EUROPA > DG Health and Consumer Protection > Public Health  Contact | Search | What's New? | Subscribe | Site Map | Index 
Depleted Uranium home
Source document:
SCHER (2010)

Summary & Details:
Media Consulta

Depleted Uranium

6. What harmful effects can uranium have in the environment?

The SCHER Opinion states

Limited data on the ecotoxicity of U are available. In the US EPA ECOTOX database, only 46 records are available for U toxicity to aquatic species. LC50 values range from 21- 32,700 μg/L in crustaceans, 36,300 for an algal species, 4,000 – 100,000 μg/L in fish and 2,900-3,900 μg/L in an invertebrate species (H. viridissima). No data are recorded in the ECOTOX database for U toxicity in terrestrial species. For U oxide, four records are available in the US EPA ECOTOX database, all for the water flea C.dubia. The reported NOEC level is 30 μg/L and the LC50 is 50 μg/L (US-EPA, 2009).

Uranium in the aqueous environment generally occurs as the uranyl ion (UO22+). In freshwater at a pH > 6, the uranyl ion forms complexes with carbonate ions (Poston et al., 1984).
The ECOTOX database contains data for uranyl sulfate (55 records, 9 species) and uranyl nitrate (105 records, 14 species). For uranyl nitrate, a (90-120 d) NOEC of 2,000 μg/L was recorded in alga. The 48 h EC50 in D.magna ranges between 4,000 and 74,000 μg/L. The 48 h LC50 in C.dubia ranges between 60-89 μg/L, whereas the (7 d) NOEC ranges between 1.5 and 8 μg/L. In fish, the 96 h LC50 values are above 3 mg/L. In duckweed, a NOEC of 500 μg/L was recorded. No data are available for uranyl nitrate in terrestrial organisms. For uranyl sulfate, the lowest (5 d) NOEC value reported was in the daphnid M. macleayi at 10 μg/L. The lowest reported LC50 in fish is 2.5 mg/L, and the lowest (4 d) NOEC is 560 μg/L. In the invertebrate H. viridissima, a (5 d) NOEC of 150 μg/L is reported. For the terrestrial environment, the (0.5 h) LOEL in reindeer lichen is 0.1 M.

The Dutch RIVM has summarized information on the occurrence and toxicity of U in the environment (Van de Plassche et al., 1999). On the basis of aquatic and terrestrial ecotoxicity data reviewed, a maximum permissible addition to background levels of 1.0 μg U per L in both seawater, freshwater and groundwater was proposed. For soil, a background concentration of 2.9 mg/kg was derived and a maximum permissible concentration of 28.3 mg U/kg of soil was proposed.

These risk limit values were proposed based on toxicity data taken from the literature. Several, but not all of the studies corresponded to the ones used in the US-EPA ECOTOX database. Chronic toxicity of U to freshwater crustaceans ranged from 10 – 1,290 μg/L (NOEC, 2 species). Acute toxicities in crustaceans ranged from 400 to 30,000 μg/L, whereas, in fish, the LC50 ranges from 730 to more than 100,000 μg/L. For the terrestrial environment, the RIVM study (Van de Plassche et al., 1999) quoted Sheppard et al. 1992 who reported a NOEC for plants of 254 mg U per kg dw of soil, and a LC50 for the earthworm L. terrestris of more than 1000 mg/kg (Sheppard et al., 1992).

Sheppard and collaborators (2005) later reviewed the chemical toxicity of U and proposed a suite of ecotoxicity thresholds for U (Table 4). The most sensitive organisms in this evaluation appeared to be the freshwater invertebrates and freshwater plants, for both of which a PNEC of 5 μg/L was proposed (Sheppard et al., 2005). They also concluded that in human risk assessments the chemical toxicity of U is the focus, and that the same is expected for non-human biota.

Terrestrial plants

250 mg U/kg dry soil

Based on one study with multiple plant species and soils.

Other soil biota

100 mg U/kg dry soil

There is some evidence that certain other soil biota and processes are more sensitive than plants and effect concentrations at this level have been reported.

Fresh water invertebrates

0.005 mg U/L

Derived as the 5th percentile of the distribution of observed effect concentrations, with the implication that 95 % of biota would be protected using this as a guideline concentration.

Freshwater benthos

100 mg U/kg dry sediment

Based on the LEL approach of observed benthic populations in U impacted sediments.

Freshwater fish

0.4 mg U/L in hardness <10 mg/L
2.8 mg U/L in hardness 10–100 mg/L
23 mg U/L in hardness >100 mg/L

There was a good relationship between effect concentrations and water hardness from a number of studies, the functional expression (units of mg/L) was: effect concentration = 0.26 (hardness).

Freshwater plants

0.005 mg U/L

Equivalent to the GM effect concentration for Chlorella, with a safety factor of about 10-fold. Because this resulted in a value very similar to that proposed for aquatic invertebrates, that number was used.


Same as mammals

Only one study, which concluded that birds were 100-fold less sensitive than small mammals.

Mammals, renal damage

0.05 mg U/kg x d, body mass 1 kg
0.01 mg U/kg x d, body mass 1000kg

Based on 3 studies from the same laboratory. Extrapolation to 1000 kg animal is based on relationship of body mass to the power 0.75. It is not clear if this renal damage would have an ecological consequence.

Mammals, growth and development

0.1 mg U/kg x d, body mass 1 kg
0.02 mg U/kg x d, body mass 1000kg

Based on 3 studies from the same laboratory with a 10-fold safety factor. Extrapolation to 1000 kg animal is based on relationship of body mass to the power 0.75.




Table 4. Ecotoxicity thresholds proposed for U by Sheppard et al. (2005).

Environment Australia (2000) has proposed a freshwater low reliability trigger value of 0.5 μg/L that was calculated for U using an AF (assessment factor) of 20 on limited chronic data. No marine data were available to calculate a guideline value. This value should only be used as an indicative interim working level (Environment-Australia, 2000).
The OEHHA (California Office of Environmental Health Hazard Assessment) has withdrawn the previously established PHG for U of 2 picocuries per L of water, and announced to develop and adopt a new PHG in accordance to Health and Safety Code, Section 116365. Based on the current review of the new information, it can be concluded that relatively few data are available for the ecotoxicity of U and that hardly any such data are available for the terrestrial environment (OEHHA, 1998).

The GreenFacts Three-Level Structure used to communicate this SCHER Opinion is copyrighted by Cogeneris SPRL.