human activity (Paulin, 1997). Four principal sources of industrial activity enrich certain compartments, including soil, sediment and water, of the biosphere with uranium:
- nuclear fuel cycle from uranium mining to waste processing: France has only 3% of reserves and no more domestic production of uranium. In general, nuclear fuel is an alloy of uranium, either uranium dioxide, a mix of uranium and plutonium oxide, or uranium carbide;
- military use of depleted uranium (natural uranium in which the 235U content has been reduced by 0.7 to 0.2%). The metal is used for its pyrophoric properties and sites subject to bombardment by this type of weapon are enriched with fine particles of UO2(solid) deposited near the explosion;
- use of coal, which emits uranium into the atmosphere during combustion;
- agricultural use of phosphate fertilisers produced from natural phosphates that are particularly rich in uranium 238.
Conversion rule. Radioactivity/weight conversions given below are performed using the assumption of the measurement of uranium 238 in the expected ratio for natural uranium, with 1 Bq of 238U corresponding to 0.08 mg of natural uranium. These conversions are for purposes of information and comparison. In fact, due to possible imbalances between uranium isotopes in the sampled compartments, only measurement is valid.
Uranium-related radioactivity for the principal components of the continental environment – air, plants, animals, surface water and sediments, underground water – is related to that of soil, which is itself related to adjacent geological formations.
Overall, soils in sedimentary basins and limestone formations have less uranium than those of granite massifs. On a more local level, specific geological conditions may also lead to unusual levels of uranium, particularly in rivers and ground waters (IRSN 2009). In Europe, median uranium concentration in soil is estimated to be approximately 2 mg/kg, with a range of variation from less than 0.1 to more than 50 mg/kg (De Vos and Tarvainen, 2006). The highest concentrations in France are found in the Massif Central in conjunction with secondary hydrothermal alteration of granite domes, vein deposits or the presence of Autunian black schists. These results are coherent with measurements of radioactivity from uranium 238 for the majority of French soil, which vary between several Bq/kg dry weight (d.w.) and several hundred Bq/kg d.w. (several tenths to tens of mg/kg d.w.). They may reach a thousand Bq/kg d.w. (a hundred mg/g) in uraniferous granitic soils. Le Roux (2007) proposes an overall average (including all regions, sites and soil types) of 40 Bq/kg d.w. (or approximately 3 mg/kg d.w.). Nevertheless, such an average is necessarily relative because there is significant influence from the representation of soils analysed with regards to various soils found in France. In this instance, samples of soil found on sedimentary substrate are probably overrepresented. In a compilation of French data from various sources, Picat et al. (2002) indicate that an average of 98 Bq/g d.w. (8 mg/kg) around facilities upstream of the nuclear cycle, for a range of variation from 25 to 173 Bq/kg d.w. (2 to 14 mg/kg).
The United Nations Scientific Commission on the Effects of Atomic Radiation (UNSCEAR) assesses radioactivity in the air from 238U to be around 1 µBq/m3 (UNSCEAR, 2000) based on some references that measure uranium outside the influence of nuclear facilities, as in Poland (1 to 18 µBq/m3; or 0.08 to 1.5 ng/m3) and in the USA (0.9 to 5 µBq/m3; 0.07 to 0.4 ng/m3). At the Cadarache site in France, where the substrate is sedimentary by nature, a lower value of 0.1 µBq/m3 (0.008 ng/m3) corresponding to local background was recently measured.
In Europe, median uranium concentration in sediments is estimated to be approximately 2 mg/kg, with a range of variation from less than 1 to more than 90 mg/kg (De Vos and Tarvainen, 2006). As with soils, sediments from rivers in the Massif Central may have high uranium concentrations of up to 59 mg/kg. These results are consistent with those obtained from measuring 238U. Indeed, according to Le Roux (2007), radioactivity from uranium in sediments from rivers in France resembles that in soil: from several Bq/kg d.w. to several hundred Bq/kg d.w. with a similar average of around 40 Bq/kg (or approximately 3 mg/kg d.w.).
In Europe, the average geochemical background of continental surface water is on the order of 0.3 µg/L while it is estimated to be 0.5 µg/L worldwide (De Vos and Tarvainen, 2006). Variation in uranium concentration in river water extends over four orders of magnitude (average content between 0.02 and 6 µg/L; Bonin and Blanc, 2001). It can attain 11 µg/L and on very rare occasions, it can exceed 20 µg/L. In France, the uranium background in river water is estimated to be 0.44 µg/L in sedimentary areas and 0.15 µg/L in the Hercynian shelf and alpine areas (Salpeteur and Angel, 2010). This inverse distribution from that found in soil is related to water chemistry, particularly acidity. Measured radioactivity of 238U corroborates these variation ranges, since it is estimated to range between 0.01 and 0.1 Bq/L in France (approximately 1 to 10 µg/L). Higher radioactivity is regularly measured in drinking water in France, usually from underground sources. While 95% of water supply samples analysed between 2005 and 2007 have a 238U concentration less than 0.1 Bq/L, 2.5% are between 0.1 and 0.2 Bq/L and 1% exceed this value, with a maximum of 1.5 Bq/L (Loyen and Thomassin 2009, ASN-IRSN, 2009).
In fish from the Rhone River, the concentration in 238U (average from 1,000 samples) was estimated at 1.0 ± 0.2 Bq/kg wet weight (w.w.), or 0.08 mg/kg w.w. (Lambrechts et al., 1992). For aquatic plants in the river, this study gives average concentrations for 238U varying from 71 ±19 Bq/kg d.w. (5.8 mg/kg d.w.) for aquatic moss up to 41 ±7 Bq/kg d.w. (3.3 mg/kg d.w.) for underwater plants.
There are few international publications on uranium radioactivity for air and biological components of the environment, plants and animals, and foodstuffs. UNSCEAR produced a compilation (UNSCEAR, 2000). The table below gives value ranges for Europe (Italy, Germany, United Kingdom, Romania and Poland) and the rest of the world, as well as the adopted reference value.
By default of enough data in France for areas not affected by nuclear facilities, the above values should serve as a reference.
In the marine environment, uranium activity concentrations are more uniform. Mean concentration for this element is 3.3 µg/L, with uranium being often part of complexes or associated with suspended matters (Paquet et al., 2009). Uranium concentrations found in sediments from shallow water differ little from the earth’s crust: approximately 3 mg/kg, or 37 Bq/kg; uranium concentrations in deep marine sediments have the same order of magnitude. Residue from the production of phosphoric acid (phosphogypses) can cause higher concentrations locally, on the order of 100 mg/kg.
Bioaccumulation of uranium in marine organisms is generally low. Concentrations in algae and plankton are on the order of 1 to 10 µg/g d.w. (Knauss and Ku, 1983; Szefer, 1987). Concentrations between 0.1 and 0.3 µg/g d.w. have been reported for molluscs and shellfish in the Baltic (Szefer and Wenne, 1987). Lower or equivalent concentrations (shellfish, 0.003-0.04 µg/g; molluscs, 0.002-0.3 µg/g) have been measured in the coastal areas of Japan that are supplied by estuaries (Takata et al., 2011).
In the area around an uranium processing plant (less than 2 km), activity concentrations of uranium vary widely in the air from 1 to 900 µBq/m3 (Ahier and Tracy, 1997). Near an open air mine in operation, air concentration reaches 33 µBq/m3 (Thomas, 1997). At sites in Serbia and Montenegro where munitions containing depleted uranium were used, activity concentration in the air varies between 2 and 42 µBq/m3 (Jia et al., 2005). The highest value was measured at a site where soils contain a high amount of natural uranium.
Metrology, analytical techniques and detection limits
Quantification of uranium in a sample from the environment can be performed using either weight or isotope methods (Wagner and Vian, 1999; Augeray et al., 2008).
Depending on the case, analysis of liquid or solid samples may be performed directly or after processing the sample (mineralisation, purification and filtration).
Methods using weight
Total uranium contained in a sample is characterised by its weight, expressed in µg, per unit of volume or weight of the sample.
Uranium salts dissolved in water or obtained from a solution after mineralisation are concentrated using evaporation and separated by paper chromatography. A pellet prepared by calcination and fusion with a mixture of fluoride and anhydrous sodium carbonate is subject to ultraviolet radiation with an excitation wavelength of approximately 360 nm. Uranium salts emit fluorescence that is located in the visible spectrum whose measured intensity is comparable to a stock solution (AFNOR, 2003a). Fluorometry characterises uranium content on the order of 1µg/L in water, measurement uncertainty is on the order of 20% (k=2).
The relatively restricted performance of this technique often renders it useful for non quantitative analyses intended for rapid testing for the presence of a strong concentration of uranium.
- Inductively coupled plasma atomic emission spectrometry (ICP-AES)
After mineralisation of solids and filtration and purification if necessary for liquids, the sample is nebulised. The aerosol product is transferred in a plasma which, using atomic excitation, emits spectrum lines that are characteristic of the element being researched. Emission intensity is proportional to concentration (AFNOR, 2003b). Depending on the type of equipment and nebulisation system performances, ICP-AES quantifies minimum concentrations from 10 to 100 µg/L in water.
Comment: perfectly clear water may not require prior treatment.
Methods using isotopes
Total uranium contained in a sample is characterised by the radioactivity of its various isotopes (234U, 235U and 238U), expressed in Bq per unit of volume or weight of the sample.
Quantification using gamma spectrometry with sufficient accuracy for uranium isotopes depends on:
- energy and intensity of gamma radiation emitted by each isotope;
- specific activity of sample;
- as appropriate, applied concentration coefficient (when the sample is subject to evaporation or calcination);
- counting geometry;
- sample density;
- type of detector used;
- the environment in which measurement is taken.
Identification and quantification of non gamma-emitting, low-intensity gamma-emitting, and gamma-emitting isotopes in an interface area with the emission of photons by other radionuclides, may be performed through the intermediary of their short-life decay products, which are also gamma emitters as soon as the radioactive decay equilibrium is reached. This equilibrium shall be very carefully monitored (Papachristodoulou et al., 2003). Causes for disequilibrium are numerous and varied, and include solubility differences, fixation in living organisms, etc.
The isotope emits only one low intensity gamma ray (0.068%) at low energy (49.6 keV) whose use leads to detection limits on the order of several hundred Bq/kg. However, using the assumption of radioactive equilibrium, spectral lines emitted by its decay products may be used, with the most commonly used being 234Th (63 keV (4.8%), 92.4 keV (2.81%) and 92.8 keV (2.77%)). Beyond 226Ra, the equilibrium is potentially very disturbed by the behaviour of the gaseous decay product 222Rn.
The most intense line of this isotope (57.1% at 185.7 keV) is close to the line of 226Ra at 186.2 keV (3.59%). Its lines without interference (143.76 keV (10.96%), 163.3 keV (5.08%) and 205.3 keV (5.01%)) should be avoided due to disturbance from summation effects. The line at 185.7 keV shall thus be used and the contribution of 226Ra deduced, using the knowledge of its decay products located beyond 222Rn, particularly 214Pb at 295.22 keV (18.5%), whose equilibrium must be guaranteed.
Just as with uranium 238, this isotope emits only one low intensity gamma ray (0.1253%) at low energy (53.2 keV) whose use also leads to detection limits on the order of several hundred Bq/kg. It has immediate short-life decay products that can be used to quantify it. For its decay product 230Th, an emitter at 68 keV (0.38%), the detection limit remains on the order of several tens of Bq/(L or kg).
However, even if the method does not overall perform as well (10 to 100 Bq/(L or kg)) as alpha or mass spectrometry, it is frequently used as a preliminary measure to avoid needless use of heavier methods and optimise the concentration of tracer required for these two methods.
This technique is also used to quantify each of the uranium isotopes. It has the advantage of reaching detection limits that are lower than those obtained using gamma spectrometry. It does however require chemical preparation of the sample prior to measurement.
After concentration if required, the preparation consists of separating uranium to be measured in a solid or liquid sample, then producing a thin source that is used to detect and quantify alpha particles.
The principal phases of analysis are (AFNOR, 2005a):
- addition of a tracer;
- mineralisation of sample;
- concentration using precipitation of phosphates;
- chromatographic separation on resin;
- co-precipitation of cerium fluoride;
- alpha spectrometry.
Uranium 232 is used as a tracer because no alpha interference is observed with uranium isotopes 234, 235 and 238. Alpha emission energies from these four isotopes, respectively 5320, 4775, 4397 and 4198 keV, are easily distinguished.
Measurement of radioactivity and determination of overall efficiency of the analyse is done with alpha spectrometry. Spectrometry may be performed using gas detectors (grid ionization chamber) or silicon semiconductors, depending on whether counting efficiency or energy resolution is sought.
Typical detection limits are on the order of 0.01 Bq/(L or kg) or approximately 1µg/(L or kg).
Measurement using inductive-coupled plasma mass spectrometry (ICP-MS) quantifies the number of atoms of each isotope of elements present in the sample (AFNOR, 2005b; ISO, 2007). Conversion in radioactivity is then possible using the value for specific activity (Bq/g) for each radioactive isotope.
If radioactivity level is equivalent, the most easily detected isotopes using mass spectrometry are 238U and 235U because their specific activity is also markedly lower than that of 234U. ICP-MS also detects the presence of 236U, which indicates whether the source of a uranium sample is natural. Approximate values are:
- uranium 238: 1 µg/kg = 0.01 Bq/kg
- uranium 235: 1 µg/kg = 0.1 Bq/kg
- uranium 234: 1 µg/kg = 230 Bq/kg
- uranium 236: 1 µg/kg = 2.4 Bq/kg
Uranium 233 is used as a tracer because no mass interference is observed with isotopes 234, 235, 236 and 238. However, if the quantity of thorium 232 is significant, recombination of thorium 232 with 1 hydrogen may interfere with the atomic weight of 233 of the tracer. Purification using ion exchange resin overcomes the problem by eliminating thorium.
The advantage of ICP-MS over alpha spectrometry is speed of measurement and the possibility of taking direct measurement (without chemical purification) for low-salt liquid matrices.
For uranium 238, quantification limits calculated with routine protocols and using quadripolar ICP-MS are on the order of the mBq/kg per ash sample for solid matrices and hundredths of a mBq/L, or on the order of ng/L, for water or liquids that have undergone radiochemical treatment. For water or liquids measured immediately after dilution – without radiochemistry – mass spectrometry has a detection limit on the order of a hundred ng/L, or on the order of the mBq/L.
Mobility and bioavailability in terrestrial environments
The behaviour of uranium in terrestrial ecosystems is closely related to environmental redox conditions (Cuney et al., 1992; Gueniot et al., 1988). In an oxidant environment, uranium has +VI valence (UO22+ uranyl ion), which is the most stable and mobile form. In anaerobic conditions, it may be reduced to the +IV state in U(OH)4 and U(OH)3 or react with a sulphide.
Uranium mobility in soil is medium. In aerobic conditions, it easily forms complex compounds with organic matter, carbonates, phosphate and sulphates. These compounds, which are more or less soluble, and iron oxy-hydroxides determine for the most part uranium mobility in soil. In particular, these phenomena contribute to the existence of accumulation areas in horizons that have much organic matter. Clay minerals play only a secondary role in soil retention of uranium. The often observed correlation between a soil having many fine particles and its uranium content is probably due to the adsorption on the surface of particles rather than ionic-type or intrafoliar retention. In a reductive environment (e.g., in soil that is flooded or in contact with a water table), uranium has a valence of +IV and thus a tendency to precipitate.
Since it is found in all soils in a form that is partially available, uranium is found in all plants (Mordtvedt, 1996). The average concentrations measured, except for plants from uraniferous regions, are on the order of several millibecquerels per kilogramme of fresh matter with significant variability depending on soil type. In general, lower plants absorb more uranium than higher plants. For the latter, there is a strong correlation between the soil’s uranium content and that found in plants.
Root transfer appears to be the prevailing pathway for terrestrial plant contamination (Paquet et al., 2009). Root absorption depends on the same parameters that affect uranium mobility in soil, i.e., amount of organic matter they contain and the presence of phosphates, sulfates and carbonates. Various studies have demonstrated a competition effect on soil-plant transfer of calcium and magnesium ions when there is a significant concentration in the soil. The phenomenon becomes more pronounced as pH increases. Average root transfer factors vary between 5 ´ 10-3 and 1 Bq/kg of d.w. plant per Bq/kg of d.w. soil. The highest values are measured in tropical climates (IAEA, 2010). On the other hand, uranium is found in higher proportions in the stems and leaves of plants than in cereals and fruit.
There is no data on foliar transfer of uranium to plants. By default, values for plutonium are used, which, given the higher mobility of uranium compared with plutonium, probably results in underestimating the transfer.
Uranium found in meat and dairy products results from the consumption of plants by animals, including food supplements prepared with natural phosphates given to dairy cows (Paquet et al., 2009). Ingestion of soil particles, either directly or through grass covered with soil, may be a significant component of livestock contamination. The transfer parameters of natural uranium are known for the principal species used for meat (cattle, sheep and pigs) as well as cow’s milk. They vary between 4x10-4 and 4.4x10-2 j/L for cow’s milk and between 3.9x10-4 and 7.5x10-1 j/L for beef and chicken respectively.
There is no data on uranium transfer during food processing.
Mobility and bioavailability in continental aquatic environments
Total uranium concentration is controlled by the U(VI) form in oxidant environments and element mobility is related to colloidal fraction. Uranyl ion UO22+ is the dominant species in surface water in oxidant environments up to pH 6. Beyond this, hydroxylated forms appear and then carbonated forms for pH higher than 8.In addition to pH and redox potential, speciation of uranyl in freshwater may be affected by organic ligand concentration — especially humic substances that form stable uranyl complexes — and contributes to the element’s migration in aquatic systems (Moulin et al., 1992). In rivers, more than 90% of uranium may be associated with colloidal fraction, either in relation to iron colloids or through interactions with humic or humin acids. Colloidal fraction governs transport mechanisms of uranium in hydrosystems and tends to diminish linearly as salinity increases. In estuarial environments, less than 5% of uranium is associated with this fraction for salinity on the order of three parts per thousand (Porcelli et al., 1997; Andersson et al., 1998).
Retention of uranium by matter in suspension and sediment decreases as oxidant and alkaline conditions increase (Bird and Evenden, 1996). In such pH-EH conditions, dominant forms are electrically neutral or negatively charged. Highly carbonated water causes a decrease in uranium adsorption on particles due to an increase in the solubility of the element (dominant carbonated uranyl complexes). With acidic pH, the properties of cationic exchanges of sediments may contribute to retention of positively-charged uranyl ions. Ca2+ ions may enter into competition in the ion exchange on potential mineral ligands such as calcite in the solid phase, for example. Phosphate concentration may cause precipitation of uranyl ions and thus their transfer to the sedimentary compartment. This may also be governed by the adsorption process on the organic fraction of solids and/or, in a reductive environment, result from a reduction of U(+VI) to U(+IV), followed by precipitation of the latter. Uranium has a tendency to combine with iron hydroxides and is mainly found during typical sequential extractions in the fraction solubilised by an acid attack. Clay content affects uranium retention on the solid phase.
There has been little study of transfer to aquatic plants. In plants, uranium has a high affinity for protein and lipids with intracellular pH, limiting reversibility of absorption processes. The adsorption phenomena on the surface of cells may be dominant depending on species. The concentration factor extends from 100 to 1,000 (using wet weight) for a species of unicellular green algae (Scenedesmus quadricauda) (Pribil and Marvan, 1976). The accumulation of uranium decreases with concentrations of calcium, magnesium, phosphate and organic ligands and is pH-dependent (Fortin et al., 2007). Primary production may affect contamination of superficial sediment through deposit and cellular lysis.
The literature provides little data relating to macrophyte contamination, even though their role in the food chain is significant. Various research reviews give values for concentration factors (wet weight) from 4 to 2900 for plants in general. The highest transfer factor concerns water milfoil, a species of macrophytes (Paquet et al., 2009).
In animal organisms (crustaceans, molluscs and fish), the concentration factor from water remains very low. In fish, pelagic species accumulate approximately ten times less than benthic species and variation ranges encountered in the literature for overall concentration factors (direct and trophic pathways) extend from 0.01 to 20 (wet weight). In invertebrates, the concentration factor varies between 48 and 1700. The target tissues are bone and kidneys, and to a lesser extend, liver, gills and muscles, digestive system and gonads (Cooley and Klaverkamp, 2000; Simon et al., in press). The range of the concentration factor for muscles extends from 1.55 to 24.3 (L/kg w.w.) compared with 200 to 8,000 for bone tissue (Clulow et al., 1998). Direct (from the water column) and trophic transfer values for uranium do not exhibit the phenomenon of bioamplification (Paquet et al., 2009).
In animal cells, uranium is found mostly in spherocrystals and lysosomes where it co-precipitates with phosphorus (phosphate granules). This phenomenon is observed especially in the digestive gland and the exoskeleton/shell of shellfish and molluscs, and the gills and liver of fish. There is also formation of dense intranuclear granules whose structure resembles that observed when there is intoxication by stable metals (Ribera et al., 1996). Nearly 80% of accumulated content in the freshwater bivalve Corbicula fluminea is found in the insoluble sub-cellular fraction (cellular debris, granules, organelles). In the gills of this mollusc, accumulated concentrations (after 42 days) remain constant after 60 days of depuration (Simon and Garnier-Laplace, 2004).
Uranium is found preferentially in two organs, the kidney and bone/skeleton. The soluble fraction of uranium (composed of complexes of low molecular weight), filtered through the renal glomerulus is eliminated after several weeks (80-90%). Substitution of calcium with uranium in the hydroxyapatite matrix favours trapping in the bones/carapace of animals. The physiology of organisms (vertebrates versus invertebrates) may thus explain variation of uranium organotropism.
Mobility and bioavailability in marine environments
Uranium liberated by continental surface erosion is carried to the ocean essentially in solution in the form of uranyl ion UO22+ (uranium(VI)). It reacts slightly with sedimentary particles such that its residence time in the ocean (2 to 4 x105 years; Ku et al., 1977; Chen and Laurence, 1986) far exceeds the mixing time of the ocean itself (some 103 years). Its concentration, which is relatively uniform overall in ocean basins, is on the order of 3 µg/L, or 37 mBq/L. Uranyl ions may combine with the dissolved organic matter; these complexes may represent up to 20% of the dissolved fraction of uranium in seawater (Mann and Wong 1993).
Dissolved uranium behaves conservatively in numerous estuaries: its concentration varies with the freshwater/saltwater mixture, usually tracked by salt content. Transfers of dissolved uranium from the water column to particles in suspension were demonstrated in the Delaware and Chesapeake estuaries (Sarin and Church, 1994) as resulting from the association of uranium with phosphates and humic substances, coupled with deficits in dissolved oxygen and alkalinity in the area of low salt content.
The most efficient mechanisms for trapping uranium from the water column are found in highly organic sediments in which it is reduced from the U(VI) state to the U(IV) state. Uranium penetrates interstitial water through diffusion from overlying water or is liberated there through the dissolution of iron oxyhydroxide coatings. Uranium may then be associated with sedimentary sulfides. The re-suspension of such sediment is accompanied by remobilization of uranium to free water by means of dilution of interstitial water enriched with uranium (McKee et al., 1987) or dissolution of bearing phases (Barnes and Cochran, 1993).
The transfer of uranium to marine plants demonstrates significant variability within a single group. A uranium concentration factor of 270, obtained in the laboratory, was reported by Hosseini et al. (2008) for marine phytoplankton. This value is close to that measured in situ at diatoms in the Pacific northwest (213; Miyake et al., 1970). For the same periods, other sources recommend values less than an order of magnitude: 10 for Cherry and Shannon (1974) and 20 for the IAEA (2004). In marine macro-algae, similar dispersion in the values of concentration factors may be observed, between 100 (GRNC 1999, IAEA 2004) or 120 (Hosseini et al., 2008) and 700 (Holm and Persson, 1980; Nilsson et al., 1980). Indeed, uranium bio-concentration appears to vary according to algae group. Miyake et al. (1970) thus measured the concentration factors of 61, 79 and 248 for red, green and brown algae, respectively.
Lastly, a uranium concentration factor of 350 was measured in the marine seed plant Heterozostera tasmanica by Ahsanullah and Williams (1986).
As for plants, a variety of concentration factors are reported for marine zooplankton, from 5 (Cherry and Shannon, 1974) to 71 (Miyake et al. (1970). According to certain sources (GRNC, 1999, IAEA, 2004, INERIS, 2010), fish do not concentrate uranium, while others (Hosseini et al., 2008) give a concentration factor of 14. On the other hand, a certain consensus seems to exist for other animal groups. A value of 10 is thus recommended for benthic macro-crustaceans (GRNC, 1999; IAEA, 2004; INERIS, 2010). For molluscs (except cephalopods), a concentration factor of 30 is cited (GRNC, 1999; IAEA, 2004; INERIS, 2010, Hosseini et al., 2008).
Mobility and availability in semi-natural ecosystems
This section is based on an international review of the literature as part of the revision of the IAEA Handbook relating to parameter values for predicting radionuclide transfer in temperate continental terrestrial and aquatic environments (IAEA, 2010).
There is no specific information available on the mobility and bioavailability of uranium in forest ecosystems.
There is no specific information available on the mobility and bioavailability of uranium in arctic ecosystems.
There is no specific information available on the mobility and bioavailability of uranium in alpine ecosystems.
The effects of exposure to ionising radiation depend on the quantity of energy absorbed by the target organism, expressed by a dose rate (µGy/h). This dose rate is obtained by applying dose conversion coefficients (DCC, µGy/h per Bq/unit of weight or volume) to radionuclide concentrations in exposure environments or organisms (Bq/unit of weight or volume).
The characteristic DCCs for uranium 238 were determined without considering decay products and relative biological effectiveness (RBE) using EDEN software (Beaugelin-Seiller et al., 2006) version 2.2, by considering form, dimension and chemical composition of organisms and their living environments, as well as geometric relationships. Model species considered were chosen as examples.
With the exception of fescue (10-1 µGy/h per Bq/kg), internal exposure of organisms to uranium 238 is generally characterised by DCCs with an order of magnitude that varies from 10-4 to 10-3 µGy/h per Bq/kg.
External exposure is less significant and is characterised by significant variability of DCCs depending on the organism, with orders of magnitude ranging from 10-10 to 10-3 µGy/h per Bq/unit of weight or volume.
For more details on how to calculate DCC, see the Environmental Dosimetry Sheet.
Data relating to uranium ecotoxicity in terrestrial and aquatic ecosystems was reviewed by Sheppard et al. (2005) with the purpose of estimating predicted no effect concentration (PNEC) for various types of relevant biota. They propose a PNECsoil of 250 mg/kg for terrestrial plants and 100 mg/kg for soil organisms. In France, the provisional environmental quality standard (EQS) proposed for uranium in a continental aquatic environment, the result of work by INERIS, is obtained summing the geochemical background and the PNECfresh water (0.3 µg/L). IRSN proposed a revision of this PNEC value in 2010 (Beaugelin-Seiller et al, 2010), integrating the speciation of uranium according to the physicochemical properties of the watercourse. The grid of values obtained varies for chronic exposure from 0.3 to 3,510 µg U/L.
Concerning effects on soil microorganisms, the lowest reported effect concentration is 5 mg/kg (dry weight of soil) and concerns nitrogen fixing bacteria. Indirect effects on soil respiration or decomposition rates of organic matter were also demonstrated but for distinctly higher levels of uranium content (Meyer et al., 1998).
Little data exists for uranium phytotoxicity. In particularly uraniferous regions, modification in pigmentation in certain plant species and changes in their growth were observed. Macroscopic effects have been described, such as a reduction in biomass efficiency of plants and seed production (Sheppard et al., 2005) and elongated roots (Panda et al., 2001). Photosynthetic activity in lichens is also affected (Boileau et al., 1985), while oxidant stress and genotoxicity were observed for various higher plant species (Panda et al., 2001; Vandenhove et al., 2006; Vanhoudt et al., 2008). Toxic levels nevertheless vary significantly according to plant species, since Aery and Jain (1998) report an effect on the production of seeds in Triticum aestivum starting at 0.5 mg/kg of soil while other research observes no phytotoxic effect below 1,000 mg/kg of soil (Sheppard et al., 2005).
Data relating to animal organisms in terrestrial ecosystems concerns soil invertebrates, springtails and earthworms (Sheppard et al., 2005). The lowest effect concentration observed concerns survival rate and reproduction in sandy soil of Onychiurus folsomi (species of springtail that consumes roots), which is affected beginning at 92 mg/kg of soil. Other organisms tested appear distinctly less sensitive to uranium. In general toxicity is stronger on fine sand-type soil than humus-type soil.
Concerning freshwater organisms, more than 150 indications of acute ecotoxicity on algae, cnidarians, shellfish, insects, molluscs, plants and fish were found in the literature for approximately 50 from characterisation of chronic ecotoxicity from uranium (Beaugelin-Seiller et al., 2010). The latter basically include the three taxons of lower algae, shellfish and fish, and some rare indication on amphibians, molluscs and cndiarians.
Uranium toxicity in freshwater algae (Chorella) appears pH-dependent (Franklin et al., 2000). Lethal concentrations for 50% of the population exposed (LC50) after 72 hours vary over an order of magnitude between 0.04 and 0.3 mg U/L. For higher plants, the sole LC50 reported (Charles et al., 2006) is 0.8 mg U/L. The no observed effect concentration (NOEC) are in general lower but partially overlaps the range of LC50 (0.009 to 0.22 mg U/L).
For aquatic animals, chemical toxicity of the uranyl ion diminishes as hardness, alkalinity and concentration of organic matter dissolved in water increase (Markish et al., 1996). For Cladocera, LC50 at 24 hours varies between 0.41 and 23 mg of uranium per litre depending on species, pH and hardness of the water used, while NOECs tier between 0.02 and 0.06 mg U/L (Beaugelin-Seiller et al, 2010). For fish, LC50 after 96 hours ranges from 0.7 to 135 mg of uranium per litre and depend on biotic (species, stage of maturity) and abiotic (temperature, hardness, pH) factors. NOECs vary between 0.3 and 6.1 mg U/L. Symptoms describing uranium intoxication progress from an increase in respiratory ventilation to disorganised swimming, loss of equilibrium and colour, haemorrhaging of fins and death. An attack on the gills of fish exposed to metal in water is frequently described in the literature. Uranium toxicity is also due to ionising radiation of its isotopes and the products of their radioactive decay.
Radiotoxicity of radioactive isotope 238U
Uranium is a chemio- and radiotoxic element. Its radiological toxicity is only considered preponderant for mixtures of uranium enriched with the 235 isotope. Several studies, including recent theoretical work by Mathews et al. (2009) support this notion. The PNEC defined in this last study (3.2 µg U/L) appears protective from a radiological perspective for a typical aquatic ecosystem exposed to depleted, natural or slightly enriched uranium after a period of decay of no more than a century. On the other hand, the radiological protection criterion of 10 µGy/h appears to protect this type of ecosystem chemically only during chronic exposure to uranium highly enriched in radioactive equilibrium.
As an alpha emitter, uranium 238 has very weak penetrating power (several centimetres in air, stopped by the horny layer of the epidermis or a sheet of paper) which causes radiation stress mainly related to internal radiation when the radionuclide is incorporated. But as an alpha emitter, uranium 238 demonstrates the greatest biological efficiency among the various types of ionising radiation, which could be characterised among mammals by a relative biological effectiveness (RBE) of 10, value already largely under discussion. Radioactive emission of this radioisotope may thus interact in cells directly or indirectly with biological molecules. Damage to DNA, proteins and lipids from alpha particles emitted by uranium has been observed (Miller et al., 2002). Effects on fish (hatching of eggs, percentage of damage to DNA) have likewise been correlated to the radiotoxicity of uranium (Barillet et al., 2007). While the kidney, the target organ of uranium in animals, is considered slightly radiosensitive, the radiotoxic action of this element would performed preferentially on its second target tissue, bone tissue, which is the primary organ for long-term fixation (Legget, 1994).
Data for uranium radiotoxicity is insufficient to validate the pertinence of the threshold criterion of 10 µGy/h for uranium 238 for protecting ecosystems that has been set at the European level for chronic exposure to external gamma radiation.