 Commonwealth of Australia 2010

Acute lethality to birds

An acute oral LD50 study was also conducted with mallard ducks (Anas platyrhynchos), reported by Henny et al. (1994), with further data evidently relating to this study presented by Hagelstein and Mudder (1997a). Mallard ducks were used for this study because waterfowl and shorebirds constituted over 70% of avian mortality at Nevada gold mines, and no data were available for aquatic bird species.

In these studies, sodium cyanide was administered to the birds in a single dose by gelatin capsule (as 99.4% reagent grade for the Wiemeyer et al., 1986 study and 96.7% technical grade for the Henny et al., 1994 study). The experimental methods used for the LD50 studies were based on those described by Hill and Camardese (1984), who referred to the (then) avian single-dose oral LD50 (median lethal dose) protocol of the US Federal Insecticide, Fungicide and Rodenticide Act (FIFRA, US Environmental Protection Agency, 1978). LD50 values were estimated by probit analysis.

Of the avian toxicity test data available, only the mallard duck (Anas platyrhynchos; a close relative of the local Pacific Black Duck) is a standard OECD test species for oral toxicity tests. Data are occasionally generated on the Japanese quail, an alternative OECD test species, whereas the starling and domestic chicken are generally considered to be relatively insensitive species. Thus, the available data includes species known to range widely in sensitivity from toxicity studies conducted with other chemicals, and also included several raptor species not normally tested in such studies.

Thus, these studies were based on a satisfactory guideline and were analysed by an appropriate statistical procedure. They are considered acceptable, but cannot be considered fully reliable as:

• 
  in most cases, only 3-5 birds per dose level were used, whereas current guidelines (USEPA 850.2100) specify a minimum of 10;
  
• 
  important information is lacking from the available reports, including the actual doses tested in several studies, various other details of the methods and details of the results, so that endpoints cannot be independently confirmed;
  
• 
  the birds used in the Wiemeyer et al. (1986) study were in several cases wild caught, non-standard test species, with no indication as to how they were habituated to the test facilities or how they reacted to the stresses involved in capture, handling and dosing;
  
• 
  while the Wiemeyer et al. (1986) study related to sodium cyanide, its purpose and design related to baiting uses of the substance rather than gold mining, though various raptor species are also present at mine sites.
  

Vomiting soon after dosing is a potential problem in oral toxicity studies as it can reduce the dose received. Wiemeyer et al. (1986) commented that neither passerine (including starlings) nor galliform (including quail and domestic chickens) bird species have well developed vomiting reflexes. They stated that even though vomiting was noted in two black vultures that eventually survived exposure, its prevalence was difficult to ascertain in these studies because birds were deprived of food before exposure, but presumably regurgitation of the capsule casing would have been readily evident. Mallard ducks are a species where the gagging reflex is sometimes observed in toxicity studies, but Henny et al. (1994) do not mention this occurring, though they did observe the treated birds (which were fasted overnight before dosing) continuously for 1 hour after dosing. Vomiting did need to be allowed for in studies with pigeons by Cooper (2003), but in that case liquid was delivered to the birds rather than a capsule.

Wiemeyer et al. (1986) noted that the black vultures, kestrels, owls and quail reacted much more violently to NaCN exposure than did the chickens or starlings. The first signs of toxicosis (i.e. illness resulting from exposure to a toxic substance, in this case cyanide) at nearly all dosages occurred between 30 sec and 5 min post exposure, and death usually followed in 15-30 min. Birds alive at 1 hour usually recovered, e.g. surviving black vultures were standing and reasonably alert at 1 hour after dosing. LD50 values from their study ranged from 4.0 to 21 mg/kg bw (equivalent to 2.1 to 11.1 mg CN/kg bw) in the species tested. There were no sex differences in the two species where this was tested (quail and starlings). Marked differences in acute toxicity of NaCN were found, with the three flesh eaters being more sensitive than the three plant material consumers, and species differences in sensitivity to NaCN were not related to body size. As expected, the chickens and starlings were relatively insensitive compared to the quail (Table 9.).

The results for mallard duck from the Henny et al. (1994) study indicated it to be the most sensitive bird species of those that have been tested in acute oral studies with NaCN (LD50 = 2.7 mg/kg bw, equivalent to 1.4 mg CN/kg bw). This result is considered acceptable, though not fully reliable, as only 6 birds per dose were tested and the full study report is not available. It is consistent with that of 2.9 mg NaCN/kg bw (equivalent to 1.5 mg CN/kg bw) reported by Clark et al (1991) in a study that is available as an abstract only and hence cannot itself be evaluated adequately.

Data to indicate the NOEL values for lethal effects (NOEL_mortality) are available for some species in the above studies. These data indicate NOEL_mortality values ranging from 0.53 mg CN/kg bw for mallard ducks to 6 mg CN/kg bw for domestic chickens, with the corresponding LOELs being 1.1 mg CN/kg bw and 12 CN/kg bw, respectively (Table 9.). A higher NOEL_mortality value for mallards (the most sensitive species for which data are available) of 0.8 mg/kg bw was obtained in the biochemical studies by Pritsos and Ma (1997) and Ma and Pritsos (1997) discussed below (Table 9.).

Acute lethality to mallard ducks

In the first *Fletcher (1986) study, different sets of five male and five female 20 week-old mallard ducks were exposed to a range of concentrations of free cyanide in tap water (0-265 mg/L) in a single dose (it is not clear whether this was administered by gavage, or was in fact simply provided on a single occasion for a limited period in drinking water). Following administration of the single dose, the ducks were observed for a period of seven days. The LC50 value for this single dose exposure for the combined male and female ducks was 181 mg total CN/L. Assuming a bodyweight of 1.0 kg per duck, this was stated to equate to an LD50 of 2.5 mg NaCN/kg (the calculated figure is stated to be the ‘acute oral toxicity of cyanide in the tap water or the actual mining solution when expressed in terms of mg/kg’, but is then compared directly to the other toxicity values for NaCN presented in Table 9.). Expressed in terms of total CN, this LD50 would be 1.3 mg/kg.

In a second similar study with mallard ducks by *Fletcher (1986), solution collected from an effluent holding pond at a gold mining operation in Nevada was used instead of tap water. The total cyanide levels were adjusted through the addition of NaCN to produce concentrations ranging from 107-332 mg/L. The result, again provided as a LC50 value following a single dose, was calculated to be 212 mg total CN/L. Assuming a bodyweight of 1.0 kg per duck, this equated to an LD50 of 3.2 mg NaCN/kg, or 1.7 mg CN/kg.

All of the mortalities occurred within the first three hours following treatment and all surviving birds appeared normal after seven days. Smith and Mudder (1991) noted that the effluent pond result was similar to that obtained with tap water and suggested that the overall results indicated that cyanide (rather than other toxins) was the primary concern with toxicity from mining solutions. Hagelstein and Mudder (1997a) noted that in general, the mallard tolerates slightly higher concentrations of cyanide when it is in solution with other effluent or process solution ions. However, there is inadequate evidence to confirm this from the slightly lower LC50 value for the effluent pond water in this single comparison. It is also evident that the way in which the solutions were prepared left much of the cyanide present as free CN.

If in fact provided as a direct single dose, these studies (*Fletcher, 1986) appear to have been conducted based on standard USEPA acute toxicity guidelines, though the results have unconventionally been expressed as LC50 values. It may also be that in fact they were short term exposure drinking water studies, which would explain the presentation of the results as LC50 values and the lack of certainty as to the volume received by each bird. Without the original reports to verify the actual procedures used and the full results, including the measured concentrations and dose volumes given, actual bird weights, and sublethal effects, these data cannot be confirmed or fully interpreted. With a ‘single dosing’ regime it is likely that the experimenters were able to achieve the concentrations stated, whether administered by gavage or in drinking water. However, the conversions of the LC50 data to LD50 values appear to be approximations, as the bodyweights used were evidently approximations, and if administered voluntarily via drinking water, the volume taken was also uncertain. Hence these studies are considered as supplementary information rather than reliable or acceptable toxicity studies.

However, the LD50 results for mallard ducks inferred from these studies are supportive of the LD50 result for mallard duck obtained by Henny et al. (1994) discussed above. The latter value will therefore be used as the endpoint for that species, i.e. 1.4 mg CN/kg.

The volume of solution given per bird or assumed to have been consumed by each bird in the *Fletcher (1986) studies was stated to be ~10 mL, but this is not fully consistent with the calculated results. Using the corresponding LC50 and LD50 figures (correcting the LD50 value from NaCN to total CN), back calculations indicate that the average doses or assumed intake for these studies were ~7.3-8.0 mL per bird (i.e. ~0.25 US fl oz). This is approximately one eighth of the estimated daily drinking water consumption of ~60 mL for a bird weighing 1 kg using the allometric Avian Water Intake Rate equation of Calder and Braun (1983) (Section 29.1.3). This was evidently envisaged by the researcher as representative of water consumption by mallard duck in a single drinking event. Studies by Pritsos and Ma (1997), Ma and Pritsos (1997), Cooper (2003), Brasel et al. (2003) and Cooper and Pritsos (1999) with mallard ducks and homing pigeons (below) used a similar strategy, with forced gavage of 10 mL of test solution varying in concentration intended to represent exposure to a single drink of CN-contaminated water in investigations of sub-lethal effects.

The LC50 results are also considered acceptable as an indication of the concentration of cyanide which is toxic to mallards from a single drinking event. However, some caution in interpreting these single event LC50 concentrations is necessary, as field and laboratory observations make it clear that individual birds within a species differ significantly in their drinking behaviour (see ‘Short term oral and contact exposure’ discussion below), as well as there being wide differences between species. In particular, mallards are a waterbird which can be expected to drink at several times over a day and therefore to take a smaller proportion of their daily water intake with each drink. In contrast, species which visit a waterbody on one to a few occasions per day to drink would be expected to take a larger proportion of their daily consumption on each occasion they visit. Caution is also necessary in interpreting how the results from a single exposure event apply to continuing exposure after an initial drink, as there may be cumulative effects due to incomplete detoxification and recovery between drinking events, and/or a decrease in the ability of detoxification mechanisms to operate. The LC50 results are somewhat higher than the LC50 of 115 mg CN/L from the brief test with 2 h exposure described below, which could be a result of greater water ingestion and/or additional means of exposure in the test below.

An evaluation of the mortality from the *Fletcher (1986) studies together with those by *Fletcher (1987) (Section 34.1.1) was used to estimate a LC1 (concentration in the water necessary to kill 1% of the birds) for free CN of ~50 mg/L. This was stated to be equivalent to ~0.73 mg/kg bodyweight in a single dose, i.e. below reported LD50 values in the literature (Table 9.). Note that if a 1 kg duck consumes 8-10 mL of a 50 mg WAD CN/L solution, this would result in a body dose of 0.4-0.5 mg CN/kg, which is at or below the lowest NOEL for lethal effects to mallard ducks of 0.5 mg CN/kg (see Table 9.: evidently from the Henny et al., 1994 study).

Table 9.. Avian acute oral toxicity studies with sodium cyanide