Pharmacokinetics and methaemoglobin reductase activity as determinants of species susceptibility and non-target risks from sodium nitrite manufactured feral pig baits
Steven J. LapidgeAD and Charles T. EasonBC
AInvasive Animal Cooperative Research Centre, 48 Oxford Street, Adelaide, SA 5061, Australia.
BBio-Protection and Ecology Division, Lincoln University, PO Box 84, Lincoln 7647, Canterbury, New Zealand.
CConnovation Ltd, PO Box 58613, Manukau 2141, New Zealand.
DCorresponding author. Email: firstname.lastname@example.org
Abstract. Vertebrate pesticides are used in Australia to manage populations of invasive pest species, including feral pigs, which threaten native ecosystems, damage crops and spread disease. Feral pigs are currently poisoned using actives that are considered by some to be inhumane and more suitable agents are being sought. Pigs are susceptible to methaemoglobin forming compounds, due to innately low levels of methaemoglobin reductase, and research has identified that sodium nitrite is a promising humane alternative active. To attain registration, a new vertebrate pesticide must however be species-specific in its toxicity (rare), or be presented in such a manner to demonstrate that it is safe for use around non-target species. Furthermore, it should be shown to represent a low risk of bioaccumulation and secondary poisoning, as is the case with sodium nitrite which has a half-life of one hour or less in those species tested. Here we present a risk analysis for the use of sodium nitrite used in manufactured feral pig baits. As pharmacokinetics of sodium nitrite are similar in different species, and it does not require biotransformation to an active metabolite, methaemoglobin reductase activity in red blood cells from different species is directly correlated with differences in acute toxicity. Primary poisoning risks, or susceptibility, have been calculated for 28 marsupial and nine eutherian mammal, four reptile and two bird species based on published doses and methaemoglobin reductase activity levels. Those species that have been previously recorded to sample manufactured feral pig baits, and will be susceptible to the likely bait dose used, are identified, along with the percentage of a bait that they will need to consume. The current steps being taken to limited potential non-target poisonings, including an appropriate active to matrix ratio, are discussed.
Currently, sodium fluoroacetate (1080), warfarin and yellow phosphorus are the toxins used to poison feral pigs, Sus scrofa, in Australia. However, the use of warfarin and yellow phosphorus has recently been assessed as inhumane (Cowled and O’Connor 2004; Sharp and Saunders 2004). Sodium fluoroacetate is the most commonly used toxin, with it being added to various palatable bait substrates for poisoning pigs, such as grain, pellets, meat and manufactured baits (Hone and Kleba 1984; Twigg et al. 2005; Cowled et al. 2006a). But whilst it appears markedly more humane than other current pig toxins (Cowled and O’Connor 2004; Sharp and Saunders 2004), it has come under scrutiny and is not favoured by animal welfare groups (Sherley 2004 and 2007; Cooper et al. 2007).
In 2005 a search was undertaken for other potential chemicals that would be more suitable for feral pig management, with sodium nitrite proving to be the lead candidate (Cowled et al. 2008a). Proof-of-concept pen trials showed that it was highly efficacious in euthanasing pigs whether given by gavage or eaten in PIGOUT® baits (Cowled et al. 2008a). Ironically, it is the same chemical that is used to preserve pig meat that is also highly toxic to the live animal. Because of the widespread use of sodium nitrite as a human food preservative there is a considerable amount of data on the chemistry and toxicology of this compound. Nitrite acts by the humane (Porter and Kuckel 2009), and Royal Society for the Prevention of Cruelty to Animals (RSPCA) supported (Sherley 2007), mode of action of methaemoglobinaemia, and so is preferred to 1080 on welfare grounds (Marks et al. 2004; Cowled et al. 2008a). Another methaemoglobin former, para-aminopropiophenone (PAPP), is currently being developed for foxes, wild dogs and feral cats in Australia (Marks et al. 2004; Fleming et al. 2006), and stoat, ferrets and feral cats in New Zealand (Fisher et al. 2005; Fisher and O’Connor 2007; Murphy et al. 2007). Pigs are highly susceptible to the effects of methemoglobinaemia as they contain uniquely low levels of methemoglobin reductase (Smith and Beutler 1966; Agar and Harley 1972), the enzyme required to reverse the methemoglobin formation process. This inherent weakness has previously resulted in numerous reported fatal cases of domestic pigs being poisoned with nitrite (Robinson 1942; Gwatkin and Plummer 1946; Winks et al. 1950; London et al. 1967).
As with PAPP, nitrite kills through terminal hypoxia caused by methaemoglobinaemia. High methaemoglobin levels reduce the oxygen-carrying capacity of the blood and at toxic doses hypoxia and central nervous system depression precede death. Time to death for feral pigs is 1-2 hours with few visual symptoms (Cowled et al. 2008a; Lapidge et al. 2009). This is a marked improvement on conventional vertebrate poisons. A further benefit of nitrite is that it is known to break down readily in the environment through biological reduction (e.g. Wanntorp and Swahn 1953; Sofia et al. 2004), limiting non-target poisoning risks and environmental contamination. A further benefit of the chemical is that methylene blue will reverse methaemoglobinaemia induced by sodium nitrite, or PAPP, and is an established antidote for nitrite poisoning in livestock (Wendel 1939; Wright et al. 1999).
Understanding mechanisms of toxicity, pharmacokinetics and receptor level effects are important for a number of reasons (Eason et al. 1990). These include development of antidotes, risk assessment, defining the susceptibility of non-target species, defining the risk of secondary poisoning or bioaccumulation, and also the risk of residues persisting in non-target species, including wildlife and livestock inadvertently exposed to sub-lethal doses. In the following sections we firstly examine the basis for species variation in response to xenobiotics, and then review the pharmacokinetics of sodium nitrite and then receptor level processes as determinants of species susceptibility. This risk assessment is based to some extent on the research into the sensitivity of Australian animals to 1080 poison (McIlroy 1981, 1986).
It is planned that formulated sodium nitrite will be delivered using the new HOG-GONE® matrix (Animal Control Technologies Australia P/L). The concept is based on the successful PIGOUT® bait matrix which is currently used to deliver 1080 to feral pigs (Cowled et al. 2006a,b), but was also designed as a potential vaccine carrier (Campbell et al. 2006; Cowled et al. 2008b). Improvements to the target-specificity of feral pig baiting campaigns have been substantial in Australia when compared to the use of grain and meat baiting materials (Cowled et al. 2006b). Preliminary field trial results for the HOG-GONE® matrix, which is harder and less pungent, indicate that the target specificity has been increased further (S. Lapidge, unpub. data; A. Bengsen, unpub. data). Notwithstanding, a thorough risk analysis is required for the potential field use of sodium nitrite if the chemical is to gain registration as a new vertebrate pesticide. This paper predicts the innate susceptibility to sodium nitrite for 28 marsupial and nine eutherian mammal, four reptile and two bird species based on published oral doses and methaemoglobin reductase activity levels. It highlights non-target species potentially at risk for the use of HOG-GONE® baits, and the methods being employed to minimise such risk. The risks of secondary poisoning, either through the consumption of sub-lethally dosed or poisoned animals are also discussed.
Pharmacokinetics and species variation
The metabolism and pharmacokinetics of a vertebrate pesticide are often determinant factors in its ultimate manifestation of toxicity, non-target safety and residue profiles. The metabolic deactivation of toxins results from the actions of enzymes whose primary function is to protect the body against the accumulation and undesirable effects of foreign compounds naturally present in food and in the environment. Metabolism and excretion are protective processes attempting to limit persistence in the body.
Species differences in sensitivity to an individual chemical are linked to variation in the pharmacokinetic differences. Savarie et al. (1983) clearly demonstrated this with PAPP, with a seventy-fold difference occurring between the LD50 of a coyote and a Striped skunk or Golden eagle. Wood et al. (1991) later detailed that this was due to the metabolic activation of PAPP to a reactive, and more toxic, intermediate metabolite in susceptible species such as canines. This is also the case with 1080 and cholecalciferol. However, as is normal with toxic intermediates of other drugs or pesticides, metabolic processes in the body then go on to detoxify these intermediate compounds as well. In the case of 1080 and sodium nitrite the parent compound is itself highly water soluble; hence the normal metabolic processes that render a toxin more water soluble are less important with regard to their elimination from the body.
Receptor site interactions and species susceptibility
The pharmacokinetics of the direct methaemoglobin (MetHb) former sodium nitrite is similar in different species as it does not require biotransformation to an active metabolite (Wright et al. 1999). As with the drug amrinone (Eason et al. 1986), there are rapid and similar patterns of absorption and excretion across a range of different species. Hence, extrapolation of chemical-receptor interactions can be used with greater confidence to predict the innate risk of sodium nitrite to non-target species. The receptor site for sodium nitrite poisoning is haemoglobin in the red blood cell. The mode of action of nitrite is the oxidization of the haem iron in red blood cells from the ferrous state (Fe++) to the ferric state (Fe+++) to form MetHb. MetHb is incapable of carrying oxygen and cyanosis results, with death occurring if the dose is high enough (Egyed and Hanji 1987). The pattern of methaemoglobinaemic response induced when erythrocytes are exposed to sodium nitrite oxidant challenge will be a balance between MetHb formation and its subsequent reduction back to haemoglobin by the protective enzyme MetHb reductase.
The activity of the enzyme MetHb reductase varies in different animals, and is known to determine a species direct sensitivity to a direct methaemoglobin former (Smith and Beutler 1966; Stolk and Smith 1966; Agar and Harley 1972; Board et al. 1977; Lo and Agar 1986; Whittington et al. 1995; Agar et al. 2000; Rockwood et al. 2003). Species differences in MetHb reductase are therefore critical when evaluating the risk to different species from MetHb formers such as sodium nitrite. Under normal conditions this enzyme is the only system within the erythrocyte that maintains haemoglobin in its oxygen-carrying reduced state. Toxicologically, MetHb reductase will therefore be the rate-limiting enzyme controlling the toxicodynamics of sodium nitrite’s effect on the red blood cells. In theory, species with lower MetHb reductase activity convert MetHb back to haemoglobin more slowly than do species with higher activity, and will therefore be more susceptible to sodium nitrite. Conversely animals with greater MetHb reductase activity will be at less risk of sodium nitrite induced toxicity.
Correlating methaemoglobin reductase activity, body size, diet and risk
To test this hypothesis we collated all published acute oral sodium nitrite toxicity data and MetHb reductase activity, as generally determined through nitrite challenge. Data for seven eutherian mammal species were obtained (Table 1). In those species for which published data exists on both gavage acute toxicity and on red blood cell MetHb reductase activity there is a strong correlation between susceptibility and MetHb reductase activity (r=0.894). Simple linear regression analysis revealed a highly significant (F1,5= 21.7; P = 0.006) relationship, with a line of best fit accounting for 81% of the variance with the equation y=0.2614x-10.149 (Fig. 1). The equation was subsequently used to predict the lethal oral dose of sodium nitrite for all Australian species with published MetHb reductase activity levels. MetHb reductase activity values were obtained from Agar et al. (2000), except for the platypus and echidna (Whittington et al. 1995). Reptile and bird values were obtained from Board et al. (1977), and eutherian mammal values from Lo and Agar (1986) and Rockwood et al. (2003). Predictions were based on the equation:
Estimate minimum lethal gavage dose = (MetHb reductase activity + 10.149) / 0.2614)
and results are presented in Table 2.
Estimated lethal bait dose is based on a correction factor of 3 times, as indicated between the minimum effective gavage and bait delivered sodium nitrite doses for feral pigs (Cowled et al. 2008a). The correction factor takes into account the destruction of nitrite by stomach acid and slower absorption when delivered in the physically dense PIGOUT/HOG-GONE® bait matrix. This correction factor would change if another bait type was used, a smaller volume of matrix was used, or a high active to matrix ratio was used. It may also differ between species depending on the digestive physiology of the species.
The indicative body weights used in Table 2 to estimate lethal bait-delivered sodium nitrite doses are the average adult body mass of males and females as published in Strahan (2004), as well as various internet sources. It is acknowledged that juveniles of each species will require smaller doses of sodium nitrite to those predicted for adults. Species recorded consuming PIGOUT® baits were reported by Cowled et al. (2006b) or are unpublished data. Early indications are that the HOG-GONE® bait, which will carry sodium nitrite, is significantly more target-specific with introduced red foxes (Vuples vulpes)and Australia ravens (Corvus coronoides) the only non-target species recorded consuming toxic bait during preliminary fieldtrials. Although many of the species listed in Table 2 are highly unlikely to consume a manufactured feral pig bait, they have been included in case other bait matrices are considered for the delivery of sodium nitrite in the future, such as grain.
Insert Table 1 -
Insert Figure 1 –
Insert Table 2 -
Proofing model: direct non-target species testing
To examine the validity of the lethal dose predictions in Table 2, and to ascertain the direct risk to two key potential non-target species that have previously been observed to consume PIGOUT® baits (Lapidge, pers. obs. and Cowled et al. 2006b), direct toxicity testing on Common brushtail possums and Tammar wallabies was undertaken by Landcare Research (New Zealand) and Connovation (New Zealand) respectively. Trials occurred in New Zealand where both species are invasive and the use of sufficient animals for potentially lethal trials was more likely to be ethically approved. Both trials were approved by each organisations respective Animal Ethics Committee.
Pen trials were undertaken on 12 individually caged brushtail possums at the Landcare Research Animal Facilities in Lincoln, New Zealand. Although the trial was to be conducted using non-toxic and toxic HOG-GONE® baits, the use of the animal protein-containing baits was blocked by the New Zealand Ministry of Agriculture and Forestry, and as such a substitute but very similar bait matrix (Connovation Ferafeed 213) was used for the trials. Formulated sodium nitrite was hand-incorporated into the bait matrix at the same 10% toxin ratio as HOG-GONE® so the results were directly applicable. Possums were acclimatised to the facility for four weeks prior to the trial commencing.
On the day prior to toxic testing each possum was offered 125g of non-toxic bait material, of which they consumed 32±1g on average over 22 hours. The following day each possum was offered four 25g toxic matrix balls (notably easier to consume than one 200g HOG-GONE® solid bait encased in a cellulose skin), of which they ate 3.8±0.5g on average over 11.5 hours. The total amount of each bait type was substancially different, indicating that the toxic bait was less acceptable to possums. Maximum consumption by any individual possum was 8g of matrix (0.8g active) or 229 mg/kg of sodium nitrite. Although five of the 12 possums showed visible signs of methaemolobin formation (blue/cyanotic nose), none of the animals were affected behaviourally in terms of responses to stimuli, and as predicted no possums received a lethal dose and died. The model predicted a minimum lethal dose of 393 mg/kg or 1.6g of active (Table 2). All possums were monitored for seven days following the trial, with the maximum weight changes being -3.3% to +10.5%, with all possums appearing healthy throughout the observation period (Fisher et al. 2009).
Tammar wallaby trials were undertaken in outdoor pens in a new animal research facility at Rotorua. Twelve individually-housed wallabies were fed a single 10g ball of Ferafeed (permission to us HOG-GONE was again denied), instead of their 100g of stock pellets, on three days leading up to toxic testing. Water and grass was available ad libitum. All wallabies were observed to consume their 10g ball of Ferafeed within 45 min. Toxic trials consisted of the presentation of 50g of the 10% nitrite matrix presented in four small balls. Nine of 12 wallabies would not consume any toxic matrix after sniffing the mixture. Three wallabies that consumed material (body weights 1.9, 3.0 and 4.0 kg) ate 17, 10 and 25g of matrix respectively, resulting in nitrite doses of 894, 335 and 625 mg/kg respectively. All doses were lethal, as predicted in Table 2 (minimum lethal dose 245 mg/kg), with the mean time to first symptoms being 63 min and 157 min to death. Of note is that one juvenile 3kg female consumed only 1g of sodium nitrite, which is less that the estimated lethal bait dose of 1.6g predicted in Table 2. This discrepancy is due to the adult tammar wallaby weight of 6.5kg used to make the prediction (from Strahan 2004). The three wallabies that died all displayed lethargy, shallow breathing, slight leg spasms, and unconsciousness before death (Shapiro and Eason 2009).
Additional non-toxic trials were undertaken by the Tasmanian Department of Primary Industries and Water to assess the attractiveness of the HOG-GONE® baits to Bennett’s wallaby (Macropus rufogriseus; n=5) and Tasmanian pademelon (Thylogale billardierii; n=17) held at the departments Launceston facilities. Twelve HOG-GONE® baits were placed in the wallabies 0.8ha enclosure and monitored using motion sensitive video cameras continually for 5 nights. Although most wallabies investigated the baits at least once, they generally recoiled abruptly once smelling the bait material (Fish and Statham 2009). No wallabies showed any inclination to consume the baits.
Although low sample sizes for possums (n=5) and tammar wallabies (n=3), results from toxic trials supported the lethal dose predictions made in Table 2, and provide some validity to the model approach presented. Results from the overall trials do however clearly indicate that marsupials are generally repelled by the HOG-GONE® bait matrix itself or the smell of formulated sodium nitrite. Despite formulation, the nitrite smell is still present in the toxic matrix and the chemical is extremely salty to taste. This is likely to be accentuated in the final product when the two are combined, although additional masking formulation research is still ongoing. Notwithstanding, the primary poisoning risks posed by a toxic HOG-GONE® baiting campaign are likely to be acceptable, particularly considering that in a recent HOG-GONE® field trial on Kangaroo Island, South Australia, no free-living tammar wallabies were recorded consuming ground-laid and unprotected HOG-GONE® baits, despite the species being locally abundant (Lapidge, unpub. data).
Pharmacokinetics of sodium nitrite
There is no evidence that sodium nitrite exerts selective toxicity based on classical species variation in the metabolism or pharmacokinetics of this compound in mammals. Analysis of the publications on the fate of sodium nitrite in animals and humans therefore enables comparisons across different species. All these publications indicate that sodium nitrite is rapidly eliminated by different animals and humans. Since sodium nitrite does not require biotransformation to be active there is a close correlation between the concentration of sodium nitrite in the blood and methaemaglobin induction. One of the more detailed recent publications in this field by Kohn et al. (2002) examines time and the dose-dependent relationship between exposure to nitrite and induction of methemoglobinemia in rats. The authors report on the toxicodynamic processes as absorption of nitrite following oral ingestion, elimination from the plasma, partitioning between plasma and erythrocytes, binding of nitrite to haemoglobin and methaemoglobin, and the free radical chain reaction for haemoglobin oxidation. Peak plasma levels of nitrite were achieved in both sexes of rats approximately 30 minutes after oral exposure, and peak methaemoglobin levels were achieved after 100 minutes. The t(1/2) for recovery from methemoglobinaemia was 60 to 120 minutes depending on dose and route of administration.
We have reanalysed the raw data from Kohn et al. (2002) to elucidate the time course of plasma nitrite in rats at different doses and this is summarised in Table 3. The plasma elimination t(1/2) for sodium nitrite in rats ranges from 42.0 to 62.5 minutes after oral dosing. These values are similar to those quoted in humans (Dejam et al. 2007), who showed a plasma elimination half-life of 42 minutes. They also correlate closely to plasma elimination t(1/2) values of 29.0, 30.0 and 34.0 minutes in sheep, dog and ponies reported by Schneider and Yeary (1975).
- Insert Table 3 -
In an unpublished report (MRI 2004) confirmation that sodium nitrite is rapidly eliminated from the blood is provided in both rats and mice. In rats orally dosed with 80 mg/kg peak plasma concentrations were achieved at 30 mins after dosing and these decreased to below the limit of detection after 8 hours. In mice peak plasma concentrations occurred after 10 mins and sodium nitrite was undetectable in the blood after 4 hours. Earlier publications on the biotransformation of sodium nitrate indicate that conversion to nitrite is important for toxicity, however nitrite is very quickly converted and may not be readily detected, therefore methaemoglobin formation was often used as an indicator of nitrite formation (Ward et al. 1986). The rapid elimination of nitrite predicted by Ward et al. (1986) has been confirmed by Kohn et al. (2002) and other research groups. Nitrite is reportedly broken down to hydroxylamine, ammonia and urea (Mascher and Marth 1993). Kovacs et al. (1960) reported that the concurrent administration of sodium bicarbonate was found to increase the lethality of sodium nitrite, probably due to the creation of low acid conditions in the stomach, reducing breakdown to ammonia and aiding absorption.
The pharmacokinetic data on sodium nitrite in such diverse species as mice, rats, sheep, dog, ponies and humans, coupled with information on the toxicodynamics of sodium nitrite, suggests similar Cmax, t(1/2), and rapid clearance would occur in other animal species and we could expect elimination of sodium nitrite following sub lethal exposure within 12 hours.
To help distinguish between different vertebrate pesticides and add some clarity to risk assessment, Eason et al. (2008) has recently classified compounds used for animal pest control into four groups, based on their persistence in sub-lethally exposed animals. The criteria for the four groups, and the allocation of different compounds to these groupings, are described below and in Table 4.
Group 1 – Sub-lethal doses of these poisons are likely to be substantially excreted within 24 hours (e.g., cyanide, zinc phosphide, PAPP and 1080). In the case of 1080, complete excretion of all residues may take up to 4-7 days. On the basis of what we can identify in the literature regarding sodium nitrite we believe it appropriate to include it in Group 1 and distinguish it from the more persistent compounds in Groups 2-4.
Group 2 – Residues resulting from sub-lethal doses of these poisons are likely to be substantially cleared from the body within 2 to 4 weeks (e.g., pindone and diphacinone).
Group 3 – Residues resulting from sub-lethal doses of these poisons are likely to be cleared from the body within 2 to 4 months (e.g., cholecalciferol and coumatetralyl).
Group 4 – Residues resulting from sub-lethal doses of these poisons may not ever be completely cleared from the body (e.g., bromodiolone, brodifacoum, difenacoum, and flocoumafen).
- Insert Table 4 -
Risk to non-target species from vertebrate pesticides can be exacerbated if they bioaccumulate in the food chain. Pharmacokinetic analyses in different animals and humans have shown that unlike some common rodenticides e.g. brodifacoum, sodium nitrite is unlikely to cause secondary poisoning due to bioaccumulation. Furthermore it is unlikely to cause any significant food web contamination when used repeatedly as a vertebrate pesticide because it will be rapidly excreted.
This paper has demonstrated that the toxicity of sodium nitrite is directly related to the activity of methaemoblobin reductase in each species. This strong relationship has meant that a risk analysis for 43 species of mammals, birds and reptiles residing in Australia has been able to be undertaken without the need for extensive lethal trials, as often previously occurred in the past (McIlroy 1981). Although not as extensive as the risk assessment research documented by McIlroy (1986) for 1080, c. 171 species, the analysis provides a clear indication of those species that are likely to be at risk.
Results from this analysis indicate that sodium nitrite is highly toxic to most species. Although feral pigs are one of the more susceptible species on a mg/kg basis (Table 1; Cowled et al. 2008a), they are also a large animal compared to most potential non-target species, and therefore the amount of chemical required to humanely euthanase an adult feral pig will always mean that non-target species will be at risk unless the active can be delivered in a species specific manner. The HOG-GONE® bait has been demonstrated to almost achieve this, with the exception of foxes and ravens. Obviously the same risk analysis will need to be undertaken for any other manufactured baits, or if a concentrate version of the chemical was made for presentation in grain or meat baits.
This risk assessment has involved all marsupial and eutherian mammal, reptile and bird species occurring in Australia for which methaemaglobin reductase activity has been published. Besides foxes and ravens, the risk analysis covers most species that have previously been recorded to take manufactured feral pig baits. Obviously many of the species listed in Table 2 are highly unlikely to consume manufactured feral pig baits, but have been included should predictions ever be required. Potential exists for other species to be included if data on direct susceptibility or methaemaglobin reductase activity is found. Furthermore, non-lethal assessment of methaemaglobin reductase activity can be easily undertaken to assess other species innate sensitivity (Power et al. 2007). Such a process will need to be used if sodium nitrite is to be registered in other countries where less information is available on methaemaglobin reductase activity in potential non-target species.
Based on their previous consumption of PIGOUT® baits the non-target risk posed by a HOG-GONE® baiting campaign are principally to brush-tailed possums (9% of a bait), northern brown bandicoots (6%), swamp (17%) and tammar (8%) wallabies, foxes (unknown) and wild dogs (18%). This is however likely an over-estimation, as recent HOG-GONE® pen and field trials have indicated that wallabies (swamp, tammar, Bennett’s and Tasmanian pademelon) and wild dogs are less inclined to consume the new HOG-GONE® bait matrix. Furthermore, given the high intake of non-toxic HOG-GONE® baits in field trials conducted thus far, scope exists to reduce the individual bait loading to 10g of formulated nitrite or a 5% ratio. This would effectively double the predicted bait consumption percentages required in Table 1, and provide a further safeguard. This is because MetHb needs to be induced rapidly to case death. Target species such as pigs devour their food quickly which aids induction of lethal MtHb levels, whereas small mammals or birds that nibble at baits would be more likely to experience a transient sub-lethal MtHb rise. For example, an adult possum that requires a 1.9g nitrite dose will need to eat 40g of toxic HOG-GONE® matrix in one short sitting to receive a lethal dose. Results from the Landcare Research trial indicated that only the largest of the possums would generally do this with non-toxic matrix, with 59g being the highest consumption over a 22hr period. Furthermore, an animal must consume the dose before it starts feeling lethargic (the first symptom of nitrite poisoning; Cowled et al. 2009), after which animals generally stop eating. The slow consumption of baits is unlikely to be lethal in most species as methaemoglobin reductase is able to keep pace with the conversion of methaemoglobin back to oxyhaemoglobin. Notwithstanding, methods being used to limit the risk include:
the 200g size and formulation of the HOG-GONE® bait itself, whereby most possums and wallabies are adverse to consuming the bait, particularly when it contains sodium nitrite;
adult feral pigs are known to consume multiple HOG-GONE baits, up to 15 in one sitting, and therefore the lethal dose for particularly large (100kg+) boars can be spread over multiple baits; and
where necessary, physical exclusion of non-target species is possible, either using large rocks, over-turned buckets pegged into the ground (as occurs in Namadgi National Park), or through a bait hopper, such as the Boar Buffet® (Lapidge et al. 2009), that physically excludes non-target species.
When used as a package there is little primary poisoning risk from the use of HOG-GONE® to non-target wildlife species or domestic stock. Any baiting campaign will however hold some risk, as individual behaviour of animals can not always be predicted.
Phamacokinetic analyses of sodium nitrite has demonstrated that it is rapidly eliminated from the body. Lack of persistence is an important feature in vertebrate pesticide risk assessment. Risk to non-target species from vertebrate pesticides can be exacerbated if they bioaccumulate in the food chain. Pharmacokinetic analyses in different animals and humans have shown that, unlike some common rodenticides (e.g. brodifacoum), sodium nitrite is unlikely to cause secondary poisoning due to bioaccumulation. The plasma elimination half-life in a wide range of animals, including mice, rats, dog, sheep and ponies and in humans is approximately 1 hour or less, and residues will not persist in sub-lethally exposed animals for >12 hours. Given this remarkable consistency in the pharmacokinetics of sodium nitrite, and the fact that biotransformation is not necessary for sodium nitrite to be active or toxic, it is possible to elucidate species differences in terms of receptor differences.
In this paper we have reviewed risk to non-target species linked to pharmacokinetics and also MetHb reductase activity, animal size and behaviour to determine non-target risk when sodium nitrite is potentially used for feral pig control in manufactured baits. One aspect we have not completely covered is the persistence of nitrite in poisoned pigs’ carcasses and the rate of decay of nitrite. This will require further study to ensure persistence of nitrite is not a feature of its use. This is a recognised hazard associated with 1080 and clarification of the risk to non-target species form poisoned carcasses will be needed. Residue data will be ascertained from field poisoned feral pigs during large-scale HOG-GONE® field efficacy trials and will be published at a later date. As the reaction of nitrite with deoxyhemoglobin results in the production of nitric oxide and methemoglobin, most nitrite is converted during the toxicosis, and the chemical has a short half-life, it is predicted that residues will be negligible.
Sodium nitrite has the potential to become a new, humane (Porter and Kuckel, 2009) vertebrate pesticide for the management of feral pigs and potentially other species. However, the active is generally toxic to most species, and will therefore require presentation in species-tailored baits. Bengsen et al. (2008) details a method for achieving this. As outlined above, given an appropriate presentation the active can be delivered in a highly species-tailored fashion that posses few risks for potential non-target species. The short half-life of the compound and rapid elimination from the body further reduce this risk. Similarly, nitrite is known to break down readily in the environment through biological reduction (e.g. Wanntorp and Swahn 1953; Sofia et al. 2004), and is therefore unlikely to cause persistent environmental residues.
The preparation of this risk analysis, including the contract testing of possums (Landcare Research, New Zealand) and tammar wallabies (Connovation, New Zealand), was funded by the Federal Department of the Environment, Water, Heritage and the Arts. The authors thank Michelle Smith and the team at Animal Control Technologies Australia P/L, Penny Fisher and team, Lee Shapiro, Duncan MacMorran, Mick Statham and Rebecca Fish for being directly involved in pen trials. Associate Prof Chris Frampton, University of Otago, is thanked for his advise on pharmacokinetics analyses on the Kohn et al. (2002) paper, and Jason Wishart, IA-CRC, is thanked for extracting the data from this paper which greatly assisted in our understanding of sodium nitrite pharmacokinetics.
Agar, N.S. and Harley, D. (1972). Erythrocytic methaemoglobin reductases of various animal species. Experientia 28, 1248-1249.
Agar, N.S., Reinke, N.B., Godwin, I.R., and Kuchel, P.W. (2000). Comparative biochemistry of marsupial erythrocytes: a review. Comparative Haematology International 10, 148-167.
Bartik M. and Piskac A. (1981). Veterinary Toxicology. Elsevier, Amsterdam. 346 pp.
Bengsen, A.J., Leung, L.K-P., Lapidge, S.J. and Gordon, I.J. (2008). A theoretical framework for the design of target-specific vertebrate pest control in complex communities. Ecological Management & Restoration 9, 209-216.
Berlin, C.M. (1970) Treatment of cyanide poisoning in children. Pediatrics 46, 793-796.
Board, P.G., Agar, N.S., Gruca, M. and Shine, R. (1977). Methaemoglobin and its reduction in nucleated erythrocytes from reptiles and birds. Comparative Biochemistry and Physiology B 57, 265-267.
Boink, A. and Speijers, G. (2001). Health effects of nitrates and nitrites, a review. Acta Horticulturae 563, 29-36.
Campbell, T.A., Lapidge, S.J. and Long D.B. (2006). Baits to deliver pharmaceuticals to feral swine in southern Texas. Wildlife Society Bulletin 34, 1184-1189.
Cooper, D., Larsen, E., and Shields, J. (2007). 1080 and wildlife: ethical issues raised by its use on Australian animals Pp. 229-232 in Lunney, D. and Hutchings, P. (eds) Pest or guest: the zoology of over abundance. Royal Zoological Society of New South Wales.
Cowled B.D., Lapidge S.J., Smith, M. and Staples, L. (2008b). Vaccination of feral pigs (Sus scrofa) using iophenoxic acid as a simulated vaccine. Australian Veterinary Journal 86, 50-55.
Cowled, B. and O’Connor, C. (2004). A project that investigates current options for managing feral pigs in Australia and assesses the need for the development of more effective and humane techniques and strategies – Stage 3 Report. Pest Animal Control Cooperative Research Centre, Canberra, Australia. Available from http://www.environment.gov.au/biodiversity/invasive/publications/feral-pig/stage3.html
Cowled, B.D., Elsworth, P. and Lapidge, S.J. (2008a). Additional toxins for feral pigs (Sus scrofa) control: identifying and testing Archilles’ Heel. Wildlife Research 35, 651-662.
Cowled, B.D., Gifford, E., Smith, M., Staples, L. and Lapidge, S.J. (2006a). Efficacy of manufactured PIGOUT® baits for localised control of feral pigs in the semi-arid Queensland rangelands. Wildlife Research 33, 427-437.
Cowled, B.D., Lapidge, S.J., Smith, M. and Staples, L. (2006b). Attractiveness of a novel omnivore bait, PIGOUT®, to feral pigs (Sus scrofa) and assessment of risks of bait uptake by non-target species. Wildlife Research 33, 651-660.
Dejam, A., Hunter, C.J., Tremonti, C., Pluta, R.M., Hon, Y.Y., Grimes, G., Partovi, K., Pelletier, M.M., Oldfield, E.H., Cannon, R.O., Schechter, A.N. and Gladwin, M.T. (2007). Nitrite infusion in humans and nonhuman primates: endocrine effects, pharmacokinetics, and tolerance formation. Circulation 116, 1821-1831.
Dollahite, J.W. and Rowe, L.D. (1974). Nitrate and nitrite intoxication in rabbits and cattle. Southwestern Veterinarian 27, 246-248.
Druckery, V.H., Steinhoff, D., Beuthner, H., Schneider, H. and Klärner, P. (1963). Screening of nitrite for chronic toxicity in rats. Arzneimittelforschung 13, 320-323.
Eason, C.T., Bonner, F.W. and Parke, D.V. (1990). The importance of pharmacokinetics and receptor studies in drug safety evaluation. Regulatory Toxicology and Pharmacology 11, 288 307.
Eason, C.T., Ogilvie, S., Miller, A., Henderson, R., Shapiro, L., Hix, S., MacMorran, D. and Murphy, E. (2008) Smarter pest control tools with low-residue and humane toxins. Vertebrate Pest Conference 23, 148-153.
Eason, C.T., Pattison, A., Howells, D.D., Mitcheson, J. and Bonner, F.W. (1986). Platelet Population Profiles : Significance of Species Variation and Drug Induced Changes. Journal of Applied Toxicology 6, 437 41.
Egyed, M.N. and Hanji, V. (1987). Factors contributing to recent outbreaks of acute nitrate poisoning in farm ruminants. Israel Journal of Veterinary Medicine 43, 50-55.
Fish, R and Statham, M. (2009). The atrractiveness of the feral pig bait HOG-GONE® to Bennett’s wallaby (Maropus rufogriseus) and Tasmanian pademelon (Thylogale billardierii). Unpublished report, Tasmanian Insititute of Agricultural Research, Launceston, Tasmania.
Fisher, P. and O’Connor, C. (2007). Oral toxicity of p-aminopropiophenone to ferrets. Wildlife Research 34, 19-24.
Fisher, P., Brown, S. and Arrow, J. (2009). Possum response to ingestion of a new toxin for feral pig control. Unpublished report, Landcare Research, Lincoln, New Zealand.
Fisher, P.M., O’Connor, C.E. and Murphy, E.C. (2005). Acute oral toxicity of p-aminopropiophenone to stoats. New Zealand Journal of Zoology 32, 163-169
Fleming, P.J.S., Allen, L.R., Lapidge, S.J., Robley, A., Saunders, G.R. and Thomson, P.C. (2006). A strategic approach to mitigating the impacts of wild canids: proposed activities of the Invasive Animals Cooperative Research Centre. Australian Journal of Experimental Agriculture 46, 753-762
Gwatkin R. and Plummer P.J.G. (1946). Toxicity of certain salts of sodium and potassium for swine. Canadian Journal of Comparative Medicine 10, 188-190.
Hone, J. and Kleba, R. (1984). The toxicity and acceptability of warfarin and 1080 poison to penned feral pigs. Australian Wildlife Research 11, 103-111.