How are operations monitored?
The operational effectiveness of most pest control operations is assessed using some form
of population index. In the case of possum, control agencies typically use the "Residual
Trap Catch" (RTC) method. The "residual" component of the method refers to the fact
that the catch rate is usually determined after a control operation and hence the residual
population is being assessed, however, the method can be used at any time. If catch rates
are obtained before and after an operation the difference can be expressed as a kill rate
e.g. a 90% reduction could be referred to as a 90% kill.
The RTC method involves lines of traps that are set according to a fixed protocol with the
number of traplines related to the size of the operation being assessed. Trap catch results
are usually presented in the form of a percentage that represents the average catch rate
per trap line per night. E.g. if the traplines are 20 traps long and each trapline catches on
average 1 possum per night then the rtc figure would be 5%. Catch rates are sometimes
also used to index other pest populations e.g. mice, rats and stoats but for most DOC
operations these pests are assessed using "tracking tunnels". Tracking tunnels record the
presence of a pest animal by foot prints left on paper after the pest has walked across an
inked pad. Results are usually presented as a percentage of the tracking tunnels that have
been visited by the particular pest.
The physical ability of contractors to undertake trapcatch monitoring in most terrain does
not equate to similar contractors or staff being able to effectively and efficiently control
pests in that same terrain. While trapping and other ground based control methods can be
used in most situations they are generally less effective for any given amount of resource
than aerial baiting undertaken with 1080. The Department of Conservation typically
estimates that costs for effective control of possums alone using ground based methods in
back-country areas is 2-4 times more expensive than aerial baiting. Effective ground
based control of the full suite of pests now being controlled with 1080 (possums, rats,
stoats, ferrets) is more expensive again. A recent quotation from an experienced
contractor estimated that ground-based control (using 1080 in bait stations) of rats, stoats
and possums in the Owhango water supply catchment would cost $109/hectare whereas
the Department's typical operational cost for aerial 1080 baiting in this area is about
$15/hectare.
While initiatives such as alternative toxins, better traps and self-resetting traps will all
improve the effectiveness and therefore the efficiency of ground-based control we also
note that research and development is also leading to increased effectiveness and lower
costs for aerial baiting."
IN THE MATTER OF the Resource Management Act 1991
AND
IN THE MATTER OF an application by CANTERBURY REGIONAL
COUNCIL for resource consent to discharge sodium monofluoroacetate and pindone to land and to water via aerial application, and to land within the bed of a river, lake or artificial watercourse via ground control methods
STATEMENT OF EVIDENCE OF CHARLES THOMAS EASON IN SUPPORT OF APPLICATION
THE 6th DAY OF SEPTEMBER 2010
WYNN WILLIAMS & CO Applicant’s Solicitors
SOLICITORS Level 7, BNZ House, 129 Hereford St
CHRISTCHURCH PO Box 4341, DX WP 21518
CHRISTCHURCH
Solicitor: Margo Perpick Tel 0064 3 3797 622
1. My full name is Charles Thomas Eason. I am a research scientist and Professor of Wildlife Management at Lincoln University and specialise in Toxicology. I am also the Research Director for Connovation Ltd a small company in Auckland which makes pest control products. Connovation Ltd does not manufacture 1080 or pindone based products, and its focus is on alternatives to 1080. My current research focus is on alternatives to 1080.
Code of Conduct for Expert Witnesses
2. I acknowledge that I have read the code of conduct for expert witnesses contained in the Environment Court’s Practice Note dated 31 March 2005. I have complied with it when preparing my written statement of evidence and I agree to comply with it when I give any oral evidence.
3. Except where I state that I am relying on the specified evidence of another person or agency, the evidence in this statement is within my area of expertise.
Qualifications and Experience
4. I have a PhD from the University of Surrey, UK. My scientific expertise is in toxicology, with specialist expertise in the toxicology of vertebrate pesticides. I have researched, over a 20 year period, many aspects of vertebrate pesticides including anticoagulant toxins, 1080 and their fate and effects in the environment and I have published over 40 papers on 1080 including research and literature review papers.
Ambit of My Evidence
5. This evidence provides information about:
a. the substances sodium fluoroacetate ("1080") and pindone and their use as vertebrate pesticides in NZ;
b. the outcome of the reassessment of 1080 by the Environmental Risk Management Authority (ERMA);
c. the ecotoxicology of 1080, including its fate on land, in plants and in water and its effects on non-target terrestrial and aquatic species, and target-species possums and rabbits, including its humaneness;
d. the potential for human exposure to 1080 and possible effects on human health; and
e. the ecotoxicology of pindone and its effects on non-target species.
Facts Relied Upon and Formation of Opinions
6. The data, information, facts, and assumptions I have considered in forming my opinions are set out in the part of the evidence in which I express my opinions, together with the References and Bibliography section of my evidence.
7. I have not omitted to consider material facts known to me which might alter or detract from the opinions I have expressed.
8. The literature or other material which I have used or relied upon in support of my opinions is set out in the References and Bibliography section of my evidence.
The substances sodium fluoroacetate ("1080") and pindone and their use as vertebrate pesticides in NZ.
Sodium Fluoroacetate ('1080")
9. The toxicity of fluoroacetate containing compounds was first noted in the 1930s, leading to their patenting as potential rodenticides prior to the Second World War. US led research programmes in the early 1940s identified sodium fluoroacetate as the most promising candidate (Atzert 1971). Sodium fluoroacetate is a ‘salt’ form of fluoroacetate, used because it is more soluble in water and thus readily formulated into bait products. Once in water, the sodium bond dissociates, leaving fluoroacetate in solution. It received the number 1080 at a Research Centre in the USA and is now commonly known as Compound 1080, or just 1080. Its chemical formula is CH2FCOONa. It is an odourless, white, non-volatile powder that is highly
water soluble.
10. When ingested, 1080 is converted in cells to fluorocitrate. Fluorocitrate inhibits enzymes in the Krebs cycle, which is a metabolic pathway that converts food to the energy used for normal cell functions. Sufficient inhibition of energy production in this way results in either cardiac or respiratory failure (Savarie 1984).
11. Up to the 1940s no one suspected that carbon-fluoride compounds had existed in nature before they were synthesised by chemists. In 1943 fluoroacetate was identified as the toxic component of a poisonous plant in South Africa (Dichapetulum cymosum) (Marias 1944; Peters 1957). The leaves of this plant are very toxic to animals. Manufactured 1080 for use in toxic baits is toxicologically identical to the fluoroacetate found in some poisonous plants (de Moraes-Moreau et al 1995).
12. Fluoroacetate also occurs naturally in some 40 plant species in Australia (Twigg 1994; Twigg et al. 1996a, b; 1999). The highest fluoroacetate
concentration so far reported from a living source is 8.0 mg/g (8,000 ppm) in the seeds of the African plant Dichapetalum braunii (O'Hagan et al. 1993). This is higher than the concentration of 1080 in 0.15% possum baits (1,500 ppm). Fluoroacetate appears to be one of the many secondary plant compounds that have evolved at high concentrations as a defence
mechanism against browsing animals (Twigg 1994). Most studies assessing fluoroacetate concentrations in plants have focused on those species that are overtly toxic to mammals. However, it would appear that the ability of plants to synthesise fluoroacetate may be more widespread than generally supposed, since fluoroacetate occurs at extremely low concentrations in some Finnish plants (Vartiainen & Kauranen 1980), in tea leaves (Vartiainen & Kauranen 1984), and guar gum (Vartiainen & Gynther 1984; Twigg et al. 1996b).
13. 1080, the water soluble sodium salt of fluoracetate is manufactured in the US, and in New Zealand is made into baits or applied as a stock solution to carrots. 1080 has been used in New Zealand for pest control since the 1950s. It is registered for aerial control of possums and rabbits. Baits and stock solutions are prepared by Animal Control Products based in
Wanganui. The concentrations of 1080 used in bait for rabbit control are 0.04% in pellet bait and 0.02% in carrot bait, less than the 0.15% concentration used in possum bait.
14. Its use is a reflection of unique pest problems within New Zealand and its effectiveness at reducing the unwanted impacts of introduced animals. The most extensive use of 1080 occurs in New Zealand, followed by Australia (PCE 1994; Cavanagh & Fisher 2005). 1080 is registered for use in Australia, Canada, the United States and Israel. All of these countries have valued non-target species, such as native mammals, that would be at risk from 1080 use. Consequently they use less 1080 than in New Zealand. 1080 is restricted in many countries including the United States, because of its effects on native mammals and predators during poisoning operations. In
New Zealand, we do not have these complicating factors to such an extent. There are only two native mammals (bats).
15. In New Zealand, 3500 tonnes of agrochemicals (active pesticide ingredient) are used annually, most of which are herbicides (Manketelow et al. 2005). When compared to overall agrochemical use the amount of 1080 active ingredient used per year is small (1.0 to 3.5 tonnes (Innes & Barker 1999)), less than 0.1% of the total pesticide used per annum in New Zealand.
Pindone
16. Pindone is less widely used than 1080. It is the common name for 2-pivaloyl-1,3-indandione. The formula for pindone is C14H14O3.. It is a yellow crystalline
powder with low solubility in water. The sodium salt (‘Pival’ or pindone-sodium) has the formula C14H13NaO3 and is water soluble.
17. Pindone was synthesised in 1937 (Beauregard et al. 1955) and developed as a pesticide in the early 1940s. It has the same mode of action as warfarin. Warfarin is a rodenticide and a drug with the trade name (Marevan®). Both warfarin and pindone are anticoagulants or blood thinning agents. They inhibit the production of Vitamin K-dependent blood clotting agents in the liver which can lead to fatal haemorrhages in animals when sufficient is ingested.
18. Pindone has been used worldwide to control rodents, though its use for the control of rats and mice has decreased following the introduction of more potent anticoagulant rodenticides, such as brodifacoum. In New Zealand pindone has been used to control wallabies and possums, but in both Australia and New Zealand it has proved most effective for rabbit control (Eason & Jolly 1993).
19. Pindone, unlike 1080, has a delayed onset of action, and poisoning may not become evident for some days after ingestion of a lethal dose. Poisoning with pindone can occur with a large single dose, but it is more effective when rabbits ingest bait at smaller doses over a period of four to 12 days. Deaths following pindone poisoning are likely to occur after a week or two. Because pindone is slow acting, Vitamin K1 can be administered as an
effective antidote if accidental poisoning occurs.
20. Formulations containing pindone are registered for possum and rabbit control in New Zealand. Cereal pellet formulations for rabbit control contain 0.25%-0.5 % (0.25g/kg) pindone. A liquid concentrate formulation containing
the sodium salt form of pindone is also registered, for addition to chopped carrot or oat baits for rabbit control. This contains pindone at 3.4% (34g/litre).
21. Both pindone and 1080 baits are sensible choices for aerial application in bait for the control of rabbit as they are effective and not persistent.
22. The use of strychnine and arsenic for rabbit control has been discontinued for reasons of persistence, humaneness and health hazards to operators preparing baits. Other compounds such as brodifacoum have been trialled in the past for the control of rabbits (Williams et al. 1986), and whilst
brodifacoum is as effective, it is extremely persistent. Pindone and 1080 are a better choice as they will not bioaccummulate in the food-chain (Eason et al 1994a; Fisher et al. 2003).
23. New low residue toxins are being developed, primarily for possums and predators. These include novel rodenticides, para-aminopropiophenone (PAPP) for predators, such as stoats, and zinc phosphide and sodium nitrite for possums (Eason and Ogilvie 2009). None of these new toxins, which are being researched as alternatives to 1080 are yet registered. When or if they are approved by ERMA they will be registered for ground control and not for rabbits. These toxins are being researched for possums and predators in the first instance and for ground control. None of the new alternatives are likely to be available for aerial application for at least 6 years.
The outcome of the reassessment of 1080 by the Environmental Risk
Management Authority (ERMA):1
24. In March 2002, ERMA decided that there were grounds for reassessment of 1080. The grounds for reassessment of 1080 were: the increase in the amount of 1080 being used and planned for use, the completion of significant research on 1080 since it was first registered, and the level of public concern regarding its use. The Animal Health Board (AHB) and Department of Conservation (DOC) submitted an application in 2006, seeking approval for the continued use of 1080 for the control of possums, rabbits, wallabies and stoats.
25. ERMA completed an Evaluation and Review (E and R) report based on the information provided by the applicants and public submissions were sought.
1
information derived from a report produced from ERMA (2007) and details of the reassessment process and outcomes from the ERMA New Zealand web site.
A total of 1406 submissions were received and reviewed and public hearings were held throughout New Zealand.
26. In August 2007 the ERMA Committee released their decision approving the continued use of 1080. In reaching its final decision, the Committee considered the application, which covered 1080 toxicology, ecotoxicology, fate and effects, including both positive and adverse effects and the rationale for its use, all the written submissions, oral presentations made at the hearings, the E&R report and a report produced by ERMA’s Māori Advisory Committee.
27. The ERMA Committee considered that the use of 1080 was necessary to manage the threat of possums to native plants and birds and to control the spread of bovine tuberculosis. While the focus of the ERMA assessment was the use of 1080 to eradicate possums, its use to control other pests such as rabbits, rats and stoats was also acknowledged. The Committee recognised the environmental benefits from the use of 1080 to control possums and others pests in terms of protection of native fauna and flora and benefits for the market economy and farming communities from their reduced incidence of bovine TB and reduced competition for grazing from pest species, primarily rabbits. The committee also recognised that adverse effects and potential adverse effects could occur to non-target species, hunting resources and Maori values.
28. The Committee concluded that overall the benefits of using 1080 outweighed the adverse effects, with continued use of 1080 expected to have benefits for New Zealand’s environment. The Committee decided to allow the continued use of 1080. Additional controls were placed on the use of 1080 especially around aerial application of bait, and a tighter management regime is now required for aerial application of 1080. There were four key parts to this tighter management regime:
a. The establishment of a watch list, requiring reports on aerial
operations to be provided to ERMA to enable active monitoring of all future 1080 aerial operations;
b. the strengthening of existing controls to further mitigate risks involved in 1080 aerial drops;
c. the promotion of best practice in relation to pre-operation planning, consultation and notification, and management of 1080 aerial operations;
d. recognition that further research should be undertaken, both on alternatives to 1080, and on the effects of 1080 in specific topic areas where there remains a lack of knowledge and a degree of public concern.
The need for improved consultation was emphasized (ERMA 2007; Ogilvie et al 2010).
The ecotoxicology of 1080, including its fate on land and in water and effects on non-target terrestrial and aquatic species, both animal, and plant and its effects on target species, possums and rabbits
i) Fate and effects in soil
29. Two factors will affect the fate of 1080 in baits and on land, firstly dispersion and dilution of 1080 in decaying baits by rainwater and secondly
biodegradation.
30. 1080 is very water soluble and can be contrasted with persistent organic compounds such as DDT which can be in soils for many years. After leaching from baits, 1080 will be dispersed by water (Parfitt et al. 1995; Booth et al 1999a). The leaching of 1080 from bait occurs rapidly, when compared to other toxins because of its high water solubility. In a trial with cereal bait, leaching began after 20 mm of rain. If rainfall follows the use of 1080 baits, dilution to very low concentrations, below current limits of detection in bait or in soil, may precede biodegradation. Soil concentrations under baits reached a maximum concentration after 100 mm and these soil concentrations declined to near the limit of detection after 250 mm. In comparison to cereal bait, 1080 is retained in carrot baits for longer. In a simulated rainfall experiment carrot baits showed no decline in 1080 concentration after 200 mm of rain, indicating 1080 will slowly leach from carrots into the soil (Bowen et al. 1995).
31. Thomas et al. (2004) subjected 12 g carrot baits containing 1.5g/kg 1080 to two different simulated rainfall treatments. In the first treatment the 1080 in the carrot leached out, with the carrot losing approximately 74% of the 1080 after 10 mm of simulated rainfall. The second treatment, which was designed to be more representative of field operations, involved starting the simulated rainfall 48 hours after the 1080 was applied to the carrot. The carrot in this treatment retained more than 60% of its 1080 concentration after 500 mm of simulated rainfall. This is consistent with the early work by Bowen et al. (1995) and confirms extra care is needed when restocking after carrot baiting.
32. When 10-12 g of 0.15% 1080 Wanganui #7 cereal pellets were exposed to a simulated rainfall of 20 mm/hour, most of the 1080 concentration was retained after exposure to 50 mm of rain. The 1080 concentration rapidly declined in the pellets over the following 50 mm of rainfall. By comparison, the 1080 concentration in 10-12 g of 0.15% RS 5 pellets declined at a steady rate. By 100mm the 1080 had completely leached out of both types of pellets (Thomas et al. 2004). The 10-12 g cereal pellets in this study retained more 1080 when exposed to <100 mm of simulated rain than the 6 g cereal pellets examined by Bowen et al. (1995). Ogilvie et al. (2004) reported that
Wanganui #7 pellets lying on the ground in the field had a 99% reduction in their 1080 concentrations after 56 days. Over this time period 110 mm of rain fell.
33. Thomas et al. (2004) analysed bait breakdown rates from data collected during 19 operations using 0.15% 1080 Wanganui #7 cereal pellets and 11 operations using 0.15% 1080 RS 5 cereal pellets. Bait sizes used in the operations ranged from 3 – 12 grams. Most of the 1080 content, of both bait types, was removed following 150 – 200 mm of natural rainfall.
34. Once baits have degraded and 1080 leached into the soil there should be very low risks to non-target species. Not surprisingly residues have been found in soil under decaying baits.
35. 1080 can be metabolised (broken down) by some soil bacteria and fungi (Bong et al. 1979; Walker & Bong 1981; King et al. 1994). Not all micro-organisms can defluorinate fluoroacetate and the rate of metabolism differs with different species of soil bacteria and fungi (King et al. 1994). Enzymes capable of degrading fluoroacetate have been isolated from a number of micro-organisms. The fluoride-carbon bond is cleaved and ultimately, enzyme-bound intermediates form non-toxic metabolites such as glycolate (O’Hagan & Harper 1999).
36. All soils and natural freshwater so far examined in New Zealand possess populations of bacteria or fungi capable of detoxifying 1080 by release of fluoride ions from 1080. Pseudo-monad bacteria are also capable of degrading fluorocitrate which is responsible for the toxic effect of 1080 in animal cells. The breakdown to release fluoride ions from fluoroacetate and fluorocitrate occurs within the bacteria and fungi. The defluorination of 1080 by micro-organisms would probably elevate the fluoride levels in water or soil, but not to a measurable extent.
37. Under favourable conditions, such as 11–20ºC and 8–15% moisture, 1080 may be significantly defluorinated to non-toxic metaboilites in 1–2 weeks. In
less favourable conditions, breakdown might take several weeks and in extreme cold and drought, 1080 residues might persist in baits or in the soil for several months (King et al. 1994). Recent research has confirmed temperature dependent breakdown over the range 5-20ºC and has identified that the reduction of 1080 concentrations in soil is principally through transformation to hydroxyacetic acid (Fisher and Northcott 2010).
38. 1080 has been measured in soils after aerial application. The highest concentration of 1080 recorded in soil or leaf litter samples after aerial baiting was 0.19 mg/kg (ppm) (Wright et al. 2002), but mostly when 1080 could be detected concentrations were below 0.01 mg/kg (ppm). In a later field trial concentrations of up to 16 mg/kg (ppm) were found in leaf litter (Spurr et al 2002) though most samples had lower concentrations. The toxicity of 1080 to earthworms has been evaluated in separate laboratory studies. A lowest observable effect concentration (LOEC) of 90 mg/kg/soil (ppm) is much higher than that likely to occur in practice. No mortality was observed in worms exposed to a 1080 concentration of 865 mg/kg (ppm), and concentrations of 1,000 mg/kg (ppm) in soil did not affect mineralisation of nitrogen by soil micro-organisms (O’Halloran et al. 2005).
39. In conclusion to this section of my evidence, 1080 reaching soil from baits will be degraded or be diluted, it will not persist and will, when present in soil, be at low concentrations that will not adversely affect soil organisms.
ii) Fate and effects in plants
40. 1080 that has leached into soil may be absorbed by plants (Atzert 1971; Rammell & Fleming 1978). Research on plants has focused on determining whether 1080 residues might have an adverse effect on plant growth and whether or not uptake of 1080 by food plants could be at concentrations that would adversely affect animal or human health. In laboratory studies lettuce seedlings were adversely affected at 1080 concentrations in soil of 7 mg/kg (ppm), a significantly greater concentration than found in soil after the aerial application of bait (O’Halloran et al. 2005). To investigate secondary poisoning risks to livestock if they consume plants that have taken up 1080 leached from bait, concentrations of 1080 were assessed in broadleaf and perennial ryegrass following simulated baiting (Ogilvie et al. 1998). Both species absorbed 1080, which reached a peak around 10 days and then declined in 5-6 weeks to below the limit of detection, supporting previous findings that plants can degrade 1080 (Preuss & Weinstein 1969; Ward & Huskisson 1969). The concentration achieved in broadleaf and ryegrass
would be most unlikely to cause poisoning of livestock and other browsing herbivores (Ogilvie et al. 1998).
41. Research has been undertaken to determine whether plants used by Māori for food or medicines can take up 1080 and whether 1080 will persist in these plants. Cereal baits containing 1080 were placed at the base of pikopiko and karamuramu. Plants were sampled at various times up to 56 days and analysed for 1080 content. No 1080 was detected in any of the pikopiko samples, but it was detected in karamuramu at a maximum concentration of 5 (µg/kg) ppb after 7 days, and 2.5 (µg/kg) ppb after 14 days. Concentrations declined to zero at 28 days (Ogilvie et al. 2006). A similar study was carried out on two commonly eaten plant species, puha and watercress (Ogilvie et al. 2006). Samples of plant tissue were taken at various times up to 38 days for puha, and up to 17 days for watercress. 1080 was detected at a maximum concentration of 15 (µg/kg) ppb in a single puha sample after 3 days, and at a maximum of 63 (µg/kg) ppb from a single watercress sample on day 8. These concentrations are a very small fraction of the concentrations found in poisonous plants namely 8.0 mg/g (8,000 ppm) or in standard 0.15% possum baits (1,500 ppm). By day 38, no 1080 could be detected in puha; 1080 was not detected in watercress after day 8 (Ogilvie et al. 2009). Trace amounts of fluoroacetate were found in some puha samples taken prior to the experiment, indicating that fluoroacetate may occur at very low concentration naturally (Ogivlie et al. 2009), as it does in tea and guar gum.
42. In conclusion to this section New Zealand plants may absorb residues of 1080, but it will not persist in them and concentrations of 1080, after aerial or ground baiting, will not be high enough to affect the plants themselves or animals or people feeding on the plants.
43. I also understand that as a part of the proposed conditions of consent warning signs are to be erected at every place where people normally obtain access to areas which have been baited, and are to be maintained until all poisoned baits have completely broken down and for a minimum period of 6 months following baiting. I consider this and procedures outlined in the evidence from Steven Palmer to be appropriate to minimise risk of livestock or human exposure to 1080 residues from baits, soil or plant material.
iii) Fate in waterways
44. Drinking Water Standards were issued in 2000 by the Ministry of Health (MoH); the provisional maximum acceptable value (PMAV) for 1080 in water
is 3.5 ppb (µg/L). As a precautionary measure, MoH also recommended that water taken from catchments sown with 1080 baits should not be used for human supply until tests show that the concentration of 1080 is below 2 ppb (µg/L).
45. 1080 is highly water-soluble (Parfitt et al. 1995). Laboratory studies have shown that 1080 is biodegraded by aquatic plants and micro-organisms in biologically active water in 2-6 days, Parfitt et al. (1994). Eason et al. (1993) showed that 1080 declined by approximately 70% in 1 day and to below detectable limits in 4 days in aquaria containing plants and invertebrates. Research has also indicated that residues of 1080 do not persist in aquatic plants (Ogilvie et al. 1996, 1998; Booth et al. 1999), indicating degradation of 1080 within the plants. Some of these studies (Ogilvie et al. 1996; Booth et al. 1999) deliberately used solutions containing high concentrations of 1080 to simulate worst case scenarios.
46. Water monitoring of streams and waterways after aerial applications of 1080 baits was initiated by research teams studying the fate of 1080 in soil, invertebrates and water in the early 1990s (Eason et al. 1992, 1993). Results of the initial research and subsequent monitoring demonstrated that there has been no evidence of 1080 presence in reticulated water and no evidence of significant or prolonged 1080 contamination in surface waters (Parfitt et al. 1994; Eason et al. 1992, 1993; Hamilton & Eason 1994;
Meenken & Eason 1995; Booth et al. 1997; Eason et al. 1999; Meenken et al 2000; Eason & Wright 2001; Wright et al. 2002; Booth et al 2007; Eason & Temple 2008). Between 1990 and 2008, Landcare Research analysed over 2,000 water samples. Of the samples taken, 96% were free of detectable 1080. Of the 4% of water samples that contained detectable 1080 residues, most had 1080 concentrations of less than 1 µg/L (ppb). When 1080 was found in samples in monitoring programmes, it was at very low
concentrations, typically in small streams in remote locations. Often baits were seen near where samples were taken. The contamination was transient and not usually picked up in repeat samples.
47. NIWA researchers have deliberately spiked small streams with 1080 baits. They monitored the fate of 1080 (Suren & Lambert 2004, 2006; Suren 2006; Suren & Bonnett 2006). In these field experiments, quantities of baits were placed in small streams <3 metres wide. 1080 was detected in water samples for a short period thereafter (<24 hours). 1080 concentrations 10 metres downstream from the site where baits were added were higher than those 100 metres downstream, illustrating the influence of dilution. Observed concentrations were below the MoH guidelines of 2 µg/L (ppb).
48. Suren from NIWA (2006) recommended that where there were concerns, water samples should be taken within 4-8 hours of streams being potentially contaminated with bait, and suggested that later monitoring was likely to be of limited value. All the research groups have reported that when
contamination does occur, only low concentrations are detected and only for a short period. I understand that ECan is proposing a consent condition whereby samples of surface water shall be taken from a representative site where bait are distributed within 200 metres of a domestic or community water supply may enter water, within 4 to 8 hours after the discharge event has ceased or as determined by the Medical Officer of Health. I consider this and the precautions outlined in Stephen Palmer’s evidence to be appropriate.
49. In summary, there are two means by which any 1080 residues present in waterways will be reduced to undetectable and toxicologically insignificant amounts: i) dilution, and ii) biodegradation. Even though biodegradation can occur over the first 24 hours of 1080 entering water, the effect of dilution will be immediate and more profound. The research by NIWA scientists led by Dr Alastair Suren indicates that 1080 will have negligible effects on water quality.
50. It is noteworthy that in a worst case scenario, if 50 kg of 1080 bait were accidentally dropped into a farm pond during a helicopter aerial operation, 75 grams of poison would enter the water. If the pond was 100 square metres in area, with an average depth of one metre, an adult would have to drink more than 400 litres of the pond water at one time in order to receive a lethal dose.
iv) Effects on aquatic species
51. The concentrations described above in “real world” operation settings, as opposed to laboratory investigations are consistently well below those that will impact on aquatic organisms. Aquatic toxicity tests were completed in the USA. The first estimated the acute toxicity of 1080 to bluegill sunfish. No mortality or sub-lethal effects were observed at any concentration tested, with a highest NOEC (no observed effect concentration) of 970 mg/L (ppm). Based on the results of this study and criteria established by the US Environmental Protection Agency (EPA), 1080 would be classified as practically non-toxic to bluegill sunfish. The second test on rainbow trout used the same test conditions as the bluegill sunfish studies. The NOEC was 13 mg/L (ppm), which the US EPA classifies as slightly toxic to rainbow trout.
The third test estimated the acute toxicity (EC50) of 1080 to the small
freshwater invertebrate Daphnia. The 48-hour EC50 value for daphnids
exposed to 1080 was 350 mg/L (ppm) and the NOEC was 130 mg/L, making 1080 practically non-toxic to Daphnia by US EPA classification standards (Fagerstone et al. 1994). These concentrations of 1080 described above are many times higher than transient residue concentrations, found rarely after 1080 use.
52. Suren & Lambert (2006) from NIWA studied the responses of an invertebrate community and monitored caged longfin eels, koaro and upland bullies after deliberately spiking small streams. No effects of 1080 were detected on any of the invertebrate species, including caddisflies, midges and mayflies, or the three caged fish species. Further work on eels and freshwater crayfish when the animals were fed 1080 baits under experimental conditions designed to mimic real-site situations, also showed no effects (Lyver et al. 2005; Suren & Bonnett 2006) and residue concentrations decreased quite quickly over time as they do in mammals and birds (see below).
53. In conclusion there should be no toxic effects on fish or their food.
v) Effect, welfare and fate in animals
54. Any substance is toxic if taken in a high enough concentration or dose. Common salt is toxic to humans at a dose of about 4g per kilogram. The toxicity of a substance to an animal is generally defined in terms of LD50. The
LD50 is a statistically derived estimate of the smallest amount of poison
required per animal to kill 50% of the population. For example, the LD50 of
1080 for possums is usually quoted at 0.8 mg per kg. This implies that if 100 possums are each fed 0.8 mg of 1080 per kilogram body weight then 50 of the animals would be expected to die as a result. A milligram (mg) is one thousandth of a gram, and a kilogram (kg) is a thousand grams. The LD50 of
1080 for animals varies widely from as low as 0.05mg/kg for the Texan pocket gopher to over 500mg/kg for the South African clawed toad. Dogs are especially susceptible to 1080 with an LD50 of 0.07mg/kg. For herbivores
including sheep, cows and goats, the LD50 is 0.4 to 0.6mg/kg.
55. In pest species, the LD50 of sodium fluoroacetate is usually 1 mg/kg or less,
whereas in humans, LD50 doses have been estimated to be 2-5 mg/kg
(Rammell & Fleming 1978). Livestock can also succumb to 1080 poisoning (ERMA 2008) and must be kept well away from baits. Even partially degraded baits should be regarded as hazardous to sheep and cattle. Where livestock deaths have been reported, it has generally been due to
animals being returned too early to areas where 1080 application has occurred (Eason 2002). Sheep exposed to single sub-lethal doses of 1080 have not been reported to experience any long term adverse affects (Eason 2002).
56. Other mammals are also susceptible and dogs are extremely susceptible to 1080 (Atzert 1971; Rammell & Fleming 1978; Eisler 1995). Non-target animals including pets, livestock and working dogs must be kept clear of bait and dogs must not be allowed contact with poisoned rabbit or possum carcasses. There is no antidote to 1080. Veterinary treatment of poisoned dogs is challenging, frequently not successful and relies on symptomatic and supportive care (Parton et al. 2006). For example, poisoned dogs are heavily sedated to reduce the possibility of seizures, and fluid levels are maintained to offset electrolyte imbalances (Parton et al. 2006).
57. The mitigation measures that ECan is proposing as conditions of consent, and procedures described by Steven Palmer, such as notices and
consultation and also the fact that the application is for discharge of baits on private, not public land, are appropriate. Landowners are able to manage and keep pets, livestock and working dogs away from baited areas.
58. 1080 poisoning can be lethal or sub-lethal. In mammals, signs of poisoning usually become evident between ½ and 3 hours after ingesting 1080, related to the time taken for 1080 to be absorbed by the gut, and transported through the blood and to reach mitochondria within cells. Animals receiving a sub-lethal dose show mild symptoms, and metabolise and excrete 1080 over a number of days and recover (Eason 2002). Animals receiving a lethal dose show more severe symptoms, and in general herbivores appear to die of cardiac failure, whereas carnivores die of respiratory failure associated with central nervous system disturbances and symptoms such as convulsions (Egeheze & Oehme 1979).
59. The humaneness of 1080 has been debated (Sherley 2007; Twigg and Parker 2010). 1080 is not as humane as cyanide (Sherley 2007). The symptoms of 1080 poisoning depend on the species. Researchers that have reported 1080 to be a humane poison do so in part based on observation of subdued behaviour in herbivores and death from cardiac failure (Batcheler 1978; Morgan 1990). In dogs, central nervous system disturbances are marked, and dogs appear distressed with convulsions and seizure. Possum and rabbit bait formulations are optimised to induce as swift a death as possible. While 1080 is not as humane as cyanide, it is a lot preferable to
anticoagulants like brodifacoum for possum control from a welfare perspective (Littin et al. 2002; Littin et al. 2009).
60. 1080 is absorbed through the gastrointestinal tract, or via the lungs if inhaled. It is not readily absorbed through intact skin (Saunders et al. 1948; Pattison 1959). Prolonged persistence of 1080 in animals after sub-lethal exposure will not occur. In studies on rabbits, goats, possums, and sheep receiving sub-lethal doses (Gooneratne et al. 1994; Eason et al. 1994a) 1080 was excreted within a few days. If recommended practices are followed in possum or rabbit control operations, 1080 is unlikely to be present in livestock, and hence meat for human consumption (Rammell 1993).
61. In these studies highest concentrations occurred in the blood, with moderate levels in the muscle and kidneys, and the lowest concentration in the liver. In sheep, the highest concentrations in blood occurred 2.5 hours after dosing and there were negligible amounts of 1080 in tissue and plasma 4 days after dosing. The elimination half-life in blood is 11 hours or less in sheep (11.0 hours), goat (5.5 hours), rabbit (1.1 hours), and possum (9.0 hours) (Eason et al. 1994a), and contrasts with longer elimination half-lives in blood and liver for brodifacoum in possums (Eason et al. 2006).
62. Mallard ducks dosed with an 8 mg 1080/kg sub-lethal dose substantially eliminated the 1080 from heart muscle and blood within 24 hours (Ataria et al. 2000)
63. To further examine the persistence of 1080 and its potential to contaminate milk and meat, two groups of lactating ewes were fed 1080 baits (“high dose”, 0.25 mg/kg body weight; “low dose”, 0.01 mg/kg body weight) (Milne et al. 2002). 1080 concentrations in blood and milk were examined 4 – 96 hours after dosing. Four hours after dosing, 1080 concentration in blood (0.56 µg/ml) was 30 times greater than in milk (0.017 µg/ml) in the high dose group. Lower concentrations of 1080 were detected in the blood after 72 hours (0.020 µg/ml), and in milk, concentrations were near the analytical level of detection (0.0006 µg/ml or ppm). The low dose ewes had no detectable 1080 in blood or milk 72 hours after dosing (Milne et al. 2002).
64. While 1080 is comparatively rapidly eliminated from living animals, it can persist in carcasses for many months in cool or dry conditions where it will break down more slowly, and carcasses of possums or rabbits will pose a risk to dogs for some time after a possum or rabbit control operation (Meeken and Booth 1997, Wright 2004).
65. There are understandable concerns that meat will contain 1080 if livestock or game eat bait and receive a sub-lethal dose. Unlike organochlorine
pesticides, such as DDT or other animal toxins like anticoagulant
brodifacoum, 1080 is water soluble, not fat soluble, and therefore does not accumulate in fatty or other tissues of animals.
66. The risk of livestock being exposed to 1080 will be managed by farmers who are property owners. Application of bait on their land will not occur without their permission, allowing time for farm managers to remove stock from areas to be baited. Risk of contaminated game being procured is minimised by the rapid excretion of sub-lethal doses of 1080 by any animal exposed to bait, and by landowner involvement in baiting programmes, minimising the shooting of game when baiting has recently been completed.
67. In conclusion 1080 is effective at killing possums and rabbits. It is
moderately humane, when compared with other toxins for controlling animal pests. It is eliminated rapidly from living animals receiving a sub-lethal dose. Human exposure through consumption of contaminated meat could only occur if meat is procured from game or livestock within a relatively short time after an animal has eaten baits. Consent conditions and notification should minimise the risk of 1080 poisoning through the consumption of meat.
68. However, degradation in carcasses will be much slower in cold or dry conditions, and dogs must be kept clear of poisoned rabbits or possums. The proposed notice requirement which states that notices shall be maintained until all poisoned baits and carcasses have completely broken down or for a minimum of 6 months is appropriate to minimise risk to dogs.
vi) Effects on birds
69. 1080 is toxic to birds and bird deaths have been reported during 1080 pest control operations almost from when they were first conducted in New Zealand. The first monitoring of bird deaths occurred in 1970’s when raspberry-lured carrot bait that had a high percentage of small fragments or “chaff” was being used (Harrison 1978). This led to significant changes in the baits being used.
70. Fewer birds are killed today for a number of reasons:
a. increased use of cereal baits, shown by research to be less attractive to some bird species than carrot bait;
c. raspberry lure has been banned;
d. cinnamon oil is now used as a lure for possums and this is unattractive to birds;
e. baits are dyed green as this colour is less attractive to birds;
f. bait sizes are made too large for a bird to easily ingest whole; and
g. carrot is screened and application rates of baits have been reduced.
71. Loss of individuals in a population of native species as a consequence of 1080 poisoning can have variable significance to the long term viability of the population depending on the context. Those birds with a large population and/or a high rate of reproduction increase can compensate for small losses. Poison-related mortality may be replacing deaths from predation or winter starvation. Monitoring has focused on the impacts 1080 has on native bird species at a population level.
72. North Island robins suffered c. 50% mortality during a 1996 poison operation using 1080 carrots. However, the reduction in predators resulted in improved breeding of robins. A year later the robin population contained more birds and a greater proportion of females than just prior to the operation
(Powlesland et al. 1999). Similarly, during a 1080 carrot operation in 1997, mortality of tomtits was estimated to be 79% (Powlesland et al. 2000). However, the tomtits had enhanced nesting success following the 1080 operation, with pairs rearing two, and in some cases, three broods the following season. More recent research indicates that cereal bait operations using low sowing rates and large baits have little if any impact on tomtit populations (Westbrooke & Powlesland 2005).
73. A number of studies using individually marked birds have been undertaken on threatened native non-target populations to determine the effects of aerial poisoning, using carrots and cereal baits. Threatened species usually have a poor ability to recover from additional mortality, hence the focus on these species.
74. Most intensive monitoring has been of kokako, with four of 366 birds monitored through 1080 operations disappearing, giving a calculated mortality rate of 1.4% deaths per operation (Flux & Innes 2001). Over 200 radio-tagged adult kiwi have been monitored through 1080 operations with no deaths recorded (H. Robertson, pers. comm.). All 73 kaka, 19 blue ducks and 15 kereru monitored through aerial poisoning operations using radio transmitters have survived. During monitoring, one of 40 weka died, and four
of 23 fernbirds disappeared. None of these studies have identified population level mortality which threatens the viability of a species (Broome et al. 2009).
75. Seaton et al. (2009) collected productivity data from 87 falcon nests in the Kaingaroa pine plantation during three breeding seasons, 2003-2006. During this time 1080 pellets and carrots were ground laid or aerially applied in forest compartments where falcon bred. During the study the breeding falcon population increased from 20 to 36 pairs, leading to the authors concluding that 1080 did not have a negative impact on falcon, and probably had a positive impact by reducing predation pressure on the falcon.
76. Australasian harrier have not been monitored individually. However no detectable impact could be determined through five minute bird count monitoring before and after an aerial 1080 operation using cereal pellets on Rangitoto Island, and “the small resident population was still
seen…throughout the year following the poisoning” (Miller & Anderson 1992). Additionally, Pierce and Maloney (1989) found no evidence of dead harriers after aerial 1080 poisoning of rabbits in the McKenzie basin.
77. Blue duck are unlikely to eat cereal pellet baits and their aquatic invertebrate prey are unlikely to be contaminated by 1080. However studies have been done to determine their survival following aerial 1080 operations. There was no reduction in visual counts of blue duck in the Otira valley after application of 0.15% 1080 pellets at 6 kg/ha in 1989 (Spurr & Powlesland 1997). Additionally, all 19 radio-tagged blue ducks in Waihaha survived for at least four weeks following aerial application of carrot bait (0.08%) at 15 kg/ha (Greene 1998).
78. Kereru have been monitored during 5 aerial 1080 operations using cereal pellets and no population changes were detected using the five minute count method (Spurr & Powlesland 1997). Additionally, all 15 radio tagged birds exposed to an aerial 1080 operation using carrot bait survived (Powlesland et al. 2003).
79. During kea monitoring in 2008, ten radio-tagged birds survived an aerial 1080 operation in the Arawhata Valley and both kea radio-tagged in Horonu Range survived a second operation. However, seven out of seventeen kea were killed in a third operation, near Franz Josef. A research programme is underway to determine what the likely risk factors are for kea and to develop operational specifications that will eliminate or minimise kea deaths. Trials in winter 2009 showed there was no mortality in kea after operations at Mt Arthur (13 birds monitored) and in the Hawdon Valley (10 birds monitored) following adherence to these new operational procedures (pers comm. Josh
Kemp). Furthermore there is a kea habitat diagram in the application and this indicates that these habitats are not likely to be baited so the risk to kea should be reduced.
vii) Fate and effects on invertebrates
80. 1080 has been used experimentally as an insecticide for fleas, aphids and wasps and is toxic to invertebrates (David 1950; Notman 1989; Spurr 1991). Use of apple paste (‘jam’) bait formulations of 1080 found to be attractive to bees (e.g. Goodwin & Ten Houten 1991) was discontinued in 1995 because of the potential risks of bee mortality and contamination of honey.
81. Invertebrates have been seen eating baits. Hence research has been completed on the persistence of 1080 in invertebrates and effects on populations and individual insects. In a laboratory study tree weta eliminated residues over 6–10 days after exposure. Similar results were obtained for a native ant (Booth & Wickstrom 1999). Insects have been monitored in forests for 1080 residues after aerial sowing of toxic baits for possum control. No 1080 was found in earthworms, spiders, beetles, millipedes, or centipedes. 1080 was found in some cockroaches, bush weta and cave weta during the period the baits were on the ground. After 3–4 weeks (when baits could have been eaten at any time) all invertebrate samples were free from 1080 residues (Eason et al. 1993).
82. Persistence of 1080 in invertebrates appears to be short-lived, so the risk to insectivorous birds or other predators will also be confined to a short period after sowing baits. However, large invertebrates will eat bait (Spurr & Drew 1999) and, since species like weta can contain large amounts of bait, secondary poisoning via this route is possible.
83. In surveys conducted in the 1990s, no negative impact was detected on populations of weta in Waipoua Forest, a range of invertebrate species on Rangitoto Island, predatory insects in Mapara Reserve, or ground-dwelling invertebrates in Puketi Forest and Titirangi Reserve (Spurr 1994).
84. Similar results were obtained by Spurr & Berben (2004) using artificial refuges, (weta houses) and mark-recapture techniques. No impacts on weta, slugs, cockroaches and spiders were identified. Powlesland et al. (2005) has since confirmed that aerial 1080 will not have a detrimental effect on forest invertebrates.
85. Observations of the number of species and number of individual invertebrates found feeding on 1080 baits has led to the prediction that
vertebrate pest control operations are unlikely to have long-term deleterious impacts on invertebrate populations(Sherley et al. 1999; Spurr & Drew 1999). This conclusion is consistent with the monitoring of Aspen et al. (1999), which showed no impacts on the number of ground-dwelling insects up to 1 year after aerial 1080 application, and more recent research by Powlesland et al 2005, monitoring cave and tree weta, cockroaches and spiders.
viii) Effects on deer
86. Red deer kill during aerial 1080 poisoning operations has been monitored in several different areas, and was found to range from 5% and up to 93%, recorded after an aerial carrot operation in Pureora in 1997 (Fraser & Sweetapple 2000). Two-thirds to three-quarters of the fallow deer population in the Blue Mountains were killed during an aerial pellet operation (Nugent & Yockney 2004), the kill appearing to be greatest for the youngest (smallest) deer in areas with the highest deer density and the most open under-storey. There appears to be no consistent pattern in the number of deer deaths based on bait type, sowing rate or toxic loading suggesting other factors, such as deer density, are important (Nugent et al. 2001).
87. Deer by-kill can now be avoided, or substantially reduced, in practice by including a deer repellent in bait that does not deter possums (Morriss 2007). Crown land where most feral deer are found and hunted are not included in this application. I understand further details on the issue of deer by-kill are covered in Graham Sullivan's evidence.
ix) Potential for human exposure
88. Being highly toxic, 1080 has the potential to kill and cause sub-lethal effects in humans, but exposure of individuals or communities to amounts of 1080 that would cause such effects is most unlikely.
89. There is little information specifically about the effects of 1080 exposure on people, and most of the available data concerns relatively high exposures. The oral toxicity of 1080 to adult humans is estimated at 2 to 10 mg/kg.
90. A 15 kg child would have to eat about one to two 12 gram cereal baits containing 0.15% 1080 to receive a lethal dose. An 80 kg adult would have to eat eight to nine 12 gram baits containing 0.15% 1080 to receive a lethal dose. Although 1080 is potentially dangerous, strict safety procedures are in
place, and there have been no recorded deaths in New Zealand from accidental 1080 poisoning.
91. These effects of 1080 have been extensively studied. The organs most affected are those with high energy demands such as the heart and testes.
92. Animal studies and studies in cell cultures are used to address questions of sub-lethal toxicity and possible adverse effects on human health. For example, results of three different laboratory tests with bacteria and cultured cell lines indicate that 1080 is not mutagenic, and is thus not anticipated to be carcinogenic (Eason et al. 1999).
93. Cardiac muscle (heart) is also recognised as a target organ following sub-lethal exposure of mammals to 1080, with cardiac histopathology reported in herbivores especially (e.g. Gooneratne et al. 2008). When rats were orally administered sub lethal doses of 1080 for 90 days, there were 1080-related changes in heart weights of male and female rats so that the no-observable-effects level (NOEL) was 0.075 mg/kg/day, and the lowest-observable-effects level (LOEL) dose was 0.25 mg/kg/day (Eason & Turck 2002).
94. Toxicity data from animals shows that single or repeated sub-lethal doses of 1080 can have reproductive and/or developmental toxicity. 1080 caused developmental defects in foetal rats when pregnant females were exposed to relatively high doses of 1080 on a daily basis during days 6 through to 17 of gestation, when a significant proportion of foetal development occurs (Eason et al. 1999). Teratogenic effects of 1080 occurred at 0.75 mg/kg/day dose, but not at lower doses, so that the developmental no-observable-effects level (NOEL) was 0.1 mg/kg/day 1080 (Eason et al 1999).
95. 1080 has been described as a “male reproductive toxicant” because testes are a target organ in mammals (Eason & Turck 2002) The NOEL for testicular damages in rats administered oral sub-lethal doses of 1080 for 90 days was 0.075 mg/kg/day. The lowest-observable-effects level (LOEL) dose was 0.25 mg/kg/day (Eason & Turck 2002). For humans to get an equivalent dose a person would need to eat 1 to 2 baits per day.
96. Endocrine Disrupting Chemicals (EDCs) cause special concern and are classified by toxicologists as chemicals that have oestrogenic or anti-androgenic activity. EDCs like DDT and dioxins are complex molecules that mimic natural hormones by binding to their receptors in the body. Neither 1080 or fluorocitrate, the toxic metabolite displays oestrogenic or anti-androgenic activity (Tremblay et al 2004 & 2005, Twigg and Parker 2010).
97. Nevertheless it is apparent that 1080 exposure could have a multitude of effects resulting from aconitate hydratase inhibition and secondary
biochemical perturbations. Chronic toxicology studies in animals as human substitutes, and mutagenicity tests, enable elucidation of potential
toxicological effects in humans if they are exposed. As indicated earlier, pathological changes observed in chronic animal studies are the
consequences of primary and secondary effects of 1080 on the body. Animal toxicology studies by Eason & Turck (2002) have been used to define conservative no-effect levels and more recently, tolerable daily intake (TDI) values of 0.03 µg/kg/day (Faronda et al. 2007a, b) have been proposed based on a NOEL of 0.075 mg/kg/day and a LOEL of 0.25 mg/kg/day (Eason & Turck 2002). The TDI proposed by Faronda et al. (2007 a, b) is of
relevance to workers at the bait manufacturing plants but not to the broader community, which should not be exposed to 1080 under normal baiting practices. Weaver (2003) suggested additional toxicology studies would be valuable. Equally, or probably more importantly in a practical sense, the use of 1080 must continue to include safeguards that focus on those individually handling 1080 or 1080 baits to ensure they do not ingest, inhale or absorb 1080.
98. Risk to humans can be minimised by ensuring that exposure does not occur. Thus monitoring of the occurrence of 1080 residues in potential
environmental pathways of human exposure, such as drinking water and food, has been important. Communities should not be exposed, and workers in the industry must adhere to strict safety procedures and high standards have been set in terms of exposure limits (Faronda et al. 2007 a and b). Water monitoring conditions and conditions that place buffer zones around waterways and the precautions described in Steven Palmer’s evidence should provide further reassurances.
99. Animal studies have shown effects when animals were exposed to sub-lethal doses that underpin the strict safety procedures required for workers at factories where 1080 baits are manufactured and for pest control workers using 1080 in the field. Strict safety precautions are enforced, including the use of protective clothing and pesticide handling rules to protect workers in the pest control industry. Equally, exposure to 1080 by any member of the community should be avoided. Exposure of workers in the pest control industry, who are the members of the community who could be exposed to traces of 1080, has been monitored. Blood and urine samples have been analysed for traces of 1080 to test effectiveness of these safety procedures (Beasley et al, 2009).
100. In summary, fluoroacetate, the active ingredient of 1080, occurs naturally in toxic plants in Australia, South Africa, and South America. Manufactured 1080 is used to control introduced mammals in New Zealand and Australia. Since 1080 is highly water soluble, it will be dispersed and diluted in the environment by rain and stream water. Some micro-organisms in the soil, such as Pseudomonas species, may defluorinate 1080. Water-monitoring surveys, conducted during the 1990s, have repeatedly confirmed that significant contamination of waterways following aerial application of 1080 bait is most unlikely, and in most instances no traces of 1080 can be found in water. 1080 acts by interfering with the energy-producing tricarboxylic acid cycle in the mitochondria. Dogs are extremely susceptible to poisoning. If an animal has ingested a sub-lethal dose of 1080, toxin residues will not persist in meat, blood, the liver, or fat (in contrast to brodifacoum). High-quality baits reduce non-target impacts on birds. Current evidence suggests that
populations of common bird species and invertebrates are not adversely affected.
The ecotoxicology of pindone and its effects on non-target species Baits and soil
101. The fate, effects and ecotoxicology of pindone has not been as extensively studied as the fate and effects of 1080.
102. The fate of pindone in baits, like that of 1080 will also be affected by dispersion and dilution and secondly degradation. Nevertheless, it is likely that in wet weather cereal baits will degrade in a few days and carrots in a few weeks. In cold or dry conditions, baits containing pindone can remain toxic for months (Fairweather and Fisher 2009).
103. Unlike 1080 there is limited published evidence for biodegradation of pindone in soil or baits. Photodegradation is probable.
104. Pindone has low solubility in water. The sodium salt (‘Pival’ or pindone-sodium) is water soluble.
105. The sodium salt of pindone (Pival) used in liquid formulations for application to carrot, will be more readily washed out of bait and dispersed in soil than pindone. My understanding is that pindone rather than the sodium salt is used in cereal baits.
106. For pindone itself, photodegradation in baits and soil will be important for decreasing the concentration of residues in the environment. For the sodium
salt (‘Pival’ or pindone-sodium) dilution by water will contribute to decreases in concentrations in bait and soil.
107. In a comparative study, pindone was more slowly leached from oat grain bait for rabbits when compared with 1080, which is consistent with the lower solubility of pindone in water (Wheeler & Oliver 1978). In a rain-simulated study on pindone in cereal pellets pindone concentrations declined approximately 20% after 400 mm of rain. In the same study 1080 was substantially leached from cereal bait after 100 to 200 mm of rain. (Booth et al. 1999).
108. Pival will leach from the bait and be dispersed by rainwater. In a field trial in 1994 fresh carrot baits had concentrations of 185 mg/kg that declined to 108 mg/kg after a day in the field and to 88 mg/kg after 2 days. Residue
concentrations then increased to 134-135 mg/kg after 4 and 7 days, in dry conditions, with a subsequent decline to 71 mg/kg after 11 days following 26 mm of rain (Boswell 1995).
109. Once baits have degraded and pindone leached into the soil there should be very low risks to non-target species. Not surprisingly residues have been found in soil under decaying baits. The persistence of pindone or pival, the sodium salt in soil has not been studied. However results for
chlorophacinone and diphacinone, two closely related compounds have been used to estimate the half-life of pindone in soil to be approximately one month (NRA 2002). Residues of the water soluble sodium salt of pindone (Pival) are likely to be even less persistent in soil, particularly after rainfall.
Fate and effects in plants
110. There are no studies on the fate and effects of pindone on plants. Uptake by plants is unlikely for the non-water soluble pindone acid, but more likely for water soluble sodium salt of pindone (Pival). As is the case with 1080, if this does occur the concentrations in plants are likely to be very low and not likely to present a risk to livestock.
Water
111. Significant contamination of waterways following aerial application of pindone is unlikely, in the same way that significant contamination of
streams or rivers after aerial application of 1080 is unlikely. A small survey of water samples in a catchment area where pindone baits had been aerially
sown for rabbit control was completed in 1994. No pindone residues were detected. (Eason and Wickstrom 2001).
112. There should be low risk to aquatic species and humans where safety procedures and good baiting practices are adhered to.
113. There is information based on laboratory studies that pindone is toxic to some aquatic invertebrates and fish. For example, aquatic toxicity data shows that pindone is toxic to Daphnia at 5.6mg/L (ppm) over 48 hours (http:// msds.chem.ox.ac.uk/PI/pindone.html) and has moderate toxicity to bluegill sunfish and is highly toxic to rainbow trout, in tests where the fish were exposed over an 8 hour period to water containing different
concentrations of pindone (British Crop Protection Council 2000). The LC50 for rainbow and bluegill trout is 210.0 and 1600.0 µg/L (ppb) (Fairbrother and Fisher 2009). However, these results are not relevant to normal baiting practice when significant contamination of waterways with pindone is unlikely.
114. The proposed monitoring conditions are appropriate for pindone and should provide further reassurance.
Effects, welfare and fate in animals
115. Pindone is toxic to all mammals, and potentially toxic to birds. Its oral toxicity is substantially increased when it is ingested in repeated, consecutive daily exposures, with less pindone required for a lethal dose when oral exposure occurs over a number of days. Formulations containing pindone are registered for possum and rabbit control in New Zealand. Pindone acts like the other anticoagulant toxicants by interfering with the normal synthesis of vitamin K-dependent clotting factors in the liver.
116. Pindone is less potent than second generation anticoagulants, such as brodifacoum and this is related to a generally lower binding affinity to the liver when compared to second-generation compounds (Parmar et al. 1987; Huckle et al. 1988). This is significant since brodifacoum was researched as a potential rabbit control tool (Williams et al. 1986). Pindone is a better and safer choice.
117. For first-generation anticoagulants such as pindone, either very large single doses or repeated smaller doses are generally needed to induce death. A single dose of approximately 18 mg/kg is, however, sufficient to kill rabbits. In rabbits the repeat dose (7 days) LD50 is 0.52 mg/kg/day, while rabbits
receiving 1 mg/kg/day for 7 days will die. Three out of six rabbits dosed with a single higher dose of 25 mg/kg pindone died within 6 days (Eason & Jolly 1993).
118. By contrast, pindone doses of up to 12 mg/kg/day did not cause clinical or post-mortem haemorrhage in sheep (Oliver & Wheeler 1978). Possums appear to be even more resistant to pindone than sheep. None of 12 possums dosed at 8 and 16 mg/kg/day for 5 days died. One of 12 possums died when dosed with 32 mg/kg/day for 5 days, and 9 of 14 possums died when dosed with 64 mg/kg/day for 5 days. From these data an LD50 of 51 mg/kg/day for 5 days was calculated (Jolly et al. 1994). As a result pindone is more useful for rabbit control than for possum control.
119. Clinical signs of poisoning in animals will usually reflect some manifestation of haemorrhage. Symptoms include anaemia and weakness and
haemorrhaging may be visible around the nose, mouth, eyes, and anus. Anticoagulants are not considered to be as humane as 1080 or cyanide (Littin et al. 2002; Twigg and Parker 2010). Swollen, tender joints and convulsions can occur dependent on the site of the haemorraging. Poisoned animals will die usually of multiple causes associated with loss of blood.
120. I could not locate much toxicology data relating to the potential carcinogenicity or tetratogencity of pindone. A series of Ames tests, a recognised standard toxicology screening test, with and without metabolic activation, at different doses (0.1 to 100 µg/plate), all had negative results (Zeiger et al. 1987), indicating lack of genotoxicity.
121. Pregnant sheep exposed to pindone had significantly more stillborn lambs, or lambs failing to survive more than two days compared to untreated ewes. Four lambs (out of 127) produced by the pindone treated ewes had
moderate to significant deformities. Pindone also affected the motility of the sperm of treated rams (Robinson et al. 2005). Toxicological information of this type is of relevance to farmers, taking care not to restock too early and to workers in the industry to take care not to expose themselves to pindone. It has been estimated that sheep (approx 50 kg) could eat > 7 kg of 0.5 g/kg pindone bait before they received a lethal dose, and an 8 kg dog > 60 gm if the bait were eaten as a single feed. However if the animals ate the baits daily the amount of bait needed to cause toxicity would be substantially reduced, to slightly over 1 kg/day for sheep and 4 gm of bait per day for dogs (Fairweather and Fisher 2009). Despite these safety margins livestock and pets should be kept clear of baits, and workers should protect themselves from exposure. Other members of the community should not be exposed to
pindone. The proposed conditions of this application including consultation, notices and the fact that the consent application is only for private land should limit the opportunity for contact with bait.
122. Pindone when eaten in bait is absorbed through the gastrointestinal tract. Pindone is far less persistent than second-generation anticoagulants such as brodifacoum. In sheep, the highest concentrations occur in the liver
(Robinson et al. 2005). Residues have been detected in the liver and fat of animals dosed with 20 mg/kg for 8 days, but at 2 weeks none was detected (Nelson & Hickling 1994). The elimination half-life of pindone in rat liver has been estimated as 2.1 days, compared to 113.5 days for brodifacoum (Fisher et al. 2003). Recent research suggests that anticoagulants are more persistent in cattle than in other species (Eason et al. 2010 publication in preparation).
123. This information that pindone is not as persistent as brodifacoum is important for two reasons:
a. firstly contamination of meat, liver or kidney for human
consumption would be relatively short lived if sheep ingest a sub-lethal dose; and
b. secondly, reduced persistence of pindone in animal tissues which could represent a hazard to non-target predators and scavengers, means there is less risk of bioaccumulation of pindone versus brodifacoum. Brodifacoum is used, principally in bait stations for possum and rodent control but not for rabbit control. The lower risk of bioaccumulation with pindone versus brodifacoum does not mean that there is no risk and birds eating carcasses are at some risk of secondary poisoning, particularly if they eat a number of carcasses over a period of a few days.
124. Residues of pindone have been found in non-target species found dead after pindone baiting. These have included Southern black-backed gulls, an Australian harrier, a Moko skink and a Green gecko (Glentworth and Sullivan 1994; Fairweather and Fisher 2009). Pindone residues were detected in the liver of 3 hawks and 3 seagulls found dead after baiting, indicating that secondary poisoning had occurred (Glentworth & Sullivan 1994). These results should not be surprising. Some mortality in birds of prey must be expected. When birds or other non-targets (including reptiles) ingest sub lethal amounts of pindone and survive, the residues would be excreted.
Effects on non-target species
125. Non-target research conducted in Australia provides information on the susceptibility of horses, cattle, goats, chickens, dogs, and cats. All these species were less susceptible than rabbits. Daily doses of pindone, ranging from 0.3 to 2.5 mg/kg, were administered for 5 days. No mortalities occurred and susceptibility was assessed by using extension of prothrombin time, an indicator of blood thinning, as a biomarker of poisoning. In this study, cattle and cats appeared most susceptible out of the six species tested, and horses least susceptible to pindone toxicity (Martin et al. 1991). Nevertheless, the rabbit is the most susceptible mammalian species evaluated to date.
126. Risk is heightened by multiple consecutive exposures over a number of days and requires consideration if non-target species may find and eat bait or pindone-contaminated carcasses over a number of days. The LD 50 for grey
ducks is reported to be 5 mg/kg, over seven days (Twigg et al 1999b). This LD50 value equated to a duck eating 10-30g of bait per day, depending on
the loading concentration used in bait 0.17 to 0.5 g/kg (Fairbrother and Fisher 2009). Reptiles are known to be susceptible to anticoagulants but in a study with native skinks held in captivity the skinks ate 0.02 g of Agtech pindone bait over two days (a maximum dose of 15 mg/kg) without adverse effects (Freeman et al 1996).
127. A moko skink was found dead following hand laying of pindone rabbit pellets at 3.3 kg/ha at Whangapoua Conservation Area, Great Barrier Island in 2007. Two green geckos were found dead during pindone pellet (0.5 g/kg pindone) bait station operations at Boundary Stream, Hawkes Bay in 2002 and 2003 (Fairweather and Fisher 2009).
128. Some non-target bird and reptile deaths are inevitable after aerial application of pindone. Bird species reported as killed following New Zealand pindone baiting operations include plovers, quail, rails, wrybills, silvereyes, grey warblers, black-backed gulls and Australian harriers (Sullivan 1994). Long-term population level impacts have not been studied but are not considered likely.
129. Some bird species may eat bait but gulls and harriers are likely to have died from scavenging the carcasses of animals that have died from pindone poisoning. Six black-backed gulls were found dead following mechanical spreading of carrot baits (0.17 g/kg pindone in two applications at 6.7 kg carrot/ha/application) on Grays Hill, Mackenzie Basin in August 1994 (Glentworth & Sullivan 1994).
