REDONDA RESTORATION PROGRAMME: Technical report on the eradication of black rats ( Rattus rattus) from Redonda, Antigua & Barbuda

Cover illustration: Magnificent frigatebirds, Fregata magnificens, on their nest sites on Redonda (Photo: Elizabeth Bell, WMIL).

THE OPINIONS OF THE AUTHORS DO NOT NECESSARILY REFLECT THE OPINIONS OF THE PARTNER ORGANISATIONS OF THE REDONDA RESTORATION PROGRAMME.

PUBLISHED BY: Redonda Restoration Programme, St. John’s, Antigua

RECOMMENDED CITATION: Bell, E.A., Ibbotson, J. & Daltry, J.C. (2017) Technical report on the eradication of black rats Rattus rattus from Redonda, Antigua and Barbuda. Report from Wildlife Management International Ltd and Fauna & Flora International to the Redonda Restoration Programme, St John’s, Antigua and Barbuda.

Author details: Ms. Elizabeth ‘Biz’ Bell Dr Jenny Daltry, FRGS Senior Ecologist Head of Caribbean and Senior Conservation Biologist Wildlife Management International Ltd Fauna & Flora International PO Box 607, The David Attenborough Building Blenheim 7240 Pembroke Street, Cambridge CB2 3QZ New Zealand United Kingdom United Kingdom www.wmil.co.nz www.fauna‐flora.org Email: [email protected] Email: jenny.daltry@fauna‐flora.org

Mr. Jack Ibbotson Loddon, Main Street Long Riston East Yorkshire United Kingdom Email: [email protected]

All photographs in this report are copyright ©WMIL unless otherwise credited, in which case the person or organization credited is the copyright holder.

The Redonda Restoration Programme was launched in 2016 by the Government of Antigua and Barbuda, Environmental Awareness Group, British Mountaineering Council, Fauna & Flora International, Island Conservation and Wildlife Management International Ltd. This partnership is working to rehabilitate Redonda into a healthy island ecosystem that is sustainably managed for the conservation of indigenous biodiversity, and to preserve Redonda’s important historical values and facilitate the sustainable use and enjoyment of Redonda by present and future generations. The development of this plan was funded by the National Fish and Wildlife Foundation and Darwin Initiative (#23‐003: Eradicating Invasive Species from the Highest Priority Caribbean Island).

Bell, Ibbotson & Daltry i

EXECUTIVE SUMMARY 1. Redonda has been internationally recognised as an Important Bird Area and supports rare and unique animals and plants of both national and global importance, including seabirds such as magnificent frigatebird Fregata magnificens, masked booby Sula dactylatra, red‐ footed booby S. sula and brown booby S. leucogaster, as well as smaller numbers of red‐ billed tropic bird Phaethon aethereus, brown noddy Anous stolidus and bridled tern Sterna anaethetus and endemic reptiles such as the Redonda ground lizard Ameiva atrata and Redonda tree lizard Anolis nubilus and an unnamed dwarf gecko (Sphaerodactylus sp.) 2. The Redonda Restoration Programme is a collaborative programme of the Government of Antigua & Barbuda, Environmental Awareness Group (EAG), Fauna & Flora International (FFI), British Mountaineering Council (BMC), Wildlife Management International Ltd (WMIL) and Island Conservation (IC), with support from Darwin Initiative through UK Government funding, National Fish & Wildlife Foundation, Taurus Foundation, Global Wildlife Conservation, Syngenta Crop Protection AG and Caribbean Helicopters Ltd. 3. This partnership is working to rehabilitate Redonda into a healthy island ecosystem that is sustainably managed for the conservation of indigenous biodiversity, and to preserve Redonda’s important historical values and facilitate the sustainable use and enjoyment of Redonda by present and future generations. 4. Redonda has a large population of invasive alien black rats (Rattus rattus). Together with feral goats (Capra hircus, which are to be relocated), these rodents are heavily implicated in the severe and ongoing decline of the island’s vegetation and native wildlife populations, and a number of species have become extinct. The once‐forested island has reached advanced stages of desertification. 5. The eradication of black rats from Redonda was completed using a combination of ground‐ based and aerial methods between 13 February and 7 April 2017. 6. There were 327 bait stations on the top of the island completed using ground personnel. 7. There were 95 bait stations and bait points on the cliffs completed using rope‐access personnel. 8. There was approximately 26 hectares completed by aerial application of bait from a helicopter. 9. Grain‐based wax blocks (Klerat® and Final®) containing brodifacoum were used. Bait stations contained 20g blocks and 5g blocks were used for broadcasting from the helicopter or by hand. 10. The eradication of black rats from Redonda appears to be successful with the last rat sign recorded on 8 March 2017. 11. Rats consumed approximately 170 kg of bait in the bait stations on the top, beaches and cliff‐access areas of the island. 12. The rat population was high. It was estimated that there were between 5000 and 7500 rats present on Redonda. 13. In line with common international practice, Redonda can only be officially declared rat‐free if no further signs of rats are detected within two years (April 2019). 14. Bait station interference by land crabs was high.

Bell, Ibbotson & Daltry i

15. Except for one peregrine falcon, there was no evidence that other non‐target species were affected by the eradication project. 16. Redonda Restoration Programme, Environmental Awareness Group and Department of Environment staff and volunteers were trained in eradication, long‐term monitoring and rodent detection techniques. 17. Although the birds were not found, Audubon’s shearwaters were recorded calling on the ground for the first time on Redonda. 18. Long‐term monitoring of 39 permanent bait stations will be carried out by Redonda Restoration Programme, Environmental Awareness Group and Department of Environment staff and volunteers between April 2017 and March 2019. 19. Biosecurity measures such as permanent bait stations and monitoring points were established on the island following the completion of the eradication operation. These points will be monitored indefinitely by Redonda Restoration Programme, Environmental Awareness Group and Department of Environment staff and volunteers as part of the biosecurity requirements to maintain the rat‐free status of Redonda. 20. Research and monitoring of the seabirds, land birds, reptiles and vegetation should continue after the eradication programme to assess the effects of removing rats on the Redonda ecosystem. Redonda Restoration Programme, Environmental Awareness Group, Department of Environment staff and volunteers and other researchers from groups such as RSPB should undertake this work with technical support from Fauna & Flora International.

Bell, Ibbotson & Daltry ii

TABLE OF CONTENTS EXECUTIVE SUMMARY i TABLE OF CONTENTS iii ACKNOWLEDGEMENTS v TECHNICAL REPORT OF THE ERADICATION OF BLACK RATS (RATTUS RATTUS) FROM REDONDA, ANTIGUA & BARBUDA 1 1 INTRODUCTION 1 1.1 REDONDA 2 1.2 BLACK RAT, RATTUS RATTUS 7 2 GOAL AND OBJECTIVE OF REDONDA RAT ERADICATION 8 3 METHODS OF THE REDONDA RAT ERADICATION OPERATION 8 3.1 ESTABLISHMENT OF THE GROUND BAIT STATION GRID 9 3.1.1 BAIT STATION DESIGN .... 9 3.1.2 GRID DENSITY .... 9 3.2 ESTABLISHMENT OF THE CLIFF AND ROPE‐ACCESS BAIT STATION GRID 11 3.3 AERIAL APPLICATION AREA 13 3.4 POISONING PHASE OF THE REDONDA RAT ERADICATION 14 3.4.1 POISON TYPE USED FOR THE DOG ISLAND RAT ERADICATION .... 14 3.4.2 POISONING PHASE OF THE REDONDA RAT ERADICATION .... 15 3.5 TRAPPING 18 3.6 INTENSIVE MONITORING PHASE OF THE REDONDA RAT ERADICATION 18 3.7 ESTABLISHMENT OF THE PERMANENT BAIT STATION (LONG‐TERM MONITORING) SYSTEM ON REDONDA 18 4 RESULTS 19 4.1 BAIT ACCEPTANCE AND BAIT TAKE BY RATS 19 4.2 MONITORING FOR RATS 24 4.2.1 LONG‐TERM MONITORING FOLLOWING RAT ERADICATION ON REDONDA .... 26 4.3 INTERFERENCE BY NON‐TARGET SPECIES 26 4.3.1 INTERFERENCE ON BAIT AND BAIT STATIONS BY NON‐ TARGET SPECIES .... 26 4.3.2 INTERFERENCE ON MONITORING POINTS BY NON‐ TARGET SPECIES .... 31 4.4 OTHER OBSERVATIONS DURING THE RAT ERADICAITON ON REDONDA 34

Bell, Ibbotson & Daltry iii

4.4.1 LIZARD OBSERVATIONS DURING THE RAT ERADICATION ON REDONDA .... 34 4.4.2 BIRD OBSERVATIONS DURING THE RAT ERADICATION ON REDONDA .... 34 5 PROBLEMS DURING THE RAT ERADICATION ON REDONDA 34 5.1 ACCESS GULLY 35 5.2 NON‐TARGET SPECIES 35 6 VERIFYING THE RAT ERADICATION OPERATION ON REDONDA WAS SUCCESSFUL 36 7 BIODIVERSITY MONITORING 38 8 BIOSECURITY 38 9 CONCLUSIONS 40 10 RECOMMENDATIONS 41 11 REFERENCES 42 12 APPENDICES 47 12.1 TIMETABLE OF ACTIVITIES FOR THE REDONDA RAT ERADICATION (29 JANUARY TO 13 APRIL 2017) 47 12.2 BIRDS SEEN DURING THE REDONDA RAT ERADICATION (JANUARY TO APRIL 2017) 54 12.3 RATS TRAPPED OR RECOVERED ON THE SURFACE DURING THE REDONDA PRE‐ASSESSMENT VISIT (JULY 2016) AND RAT ERADICATION (JANUARY TO APRIL 2017) 56

Bell, Ibbotson & Daltry iv

ACKNOWLEDGEMENTS This work was carried out under the auspices of the Redonda Restoration Programme, a collaborative programme of the Government of Antigua and Barbuda, Environmental Awareness Group (EAG), British Mountaineering Council (BMC), Fauna & Flora International (FFI), Island Conservation and Wildlife Management International Ltd (WMIL). This programme was launched in 2016 with funding from Darwin Initiative through UK Government funding (grant #23‐003), National Fish & Wildlife Foundation and Global Wildlife Conservation, and additional technical and material support from Caribbean Helicopters and Syngenta Crop Protection AG. Additional staff time, international flights and equipment were provided by Fauna & Flora International (FFI) and Wildlife Management International Ltd. The authors would like to thank to the following people and organisations: The Government of Antigua and Barbuda—in particular the Department of Agriculture, Lands, Housing and Environment and Department of Environment—for their kind permission to conduct field work and for their interest, support and advice throughout the development of this component of the Redonda Restoration Programme. The Department of Environment graciously hosted planning meetings in 2016 and facilitated the import of equipment. Shanna Challenger, Redonda Project Coordinator, for her key on‐going logistical support role co‐ ordinating everything to do with the project, advice, assistance in the field and on Antigua, and organisation of meetings and introductions to stakeholder groups. All the Redonda Rat Eradication Interns; Chris Clarke, Thea Eldred, James Stops, Salina Janzan, Silas Walton, Bede West, John Tayton and Edward Marshall for their hard work, enthusiasm and support during the project. The success of the project so far, and in the future, is a credit to them. Sophia Steele, FFI Eastern Caribbean Coordinator, for her assistance with the logistical arrangements on Antigua, assistance in the field and ongoing support of the project. Natalya Lawrence, Coordinator, Offshore Islands Conservation Programme, EAG, for her advice, assistance and information on the project, and organisation of meetings and introductions to stakeholder groups. The EAG Board Members also provided very helpful support and advice. Greg Scott, Caribbean Helicopters, for his ongoing support of the project and expert advice and information on all helicopter aspects of the work. Thanks also to pilots, Lucas Adkins and Cordell Walker, and ground crew XX XX for their assistance with the aerial baiting and other transport requirements. Thanks to Karen and Wendy for all their logistical and scheduling assistance. Dave Turnbull and Adam Long, British Mountaineering Council, for permission to use their map of Redonda and advice on the climbing access aspects of this project. Adam Long participated in further surveys of Redonda and planning meetings in 2016. Syngenta Crop Protection AG, for the supply of over four tonnes of Klerat® bait and their interest and support of the project. Craig Reikena, Bell Laboratories Ltd., for his advice and provision of Final® bait for the Redonda rat eradication operation.

Bell, Ibbotson & Daltry v

Wildlife Management International Ltd staff, particularly Kelvin Floyd for developing GIS systems, maps, figures and databases for the eradication work on Redonda and producing some of the maps and figures for this report and others are thanked for reading and commenting on this report. Fauna & Flora International staff provided vital administrative support during the development of this report. We also thank other individuals for peer‐reviewing this plan.

Bell, Ibbotson & Daltry vi Redonda Rat Eradication: Technical Report

TECHNICAL REPORT OF THE ERADICATION OF BLACK RATS (RATTUS RATTUS) FROM REDONDA, ANTIGUA & BARBUDA

1 INTRODUCTION Redonda has been internationally recognised as an Important Bird Area and supports rare and unique animals and plants of both national and global importance, including seabirds such as magnificent frigatebird Fregata magnificens, masked booby Sula dactylatra, red‐ footed booby S. sula and brown booby S. leucogaster, as well as smaller numbers of red‐ billed tropic bird Phaethon aethereus, brown noddy Anous stolidus and bridled tern Sterna anaethetus and endemic reptiles such as the Redonda ground lizard Ameiva atrata and Redonda tree lizard Anolis nubilus and an unnamed dwarf gecko (Sphaerodactylus sp.) (Morse, 1979; Malhotra & Thorpe, 1999; Office of the Prime Minister, 2001; Daltry, 2007; Prosper et al., 2008; Sylvester et al., 2009; Devine et al., 2010; BirdLife International, 2012; Bell & Daltry, 2012; Hedges & Conn, 2012; Lowrie et al., 2012). Redonda has a large population of invasive alien black rats (Rattus rattus). Together with feral goats (Capra hircus, which are to be relocated), these rodents are heavily implicated in the severe and ongoing decline of the island’s vegetation and native wildlife populations, and a number of species have become extinct. The once forested island has reached advanced stages of desertification. Black rats are one of the most widespread and devastating invasive species around the world and are recorded on 80% of the world’s islands (Moors & Atkinson, 1984; Atkinson, 1985; Jones et al., 2008). Rats have devastating impacts on islands through predation, competition and habitat modification (Towns et al., 2006; Jones et al., 2008), but have been successfully removed from islands ranging in size from 1 to 36,000 ha throughout the world including a number of islands throughout the Caribbean region (Day & Daltry, 1996; Daltry, 2000; Varnham & Daltry, 2006; Howald et al., 2007; Witmer et al., 2007; Parks & Wildlife Service, 2008; Varnham, 2010; Bell, 2012; Parks & Wildlife Service, 2014; Martin & Richardson, 2017). These operations use techniques developed in New Zealand with methods ranging from ground‐based (bait station or hand spreading) to aerial broadcasting operations or in some cases a combination of methods (Atkinson, 1985; Towns & Broome, 2003; Howald et al., 2007). The best method depends on the physical characteristics and ecological characteristics (specifically the risk to non‐target species) of the island and the species of rat being targeted. More than a dozen black rat eradication operations have been carried out on islands around Antigua under the auspices of the Antiguan Racer Conservation Project and Offshore Islands Conservation Programme: specifically, Great Bird, Galley Major, Galley Minor, Rabbit, Redhead, Lobster, Little Lobster, Codrington, Maiden (West), Maiden (East), Green and York (e.g. Day & Daltry, 1996; Daltry, 2000; Varnham & Daltry, 2006; Witmer et al. 2007; Havery et al., 2014; Lawrence & Daltry, 2015). The largest of these is Green Island, at 45 hectares. All of these operations involved Antiguan and Barbudan personnel (staff and volunteers from the Forestry Unit and Environmental Awareness

Bell, Ibbotson & Daltry 1 Redonda Rat Eradication: Technical Report

Group) working alongside international organisations (specifically, Fauna & Flora International, Durrell Wildlife Conservation Trust and Island Resources Foundation). The Redonda Restoration Programme is a collaborative programme of the Government of Antigua & Barbuda, Environmental Awareness Group (EAG), Fauna & Flora International (FFI), British Mountaineering Council (BMC), Wildlife Management International Ltd (WMIL) and Island Conservation (IC), with support from Darwin Initiative through UK Government funding, National Fish & Wildlife Foundation, Taurus Foundation, Global Wildlife Conservation, Syngenta Crop Protection AG and Caribbean Helicopters Ltd. The feasibility study was prepared in 2012 (Bell & Daltry, 2012) and the Operational Plan was prepared in 2016 (Bell & Daltry, 2016). Elizabeth Bell (Wildlife Management International Ltd, WMIL) was contracted by FFI to lead the black rat eradication phase of the Redonda Restoration Programme. This partnership is working to rehabilitate Redonda into a healthy island ecosystem that is sustainably managed for the conservation of indigenous biodiversity, and to preserve Redonda’s important historical values and facilitate the sustainable use and enjoyment of Redonda by present and future generations. This technical report details the black rat eradication operation including methods, results, issues, logistics and timetable for the black rat eradication on Redonda. Day to day information on the eradication can be found in Appendix 12.1. 1.1 REDONDA Redonda (16°56’N; 62°21’W) is the third largest island in Antigua and Barbuda; lying 56 km southwest of Antigua, 22 km northwest of Montserrat and 32 km southeast of Nevis (Figure 1). Redonda is uninhabited and is 1.6 km long, 0.5 km wide and rises to nearly 400 m (Figure 2, Prosper et al. 2008). The island’s planar area is approximately 53 hectares, but its actual surface area is estimated to be at least 80 hectares1 (Figure 2, Morse, 1979). Redonda has a tropical dry climate with low rainfall (<1,000 mm per year). The wet season, which overlaps with the hurricane season, extends from August to November, but Redonda’s small size and lack of trees means that few clouds gather here. A remnant volcanic core, Redonda rises steeply from the sea. There are sheer cliffs around most of the island. There used to be an access route from the shore up a steep gully, but this was destroyed by a series of rock falls in March 2017 and remains inaccessible to date (Figure 2). Access is currently only possible by helicopter to several flat areas on the top of the island (Figure 2). The top of the island slopes eastward, with few flat areas. Large rock scree and boulders cover these sloping areas.

1 Published areas vary widely, with mistakes often repeated by subsequent authors. For example, Prosper et al. (2008) give the area as being up to 260 hectares, while Horwith & Lindsay (1997) contain a table giving the area of Redonda as 500 hectares. However, the island is just 1.6 km long and 0.5 km wide at its widest point, so these higher figures are not credible. The planar area of 53 hectares was measured by Dr. Brian Cooper using GIS from the best available topographic map of the island, but should be verified using a good geo‐rectified satellite image when one becomes available. The ‘three‐dimensional’ surface area is probably nearer 80 hectares owing to its steeply pointed shape, and is the estimate used in this report, but again this should be verified with an accurate GIS model.

Bell, Ibbotson & Daltry 2 Redonda Rat Eradication: Technical Report

Figure 1 Location of Redonda, Antigua & Barbuda.

N

500 m

Bell, Ibbotson & Daltry 3 Redonda Rat Eradication: Technical Report

Figure 2 Redonda, Antigua & Barbuda.

500 m

Bell, Ibbotson & Daltry 4 Redonda Rat Eradication: Technical Report

There is limited vegetation; generally restricted to a few trees growing on steep cliff faces and a flush of green herbs (mostly non‐native Cleome spp.) after rain. There is no permanent source of freshwater, although there are several rainwater seepage points, including one in Centaur’s Cave. There are several ruins and archaeological structures on the island, but no accommodation or other facilities. Redonda is stated‐owned (crown land). The island was mined for guano between 1865 and 1914, during which time up to 120 people lived and worked here (Morse, 1979; Prosper et al. 2008). Buildings and other structures, including a wharf, were constructed during this period. A post office was established, and a caretaker lived on the island in 1978 (Morse, 1979), but most structures are in ruins, many of them damaged by hurricanes. Around 70 species of plants have been recorded on Redonda over the past 70 years, but some of these have disappeared and at least 20 species are not native to the Caribbean, including ornamental plants brought by the miners (Howard, 1962; Pratt & Thomas, 2011). Historical paintings indicate that Redonda used to be forested and, even photographs from the 1960s and 1970s still show many shrubs and agave blanketing its surface (see Morse, 1979). Only a handful of woody plants remain today, and much of the island has a barren appearance. Short‐leafed fig trees Ficus citrifolia can be found clinging to the cliffs, and a single Australian casuarina tree stands near the ruins of the mine manager’s house between ‘the plateau’ and Shiel’s Summit. There is little or no evidence of tree regeneration. Further details on the vegetation of Redonda can be found in Bell & Daltry (2012) and Lindsay & Horwith (1997) with additional information on the full report on the vegetation of Redonda currently being prepared by Lindsay (in prep.). With the loss of vegetation has come the loss of topsoil, and much of the island is seriously eroded and unstable. Much of the habitat on Redonda is sparsely vegetated cliffs, talus/scree slopes and jumbled boulders (Devine et al. 2010). Redonda has been designated an Important Bird Area (no. AG001) because it holds regionally significant populations of seabirds, specifically magnificent frigatebird, masked booby, red‐footed booby and brown booby, as well as smaller numbers of red‐billed tropic bird, brown noddy and bridled tern (Prosper et al. 2008; BirdLife International, 2012). Audubon’s shearwater Puffinus iherminieri was previously listed as present on the island (Sylvester et al., 2009), but no recent records have been reported excluding a possible call heard in 2009 (Lowrie et al., 2009). The island is surprising for its lack of other small nesting seabirds, particularly terns and gulls, which may be attributable to the very high density of large rats. There are very few records of land birds on the island. The burrowing owl Athene cunicularia has been recorded on Redonda (Clarke, 1957; Office of the Prime Minister, 2001), but there have been no recent sightings. Only a handful of zenaida doves Zenaida aurita and peregrines Falco peregrinus were observed in 2012 (Bell & Daltry, 2012)), but it

Bell, Ibbotson & Daltry 5 Redonda Rat Eradication: Technical Report

is possible some additional species pass through on migration or periodically come from Montserrat. Records of birds seen during the eradication are recorded in Appendix 12.2. Redonda has six species of lizard, at least four of which are endemic to the island and all are threatened with extinction (Malhotra & Thorpe, 1999; Daltry, 2007; Prosper et al. 2008; Hedges & Conn, 2012). The Redonda ground lizard and Redonda tree lizard are the most conspicuous. An unnamed dwarf gecko is currently being identified. The Redonda skink Copeoglossum redondae (previously mis‐assigned to the genus Mabuya) is another island endemic but has not been seen for decades. An iguana (possibly Iguana iguana, I. delicatissima or an endemic species), is also feared extinct. The sixth lizard recorded on Redonda is the common house gecko Hemidactylus mabouia, an alien invasive species. No snakes or tortoises have ever been recorded on Redonda. Green turtles Chelonia mydas and hawksbill turtles Eretmochelys imbricata have commonly been seen in the waters around the island (Prosper et al. 2008), but presumably do not nest here because Redonda lacks sandy beaches. Further details on the reptiles of Redonda can be found in Bell & Daltry (2012). Although domestic sheep Ovis aries, dogs Canis lupus familiaris and cats Felis catus were on Redonda during the mining period (Morse, 1979), the only mammals still present on the island during the eradication operation were the non‐native feral goats and black rats. It is possible goats have been on Redonda since the 1600s, left by early explorers (Morse, 1979; Lindsay, 2012), although additional stock may have been introduced over time. There were approximately 65 goats on the island in April 2012 (Bell & Daltry, 2012) with numbers at a similar level when the rat eradication operation began in January 2017. No bats have been recorded on Redonda before 2017, although they may have been present historically. Two sightings of bats (suspected to be Mexican free‐tailed bat Tadarida brasiliensis) occurred during this operation (Appendix XX.X). It is likely that black rats were accidentally introduced to Redonda sometime after the 17th century when rats were first recorded in the Caribbean region (Campbell, 1991; Long, 2003); probably when ships were wrecked along the shores. Rats have been implicated as causing major impacts on island biodiversity (Towns et al., 2006; Jones et al. 2008) and they are known to affect important species on Redonda. Prosper et al. (2008) suggested that brown or Norway rats Rattus norvegicus may have invaded Redonda, and they were listed as present by Lowrie et al. (2012), but we consider this to be a misidentification of the resident black rats, which are brownish and unusually large. Rattus norvegicus are present in urban areas on Montserrat, Antigua and other large islands in this region, but none were observed on Redonda during the feasibility study (Bell & Daltry, 2012) or throughout the eradication operation. No house mice Mus musculus have ever been recorded on Redonda. There have been few studies of the invertebrate fauna of Redonda, but it is clear that their current diversity is abnormally low (Campbell, 1977; Morse, 1979). Further details on the known invertebrates of Redonda can be found in Bell & Daltry (2012). It can be reliably inferred that many native species have been lost due to deforestation and the impacts of alien species.

Bell, Ibbotson & Daltry 6 Redonda Rat Eradication: Technical Report

1.2 BLACK RAT, RATTUS RATTUS Originally from the Indian subcontinent, black rats are found throughout the world. They are relatively large, with a slender body, long scaly tail, large ears and dark hairy feet (King, 1990; Novak, 1999). Black rats can grow up to 230 mm in length and weighing up to 300 g (Cunningham & Moor, 1996). There are three colour phases, termed rattus (black back and dark grey belly), alexandrinus (brown back and pale grey belly) and frugivorus (brown back and white or cream belly). The proportion of colour phases can vary depending on the location, although frugivorus is usually the most common colour phase (Cunningham & Moor, 1996; King, 1990). The frugivorous colouring was the most common phase on Redonda (Bell & Daltry, 2012). Males are larger than females, and when mature, have prominent scrotum at the base of the tail (King, 1990; Novak, 1999). Usually only breeding females have visible nipples (King, 1990; Novak, 1999). Black rats have excellent smell, touch, taste and hearing senses (King, 1990). Black rats are omnivorous (but can also be specialist) feeders, taking advantage of any potential food source and will often cache food (King, 1990; Nowak, 1999). When on the ground, black rats prefer to eat food under cover; but in the trees rats will feed on any available flat surface (King, 1990). Black rats are well‐documented to be major predators of land and seabirds, invertebrates, lizards and native mammals, and are voracious consumers of vegetation, seeds and fruit. They have caused the extinction of many plant species, particularly those on isolated offshore islands (Atkinson, 1985; Bell, 1978; Imber, 1985; King, 1990; Blackburn et al., 2004; Towns et al., 2006; Bell et al., 2016). Although natural food normally forms a high proportion of black rat diet, human products (stores, vegetables and crops) are also targeted (King, 1990). Black rats are very agile and skilful climbers, and live both in trees and on the ground (King, 1990). They are reluctant to enter water, but have been recorded swimming between islands up to 750 m apart (King, 1990). Black rats do not often burrow, preferring to nest in trees, under thick vegetation or in rock tumbles or crevices (King, 1990; Nowak, 1999). Despite this, tracks and runs are common in areas of black rat activity (King, 1990). Black rats are usually associated with forests or vegetated areas, but can live in a range of habitats from barren ground, coasts, grassland to lush forest as well as human dwellings, buildings and farms (King, 1990; Nowak, 1999). In natural habitat, black rats do not live in colonies, preferring to disperse throughout the available area (King, 1990). However, in urban areas, a small number of adult females and one dominant male will live together in a territory that will be aggressively defended against other rats (King, 1990). Home ranges can vary from 0.1 ha to 1 ha in all types of habitats depending on food availability and habitat quality (Moors, 1985; King, 1990; Nowak, 1999). Males have larger home ranges than females, which prefer to stay close to breeding sites, but this may vary depending on habitat quality, food availability, predation pressure and other factors (King, 1990; Nowak, 1999).

Bell, Ibbotson & Daltry 7 Redonda Rat Eradication: Technical Report

Black rats construct nests out of various items, including newspaper and cardboard, but they are usually made from vegetation (twigs and leaves) and feathers, with new material added regularly (King, 1990; Nowak, 1999). They can breed throughout the year, but this generally depends on food availability and habitat (King, 1990; Nowak, 1999). Gestation is between 20 and 22 days and litter size vary from three to ten young (usually five‐six); the average annual production can be up to 40 young per year (King, 1990; Nowak, 1999). The young are weaned when they are between 21 and 28 days old (about 40 g) and can be sexually mature at three months old (King, 1990; Nowak 1999). Black rats usually live between 12 and 18 months in the wild, with females generally living longer than males (Daniel, 1972; King, 1990). Black rats are typically nocturnal and generally shy; however, this depends on habitat, predation pressure, hierarchy, disturbance and food availability (King, 1990). They explore all areas and objects within their home range, but can be cautious regarding new or strange objects within this area (King, 1990). Black rats are commonly infested with fleas and mites as well as being known carriers of several diseases, including leptospirosis and salmonellosis (King, 1990). 2 GOAL AND OBJECTIVE OF REDONDA RAT ERADICATION The goal of the operation to eradicate invasive alien rats in order to establish Redonda as a haven for native biodiversity and restore many of its natural ecological functions. The objective was to eradicate all black rats from Redonda while minimising any adverse impacts on the environment, non‐target species (including goats), archaeology and humans, for the conservation and restoration of the island’s nationally and globally important biodiversity. 3 METHODS OF THE REDONDA RAT ERADICATION OPERATION The Redonda rat eradication phase was completed using a combination of ground‐based bait and aerial methods using rodenticide blocks, distributed at bait stations at 30‐metre intervals and broadcast by hand or from the air (helicopter). There were three phases to the Redonda rat eradication operation: establishing the bait station grid and climbing routes, eradication (ground, climbing and aerial application) and intensive monitoring. The methods for each are given below. Elizabeth Bell (WMIL) and Jack Ibbotson (Consultant Rat Eradication Assistant and Mountaineer) arrived on Antigua on 29 January 2017 and Adam Long (BMC, Mountaineer) arrived on 3 February 2017. Jack and Adam visited Redonda between 4 and 8 February 2017 to establish climbing routes and check climbing bolt placement. Most of the volunteer team arrived on 6 February 2017. Detailed briefings about the project and eradication operation, first aid training and purchase and packing of equipment and supplies was completed between 8 and 12 February 2017. The team was on Redonda from 13 February to 7 April 2017. Day to day details of the operation are given in Appendix 12.1

Bell, Ibbotson & Daltry 8 Redonda Rat Eradication: Technical Report

3.1 ESTABLISHMENT OF THE GROUND BAIT STATION GRID The establishment of the bait station grid occurred between 14 and 16 February 2017. 3.1.1 BAIT STATION DESIGN Bait stations must allow ready access for rats to the bait, but must also prevent entry by key non‐target species (such as goats and land birds). Bait stations were made from clear plastic 1.5L water bottles (c. 250 mm lengths and 100 mm diameter) with wire “legs” (or rocks) to peg them to the ground to prevent movement by animals and/or wind (Figure 3). The project team prepared the bottles (i.e. removing top and bottom of each bottle) on Redonda. Bait was held in the centre of the station by a wire that passed through the station and twisted over each end (Figure 3). This ‘bottle’ bait station design is well proven in many eradication programmes in the Caribbean region (Varnham & Daltry, 2006; Witmer et al., 2007; Bell, 2012). Figure 3 Bait stations used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

The stations were placed off the ground in some locations to prevent invertebrates and hermit crabs from accessing the tubes. Access to the bait station to replace and monitor bait was via the ends of the bottle. 3.1.2 GRID DENSITY Bait stations were placed out on a 30m x 30m grid over the easily accessible areas on the top of Redonda (Figure 4). Each bait station was put into position before being individually numbered, mapped using GPS and was added into a GIS‐linked database (Figure 4). Flagging tape and spray paint was used to locate stations.

Bell, Ibbotson & Daltry 9 Redonda Rat Eradication: Technical Report

Figure 4 Bait station grid used on the top, beaches and cliff‐access sections during the black rat Rattus rattus eradication on Redonda, January to April 2017. Note: Bait stations were placed on cliffs, ruins, caves and sink holes.

Bell, Ibbotson & Daltry 10 Redonda Rat Eradication: Technical Report

There were 327 stations placed over Redonda; on the top and around the easily accessed coastline (Figure 4). In addition to the ‘bottle’ bait stations, 39 commercial Protecta™ lockable plastic bait stations were used in permanent locations around Redonda as part of the long‐term monitoring and biosecurity (see Sections 6 and 8). These stations were installed in April 2017. 3.2 ESTABLISHMENT OF THE CLIFF AND ROPE‐ACCESS BAIT POINT GRID In addition to the bait stations, areas of the island were baited by the mountaineering team using ropes and scramble routes (Figure 5 to 7). The rope‐access team also maintained bait points along harder‐to‐access areas along the beach, particularly past one point on the eastern shore requiring abseil‐access to reach the beach (Figure 4). The rope access team were also responsible for accessing the bait points on the rock stacks (Spilt Island, Figure 4). There were 95 bait points and eight fishing line drop points placed over Redonda; on the cliffs and around the rope‐accessed coastline (Figures 4‐7). These eight fishing line drop points were found not to be particularly useful and were removed early due to the risks of bird entanglement. In addition to placing 10 blocks of bait at each point or inside each station on the cliffs, 10 blocks of bait were hand‐broadcast from those sites in either direction (a total of 20 blocks at each site). Additional blocks were broadcast using catapults to spread the bait into inaccessible ledges and areas from the climbing points. Figure 5 Scramble Line 2 (X1‐20) and ‘Big Abseil’ cliff‐access stations used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 11 Redonda Rat Eradication: Technical Report

Figure 6 Scramble Line 3 (X22‐25, X39, X40, X53‐58) cliff‐access stations used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Figure 7 Rope‐access Line 4 (X526‐38) cliff‐access stations used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Several stations on the isolated beaches (i.e. north‐western end) required team members to swim to these spots in order to carry out checks. These checks were dependent on the state of the sea and as such, some of these beach stations were checked less regularly that the climbing or ground area stations.

Bell, Ibbotson & Daltry 12 Redonda Rat Eradication: Technical Report

3.3 AERIAL APPLICATION AREA The remainder of the island was baited using a helicopter (Figure 8). One team member was harnessed into the helicopter and broadcast bait by hand while standing on the helicopter skid. The ‘thrower’ was in close communication with the pilot to reach target locations across the island. Another team member was also in the helicopter to assist the ‘thrower’ by opening bait buckets, holding bait buckets in position and throwing bait on occasion. Figure 8 Aerial application area (shaded red) used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 13 Redonda Rat Eradication: Technical Report

3.4 POISONING PHASE OF THE REDONDA RAT ERADICATION The majority of the Klerat® bait was transported to Redonda on 13 February 2017, with additional small amounts to be used for the aerial application bought in weekly by helicopter. The Final® bait was transported to Redonda on in two batches; 9 and 13 March 2017. The poisoning phase of the eradication project started on 15 February 2017 and continued through to 2 April 2017. Klerat® was used between 15 February and 14 March 2017 and 18 March to 2 April 2017. Final® was used between 15‐17 March 2017.

3.4.1 POISON TYPE USED FOR THE DOG ISLAND RAT ERADICATION The poison used was during the eradication on Redonda was brodifacoum in two formulations; wax‐based Klerat® (20 g, Syngenta Crop Protection AG) and grain‐based wax blocks, Final® (24 g, Bell Laboratories Ltd.) (Figure 9). Figure 9 Bait used during the black rat Rattus rattus eradication on Redonda, January to April 2017. Where A = Klerat® by Syngenta (Photo: J. Daltry, FFI) and B= Final® by Bell Laboratories Ltd. (Photo: E. Bell, WMIL).

A B

Brodifacoum has been used successfully in over 70% of eradications completed worldwide and on most of the eradications within the Caribbean region (Howald et al., 2007; Varnham, 2010; Bell, 2012; Havery et al., 2014). Klerat® was used as the main bait for most of the project and Final® was used at the end to target any surviving rats (Appendix 12.1). The Klerat® is dyed blue (or green/blue) and the Final® was dyed red to make the baits less attractive to birds. Brodifacoum is a second‐generation anticoagulant poison that acts by reducing the animal’s ability to coagulate blood (i.e. inhibits the synthesis of Vitamin K and as a result rats die of internal haemorrhaging, Eason & Wickstrom 2001). Death usually occurs between three and ten days after consumption of a lethal dose (LD50). For a 250g black rat, the LD50 for brodifacoum is less than 5 g of bait which can be obtained in a single feed. As a result, bait shyness is avoided. Both bait formulations contained a bittering agent (Bitrex™), which is usually added to prevent the consumption by humans (particularly children). Recently there has been

Bell, Ibbotson & Daltry 14 Redonda Rat Eradication: Technical Report

research that suggests that rats may be able to detect Bitrex™ (and other such bittering agents; Veitch, 2002). During the Lundy Island eradication programme two locations had rats that would not consume bait containing Bitrex™, but did eat non‐Bitrex™ bait that was specifically made by Sorex Ltd. for the project (Bell, 2004). Towards the end of the poisoning phase (between 18 March and 2 April 2017) all waste bait from the bait stations was placed into rock tumbles to target any rats that may have been avoiding bait stations or areas that had previously high rat activity. Both types of bait were used whenever possible in these locations. 3.4.2 POISONING PHASE OF THE REDONDA RAT ERADICATION The poisoning phase of the Redonda rat eradication began on 14 February 2017 for the climbing routes and on 17 February 2017 for the ground and helicopter areas. Baiting continued until 2 April 2017. Bait was present in the stations throughout the poisoning programme and was replaced as required each day (Tables 1 and 2). Five blocks of Klerat® were placed in each ground station until 28 February 2017 when it was dropped to 2 blocks for the remainder of the poisoning phase (2 April 2017). The Klerat® bait was replaced with 2 blocks of Final® between 15 and 17 March 2017. Blocks were replaced when eaten by rats (or non‐target species) and/or damaged by weather. Throughout the poisoning phase, every ground bait station was checked either every day or every two days; there were 35 complete checks of the ground bait stations (Table 1). However, areas with difficult access such as the cliff sections, swimming sections or beach‐access areas were checked whenever possible. Climbing stations had 10 blocks at each bait point and 20 blocks hand‐broadcast from each bait point (Table 2). Bait was placed at these points between 14 February and 30 March 2017. There were eight complete checks of the cliff‐access points and beach stations (Table 1). There were ten aerial applications of bait from the helicopter between 18 February and 27 March 2017; the aerial application rate was 1.5 kg per hectare and the amount of bait used varied between drops ranging from 70‐200 kg of bait (Tables 1 and 2). The majority of these areas could not be assessed for bait take due to inaccessibility, but it is assumed that rats behaved similarly to those on the cliff and ground sections. Bait stations were lowed into position using fishing line in eight locations. However, these were found to be of limited value (i.e. no bait take recorded on any line) and were removed early due the risk of bird entanglement. Bait take was recorded in notebooks and added to a data base on the computer at base camp. Any rat carcasses were collected, dissected whenever possible, and removed in waste from the island. Active bait stations were recorded and maps (using the database) with these locations were produced. Rat bait take maps were produced in real‐time to enable the team to monitor rat activity and target any “hot spots”.

Bell, Ibbotson & Daltry 15 Redonda Rat Eradication: Technical Report

Table 1 Date range for each bait check for the ground, cliff and aerial application during the black rat (Rattus rattus) eradication on Redonda, January‐April 2017. Bait check Ground bait stations Cliff bait points Aerial application 0 17/2/17 to 18/2/17 14/2/17 to 18/2/17 1 19/2/17 20/2/17 to 26/2/17 17/2/17 2 20/2/17 27/2/17 to 1/3/17 22/2/17 3 21/2/17 2/3/17 to 5/3/17 24/2/17 4 22/2/17 6/3/17 to 8/3/17 27/2/17 5 23/2/17 9/3/17 to 12/3/17 3/3/17 6 24/2/17 14/3/17 to 19/3/17 6/3/17 7 25/2/17 21/3/17 to 23/3/17 13/3/17 8 26/2/17 25/3/17 to 30/3/17 20/3/17 9 27/2/17 24/3/17 10 28/2/17 27/3/17 11 1/3/17 12 2/3/17 13 3/3/17 14 4/3/17 15 5/3/17 16 6/3/17 17 7/3/17 18 8/3/17 19 9/3/17 20 10/3/17 21 11/3/17 22 12/3/17 23 14/3/17 23 15/3/17 25 16/3/17 26 17/3/17 27 18/3/17 28 19/3/17 29 20/3/17 30 21/3/17 to 22/3/17 31 23/3/17 32 25/3/17 33 26/3/17 34 28/3/17 35 29/3/17 to 2/4/17

Bell, Ibbotson & Daltry 16 Redonda Rat Eradication: Technical Report

Table 2 Amount of bait during each bait check for the ground, cliff and aerial application during the black rat (Rattus rattus) eradication on Redonda, January‐April 2017. Bait check Ground bait stations Cliff bait points Aerial application 0 1 70 kg (Klerat®) 2 160 kg (Klerat®) 3 100 kg (Klerat®) 10 blocks/point 4 110 kg (Klerat®) + 5 5 blocks/station 200 kg (Klerat®) 20 blocks hand‐broadcast 6 (Klerat®) 200 kg (Klerat®) 7 200 kg (Klerat®) 200 kg 8 (180 kg Klerat®; 20 kg Final®) 200 kg 9 (180 kg Klerat®; 20 kg Final®) 100 kg 10 (80 kg Klerat®; 20 kg Final®) 11 12 13 14 15 2 blocks/station 16 (Klerat®) 17 18 19 20 21 22 23 23 2 blocks/station 25 (Final®) 26 27 28 29 30 2 blocks/station 31 (Klerat®) 32 33 34 35

Bell, Ibbotson & Daltry 17 Redonda Rat Eradication: Technical Report

3.5 TRAPPING Kill traps (T‐Rex™) were used in the early stage of the project between 13 and 23 February 2017 around the camp and helicopter pad. Traps were set in pairs in natural tunnels, covered by rocks. Traps were baited with peanut butter, set only during the ours of darkness (to avoid non‐target species, particularly lizard captures) and checked daily. Where possible, rats were necropsied, and specimens collected for museums (Appendix 12.3). 3.6 INTENSIVE MONITORING PHASE OF THE REDONDA RAT ERADICATION Monitoring is one of the most important aspects of any eradication programme and on Redonda this was carried out from 1 March 2017 to detect whether any rats had survived the poisoning phase. The intensive monitoring phase continued to 2 April 2017. A 15m x 30m grid of rat‐attractive food items (i.e. flavoured wax, soap, etc.), trail cameras and tracking tunnels was established over the whole island, including offshore rock stacks. There were more monitoring stations than bait stations to ensure any activity would be identified as soon as possible. This included a smaller monitoring grid in any area of known rat activity. There were over 650 monitoring stations; one at the bait station and one in‐between two adjacent stations. All monitoring stations were marked with flagging tape and/or spray point to ensure visibility. These points were mapped and entered into the GIS‐linked database. The flavoured wax was prepared by the project team in Antigua prior to transport to Redonda. It was produced by melting down wax pellets and adding a measure of cocoa powder, peanut butter or flavouring essence (e.g. peanut, aniseed, caramel). The molten flavoured wax was then poured into ice‐cube trays or silicon muffin trays to harden and produced small wax blocks. The flavoured wax was wired to rocks or the bait stations. Each monitoring site was checked regularly (either separately or together with the poisoning bait station grid). If rat sign (usually teeth marks) was detected, an intensive poisoning programme (bait stations closed to a 20m grid around the site, traps put around the site or bait was placed under rocks) was established. Rat (or non‐target species) sign was recorded in notebooks and added to the GIS‐linked database back at camp. Active monitoring stations were recorded on real‐time maps to enable the team to monitor these “hot spot” areas. 3.7 ESTABLISHMENT OF THE PERMANENT BAIT STATION (LONG‐TERM MONITORING) SYSTEM ON REDONDA After the completion of the poisoning and intensive monitoring phase, 39 permanent bait stations were established on Redonda. These stations are a single line of lockable bait stations (Figure 10) that are 50 metres across the island and southern end.

Bell, Ibbotson & Daltry 18 Redonda Rat Eradication: Technical Report

Figure 10 Permanent bait station used during the black rat Rattus rattus eradication on Redonda, January to April 2017.

These biosecurity stations were baited with four blocks of Klerat® and a piece of flavoured wax or resin and will be checked every eight weeks by the Redonda Restoration Programme, EAG or Department of Environment staff. Biosecurity monitoring began from April 2017. 4 RESULTS 4.1 BAIT ACCEPTANCE AND BAIT TAKE BY RATS Bait acceptance was good over most of the island. The appearance of blue rat droppings (discoloured by the dye in the bait) within three days showed how quickly the rats became aware of the bait and were attracted to it. During the poisoning phase, rats accounted for 127 kg of bait consumed and visited almost every bait station or bait point (Table 3, Figures 11 and 12). Bait take cannot be estimated for the helicopter or hand‐broadcast areas, but is likely to match or be similar to the amount taken on the top of the island. Interestingly bait take was very high in the two checks and quickly dropped to a relatively low level by the 6th check (Figure 11). Bait take levelled off between the 10th and 12th checks before dropping to nothing by the 19th check on the 9 March 2017 (Table 3, Figure 8) and this pattern may relate to the dominant rats, which have consumed a lethal dose during the first part of the programme began to die, allowing sub‐dominant rats access to the bait. Throughout the poisoning phase, 93.8% of bait stations or bait points (n = 396) were visited by rats at some stage, with 76.5% active (n = 323) within the first four checks (Figure 12). The high number of active bait stations during the first four bait rounds shows that the rats quickly accepted the bait over most of the island.

Bell, Ibbotson & Daltry 19 Redonda Rat Eradication: Technical Report

Figure 11 Bait (kg) consumed by rats during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Table 3 Bait take (in kg) by rats from the ground, beach and cliff access bait stations during the black rat Rattus rattus eradication on Redonda, January to April 2017. Bait round Bait consumed (kg) Number of bait blocks consumed 1 28.999 1449.95 2 29.6399 1481.995 3 24.269 1213.45 4 20.379 1018.95 5 13.0492 652.46 6 6.439 321.95 7 2.2292 111.46 8 1.0314 51.57 9 0.649 32.45 10 0.077 3.85 11 0.0978 4.89 12 0.194 9.7 13 0 0 14 0.008 0.4 15 0 0 16 0 0 17 0 0 18 0.002 0.1 19 0 0 20‐35 0 0 Total 127.06 6353.175

Bell, Ibbotson & Daltry 20 Redonda Rat Eradication: Technical Report

Figure 12 Percentage of active bait stations and cumulative percentage of active bait stations by date as visited by rats during the black rat Rattus rattus eradication on Redonda, January to April 2017.

There were 396 bait stations that had rat activity during the poisoning phase and the remaining 26 stations were either not active or only had non‐target species bait take (mainly crab or ant activity). After the first nine checks, the percentage of active bait stations had dropped to less than 20% (Figure 12). This level of activity suggests that the rat numbers on Redonda quickly reduced after the immediate take of bait during the first few days of the eradication. There was no evidence of rats having taken the Final® bait (only used during bait check 24 and 25) in any bait station which suggested that the Klerat® bait successfully targeted all the rats on Redonda. The hand‐broadcasting Klerat® bait on the cliffs, selected sites on the top and all helicopter areas meant rats that lived in inaccessible areas of the island or would not enter bait stations were targeted successfully; the last rat sign was on 8 March 2017 (recorded on a trail camera eating bait in a station). Of these rat‐active bait stations, 14.1% were visited only once, 8.3% visited twice, 4.8% were visited three times, 4.0% were visited four times and the remaining 68.7% were visited five or more times (Figure 13). There were 220 stations that were visited more than five times (ten of which were active ten or more times each, Figure 13). The percentage of active bait stations that were visited multiple times by rats during the eradication programme is shown in Figure 13. The majority of bait stations were visited six times (19.7%), but ten bait stations were active over 10 times (Figure 10). Given that most

Bell, Ibbotson & Daltry 21 Redonda Rat Eradication: Technical Report

stations were checked daily, this suggests that the majority of the rats were targeted within 10 days. Figure 13 Percentage of times that active bait stations were visited by rats during the black rat Rattus rattus eradication on Redonda, January to April 2017.

The average number of blocks taken by rats was 15.0 (± 0.5) blocks per station (range 0 to 36 blocks) (Figures 14). This also indicates that rats were quickly removed from most sites across the island. Figure 14 Number of bait blocks taken by rats per bait station during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 22 Redonda Rat Eradication: Technical Report

Although bait take in the helicopter area cannot be estimated, as shown by Figure 15, bait take recorded in the top, beach and cliff access bait stations was not evenly distributed with the greatest level of bait take at the southern end and around the Manager’s house. Figure 15 Distribution of bait take by rats in the top, beach and cliff‐access bait stations during the black rat Rattus rattus eradication on Redonda, January to April 2017. Where rat bait take was calculated by weight (g) of bait taken by rats and warmer colours represent more blocks of bait taken. Note: Bait take is shown from the bait points (cliff‐access) and at bait stations only. Blank areas on the map were areas where bait take could not be monitored and assessed due to inaccessible terrain (these areas had bait deployed by hand‐ broadcasting, by catapult and/or from the helicopter).

Bell, Ibbotson & Daltry 23 Redonda Rat Eradication: Technical Report

An indication of rat numbers on the top 24 hectares of the island can be calculated by the total amount of bait taken (127 kg or 6354 blocks, Table 1); this relates to between 181 and 6354 rats (assuming each rat ate between four and sixty times its lethal dose (i.e. 1 block or up to 35 blocks, Figure 14). Given that camera footage showed multiple rats targeting the bait in each bait station (Figure 19), it is suspected the population of black rats on the top of Redonda was at least 3200 rats (with the majority of the 5,500 rats (70 rats/ha) as calculated during the feasibility study being found on the top of the island, Bell & Daltry, 2014). Assuming the level of bait take was similar in the aerial‐applied areas, it is likely that there were approximately 5000 to 7500 rats on Redonda. Figure 16 Footage of multiple rats targeting bait stations during the black rat Rattus rattus eradication on Redonda, January to April 2017.

The majority of rats targeted on Redonda died underground and under rocks and could not be recovered or counted. Any rat carcasses on the surface was collected and removed as waste for disposal on Antigua. There were 117 rats found on the surface of Redonda between 14 February and 6 April 2017. This was an unusually high number, but was likely to be related to the lack of burrows due to the rocky substrate of the island. 4.2 MONITORING FOR RATS The intensive monitoring phase ran for four weeks; between 1 March and 2 April 2017 (Appendix 12.1). There were 24 complete checks, with additional checks in areas that had suspicious sign reported by project members. Actual dates for the 24 complete checks and other activity during the intensive monitoring phase are given in Appendix 12.1.

Bell, Ibbotson & Daltry 24 Redonda Rat Eradication: Technical Report

Rat sign was recorded on 60 monitoring stations on Redonda on either 2 and 6 March 2017 (Figure 17) and a juvenile rat was recorded eating bait at station D6 (above the Manager’s house) on a trail camera on 8 March 2017 (Figure 18). Figure 17 Distribution of monitoring points visited by rats during the black rat Rattus rattus eradication on Redonda, January to April 2017. Where • = 2 March, • = 3 March, • = 4 March, • = 5 March and • = 6 March 2017. Note: These monitoring results are from the top, beach and cliff‐access monitoring points only. Blank areas on the map were areas where monitoring could not occur due to inaccessible terrain.

Figure 18 Juvenile rat detected on a night‐vision camera at bait station D6 on 8 March 2017 during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 25 Redonda Rat Eradication: Technical Report

All of these sites were located on the top of the island and a number were clustered around the Manager’s house. Each site was targeted more often with fresh bait and extra monitoring stations and the camera site had snap traps also placed into position. All sites were no longer active by the end of the project (7 April 2017). There was no rat activity detected on monitoring points on the offshore rock stack.

4.2.1 LONG‐TERM MONITORING FOLLOWING RAT ERADICATION ON REDONDA Long‐term monitoring began in April 2017 and will continue for two years (April 2019). This monitoring uses the permanent bait stations established across the island and additional chocolate wax monitoring points placed out randomly across the island. Monitoring to date on this long‐term monitoring phase has also shown no rat sign (S. Challenger, RRP, pers. comm.). Following international convention, either after two breeding seasons (c. 1 year in tropical areas such as Antigua & Barbuda) or after the two years of monitoring has been completed, the island can be declared rat‐free if there has been no sign of rats up to that point. Following this, these permanent bait stations will be used as part of the biosecurity protocols for Redonda and monitoring to detect rat incursions (see Section 8). 4.3 INTERFERENCE BY NON‐TARGET SPECIES Any eradication programme has an associated risk that non‐target species will be accidentally poisoned. This can be through direct consumption of the bait or secondary poisoning by eating poisoned animals. To reduce this risk, the bait stations were designed to limit access to as many non‐target species as possible. In addition to this, bait was also wired into the stations. To reduce the risk of secondary poisoning by eating poisoned rats, any carcasses that were found were collected and were incinerated; 117 rat carcasses were found on the surface. 4.3.1 INTERFERENCE ON BAIT AND BAIT STATIONS BY NON‐TARGET SPECIES There were high levels of interference by land crabs on Redonda during the baiting phase (Figures 19 & 20). Although it took a couple of days for the crabs to notice the bait, they filled the bait stations to target the bait for the remainder of the project (Figure 20). Crabs ate 68.19 kg of bait (3409.44 blocks) throughout the eradication on Redonda. There was no evidence that any crabs were affected by the bait. Nearly all the bait stations were interfered with by crabs over the entire eradication period; only 10 stations were not affected. Figure 20 shows the percentage of bait stations that were visited by crabs during each bait round and the cumulative percentage of activity over the whole eradication. It is apparent that the land crabs took a few days to get onto the bait, but once they had learnt the bait was present in the stations activity levels remained high (i.e. from bait round 9 onwards over 50% of the stations were visited by crabs) (Figure 20).

Bell, Ibbotson & Daltry 26 Redonda Rat Eradication: Technical Report

Figure 19 Hermit crabs in a bait station during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Not all bait stations were active at the same time; although crabs visited only a maximum of 43.1% of all bait stations at any one bait check, crab activity increased from 1st check (5.7%) to the 30th check (43.1%) (Figure 20) showing that crabs learnt to target bait stations as the operation continued. A number of bait stations had to be moved by team members to prevent bait being eaten by crabs instead of rats. Figure 20 Percentage of active bait stations and cumulative percentage of active bait stations by bait check as visited by crabs during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 27 Redonda Rat Eradication: Technical Report

There were 412 (97.6%) bait stations on Redonda that had crab activity during the poisoning phase at some stage (Figure 20). Figure 21 shows that the crab activity at the bait stations was evenly distributed over the top of the island. Crabs took the most bait near the Manager’s house and in three other locations each centred aroudnt he main track across the island (Figure 21). Figure 21 Distribution of bait take by crabs during the black rat Rattus rattus eradication on Redonda, January to April 2017. Where bait take was calculated by weight (g) of bait taken by crabs and warmer colours represent more bait taken. Note: Bait take is shown from the bait points (cliff‐access) and at bait stations (top and beach) only. Blank areas on the map were areas where bait take could not be monitored and assessed due to inaccessible terrain (these areas had bait deployed by hand‐broadcasting, by catapult and/or from the helicopter).

Bell, Ibbotson & Daltry 28 Redonda Rat Eradication: Technical Report

The average number of blocks taken by land crabs was 8.1 (± 0.5) blocks per station (range 0 to 40.6 blocks) (Figure 22). Crabs ate over 68 kg (over 3409 blocks) of bait throughout the eradication on Redonda with the majority eating less than 5 blocks per station (Figure 22). Given the number of crabs recorded in each station, it is likely that the amount eaten by each individual crab was much less than 5 blocks per station over most of the island, with larger quantities eaten near the Manager’s house and other locations (as shown by Figure 21). There was no evidence that any crabs were affected by the bait (E.A. Bell, pers. obs.). Figure 22 Number of bait blocks taken by crabs per bait station during the black rat Rattus rattus eradication on Redonda, January to April 2017.

There were low levels of interference by other non‐target species on bait and bait stations during the rat eradication on Redonda; 3.3 kg (165 blocks) of bait for insects (particularly ants) and another 1.6 kg (80 blocks) attributed to weather (Figure 23). The stations that were affected by weather were those along the coast which were lost during a storm event. Only one station was affected by goats with 1.75 blocks of bait taken (35 g); this was not enough to pose any risk to the animal. Goats did knock over a number of marking sticks and chewed flagging tape throughout the programme. Like other eradications in the Caribbean (Day & Daltry 1996, Daltry 2000, Daltry et al. 2001, Varnham 2003, Varnham & Daltry 2006, Varnham 2010, Bell 2012), ground lizards did not appear to eat the Klerat® bait and did not interfere with the bait at any point during the eradication. Although the ground lizards showed an interest when presented

Bell, Ibbotson & Daltry 29 Redonda Rat Eradication: Technical Report

with the crushed Final™ bait, they did not appear to consume it. The tree lizards and dwarf geckoes also showed no interest in either bait types. Figure 23 Non‐target species (excluding crabs) bait take (in kg) and weather impacts on bait (in kg) during the black rat Rattus rattus eradication on Redonda, January to April 2017.

However. in some circumstances it could be difficult to determine which species had actually taken the bait and as such, these figures are estimates only. Excluding crabs, this interference accounted for a total of 768 stations being affected over the 8‐week programme. One peregrine carcass was found on 27 February 2017 showing signs of secondary poisoning (i.e. bruising in the body cavity, blood in the body cavity and pale liver). It is suspected that this bird targeted dying rats on the surface; this is unusual behaviour for a peregrine as they hunt other birds on the wing. This may relate to the scarcity of appropriate prey on Redonda, despite the presence of high numbers of boobies which could be targeted by peregrines. Unfortunately, this is the first confirmed case of mortality to a non‐target species from any eradication operation in Antigua. In future operations where peregrines are present on the target island and the availability of regular prey species is limited, it could be possible to adapt an eradication operation to reduce the risk to this species by scaring the birds away or choosing a time of the year when the birds are not present on the island. A further two dead peregrines were found during the Redonda eradication; one on 25 February 2017 and the other on 10 March 2017. Both carcasses were too decomposed to assess for rodenticide poisoning and one of

Bell, Ibbotson & Daltry 30 Redonda Rat Eradication: Technical Report

those two carcasses showed injuries attributed to striking rocks or fighting with another raptor. It should be noted that at least five peregrines were recorded during the final week of the operation on Redonda and this suggests that the species has quickly returned to Redonda. 4.3.2 INTERFERENCE ON MONITORING POINTS BY NON‐TARGET SPECIES The monitoring also showed that there were high densities of crabs over Redonda; 2370 monitoring points were visitedby crabs over the four‐week monitoring period. Again, it took longer for the land crabs to become aware of the monitoring points, but once recognised as a food source, interference levels increased. Figure 24 and Table 4 shows that the crabs interfered with almost all the monitoring stations at some point of the eradication operation on Redonda. Figure 24 Distribution of interference by crabs on monitoring items during the black rat Rattus rattus eradication on Redonda, January to April 2017. Note: Results is shown from the monitoring points on the beach, top and cliff‐access areas only and warmer colours represent more monitoring items affected. Blank areas on the map were areas where monitoring could not occur due to inaccessible terrain.

Bell, Ibbotson & Daltry 31 Redonda Rat Eradication: Technical Report

Although crabs interfered with monitoring stations on the beaches and cliffs, most interference was recorded on the top of the island, with a similar pattern as shown by the bait take by crabs (Figures 21 and 24). crabs interfered with almost all the monitoring stations at some point of the eradication operation on Redonda. Table 4 Total number and type of monitoring station affected by non‐target species and the weather during the black rat Rattus rattus eradication on Redonda, January to April 2017. Species Flavoured wax Soap Tracking tunnel Total Ant 195 2 10 207 Crab 1760 599 11 2370 Goat 2 ‐ ‐ 2 Lizard ‐ ‐ 14 14 Weather 1063 137 2 1202 Total 3020 738 37 3795 Monitoring points were affected by crabs, ants, goats, lizards and the weather (Table 4 and Figure 25). There were 3795 stations affected over the four‐week monitoring period. Flavoured wax was interfered with the most, but this relates to the high coverage density by this monitoring tool. Goats only interfered with two chocolate‐flavoured wax monitoring stations (Figure 25). Tracking tunnels were the only monitoring tool affected by lizards (Figures 25 and 26). Figure 25 Non‐target species and weather impacts on monitoring stations during the black rat Rattus rattus eradication on Redonda, January to April 2017.

Bell, Ibbotson & Daltry 32 Redonda Rat Eradication: Technical Report

Figure 26 Distribution of interference by lizards on tracking tunnels during the black rat Rattus rattus eradication on Redonda, January to April 2017. Note: Results is shown from the monitoring points on the beach, top and cliff‐access areas only and warmer colours represent more monitoring items affected. Blank areas on the map were areas where monitoring could not occur due to inaccessible terrain.

There was a total of 3795 stations affected over the four‐week monitoring period (Table 4). It is important to note that in some circumstances it could be difficult to determine which species had actually taken the monitoring point and in many cases the monitoring point was simply missing (likely to have been taken by land or hermit crabs) and as such, these figures are estimates only.

Bell, Ibbotson & Daltry 33 Redonda Rat Eradication: Technical Report

4.4 OTHER OBSERVATIONS DURING THE RAT ERADICATION ON REDONDA In addition to monitoring rat and non‐target species bait and monitoring take; observations on other species present on Redonda were undertaken. Information on the lizard and bird observations are listed separately below and further details on bird observations can be found in Appendix 12.3. 4.4.1 LIZARD OBSERVATIONS DURING THE RAT ERADICATION ON REDONDA Ground lizards and tree lizards were common across the island; both species were seen every day. Dwarf geckos were seen around the Manager’s house and on cliff sites, particularly near or under trees on a number of occasions by various team members. Harvard University scientists also spent a week on Redonda monitoring the pre‐ eradication population and behaviour of lizards. This baseline data will be compared with other data collected on future visits to determine changes following the eradication of rats. This research is part of a wider project studying the changes to lizard densities and behaviour on islands throughout the Caribbean that have had invasive species (mainly rats) eradicated. Further details on this research can be obtained from Dr. Colin Donihue ([email protected]). 4.4.2 BIRD OBSERVATIONS DURING THE RAT ERADICATION ON REDONDA Bird species were recorded daily by the Redonda team and birds of interest or significance were reported to EAG. A complete bird list of birds seen during the rat eradication on Redonda and can be found in Appendix 12.3. A highlight of the project was hearing Audubon’s shearwater (Puffinus lherminieri) calls above the Manager’s house on 29 and 30 March 2017 (Appendix 12.3). Unfortunately, despite a number of searches in the area, the birds could not be located on the ground. Seabirds (booby species, frigatebirds and tropicbirds) have been monitored by EAG, RSPB, FFI and visiting scientists over a number of years. The eradication team conducted surveys to determine population estimates of the masked, brown and red‐footed boobies, frigatebirds and tropicbirds (Appendix 12.3). Land bird sightings were recorded through the eradication and included pearly‐eyed thrashers, a bananaquit, Zenaida doves, an American kestrel, up to five peregrines and at least two swallows (Appendix 12.3). Zenaida dove numbers increased towards the end of the operation and it is likely that these and other birds will colonise from Antigua, Montserrat and St Kitts. 5 PROBLEMS DURING THE RAT ERADICATION ON REDONDA There were few problems during the rat eradication operation on Redonda. The main issue was the terrain and the requirement for extreme care on all areas of the island, particularly the Access Gully and climb or scramble routes. Access was by helicopter which removed all difficulties for carrying gear onto the island.

Bell, Ibbotson & Daltry 34 Redonda Rat Eradication: Technical Report

5.1 ACCESS GULLY A series of landslides in the Access Gully on 16 March 2017 meant the access to the beach became extremely difficult. This route became too unstable to use safely. The temporary fixed ropes that were put in place to help ascents and descents were broken or lost in these landslides. The climb team have tried to make the gully more stable by removing a number of loose rocks, but unfortunately this was unsuccessful, and the gully remained risky. There were continuous rock falls for a number of days as the area settled after the big landslide. Luckily no one was in the gully at the time. Alternative routes requiring rope access were used by the climb team to check the bait and monitoring stations along the coast during that last two weeks of the operation. 5.2 NON‐TARGET SPECIES With the exception of land and hermit crabs, most non‐target species did not cause any problems during the eradication. Crabs constantly took bait and disturbed bait stations throughout the operation and accounted for nearly 70 kg of the bait taken. Despite this, the baiting regime, occasional moving of bait stations and additional bait dispersed under rocks meant there was enough bait available for rats throughout the eradication. No dead crabs were found on Redonda and research shows that although residues can be found in land crabs, they are not affected by brodifacoum (Pain et al. 2000). Cases of lizards being killed during rat eradication operations have been reported from around the world (Merton 1987, Thorsen et al 2000, Eason & Wickstrom 2001, Merton et al. 2002, Bell 2002, Fisher & Fairweather 2005, Wedding 2007) and although lizards are known to be susceptible to brodifacoum poisoning, the lethal dose information is unknown. There were no observations of any lizard species on Redonda consuming either Klerat® or Final™ bait during the eradication. It should be noted that most ground lizards avoided both types of bait, although were seen eating insects that were present in and around bait stations. It is expected that all the lizard populations will increase following the black rat eradication as shown by other programmes in the Caribbean and around the world (Towns et al. 2006, Daltry, et al., 2012, Lawrence, et al., 2013). Although 117 dead rats were found on the surface, only other rats, ground lizards and crabs were seen to be scavenging these carcasses. Three peregrines were found dead and one death could be attributed to scavenging dying rats. This is the first recorded non‐ target mortality event during any rat eradication in Antigua. The risk to raptors was identified in the feasibility study and operational plan and estimated as low due to the usual behaviour of peregrines (i.e. migrant visitors to Redonda and aerial hunters of birds), and it is unlikely that any change to the operation or planning (excluding using an alternative bait which would increase the risk of failure to the overall operation or moving the operation to later in the year when the migratory peregrines would be absent) could have prevented this event.

Bell, Ibbotson & Daltry 35 Redonda Rat Eradication: Technical Report

6 VERIFYING THE RAT ERADICATION OPERATION ON REDONDA WAS SUCCESSFUL Long‐term monitoring for surviving (or reinvading) rats generally continues for two years between the end of the eradication phase and declaring the island rat‐free. This is based on the average life expectancy of a wild adult rat (c. 18 months) and breeding cycles (of template zones) and is an internationally accepted method of determining the success of a rat eradication project. This can be reduced to one year in tropical situations due to the increased breeding capacity of rats in these environments. This long‐term monitoring involves maintaining and checking the set of biosecurity stations around the island at least every six weeks; 39 permanent bait stations were established on Redonda (Figure 26). Additional monitoring points (using flavoured wax, tracking tunnels and trail cameras) can also be used, particularly in areas that had high rat numbers and activity during the eradication. If these methods (i.e. flavoured wax and tracking tunnels) are used, these devices should be in place and checked daily (or every two to three days) as non‐target species can also target these points and make detection of rats difficult. As such, this additional monitoring should be undertaken when multi‐day expeditions or consecutive visits to Redonda are planned. The biosecurity stations on Redonda comprise 39 lockable stations (containing bait and flavoured wax or resin) that are 50 m apart and have been mapped using GPS (Figure 26). Monitoring and maintenance of the stations will be undertaken by RRP, EAG or Department of Environment staff and volunteers for the foreseeable future. In addition to these stations, monitoring should be carried out randomly at other sites across Redonda. It is also important to check muddy or sandy areas for rodent tracks and droppings. It is important to monitor using a range of detection devices (such as flavoured wax, tracking tunnels, chew cards, traps, rodent motels and trail cameras) and have a regular search effort, particularly on overnight trips to the island. Low numbers of rats may take longer to detect than realised. Vigilance is important to detect incursions or surviving rats. Once the one or two‐year monitoring has been performed and no rats have been detected, a further intensive island‐wide monitoring check should be completed. This involves putting a range of monitoring devices over the entire island and checking daily for up to a month. This ‘final check’ should be completed by the experienced eradication specialist. Once this check is completed and no rats have been detected, Redonda can be declared rat‐free.

Bell, Ibbotson & Daltry 36 Redonda Rat Eradication: Technical Report

Figure 27 Biosecurity stations on Redonda.

Bell, Ibbotson & Daltry 37 Redonda Rat Eradication: Technical Report

7 BIODIVERSITY MONITORING As well as long‐term monitoring for rats, the ecology of Redonda should be studied. Seabird, land bird, lizard, invertebrate and vegetation surveys should be completed using standardised, repeatable methods. Baseline surveys for some species have been completed by FFI, EAG and DOE staff and visiting scientists (such as Harvard University) during the feasibility, pre‐assessment and eradication operations and will continue. Biodiversity monitoring is a central part of the Redonda Restoration Programme and methods are being described in a separate manual which will be completed in 2018. It is important to continue and expand the monitoring projects for vegetation (with more vegetation plots and photo points) and invertebrates to detect changes of these species. It will be important to determine population recovery for all seabirds and these other species as well as monitoring other ecological changes on the island as this information can support and inform future eradication proposals. 8 BIOSECURITY With the eradication of black rats from Redonda, the priority is to ensure that they do not become re‐established on the island. A detailed Biosecurity Plan for Redonda is being developed and will be available before the end of 2017. Prevention of rat re‐infestation should be the primary aim. Precautions need to be taken not only in obvious situations, but also when the risk may be mistakenly thought to be negligible. The greatest risk of reinvasion is by rats reaching Redonda with private vessels and fishing vessels or a shipwreck. Information should be shared throughout the region to ensure that locals and visitors are aware of the rat‐free status of Redonda and ways they can assist in keeping the island rodent‐free. Quarantine measures should be designed to reduce the risk of rats being re‐introduced to a minimum, without being too much of a hindrance to ongoing EAG, Department of Environment and other agencies. It will be important to focus on advocacy and education regarding biosecurity protocols and methods. This will enable stakeholder groups and Antigua community to take ownership of keeping Redonda rat‐free. The present method of reaching Redonda by helicopter already reduces the chance of rats being able to stow away. However, quarantine measures can be improved without significant impacts. Bait stations should be maintained in store sheds and in loading areas in Antigua. Staff loading equipment and supplies onto the helicopter should check these for rat damage. Any boxes that have been stored for long periods of time should be opened and checked prior to the visit to Redonda to ensure they don’t harbour rats before loading. This should not delay visits to the island as rat sign/damage is highly visible and easily detected. The permanent biosecurity stations and monitoring devices (rodent motels, tracking tunnels and flavoured wax) have to be maintained indefinitely and regularly checked on Redonda by the RRP team, EAG or Department of Environment staff or community members. It is important to use a variety of baits (i.e. alternative toxins), lures and monitoring techniques regularly throughout the biosecurity monitoring for rodent

Bell, Ibbotson & Daltry 38 Redonda Rat Eradication: Technical Report

incursions. One type of bait should not be used long‐term. The early interception of incursions is vital, and it is recommended that surveillance (using rodent motels, traps, tracking tunnels etc.) is undertaken at least every 2 months. Any rodent caught in a trap or fresh rat carcass found on the island should be sent for DNA sampling to determine provenance (i.e. failure of eradication programme or incursion from a neighbouring island or the mainland). DNA samples were collected from rats trapped during the Redonda eradication and are currently stored by the Department of Environment. The Biosecurity Plan for Redonda should include information on invertebrate and plant pests, parasites and diseases and simple protocols to deal with these issues. Local staff from the RRP team, EAG, DOE and other stakeholder groups will have to manage the biosecurity and quarantine procedures on Redonda. Data collection and management is important (particularly if incursions are detected and subsequently eradicated); all sightings and other rodent‐related observations should be recorded and investigated. The various stakeholder groups and community should also be involved in all aspects of the biosecurity requirements for Redonda as they have to maintain vigilance following the eradication and could assist with the detection and reporting of rodent (or any other biosecurity issue) activity. Periodic audits and on‐going monitoring of these biosecurity measures should be completed as it is common for people and agencies to become complacent and have standards or support for the programme drop. It is important that all involved (i.e. stakeholders, community and relevant agencies) realise that biosecurity is a long‐term ongoing commitment. WMIL would also be available to undertake audits of any aspect of the Redonda project including on‐going monitoring and biosecurity. The Biosecurity Plan provides details to minimize the risk of accidental liberation of rodents, and what measures should be taken if a rodent is sighted or sign is found on the island. Biosecurity is important for Redonda. Regardless of a successful eradication project, it is important to prevent other invasive species (i.e. parasites and weeds) as the ecosystem could be devastated if these reached the island. Despite the best quarantine measures, there is always the chance of an accidental rat re‐introduction. Various contingency procedures outline the specific action to be taken in the event of a rat sighting or a shipwreck (usually establishing and maintaining a poison and/or trapping grid in the area). In the event of a rat sighting, action should be taken immediately. The RRP team should maintain a Rodent Contingency Kit (containing bait stations, bait, monitoring tools etc.) which will enable an immediate response to a rat sighting on Redonda. Contingency funding should also be set aside to help deal with any future incursions and covers costs such as equipment, transport, specialist personnel (i.e. rope access, eradication specialists, etc.) and rodenticide. Quick action is vital. Further biosecurity information can be found in papers and reports from the New Zealand Department of Conservation, which has been working with rat‐free islands for decades (www.doc.govt.nz). In particular, the Standard Operating Procedures for Island Biosecurity report has excellent suggestions for quarantine procedures and contingency requirements for rat‐free islands (Roberts 2003,

Bell, Ibbotson & Daltry 39 Redonda Rat Eradication: Technical Report

DOC 2005). Although much of this report (and other Department of Conservation papers) and the recommendations are for optimal situations, many of these procedures could be adapted to suit Redonda. The RRP team, EAG and Department of Environment staff and volunteers have been trained in identifying rat sign (i.e. teeth marks, chews, droppings) and implementing rat contingency procedures in case of a rat being seen on the island. WMIL would also be available to undertake refresher courses in rodent identification and detection procedures as required or undertake wider‐scale training of local agencies to build capacity in the region in these rodent identification and biosecurity procedures. The importance of Redonda as an ecological reserve will be greatly increased by the eradication of black rats. The eradication, whilst removing a non‐native predator, also opens up the opportunity to restore and preserve the seabird communities. Prevention of rat re‐infestation should be a primary aim of all parties with an interest in Redonda. 9 CONCLUSIONS The combination of a ground‐based and aerial operation on Redonda has shown that a well‐planned, adequately resourced, well‐executed programme, with the support of the local agencies and government and led by an experienced operator with dedicated volunteers means total eradication is possible. At this stage, and it is important to stress that although further long‐term monitoring is required, results point to the likelihood that black rats have been successfully eradicated from Redonda. The Redonda eradication operation appeared to have had little secondary and primary non‐target species affected by the poisoning programme despite a high level of interference by land crabs. Only one peregrine was affected by secondary poisoning. Much of the island was cleared of rats within two weeks. Bait take showed that rats were spread across the whole island. This activity pattern suggests that rats would have been having a significant effect on seabird and lizard populations on Redonda.

The eradication of introduced mammalian predators from islands has become one of the most important tools in avian conservation worldwide; significant long‐term benefits are achieved, with short‐term investment. For restoration of seabird colonies on islands, rat eradication is an important requirement. The eradication of black rats from Redonda will benefit the seabird species on the island as well as the lizard and invertebrate populations. It will be important to monitor the response of the species (plants, invertebrates, mammals, reptiles and birds) present on Redonda following the eradication of black rats. Most species have benefited from rat eradications on islands, but there have also been unforeseen and negative impacts recorded in several projects around the world (Courchamp, et al., 2003; Towns, et al., 2006; Witmer, et al., 2007; Varnham, 2010; Bell, et al., 2011; Bell, 2012).

Bell, Ibbotson & Daltry 40 Redonda Rat Eradication: Technical Report

The successful eradication of black rats from Redonda demonstrates how combining the requirements and techniques of both ground‐based and aerial programmes and working through any difficulties that arose during the operation can be utilised on larger islands throughout the Caribbean region. It is important to stress that keeping Redonda free of rats will require constant vigilance and an ongoing commitment from all agencies, interested parties and the Anguillan community to prevent, detect and respond to any incursions. 10 RECOMMENDATIONS WMIL recommends: 1. That long‐term monitoring checks for rats continue every two months over the next two years (until April 2019) in the permanent bait stations by RRP, EAG or DOE staff and/or volunteers. 2. That the permanent bait and monitoring stations are maintained indefinitely as part of the biosecurity requirements for Redonda. 3. That Klerat® bait and flavoured wax or resin is used in the permanent bait stations for the ongoing biosecurity procedures on Redonda. 4. That WMIL return to the island to undertake an intensive final monitoring check in 2019 (February/March 2019) before the island is declared rat‐free. 5. That RRP, EAG and DOE be responsible for the maintenance of permanent bait and monitoring stations and other management requirements for the biosecurity on Redonda. 6. That RRP, EAG and DOE staff and volunteers are trained in identifying rat sign (i.e. teeth marks, chews, droppings) and implementing biosecurity and incursion procedures in case of a rat being seen on the island. 7. That the seabird populations on Redonda continues to be monitored regularly by RRP, EAG and DOE and other agencies such as RSPB and FFI to determine and document recovery and/or changes. 8. That the lizard populations on Redonda continues to be monitored every five years by RRP, EAG and DOE and other agencies such as Harvard University and FFI to determine and document recovery and/or changes. 9. That the vegetation on Redonda continues to be monitored RRP, EAG and DOE and other agencies such as FFI to determine and document recovery and/or changes. 10. That WMIL undertakes regular audit inspections of the biosecurity procedures, protocols and systems and updates training of RRP, EAG and DOE and other interested parties in rodent identification, monitoring and detection.

Bell, Ibbotson & Daltry 41 Redonda Rat Eradication: Technical Report

11 REFERENCES Atkinson, I.A.E. (1985). The spread of commensal species of Rattus to oceanic island and their effects on island avifauna. Pages 35‐81 in Moors, P.J. (ed.). Conservation of island birds. ICBP Technical Report No. 3. International Council for Bird Preservation. Cambridge, UK. Bell, B.D. (2002) The eradication of alien mammals from five offshore islands, Mauritius, Indian Ocean. In Turning the Tide: the Eradication of Invasive Species (eds D. Veitch & M. Clout), pp. 40‐45. IUCN SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and Cambridge, United Kingdom. Bell, E., Boyle, D., Floyd, K., Garner‐Richards, P., Swann, B, Luxmoore, R., Patterson, A. and Thomas, R. (2011) The ground‐based eradication of Norway rats (Rattus norvegicus) from the Isle of Canna, Inner Hebrides, Scotland. In Island Invasives: Eradication and Management (C.R. Veitch, M.N. Clout & D.R. Towns), pp. 269‐274. IUCN, Gland, Switzerland. Bell, E.A. and Daltry, J.C. (2012) Feasibility Study for the Eradication of Black Rats Rattus rattus from Redonda, with New Observations on the Island’s Biodiversity and Ecology. Report from Wildlife Management International Ltd and Fauna & Flora International to the Offshore Islands Conservation Programme, St John’s, Antigua and Barbuda. Bell, E.A. (2012) Dog Island Restoration Project: the Eradication of Black Rats (Rattus rattus) from Dog Island, Anguilla. Unpublished technical report prepared for Anguilla National Trust, Department of Environment, Government of Anguilla and Fauna & Flora International. Bell, E.A. (2004). Lundy Seabird Recovery Project 2002‐2004: the eradication of black and brown rats from Lundy Island. Unpublished report to Royal Society for the Protection of Birds. South‐west Regional Office, Devon, UK. Bell, E.A. (2012). Dog Island Restoration Project: Technical report for the eradication of black rats Rattus rattus from Dog Island, Anguilla. Unpublished technical report to the Anguilla National Trust, Fauna & Flora International, Department of Environment and Royal Society for the Protection of Birds. Bell, E.A., Bell, B.D. and Merton, D.V. (2016). The legacy of Big South Cape: rat irruption to rat eradication. New Zealand Journal of Ecology 40(2): 212‐218. BirdLife International (2012) Important Bird Areas Factsheet: Redonda. Http://www.birdlife.org/datazone/ sitefactsheet.php?id=19939on [Accessed 12 October 2012]. Blackburn, T.M., Cassey, P., Duncan, R.P., Evans, K.L. and Gaston, K.J. (2004) Avian extinctions and mammalian introductions on oceanic islands. Science 305: 1955‐ 1958.

Bell, Ibbotson & Daltry 42 Redonda Rat Eradication: Technical Report

Campbell, E.W. (1991) The effect of introduced roof rats on bird diversity of Antillean cays. Journal of Field Ornithology, 62, 343‐348. Courchamp, F., Chapius, J‐L., and Pascal, M. 2003. Mammal invaders on islands: impacts, control and control impact. Biological Review 78: 347‐383. Cunningham, D.M. and Moors, P.J. (1996) A Guide to the Identification and Collection of New Zealand Rodents. Occasional Publication No. 4. New Zealand Wildlife Service, Department of Internal Affairs, Wellington. New Zealand. Daltry, J. (2000). The second rat eradication on Praslin Island, Saint Lucia: August – September 2000. Unpublished report to the Saint Lucia Forest and Lands Department and the Durrell Wildlife Conservation Trust. Daltry, J.C. (2007) An introduction to the herpetofauna of Antigua, Barbuda and Redonda, with some conservation recommendations. Applied Herpetology, 4, 97–134. Daltry, J.C., Bloxam, Q., Cooper, G., Day, M.L., Hartley, J., Henry, M., Lindsay, K. and Smith, B.E. (2001) Five years of conserving the ‘world’s rarest snake’, the Antiguan racer Alsophis antiguae. Oryx, 35, 119‐127. Daltry, J.C., James, K.J., Otto, A. and Ross, T.N. (2012) Evidence that eradicating black rats (Rattus rattus) has boosted the recovery of rare reptiles and sea birds on Antiguan islands. Island Biodiversity: Flora, Fauna and Humans in the Lesser Antilles, 6‐8 November, Martinique, France. Day, M. and Daltry, J. (1996) Rat eradication to conserve the Antiguan racer. Aliens, 3, 14‐ 15. Department of Conservation (DOC). 2005. Island Pests. Brochure published by Department of Conservation, Wellington, New Zealand. Available via internet link as follows: www.doc.govt.nz/Regional‐Info/008~Wellington/pdf/Island‐Pests‐Brochure.pdf Devine, B., Drayton, N., Lindsay, K., Thomas, T., Cooper, B. and Nieves, P. (2010) Protected Areas Analysis for Antigua and Barbuda: Assessment and Mapping of Antigua and Barbuda Ecosystem Resources and Promoting a System of Protected Areas for Antigua and Barbuda. Unpublished technical report submitted to Ministry of the Environment, Government of Antigua and Barbuda. Eason, C.T. and Wickstrom, M. (2001) Vertebrate Pesticide Toxicology Manual (Poisons). Department of Conservation Technical Series 23. Wellington, New Zealand. Fisher, P. and Fairweather, A. (2005) Brodifacoum: A Review of Current Knowledge. Department of Conservation Pesticide Information Review No. 6 (No. HAMRO‐ 106522). Wellington, New Zealand. Havery, S.J., Lawrence, S.N., Daltry, J.C. and Bell, E.A. (2014) Technical Report for the Eradication of Black Rats Rattus rattus and Small Asian Mongooses Herpestes javanicus from Pelican and Codrington Islands, Antigua. Report for the Offshore Islands Conservation Programme, St John’s, Antigua and Barbuda.

Bell, Ibbotson & Daltry 43 Redonda Rat Eradication: Technical Report

Hedges, S.B. & Conn, C.E. (2012) A new skink fauna from Caribbean islands (Squamata, Mabuyidae, Mabuyinae). Zootaxa, 3288, 1‐244. Horwith, B. & Lindsay, K. (1997) A Biodiversity Profile: Antigua, Barbuda and Redonda. East Caribbean Biodiversity Programme Publication 3. Island Resources Foundation, USA. Howald, G., Donlan, C.J., Galvan, J.P., Russell, J.C., Parkes, J., Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. and Tershey, B. (2007) Invasive rodent eradication on islands. Conservation Biology, 21, 1258‐1268. Howard, R.A. (1962) Botanical and other observations on Redonda, the West Indies. Journal of the Arnold Arboretum, 63, 51‐66. Jones, H.P., Tershy, B.R., Zavaleta, E.S., Croll, D.A., Keitt, B.S., Finkelstein, M.E. and Howald, G.R. (2008) Review of the global severity of the effects of invasive rats on seabirds. Conservation Biology, 22, 16‐26. King, C.M. (1990) The Handbook of New Zealand Mammals. Oxford University Press. Auckland, New Zealand. Lawrence, N., James, K., Otto, A., Ross, T.N. and Daltry, J.C. (2013) Is eradicating rats worth it? Impacts observed on Antigua’s offshore islands. Program and Abstracts: Society for the Conservation and Study of Caribbean Birds, 19th Regional Meeting, July 26‐ 31. Georgetown, Grenada. Lawrence, S.N. and Daltry, J.C. (2015) Antigua announces its 15th island cleared of invasive alien mammals. Oryx, 49, 389. Lindsay, K. and Horwith, B. (1997) A Vegetation Classification of Antigua, Barbuda and Redonda. East Caribbean Biodiversity Programme Publication 2. Island Resources Foundation. Long, J.L. (2003) Introduced Mammals of the World: Their History, Distribution and Influence. CABI, Wallingford, UK. Lowrie, K., Lowrie, D. and Collier, N. (2012) Seabird Breeding Atlas of the Lesser Antilles. Environmental Protection in the Caribbean, Sint Maarten, Netherlands Antilles. Malhotra, A. and Thorpe, R.S. (1999) Reptiles & Amphibians of the Eastern Caribbean. Macmillan Education Ltd., London, UK. Martin, A.R.; Richardson, M.G. 2017. Rodent eradication scaled up: clearing rats and mice from South Georgia. Oryx: 1‐9. DOI: 10.1017/S003060531700028X Merton, D. (1987) Eradication of rabbits from Round Island, Mauritius: a conservation success story. Dodo, 24, 19‐43. Merton, D., Climo, G, Laboudallon, V., Robert, S. and Mander, C. (2002) Alien mammal eradication and quarantine on inhabited islands in the Seychelles. In Turning the Tide: The Eradication of Invasive Species (eds C.R. Veitch & M.N. Clout), pp. 182‐198. IUCN SSC Invasive Species Specialist Group, Gland, Switzerland.

Bell, Ibbotson & Daltry 44 Redonda Rat Eradication: Technical Report

Moors, P.J. and Atkinson, I.A.E. (1984). Predation on seabirds by introduced animals and factors affecting its severity. Pages 667‐690 in Croxall, J.P, Evans, P.G.H. and Schreiber, R.W. (Eds.). Status and conservation of the world’s seabirds. ICBP Technical Publication 2. Morse, A.R. (1979) The Quest for Redonda: the Quest for M.P. Shiel’s Realm of Redonda. The Reynolds Morse Foundation, Cleveland, Ohio, USA. Nowak, R.M. (1999) Walker’s Mammals of the World: Volume II. The Johns Hopkin Universtity Press, London, United Kingdom. Office of the Prime Minister (2001) Biodiversity Strategy and Action Plan for Antigua and Barbuda. Government of Antigua and Barbuda, St John’s, Antigua and Barbuda. Pain, D.J., Brooke, M.D., Finnie, J.K and Jackson, A. (2000) Effects of brodifacoum on the land crabs of Ascension Island. Journal of Wildlife Management, 64, 380‐387. Parks & Wildlife Service. (2008) Macquarie Island Pest Eradication Plan. Part A: Overview. Department of the Environment, Parks, Heritage and the Arts, Hobart, Tasmania. Available for download from http://www.parks.tas.gov.au/index.aspx?base=6186. Parks and Wildlife Service. (2014). Evaluation Report: Macquarie Island Pest Eradication Project, August 2014. Department of Primary Industries, Parks, Water and Environment. Hobart Tasmania. Prosper, J., Joseph, V., Otto, A. and Prosper, S. (2008) Antigua and Barbuda. In Important Bird Areas in the Caribbean (eds D.C. Wege & V. Anadon‐Irizarry), pp. 35‐46. BirdLife International, Cambridge, UK. Raffaele, H., Wiley, J., Garrido, O., Keith, A. and Raffaele, J. (2011). Birds of the West Indies. Helm Field Guides, London. Roberts, A. (2003). Island Biosecurity: Standard Operating Procedures: Best Practice Manual. Department of Conservation, Southland, New Zealand. Sylvester, I., Joseph, V., Prosper, J., Prosper, S. and Hay, B. (2009). Antigua‐Barbuda‐Redonda and Selected Offshore Islands. An Inventory of Breeding Seabirds of the Caribbean (eds. P.E. Bradley & R.L. Norton), pp. 169‐175. University Press of Florida, USA. Thorsen, M., Shorten, R., Lucking, R. and Lucking, V. (2000) Norway rats (Rattus norvegicus) on Fregate Island, Seychelles: the invasion, subsequent eradication attempts and implications for the island's fauna. Biological Conservation, 96, 133‐ 138. Towns, D.R. and Broome, K.G. (2003) From small Maria to massive Campbell: forty years of rat eradications from New Zealand islands. New Zealand Journal of Zoology, 30, 377‐398. Towns, R.T. Atkinson, I.A.E. and Daugherty, C.H. (2006) Have the harmful effects of introduced rats on islands been exaggerated? Biological Invasions, 8, 863‐891.

Bell, Ibbotson & Daltry 45 Redonda Rat Eradication: Technical Report

Varnham, K. and Daltry, J.C. (2006) Eradication of Invasive Black Rats (Rattus rattus) from Green and York Islands, Antigua: June‐July, 2006. Unpublished technical report to the Mill Reef Club and partners of the Antiguan Racer Conservation Project. Varnham, K. (2003) Eradication of Black Rats (Rattus rattus) From Sandy Cay, British Virgin Islands. Unpublished technical report for Island Resources Foundation, US Virgin Islands. Varnham, K. (2010) Invasive Rats on Tropical Islands: Their History, Ecology, Impacts and Eradication. RSPB Research Report No. 41. Royal Society for the Protection of Birds, Sandy, United Kingdom. Veitch, C.R. (2002) Eradication of Pacific rats (Rattus exulans) from Fanal Island, New Zealand. In Turning the Tide: The Eradication of Invasive Species (eds C.R. Veitch & M.N. Clout), pp. 357‐359. IUCN SSC Invasive Species Specialist Group. IUCN, Gland, Switzerland and Cambridge, United Kingdom. Wedding, C.J. (2007) Aspects of the impacts of mouse (Mus musculus) control on skinks in Auckland, New Zealand. Unpublished MSc thesis, Massey University, Auckland, New Zealand. Witmer, G.W., Boyd, F. and Hillis‐Starr, Z. (2007) The successful eradication of introduced roof rats (Rattus rattus) from Buck Island using diphacinone, followed by an irruption of house mice (Mus musculus). Wildlife Research, 34, 108–115.

Bell, Ibbotson & Daltry 46 Redonda Rat Eradication: Technical Report

12 APPENDICES 12.1 TIMETABLE OF ACTIVITIES FOR THE REDONDA RAT ERADICATION (29 JANUARY TO 13 APRIL 2017) ACTIVITY DATE Planning and preparation  Elizabeth (Biz) Bell (Rat Eradication Team Leader), Jack Ibbotson (Mountaineer Advisor) 29/1/2017 and Jenny Daltry (FFI) arrive  Meetings and equipment purchases 30/1/2017  One feral goat transport off Redonda  Equipment check at Seaton’s shed 31/1/2017  Meetings to finalise equipment and operational planning  Jenny departed Antigua  One feral goat transported off Redonda 1/2/2017  Equipment check at Seaton’s shed  Sorted equipment 2/2/2017  Climbing equipment checked  Adam Long (Mountaineer Advisor) arrived  Equipment organised 3/2/2017  Purchased food  Jack and Adam out to Redonda  One feral goat picked up from Redonda 4/2/2017  Salina Janzan (Rat Eradication Intern) off Redonda  Equipment check at Seaton’s shed 5/2/2017  Chris Clarke, Thea Eldred, Edward Marshall, James Stops, John Tayton, Silas Walton and Bede West (Rat Eradication Interns) arrive 6/2/2017  Purchase of food  Project briefing for interns 7/2/2017  Moved bait from Seaton’s to Caribbean Helicopters  Purchase of food and gas  Sophia Steele (FFI) to Redonda 8/2/2017  Jack and Adam return from Redonda  First Aid Training 9/2/2017  First Aid Training 10/2/2017 Day off 11/2/2017  Team trip to Great Bird Island Day off 12/2/2017 Phase I: Establishing bait station grid  Whole team to Redonda  Klerat® bait delivered to Redonda 13/2/2017  Established base camp  Cliff baiting 0: Rope‐access team begin baiting (10 blocks per station, plus 20 blocks hand‐ broadcast) on eastern side 14/2/2017  Lines V, A, B and C established on top of island  Cliff baiting 0 continued: Southern end and beach  Lines D, E and F established on top of island 15/2/2017  Unweaned new‐born feral goat captured  Cliff baiting 0 continued: Access gully and throw paths established 16/2/2017  Lines G, H, J, K, L, M and N established on top of island

Bell, Ibbotson & Daltry 47 Redonda Rat Eradication: Technical Report

Phase II: Poisoning  Ground baiting 0: Ground‐team begin baiting (5 blocks per station)  Helicopter baiting 1: 1.5 kg/ha, 70 kg (20g Klerat® blocks)  Cliff baiting 0 continued: West‐side scramble completed, and bolts put in place 17/2/2017  Klerat® bait delivered to Redonda  Shanna Challenger (FFI) arrived and trained in eradication techniques  One feral goat captured  Ground baiting 0 continued: Final line of beach established and baited (5 blocks per station)  Cliff baiting 0 continued: final section of rope‐accessible beaches baited 18/2/2017  James injured shoulder  Two feral goats caught  Ground bait check 1  Jenny visited Redonda  James left island for medical treatment 19/2/2017  One feral goat caught  Three feral goats transported off Redonda  Ground bait check 2  Cliff bait check 1  Shanna, Salina, Thea and Adam off Redonda 20/2/2017  One feral goat transported off Redonda  Progress report 1  Ground bait check 3  Cliff bait check 1 continued 21/2/2017  Blue rat droppings recorded  Ground bait check 4  Cliff bait check 1 continued 22/2/2017  Helicopter baiting 2: 1.5 kg/ha, 140 kg (5g Klerat® blocks) + 20 kg waste bait  Klerat® bait delivered to Redonda  Ground bait check 5  Cliff bait check 1 continued 23/2/2017  Dead rat carcasses collected on surface  Humpback whales seen off island  Ground bait check 6  Cliff bait check 1 continued  Helicopter baiting 3: 1.5 kg/ha, 80 kg (5g Klerat® blocks) + 20 kg waste bait  Klerat® bait delivered to Redonda 24/2/2017  Salina, Thea and James on, Ed, Silas and Bede off  Michai (DOE) on  Dead rat carcasses collected on surface  Ground bait check 7  Cliff bait check 1 continued  Harvard University researchers on  Antiguan hunters with dogs arrived to capture goats (one female caught, but left on island with eradication team) 25/2/2017  Two feral goats caught  Two feral goats transported off Redonda  Peregrine falcon carcass found – no symptoms of poisoning or injuries observed due to decomposition  Dead rat carcasses collected on surface

Bell, Ibbotson & Daltry 48 Redonda Rat Eradication: Technical Report

 Ground bait check 8 26/2/2017  Cliff bait check 1 continued  Ground bait check 9  Cliff bait check 2  Helicopter baiting 4: 1.5 kg/ha, 90 kg (5g Klerat® blocks) + 20 kg waste bait  Klerat® bait delivered to Redonda 27/2/2017  Peregrine falcon carcass found – confirmed symptoms of secondary poisoning  Silas, Ed & Bede on, John, Jack & Chris off  Dead rat carcasses collected on surface  Progress report 2  Ground bait check 10: bait reduced to 2 blocks per station  Cliff bait check 2 continued 28/2/2017  Fresh dead rat found  Ground bait check 11  Monitoring (soap) put out at and between all ground stations 1/3/2017  Monitoring (soap) put out of part of the climbing routes  Cliff bait check 2 continued  Ground bait check 12  Monitoring check 1 2/3/2017  Cliff bait check 3  Ground bait check 13  Monitoring check 2  Cliff bait check 3 continued  Helicopter baiting 5: 1.5 kg/ha, 180 kg (5g Klerat® blocks) + 20 kg waste bait 3/3/2017  Jack, John & Chris on, Biz, James & Thea off  One feral goat captured  One feral goat transported off Redonda  Ground bait check 14  Monitoring check 3  Cliff bait check 3 continued 4/3/2017  One feral goat captured  Harvard University researchers off  Ground bait check 15  Monitoring check 4: Chocolate wax replaced soap  Cliff bait check 3 continued 5/3/2017  One feral goat captured  Progress Report 3  Ground bait check 16  Monitoring check 5  Cliff bait check 4 6/3/2017  Helicopter baiting 6: 1.5 kg/ha, 180 kg (5g Klerat® blocks) + 20 kg waste bait  Klerat® bait delivered to Redonda  Biz, James & Thea on, Silas, Salina & Bede off  Ground bait check 17  Monitoring check 6: monitoring only placed in‐between stations only 7/3/2017  Cliff bait check 4 continued  Ground bait check 18  Monitoring check 7 8/3/2017  Cliff bait check 4 continued  Rat recorded on trail camera (D5 and E6)

Bell, Ibbotson & Daltry 49 Redonda Rat Eradication: Technical Report

 Rat traps placed out at active site (D5 and E6)  Ground bait check 19  Monitoring check 8  Cliff bait check 5  Silas, Salina & Bede on, John & Ed off  Klerat® and Final® bait delivered to Redonda 9/3/2017  Jenny arrives in Antigua  One feral goat captured  One feral goat transported off Redonda  No rats caught in traps at D5/E6, fresh take on bait in those stations  Ground bait check 20  Monitoring check 9  Cliff bait check 5 continued 10/3/2017  Brown booby tangled in bait throw line, released unharmed, throw lines removed  Peregrine falcon carcass found – no symptoms of poisoning or injuries observed  Ground bait check 21  Monitoring check 10: aniseed wax replaced chocolate wax 11/3/2017  Cliff bait check 5 continued  One feral goat captured  Ground bait check 22  Monitoring check 11  Cliff bait check 5 continued 12/3/2017  One feral goat died in corral  Progress report 4  Helicopter baiting 7: 1.5 kg/ha, 180 kg (5g Klerat® blocks) + 20 kg waste bait  Final® bait delivered to Redonda  John, Ed & Jenny on, James, Thea & Chris off 13/3/2017  Feral goat dissected (case of death appeared to be twisted intestines)  Camp tidied and packed up (due to forecast winds)  Ground bait check 23  Monitoring check 12: monitoring put back at and in‐between stations 14/3/2017  Cliff bait check 6  Ground bait check 24: Final® put out (2 blocks)  Monitoring check 13: chocolate wax replaced aniseed wax  Cliff bait check 6 continued  One feral goat captured 15/3/2017  Lizard point counts completed  Bananaquit seen  Access gully avalanche (access route destroyed)  Ground bait check 25: Final® (2 blocks)  Monitoring check 14  Cliff bait check 6 continued  One feral goat captured 16/3/2017  Lizard point counts completed  Malaise trap put out  Further collapses in the access gully avalanche (access route destroyed)  Ground bait check 26: Final® replaced with Klerat® (2 blocks)  Monitoring check 15 17/3/2017  Cliff bait check 6 continued  Jenny, Biz, Silas, Salina & Bede off, James, Thea & Chris on

Bell, Ibbotson & Daltry 50 Redonda Rat Eradication: Technical Report

 One feral goat transported off Redonda  Ground bait check 27  Monitoring check 16 18/3/2017  Cliff bait check 6 continued  Ground bait check 28  Monitoring check 17  Cliff bait check 6 continued 19/3/2017  Bede leaves Antigua  Progress Report 5  Ground bait check 29  Monitoring check 18: peanut wax replaced chocolate wax  Helicopter baiting 8: 1.5 kg/ha, 180 kg (5g Klerat® blocks) + 20 kg waste bait (Final™) 20/3/2017  Biz, Silas & Salina off, Jack, Ed & John off  Two feral goats captured  Ground bait check 30  Monitoring check 19 21/3/2017  Cliff bait check 7  Ground bait check 30 continued  Monitoring check 19 continued  Cliff bait check 7 continued 22/3/2017  Department of Environment, National Parks and EAG personnel arrive for day visit  One feral goat captured  Ground bait check 31  Monitoring check 20  Cliff bait check 7 continued 23/3/2017  Possible rat sign reported (‘fresh” droppings) on cliff station (X20)  Traps, cameras and fresh monitoring put out at X20 and surrounding area  One feral goat captured  Helicopter baiting 9: 1.5 kg/ha, 180 kg (5g Klerat® blocks) + 20 kg waste bait (Final™)  No sign of rat at X20 or surrounding area  Four permanent biosecurity stations put in place  Pitfall traps put out 24/3/2017  One feral goat captured  Four feral goats transported off Redonda  Jack, Ed & John on, Chris, James & Thea off  Ground bait check 32  Monitoring check 21: chocolate wax replaced peanut wax  Cliff bait check 8: removed one line 25/3/2017  No sign of rat at X20 or surrounding area  Three feral goats captured  One feral goat transported off Redonda  Ground bait check 33  Monitoring check 22  Cliff bait check 8 continued: removed scramble lines  No sign of rat at X20 or surrounding area 26/3/2017  Permanent biosecurity station locations marked out  Eleven feral goats captured  American kestrel seen  Progress report 6  Helicopter baiting 10: 1.5 kg/ha, 100 kg (80 kg 5g Klerat®, 20 kg Final®) + 20 kg waste bait 27/3/2017

Bell, Ibbotson & Daltry 51 Redonda Rat Eradication: Technical Report

 Pitfall traps collected in  Permanent biosecurity stations drilled  Department of Environment, National Parks and EAG personnel arrive for day visit  Chris, Thea & James on, Silas & Salina off  Seven feral goats transported off island  Ground bait check 34  Monitoring check 23 28/3/2017  Cliff bait check 8 continued: removed eastern beach lines  Permanent biosecurity station locations marked out  Ground bait check 35: pulled in D, J, K & V  Monitoring check 24: pulled in D, J, K & V  Cliff bait check 8 continued: removed big ab and northern beach lines  One feral goat captured 29/3/2017  Euthanised an injured juvenile feral goat (found after it had fallen off cliff)  Audobon’s shearwater heard near camp  Tail‐less whip‐scorpion noted in small cave below Manager’s house  Bat seen flying over camp  Ground bait check 35 continued: pulled in E, F, N & M  Monitoring check 24 continued: pulled in E, F, N & M  Cliff bait check 8 continued: removed scramble lines  Permanent biosecurity stations drilled 30/3/2017  Two feral goats captured  Audubon’s shearwater heard near camp  Ed injured finger  Climbing equipment cleaned and removed from island  Permanent biosecurity stations drilled  EAG personnel arrive for day visit  One feral goat captured 31/3/2017  Seven feral goats transported off Redonda  Silas & Salina on, Ed & John off  Bat seen flying around camp  Ground bait check 35 continued: pulled in G, H & L  Monitoring check 24 continued: pulled in G, H & L  Permanent biosecurity stations steel bar put in place 1/4/2017  Pit fall traps put out  Bat seen flying around camp  Ground bait check 35 continued: pulled in A, AA, B & C  Monitoring check 24 continued: pulled in A, AA, B & C  Permanent biosecurity stations cemented in place 2/4/2017  One feral goat captured  Progress Report 7  Media members, goat host, Department of Environment and Shanna visit  One feral goat transported off island 3/4/2017  John & Ed on  Booby survey completed  Lizard survey completed  Lichen samples collected 4/4/2017  Permanent biosecurity stations put in place  Pitfall traps taken in  Lizard survey completed 5/4/2017

Bell, Ibbotson & Daltry 52 Redonda Rat Eradication: Technical Report

 Lichen samples collected  Permanent biosecurity stations put in place  One feral goat captured  Lizard survey completed  Repeated historic photographs 6/4/2017  Packed and moved gear to helipad  One feral goat captured  Whole team off Redonda  Aerial photography of island completed 7/4/2017  One feral goat captured  Seven feral goats transported off Redonda Project debrief  Equipment cleaned and sorted 8/4/2017  Debriefing meeting with Shanna and Sophia Day off 9/4/2017 Day off 10/4/2017  Team trip to Green Island  Rat Eradication Interns depart from Antigua  Jenny and Pete arrive in Antigua 11/4/2017  Meetings regarding the eradication  Steering Group Meeting 12/4/17  Steering Group Meeting 13/4/17  Biz departs from Antigua

Bell, Ibbotson & Daltry 53 Redonda Rat Eradication: Technical Report

12.2 BIRDS (AND BAT) SEEN DURING THE REDONDA RAT ERADICATION (JANUARY TO APRIL 2017) [Note: Common and scientific names follow Raffaele et al. 2011; however, some of these are now dated] Birds: Audubon’s Shearwater Puffinus lherminieri Heard calling above the Manager’s House on two occasions, 29 March 2017 and 30 March 2017. Searches of the area could not locate the bird(s) on the ground. Peregrine Falco peregrinus This species was very conspicuous being seen daily throughout the project, often seen interacting noisily and on one occasion up to five birds were noted together. Three peregrines were found dead during the operation; one had confirmed symptoms of anticoagulant rodenticide poisoning (found on 27 March 2017), one had injuries suggesting a fight between two peregrines (10 March 2017) and one was too decomposed to confirm cause of death (25 March 2017). American kestrel Falco sparverius Seen once at the northern end of the island on 26 March 2017. Lesser Antillean bullfinch Loxigilla noctis Seen in the top of the Access Gully feeding in fig tree on 25 March 2017. Caribbean Eleania Elaenia martinica Seen in the top of the Access Gully in fig tree on 25 March 2017. Zenaida Dove Zenaida aurita Recorded across the Island throughout; breeding pairs noted towards the end of the operation. Now resident on Redonda. Pearly‐eyed thrasher Margarops fuscatus Seen at the northern end of the island in fig tree on a number of occasions between 6 and 21 March 2017 and in the top of the Access Gully feeding in fig tree on 25 March 2017. Barn Swallow Hirundo rustica Noted towards the end of the operation in late March/early April, between 25 March and 6 April 2017. Either single birds or 2‐3 individuals flying around the Access Gully. Bananaquit Coereba flaveola One individual recorded at the Manager’s house on the Casuarina tree on 15 March 2017 and was seen for a day or two following. Red‐billed Tropicbird Phaethon aethereus Seen most days throughout flying and in nest crevices. Breeding was recorded with both eggs and chicks noted. It is estimated that there are at least 30 breeding pairs on Redonda. Red‐footed Booby Sula sula Breeding noted on six fig trees on the cliffs; at the southern cliffs beyond the access gully, in the access gully itself, eastern cliff above the Manager’s house, two on the western

Bell, Ibbotson & Daltry 54 Redonda Rat Eradication: Technical Report

cliffs, and at the northern end of the island. Both large chicks and adults incubating eggs being present throughout the duration of the project. Additional pairs in the early stages of nest building were also observed towards the end of the period. Brown Booby Sula leucogaster Breeding noted across the island; both large chicks and adults incubating eggs being present throughout the duration of the project. Additional pairs in the early stages of nest building were also observed towards the end of the period. At least 150 breeding pairs counted in April 2017 surveys, but estimated to have a higher overall population on Redonda. Masked Booby Sula dactylatra Breeding noted across the island; both large chicks and adults incubating eggs being present throughout the duration of the project. Additional pairs in the early stages of nest building were also observed towards the end of the period. At least 160 breeding pairs counted in April 2017 surveys, but estimated to have a higher overall population on Redonda. Magnificent Frigatebird Fregata magnificens Breeding noted on fig trees on the cliffs across the island; both large chicks and adults incubating eggs being present throughout the duration of the project. Additional pairs in the early stages of nest building were also observed towards the end of the period. At least 110 breeding pairs counted in April 2017 surveys. Bat: Bat – suspected to be Brazilian free‐tailed bat Tadarida brasiliensis Seen on the evening of 31 March 2017 and 1 April 2017 flying around the Manager’s house

Bell, Ibbotson & Daltry 55 Redonda Rat Eradication: Technical Report

12.3 RATS TRAPPED OR RECOVERED ON THE SURFACE DURING THE REDONDA PRE‐ASSESSMENT VISIT (JULY 2016) AND RAT ERADICATION (JANUARY TO APRIL 2017) Where under Colour phase: Frug = frugivorous; Alex = alexandrinus, Rat = rattus; and under Sex: M = male; F = female; U = unknown Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw P1 3/07/2016 Trap Helipad frug. M 246 204 235 23 35 37 large bald testes good vegetation lactating, prickly pear, P2 3/07/2016 Trap Helipad frug. F 218 190 221 20 35 38 good swollen ovaries vegetation P3 3/07/2016 Trap Helipad alex. F 222 197 226 26 36 39 lactating good seeds, vegetation lactating, prickly pear, P4 3/07/2016 Trap Helipad frug. F 188 185 214 21 33 36 good swollen ovaries vegetation P5 3/07/2016 Trap Helipad alex. F 204 186 227 26 38 41 lactating very good seeds, lizard lactating, vegetation, seeds, P6 3/07/2016 Trap Helipad frug. F 220 200 223 26 38 41 very good swollen ovaries prickly pear lactating, prickly pear, seeds, P7 3/07/2016 Trap Helipad frug. F 260 204 231 23 35 39 very good pregnant (3/3) slug vegetation, goat P8 3/07/2016 Trap Helipad frug. M 176 189 215 26 36 39 large bald testes very good faeces P9 3/07/2016 Trap Helipad frug. M 180 196 213 25 34 37 large bald testes good bile lactating, P10 3/07/2016 Trap Helipad frug. F 178 192 215 22 33 36 good vegetation pregnant (3/3) lactating, vegetation, lizard, P11 4/07/2016 Trap Helipad frug. F 174 183 222 22 35 38 good pregnant (1/1) seeds vegetation, goat P12 4/07/2016 Trap Helipad alex. M 164 175 201 23 34 36 large bald testes good faeces Detex®, prickly pear, P13 4/07/2016 Trap Helipad frug. F 216 198 217 23 35 38 lactating good seeds, vegetation slug, prickly pear, P14 4/07/2016 Trap Helipad frug. M 160 182 209 21 34 36 large bald testes good vegetation, seeds lizard, seeds, P15 4/07/2016 Trap Helipad frug. M 234 198 228 20 34 36 large bald testes good vegetation, blood and flesh 1 13/02/2017 Trap Managers house alex. M 182 211 26 38 40 large bald testes excellent seeds, plant material insects, plant 2 13/02/2017 Trap Managers house frug. M 193 231 23 38 41 large bald testes excellent material, goat faeces 3 13/02/2017 Trap Managers house frug. F 176 228 25 38 40 lactating good vegetation 4 13/02/2017 Trap Managers house frug. F 200 236 26 38 41 lactating good plant material, seeds

Bell, Ibbotson & Daltry 56 Redonda Rat Eradication: Technical Report

Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw 5 14/02/2017 Trap Managers house frug. M 188 219 24 39 41 large bald testes excellent plant material seeds, goat faeces, 6 14/02/2017 Trap Managers house frug. F 178 21 37 39 inactive poor plant material, fibre stomach ulcer, 7 14/02/2017 Trap Managers house alex. M 193 239 22 38 40 large bald testes fresh plant material pale liver 8 14/02/2017 Trap Managers house frug. F pregnant (4/0) decomposed empty 9 14/02/2017 Trap Managers house frug. M 193 216 21 37 39 large bald testes malformed testes plant material small testes peanut butter, plant 10 14/02/2017 Trap Managers house alex. M 160 207 21 36 38 ear chewed (immature) material 11 15/02/2017 Trap Managers house frug. M 183 233 23 39 41 large bald testes excellent plant material 12 15/02/2017 Trap Managers house frug. F 164 220 24 36 39 lactating good fibrous plant material 13 16/02/2017 Trap Managers house frug. M 190 222 19 37 39 large bald testes excellent fresh plant material seeds, fibrous plant 14 16/02/2017 Trap Managers house frug. M 188 265 21 31 33 large bald testes good material, leaves, insect carapaces lactating, lizard flesh, plant 15 16/02/2017 Trap Managers house frug. F 183 216 24 33 35 good pregnant (2/2) material, seeds old bite injury to 16 17/02/2017 Trap Managers house frug. M 179 221 24 35 37 large bald testes plant material shoulder unperforated, peanut butter, plant 17 17/02/2017 Trap Managers house frug. F 152 198 21 32 36 oviduct straight, white tip to tail material immature peanut butter, plant 18 18/02/2017 Trap Managers house alex. M 159 211 23 34 36 large bald testes good material 19 18/02/2017 Trap Managers house frug. M 168 237 23 38 40 large bald testes chewed left ear plant material 20 18/02/2017 Trap Managers house frug. M 184 269 26 37 39 large bald testes excellent lichen, plant material Tail damaged, peanut butter, plant 21 18/02/2017 Trap Managers house frug. M 202 251 24 36 39 large bald testes lump on side material egg (scavenged), 22 18/02/2017 Trap Managers house frug. F 185 267 23 38 40 pregnant (4/5) decomposed peanut butter, plant material pale liver, tail 23 19/02/2017 Surface near X3 frug. F 191 26 35 38 pregnant (4/3) shortened, bruising plant material, bait on skin and legs lactating, very pale liver, 24 19/02/2017 Surface near X9 frug. F 204 243 27 36 39 bait pregnant (2/2) bruised skin 25 19/02/2017 Surface near X6 frug. F 192 243 26 37 39 recently given pale liver, bruised bait, plant material

Bell, Ibbotson & Daltry 57 Redonda Rat Eradication: Technical Report

Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw birth, not pregnant, lactating oviduct straight, immature, but perforated, bruising but dark peanut butter, plant 26 19/02/2017 Trap Managers house frug. F 157 220 26 36 39 coming into liver material breeding condition mites, good, 27 20/02/2017 Trap Managers house alex. M 166 203 24 35 38 large bald testes greenish tinge to plant material fur 28 20/02/2017 Trap Managers house alex. M 184 217 24 36 39 large bald testes chewed ears bait, plant material 29 20/02/2017 Trap Managers house frug. M 184 217 24 36 39 large bald testes chewed ears bait, plant material head crushed by 30 20/02/2017 Trap Managers house frug. F 219 24 36 38 pregnant (3/4) bait, plant material trap 31 20/02/2017 Trap Managers house alex. M 180 222 25 35 37 large bald testes excellent plant material 32 20/02/2017 Trap Managers house frug. F 163 193 23 25 37 ready to breed pale liver bait, plant material lactating, slightly chewed prominent ear, bruising, pale 33 21/02/2017 Surface near Y55 frug. F 163 209 24 31 33 bait, plant material nipples, recently liver, blood in given birth stomach cavity 34 21/02/2017 Trap Managers house frug. M 173 233 24 35 37 large bald testes pale liver plant material getting into 35 21/02/2017 Surface near H18 alex. F 171 220 24 33 34 breeding lump on tail bait, plant material condition 36 21/02/2017 Surface near A7 frug. M 174 223 18 35 37 large bald testes chewed ear bait smashed head 37 21/02/2017 Trap Managers house frug. M 227 23 37 39 large bald testes bait, plant material from rock 38 23/02/2017 Surface D24 frug. F 183 241 21 38 41 bait 39 23/02/2017 Surface D16 frug. decomposed bait not in breeding 40 23/02/2017 Surface D3 frug. F 173 238 24 38 41 decomposed condition not in breeding pale liver, bruised 41 23/02/2017 Surface E7 frug. F 202 238 22 37 40 bait condition skin not in breeding pale liver, bruised rat flesh, bait, plant 42 23/02/2017 Surface E30 frug. F 187 233 23 38 41 condition skin material

Bell, Ibbotson & Daltry 58 Redonda Rat Eradication: Technical Report

Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw not in breeding pale liver, bruised 43 23/02/2017 Surface E43 frug. F 195 216 25 37 40 bait condition skin pale liver, bruised 44 23/02/2017 Surface goat tank frug. M 183 241 24 34 37 large bald testes bait, plant material skin not in breeding pale liver, bruised 45 23/02/2017 Surface B15 frug. F 189 209 25 36 38 bait condition skin 46 24/02/2017 Surface N1 frug. M 193 246 22 38 40 large bald testes decomposed bait 47 24/02/2017 Surface L9 frug. F 180 233 21 35 37 just given birth decomposed bait eaten by lizards, 48 24/02/2017 Surface helipad frug. M 198 235 24 37 40 large bald testes bait partly decomposed pregnant (3/3), 49 24/02/2017 Surface bait depot frug. F 186 219 23 34 36 decomposed bait lactating 50 24/02/2017 Surface helipad frug. M 176 230 23 35 37 large bald testes good bait, plant material 51 24/02/2017 Surface A22 frug. M decomposed 52 24/02/2017 Surface C4 frug. F decomposed 53 24/02/2017 Surface C12 frug. F decomposed 54 24/02/2017 Surface L9 frug. F decomposed 55 24/02/2017 Surface AA2 frug. M decomposed 56 24/02/2017 Surface V14 frug. F decomposed 57 24/02/2017 Surface G4 frug. F decomposed 58 24/02/2017 Surface D24 frug. M decomposed 59 25/02/2017 Surface C3 frug. M decomposed 60 25/02/2017 Surface E10 frug. F decomposed 61 25/02/2017 Surface D12 frug. M decomposed 62 25/02/2017 Surface S2 frug. M decomposed 63 25/02/2017 Surface S2 frug. U decomposed 64 25/02/2017 Surface X54 frug. M decomposed 65 25/02/2017 Surface D8 frug. M decomposed 66 25/02/2017 Surface X57 frug. F decomposed 67 25/02/2017 Surface X54 alex. M decomposed 68 25/02/2017 Surface V27 frug. M decomposed 69 25/02/2017 Surface D4 frug. F decomposed 70 25/02/2017 Surface D2 frug. F decomposed 71 25/02/2017 Surface X24 frug. M decomposed 72 25/02/2017 Surface X39 frug. M decomposed 73 25/02/2017 Surface N3 frug. F decomposed 74 25/02/2017 Surface F10 frug. M decomposed

Bell, Ibbotson & Daltry 59 Redonda Rat Eradication: Technical Report

Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw 75 25/02/2017 Surface B20 frug. F decomposed 76 25/02/2017 Surface E1 alex. F decomposed 77 26/02/2017 Surface G21 frug. F decomposed 78 26/02/2017 Surface D32 frug. M decomposed 79 26/02/2017 Surface B14 frug. F decomposed 80 26/02/2017 Surface F5 frug. F decomposed 81 26/02/2017 Surface C12 frug. F decomposed 82 26/02/2017 Surface B19 frug. F decomposed 83 26/02/2017 Surface A14 frug. U decomposed 84 26/02/2017 Surface A14 frug. M decomposed 85 26/02/2017 Surface B15 frug. U decomposed 86 26/02/2017 Surface Y65 frug. F decomposed 87 26/02/2017 Surface Y67 frug. F decomposed 88 26/02/2017 Surface Y44 frug. M decomposed 89 26/02/2017 Surface L16 frug. M decomposed 90 26/02/2017 Surface G7 frug. F decomposed 91 26/02/2017 Surface G18 frug. M decomposed 92 26/02/2017 Surface G21 frug. M decomposed 93 26/02/2017 Surface G24 frug. M decomposed 94 26/02/2017 Surface H30 frug. F decomposed 95 26/02/2017 Surface Camp frug. M decomposed 96 26/02/2017 Surface E6 frug. U decomposed 97 26/02/2017 Surface N1 frug. M decomposed 98 27/02/2017 Surface L9 frug. M decomposed 99 27/02/2017 Surface D5 frug. M decomposed 100 28/02/2017 Surface V20 frug. F decomposed 101 28/02/2017 Surface H4 alex. M decomposed fresh, pale liver, lactating, blood in body 102 28/02/2017 Surface E8 frug. F 196 247 25 37 40 recently given bait cavity, bruising on birth body walls 103 1/03/2017 Surface X20 frug. U decomposed 104 1/03/2017 Surface Y47 unknown U decomposed 105 1/03/2017 Surface A16 frug. M decomposed 106 5/03/2017 Surface C7 frug. U decomposed 107 5/03/2017 Surface C16 frug. U decomposed 108 5/03/2017 Surface C9 frug. U decomposed

Bell, Ibbotson & Daltry 60 Redonda Rat Eradication: Technical Report

Hind Rat Surface Colour Hind foot Date Trap/Location Sex Weight HBL Tail Ear Breeding status Condition Stomach contents no. or Trap phase foot + claw 109 7/03/2017 Surface B13a frug. U decomposed 110 7/03/2017 Surface V19a unknown U decomposed 111 7/03/2017 Surface L6a unknown U decomposed 112 8/03/2017 Surface Bait depot frug. U decomposed Water 113 8/03/2017 Surface frug. U decomposed catchment (D1) 114 12/03/2017 Surface Y66 frug. M decomposed

Bell, Ibbotson & Daltry 61