Ammonium hydrogendifluoride

Short-term toxicity to fish

Description of key information

Data are availabe for AMBI, other soluble fluoride compounds and ammonium compounds.  The acute LC50 of AMBI in fish is reported to be 422 mg/L; the value is consistent with data for other fluoride compounds,indicating that the acute toxicity of the substance to fish is attributable to fluoride.  Data for ammonia and ammonium conpounds report very low LC50 values when expressed as non-ionised ammonia (i.e. NH3) but much higher values when expressed as total ammonia (i.e. NH3 and NH4+).  AMBI will be ionised under aqueous conditions to form ammonium ions (NH4+) with very little non-ionised ammonia (NH3) present, therefore LC50 values reported for NH£ are not relevant.

Key value for chemical safety assessment

Fresh water fish

Fresh water fish

Effect concentration:

422 mg/L

Additional information

The acute LC50 of the substance (AMBI) in fish is reported to be 422 mg/L (Hamburger & Hodgeschwender, 1987); this data is consistent the with results of other studies with fluoride compounds.

A number of further studies have investigated the acute toxicity of the water-soluble substance sodium fluoride. LC50 values of 107.5, 92.4, 118.5, 105.1, 64.1 ppm at 96, 120, 144, 168 and 192 h respectively are reported for rainbow trout (Camargo & Tarazona, 1991). The same authors report LC50 value for brown trout of 164.5, 135.6, 118.5, 105.1 and 97.5 ppm after 96, 120, 144, 168 and 192 h respectively. The EU RAR for hydrogen fluoride reports additional LC50 values of 299 mg/L (48h in Leuciscus idus); 51 mg/L in (96h in Onchorynkus mykiss) and 340 mg/L (96h in Gasterosteus aculeatus). The RIVM Integrated Criteria Document reports additional data, with LC50 values ranging from 128 -460 mg/L (Sloof et al,1988). The RIVM document also reports LC50 values of >100 mg/L (the limit of solubility) for sodium fluoride in seawater.

The acute toxicity of ammonia to fish has also been well investigated. A number of studies are available for aqueous ammonia and the read-across substance, ammonium chloride. In aqueous environments, aqueous ammonia (ammonium hydroxide) will exist predominantly as the ammonium (NH4+) ion. Read-across to other water-soluble inorganic ammonium salts such as ammonium chloride is therefore justified. The results of the studies demonstrate that ionised ammonia (NH4+) is of much lower toxicity than the non-ionised form (NH3). At environmentally relevant pHs, however, the substance will exisit predominantly as the ionised form (NH4 +); the ammonium ion. The toxicity of ammonia is therefore dependent on pH, as pH will influence the proportion of total ammonia prsent as the non-ionised form. The toxicity of ammonia was investigated in the Fathead minnow (Pimephales promelas), using ammonium chloride (Thurston et al, 1983). The 96-h LC50 ranged from 0.75-3.4 mg/l unionized ammonia (34-109 mg/l total ammonia-N). The toxicity of ammonia decreased as temperature increased from 12-22⁰C. There was no significant relationship between ammonia toxicity and dissolved oxygen concentration, over the range of the latter from 3-9 mg/l. Toxicity was not related to the size or source of test fish. Thurstonet al (1981) exposed S. gairdneri rainbow trout) and S. clarki (cutthroat trout) to ammonium chloride for 96 hours. Concentration levels were either fixed or fluctuating. Fish were more tolerant of a fixed concentration than they were of fluctuating concentrations. The larger rainbow trout (> 2 kg) were more vulnerable to acutely toxic concentrations of ammonia than were smaller fish (20-300 g). The 96 hour LC50 for fixed ammonia concentrations in rainbow trout was 0.163 -0.500 mg unionised NH₃/L (equivalent to 21.6 to 31.6 mg total ammonia-N/L). The 96 hour LC50 for fixed ammonia concentrations in cutthroat trout was 0.296 -0.327 mg unionised NH₃/L (equivalent to 26.3 to 29.1 mg total ammonia-N/L). Under fluctuating conditions, LC50 values for non-ionised ammonia of 0.099 -0.292 g/L in rainbow trout and 0.194 -0.217 mg/L in cutthroat trout were measured; equivalent total ammonia concentrations are 10.5 -22.3 mg/L in rainbow trout, 17.3 -19.3 mg/L in cutthroat trout. The toxicity of ammonia was assessed in different developmental stages of rainbow trout (Salmo gairdneri(Oncorhynchus mykiss)) under flow-through conditions, with exposure to ammonium chloride (Calamariet al, 1981). The 96 - hour LC50 of ammonia in flow through conditions to rainbow trout at development stages from egg to hatch was > 0.486 mg unionised NH3/L, in fry stages was 0.160 - 0.370 mg unionised NH3/L, and in fingerling stages was 0.440 mg unionised NH3/L. The acute toxicity of ammonium chloride was investigated inSalmo namaycushandSalmo salar(Soderberg & Meade, 1992). The effects of calcium (present as CaCl2) and sodium (present as NaCl) on the toxicity of ammonia were also investigated in this study. Calcium did not reduce ammonia toxicity to Atlantic salmon fry or smelts. Sodium reduced toxicity to salmon smelts, but had no effect on toxicity to fry. Both sodium and calcium reduced toxicity to lake trout fingerlings, but had no effect on lake trout fry. Results suggest that mitigating effects of solution cations on un-ionized ammonia toxicity may be related to species, size, and life stage. The influence of pH and temperature on the toxicity of ammonia was investigated by Dabrowski & Skiora (1986). Based on acute toxicity studies with common carp (Cyprinius carpio) the NH₄Cl 48 hour LC₅₀mean range was 1.60-1.96 mg un-ionized NH₃/L(equivalent to 103-109 mg total NH₃/L). Ammonia toxicity increased with higher pH and lowered with increasing temperature. The toxicity of ammonia to green sunfish (L. cyanellus) was studied by McCormicket al(1984). The 96 h LC50 concentrations at pH values of 6.6, 7.2, 7.7 and 8.7 were 0.5, 1.06, 1.34 and 1.73 mg/l NH3 or 272, 139, 57 and 9 mg/l NH3-N, respectively. The acute toxicity of ammonia to hatchery reared rainbow trout was measured in 86 flow-through tests, 96h -35 days long (Thurstonet al, 1983). Fish ranged from 1 day old fry (<0.1 g) to 4 year old adults (2.6 kg). The 96h LC50 was 0.6-1.1 mg/l unionized ammonia (11-48 mg/l total ammonia N). Susceptibility to ammonia decreased as the fish developed from sac fry to juveniles and increased thereafter. Acute toxicity decreased as temperature increased from 12-19 °C. No significant differences in toxicity were observed in tests with different ammonium salts. The LC50 values obtained for 12 and 35 days were not different from those obtained earlier on in the test.

The toxicity of ammonia to fish is attributable mainly to the un-ionised NH3 molecule. The proportion of un-ionised ammonia increases with increasing temperature and pH, but decreases with increasing salinity. At pH 8.5, the proportion of un-ionised ammonia is approximately 10 times that at pH 7.5 and, for every 9°C increase in temperature, the proportion of un-ionised ammonia approximately doubles. A number of mechanisms for the effect of un-ionised ammonia on the growth of fish have been proposed including:

• gill damage eventually causing suffocation;

• alteration of biochemical mechanisms including stimulation of glycolysis and suppression of the Krebs cycle leading to a build-up of acidic metabolites causing death by acidosis;

• alteration of biochemical mechanisms leading to death from a depletion of ATP in the basilar region of the brain

• osmoregulatory disturbance

• severe electrolyte imbalance;

• reduction in cellular K+ levels and hyperexcitability;

• inhibition of ATP production by uncoupling oxidative phosphorylation;

• increase in cerebral glutamine levels leading to a decrease in the neuro-inhibitor GABA (γ-aminobutyric acid) causing hyperexcitability.

No single definitive hypothesis for the toxicity of un-ionised ammonia to fish has been established.

The large dataset on the acute toxicity of ammonia to fish has recently (2007) been reviewed by the UK Environment Agency, in order to derive Environmental Quality Standards (EQSs) for this substance. The report states that the key study of acute fish toxicity is that of Rice & Bailey (1980) with ammonium sulphate, which identified a 96 -hour LC50 value in Oncorhyncus gorbuschka of 0.067 mg/L (NH3).