Mortality of North Island tomtits (

Aerial poisoning operations with carrot or cereal baits are used to control brushtail possum (Trichosurus vulpecula) populations in New Zealand forests for ecosystem conservation and to stop the spread of bovine tuberculosis to cattle and deer herds on adjacent farmland. Although various measures have been implemented to reduce the incidence of bird kills, dead birds continue to be found after poison operations. Colour- banded North Island tomtits (Petroica macrocephala toitoi) were monitored in treatment and non-treatment areas in Pureora Forest Park to determine the costs and benefits of aerial 1080 possum poisoning operations to tomtit populations. The August 1997 operation (carrot baits with very little chaff, 0.08 % w/w 1080, 10 kg ha(-1)), resulted in 11 (79%) of 14 tomtits disappearing, but none of nine from the non-treatment area. Whether the birds died of primary or secondary poisoning is unknown. No tomtits in either treatment or non-treatment areas disappeared following the August 1998 operation (cereal baits, 0.08% w/w 1080, 5 kg ha(-1)). The carrot bait operation resulted in almost all possums and rodents being killed, but a few possums and rodents survived the cereal bait operation, apparently because of a gap in bait distribution. During the 1997/98 nesting season, tomtit pairs in the 1997 treatment area had high nesting success (80% of nests fledged chicks, mean of four fledglings per nest). Even so, by the following spring it seemed that the population had not recovered to its pre-poison level. Further research on this topic is warranted, the priority being to monitor tomtit mortality during more aerial 1080-cereal bait operations in order to assess the likely risks of using those baits.     

Comparing methods for assessing mortality impacts of an aerial 1080 pest control operation on tomtits (

This study aimed to estimate the level of mortality of North Island tomtits (Petroica macrocephala toitoi) during an aerial 1080 possum poisoning operation in Tongariro Forest, New Zealand, and to evaluate transect-based alternatives to banding for monitoring tomtit populations. The operation used 12 g toxic (1080 at 0.15% weight/weight) cereal baits sown at 3 kg/ha. Transects were established at three neighbouring sites; two within the 1080 poison area, and one outside. The re-sighting of 14 out of 15 banded male tomtits at one site within the 1080 operation indicated that mortality was low. This was backed up by results from a before-after-control-impact (BACI) design to analyse density estimates from distance sampling along transects. We analysed the change in counts of territorial males before and after the operation based on the same transect surveys. This also showed little impact of poisoning on tomtits, and indicated that loss rates greater than 8.4% due to 1080 were incompatible with the data (95% one-sided confidence bound). Counts of territorial males gave a much tighter confidence bound than the banding or distance sampling results. Of the techniques applied, the counting of territorial males appears to have the most promise for providing high-precision estimates of short-term impacts, by taking full advantage of the territorial habits of male tomtits in spring. However, distance sampling shows potential for providing the basis for longer-term monitoring of tomtit populations. The transect-based approaches involved substantially fewer resources than banding for estimating short-term impacts, and offer a considerably less-intensive means of longer-term monitoring of tomtits.     

Hoarding behavior by ship rats (Rattus rattus) in captivity and its relevance to the effectiveness of pest control operations

Hoarding of food items is well known among muroid rodents, but evidence for hoarding behavior among ship rats (Rattus rattus) is scant. Here, we characterize hoarding behavior in ship rats maintained in captivity after capture from the wild. After acclimatization to captivity, 40 ship rats (21 females, 19 males) were presented with baits in experiments designed to emulate a typical poison control operation for vertebrate pests in New Zealand: this involved first offering rats nontoxic cereal baits (of 2- or 6-g size) as a prefeed for three nights consecutively, followed by 6- or 12-g cereal baits laden with 0.15% 1080 on the fourth night. Seventy-eight percent of rats (31/40) hoarded food in distinct cache sites when presented with nontoxic baits although there was no significant effect of bait size or type on hoarding behavior and nor did hoarding behavior vary according to rat gender. When rats were presented with 1080-laden baits, the incidence of hoarding was reduced to 40%, due to the onset of toxicosis. This study indicates that R. rattus will show hoarding behavior analogous to other rat species when presented with an excess of cereal-based baits, at least under conditions of captivity and free from competition. This finding may have practical relevance: since 1080 is the principal toxin used against the major vertebrate pest species in New Zealand (the brushtail possum, Trichosurus vulpecula), ship rats have the potential to deplete supplies of prefeed and/or toxic baits intended for possum control. However, based on typical rat densities recorded in New Zealand native forest (c. 5 rats/ha), the degree of removal and manipulation of toxic baits observed by ship rats here is unlikely to impact adversely on the efficacy of possum control operations.     

The toxicity of sodium monofluoroacetate (1080) to

Although many species of native invertebrates have been identified on toxic baits containing sodium monofluoroacetate (1080) following aerial operations for possum control, few quantitative data are available to determine the risk of primacy or secondary poisoning that may result from these exposures. This paper reports on a series of studies conducted to determine the risk of 1080 exposure to one such non-target insect, the native ant Huberia striata. Subsequent risk of secondary poisoning to insectivorous animals is extrapolated. Ants were exposed in the laboratory to cereal baits containing 0.15% sodium monofluoroacetate, with and without alternative sources of food, and mortality was compared with controls after 24 and 48 h. Acute mortality was significantly greater in both exposed groups and ants that died contained 1080 residues, indicating that some ants could consume lethal amounts of 1080 in cereal baits. However, the increase in mortality was relatively small (7% after 24 h and 12% after 48 h), indicating either that palatability of 1080 bait to this species is low, or that the toxin is not readily absorbed from this matrix. Additional trials were conducted to expose ants to 1080 in sugar-water, in order to determine the acute toxicity dose range. The oral LD50 at 48 h was 32 mg kg(-1), which was comparable to previous results obtained for the large-headed weta, Helmideina crassidens. The persistence of 1080 in ants consuming a sub-lethal dose was also determined, in order to assess risks of secondary poisoning to insectivores. Residues in ants receiving approximately 36 mg kg(-1) of 1080 declined rapidly, from a Peak of 5.51 mg kg(-1) 1 day after exposure, to 0.27 mg kg(-1) after 7 days. Potential risks to insectivorous birds were calculated using worst-case exposure assumptions, and were determined to be negligible.     

Do colours that deter birds affect cereal bait acceptance by possums (

Poisonous baits used for pest control in New Zealand commonly contain green dye and cinnamon oil to make them less attractive to birds. Research aimed at reducing the impact of poison based pest control on birds has shown that some birds are initially deterred from feeding on blue or, to a lesser extent, green coloured food and are attracted to yellow or red food. We determined whether colours that deter or attract birds affected the acceptance of non-toxic and toxic cereal baits by captive brushtail possums (Trichosurus vulpecula). Individual possums were offered, daily, a choice between a standard green dyed non-toxic cereal bait and either a blue dyed(17 possums) or yellow dyed non-toxic bait (16 possums) for 10 days. Following this, for the first group of 17 possums, 1080 toxin was added to either the green bait (9 possums) or blue bait (8 possums) and possums were offered the green versus blue choice again. Two additional groups that had not previously been fed cereal baits were also given a choice between blue and green baits, one of which was toxic. All possums offered non-toxic bait ate less on the first day of presentation than on subsequent days. There was no difference in acceptance of either blue or yellow coloured non-toxic bait compared to the standard green non- toxic bait on any days. Bait colour appeared to be unimportant in cereal bait choice and did not deter possums from eating any of the baits. The addition of toxin to baits did not significantly alter bait choice in any groups, although some individuals which had no previous experience with baits ate more toxic than non-toxic bait. These data suggest that adding a stronger bird deterring colour (i.e., blue) to poisonous baits is unlikely to adversely affect the acceptance of baits by possums.     

Costs and benefits of pre‐feeding for possum control

Abstract

Four control operations, each with paired pre‐feed and no‐pre‐feed treatments, and comprising two baited with carrot and two with cereal, provided strong support for the hypothesis that pre‐feeding produces higher possum kills than no pre‐feeding. In these replicates, even where high kills were achieved, the modelled two‐ to three‐fold extension of the period of population depression given by pre‐feeding was sufficient to warrant the expense of the additional bait sown. The benefits of pre‐feeding were clear‐cut, and were greatest where poorest kills were achieved, irrespective of bait type. Pre‐feeding is thus likely to reduce the probability of failure and should be considered a form of operational insurance.     

Invertebrates feeding on baits used for vertebrate pest control in New Zealand

This study was initiated in response to concerns that vertebrate pest control operations in New Zealand may be having deleterious impacts on invertebrate populations and, secondarily, on insectivorous non-target vertebrate populations. Invertebrates feeding on non-toxic baits of the types used for vertebrate pest control were collected and identified. The bait types were diced carrots and three types of cereal-based baits (No.7, RS5, and AgTech). The study was conducted in two rata/kamahi dominated forests (Bell Hill Scenic Reserve and Kopara Forest, West Coast), in July and September 1996. The most common species found on baits was the ant Huberia brouni (Hymenoptera: Formicidae). Other common taxa were Orthoptera (at least eight species of weta including Zealandosandrus aff. gracilis, Gymnoplectron sp., and Pleioplectron sp.). Coleoptera (at least nine species of beetles including Saphobius nitidulus, Nestrius sp., and Phrynixus sp.), Dermaptera (at least one species of the earwig Parisolabis sp.), Opiliones (at least three species of harvestmen), and Acarina (at least three species of mites). The ants and weta were found predominately on cereal-based baits, and the beetles, earwigs, harvestmen, and mites predominantly on carrot baits. More invertebrates were found on carrot and RS5 cereal-based baits than on the other two bait types, and more on baits at night than during the day. Fewer invertebrates were found on cinnamon-flavoured baits (used for 1080-poisoning of possums) than on plain baits (used for brodifacoum-poisoning of rodents). The number of species and number of individual invertebrates found on baits were a small proportion of the number likely to be present in the forest litter. We predict that vertebrate pest control operations are unlikely to have any long term deleterious impacts on invertebrate populations. This prediction should be tested by monitoring populations of invertebrate species, found to ear baits, during vertebrate pest control operations.     

Brodifacoum residues in possums (

The presence of brodifacoum residues in possum (Trichosurus vulpecula) livers following routine possum control was investigated. Possums were poisoned in six nature reserves in the Wellington Region, New Zealand, using cereal baits containing 20 mg kg(-1) brodifacoum dispensed from bait stations. Thirty-five surviving possums, and five dead possums were sampled from the reserves following poisoning, and their livers analysed for the presence of brodifacoum. The majority (83%) of samples contained brodifacoum at concentrations ranging from 0.007 mg kg(-1) to 6.2 mg kg(-1). The presence of significant quantities of brodifacoum in possum carcasses following poisoning operations creates potential for secondary and tertiary poisoning of non-target species.     

Controlling small mammal predators using sodium monofluoroacetate (1080) in bait stations along forestry roads in a New Zealand beech forest

A single five night pulse of sodium monofluroacetate (0.15% 1080) applied in bait stations at two different spacing intervals, 100 and 200 m, along forestry roads in New Zealand beech forest, killed all four of the resident radio-tagged stoats (Mustela erminea) and all three of the resident radio- tagged wild house cats (Felis catus) by secondary poisoning. Gut contents of predators indicated that house mice (Mus musculus), ship rats (Rattus rattus) and bushtail possums (Trichosurus vulpecula) were important sources of the toxin. High kills of predators, possums and rats at both 100 and 200 m spacing regimes suggest that greater efficacy of controlling these pests would be achieved with the latter method. Evidence suggests that routine management of possums and rats using 1080 and brodifacoum has resulted in widespread control of small mammalian carnivores by secondary poisoning in New Zealand forests. However, aerial application of poison can kill large numbers of tomtits (Petroica macrocephala) and robins (Petroica australis) and few other native bird species have been adequately monitored through such operations. Reducing risks to native wildlife is responsible ecological management. Use of bait stations along forestry roads or tracks may be fundamental in mounting cost-effective :Large-scale ground-based protection of native wildlife through safer predator controls.     

A review of recent regulatory and environmental toxicology studies on 1080: Results and implications

Sodium monofluoroacetate (1080) is a highly toxic vertebrate pesticide that has been widely used for possum and rabbit control in New Zealand since the 1950s. Because of its importance in pest control and the highly toxic nature of this compound, its environmental fate, persistence, non-target impacts and general toxicology have been and continue to be extensively studied. A series of in vitro (cell culture) and laboratory animal studies (in rats and mice) have recently been undertaken to update the regulatory toxicology database for 1080. Results of three different, complementary tests indicate that 1080 is not mutagenic, and therefore unlikely to cause cancer. Results of developmental toxicity studies indicate that 1080 causes developmental defects in rats when pregnant females are exposed to relatively high doses (0.33 and 0.75 mg kg(-1) day(-1)) on a daily basis during the period of organogenesis (from days 6 through to 17 of gestation). The developmental abnormalities observed were mild skeletal effects: slightly curved forelimbs, and bent or "wavy" ribs. These results highlight the highly toxic nature of 1080 and the need for extreme care when handling this pesticide during the manufacture and distribution of bait, but do not preclude its proper use. The morphological changes in foetuses observed in the recently completed rat developmental toxicity studies are likely to raise concern regarding the exposure of humans (particularly pregnant mothers) to 1080 through contamination of waterways after the aerial sowing of 1080 baits. However, ongoing monitoring (1990-98) of waterways from areas where 1080 baits have been aerially sown confirms that there is negligible risk of human contact with 1080 through this route of exposure.     

Evaluation of non-toxic bait interference for indexing brushtail possum density

Percentage non-toxic bait interference is currently used by local authorities to monitor brushtail possum control operations but the validity of the method has not been established. Two models have been proposed to estimate an index of possum density (possums per bait) from a log-transformation of percentage bait interference. In two trials, percentage bait interference and the density index derived from percentage bait interference using the Bamford (1970) model usually increased from night to night. Variation in the number of possums per bait was reduced by dividing the Bamford index by the number of nights baits were put out, as proposed by Jane (1981). The spatial distribution of bait interference along lines of bait stations was random for the first night and usually for at least 5 nights. Rhodomine dye and cyanide poison, alternated at bait stations after at least 5 nights non-toxic baiting, showed that about 83% of possums in one trial and 29% of possums in the other trial interfered with more than one bait on the same night. Increasing the spacing between bait stations reduced the proportion of possums interfering with more than one bait. The number of possums per line killed by cyanide baits after at least 5 nights of non-toxic baiting was strongly correlated with percentage bait interference, with and without logarithmic transformation, on the first night. The possums' apparent ability to remember the location of bait stations from one night to the next means that, in the absence of better advice, the original Bamford (1970) model should be modified as proposed by Jane (1981) to estimate possum density and change in possum density (e.g., from pre- to post-poison).     

Large-Scale Poisoning of Ship Rats (

This paper describes the impact of nine poison operations on ship rats in four areas (35 ha to 3200 ha) of North Island forest. Poisoning with 1080, brodifacoum, or pindone killed 87- 100% of rats, based on trapping and tracking-tunnel indices. Rat populations took 4-5 months to recover. Operations to protect nesting birds should therefore coincide with the onset of nesting and be rePeated each year, although not necessarily with the same methods. Population reduction declined each year at Mapara, King Country, during three annual 1080 operations which used the same lures and baits, but remained high at Kaharoa, Bay of Plenty, where poison toxicity was higher, non- toxic bait was pre-fed, and poisoning methods varied each year. Mouse tracking rates increased in poisoned forests 3-6 months after poisoning if the initial kill of rats exceeded 90%, Peaked 7-9 months after poisoning, then declined to pre-poison levels. Future research should focus on how prey and non-prey species within a forest community respond to a temporary reduction in rat numbers, and on methods to maintain low rat densities after initial knock-down.     

Life history traits contribute to decline of critically endangered lizards at Macraes Flat, Otago

To measure the costs and benefits of an aerial 1080 possum control operation to kereru and kaka in Whirinaki Forest Park, individuals of both species were radio-tagged from October 1998 to June 2002. We monitored birds in one treatment and one non-treatment study area to compare toxin-related mortality, nesting success and survival. The poison operation involved the spreading of non-toxic carrot baits on 1 May 2000, and the toxic baits on 17/18 May 2000. Possums and rats were moderately abundant in both study areas prior to the poison operation, but afterwards few possums and rats remained in the treatment area. All radio-tagged kaka and kereru in the treatment area survived the poison operation. No radio-tagged kereru and too few radio-tagged kaka bred in either study area during the 2000/01 nesting season to show whether reduced possum and rat populations would enable the birds to nest more successfully. A reduction in possum and rat densities in the non-treatment area (and an increase in densities in the treatment area) during 2001/02 meant that during the second nesting season after the poison operation, possum and rat densities were similar in the two study areas. The nesting effort and success of kaka and kereru is described for each of four nesting seasons, with the main cause of nesting failure for both species being predation. While no radio-tagged adult male kaka died during the study, 6 females did, giving them a mean life expectancy of 9.5 years. In contrast, radio-tagged adult kereru suffered high mortality, resulting in a mean life expectancy of just 1.5 years. Predation by introduced mammalian predators was the main cause of mortality of kaka eggs, chicks, fledglings and adult females, and of kereru eggs, chicks, fledglings and adults. Effective control of introduced mammalian predators, including control by aerial 1080 operations, just before mast fruiting events that invariably promote prolific kaka and kereru breeding, should benefit these bird populations.     

Sodium Monofluoroacetate (1080) Water-Residue Analysis after Large-Scale Possum Control

Successful possum control operations were conducted in 1990, using aerially-sown sodium monofluoroacetate (1080) possum baits, at Waipoua (85% kill) and on Rangitoto Island (93% kill). In Waipoua Forest Sanctuary, streams and rivers were monitored for 4 months after 100 tonnes of 1080 possum baits were sown over 17 000 ha of forest. At Rangitoto Island, adjacent to Auckland, surface and ground water samples were analysed for 6 months after 20 tonnes of 1080 possum baits were sown over the 2300 ha island. No 1080 was detected in the streams and rivers of the Waipoua forest or in the surface or ground water of Rangitoto Island.     

Combining Aspirin with Cholecalciferol (Vitamin D3) – A Potential New Tool for Controlling Possum Populations

The introduced Australian brushtail possum is a major vertebrate pest in New Zealand, with impacts on conservation and agriculture being managed largely through poisoning operations. Cholecalciferol (vitamin D₃) is registered for use in controlling possums and despite its many advantages it is expensive and relatively inhumane. Combination of a high proportion of aspirin with a low proportion of cholecalciferol was effective in killing high proportions of groups of acclimatised, caged possums: this is attributed to both an unexpectedly high toxicity of the type of cholecalciferol used, and a proposed synergistic mechanism between the two compounds. Death was caused by localised damage to heart ventricles by aspirin, and inhibition of tissue repair by both aspirin and cholecalciferol. The observed toxicosis had lower impact on the welfare of possums than either compound administered alone, particularly aspirin alone. Residue analyses of bait remains in the GI tract suggested a low risk of secondary poisoning by either compound. The combination of cholecalciferol and aspirin has the potential to meet key requirements of cost-effectiveness and humaneness in controlling possum populations, but the effect of the combination in non-target species has yet to be tested.     

Breeding success of New Zealand falcons (

Productivity data on the New Zealand falcon (Falco novaeseelandiae) were collected from 87 nest sites in Kaingaroa pine plantation during three breeding seasons, 2003 to 2006. On average, 1.81 chicks were successfully fledged per nest, with young reared successfully at 71% of nests. Breeding occurred between August and March, with most eggs laid before December and most chicks fledged by February. Fifteen percent of nests were depredated, 9% contained eggs that failed to develop and 4% failed owing to forestry operations disturbing or destroying nests. No negative impact of 1080 bait or of desiccant or release spray application was recorded on falcon productivity. The population of falcons in Kaingaroa Forest increased during a period of pest control using 1080 bait so we see no reason to discontinue its use. Although impacts from forestry operations were low and restricted to land preparation and harvesting operations, there is potential for adverse impacts to increase. Mechanical forestry operations can continue without negative impacts by avoiding a buffer zone of 100?200 m around an active falcon nest.     

Effects of an aerial 1080 possum poison operation using carrot baits on invertebrates in artificial refuges at Whirinaki Forest Park, 1999– 2002

The effects of an aerial 1080 possum poison operation using carrot baits on invertebrates in Whirinaki Forest Park are described from an un-replicated study of artificial refuges attached to tree trunks. Auckland tree weta (Hemideina thoracica), cave weta (Pharmacus sp. and Isoplectron sp.), cockroaches, spiders and harvestmen, and leaf-veined slugs (Athoracophorus bitentaculatus) were the most frequent occupants, but snails, millipedes, centipedes, flatworms, lepidopteran larvae, glowworm larvae (Arachnocampa luminosa), peripatus, slaters and beetles were also present occasionally. Invertebrate numbers were monitored every second or third month for a year before the poison operation, and for two years afterwards. Numbers of tree weta, cave weta, cockroaches, spiders and harvestmen, and leaf-veined slugs did not decline substantially in refuges in the treatment area relative to those in the non-treatment area immediately after the poison operation. Our results, and those from two other similar studies, suggest that aerial 1080 poison operations are unlikely to have a detrimental effect on invertebrates that occupy cavities above ground.     

Poisoning mammalian pests can have unintended consequences for future control: Two case studies

Vertebrate pest control operations using toxic baits can have unintended consequences for nontarget species. some of which may themselves be pests. Learned avoidance behaviour (termed 'aversion') can be induced by sublethal dosing, which can arise when species with high and low susceptibilities to a toxin co- exist in the same area. In such cases the less-susceptible species (e.g., possums Trichosurus vulpecula) may be sublethally poisoned by control work targeting the more- susceptible species (e.g., rabbits Oryctolagus cuniculus). A case study of rabbit control on North Canterbury farmland is presented to demonstrate this effect. When control is being rePeated at frequent intervals, it is prudent to vary the control methods used. Nevertheless, aversion induced by the use of one toxic bait (e.g., cyanide paste) can in some situations 'generalise' so that the efficacy of control using other toxins (e.g., 1080 and cholecalciferol in cereal baits) is also compromised. A case study of initial and follow-up possum control in four discrete areas of Canterbury forest provides an example of this problem. The implications of these findings for future pest management in New Zealand are discussed.     

An assessment of the probability of secondary poisoning of forest insectivores following an aerial 1080 possum control operation

Assays for the toxin sodium monofluoroacetate (compound 1080) were undertaken on arthropods collected from toxic baits after a brushtail possum (Trichosorus vulpecula) control operation in Nothofagus forest in central North Island, New Zealand. The 1080 concentrations measured (mean 57 mu g per g, max 130 mu g per g) are considerably higher than those reported by other researchers who collected arthropods randomly after control operations. These data, together with published information on sensitivities to 1080, as well as diet and consumption rates, were used to calculate the median lethal doses of arthropods that have fed on 1080 baits for a number of vertebrate insectivores found in Nothofagus forest. The results indicate small insectivores that feed on, or close to, the ground (e.g., tomtit Petroica macrocephala, robin P. australis, hedge sparrow Prunella modularis, and the short-tailed bat Mystacina tuberculata) may be vulnerable to secondary poisoning. For instance, a tomtit will receive the median lethal dose of 1080 from 1.32 g (i.e., 14.7% of its daily food intake) of arthropods containing 57 mu g per g of 1080. Because of their greater sensitivity to 1080 poisoning, bats are at much greater risk; a short-tailed bat will receive the median lethal dose of 1080 from as little as 0.04 g (0.7% of its daily food intake) of arthropods containing 57 mu g per g of 1080.     

Assessment of non-target impact of 1080-poisoning for vertebrate pest control on weta (Orthoptera: Anostostomatidae and Rhaphidophoridae) and other invertebrates in artificial refuges

Artificial refuges and mark-recapture techniques were used to monitor the non-target impacts of handbroadcast application (simulating aerial application) of Wanganui No.7 cereal-based baits containing 0.15% (1500 µg g-1) 1080 on populations of weta and other invertebrates in Tararua Forest Park, North Island, New Zealand. Wellington tree weta (Hemideina crassidens) and a cave weta (Isoplectron sp.) were the only species of weta that occupied the refuges. Flatworms, slugs, spiders, harvestmen, amphipods, millipedes, centipedes, cockroaches, and beetles also occupied the refuges. Invertebrate numbers in the refuges were monitored for 12 months before and 4 months after bait application on 22 August 2000. Bait application had no significant impact on the numbers of either species of weta, or on slugs, spiders, and cockroaches, the most numerous other invertebrates occupying the refuges. Bait application also had no effect on the number of individually marked tree weta resighted in the refuges. Few weta or other invertebrates were observed on baits at night. The concentration of 1080 in a cave weta collected alive from a bait, and in a tree weta collected alive from outside an artificial refuge, was less than 10% of the average lethal dose. The results indicate that 1080-poisoning for vertebrate pest control is unlikely to have any negative impact on populations of weta or the other invertebrates monitored.