Registration Dossier

Ecotoxicological information

Endpoint summary

Administrative data

Description of key information

Additional information

There are several excellent reviews on the aquatic toxicity of copper, e.g. U.S. EPA (2007) Update, Ambient water qualitiy criteria Copper; WHO (1998) EHC 200 Copper; or ECI (2008) European Union Risk Assessment Report, Voluntary Risk Assessment of copper, Copper(II)sulphate pentahydrate, Copper(I)oxide, Copper(II)oxide, Dicopper chloride trihydroxyde.

“Copper toxicity varies markedly due to various physicochemical characteristics of the exposure water, including temperature, dissolved organic compounds, suspended particles, pH, and various inorganic cations and anions, including those composing hardness and alkalinity. Many of these physicochemical factors affect copper speciation, and their effects on copper toxicity therefore could be due to effects on copper bioavailability. That bioavailability is an important factor is evident from uptake of copper by aquatic organisms being reduced by various organic compounds and inorganic ligands known to complex copper.” (U.S. EPA 2007)

Due to the various physico-chemical factors influencing copper- availability/-toxicity, the assessment of copper toxicity represents a highly complex process. For instance it is generally acknowledged, that copper toxicity decreases with increasing water hardness, increasing pH and high DOC concentrations (WHO 1998). There are numerous amounts of studies available on copper toxicity performed under various environmental conditions. In order to reach a comparability of these studies it is necessary to adjust these onto certain physicochemical standard conditions. U.S. EPA (2007) and ECI (2008) provided such an approach, adjusting a peer reviewed collection of toxicity data (acute and/or chronic values) to “reference exposure conditions”. Therefore, for further informations on copper toxicity, these two assessments are by thereby recommended.

Reference: European Copper Institute, ECI (2008). Voluntary Risk Assessment Report on Copper and its compounds.

U.S. EPA (2007) Update, Ambient water qualitiy criteria Copper.

The following assessment refers the European Risk Assessment performed by the European Copper Institute (2008):

Aquatic effects

Information on the mode of action of copper exposure indicated that the target tissue for copper toxicity were the water/organism interface with cell wall and gill-like surfaces acting as target biotic ligands in all species investigated.


For the freshwater pelagic compartment, 139 individual NOEC/EC10 values resulting in 27 different species-specific NOEC values, covering different trophic levels (fish, invertebrates and algae) were used for the PNEC derivation. The large intra-species variabilities in the reported single species NOECs were related to the influence of test media characteristics (e.g., pH, dissolved organic carbon, hardness) on the bioavailability and thus toxicity of copper. Species-specific NOECs were therefore calculated after normalizing the NOECs towards a series of realistic environmental conditions in(typical EU scenario’s, with welldefined pH, hardness and DOC). Such normalization was done by using chronic copper bioavailability models (Biotic Ligand Models), developed and validated for three taxonomic groups (fish, invertebrates and algae) and additional demonstration of the applicability of the models to a range of other species. The species-specific BLM-normalized NOECs were used for the derivation of log-normal Species Sensitivity Distributions (SSD) and HC5-50 values (the median fifth percentile of the SSD), using statistical extrapolation methods. The HC5-50 values of the typical EU scenarios ranged between 7.8 to 22.1 μg Cu/L. Additional BLM scenario calculations for a wide range of surface waters across Europe further demonstrated that the HC5-50 of 7.8 μg Cu/L, is protective for 90% of the EU surface waters and can thus be considered as a reasonable worst case for Europe in a generic context.

Copper threshold values were also derived for three high quality mesocosm studies, representing lentic and lotic systems. The mesocosm studies included the assessment of direct and indirect effects to large variety of taxonomic group and integrate potential effects from uptake from water as well as from food. BLM-calculated HC5-50 values (Assessment Factor (AF)=1) were used as PNEC for the risk characterisation. The HC5-50 (AF=1) of 7.8 μg Cu/l was used as reasonable worst case PNEC

forin a generic context in absence of site-specific information on bioavailability parameters (pH, DOC, hardness). The AF=1 was chosen in relation to the uncertainty considerations covering 1) the mechanism of action; 2) the overall evaluation of the database; 3) the robustness of the HC5-50 values; 4) corrections for bioavailability (reducing uncertainty); 5) the sensitivity analysis with regards to DOC and read-across assumptions; 6) the factor of conservatism “built in into” the data and assessment (such as no acclimation of the test organisms and no pre-equilibration of test media); 7) results from multi-species mesocosm studies and 8) comparison with natural backgrounds and optimal concentration ranges for copper, an essential metal.


For the STP compartment, high-quality NOECs from respiration or nitrification inhibition studies, relevant to the functioning of a Sewage Treatment Plant (STP), resulted from biodegradation/removal studies and NOECs for ciliated protozoa were used to derive the PNEC for STP micro-organisms. The lowest reliable observed NOEC value was noted for the inhibition of respiration (0.23 mg/l expressed as dissolved copper) and carried forward as PNEC to the risk characterisation.


For the marine PNEC derivation, 51 high-quality chronic NOEC/EC10 values, resulting in 24 different species-specific NOEC values covering different trophic levels (fish, invertebrates, algae), were retained for the PNEC derivation. NOEC values were related to the DOC concentrations of the marine test media. Species-specific NOECs were therefore calculated after DOC normalizing of the NOECs. These species-specific NOECs were used for the derivation of species sensitivity distributions (SSD) and HC5-50 values, using statistical extrapolation methods. Considering that the log-normal distribution had a poor data fit according to goodness of fit tests, HC5-50 values, obtained by using the best-fitting parametric distribution, were considered for the PNEC derivation.


The organic carbon normalisation was carried out at a DOC level typical for coastal areas (2 mg/l) and resulted in an HC5-50 value of 5.2 μg Cu/L. Additionally a semi-parametric statistical analysis of the NOECs distribution was performed and a HC5-50 derived. Some Member States preferred this option of derivation of HC5-50. Because the difference between the HC5-50 by either approach were similar, the HC5-50 derived by the parametric curve fitting (using the fit function that fitted best) was used. The evaluations of lower-quality NOECs and EC50s from single species and multi-species marine studies added weight to the HC5-50 value derived from the best-fitting distribution. In the absence of a high-quality mesocosm, an AF of 2 has been applied on the HC5-50 and a marine PNEC of 2.6 μg Cu/L is carried forward to the risk characterization.


Reference: European Copper Institute, ECI (2008). Voluntary Risk Assessment Report on Copper and its compounds.