Aluminum sodium oxide

Short-term toxicity to fish

Additional information

No reliable studies on the acute toxicity of sodium aluminate to fish are available. The only available test conducted by Wörmann & Lamberty (1989) using a sodium aluminate product was considered not reliable due to the absence of information on test conditions. Therefore, a weight of evidence approach is applied and data are read-across to various aluminium compounds based on an analogue approach.

Two limit tests were conducted by NIVA (1996a, 1996b) using aluminium oxide and aluminium hydroxide (supernatants) as source chemicals for aluminium. Both tests were conducted according to GLP standards and OECD guideline 203 using Salmo truttaas test species. PH-values ranged between 5.9 and 6.7 in both studies. No mortalities or sublethal effects were observed and LL50s were determined to be >100 mg/L (nominal for Al₂O₃and Al(OH)₃) and LC50s > 74 µg/L and > 72 µg/L (dissolved Al for Al₂O₃and Al(OH)₃).

Aluminium chloride as source for aluminium was used in the studies of Parametrix (2009), ENSR (1992), Callet al.(1984), and Gundersenet al.(1994). The GLP study of Parametrix (2009) was conducted withPimephales promelasaccording to ASTM 2000 guideline. Nominal pH and water hardness were 8 and 50 mg/L, respectively. After dissolving the test item, a part of the aluminum solution was filtered using a 2.0 µM membrane filter. The fish was exposed to unfiltered and filtered test media containing nominal concentrations of 18.75, 35, 75, 150 and 300 mg Al/L for 96 h, separately. After exposure, the dissolved and the total aluminum in each test medium were determined by using AAS. Measured concentrations in the unfiltered test medium ranged from <30 to 218644 µg/L of total Al and from <30 to 561.2 µg/L of dissolved Al. In the filtered test medium measured concentrations ranged from <30 to 1949 µg/L of total Al and from <30 to 638.1 mg/L of dissolved Al. Concentrations of dissolved Al decreased as nominal concentrations of Al increased in both test systems. Significant effects on survival were observed at the three highest concentrations in the unfiltered media. No significant effects were observed by using the filtered test media. As mentioned above, concentrations of dissolved Al were within the same range in the filtered and unfiltered test media. Concluding from the results of that study it could be stated that the survival of fathead minnows was affected by the insoluble Al species and not by the dissolved Al as filtering of test media result in no acute effects of Al on the fish tested.

In another test by ENSR (1992)Pimephales promelaswere exposed at different water hardness: 25, 50, 100, and 200 mg/L (as CaCO₃) at pHs 7.6, 7.8, and 8.1, respectively. The study was conducted under GLP and according to USEPA 1985 guideline. The LC50-values determined based on measured values for total Al were 1.16, 8.18, 20.3, and 44.8 mg/L at water hardness of 25, 50, 100, and 200 mg/L (as CaCO³), respectively. The results demonstrate a decreasing toxicity of total Al to fish with increasing water hardness.

In series of non-GLP studies Callet al.(1984) examined the toxicity of aluminium chloride to five species of fish. Tests were conducted at a water hardness of 47.4 mg/L and at nominal pH values of 6.5, 7.5, 8.2 and 8.5. Tests withOncorhynchus mykiss(formerlySalmo gairdneri) resulted in LC50-values of 7.4 and 14.6 mg/L of total Al at nominal pH of 6.5 and 7.5, respectively. LC50-values determined in the test conducted withPimephales promelaswere >48.2 and >49.8 mg/L at pH 7.5 and 8.2, respectively. LC50-values determined for all other species, i.e.Ictalurus punctatus, Lepomis cyanellus, andPerca flavescens, were >50 mg/L at nominal pH 7.5. Although the studies did not following a certain protocol, based on the test conditions stated, the studies are considered to be reliable.

In a further non-GLP and non-guideline study Gundersenet al(1994) exposedOncorhynchus mykissto AlCl₃at four different water hardness concentrations (10, 30, 80 and 120 mg/L as CaCO3) and two pH values (7.3 and 8.3). Additional humic acid was added to some test series. In this test the different parameters, pH, water hardness and humic acid, had hardly an effect on the determined LC50 values. For total Al they were found within a range of 3.75 to 6.17 mg/L without distinct correlation to parameters. The same was found for dissolved Al were LC50-values were in the range of 0.36 to 0.79 mg/L. The study is considered to be reliable.

Aluminium sulphate was used in a non-GLP non-guideline study with Salmo salar by Roy & Campbell (1997). In this study fish were exposed to a nominal pH of 5 and some fish were additionally exposed to fulvic acid. Also in this study LC50-value in both experiments were in the same range of magnitude from 78.3 to 224.1 µg/L.