The illegality of Norway rat management on Ulva Island Aerial brodifacoum drop 2011

There is substantial evidence that the Department of Conservation’s (DoC’s) management of the Norway rat invasion of Ulva Island contravenes its obligations under the Conservation Act (1987) to preserve natural resources and freshwater fisheries, as well its as obligations under the National Parks Act (1980) to maintain this area in a natural state.

Furthermore there is evidence that Environment Southland has approved the resource consent application for this operation without an adequate assessment of adverse effects, alternatives or mitigation techniques, or a monitoring programme that is appropriate to the risk and scale of the proposed operation.

The evidence is as follows.

1. Background

Ulva Island is a nationally and internationally significant island sanctuary of 269 hectares (including numerous islets and rock stacks), situated within Paterson Inlet. It is one of the Southland’s best remaining examples of a lowland forested ecosystem. It is highly valued for its biodiversity, its ease of access and its importance to the tourism industry. Some of the birds present are nationally threatened species, introduced after rat eradication in 1997 (using poison in bait stations followed up with trapping and poisoned maize). Most of the island is a National Park managed by DoC, and it is surrounded predominantly by a DoC marine reserve and a Mataitai reserve. It is low-lying and situated in an area of high rainfall and frequent gales. There are seven streams, five less than 1 metre wide and two about 3 metres wide (DoC, 2011).

2. DoC has failed to control rats

Under the Rakiura National Park Management Plan, DoC is required to take all possible steps to remove any future introductions of introduced animals to Ulva Island. Since 1997, traps and bait stations have been used to catch re-invaders, with an average of one rat per year caught in the traps. In 2010 these DoC biosecurity measures failed, and a population of Norway rats became established. Four rats were caught in May and June 2010, then in January and February 2011, 70 rats were trapped. DoC then shut the traps and has been planning a poisoning operation, using aerially applied brodifacoum poisoned cereal baits, ever since. DoC considers that the rats will impact significantly on birds, invertebrates and plants if not eradicated. However it admits that aerial brodifacoum might not work: “Eradicating a population with lots of food hasn’t been attempted before and hence there is a very real risk of failure” (DoC, 2011).

Further to this, in the Communication Update notes at the back of the AEE (Assessment of Environmental Effects, submitted by DoC in its Resource Consent application), it is clarified that “The rats have unlimited food and may not have any interest in eating the bait….this situation has only been encountered once before, on Fregate Island..rats invaded and they attempted to eradicate them using bait stations. This failed…rats were eventually eradicated from the island 5 years later by an aerial baiting operation after the rat population had peaked” (DoC, 2011).

Thus in the first instance, DoC has not been competent in preventing rats becoming established. Secondly it failed to step up its control measures to eradicate the rats once established; instead it shut the traps and allowed rat numbers and effects of rats to continue to build up. Some highly valued rare species, including the ground-dwelling Saddleback, cannot co-exist with rats and numbers are expected to have been severely impacted by them already (Ian Jamieson, pers. comm.). Thirdly, by its own admission DoC’s plan to use aerial poison to eradicate the rats has a good chance of failure.

3. Brodifacoum will cause severe ecological harm

a) DoC’s ecological responsibility DoC is legally required to preserve natural resources, which as defined under the Conservation Act, include plants and animals of all kinds, systems of interacting living organisms, and their environment. Under the National Parks Act, areas of New Zealand that contain ecological systems that are unique, or scientifically important, must be preserved as far as possible in their natural state.

b) Effects on birdsIn the AEE of the proposed poisoning, DoC admits that there is little information on the toxicity of brodifacoum to native species, but that most vertebrate species will be susceptible to primary or secondary poisoning (DoC, 2011). Studies on effects of brodifacoum have mainly concerned birds but scientific data are scarce with most studies relying on “five minute counts” which have been discredited as being unreliable (ERMA, 2007a). Nevertheless, birds from a wide range of species have been recorded killed in aerial operations, with severe reductions in populations of at least four (DoC, 2011). Figures from previous operations include deaths of 20% of the kaka present, 86% of fernbirds, 65% of robins, over 90% of pukeko and 98 % of weka. Given that the rats might not eat the bait, bird kills on Ulva Island may be even higher than any observed previously.

The severity of the expected effects on the bird populations on Ulva Island are highlighted by comments from ornithologist Ian Jamieson, who has Marsden funding to study genetics of the birds on the island.

He considered in June that the poisoning, coming on top of the effects of the uncontrolled rats, may kill off the saddlebacks (Ian Jamieson, pers. comm.). Jamieson communicated to DoC in March that his research was likely to be sustainable if the poison drop went ahead as soon as feasible, otherwise he was at risk of losing the benefits of his study (AEE correspondence record).

For surviving species that suffer from severely reduced populations after the poisoning, a lack of genetic diversity may make them vulnerable to future stressors such as new diseases or competitors (Innes et al., 2010).

The deaths of a large number of birds would be expected to have a severe negative impact on tourism. Frequent encounters with birds are a major selling point to tourists. For example, the Ulva Island guided walks website–Kiwispotting.asp states that “The morning is spent on Ulva Island; an amazing insight into the natural history, stunning bird and plant life in the southern-most bird sanctuary in the world.  Birds to watch out for are the Stewart Island Weka, South Island Saddleback, Tomtit, Rifleman, Stewart Island Robin, Brown Creeper, Tui, Kaka, Fantail, NZ Pigeon, Parakeet (yellow and red-crowned), Bellbird and Yellowhead.”

Following DoC’s poisoning operation, nature-loving visitors will not only have a significantly diminished experience due to a lack of birds, they will also be subjected to poison warning signs and (very likely) the sight of dying and debilitated birds. A targeted operation instead at just the rats would be an added attraction and demonstrate to international visitors that DoC was capable of responsible behaviour.

c) Effects on terrestrial invertebrates
Effects of brodifacoum on terrestrial invertebrates were largely discounted in the AEE, due to research showing blood clotting systems (that brodifacoum acts upon) were different from vertebrates, and due to a small number of studies on introduced species and a Tree Weta that indicated these animals might be comparatively resilient. However it was admitted that research had been done that showed mollusc (Conus) and insect (Drosophila) species indeed could be affected by brodifacoum. In fact this same research, cited by DoC, concluded that the vulnerable chemical process found in Conus and Drosophila was probably widely distributed among biological systems (Walker et al., 2001). Negative effects of brodifacoum on invertebrate species richness and abundance were found on Kapiti Island (Sinclair et al., 2005) and negative impacts on invertebrates were also found in the Perlorus Bridge study cited by DoC, even though that study only had the power to detect population decreases of 67% or more (Spurr, 1996).

It should have been of great concern to DoC and highlighted (rather than given passing mention) in the AEE that aerial brodifaoum treatment to eradicate rats on Fregate Island, Seychelles, was thought to have caused significant declines in two terrestrial snail species, one of which become endangered and the other probably extinct (Gerlach, 2005). Another species of snail, from Mauritius, was also thought to be vulnerable brodifacoum (Gerlach & Florens, 2000).

Brodifacoum binds strongly to soil and persists for many months (Weldon et al., 2008). Effects of this on soil communities were completely ignored in the AEE. However soil is an important part of the ecosystem and its inhabitants are diverse and essential to the recycling of nutrients. “Colourful flatworms (phylum Platyhelminthes), slugs and snails of all sizes (phylum Mollusca), and numerous crustaceans, insects, spiders, myriapods and mites (all members of the phylum Arthropoda) make up the populations of New Zealand soils…A square meter of forest floor may be inhabited by hundreds of species of invertebrates, many barely visible to the naked eye. These are the organisms whose identities and life cycles are often unknown. Many native species of New Zealand soil fauna are still undescribed, and the intricate relationships among individual species remain ‘terra incognita’” (

The uniqueness and diversity of the invertebrate community present on Ulva Island was demonstrated in a study of species living in ground litter and tree trunks. Sixty-two different species were found, and were distinct from those on the mainland (Michel et al., 2008).

d) Effects on reptiles
Skinks are known to eat cereal baits (ERMA, 2007b), two NZ gecko species have been observed eating brodifacoum baits, and brodifacoum bait consumption killed some Telfair’s skinks on Round Island, Mauritius (Hore & Hare, 2006). DoC’s AEE claims that rat eradication from Ulva Island will have a positive effect on lizards by removing predators, and that no threatened reptiles are known to occur on the island. Nevertheless poisoning is likely to occur because bait consumption is likely, and no study has examined what reptiles are present on the island. Sub-lethal effects of brodifacoum on reptiles may also occur, e.g. a decreased ability to thermoregulate (Hore & Hare, 2006).

e) Effects on freshwater ecosystems
Binding strongly to sediment and soil (Weldon et al., 2011), brodifacoum would be expected to enter and significantly contaminate the stream beds in this high rainfall area. However DoC’s AEE ignores all freshwater species except for fish. This is despite their legal duty under the Conservation Act to be concerned with all shellfish of the Classes Mollusca and Crustacea that inhabit freshwater. Broadifacoum has been described as a molluscicide (Gerlach, 2005) and is rated as highly toxic to aquatic organisms on the basis of its toxicity to the freshwater crustacean Daphnia magna (USEPA, 1998). Therefore it could have devastating effects on the freshwater communities on the island. New Zealand freshwater habitats contain about 450 formally identified insect species and at least 200 other kinds of invertebrate (including crustaceans, molluscs and various worm phyla) (

In his statement of evidence at the Resource Consent hearing, DoC’s Stewart Island biodiversity programme manger Brent Beaven made the incredible claim that the streams on Ulva Island had limited fauna, based on his having put fish traps out in two of the streams in May 2011 which “only caught freshwater crayfish”. These crayfish may well be unique species. This manager’s seemingly complete ignorance of the diverse species to be found in New Zealand’s freshwater environments is revealing of the depth of incompetence within the organisation.

Even effects on fish were very poorly addressed in the AEE. Brodifacoum is known to be toxic to trout but there is no information on the toxicity to New Zealand’s native fish. Possible effects were discounted in the AEE on the assumption that galaxids and bullies would not feed on stationary baits, the fact that the pellets would break down rapidly, and eels and galaxids had not been eradicated in three other brodifacoum operations. Some glaring omissions in this reasoning are that particles from breaking down baits would not be stationary, and that invertebrates and poisoned carcasses are likely to cause secondary poisoning to fish.

f) Ecological effects
The AEE states that “Ecosystem effects are expected to be minimal with an aerial operation, due to the limited exposure time of the ecosystem to the toxin.” This is in spite of all the evidence that brodifacoum contaminates organisms, soils and sediments for many months after exposure and is readily passed up the food chain (Weldon et al., 2011), and despite the profound effects expected on some members of the ecological community (notably birds, the organisms that have received the most attention from DoC). Bird deaths alone will have multiple ecological effects including a decrease in genetic diversity, altered population structure (in a recovering population there will be more juveniles), vacant resources such as food and nesting sites, and alterations in pollination rates, seed dispersal, and predation on invertebrates. Ecologists have been sounding warnings for years that to manage responsibly, DoC needs to come to terms with the fact it is dealing with systems of interacting organisms (Innes & Barker, 1999; Hore & Hare, 2006). The Ulva Island AEE shows that these warnings have so far fallen on deaf ears.

4) Marine environment

a.     The discharge of baits into the sea was omitted from Section 5 of the application for resource consent: Description of the activity. Several adjacent owners and occupiers (Section 8 of the application) and potentially affected parties (Section 9) were missed off the application form, namely the DoC Marine Reserve, Mataitai Committee, Ministry of Fisheries, and the Council of Recreational Fishing.

b.    Effects identified in the AEE
Effects on the marine environment were almost totally ignored in the AEE. Points that were addressed in the AEE, concerning (or relevant to) the marine environment, were (in order in which they appeared):
Page 10: The island has an indented rocky coast with numerous rock stacks both close to the shore and several hundred meters away (Tamihau, etc) which are also to be treated with aerial brodifacoum.
Page 16: The island is surrounded by the seawater of Paterson Inlet, the Ulva Is Marine Reserve and a Mataitai reserve. There are a few small permanent streams. Rainfall is estimated to be 1650mm annually and winds often exceed gale force.
Page 25:  Some baits will enter the marine environment.
Page 30: Brodifacoum becomes strongly bound to soil, then slowly degraded over weeks to months. Erosion of soil will cause the brodifacoum to enter water, where it is likely to remain bound to organic material and settle out in sediments. There is no knowledge of how long it will take to degrade in sediments.
Page 31: Brodifacoum is not expected to be found in sea water, due to its insolubility, “coupled with the small amount of bait being laid in coastal areas…the low concentrations of brodifacoum in the bait… the small amount of bait expected to end up in the coastal area and the tidal and wind action.”
Page 37: Invertebrates such as molluscs and arthropods may be affected by brodifacoum. This includes Conus [a genus of carnivorous marine snail].
Page 38: “Based on estimates drawn up for the Kapiti Is rat eradication [data not provided] the amount of toxin assimilated into the marine environment will be many orders of magnitude lower than the lethal limits known to be toxic to some fish species”…”toxicity of brodifacoum to native fish species is unknown”…”it is possible that some marine fish will consume whole baits….however the amount of bait presented, in comparison to the marine area exposed, will see no species risk at a population level”
Page 39: Studies made on fish by Empson & Miskelly (1999) were described in which Spotties, banded wrasse and triple fins all ate cereal bait, and some fish in aquarium trials exposing them to brodifacoum baits for one hour died from eating bait and/or exposure to the toxin. But on Kapiti Island these authors found “there was no evidence of a population decline in Spotties as a result of aerial application of Talon.”
Page 39: After DoC accidentally dumped 18 tonnes of brodifacoum bait in the sea at Kaikoura, samples of mussels and paua were contaminated for 31 months, within a 100m2 area.
Page 40: A handful of other studies on contamination of marine life (dolphins, penguins, pilchards and molluscs) were described but sampling details were not given and there were no published references.
Page 42: The risk of exposure to the public was summed up as “Exposure of any significance (i.e. which may cause harm) is only likely if baits or contaminated animals are eaten in substantial quantities”
Page 46: “Ecosystem effects are expected to be minimal…due to the limited exposure time of the ecosystem to the toxin”
Page 69: (Departmental update, 7 March 2011) “Some bait will indeed enter the marine environment. This will be in the order of two pellets per metre of coast, or less than half a kilogram of bait in a 100m stretch. The well monitored Kaikoura bait spill dropped 18 tonne in a 100m stretch and effects were limited to that 100m stretch. While the marine reserve is obviously not the local food basket, we will be discussing concerns about the bait entering the environment with the Mataitai committee. One option that has been raised has been for a Rahui to be placed around Ulva Island to completely eliminate any possibility of eating fish that may have had contact with the bait. Exploring potential solutions such as these may prove a way forward. We are also seeking the advice from national marine and fisheries experts.”  (It was subsequently found that a Rahui was not a legal option.)

  • Major omissions of the marine section of the AEE
  • How much brodifacoum will enter the marine environment, from baits, bait dust, and in sediments washing out from the land (from baits, bait dust, sediment runoff and carcasses entering the streams).

The island receives high rainfall and frequent gale force winds so significant movement of soil and sediment into streams and the sea is to be expected.

The coastline is very jagged and the numerous rock stacks make up a total of 5 hectares, so there will be a very large area of coastline poisoned with baits. Even if the greatest of care is taken these baits will land in the sea and bounce or roll off rocks into it (some of the rock stacks are described as steep and rugged, and the pellet baits are lightweight, small and rounded). In a North American aerial brodifacoum operation that DoC advised on, a deflector bucket was used for upper areas of cliffs and the lower areas and coastline were sown by hand, yet divers still found numerous baits in the sea (Howald et al., 2010).

In addition the baits are to be loaded from the Foveaux Freighter at Sydney Cove, Ulva Island. A high level of contamination from dust and spilled baits is to be expected in this area, however this likely effect has been totally overlooked in the AEE. (Apparently this plan has now been abandoned and an area of vegetation has been cleared on the Island for the bait loading part of the operation.)

The only study of toxic dust associated with the aerial dropping of poisoned cereal baits found that the dust spread at least 1 km, and was present for at least 5 days (Wright et al., 2002).

  • How long the toxin will remain in marine sediments.
  • Movement patterns of marine and freshwater sediments in the area.
  • Effects of toxic sediment on the local marine ecosystem, which includes important nurseries for commercial and recreational fisheries (including fish and molluscs), and is internationally recognised for its brachiopods and seaweeds.

According to the DoC website “Paterson Inlet is also an important habitat and nursery for at least 56 species of marine fish..Brachiopods or lamp shells are the most ancient of filter feeding shellfish…living examples are comparatively rare. Paterson Inlet is home to brachiopod species that live both on rock and sediment, thriving at depths of less than 20 metres. This makes the inlet one of the richest and most accessible brachiopod habitats in the world. Stewart Island/Rakiura has more varieties of seaweed than anywhere else in New Zealand. Paterson Inlet is home to 70% of them, including 56 brown, 31 green and 174 red kelp species. Seaweed forests provide habitat, protection and food to support diverse populations of fish invertebrates and marine mammals. Meadows of small red seaweed grow on the sand. They help to stabilise sediment as well as providing an important shelter for scallops, and a surface for spat and larvae to settle.”
Toxicity of brodifacoum to NZ’s marine invertebrates has not been investigated, although it was noted that paua were initially absent and slowly repopulated the area over the year following the spill at Kaikoura (Primus et al., 2005). In the Kaikoura spill monitoring, sediment sampling was only reported for nine days after the accident, and only 2 samples of seaweed in total were taken (Primus et al., 2005). Therefore DoC did not use that opportunity to gather much-needed information on the ecological effects on brodifacoum and they are largely unknown.

The AEE’s claim that “there was no evidence of a population decline in Spotties as a result of aerial application of Talon” (Empson & Miskelley, 1999) was misleading. The authors of the study admitted that it would only pick up a decline of 50% or more and their counts were highly variable between sites and sampling dates. Furthermore at 2 of 3 sites a short term decline was in fact apparent.

Animals living within the sediment in Paterson inlet include oysters, scallops, bryozoans, brachiopods and tubeworms (Wing, unpubl.).

  • The risk of exposure to the public, summed up as only significant if baits or contaminated animals are eaten in “substantial quantities” ignores that fact that brodifacoum is very persistent in animal tissue and cumulative (Weldon et al., 2011).  Being top predators, humans will be at risk from the toxin for many months.
  • There is no mention at all of sea birds which are an important part of marine ecology. Very large numbers of sea birds were killed in the aborted MacQuarie Island aerial brodifacoum operation and these same effects will be expected.


The AEE considered only alternative poisoning methods, using hand-broadcast baits or bait stations. Other control methods were discounted as unviable because they would require ongoing resources (as will any method because the threat of reinvasion is constant), prohibitive costs and by-kill (although I was assured the rat traps used on Ulva Island had no by-catch (Brent Beaven, pers. comm.)) and “elsewhere have not demonstrated the ability to recover vulnerable ground-dwelling species such as reptiles and invertebrates” (DoC, 2011).

However the relatively small size of the island, easy topography and ecology of Norway rats mean that alternative control methods probably are viable and should have been considered seriously. A 100 x 100m grid of bait stations was used on the island in the initial eradication of rats (followed up with trapping and poisoning), proving the viability of ground control and leaving a grid system that could be used again. The island is highly accessible and has accommodation. There is also a community group on Stewart Island keen to help with, and experienced in, rat control (Submission on application from Stewart Island/Rakiura Community and Environment Trust, 25th April 2011). Furthermore rat plagues are a problem for DoC (Sweetapple et al., 2006, Sweetapple & Nugent, 2007) and Ulva Island currently provides an ideal environment for developing control techniques that are safe for non-target species.

Norway rats are the largest of our rat species and prefer habitats near water. They build underground nests. The nest contains chambers for storing food, and for rearing young. Nests are surrounded by pathways that are habitually used by the rats, that leave urine trails that indicate sexual status among other things. In plentiful times, Norway rats are polygynous. Young male rats disperse after weaning, while females are likely to stay in the same nest ( Rats are highly neophobic, and avoidance of a new object can last from a few days to weeks, with some individuals never making contact with something foreign.

There is considerable knowledge of the best ways to trap Norway rats (Russell et al., 2008). Additional viable means of control include using detector dogs to find burrows, fumigating burrows with cyanide,, releasing radiocollared “judas” rats to find others, and using a confined (Delilah) rat to attract others. There is a definite need for field-trialling these techniques to refine them (Russell et al., 2008).

  • Monitoring
    In the AEE no monitoring of effects of the operation on non-target species was proposed. Effects on robins were expected to be recorded as part of the existing bird research on the island. Subsequently “mitigation measures” were proposed for the marine environment (Roberts, 2011) that included monitoring. However this proposed marine sampling protocol lacks any scientific basis whatsoever. Specific concerns are:
  • The monitoring is extremely narrow in scope. It addresses only brodifacoum contamination, only in adult forms, of very few species. It does not address possible effects on populations, or the likely cumulative effects that will arise in the food chain over time, or a wide range of organisms, or a range of life stages.
  • The monitoring has not been designed to provide good quality data that can be used to estimate levels of brodifacoum in the population the samples have been drawn from. The programme should be designed by a qualified marine biologist in consultation with a statistician. For example variables that may have large effects on brodifacoum content include spatial variation within each of the three sampling areas (eg between low and mid tidal zones, between one end of the defined area and the other end, between depths at the same tidal zone); variation with weather patterns (more sediment is likely to be in contact with the grazing and filtering species in rough weather); variation with feeding patterns and animal tissues (eg more sediment will be in the gut during feeding times). A properly designed monitoring programme will identify important variables and specify conditions and numbers of samples to be taken so that these variables are taken into account and the results are therefore meaningful. The sampling and analyses should be carried out by competent, independent technical personnel.
  • There is a danger of gathering no information at all according to this protocol due to part 4 “samples of particular species may not be taken in some zones if suitable habitat is not present” and part 5 “up to 10 samples of each fish species will be collected…if physically possible”. A properly designed sampling programme as recommended in (2) (above) will remove this problem.
  • There are 18 permanent DoC marine monitoring sites, for a range of indicator species, at depths of 1-12 m in the Ulva Island Marine Reserve (Wing, unpubl.). It is intended that “Information gathered following the initial survey will now provide the department with a check that its management of the marine reserve is effective and if the reserve is in a healthy state” (DoC website). Therefore if the poison operation was to go ahead, effective, informative and urgently needed monitoring on the effects of brodifacoum on marine life in the area could be carried out at these sites, using competent, independent technical staff.



DoC, 2011. Application for a Resource Consent. Environment Southland File No. D036-091. 80 pp.

Empson, A., Miskelly, C., 1999. The risks, costs and benefits of using brodifacoum to eradicate rats from Kapiti Island, New Zealand. New Zealand Journal of Ecology 23 (2): 241-254.

ERMA, 2007a. Agency’s Appendix F, p 482.

ERMA, 2007b. Agency’s Appendix N, p754.

Gerlach, J., Florens, V. 2000. Considering molluscs in rodent eradication projects. Tentacle- The Newsletter of the IUCN/SSC Mollusc Specialist Group 9: 7-8

Gerlach, J. 2005. The impact of rodent eradication on the larger invertebrates of Fregate island, Seycelles. Phelsuma 13: 44-54.

Hoare, J.M. Hare, K.M. 2006. The impact of brodifacoum on non-target wildlife: gaps in knowledge. New Zealand Journal of Ecology 30 (2): 157-167.

Howald, G., Donlan, C., Faulkner, K., Ortega, S., Gellerman, H., Croll, D., Tershy, B. 2010. Eradication of black rats Rattus rattus from Anacapa Island. Oryx 44: 30-40.

Innes, J., Barker, G. 1999. Ecological consequences of toxin use for mammalian pest control in New Zealand- an overview. New Zealand Journal of Ecology 23: 111-127.

Innes, J., Kelly, D., Overton, J. McC., Gillies, C. 2010. Predation and other factors currently limiting New Zealand forest birds. New Zealand Journal of Ecology 34 (1): 86-114.

Michel, P., White, H., Dickinson, K., Barratt, B., Fitzgerald, M., Johns, P., Nunn, J., Eyles, A., Barker, G., Andrew, I. 2008. Invertebrate survey of coastal habitats and podocarp forest on Ulva Island, Stewart Island, New Zealand. New Zealand Journal of Zoology 35: 335-349.
Primus, T., Wright, G., Fisher, P. 2005. Accidental discharge of brodifacoum baits in a tidal marine environment: A case study. Bull. Environ. Contam. Toxicol. 74: 913-919.

Roberts, A. 2011. Agreement on mitigation measures for the Ulva Island Rat Eradication. Memo to Te Whaka a Te Wera/Paterson Inlet Mataitai Committee, 28 April 2011.

Russell J.C., Towns D.R., Clout M.N., 2008. Review of rat invasion biology. Implications for island biosecurity. Science for conservation 286. 54 pp

Sinclair, L., McCartney, J., Godfrey, J., Pledger, S., Wakelin, M., Sherley, G. 2005. How did invertebrates respond to eradication of rats from Kapiti Island, New Zealand? New Zealand Journal of Zoology 32: 293-315.

Spurr, E.B., 1996. Environmental effects of rodent Talon baiting. Part II. Impacts of invertebrate populations. Science for Conservation 38: 12-27

Sweetapple, P., Nugent, G., 2007. Ship rat demography and diet following possum control in a mixed podocarp–hardwood forest. New Zealand Journal of Ecology 31(2): 186-201

Sweetapple, P., Nugent, G., Poutu, N., Horton, P., 2006. Effect of reduced possum density on rodent and stoat abundance in podocarp-hardwood forests. DOC Research and Development Series 231. 25 pp.

United States Environmental Protection Agency (USEPA). 1998. Reregistration eligibility decision (RED) rodenticide cluster. Prevention, Pesticides and Toxic Substances (7508W). EPA738-R-98-007.

Walker, C.S., Shetty, R.P., Kazuko, S.G., Letsou, A., Olivera, B.M., P.K., B. 2001. On a potential global route for vitamin K-dependent carboxylation in animal systems- evidence for glutamyl carboxylase in Drosophila. Journal of Biological Chemistry 276: 7769-7774.

Weldon, g., Fairweather, A., Fisher, P. 2011. Broadifacoum. A review of current knowledge. Department of Conservation Pesticide Information Reviews Series Dme No. DOCDM-25436. 73pp.

Wing, S. Baseline ecological monitoring of the Ulva Island / Te Wharawhara Marine Reserve. Dept of Marine Science, Otago University. 64 pp

Wright, G., Booth, L., Morriss, G., Potts, M., Brown, L., Eason, C. Assessing potential environmental contamination from compound 1080 (sodium monofluoroacetate) in bait dust during possum control operations. New Zealand Journal of Agricultural Research 45: 57-65.