Colonialism and Conservation

This article describes and analyses a recent island restoration project in which dingoes were used as ‘biocontrol’ agents to eradicate goats. Before they were taken to the island, the dingoes were implanted with poison capsules that were intended to kill them after they had killed the goats. Subsequently, freehold land on the island was marketed as a tourism development site. Our aim here is to elucidate how the creation of areas of ‘play’ for tourists (Part I) is prefaced by forms of ‘restoration’ that are also a colonial re-storying. We contextualise the Pelorus Island goat eradication project (Part II) with reference to ideas about the special role that islands play in conservation (Part III) and within the cultural–political history of carceral colonialism in Australia (Part IV). By considering the recent history of this ‘bizarre’ experiment, as it was called by Queensland Environment Minister Steven Miles (QP QWN 2016: p. 2976), with the surrounding islands’ carceral histories and their intimacy with tourism ventures, we seek to understand how the seemingly disconnected realms of settler colonialism, pest eradication programs and tourism are sutured together by fantasies of ‘restoration’: restoring places to a ‘pristine’ pre-invasion state and re-storying First Nations’ possession as a perpetual absence.

The concepts of biopolitics theorised by Michel Foucault (1978), and necropolitics and territorialisation from Achille Mbembe (2003) underpin our analysis. In his work of the creation of ‘vermin beings’, Chakanesta Clapperton Mavhunga further develops the concepts of biopolitics, necropolitics and territorialisation to make visible the connections between the construction of human and animal ‘pests’ and ‘pesticide’ (Mavhunga 2011). Mavhunga’s analysis enables us to consider the dingo as both pest and pesticide and assists our understanding of how territorialisation relates to islands as sites of tourism, restoration conservation and incarceration. This article explains how, in the Pelorus experiment (as we call it), particular animals were instrumentalised as pests and biocontrol tools. Following Mavhunga, we consider how territorialisaton, and biopolitical and necropolitical practices are utilised in the management not just of humans, but of animals too. By contextualising the Pelorus experiment with reference to the logic of elimination that constitutes Australian settler colonialism (Wolfe 2006) our intention is to elucidate how this form of so-called restoration conservation is part of and continues violent colonial, ecological and racial histories in Australia.

This violence is simultaneously hidden and appears in plain sight. Two Right to Information applications made by Animal Liberation (AL) form the empirical basis of our inquiry. The initial application was made in 2016 and the data obtained includes emails between the Pelorus project stakeholders, minutes of meetings of the Animal Ethics Committee (AEC) (largely redacted), Animal Ethics (AE) applications and reports, internal Queensland Department of Agriculture and Fisheries (DAF) situation reports, maps and media reports (Files A, B, C and E 2016).

This data is referenced in the text as File A, File B, File C and File E (all 2016). These files are uploaded on the Queensland Government Department of Agriculture and Fisheries, Current Disclosure Log: https://www.daf.qld.gov.au/our-organisation/right-to-information/disclosure-log/current, Date 27 October 2016, Application number 16-168.

Elsewhere we have addressed the instrumentalisation and discursive capture of the goats and dingoes involved in the Pelorus experiment as ‘ferals’ and ‘pests’ (Probyn-Rapsey and Lennox 2020), but their existence and experiences, for their own sakes, beyond being discursive figures for ‘ferality’, also motivates our research. We are animal studies researchers who work on dingoes, ferality and violence (Lennox 2013; 2014a; 2016; 2017; 2019a and b; 2021a and b; Probyn-Rapsey 2016; 2017; 2018; 2020; Probyn-Rapsey and Lennox 2020). Our emphasis on what happened to the goats and dingoes is a way of speaking back to conservation projects that claim their legitimacy through being able to ‘restore’ environments to their so-called original or unspoilt state while ignoring the perspectives of other inhabitants, human and non-human. Focusing on the experiences of the goats and dingoes is important because the violence inflicted on these animals is, we maintain, part of the larger violence of settler colonialism and its logic of elimination (Wolfe 2006), which includes introduced animals within its purview. The Pelorus experiment illustrates how illusions of idyllic island sanctuaries, which appeal to contemporary tourism tropes of seclusion within a ‘pristine’ environment, are predicated on the violence inherent to, but obfuscated within, settler-colonialist visions of land for the taking.

In August 2017 Chris Hemsworth, a Tourism Australia ambassador and actor who plays Norse god Thor in the Marvel superhero movies, celebrated his 34th birthday on Orpheus Island in the Great Barrier Reef with his partner Elsa Pataky, an actor and film producer, and a small group of friends. They stayed at the Orpheus Island Lodge, an exclusive resort with 14 rooms, villas and suites that accommodate just 28 people. According to its website the lodge offers guests ‘the ultimate in secluded, unspoilt tropical paradise’ (Northern Escape Collection, n.d.). Photographs of Hemsworth’s and Pataky’s weekend on Orpheus were posted on Hemsworth’s Instagram feed. The two stars perch on rocks, snorkel, paddleboard, embrace on the beach and relax by an infinity pool (Price 2017). Orpheus Island Lodge managers took them to another island, Pelorus, 800 metres north of Orpheus. They must have enjoyed their visit because at the end of April 2018, the Daily Mail reported that Hemsworth and Pataky were planning to invest four million dollars in a tourist development on Pelorus (Rolfe 2018), an island even more secluded than Orpheus and now made visible as a ‘destination’ via their 11.9 million Instagram followers.

Pelorus lies 15 kilometres off the coast of Lucinda, between Townsville and Cairns, in north Queensland. It is part of the Great Barrier Reef Marine Park and the northernmost of the Palm Island Group. Relatively small, four square kilometres, with a maximum elevation of 251 metres, its Indigenous name is Jabugay Yanooa and its traditional custodians are the Manbarra Aboriginal people whose songline tells the story of the creation of the islands in the Greater Palm Island Group ‘when an ancestral spirit, the Rainbow Serpent, broke up and left fragments of its body’ (Hooper 2008, p. 10). The registered cultural heritage body of the Manbarra people is the Manbarra Nanggarra Wanggarra Aboriginal Corporation, which includes Yanooa/Pelorus and Orpheus (its Indigenous name is Goolboddi) (QG DATSIMA 2008). But, unlike the other islands of the Palm Island Group, which are governed by the Palm Island Aboriginal Shire Council, Pelorus and Orpheus are part of the Hinchinbrook Shire Council (HSC) local government area. While Orpheus is national park, and the site of the Orpheus Island Lodge and a James Cook University research station, Pelorus is not designated national park; it hosts an Australian Maritime Safety Authority (AMSA) helipad and Aid to Navigation tower, which is apt because a ‘pelorus’ is a navigational aid, patented in 1854, used in conjunction with a magnetic compass or a gyrocompass for measuring the relative bearing of observed points. It was not until the 1880s that Jabugay Yanooa was renamed ‘Pelorus’ after British corvette HMS Pelorus, a flagship on the Australia station in 1860–1862.

The erasing of the Manbarra name Yanooa was preceded by the 19th-century frontier wars in which colonial settlers shot and poisoned Aboriginal people on and around the Palm Island Group and elsewhere. The Manbarra were ‘forcibly removed to the mainland’, though some remained on and eventually returned to Palm Island (Watson 2010, pp. 31, 40). European livestock—cattle, sheep and goats—were part of the colonial usurpation of Aboriginal land. Writer Bruce Pascoe describes how sheep destroyed Aboriginal yam pastures in Victoria, by grazing on the crop’s basal leaves, and compacted the soil (Pascoe 2014, p. 17). Settlers attacked Aboriginal people when they were harvesting; ‘croplands were mown down by sheep and cattle and people were prevented from protecting and utilizing their crops. No better device, short of murder, could ensure the weakening of the enemy’ (Pascoe 2014, p. 18). The goats brought to Pelorus in the 19th century were part of this colonial invasion. They were purportedly intended to provide meat, milk and fibre for mariners and lighthouse keepers. By winter 2016 there was a population of either 120-150 or 300 goats on the island—estimates differ, and no doubt numbers fluctuated according to breeding seasons and availability of resources.

In July-August 2017, a 4000 square metre (one acre) parcel of freehold land with a fully furnished two-bedroom beach house near a sandy beach on the south-east coast of the island, along with a speedboat and tractor, became the property of the Morris Group (Riley 2017), owners of the Orpheus Island Lodge as well as other tourism interests, including a company called Colonial Leisure, which owns hotels. Chris Morris, founder of the Morris Group, made his money from Computershare, a stock transfer company. The Morris Group website claims that ‘[t]echnology and innovation propelled our journey in the beginning and continue to drive everything we do’. They believe ‘economic success and environmental sustainability go hand-in-hand. We aim to exceed a “tread lightly” approach by strengthening and enriching the environments in which we work; actively restoring, repairing, and enhancing the natural ecosystems we operate within’ (Morris Group 2018). In February 2018 Hinchinbrook Shire Council invited expressions of interest for a ‘small scale, low intensity, eco-tourism development on Pelorus Island’ (HSC 2018).

In 2019 the beach near the house was being used as a snorkelling and picnic stop on Coral Expeditions’ Great Barrier Reef cruises. On their website Coral Expeditions describes Pelorus as their ‘very own tropical paradise’. On their scheduled afternoon stopover tourists are encouraged to relax and enjoy a barbecue ‘on the long stretch of golden beach’, take a guided walk through the rainforest, snorkel over a giant clam garden, see colourful corals from a glass-bottomed boat and take an ‘unforgettable’ scuba dive (Adventure Life 2019). In January 2020 the Morris Group were planning to build an exclusive boutique eco-lodge on the private beach ( Townsville Bulletin 2020). This proposed tourist development is only the most recent example of the island’s ‘re-storying’ from Yanooa to Pelorus; the species cleansing necessary to the exclusivity of the resort is not mentioned in the publicity.

The project of emptying Pelorus Island of its goat population was a trial scoped out by a university, a shire council, a state government, and the federal government’s Great Barrier Reef Marine Park Authority (File E 2016, p. 3). In July 2016 two male dingoes were brought separately to the south-west corner of the island and released. The first was trapped on 25 June on a grazing property on the mainland near Ingham after a wild dog management workshop run by University of Southern Queensland (USQ) wildlife ecologist Ben Allen; his father Lee Allen, a zoologist who works for Biosecurity Queensland, which is part of Queensland’s Department of Agriculture and Fisheries (DAF); and Hinchinbrook Shire Council (HSC) officers. Twenty-five people attended the workshop and, ‘to the delight of many land holders’, they caught two male and 14 female dingoes (File C 2016, p. 50).

The organisers held the ‘wild dog management’ workshop ostensibly to train land holders in dingo-trapping techniques. But the other reason was to find some very specific dingoes. On their Animal Ethics (AE) application the Pelorus experiment organisers outlined how ‘four young, healthy, male dingoes (preferably associates of each other or siblings) will be chosen to be released on Pelorus. All excess animal [sic] will be humanely euthanized by firearm by professional trappers, wild dog control officers and other authorised and skilled persons’ (File B 2016, p. 10). The AE application, called ‘Capture, handling, preparation, transport and release of dingoes from mainland North Queensland and onto Pelorus Island and the subsequent non-invasive monitoring of dingoes and goats’ describes how, after the ‘chosen’ dingoes were transported to the pound in Ingham, they would be sedated, given a health check and tick, flea and worm treatment, and:

The AE application proposes that at the completion of the project ‘All animals will either succumb to the toxin 1080 at 18–24 months via the controlled delivery device (microchip) or if the project is completed earlier than expected, then humanely euthanized by licenced firearm operator’ (File B 2016, pp.10–12). In May 2016 the DAF AEC approved the application, with conditions, which gave the organisers permission to use up to ten dingoes. (See Probyn-Rapsey and Lennox 2020 for more contextualisation of the AEC’s decision.)

The lack of males caught in the trapping workshop delayed but did not stop the implementation of the program. A black and tan dingo, one of the two males trapped on 25 June, was taken to Tropical Vet Services in Ingham where he was checked by a vet, neutered, and fitted with a GPS collar. On 1 July he was also implanted with a capsule that contained 10 milligrams of 1080 poison, ‘effectively setting his life to 600 days’ (File C 2016, p. 50). On 21 July a very young-looking golden dingo, who had also been trapped, held in the pound, checked, neutered, and fitted with a GPS collar, was implanted with a capsule of 1080 poison. He too was then released on the island.

A photograph in an online Australian Broadcasting Corporation (ABC) rural affairs Landline report shows a skinny, nervous-looking black and tan dingo wearing a very large satellite tracking collar running away from cameras. Another photograph in the report shows a family of goats gathered on a rocky outcrop over turquoise water. Two adult goats look calmly toward USQ researcher Ben Allen, who is taking the photo; two kids, close to their mother, gaze in profile over the Coral Sea (Schwartz 2016a). In another picture in the same report father-and-son dingo experts Ben and Lee Allen stand on a beach and smile into the sunshine. They look happy that they have a space in which to demonstrate their ‘novel’ method—as Peter Fleming, one of Ben Allen’s referees, describes it—of island restoration (File A 2016, p. 60). In the pictures the goats and the Allens look oblivious to the drama about to unfold.

The Landline report was designed to give the project positive publicity but the framing of the conservation project as an act of environmental rescue did not go to plan. Almost immediately, there was a public outcry about the cruelty of the scheme to both dingoes and goats, led by Animal Liberation and supported by dingo advocates and the RSPCA, which expressed concern about the welfare of goats under organised predation. RSPCA CEO Mark Townend said that the goats would be ‘partly eaten and then left to die a horrible painful death’ (Townend quoted in Schwartz 2016b). The Australian Conservation Foundation praised the ambition to control goats but ‘not at the destruction of these dingoes which they are using’ (Schwartz 2016b). In Queensland parliament, Steven Miles, Labor Minister for Environment and Heritage Protection and Minister for National Parks and the Great Barrier Reef, received a petition against the use of the 1080 capsules signed by 4645 people (STOP 2016). Another petition against implanting dingoes ‘with the horrendously crueltime-activated poison, 1080’, signed by 928 people, was sent to the RSPCA and Queensland police calling for the Allens, HSC and HSC Mayor Ramon Jayo to be charged with breaching Queensland’s animal cruelty laws (Prevent 2016). In July, in the midst of this public outcry, John Robertson, general manager of invasive species with Biosecurity Queensland, wrote in an email to the DAF media manager, ‘I think the project should run its course and prove itself before further commentary is made’ (File A 2016, p. 38). The project did end up running its course despite ongoing opposition: animal welfare groups and other institutions lodged complaints and grievances about its cruelty on ‘nearly all 365 days of 2017 and on 12 occasions in 2018’ via social media and in written form (HSC 2019, p. 10).

In mid-August 2016, Queensland Environment Minister Steven Miles intervened to order that no more dingoes should be released (there were plans to release two more), and that the dingoes already on Pelorus should be removed within 14 days. The reason given for the order, according to the minister, was because the dingoes’ presence would further endanger a population of vulnerable shore birds, the beach stone curlew. The day after Steven Miles’ Interim Conservation Order (ICO) was tabled in parliament, HSC issued a notice prohibiting public entry to Pelorus from 22 August to 5 September while they proposed to kill the dingoes by shooting them from helicopters or, reluctantly, because of its lack of likelihood to succeed and its high occupational health and safety risks to staff because of the terrain, by shooting them from the ground. At first the Pelorus organisers were also considering laying poison baits to kill the dingoes, but this idea was abandoned because it would require the approval of an Animal Ethics Committee. Australian Maritime Signals Association (AMSA), as a neighbouring property owner, and the caretakers living on Pelorus were notified that HSC would be conducting pest control with firearms. National parks were aware of the operations. Mariners were requested to observe a one nautical mile exclusion zone. In correspondence that reveals uncertainty about how isolated or interconnected Pelorus is, the exterminators were unsure of their obligations to notify residents of neighbouring islands (File C 2016, pp. 4–8).

It appears that an HSC pest controller flew to Pelorus on 24 August to ‘track and humanely destroy the dingoes’ (File C 2016, p. 11). But there were obstacles. After acquiring the VHF trackers, tablets, ammunition, and the GPS required, a vessel, the preferred form of transport, was not available. So the pest management officers had to use a helicopter. The helipad on the island was leased by Canberra-based AMSA and, to confound matters more, AMSA had scheduled maintenance on their Aid to Navigation tower on the island. AMSA wanted assurances from HSC and DAF that the discharge of firearms would not damage equipment at the site and that consideration had been made of the contractors working at the AMSA lease and tower area (File C 2016, p. 37). HSC assured AMSA that the ground-based shooting would be carried out by a ‘skilled licensed operator, who will have in his possession a GPS with the location of the tower and 2 spotters accompanying him’ and that ‘discharge of any firearms will be in the opposite direction of the tower at all times while contractors remain on site’ (File C 2016, p. 38).

Because of the cost and time associated with obtaining a permit to use the helipad, HSC tried aerial shooting first. By 31 August they had undertaken 14 hours of aerial shooting without managing to kill the dingoes, who were hidden in the thick canopy, dense vegetation, and rugged terrain (Honnery 2017), though they did shoot 45 goats from the air (Allen et al. 2020, p. E). After HSC obtained a permit for the helipad they tried ground shooting, also without success. Three days of bad weather prevented them from aerial and ground shooting (File C 2016, p. 43). They managed another seven hours of ground shooting on 1 September, also without success.

On 13 September an HSC pest controller informed AMSA that HSC staff would be ‘undertaking ground-based control activities’ from noon on 14 September until at least noon on 16 September but they would be landing on the front north-east face of the island, not the AMSA helipad (File C 2016, p. 52). This ‘very experienced’ ground shooter was assisted by a Hinchinbrook pest management officer. It was planned they would camp on the island for up to three days and ‘try all necessary means approved by the ethics committee to try and remove the dogs’ (File C 2016, p. 62). The ground shooters and spotters were unsuccessful in their attempts to kill the dingoes (File C 2016, p. 56) but they did shoot three goats (Allen et al. 2020, p. E). They returned to Pelorus to set foot-hold traps at a number of locations, to be monitored daily via a helicopter with an aerial shooter (File C 2016, p. 72). But the dingoes, after having been trapped two months earlier were wiser now and avoided the ten soft-catch traps that were laid out. One of HSC’s pest management officers thought that they might be staying away from the traps because of the daily flights of the helicopter to monitor them; he proposed sending the helicopter to fly over the south-western side of the island to herd the dingoes toward the traps (File C 2016, p. 14).

After HSC’s request for a second extension of the ICO was declined (File C 2016, p. 56), the council employed six sporting shooters who camped on the island for five nights in late September but also failed to kill the dingoes. At a meeting on 5 October 2016 in Ingham between Environment Minister Steven Miles, HSC Mayor Ramon Jayo and HSC CEO Dan McKinlay to discuss the Interim Conservation Order, Mayor Jayo told Miles that ‘the GPS units installed apparently weren’t designed to let them track the dogs in real time’ (Steven Miles quoted in Honnery, 2017). By 10 October the council, after having spent $60,000 trying to exterminate the dingoes, ceased their efforts, though they did shoot four more goats in November 2016 (Allen et al. 2020, p. E). After the ICO expired in December 2017, attempts to ‘humanely remove’ the dingoes ceased (HSC 2019, p. 3). In July 2018, HSC Mayor Ramon Jayo claimed that the Queensland government would be recompensing HSC $86,000 for their failed attempts to remove the dingoes, which, he said, vindicated the goat eradication program and showed it was legal (Bates 2018).

The preceding account of the Pelorus Island goat eradication scheme and its fallout has been reconstructed from data made public by AL’s RTI applications, media reports and academic research published by Ben Allen and coauthors. As outlined in the introduction, there are discrepancies between these sources. Inconsistencies between media reports (Bates 2018), communiqués (HLMAG 2018), Ben Allen and co-authors’ published research (Allen et al. 2020) and HSC’s annual report to the Animal Ethics Committee (AEC) (HSC 2019), which was made available by AL’s RTI application, are particularly salient in regard to what actually happened to the goats and dingoes, which we outline below. Also salient is how the availability or lack or availability of various technologies, and the technologies’ success or failure play important roles in both creating expertise and authority for the project’s organisers, and also demarcating the kinds of agency and resistance the dingoes and goats could (intentionally or inadvertently) exert. Motion-sensing cameras, satellite tracking collars and aerial surveillance are part of the kinetics of territorialisation of the experiment. Helicopters, firearms and the 1080 implants gave the project stakeholders seemingly omnipotent power over the dingoes and goats. Yet batteries ran out, the poison implants did not do what they were intended to do, cameras were stolen and required maintenance (HSC 2019, p. 5). Despite the data we do have about the dingoes and goats, these expert technologies provide a distant, mediated and, ironically, constrained view; the animals’ movements appear as straight lines on a satellite map; their bodies are flattened into glimpsed screen representations and packaged as photo opportunities for media consumption.

After their release on the island, the two dingoes were monitored through their GPS tracking collars and motion-sensing cameras. At first their collars were set to provide two GPS positions per day (Honnery 2017). A satellite map from 22 August shows them arriving on the south-west corner of the island and tracks their movements on the eastern slopes of the island, habitat for the highest densities of goats (Figure 1). They were released a few days apart but, according to the project’s organisers, they moved to the eastern slopes within hours and, after a few weeks, they found each other and began associating with each other. For several weeks they stayed on the eastern slopes (Allen et al. 2020, p. F).

Figure 1

By 6 September 2016 the dingoes had found food, water, and shelter. They were ‘successfully removing the feral goats’ and they were being monitored daily (File C 2016, p. 49). A satellite map dated 19 September shows their movements restricted to the centre of the island, possibly because of the aerial surveillance (File C 2016, p. 14) (Figure 2).

Figure 2

GPS collar monitoring was reduced to weekly in 2017 and monthly in 2018 to preserve battery life. Satellite images from January 2017, January 2018 and September 2018 show the dingoes traversing the island, visiting ‘feral goat harbourage areas’ and water sources (Figure 3) (HSC 2019, p. 2). In their 2019 report to the DAF Animal Ethics Committee, the project team call the dingoes ‘Black Dog’ and ‘Yellow Dog’ (HSC 2019, p. 3). We will call them Black Dingo and Yellow Dingo. Red lines represent the movements of Black Dingo, the black-and-tan dingo released first; yellow lines represent those of Yellow Dingo, the golden-coated dingo released second. In January 2017 Yellow Dingo ranges around the central eastern third of the island. Black Dingo ventures further north, south, and west. Neither visits the south-western corner, the dingoes’ point of ingress to and release on the island. In January 2018, Yellow Dingo travels to the northern and western part of the island. Black Dingo’s movements are more concentrated in the southern portion. Combined, their lines cover a greater proportion of the island than they did in January 2017. We have little information about how the two dingoes, social animals who usually live in family groups, interacted, co-operated and shared territory, though from the outset the project team intended for them to associate with each other. By September 2018 their combined lines traverse the whole island, apart from the south-west tip. They both visit the north-east coast. Black Dingo crosses the island on its north–south axis, Yellow Dingo on its east–west. The project’s organisers explain, ‘[t]he fact that larger distances are being covered in 2018 is probably a result of the abundance of prey’ (HSC 2019, p. 3), that is, they must travel further to find food. After March 2018, according to the project organisers, there were perhaps only two goats left on the island and, according to the team’s report to the AEC, one of them was still alive in December 2018 (see discussion below). After Black Dingo’s satellite collar ‘died’ on 26 September 2018 (HSC 2019, p. 3), the satellite image from January 2019 shows only Yellow Dingo’s movements, concentrated more in the south-east and clustering around a cove that was visited by neither dingo in January 2017 and 2018 and September 2018. Could Yellow Dingo be checking out the Coral Expeditions tourists?

Figure 3

Black Dingo survived for more than two years after his release, until at least 2 October 2018 when he was seen on camera. Organisers had programmed him to die on 19 February 2018 or 19 June (allowing for a 20 per cent margin of error). Yellow Dingo appeared on camera on 13 July 2019, well over a year after his programmed death date of 11 March 2018 (HSC 2019, pp. 2-3). Instead of ‘succumbing’ to the 1080 toxin as planned, the dingoes’ health improved (HSC 2019, p. 9). When they were last caught on cameral ‘both were in excellent body condition and appeared healthy in all respects, having increased their weight and general body condition from the time they were released’ (Allen et al. 2020, p. F).

The goats did not fare so well. In September 2019 Ben Allen explained how the dingoes ‘removed’ the kids rapidly, then the females, then the males (Allen 2019). By March 2017, after being predated on by the dingoes and being shot by the project team, a group of only about seven adult males were repeatedly observed on camera (Allen et al. 2020, p. E). In March 2018 five of them were shot from a helicopter at the base of a steep, inaccessible cliff. The shooters did not manage to kill two other goats observed at the time. During a ground survey in April 2018 the ‘last goat observed on Pelorus’ was a large male on the north-east slopes (Allen et al. 2020, p. E). In a published article the project team claim no goats ‘have been observed since that time on any camera trap, during an aerial survey in mid 2018, during a ground survey in November 2018, or during six boat surveys circumnavigating the entire island searching the coastal cliffs where goats had previously been most visible’ (Allen et al. 2020, p. F). This assertion is inconsistent with the team’s report to the AEC, which shows a black billy goat photographed on 4 December 2018 (HSC 2019, p. 5) and states he had been spotted a number of times during aerial surveillance in mid 2018, but had not been seen since. Information about how long the ‘last goat caught on camera’ (HSC 2019, p. 5) had been alone is inconsistent too. According to a communiqué from Michael Nash, chair of the Hinchinbrook Local Marine Advisory Group, there was only one goat left in April 2018 (HLMAG 2018). But in July 2018 HSC Mayor Ramon Jayo claimed there were a handful of goats left (quoted in Bates 2018).

Although the Pelorus organisers ignore animal cruelty when they claim that the project was success and that they need to conduct more ‘manipulative experiments’ (Allen et al. 2020), they were aware of the cruelty of the scheme. In a research article about the animal welfare considerations of using large carnivores as biocontrol tools the Allens and their coauthors write: ‘In our assessment we assumed that in most (if not all) individual predator-prey altercations the prey experiences an extreme level of suffering during both the chase and kill components … We also assumed that prey are aware of their impending predation’ (Allen et al. 2019, p. 165). In another article Allen et al. describe how, over the course of the experiment, goat behaviour changed in relation to human observers: at first the goats showed ‘casual disregard’, then ‘caution and wariness’, then ‘instant fear and flee responses’ (Allen et al. 2020, p. F).

In another erasure of the project’s failures, even on its own terms, organisers have provided no explanation of the failure of the so-called ‘1080 controlled release delivery system’. Indeed, when Scientec, the company that manufactured the 1080 capsules, was contacted by the Pelorus team about the failure of the implants to kill the dingoes, they responded that the capsules had not failed, that is, they had not slowly released a sub-lethal amount of 1080, and that they were still in use and probably have a much greater life expectancy than 720 days, ‘as this was the first time they had developed a capsule for such an extended period of time’ (HSC 2019, p. 3). Instead, the organisers claim that as goat numbers decreased from dingo predation, ‘island flora … thrived’ (HSC 2019, p. 5). In September 2019 at a public symposium Ben Allen showed pictures of vegetation growing on Pelorus and said that the slopes where the goats used to live ‘look fantastic now’, more like ‘the Garden of Eden’, and that the project had been a ‘successful island recovery program’, cheap and quick (Allen 2019). The concept of ‘recovery’ and ‘restoration’ in this context is a means of covering up and re-storying an inherently violent and problematic project.

Tropical islands have been ‘Marvelised’ since long before Hemsworth and Pataky spruiked Pelorus for the Morris Group, meaning that they have been settings for the creation of spectacles for consumption that are larger-than-life and may incorporate tropes of superheroes and cartoon-like, righteous violence, but these spectacles simplify or obscure the imperial and colonial history and power relations that produce the ‘island getaway’. The island that Prospero and his party of shipwreck survivors find themselves on in Shakespeare’s Tempest (c. 1611) is both a site that offers opportunity for rebirth, or a Golden Age as one of the characters puts it, and a strange risky place inhabited by barbarians. In June 2019 the surviving goat on Pelorus underwent a form of Marvelisation when he featured in a Courier Mail article describing the last dingo and the last goat on Pelorus facing off in the ‘ultimate survivor title contest’ (Michael 2019). The headline ‘Dingo and goat in fatal face-off on survivor island’ draws on the long-running television show Survivor, and The Hunger Games movies and books, to re-story the goat’s situation as a gladiatorial contest presented for human entertainment.

The island is, as environmental historian Richard Grove explains, a ‘symbolic location for the idealised landscapes and aspirations of the western imagination’ (Grove 1995, p. 3). Grove argues that tropical islands are central to the growth of environmentalism as a science and as an ethos or outlook because they are knowable environments; they can be physically, and mentally, circumscribed. European perceptions of islands were organised not just through metaphors and images, but through observations of the ‘economic impact of the coloniser on the natural environment’ (Grove 1995, p. 5), such as deforestation and the plantation system. Early French, Portuguese and Spanish visitors to the Azores, Canary Islands and Madeira observed the effects of deforestation and the decimation of Indigenous populations (Grove 1995, pp. 29-30). In 1492 Columbus noted the link between deforestation and changing rainfall in the Caribbean (Grove 1995, p. 30). Colonial enterprises not only produced large-scale environmental changes, they were also ‘conducive to rigorous analytical thinking about the actual processes of ecological change as well as thinking about the potential for new forms of land control’ (Grove 1995, p. 7). The visibility of environmental changes wrought by colonisation created new sensibilities about ‘wilderness’ and its protection. As the scale of environmental change accelerated from the 17th century, increasing human interest in ‘mechanistic analysis and comparison’ (Grove 1995, p. 51) enabled unprecedented empirical and administrative responses. The experimental conditions that islands afforded meant that decay and destruction could be measured: ‘[t]he geography of an island actually offered a contradictory set of opportunities: the social opportunity for redemption and newness as well as an encapsulation of problems posed by the need for physical and mental survival and health’ (Grove 1995, pp. 32-3). Into the 20th century islands were understood to be ‘delicately balanced’ environments where, as Rachel Carson writes, ‘nature has excelled in the creation of strange and wonderful forms’ (Carson 2014, pp. 109, 107).

In general, Grove developed his ideas on the relationship between colonial expansion and environmentalism by examining European responses to deforestation and changing climate. Here we extend Grove’s ideas to animals and their introduction to islands to map the coterminous agendas of colonial and conservation interests. Ideas of perfectibility have been implemented in the name of conservation on many islands. As the 2019 rat eradication project on Lord Howe Island demonstrates, violent conservation measures with considerable ill effects for many species have been considered justifiable to protect the flora and fauna of islands from ‘alien invaders’ and to return island environments to their pre-colonial ‘pristine’ state (Tiffen 2019). But Pelorus stands out because of the spectacular and novel ways it demonstrates how conservation and carceral colonialism share forms of so-called ‘necessary’ violence for such idealised visions to be made manifest.

The Pelorus Island goat eradication project organisers drew on millennia-old understandings of the impacts of goats when they wrote on their Animal Ethics (AE) application that feral goats ‘are a serious pest on many offshore islands worldwide, causing severe overgrazing, erosion, vegetation changes and ultimately flora and fauna extinctions’ (File B 2016, p. 12). In 6000 BC in the Southern Levant goats’ grazing on seedlings, saplings and shrubs stopped regrowth, which led to soil erosion on steep hillsides (Grove 1995, p. 18). In the 16th century goats were brought by the Portuguese to St Helena. The goats multiplied, trampled young trees, ate seedlings, and, with the help of humans, decimated the island’s gumwood, ebony and brazilwood forests leaving what Alfred Wallace described in the 1800s as a ‘rocky desert’ (Carson 2014, p. 111). Goats deforested South Trinidad after they were brought there by the astronomer Halley in about 1700, leaving ‘ghost forests’ and causing erosion of the island’s soft volcanic soil (Carson 2014, p. 112). On Pelorus the goats’ overgrazing led to changes in vegetation and ground cover that resulted in erosion, an increase in woody weeds and competition with native fauna (File C 2016, p. 61).

On their AE application the Pelorus organisers stated that there were no alternatives to using live animals as ‘biocontrol’ tools because other methods of attempting to remove goats from islands—such as aerial shooting, trapping, hunting, and using Judas goats followed by shooting—had failed. (See Rose 2008; and Celermajer and Wallach 2019 for more on using ‘Judas’ techniques to ‘manage’ feral animals.) They noted that ‘[a]cceptable examples for biological control for mammals are rare, and no natural pathogen to manage feral goats in Australia is available and acceptable … Using a native predator (the dingo) to control an introduced pest (feral goats) is a novel exception’ (File B 2016, p. 14).

Although the use of dogs and dingoes as ‘biocontrol’ agents is not common, human-orchestrated predation of dingoes on goats was not ‘novel’. A 2005 worldwide review of feral goat eradication programs reports that there have been three eradication programs involving canids; in the 1950s on Sunday Island in Victoria, Australia; in the 1990s on Klein Bonaire in the Netherlands; and in the 1990s on Townshend Island in Queensland (Campbell and Donlan 2005). Possibly the reason that eradication programs do not use canids as a biocontrol method is because dingoes and dogs do not follow the eradicators’ project plan. For example, the dogs on Sunday Island killed goats but also wallabies. In 1993 zoologist Lee Allen was involved in the project on Townshend Island in which 20 dingoes were released to eradicate a population of approximately 3000 goats. According to Lee Allen the dingoes killed the goats in a couple of years but they also established a breeding population that took 10–15 years to eradicate, and ate sugar gliders, lizards, insects, fish and crabs as well as goats (Allen, L. et al. 2012, p. 214). They also dispersed to Leicester Island, 500 metres away from Townshend (Allen et al. 2020, p. F).

In relation to the Townshend project, it was noted that dingoes are ‘not appropriate as a “biological control”…because dingo predation is not a target specific measure’ (DEWHA 2008, p. 8). To avoid these failures, the male dingoes of Pelorus were implanted with a ‘fail-safe’ (File A 2016, p. 4), the subcutaneous capsule of 1080 poison designed to release into their bodies and kill them in 600 days. But the ‘fail-safe’ failed and, in an echo of what happened on Townshend Island, the team’s report to the AEC reveals that the Pelorus dingoes’ diet became broader as ‘the remaining goats became more difficult to hunt’; the dingoes’ scats reveal that fruit and other small mammals were ‘slowly making up a larger proportion of their diet’ (HSC 2019, p. 4). There is no information in published reports or the RTI data we have had access to that outlines exactly what the dingoes did eat from March 2017 to March 2018 while it appears the number of goats on Pelorus remained constant at seven, or after March 2018 when the number of goats was possibly as low as two (Allen et al. 2020, p. E).

On their AE application, the Pelorus organisers describe dingoes as a ‘native predator’. But, as the AEC point out in their approval, on Pelorus they are ‘Class 2 pests under the current Land Protection Act and were not regarded as native animals in this area’ (File B 2016, p. 5). Consequently, under the Animal Care and Protection Act 2001, neither goats nor dingoes are entitled to the same animal welfare considerations as non-pest species (see Probyn-Rapsey and Lennox 2020). The ancestors of the Pelorus goats were domesticated; they were brought to Pelorus to be a ‘living larder’ for mariners and shipwreck survivors. They lived on the island for, maybe, 150 years. Nevertheless, they became classified as ‘feral’ because they ‘survive after deliberate release or escape from domestic herds’ (QG DAF 2016, p. 1) and as Class 2 pests because they are not ‘native to that area’ (QG 2000, p. 48). Dingoes, too, are ‘feral’ animals. In one summary, Allen et al. claim the Pelorus experiment demonstrates that ‘dingoes … exhibit the classical traits of an invasive species’ (2020, p. I). Dingoes do not fit into diochotomies of domestic/wild; native/introduced; or ecologically harmful/ecologically helpful (Lennox 2021b). Indeed, for the Pelorus experiment the role of the dingoes changed: they were enlisted as conservation saviours to kill the (feral) goats before it was assumed that they would become ecologically harmful and prey on other (non-feral) animals.

Killing ‘feral’ animals is an intrinsic part of ‘land management’ and conservation in Australia. Projects aimed at ‘controlling’ (a euphemism for killing) whole populations of many species including ‘wild dogs’/dingoes, foxes, cats, camels, horses, goats and donkeys, occur across all states and territories. Responsibilities for the management of ‘feral’ species rests at all levels of Australian society, from individuals who can be fined if found responsible for the illegal release of animals, or for not ‘controlling’ them as a condition of pastoral lease, to local councils and state governments, which are variously responsible through biosecurity and environmental protection laws. ‘Pest’ species such as the goats on Pelorus Island, and the dingoes brought over from mainland Queensland, are designated a ‘class’ status that determines where and how they can live and how they can be killed.

In these organising taxonomies, goats and, to a more limited extent, dingoes, are categorised according to a ‘nativist’ discourse that prioritises a species’ right to exist according to whether it is considered native to a region. Nativism is a particularly strong organising principle in Australian discourses of species belonging (see Franklin 2006). The nativist discourse within Australian conservation is remarkable for its role in constructing what Thom van Dooren describes as ‘exclusive ecological imaginaries’ that not only condone the death of non-natives, but also demand it ‘for the sake of any genuine conservation’ (Van Dooren 2011, p. 290). Militarised as an imperative to kill, nativism ‘renders countless species invisible, along with their unique and fascinating ecologies; it also exposes them to unfettered, unscientific, unmonitored, and unlamented mass killing programs’ (Wallach et al. 2019). In The Ethics and Rhetoric of Invasion Ecology, James Stanescu argues that the language that describes feral and ‘invasive’ species is frequently militarised (p. 17), while Subramanium argues that it also borrows heavily from racialised fear of foreign invasion (Subramanium 2017). Crowley et al. (2018) draw a typology of categorical thinking about feral animals (as having to be ‘sacrificed’ in order to ‘save’ others) to understand how they slip into a zone of being ‘killable’. In the Pelorus experiment, the killing of goats and dingoes is militarised as a defence of ‘nature’, with a logic of using ‘nature’ against itself (dingoes against goats) but fashioned as ‘biocontrol’.

Although it may seem paradoxical that a particularly macabre and cruel island restoration experiment was taking place while screen god Thor had some down time relaxing on a secluded Great Barrier Reef Island, Australia’s ‘idyllic’ islands have served as sites of violence throughout the country’s colonial history. Eradication programs aimed at controlling ‘invasive species’ share a carceral logic with this longer history. Australia itself was a carceral island after the arrival of the British in 1788 and the carceral system did not end with the arrival of the last ship of British convicts in Western Australia in 1868. Under the convict system, other islands such as Norfolk Island and Tasmania were used to house recidivists, who could be offshored yet again to places like Sarah Island in Macquarie Harbour in Tasmania. The use of islands for incarceration continues with the detention of refugees on Manus Island and Nauru. Here we examine how carceral islands have been used in the systematic attempted elimination of Aboriginal people from many parts of Australia and argue that under colonialism the treatment of Indigenous peoples and the eradication of pest animals are intimately related by an underlying ‘logic of elimination’, which includes genocide as well as other practices of replacement and displacement (Wolfe 2006).

Historian Patrick Wolfe stresses that settler colonialism is a ‘structure not an event’; its mechanisms and ‘logic’ are ongoing and complex, resulting in myriad tactics of elimination: captivity, removal, legal fictions to deny sovereignty (terra nullius), the erasure of frontier histories of violence and also the appropriation and forced inclusion of Indigenous people and cultures. Wolfe writes: ‘a logic that initially informed frontier killing transmutes into different modalities, discourses and institutional formations as it undergirds the historical development and complexification of settler society’ (2006, p. 402). He notes that the prioritising of settler colonial interests in land use, specifically mining and agriculture (2006, p. 395), are a feature of the perpetual territorial appropriation of Indigenous land by settler colonialism. Extending Wolfe’s list of institutional formations to include tourism and restoration conservation makes sense because both involve the creation of ‘ecological idylls’ that, in this case, do not acknowledge or involve Indigenous perspectives on the island itself. We found no discussion of the Indigenous significance of Pelorus in the RTI or published data, though Allen et al. acknowledge that the residents of Palm Island travel there with their canids by boat (Allen et al. 2020, p. C). The ‘pristine’ status being aimed for is, in some ways, a tacit recognition of the damage done by colonisation (which brought the goats), but the ecological restoration is ultimately to serve the HSC’s interests in tourism development, not to return it to Indigenous custodianship.

Restoration conservation is therefore implicated in the re-storying of land as somehow without both colonial history and Indigenous history. Restoration is always a retelling, a new fabrication, drawing a line in history via concepts of ecological purity that eliminates introduced animals and the reasons they are there (colonisation), thereby displacing Indigenous custodianship all over again.

In the case of Pelorus, the tourist venture or the tourist gaze is installed as the principal measure or outlook from which this ‘pristine’ status can be valued and monetized. Pelorus is not the only island to have been ‘restored’ and ‘restoried’ in such a way. Vasile Stanescu argues that the Nature Conservancy’s project to eliminate pigs from Santa Cruz Island also displaces and rewrites Indigenous histories. He writes that if ‘[t]he Nature Conservancy actually wanted to “return” the island to the same condition it existed in before the advent of western colonialism, they would have to return the control of the island over to Chumash people, who both still exist and—as late as 1984—were still suing in federal court the return of the island’ (Stanescu and Cummings 2017, p. 72). Instead, Stanescu points out, ‘[t]he Nature Conservancy is, in fact creating something new: a pristine “island paradise” free from human habitation even though this was not the case for over 10,000 years of the island’s history’ (p. 73). This perennial ‘newness’, a clean slate, is part of the discourse of elimination: denying ongoing Indigenous custodianship. Casey R. Schmidt argues that ‘restoration’ in the field of restoration ecology, promulgated by Aldo Leopold, one of its original progenitors, is shot through with Judeo-Christian myth and Biblical storeys of Eden as well as industrial modern warfare (Schmidt 2017, p. 118), making paradise and warfare its dominant motifs. These analyses fit the combat-style restoration project on Pelorus and the sort of tourist development it makes way for, which is prefaced on both the erasure of Indigenous custodianship and mass eradication of animals introduced as a feature of colonisation.

In his analysis of colonial Zimbabwe, Mavhunga extends Foucault’s concept of biopolitics (Foucault 1978, pp. 136-9) to populations beyond humans to show how settler-colonial culture shifts animals and humans into a ‘shared ontology of pesthood’ that renders both eradicable (Mavhunga 2011, p. 153); ‘pests’ are pushed not only outside the city limits, but beyond the limits of purposeful life itself. Mavhunga describes how the poison thallium sulfate was used to kill pest animals, baboons, in the 1960s and then people, African guerrilla fighters and civilians, in 1970s Rhodesia; both baboons and Africans are made to share the ontology of pesthood so that eradication can be justified (Mavhunga 2011). The shared ontology of pesthood enables the colonial state to exercise terror, or necropower or necropolitics, that is, the power of death over life (Mbembe 2003). Under colonialism, according to Achille Mbembe, there is a ‘racial denial of any common bond between the conqueror and the native. In the eyes of the conqueror, savage life is just another form of animal life’ (Mbembe 2003, p. 24). In the colony, Indigenous peoples ‘lack the specifically human character, the specifically human reality’ because colonisers regard them as behaving ‘like a part of nature’ (Mbembe 2003, p. 24). In 1883 British High Commissioner Arthur Gordon observed Queensland pastoralists’ virulent strain of this identification of Aboriginal people with animals: ‘The habit of regarding the natives as vermin to be cleared off the face of the earth has given to the average Queenslander a tone of brutality and cruelty … I have heard men of culture and refinement of the greatest humanity to their fellow whites … talk, not only of the wholesale butchery … but of the individual murder of natives, exactly as if they would talk of a days sport, of the having to kill troublesome animals’ (quoted in Lucashenko 2006).

The ability of colonial-settler states to exercise necropolitics is, Mbembe argues, intertwined with ‘colonially generated fantasies of wilderness and death and fictions to create the effect of the real’ (Mbembe 2003, p. 25). Central to colonial necropolitics is territorialisation (Mbembe 2003, pp. 25-6), that is, quarantine, reserves, prisons and fences, which are part of the ‘kinetics of colonizing and colonial settling’ (Mavhunga 2011, p. 155). On Pelorus, surveillance from helicopters and boats, and with GPS trackers and motion-sensor cameras enhances island territorialisation.

Necropolitics is evident in the 20th-century police massacres of Aboriginal people’s dogs and dingoes in the Northern Territory, which were designed to induce terror and to remind Aboriginal people of the times when they were victims of mass shootings. As anthropologist and ecocritic Deborah Bird Rose explains, ‘For people who had already been subjected to massacres, the dog shooting was a clear message of the right to kill with impunity. The power and terror show us a darker porosity to the West’s human–animal boundary: one in which humans are animalized so as to be killed with impunity’ (Rose 2011, p. 25). Whereas in the late 20th and early 21st centuries the killing of Aboriginal people more commonly happens behind closed doors—in police custody (Anthony and Cubillo 2020; Whittaker 2020)—the killing of their companions, dingoes, is open. Although there are critics of the excessive and spectacular human violence toward dingoes (dingoes are routinely poisoned in great numbers with 1080 and their bodies are hung en masse from trees and fences) violence towards dingoes is commonplace and encouraged by government policies and incentives such as bounties. Dingoes are associated with many social ills including the predation and mutilation of livestock, and the psychological effects of this predation on pastoralists, as well as loss of income to farmers. So-called scientific research on dingoes has long been dominated by agricultural interests and, until recently, much of it has been premised on the idea that it is necessary to kill them (see Lennox 2021a, pp. 219-20). With the evidence of what we know of the DAF AEC’s discussions, in some circles violence such as implanting dingoes’ bodies with 1080 poison capsules is not thought of as particularly shocking. There is no antidote to 1080. It causes a protracted and painful death. The poisoned animal’s vomit and carcass are toxic to other animals. The poison is soluble so it can leach into waterways. Rose writes, ‘1080 is transmitted though the dead bodies of animals who have ingested it, and it can keep the destruction moving in cascades of death across other species including birds’ (Rose 2012). She describes this ‘man-made mass death’ as a form of biocide, to parallel genocide. She observes that this will to destruction involves ‘imagining a future emptiness and then working systematically to accomplish that emptiness’ (Rose 2011, p. 82 emphasis added). This ‘future emptiness’ is a description of land emptied of Aboriginal people’s prior claims of sovereignty (terra nullius) and also seems apt as a description of islands primed for restoration, for re-storying and re-scripting with new owners in place.

Comparing Pelorus with another island K’gari (Fraser Island), which is also part of the tourist/carceral system, elucidates some of the differences between necropolitics and biopolitics. In another erasing of First Nations sovereignty, K’gari is still more widely known as Fraser Island. The name ‘Fraser’ comes from a sea captain who was stranded on the island with his wife and several other mariners after their ship was wrecked in 1836. Captain Fraser died on or near K’gari but his wife Eliza, and some other castaways, survived and subsequently told the infamous story of their so-called captivity at the hands of local Aboriginal people. Eliza Fraser’s narratives, along with others of the same genre, subvert the principle of colonial invasion and depict white settlers as victims of Indigenous captivity. In this way carceral logic is not reliably cast for one ‘side’ or another but can be mobilised to describe—or re-story and frame—colonial encounters in favour of settlers. The castaways noticed the dingoes of K’gari and their close relationships with Aboriginal people (Curtis 1836, pp. 39, 57). K’gari’s dingoes are now under the jurisdiction of the Fraser Island Dingo Conservation and Risk Management Strategy (Ecosure 2013). Unlike on the mainland, they are not usually killed in large numbers except in 2001, when over 30 were destroyed after two dingoes mauled a nine-year-old boy to death. Instead, now, selected dingoes are killed because of their acts of delinquency (Foucault 1995, p. 277) and because they represent ‘a kind of biological danger to others’ (Foucault 1978, p. 138) by modelling, to other dingoes, behaviour that threatens humans and by passing on what are seen to be aggressive-to-humans genes (Allen et al. 2015). The practice of targeting specific dingoes for destruction means that certain behavioural traits, such as staying away from humans, are selected (Allen et al. 2015). K’gari is also constituted as a camp, an island refuge, to protect the genetic ‘purity’ of its population of dingoes and to prevent extinction—envisaged as occurring through ‘hybridisation’ between dingoes and domestic dogs (Probyn-Rapsey 2020). On K’gari, only criminal dingoes are put to death: the old power over death is replaced by ‘the administration of bodies and the calculated management of life (Foucault 1978, p. 140). On K’gari the state conserves dingoes as a species and manages them biopolitically ‘to ensure, sustain, and multiply life, to put this life in order’ (Foucault 1978, p. 138).

Before K’gari was a camp for the biopolitical management of dingoes, it was a carceral centre for ‘blacks over whom firm control is a stern necessity’ (Archibald Meston quoted in Evans and Walker 1977, p. 83). The Aboriginal reserve on K’gari began in 1897, the same year the Queensland Aboriginals Protection Act forced the removal of First Nations people from their country to reserves. Initially, 33 Aboriginal men and youths and 18 women and girls were taken to K’gari from the mainland, ostensibly because colonial settlers complained about the spread of venereal disease from young Indigenous women to white men. The reserve became a mission in 1900. The people incarcerated there suffered from malnutrition, mumps, measles, syphilis, bronchial and chest complaints, including tuberculosis and hookworm disease. The mission was disbanded in 1904 and its surviving inmates were sent to camps in other parts of Queensland. Timber cutters took over the site and obliterated the 70 Aboriginal graves in the cemetery. That year Archibald Meston, former Southern Protector of Aborigines and head of the reserve, wrote a report for the Queensland parliament that outlines K’gari’s natural resources and, presciently, its tourist potential—which did not include dingoes: ‘If the dingoes, which are very numerous, were exterminated, the island would be an ideal spot for the preservation of all varieties of our native fauna (Meston 1905, p. 5, emphasis in original). Now, K’gari’s ‘untamed wilderness and rugged, natural charm’ (KBRV 2018), and its unique dingoes, are tourist attractions.

The fact that the mission/reserve on K’gari had terrible rates of sickness and mortality did not stop it from becoming a prototype for another more long-running island ‘penal settlement’ (Scott 1970). Palm Island lies less than 20 kilometres south of Pelorus. It is the biggest in the Palm Island group and, like Pelorus, it is, according to Aboriginal Law, also part of the body of the Rainbow Serpent and part of the registered cultural heritage body of the Manbarra Nanggarra Wanggarra Aboriginal Corporation (QG DATSIMA 2008). It was gazetted as an Aboriginal reserve in 1914. Two years later Chief Protector of Aborigines J.W. Bleakley described it as ‘the ideal place for a delightful holiday’. Its remoteness, he said, also made it suitable for use as a penitentiary for ‘individuals we desire to punish’ (quoted in Hooper 2008, p. 10). Between 1918 and 1971 ‘almost 4000 children of mixed race were forcibly removed from their families and country and … sent to Palm Island’ (Porter 2015). People from up to 57 different language groups from all over Queensland were forced to live there in appalling, overcrowded conditions. They were prohibited from speaking their languages and from conducting ceremony (Porter 2015; McQuire 2018). People had to work: ‘Even if you’re just limping along. Old men … They were bloody slaves those poor old men’ (Bill Congoo quoted in Watson 2010), and their wages were held by the government (Porter 2015). In 1974 Amnesty International described Palm Island as ‘little more than a concentration camp’ (quoted in Watson, 2010, p. 128). Along with the abuses and the paternalism was a strong tradition of resistance. As Melissa Lucashenko notes, from ‘a white perspective, Palm Island housed the “worst of the worst” blacks. For Murri people, the bravest and most outspoken were picked off and isolated on Palm Island, where they could do little damage to an oppressive system’ (Lucashenko 2006).

Palm Island, or Bwgcolman as it is known to its 2000 Aboriginal inhabitants, became notorious as a violent place, especially in the wake of the 2004 death in police custody of Indigenous man Mulrunji Doomadgee. His autopsy revealed a cut above his right eye, four broken ribs, a ruptured portal vein and a liver ‘that had been almost cleaved in two’ (McQuire 2018) but the coronial inquest ruled that his death was an accident. ‘He “had fallen on concrete” while in police custody’ (Lucashenko 2006). Mainstream press coverage of the Bwgcolman people’s ensuing outrage about police brutality was, as Amanda Porter writes, sensationalist and served to justify the heavy handed police response that followed (Porter 2015). Six helicopters were sent to the island with white police in riot gear. Houses were raided, 19 people were arrested for riot and police intimidation of locals continued for months. Melissa Lucashenko’s comment that ‘Palm Island briefly became an Arnie movie’ (Lucashenko 2006) underscores Amanda Porter’s assertion that Aboriginal deaths in custody, over-policing, harassment, and over-surveillance are all forms of neo-colonial violence (Porter 2015). But, as Amy McQuire points out, non-Indigenous Australians are rarely shocked or outraged by this violence ‘because it is normalised. It is seen as legitimate violence’ (McQuire 2020).

J.W. Bleakley’s vision of Palm Island as both penitentiary and holiday resort is not as contradictory as might first appear—indeed the project of making a place ‘safe’ for white tourism through removal and incarceration of original owners is as good a definition of colonialism as any (see Jamaica Kincaid’s A Small Place). Palm is one of a string of Queensland islands, including Eumilli/Fantome, Garoogubbee/Eclipse, Koba/Fitzroy and K’gari/Fraser, that have served as carceral sites for Aboriginal people. These islands make the fantasy of uninhabited space possible for those who are not incarcerated. Eradicating Aboriginal people from their country and disappearing them to carceral sites enables an illusion of ‘open’, terra nullius horizons for non-Indigenous people—be they colonial settlers or contemporary tourists. Islands have been constructed as spaces of separation and incarceration throughout Australia’s settler colonial history for people rendered pests and pests rendered uncontrollable. In this way, colonisation is also ecological, involving the displacement and importation of animals: ‘By the 1890s, even the wildlife was being imprisoned, with Rattlesnake Island, south of Palm off the coast at Kurukon, defined as a “a penal settlement for town goats”!’ (Meston 1895, p. 130).

While K’gari is a site of biopolitics, Pelorus is a site of necopolitics. The so-called self-culling dingoes of Pelorus are a good example of Mavhunga’s ‘pesticide’; they embody not only the tools and poisons used to kill pests, but also the theory and practice of killing them. They are both bait—pest controllers, instrumentalised to exterminate goats—and baited—expendable once their job as goat eradicators is done. The project’s organisers use this ‘natural’ predator–prey relationship to justify the experiment (File A 2016, p. 4), claiming the goats are the objects of the dingoes’ ‘affections’ (Ramon Jayo quoted in Schwartz 2016a). But they intended the Pelorus dingoes to be killed by pesticide after they had carried out their function as pesticides. In the dingoes’ bodies the poison 1080 and the theory of pest-making collapse into each other.

The Pelorus Island goat eradication project is a manifestation of the long trajectory of colonial expansion in which tropical islands have played a significant role in European idealisations of paradise and also in the origins of western environmentalism and conservation (Grove 1995, p. 13). The Pelorus experiment is part of a tradition that subscribes to the notion that ‘[p]aradise had become a realisable geographical entity, or so it seemed’ (Grove 1995, p. 51). In a re-storying of the eradication program’s cruelties and failures, and the inconsistencies of its organisers’ findings, this restoration project is nevertheless intended to be a pilot for a Golden Age of conservation projects: HSC claimed that it would ‘demonstrate best practice and make recommendations to land managers responsible for the other islands inhabited by feral goats, including other islands in the Great Palm Group’ (HSC 2016, p. 9) and, importantly, that it would be the first in a series of island restorations: ‘Once this island is successful, it will set the platform for many other island managers to follow through and carry out similar projects’ (HSC pest management officer Matthew Buckman quoted in Schwartz 2016a). The links between conservation and tourism can also be traced to the colonial expansion of Europe. Travel, too, is a product of imperialism. Travel literature became the single most sought after category of English literature in the 18th century, and, as Richard Grove points out, ‘In such literature, it might be argued, countries had become commoditised, objectified and made subjects for the European traveller, merchant and scientist’ (1995, p. 54).

But no island is an island. The precautions that the Pelorus organisers took—using only male dingoes, desexing them as well as implanting them with 1080 capsules—show that they were aware that Pelorus is not completely isolated. The Manbarra people are also known as the Wulgurukaba or canoe people. The Pelorus experiment organisers know that some of the free-ranging population of dogs on Palm Island travel by boat with people between islands (Allen et al. 2020, p. C). In this ‘sea country’ (GBRMPA 2019, p. 12) ‘sea waters enable mobilities, the forging of connections’ (Kothari 2020). Yanooa/Pelorus and Bwgcolman/Palm are, after all, part of the same body, the body of the Rainbow Serpent whose tail is at Halifax Bay and whose head can be seen in the Arcadia headland on Magnetic Island, off Townsville. Nevertheless, the notion that Pelorus, like all the other carceral islands, is a place away from the public gaze, a place where necropolitics and the public messages around it can be contained and controlled, enabled the organisers of the project to conceive and put into practice an experiment that would usually be hidden in a laboratory or scientific research facility.

The violence inflicted on the goats and dingoes of Pelorus is part of the structural violence of settler colonialism. Jabugay Yanooa became Pelorus. Dingoes became pesticide and pest. The island is territorialised as a site of eradication, ‘restoration’ and tourist development. Pelorus’s isolation enabled the Allens, HSC, DAF and the Great Barrier Reef Marine Park Authority to conceive the goat eradication project. The objective of the project, to create a pristine and secluded island paradise for tourists, appealed to the Morris Group, Hemsworth and Pataky, and their millions of Instagram followers. But the island’s isolation is a mental construct as much as it is a physical condition. Pelorus is not isolated. It is part of the body of the rainbow serpent. It sustained goats who sustained mariners. It sustains dingoes. It is part of an Australia-wide network of Aid to Navigation towers. It is planned to be the first, the pilot, in a new, cheap, quick, effective form of biocontrol. As Morris Group executive director Hayley Morris explains, the Pelorus tourism development, more exclusive and more isolated than the Orpheus Island Lodge, was possible because of the relationship between Pelorus and Orpheus: ‘It would be connected to Orpheus but be more of an exclusive place that a single group would take, either a family or three to four couples that want to be completely isolated … It would have rooms that are connected, more a home style I guess, and it would still be serviced via Orpheus, but you would have a chef and stuff like that’ (quoted in Rolfe 2018). It is isolated but connected—connected to Orpheus and to Bwgcolman and K’gari, connected to colonial histories of incarceration and pesticide.