Registration Dossier

Ecotoxicological information

Endpoint summary

Administrative data

Description of key information

Additional information

- Toxicity to freshwater organisms

Acute freshwater data:

The acute toxicity values (96h-LC50) for Pimephales promelas varied between of 40.8 µg dissolved Pb/L (at pH 5.67 and hardness of 15.9 mg/L) and 3598 µg dissolved Pb/L (at pH 7.1 and hardness of 26 mg/L). The acute toxicity values (48h-LC50) for Ceriodaphnia dubia varied between of 26 µg dissolved Pb/L (at pH 8.1 and hardness of 25 mg/L) and 3116 µg dissolved Pb/L (at pH 8.15 and hardness of 103.2 mg/L). The acute toxicity values (48h-LC50) for Pseudokirchneriella subcapitata varied between of 21 µg dissolved Pb/L (at pH 7.5 and hardness of 24.2 mg/L) and 323 µg dissolved Pb/L (at pH 7.3 and hardness of 174.4 mg/L).

Chronic freshwater data:

A Species Sensitivity Distribution (SSD) has been developed for the assessment of lead in the freshwater compartment, using the reliable species-specific chronic toxicity effect levels that have been identified in the previous section of this report. No-effect levels for dissolved lead for fish were situated between 18.9 µg Pb/L (Oncorhynchus mykiss; endpoint: scoliosis) and 1558.6 µg Pb/L (Pimephales promelas; endpoint: dry weight). No-effect levels for dissolved lead for invertebrates were situated between 1.7 µg Pb/L (Lymnaea stagnalis; endpoint: growth) and 182 µg Pb/L (Ceriodaphnia dubia; endpoint: mortality).

All high quality chronic data retained for PNEC derivation were normalised for bioavailability. Indeed, reports/publications reporting toxicity data as total Pb concenrations were in first instance converted to dissolved Pb concentrations using the equation generated by Blust (2010). An additional step in the bioavailability correction was introduced to normalize the chronic freshwater toxicity data towards the dissolved organic carbon (DOC) using the Minteq 2.61 software.

As mentioned before, the normalised Pb toxicity data have been used for the construction of a Species Sensitivity Distribution (SSD) from which the median 5th percentile, i. e. the HC5-50 with 5%-95%-confidence interval, has been derived. Using the software package BestFit, a Log-Uniform Distribution was defined as the best fitting distribution. The HC5-50 for dissolved Pb associated with the reasonable worst case DOC concentrations of 2.6 mg/L was 19.6 µg/L for dissolved lead.

Application of an assessment factor between 1 and 5 on this HC5-50 will result in the final PNEC for lead in the freshwater environment. Based on the available data, the following points have to be considered when determining the size of the assessment factor:

- The chronic Pb database covers only ecologically relevant endpoints. The selected endpoints were mortality, growth, hatching, reproduction and abnormalities,

- The NOEC/L(E)C10 data were extracted from tests performed in a variety of natural/artificial freshwaters, covering a considerable part of the wide range of the freshwater characteristics (pH value and hardness) that are normally found in European freshwaters.

- The chronic Pb database covers sensitive life stages and ‘chronic’ exposure times for all trophic levels.

- 43 individual high quality NOEC/EC10 values and 7 different species NOEC values (fish, invertebrates and algae) were compiled from the database.

- The use of the best fitting Uniform distribution minimizes the statistical uncertainties around the HC5-50 value.

- Comparison between field and mesocosm studies and the HC5-50 suggest that the HC5 -50 is sufficiently protective. However, only a limited number of data are available.

- The high sensitivity of Lymnaea sp. is not expected as the occurrence of Lymnaea sp. in European freshwaters is very common. Therefore it has been decided not to retain the low NOEC value of 1.7 µg dissolved Pb/L in the database for PNEC derivation.

However in conclusion on the subject on the choice of the assessment factor and considering all arguments above including the particular cautious approach taken for derivation of HC5-50 it is felt that the most appropriate AF would be 3. Therefore, hence the reasonable worst case freshwater PNEC is proposed to be 6.5 µg dissolved Pb/L, which will be carried over to the risk characterization.

- Toxicity to marine organisms

A Species Sensitivity Distribution (SSD) has been developed for the assessment of lead in the marine compartment, using the reliable species-specific chronic toxicity effect levels that have been identified in the previous section of this report. The selected no-effect levels for dissolved lead for fish was 229.6 µg Pb/L (Cyprinodon variegatus; endpoint: dry weight. No-effect levels for dissolved lead for invertebrates were situated between 95.9 µg Pb/L (Neanthes arenaceodentata; endpoint:: juvenile emergence) and 757 µg Pb/L (Neanthes arenaceodentata; endpoint:: adult survival).

As mentioned before, the high quality dissolved Pb toxicity data have been used for the construction of a Species Sensitivity Distribution (SSD) from which the median 5th percentile, i. e. the HC5-50 with 5%-95%-confidence interval, has been derived. Using the software package BestFit, a Log-Normal Distribution was defined as the best fitting distribution. The HC5-50 for dissolved Pb was 10.2 µg/L.

Application of an assessment factor between 1 and 5 on this HC5-50 will result in the final PNEC for lead in the marine environment. Based on the available data, the following points have to be considered when determining the size of the assessment factor:

- The chronic Pb database covers only ecologically relevant endpoints. The selected endpoints were mortality, growth, emergence, reproduction and abnormalities,

- The NOEC/L(E)C10 data were extracted from tests performed in a variety of natural/artificial marine waters, covering a considerable part of the wide range of the marine water characteristics (pH value and salinity) that are normally found in European coastal waters.

- The chronic Pb database covers sensitive life stages and ‘chronic’ exposure times for all trophic levels.

- The database includes a broad representation of temperate marine organisms, including unicellular algae, invertebrates, and fish. However, no crustaceans are included in the database. Indeed, 24 individual high quality NOEC/EC10 values and 4 different species NOEC values were compiled from the database.

- All tests were performed in artificial test media which are considered to promote metal bioavailability (i. e., worst-case conditions) due to the absence of natural compounds in the water that can form metal complexes and, hence, reduce metal bioavailability and toxicity (e. g., dissolved organic matter);

- The use of the best fitting Log-Normal distribution minimizes the statistical uncertainties around the HC5-50 value.

- No data allowed the comparison between field and mesocosm studies and the HC5-50.

However in conclusion on the subject on the choice of the assessment factor and considering all arguments above including the particular cautious approach taken for derivation of HC5-50 it is felt that the most appropriate AF would be 3. Therefore, hence the reasonable worst case marine PNEC is proposed to be 3.4 µg dissolved Pb/L, which will be carried over to the risk characterization.

- Toxicity to micro-organisms in sewage treatment systems

The protozoan tests resulted in L(E)C50 values ranging from <1.5 mg/L (LC50 for different ciliate species) to >5 mg/L (LC50 for  Opercularia minima).

The lowest observed bounded EC50 value from tests with other ciliated protozoa tested in activated sludge mixed liquor medium is 2.29 mg Pb/L (for ciliate Vorticella convallaria). On the other hand, a (visually) derived LC10 for the sludge protozoan community in the activated sludge mixed liquor of 1 mg/L dissolved Pb is estimated. Preference was given to this community endpoint because it is more relevant.

Based on the above reported results and discussion at TCNES II 07 a preference was expressed for application of an assessment factor of 10. In this case, a PNEC for micro-organisms of 0.1 mg/L for dissolved lead in effluent is calculated and, at the request of TCNES, this is the value forwarded to risk characterisation. Note that there are insufficient useful data for aquatic micro-organisms to apply statistical extrapolation.