Nickel – ion. Aquatic effects DRAFT of May 2002
Danish Environmental Protection Agency
Aquatic effect assessment for nickel
Background report on the nickel - ion
Draft
May 2002
This draft is the background document for all aquatic effect data on nickel compounds (including nickel substances not on the EU priority lists)
on which data were available.
Reports on the EU priority nickel substances will refer to this document.
Ole Christian Hansen, consultant
Danish Technological Institute, Environment
DK Rapporteur:
Danish Environmental Protection Agency
Strandgade 29
DK-1401 Copenhagen K
Denmark
+45 32 66 01 00
Henrik Tyle ()
Henrik Søren Larsen ()
Consultant:
Ole Christian Hansen ()
Danish Technological Institute
Gregersensvej
Postboks 141
DK-2630 Taastrup
Telefon: +45 72 20 20 00
Contents
3.2Effect assessment
3.2.1Aquatic compartment
3.2.2Data sources and selection
3.2.2.1Relevance
3.2.2.2Selection of data
3.2.2.3Evaluation of data
3.2.3General introduction
3.2.3.1Bioavailability
3.2.3.2Chemical speciation and complexation with DOC
3.2.3.3Chelation
3.2.3.4Abiotic factors influencing the toxicity
3.2.3.4.1Water hardness
3.2.3.4.2pH
3.2.3.5Adaptation or Tolerance
3.2.4Acute toxicity
3.2.4.1Acute toxicity to fish
3.2.4.2Acute toxicity to crustaceans
3.2.4.3Acute toxicity to algae
3.2.4.4Acute toxicity to molluscs.
3.2.4.5Acute toxicity to rotifers
3.2.4.6Conclusion on acute toxicity
3.2.5Chronic toxicity
3.2.5.1Chronic toxicity to fish
3.2.5.2Chronic toxicity to crustaceans
3.2.5.3Chronic toxicity to algae
3.2.5.4Chronic toxicity to other organisms
3.2.6PNEC water
3.2.6.1Derivation of PNECaqua using assessment factor
3.2.6.2Derivation of PNECvalues using statistical extrapolation (methods)
3.2.7Effects to microorganisms (relating to waste water treatment plants)
3.3Bioaccumulation (preliminary)
Appendix 1: References
Appendix 2 Physico-chemical properties
Appendix 3OECD TG recommended test media
Appendix 4Water quality standards
Appendix 5Aquatic ecotoxicity – acute, freshwater
Appendix 6Chronic toxicity to aquatic organisms
3.3.1Toxicity to freshwater organisms
3.3.2Marin toxicity _
3.3.3Microorganisms (STP)
Appendix 7 Measured aqueous nickel concentrations
Introduction
This draft is the background document for all aquatic effect data on nickel compounds (including also nickel substances not on the EU priority lists) on which data were available (mainly soluble nickel salts).
In the report on aquatic effects, the draft report is focused on the nickel ion. The available effect data on each substance already contained in the background report will be presented later in the equivalent chapters in the reports on the individual nickel compounds. As regards the evaluation of each substance in the individual substance reports, a brief justification will be made on why the content of the ion report applies and a reference to this evaluation will be given.
3.2Effect assessment
3.2.1Aquatic compartment
In general, the revised version of the TGD (draft version Oct. 2001) has been used when evaluating the aquatic toxicity of nickel. The evaluation has benefitted from the general experience gained by the extensive discussion at TMs on the aquatic effects assessment of zinc, cadmium and chromium.
3.2.2Data sources and selection
The available data have been listed in tables in Annex 4. The sources have been data from industry, governmental agencies (HEDSET, IUCLID, ECDIN, AQUIRE), databases (HSBD, Toxline, CESAR), reviews (to check that most available references were included) and open literature seaches. A large database on aquatic and terrestrial toxicity references prepared by industry (NiPERA 2001) has been included in the assessment. The data search by rapporteur and industry was in agreement and generally covering the same references on aquatic data.
3.2.2.1Relevance
In most references, the toxicity data are expressed as total nickel concentrations and not as the concentration of the nickel test compound (e.g. concentration of nickel salt) (However, it is indicated in the appendix on summarised results where normalisation has been performed). Individual studies may have indicated that differences were apparent between nickel species e.g. nickel chloride and nickel sulfate. For instance, nickel chloride was significantly more toxic than nickel sulfate to fish in an acute test by Muramoto (1983) and Janssen Pharmaceutica (1993)). In the latter reference, stoechiometric recalculation of toxicity relative to the Ni-ion indicated the toxicity of the Ni-ion to be of equivalent level. However, since not enough comparable studies were available and none of these were chronic data, the nickel ion has been considered to be the primary causative factor for long-term toxicity.
It is realised that the free nickel ion and other dissolved nickel species are more relevant for bioavailability and toxicity and a better indicator of toxicity than total nickel. Most studies, however, were performed without experimental measuring of the free nickel-ion. Thus, it has not been possible to use measured Ni2+-ion concentrations as the basis for the aquatic effect assessment. However, this may not be a severe limitation of this evaluation. For most studies the pH in the test solutions of the experiments range between 6.5 to 8. In the few studies where the Ni-ion has been estimated (e.g. Chen et al. 1997), nickel was calculated mainly to exist as the free nickel ion. Formation of nickel complexes does not appear to be significant for nickel speciation or bioavailability in laboratory studies and studies using filtered water where the dissolved organic matter (DOM) and dissolved organic carbon (DOC) are typically low. The dissolved fraction of a substance in water is normally defined as the fraction that passes a 0.45 µm membrane filter. Furthermore, in most studies, using well water or natural surface waters, the water has been filtered to avoid particulates. The test stem solutions are usually filtered as well.
The results of aquatic effect studies are given as total nickel and include both measured and nominal concentrations. In studies where both values were present the results indicated the difference between nominal and measured total nickel to be acceptable for using the nominal concentrations in the effect calculations (<10%). It should be noted that the use of reconstituted water and medium according to OECD TG´s includes no or very low levels of nickel.
3.2.2.2Selection of data
The data that have been used to estimate the aquatic nickel toxicity was derived from studies using soluble inorganic nickel salts as test compounds. The endpoints considered relevant were survival, growth and reproduction.
Acute data were included, as they appear to be the only available data demonstrating effect of abiotic factors such as pH, water hardness, temperature, etc.
Chronic tests were not available in sufficient amount to determine the effects of abiotic factors as pH, water hardness, temperature, complexation with organic ligands or from other parameters that might influence in the toxicity.
All aquatic data are expressed as the total concentration of the nickel ion, and not as the test compounds because the free nickel ion is considered the primary cause for toxicity. The available information was too scarse to allow for a normalisation of the total nickel concentration to the free nickel ion concentration. For aquatic organisms, especially the nickel ion is thought to be relevant for toxicity. It is, however, not fully known to which extent other nickel species may contribute to the observed toxicity. The dissolved nickel concentration in water may be a better indicator of toxicity than the total nickel concentration. However, also the dissolved fraction may contain forms of nickel that are not bioavailable or only bioavailable to a limited extent. The dissolved fraction is defined as the fraction that passes a 0.45 µm filter. The fraction includes a series of Ni-species such as Ni-ion, labile and stable inorganic and organic nickel complexes such as Ni-hydroxides, and Ni-humic and fulvic acid complexes. The final nickel-speciation depends on the water characteristics, especially the pH.
In the aquatic toxicity studies, the results were usually expressed as total nickel at nominal concentration or measured concentration. The latter usually measured by Atomic Absorption Spectrometry (AAS). It is presented in the toxicity tables whether the results are verified by measuring or calculated as nominal concentrations (m/n, respectively, in the table). It should be noted that the effect results are regarded as being dissolved nickel concentrations even if particulate fractions are included, because under laboratory conditions it was found that the main part of nickel present in the test medium was in the dissolved fraction. This was confirmed in a few studies where measurements of the dissolved nickel fraction indicated that the main fraction consisted of free nickel-ion (>80%) (cf. section on speciation).
3.2.2.3Evaluation of data
The aquatic toxicity data have been evaluated for reliability according to the following general rules
Preferred studies are studies performed according to internationally accepted guidelines such as OECD technical guidelines for the testing of chemicals. Other tests performed according to specific guidelines (standardised tests) are also considered acceptable if reference to available methods are used (e.g. ASTM, AFNOR, DIN, ISO, etc) or a sufficient description of method and the relevant variables are included in the study description (non-standardised tests).
The toxicity tests have been assigned reliability indexes (RI) modified according to the Belgian RAR on cadmium.
RI-1 / Reliable and standardised. Only standardised tests, e.g. OECD TG and accordingly.RI-2 / Reliable but non-standardised. Non-standardised tests of high quality and reliability, i.e. sufficient information available to evaluate the test including information on test medium (pH, water hardness), test circumstances (static, semi-static, flow-through, temperature), measured concentrations or indication of test nominal concentration close to measured concentration, test concentrations used, information on statistical analysis of result.
RI-3 / Not reliable. Not a standard test and insuffient information to evaluate the result.
RI-4 / Unknown reliability. Insufficient information available in description of test methodology.
In the assessment both RI-1 and RI-2 was used for the derivation of a PNEC for the aquatic environment.
For the selection of data the following approach has been taken:
Toxicological endpoints, which may affect the population level, are taken into account. In general, these endpoints are survival, growth and reproduction.
For acute tests, the results are expressed as acute LC50 or EC50 followed by the exposure period (usually 48 to 96 hours). Acute NOECs may be mentioned if available.
Long- term studies where the exposure period is longer than what is required for acute tests cover an intermediate position to chronic studies if they do not cover early life stages or a reproduction period. In relation to assigning studies to acute or chronic studies, they are considered acute long-term tests and not considered usable as chronic tests.
For assignment of a study as a chronic test, besides exposure time, considerations of the life stage and generation time of the exposed organism are considered essential. The exposure should preferentially cover at least one generation, a full life-cycle test or even more generations (the latter e.g. algae tests). NOEC values from early life stage tests (ELS) where data may indicate this to be a more sensitive life stage the data are included even if the exposure time may be shorter.
The species chronic NOEC is estimated as follows:
- If several data (chronic NOEC values) based on the same species, similar duration and toxicological endpoint are available, then for 3 data points the lowest NOEC is used, for 4 NOECs the geometric mean resulting in a “species mean” value has been calculated.
- If for one species several chronic values based on different toxicological endpoints/life stages are available, the lowest value is selected.
- Only results from tests where organisms were exposed to nickel or nickel mineral salts alone are used. Thus, tests on nickel in mixture with other metals have been excluded.
- Only tests with soluble nickel salts are included.
- Tests where no effects where observed at the highest exposure concentration used in the test, i.e. unbounded NOEC values indicated by NOEC xx, are not used. The values observed are considered informative but not usable in calculations.
Derivation of NOEC values (methods)
The methods that have been used for the derivation of NOEC values, being “real” NOEC values or NOEC values derived from effect concentrations, are those outlined in the TGD (Chapter 3 – Table 13).
If possible, “real” NOEC values were derived from the data reported, i.e. the NOEC is one of the concentrations actually used in the test. In order of preference:
- Statistical analysis: the NOEC is the highest concentration (in a series of test concentrations) at which the measured parameter shows no statistical significant effect (inhibition) compared to the control. Significance level: p = 0.05 (optional: the p = 0.01 level if reported instead of the p = 0.05 level).
- If no statistical analysis has been applied: the NOEC is the highest concentration that results in 10% inhibition compared to the control.
In both cases there must be a consistent concentration-effect relationship, i.e the LOEC is the concentration at which and above which statistical significant toxicity is found (1) or, when no statistical analysis has been applied (2), >10% inhibition is found.
If the “real” NOEC could not be derived from the data reported, the following procedure was used to derive the NOEC. In order of preference:
3)The NOEC is set at the EC10 level.
The EC10, which is calculated from the concentration-effect relationship, is used as NOEC equivalent, unless the “real” NOEC was also reported or could be derived from the data reported.
4)The NOEC is derived from the LOEC
If the EC10 was not reported and could not be calculated, the NOEC was derived from the LOEC using the following “extrapolation” factors:
a) NOEC = LOEC/2, in case inhibition is >10% but 20%.
b) NOEC = LOEC/3, in case inhibition is >20% but 30% e.g. LOEC = EC(25%).
If the percentage inhibition at the LOEC is >30% or in case the percentage inhibition at the LOEC is unknown, no NOEC is derived.
With respect to “rule 4b” it is noted that the TGD does not mention the derivation of a NOEC from a LOEC in case inhibition at the LOEC is >20%, while in this RAR the derivation of a NOEC from a LOEC up to 30% effect has been used in some (one?) aquatic toxicity studies. The use of the higher effect level is justified by the use of a higher extrapolation factor.
3.2.3General introduction
A great amount of information is available on the toxicity of nickel to aquatic organisms. In this section, a compilation is made of different studies, which provide data of nickel toxicity to different species. The tests were performed in laboratory conditions where nickel as soluble nickel salts were added to the exposure test solutions. Nickel is a transition metal that occurs in a number of oxidation states of which Ni2+ is the most common because of its stability over a wide range of pH and pE. Nickel can be present in the aquatic environment as inorganic species, including the free Ni2+-ion species. It is assumed that it is the free Ni-ion that has the main toxic effect in the studies.
The effects of metal and metal salts on aquatic animal species have been shown to correlate with the concentration of its free ion in the medium. The free ion concentration depends on the solubility of the metal (e.g. surface area cf. section 1) and the metal compounds’ ability to form complexation of the ion with organic and inorganic ligands. Thus the effect of nickel in the toxicity tests are affected by several factors such as the nickel compound used (NiCl2, NiSO4, Ni(NO3)2, etc.), the medium used, pH, water hardness, temperature etc. i.e. all factors that influence the solubility.
One hypothesis of the toxic action of nickel is that nickel toxicity is caused by its ability to replace essential metals in the metallo-enzymes, which results in the disruption of metabolic pathways (Nieboer and Richards 1980).
A considerable amount of data on the toxicity to aquatic species exists. However, due to the difference in the way the tests were performed, the medium etc (cf. above), the results are difficult to compare even when they are normalised to the total metal concentration (e.g. mg Ni/l). If measured concentrations are used to calculate the effect level, the measured element is usually the total-Ni without consideration of bioavailability, speciation etc.
The effect of abiotic factors on the availability of nickel is generally studied by observing the effect on toxicity. The next section presents some examples on such observations.
3.2.3.1Bioavailability
The bioavailability has been defined based on the available amount free metal ions present in the solution, the amount that is in reality taken up or the amount that has the potential to cause an effect. Thus, bioavailability is the amount of metal that is available for sorption or uptake by an organism in a given environment (Plette et al. 1999).
Chemical speciation and transport phenomena may affect bioavailability and subsequent toxicity of nickel. Several studies have been conducted to determine the forms of nickel that occur in aquatic systems (cf. Birge and Black 1980). The free divalent cation appears to predominate, except for waters of high pH in which nickel may form carbonates, oxides and hydroxides with very low water solubilities (cf. appendix 1).
The bioavailability is influenced by several factors and processes. Below is the variables, speciation, sorption, chelation etc. and abiotic factors such as pH, water hardness etc. that may affect these processes.
3.2.3.2Chemical speciation and complexation with DOC
Richter and Theis (1980) assume that the free divalent ion, Ni2+, is the predominant oxidation state in natural waters and the free ion dominates at the neutral pH range found in most aerobic waters (>90% in unsaturated solutions of natural waters at pH 5 to 9). However, complexes of naturally occurring ligands are formed to a small degree (OH- > SO42-, > Cl- > NH3). Under anaerobic conditions, sulfide if present will control the solubility of nickel (Richter and Theis 1980).
Data on nickel speciation in freshwater samples are limited. However, element speciation and concentration of free ion can be established with speciation models with full consideration of mole balances, relevant thermodynamic equilibrium constants, ionic strength, pH and concentration of the chelating agent introduced into the medium (Models MINTEQ Brown and Allison 1987). It should be noted that because the knowledge of modeling and chemical speciation of nickel has increased within the last years, “older” studies on the speciation of nickel should be used with caution.
Complexation of metals to organic and inorganic ligands in test media and natural environments can be estimated from metal speciation models. Speciation models for metals, including pH, hardness, DOC, and inorganic substances such as MINTEQ (Brown and Allison, 1987), WHAM (Tipping, 1994) and CHESS (Santore and Driscoll, 1995) can be used to calculate the uncomplexed and complexed fractions of the metal ions. Alternatively, the Biotic Ligand Model (BLM) allows for the calculation of the concentration of metal ion responsible for the toxic effect at the level of the organism. The BLM model has currently only been validated for a limited number of metals, organisms, and end-points (Santore et al. 2001). The models and formulae used for the characterisation of metal complexation in the media should always be clearly reported, allowing for their translation back to natural environments (OECD 2000).