Registration Dossier

Ecotoxicological information

Endpoint summary

Administrative data

Description of key information

Additional information

Information taken from EU-RAR (2008) and SIAM 27 (2008): Nickel and Nickel compounds:

Freshwater

One generic goal of the nickel risk assessment is to express the chronic freshwater toxicity of Ni in the bioavailable form. Bioavailability models were subsequently used to normalize the ecotoxicity data to sets of standard physicochemical conditions for important abiotic factors (i.e., pH, hardness, and dissolved organic carbon (DOC)). This approach allows for the comparison of intrinsic toxicity among organisms on an equal basis.

For the aquatic compartment, Biotic Ligand Models (BLMs; cf. SIAR for Zn) were used to normalize the ecotoxicity data. BLMs were developed and validated for two invertebrates (Ceriodaphnia dubia and Daphnia magna), an alga (Pseudokirchneriella subcapitata), and a fish (Oncorhynchus mykiss). Appropriate use of this bioavailability normalization necessitates that the geochemical boundaries of the BLMs are defined relative to the environmental conditions considered. In general, the BLMs cover between 10 and 90% of the pH, hardness, and DOC observed in EU surface waters (Table 1). Only reliable ecotoxicity data from tests conducted within the boundaries of the BLMs have been used for establishment of the PNEC. Definition of the relevant environmental conditions and the exclusion of otherwise reliable ecotoxicity data relative to these conditions may need to be adapted for other regions.

Table 1: Ranges of pH, hardness and Mg used for data selection:

Test organism

pH range

Hardness range (mg/L CaCO3)

Mg (mg/L)

Algae – P. subcapitata

5.7 - 8.2

20 - 480*

1.4 - 113

Higher plants – H. vulgare

4.1 - 7.5

NA

1.0 - 94

Invertebrates – D. magna

5.9 - 8.2

6.320

0.4 - 72

Invertebrates – C. dubia

6.5 - 8.2

6.320

1.1 - 72

Fish – O. mykiss

5.4 - 8.5

20 - 310

1.1 - 72

*: Hardness range for algae was based on the range of Mg concentrations used in unilateral experiments conducted to determine the relationship between Mg and Ni toxicity. This basis was used because Mg was shown to be more important than Ca in the amelioration of Ni toxicity to P. subcapitata.

 

Effects data sets selected: More than 250 individual NOEC/EC10values were collected and screened for quality and relevancy, which yielded 193 individual high quality data covering 30 different species. The selected data set covers 16 different families, different trophic levels and feeding patterns and is so far the largest data set on a metal. It should be noted that some reliable aquatic ecotoxicity data that passed the relevancy criteria were set aside because they were obtained from tests in which the geochemical parameters were outside of the BLM boundaries. These otherwise high quality data were listed in separate tables in the EU-RAR (2008), some of which will be used in a special scenario to assess risk of Ni exposure in particular alkaline waters (i.e., pH = 8.3 – 9.0) that are outside of the BLM boundary.

For algae, EC10 values of Ni for chronic exposures conducted with Pseudokirchneriella subcapitata ranged from 25.3 to 425 μg Ni/L, with a median value of 88.2 μg Ni/L (n = 47). Chronic growth inhibition data (EC10) are available for nine additional freshwater algae species. These EC10 values range from 12.3 μg Ni/L for Scenedesmus accuminatus to 51.8 μg Ni/L for Coelastrum microporum. For higher aquatic plants, chronic effects to Lemna gibba and Lemna minor ranged between 8.2 and 80 μg Ni/L.

Chronic nickel toxicity data are available for fifteen invertebrate species. The large majority of data are from crustaceans, but data from insects, hydrozoans, and molluscs are also available. The NOEC/L(E)C10 varied between 2.8 μg/L for Ceriodaphnia dubia and 1193.3 μg/L for Chironomus tentans.

Chronic nickel toxicity data are available for three species of fish, with NOEC/LC10 values ranging from 40 μg Ni/L for Brachydanio rerio to 1,100 μg Ni/L for Oncorhynchus mykiss. NOEC/L(E)C10 data are available for three species of amphibians, with values ranging from 84.5 μg Ni/L to 13,147 μg Ni/L, both values from Xenopus laevis.

In summary, NOEC/L(E)C10 values for chronic nickel toxicity to aquatic organisms range from 2.8 μg Ni/L (C. dubia) to 13,147 μg Ni/L (X. laevis).

Bioavailability correction: Many of the aquatic toxicity data were obtained from experiments designed to determine effects of water quality parameters on nickel toxicity. Factors identified to affect nickel toxicity included pH, hardness, and dissolved organic carbon (DOC). The use of the BLM performs two principal functions. First, it removes the influence of variable geochemical conditions when calculating species mean values from individual toxicity tests, which may have been performed using different combinations of pH, hardness, and DOC.

Geochemical normalization is accomplished by the chemical speciation software within the BLM (WHAM VI, in the case of the nickel BLM), and ensures that all data within a given Species Sensitivity Distribution (SSD) are evaluated on an equivalent free nickel ion basis. Second, the BLM takes into account the competitive effects of positively-charged constituents of freshwater, such as Ca2+, Mg2+, and H+, on the uptake of Ni2+ at the assumed site of action on the organisms.

Biotic Ligand Models have been developed for the three standard trophic levels for bioavailability correction covering the algae Pseudokirchneriella subcapitata, the invertebrate Daphnia magna, and the fish Oncorhynchus mykiss. An additional BLM was developed for Ceriodaphnia dubia because intraspecies variability for this particular sensitive species could not be explained by the D. magna BLM. Recalibrating the speciation model to accurately estimate nickel speciation at the low nickel concentrations (e.g., < 5 μg Ni/L) that are relevant to C. dubia was required for the development of an accurate C. dubia BLM. Effects of DOC were evaluated in toxicity tests using natural waters that represented ranges of natural DOC type (e.g., streams and ponds) and quantity (low to high DOC concentration). Therefore, the effect of varying DOC quality on Ni toxicity was implicitly addressed in the BLM development. An additional BLM was developed under hydroponic conditions for the plant Hordeum vulgaris, and this may be useful for normalizing nickel toxicity to aquatic vascular plants. Because the intra- and interspecies variability that are present among the data are largely influenced by the water quality parameters that were used in the toxicity tests, it is critical for the PNEC derivation to include a step that normalizes the NOECs/L(E)C10 values to a set of standard water quality parameters, e.g., pH, hardness, and DOC. All individual toxicity data were normalized using BLMs. In most cases, BLMs from taxonomically similar group of organisms were used to normalize the NOEC/L(E)C10 data, i.e., all fish and amphibian toxicity data were normalized by the O. mykiss BLM, invertebrate data were normalized by the more stringent of the either the C. dubia or D. magna BLM, and algae were normalized by the more stringent of either the P. subcapitata or H. vulgaris BLM. For certain groups of organisms with no BLM, the most stringent BLM was used even if it was not obvious that this BLM originated from the taxonomically most similar organism. For example, L(E)C10 data for Lemna minor, a vascular plant, were normalized using the D. magna BLM because the D. magna BLM was shown to result in the most cautious predictions. Support for the cross-species extrapolation was provided by examples from the literature for phytoplankton and fish, and from a spot-check study for invertebrates and vascular plants. The spot check study was performed on three non-crustacean invertebrates (the midge larvae Chironomus tentans, the rotifer Brachionus calicyflorus, and the snail Lymnaea stagnalis) and one higher plant (duck weed, L. minor). The results of the spot check study indicated that the available BLMs were capable of predicting toxicity to the test species used in the spot testing exercise, within a factor of 2. It was based on this concluded that “full cross-species extrapolation” as described above was the preferred approach which was therefore taken forward as the basis for normalizing the SSD according to the abiotic factors hardness, pH and DOC concentration in freshwater.

Summary of the “species mean” NOEC or EC10 values (total risk approach) in μg Ni/L (with most sensitive endpoint): 

Taxonomic group

species

Most sensitive endpoint

Species mean NOEC/EC10 value (µg Ni/L)

Algae

Scenedsemus accuminatus

Growth rate

12.3

Desmodesmus spinosus

Growth rate

22.5

Pediastrum duplex

Growth rate

23.8

Chlamydomonas sp

Growth rate

27.9

Ankistodesmus falcatus

Growth rate

28.4

Chlorella sp

Growth rate

42.0

Coelastrum microporum

Growth rate

46.2

Pseudokirchneriella subcapitata

Growth rate

92.7

Higher plants

Lemna minor

Growth

27.9

Lemna gibba

Growth rate

50.0

Rotifer

Brachionus calyciflorus

Intrinsic rate of growth

633.2

Molluscs

Lymnea stagnalis

Growth

6.8

 

Juga plicifera

mortality

124.0

Cladocerans

Ceriodaphnia dubia

Reproduction

6.9

Ceriodaphnia quadragula

Mortality

7.4

Peracantha truncata

Reproduction

8.0

Simocephalus vetulus

Reproduction & mortality

16.3

Ceriodaphnia pulchella

Reproduction & mortality

16.7

Alona affinis

Mortality

25.0

Daphnia longispina

Mortality

27.8

Daphnia magna

Reproduction

35.6

Insects

Clistoronia magnifica

Mortality

66.0

Chironomus tentans

Biomass

458.9

Hydrozoans

Hydra littoralis

Growth

60.0

Amphipods

Hyalella azteca

Mortality

29.0

Fish

Brachydanio rerio

Hatchability

40.0

Pimephales promelas

Growth

57.0

Oncorhynchus mykiss

Growth

134.0

Amphibians

Xenopus laevis

Malformation

171.6

 

Gastrophryne carolensis

Mortality

184.9

 

Bufo terrestris

Growth

640.0

 

Marine

Effect data sets: The marine chronic ecotoxicity database is represented by 15 species of marine organisms from 14 families, and includes a wide range of taxonomic groups, including unicellular algae, macroalgae, crustaceans, molluscs, echinoderms, and fish.

EC10 values for four species of marine algae are reported, ranging from 97 μg Ni/L for growth of giant kelp (Macrocystis pyrifera) to 17891 μg Ni/L for growth of the dinoflagellate, Dunaliella tertiolecta.

EC10 values are reported for nine species of marine invertebrates , ranging from 22.5 μg Ni/L for reproduction of the polychaete, Neanthes arenaceodentata, to 335 μg Ni/L for development of the echinoderm, Strongylocentrotus purpuratus. (The data for Diadema antellarum were excluded because of disease).

EC10 values are reported for two species of marine fish, ranging from 3599 μg Ni/L for growth of the topsmelt, Atherinops affinis, to 20760 μg Ni/L for growth of the sheepshead minnow, Cyprinodon variegatus.

In summary, the chronic EC10 data used in the derivation of the HC5 (50%) for the marine compartment ranged from 22.5 μg Ni/L for Neanthes arenaceodentata to 20,760 μg Ni/L for Cyprinodon variegates.

Bioavailability: No incorporation of the potential modifying effect of abiotic factors was included in this analysis. Most of the factors known to affect Ni bioavailability and toxicity in freshwater, e.g., Ca2+, Mg2+, and H+, are relatively constant in marine waters, and therefore the utility of bioavailability correction is limited. Dissolved organic carbon (DOC) can vary considerably in coastal marine waters, and has been shown to be an important factor controlling the toxicity of nickel in the freshwater environment. Therefore, DOC may also be important in the marine environment. However, the range of DOC concentrations present in the natural seawaters tested as part of this program (0.22 - 2.7 mg/L) was probably too narrow with the obtained precision and accuracy of the ECx results to accurately quantify any variation of effect concentrations attributable to DOC.