Captive Breeding and Future In-Situ Management of the Christmas Island Pipistrelle Pipistrellus Murrayi a Report to the Director of National Parks

Home , Christmas Island shrew, Pallid bat, Vespadelus

Captive breeding and future in-situ management of the Christmas Island Pipistrelle Pipistrellus murrayi A report to the Director of National Parks

Lindy Lumsden and Martin Schulz

2009

Arthur Rylah Institute for Environmental Research

Captive breeding and future in-situ management of the Christmas Island Pipistrelle Pipistrellus murrayi

A report to the Director of National Parks

Lindy Lumsden and Martin Schulz

Arthur Rylah Institute for Environmental Research Department of Sustainability and Environment 123 Brown Street, Heidelberg, Victoria 3084

2009

Arthur Rylah Institute for Environmental Research Department of Sustainability and Environment Heidelberg, Victoria

Report produced by: Arthur Rylah Institute for Environmental Research Department of Sustainability and Environment PO Box 137 Heidelberg, Victoria 3084 Phone (03) 9450 8600 Website: www.dse.vic.gov.au/ari

Citation: Lumsden, L. and Schulz, M. (2009). Captive breeding and future in-situ management of the Christmas Island Pipistrelle Pipistrellus murrayi. A report to the Director of National Parks. Arthur Rylah Institute. Department of Sustainability and Environment, Heidelberg, Victoria.

Disclaimer: This publication may be of assistance to you but the State of Victoria and its employees do not guarantee that the publication is without flaw of any kind or is wholly appropriate for your particular purposes and therefore disclaims all liability for any error, loss or other consequence which may arise from you relying on any information in this publication.

Front cover photo: Christmas Island Pipistrelle, and the only known communal roost for the species where in January 2009 there were four individuals of this species roosting (Photos: Lindy Lumsden).

Authorised by: Victorian Government, Melbourne.

Table of Contents Table of Contents...... 3 List of Tables...... 5 List of Figures...... 5 Acknowledgements ...... 6 1. Executive Summary ...... 7 1.1 The Imminent Extinction of the Christmas Island Pipistrelle...... 7 1.2 Causes of the Decline...... 7 1.3 The Urgent Need to Establish a Captive Breeding Program ...... 7 1.4 Objectives of the Captive Breeding Program...... 8 1.5 Off-shore Options for the Establishment of the Captive Breeding Facility...... 8 1.6 Recommendation to Establish the Captive Breeding Facility on Christmas Island 9 1.7 Captive Breeding Facilities ...... 10 1.8 Costings...... 10 1.9 Management and Monitoring Recommendations for the Remaining Wild Population ...... 11 2. Introduction and Background ...... 12 3 Assessment of status and population size in January 2009...... 16 3.1 Objectives...... 16 3.2 Attempts to trap bats in foraging areas...... 16 3.3 Observations at roosts...... 18 3.4 Attempts to locate new roosting areas ...... 22 3.5 Assessment of population size in January 2009 ...... 23 4. Captive management program...... 24 4.1 Review of captive breeding programs for small insectivorous bats...... 24 4.1.1 Bats in captivity...... 24 4.1.2 Enclosures...... 26 4.1.3 Feeding bats in captivity ...... 27 4.1.4 Breeding in captivity...... 29 4.1.5 Longevity ...... 30 4.1.6 Conclusion...... 30 4.2 Specific objectives of a captive breeding program ...... 32 4.3 Potential receiving facilities ...... 34 4.3.1 Singapore Zoological Gardens ...... 34 4.3.2 Territory Wildlife Park, Darwin ...... 34 4.3.3 Zoos Victoria...... 34 4.3.4 Taronga Zoo, Sydney ...... 34 4.3.5 Christmas Island...... 35 4.4 Quarantine issues...... 36 4.4.1 Singapore Zoo ...... 36 4.4.2 Australian Zoos/Wildlife Parks ...... 38 4.4.3 Christmas Island...... 39 4.5 Transport requirements ...... 41 4.6 Advantages and disadvantages of the various options ...... 42 4.6.1 Overall unknowns ...... 42 4.6.2 Advantages of taking the animals to an existing off-shore facility ...... 42

4.6.3 Risks and disadvantages of taking the animals off-shore...... 42 4.6.4 Advantages of establishing the facility on Christmas Island ...... 43 4.6.5 Disadvantages of establishing the facility on Christmas Island ...... 44 4.7 Legislative requirements...... 45 4.8 Biology and ecology of the Christmas Island Pipistrelle in the wild, relevant to a captive breeding program...... 46 4.8.1 Diet ...... 46 4.8.2 Flight pattern...... 47 4.8.3 Nightly activity patterns...... 47 4.8.4 Breeding biology...... 48 4.8.5 Longevity ...... 49 4.8.6 Social organisation ...... 49 4.8.7 Roosting habitat...... 50 4.8.8 Roosting behaviour...... 50 4.8.9 Morphometric data...... 51 4.8.10 Health condition...... 53 4.9 Captive management plan...... 55 4.9.1 Housing...... 55 4.9.2 Feeding requirements...... 58 4.9.3 Monitoring protocol ...... 59 4.10 Obtaining bats for the captive colony ...... 61 4.10.1 Recommended number of individuals to be taken into captivity...... 61 4.10.2 Timing of taking animals into captivity ...... 61 4.10.3 Techniques for catching animals to take into captivity...... 61 4.11 Funding costs and opportunities...... 63 4.12 Discussion on the feasibility and likelihood of success of a captive breeding program for the Christmas Island Pipistrelle...... 66 5 Review of existing monitoring and management programs...... 73 5.1 Stationary detector monitoring...... 73 5.2 Driving detector monitoring...... 74 5.3 Monitoring occupancy and numbers of bats in remaining roosts ...... 74 5.4 Predator-proof known roosts ...... 75 5.5 Location of potential roost sites ...... 77 5.6 Monitoring potential predators using infra-red cameras ...... 78 5.7 Artificial roosts ...... 78 5.8 Protect and enhance foraging habitat...... 80 5.9 Yellow Crazy Ant control ...... 82 5.10 Feral Cat control trial ...... 82 References...... 84

List of Tables Table 1. The size of flight enclosures used for successful captive colonies of small microbat species...... 27 Table 2. Summary of available dietary information for the Christmas Island Pipistrelle...... 46 Table 3. The weight and forearm length of individuals caught during various seasons and years...... 52 Table 4. The estimated cost of establishing and maintaining a captive breeding program for the Christmas Island Pipistrelle on Christmas Island for ten years...... 64

List of Figures Fig. 1. The trend of decline in the Christmas Island Pipistrelle from 1994 to 2008...... 12 Fig. 2. The decline in mean number of detector passes at 12 sites sampled regularly since 2006...... 13 Fig. 3. A harp trap set along the Winifred Beach Track at site L22 in January 2009...... 17 Fig. 4. A bat caught in a monofilament mist net...... 18 Fig. 5. The loose piece of bark on Roost 565 used as a roost site by Christmas Island Pipistrelles...... 21 Fig. 6. Roost tree 14 in 2005 (left) and 2009 (right)...... 22 Fig. 7. Hand-feeding a Lesser Long-eared Bat Nyctophilus geoffroyi on a mealworm...... 28 Fig. 8. A maternity roost located in 2005 under loose bark lifting off a dead Tristiropsis acutangula that was used by a colony of 32 Christmas Island Pipistrelles (Roost 14)...... 51 Fig. 9. The Parks Australia Research Station (the Pink House) in which the temporary facility could be established...... 56 Fig. 10. Fly-wire mesh shelter that could be used as temporary holding enclosures within a predator-proof room – this one is 3.5 m x 3.5 m at ground level and about 2 m high, with an enclosed floor...... 56 Fig. 11. The edge of the forest at the Pink House clearing where flight enclosures for the long-term facility could be built...... 57 Fig. 12. The protective sleeves that have been installed on roost trees and interconnecting trees...... 76 Fig. 13. A termite trail that formed on the protective sleeve on Roost 565 between 10 and 15 January 2009...... 77 Fig. 14. Artificial roost box set on a 6 m pole, that is coated with Tanglefoot...... 79 Fig. 15. One of the cleared tracks in Field 26j that provided optimal foraging habitat and areas at which to trap bats in 2005...... 81 Fig. 16. Overgrown tracks in Field 26j in 2009. These are no longer used as foraging habitat...... 81 Fig. 17. Exclusion zone for using toxic baits as part of the Feral Cat eradication trial...... 83

Acknowledgements

This project was funded by and conducted for the Director of National Parks.

LL’s visit to Christmas Island in January 2009 to reassess the status and population size of the pipistrelle was generously funded Christmas Island Phosphates.

A large number of people assisted with this project, and we thank them all.

• Marjorie Gant, Mike Misso, Michael Smith, Kent Retallick, Chris Boland from Christmas Island National Park (CINP) for their assistance, input and advice. • Dion Maple and Rebecca Reeves from CINP for their enthusiastic assistance and help during January 2009. • Kent Retallick and the Invasive Species Team, especially Dion Maple, Rebecca Reeves and Brendan Tiernan, for collecting the amazing datasets that have given us a greater understanding of the ecology, distribution and status of this species. • All the CINP staff for their interest, commitment and enthusiasm in trying to save the pipistrelle – it is great to work with such a dedicated group of people. • Peter Latch and Simon Nally, Threatened Species and Communities Section, Department of Environment, Water, Heritage and the Arts, for their input into discussions on the captive breeding program. • Graeme Gillespie, Russel Traher and Peter Courtney, Zoos Victoria, for discussions and input into the captive breeding component. • Rebecca Spindler, Research Manager, Taronga Zoo, Sydney for discussions on the captive breeding component. • Dion Wedd, Curator, Territory Wildlife Park, Darwin for discussions on the captive breeding component. • Charlene Yeong, Conservation and Research Officer, Singapore Zoological Gardens, for discussions on the captive breeding component. • Dr Kim Halpin and Dr Deborah Middleton, Australian Animal Health Laboratory, CSIRO, in Geelong Victoria for clarifying information on viruses. • Clare Jones, Biosecurity Australia who provided clarification of quarantine requirements. • Felicity Humann and Manieka Reid for compiling most of the material for the review of captive breeding programs. • Jane Sedgeley and Colin O’Donnell for providing unpublished information on captive programs in New Zealand. • David Middleton for discussions on captive breeding. • Greg Richards for facilitating LL’s visit to the island in January 2009, assisting with the field work and for many detailed discussions on how to solve the problem. • Kevin Edwards, Christmas Island Phosphates, for funding LL’s visit to the island in January 2009. • Jenny Nelson, ARI, for proof reading and commenting on a draft or this report. • Anne-Marie Delahunt, Parks Australia, for providing comments and suggestions on a draft of this report.

Field work undertaken during January 2009 was conducted under the Australian Government, Permit for an Activity in a Commonwealth Reserve Permit No. 12/08, and the Arthur Rylah Institute Animal Ethics Committee AEC 08/20.

All photographs were taken by Lindy Lumsden or Martin Schulz, except where indicated.

1. Executive Summary

1.1 The Imminent Extinction of the Christmas Island Pipistrelle The Christmas Island Pipistrelle is endemic to Christmas Island and is listed as Critically Endangered under the EPBC Act 1999. Its distribution and abundance has declined dramatically in recent years. It has undergone a 99% reduction in relative abundance since 1994, based on long-term monitoring using ultrasonic bat detectors. This long-term monitoring data predicts the species could go extinct by 2009. The current number of remaining individuals is unknown. However, based on recent observations at the only known roosting and foraging areas, it is possible that it is less than 20 individuals. Without urgent intervention there is an extremely high risk that this species will go extinct in the near future. While it can not be precisely predicted when this would occur if there was no intervention, it is highly likely that it will be within the next 6 months, i.e. by June 2009.

1.2 Causes of the Decline Despite targeted research during both the breeding and non-breeding seasons the cause/s of this rapid decline of the Christmas Island Pipistrelle remains unknown. It is possible that the difficulty in identifying the cause/s of the decline may be the result of deteriorating ecosystem health across the island and this decline is reflected in multiple synergistic stochastic effects coinciding over a short period of time. A number of potential threatening processes may be contributing to this decline, including various introduced species potentially preying on, or disturbing bats from within their roosts, such as the Common Wolf Snake Lycodon aulicus capucinus, Giant Centipede Scolapendra morsitans, Yellow Crazy Ant Anoplolepis gracilipes, Black Rat Rattus rattus and feral Cat Felis catus. Other threats may include the loss of roost sites due to the collapse of the dead trees used as roosts, reduced prey availability (e.g. as a result of the expansion in distribution of some tramp invertebrate species, such as the Yellow Crazy Ant), and disease. There may also be additional, as yet unidentified, key threatening processes.

1.3 The Urgent Need to Establish a Captive Breeding Program Given the imminent risk of extinction of this species, there is an urgent need to take into captivity all remaining individuals to establish a captive breeding program. The aim of this program would be to provide insurance against the species disappearing totally and to be a source of animals to reintroduce to the wild once the threatening processes were identified and remedial actions undertaken.

It is imperative that the establishment of a captive breeding colony is undertaken as soon as absolutely possible. The greater the time taken to commence the program the less likely it will be to succeed due to: a) the ever decreasing number of individuals remaining, increasing the risk of not being able to catch sufficient bats to form the captive colony; and b) the fewer individuals used to establish the colony the greater the risk of inbreeding depression thereby threatening the medium to long-term viability of the species. It is essential that a captive breeding colony is commenced within the next 3 months (i.e. by March-April 2009).

While it is going to take a significant commitment to rectify this dire situation, the alternative is a “do nothing” approach, which will inevitably result in the extinction of this species. If this occurs it will be the first bat in Australia to become extinct since European settlement and the fourth mammal to disappear from Christmas Island after the Bulldog Rat Rattus nativitatis and Maclear’s Rat R. macleari which disappeared in the early 1900s, and the possible extinction of the Christmas Island Shrew Crocidura attenuata trichura in the 1980s.

1.4 Objectives of the Captive Breeding Program The overall aim of the captive breeding program is to reduce the risk of the Christmas Island Pipistrelle going extinct, by breeding animals in captivity while the cause/s of the decline in the wild is identified and rectified, enabling the release of captive-bred animals to successfully re- establish a wild, self-sustaining population. Due to the current very low number of individuals, it will be necessary to maintain the captive breeding program for 10 years to be able to build the numbers up sufficiently for a release program. During this time actions will be required to identify and rectify the cause of the decline, and animals can be experimentally released to assist this process.

Although the Christmas Island Pipistrelle has not previously been kept in captivity, many other species having, including other species of pipistrelles. There have been no captive breeding programs specifically for conservation purposes to return viable populations to the wild, however, many species have been kept for the purpose of rehabilitation, research or experimentation. As a result there is a wealth of experience on the captive care and husbandry of bats that can be used to guide the Christmas Island Pipistrelle captive breeding program.

The specific objectives of the captive breeding program are as follows. • To successfully trap as many as possible of the remaining individuals in the wild and bring them into captivity. • To develop captive husbandry techniques specifically for the Christmas Island Pipistrelle to a sufficient standard that it is possible to maintain and breed animals in captivity. • To successfully breed Christmas Island Pipistrelles in captivity such that the overall number of individuals held in captivity increases, and that captive-born young breed themselves. • To ensure captive-bred Christmas Island Pipistrelles behave in a manner similar to in the wild, including forming appropriate social groups and in their ability to catch insects in flight. • To undertake genetic analysis of all individuals taken into captivity, and those born in captivity, to facilitate breeding aggregations that ensure maximum possible genetic diversity in captive-bred Christmas Island Pipistrelles. • To use the captive animals to assist in determining the cause/s of the decline, especially in relation to disease issues, by continually assessing the health of individuals and undertaking detailed post mortems of any individuals that die. • To investigate the possibility of using the captive colony to experimentally test alternate theories of the cause of decline, without jeopardising the health of the animals. • To determine roosting box preferences by trialling a number of box designs in captivity that can then be used to supplement roost sites in the wild if required. • To successfully release animals into the wild once the cause of the decline has been rectified, using an experimental, adaptive management approach. • To contribute to restoring the population in the wild to its pre-decline levels.

1.5 Off-shore Options for the Establishment of the Captive Breeding Facility The captive breeding facility could either be located on Christmas Island or at an existing off- shore wildlife facility. Four potential facilities were identified in this study: Singapore Zoological Gardens; Territory Wildlife Park, Darwin; Zoos Victoria, Melbourne; and Taronga Zoo, Sydney.

This study concludes that establishing a captive colony off-shore is not the preferred option due to:

a) The risk of introducing disease into Australia or overseas, if some unknown disease is contributing to the decline of the species on Christmas Island. b) There would be stringent quarantine requirements to take the bats to either the Australian mainland or Singapore. For the Australian mainland, an Import Risk Assessment would need to be undertaken by Biosecurity Australia to assess the risks and recommend mitigating measures. Typically this would require testing of the bats for Nipah virus and Australian Bat Lyssavirus. However, it would not be possible to extract the quantity of blood required to undertake these tests from this small species (weighing only 3 g), without seriously compromising the health of the animals. It may be possible that bats could be taken to a Quarantine Approved Premise in Australia under a ‘closed system’ regime without being tested, i.e. they would remain under strict quarantine control until re-exported to Christmas Island. However, until an IRA is completed this could not be confirmed. c) There is likely to be a significant time delay before animals could be transferred to a facility on the mainland due to the time required to undertake the IRA (possibly up to 3 years). If it was decided to wait until an IRA was completed, it is highly unlikely any animals would remain in the wild. Alternatively, animals could be taken into captivity on the island immediately, while there are still animals remaining, and held there while the IRA is completed. Facilities would thus need to be built on the island so it would be more cost effective and efficient to retain the program there. d) Due to the island’s remoteness, transportation to an off-shore facility is problematic. The cheapest option would be to use a commercial flight to the mainland. However, there would need to be a stopover in Perth, requiring additional quarantine procedures to be met before proceeding to an interstate facility. An alternative would be to use a direct military or Customs flight. However, arrangements for such an option would require high-level negotiations between the relative Commonwealth departments. e) It is unknown how well the bats will tolerate the extensive transportation period to the mainland. It is possible that there may be fatalities during this phase. A small number of individuals would need to be taken first to test their ability to be transported. Therefore multiple flights would be required, increasing the costs.

Singapore Zoo was initially interested in being considered as a potential location for the captive colony. However, after discussions on the quarantine issues and the requirements for the colony, they decided they did not have the necessary expertise and were unable to provide an ex-situ breeding site for the Christmas Island Pipistrelle.

Taronga Zoo, while interested in the problem, decided they would be unable to provide the captive breeding facilities. Zoos Victoria and Territory Wildlife Park are still interested in being involved. Territory Wildlife Park is not a registered Quarantine Approved Premise and would require additional funding to bring their facilities up to this standard. Discussions have been undertaken with Zoos Victoria and they are interested in providing captive breeding and veterinarian expertise and advice if the facility was to be established on Christmas Island.

1.6 Recommendation to Establish the Captive Breeding Facility on Christmas Island The recommendation from this study is that the captive breeding facility should be established on Christmas Island.

There are a large number of advantages in establishing the colony on the island: the program will be able to commence much sooner; there will be no quarantine requirements for the bats; individuals will not be exposed to the stress of travel and acclimatisation in a distant facility; climatic conditions are optimal for housing and to promote breeding; the captive colony could contribute to determining the cause of the decline in the wild; artificial food could be supplemented with natural food (i.e. flying insects); captive-born young could be acclimatised with natural prey and hence have a greater chance of survival on release; once the cause of

the decline has been mitigated, individuals could be released progressively and intensively monitored before the remainder were released; and it would facilitate involvement of the local community in the project.

To establish a captive breeding program on Christmas Island however, will take a significant commitment from government and other organisations. A new facility will need to be constructed and sufficient funds committed to establish the facility and the colony, including employing full-time trained personnel and expert visiting bat biologists and wildlife veterinarians. As there are no veterinarians on the island, provision would need to be made to bring an experienced veterinarian over as required (they would not be required permanently but would need to be available at short notice). The captive breeding facility could be multi- focused, also catering for the critically threatened reptile species on the island, thus sharing the costs between a number of species.

Due to the critically low number of individuals remaining in the wild and the uncertainty around being able to catch sufficient bats to start an effective captive breeding program, it is recommended that the program is undertaken in two phases: a) the initial emergency rescue and establishment of a temporary facility; and b) if sufficient animals are collected, the establishment and on-going maintenance of the longer term facility. While it will be important to commence the planning for the longer term facility now, the rescue phase needs to be undertaken immediately while animals remain in the wild. This component can not wait until the planning, funding and building of the long-term facility is complete.

1.7 Captive Breeding Facilities Temporary Holding Facility: Due to the urgency of the situation a temporary holding facility will need to be established by March-April 2009 to guard against the bat either going extinct or declining to such low numbers that it is extremely difficult (and more expensive) to capture individuals. It is proposed that a number of ‘fly-wire mesh tents’ (3.5 x 3.5 m) are set within an existing building or facility that is predator (e.g. rat, cat, wolf snake and centipede) and crab proof. This building would need to be secure and have appropriate temperature, humidity and light regimes. These tent enclosures are considered too small to provide appropriate flight space, capacity for social organisation and breeding in the long term, but would be more than adequate to be used as a temporary measure for a couple of months. One of the large rooms in the Parks Australia Research Station in the centre of the island (the ‘Pink House’) would be an ideal location for the temporary facility.

Longer term Facilities: The most important issues with respect to the longer term facility are: a) assured funding to establish and maintain the captive breeding program for the 10 year duration; b) the employment of a project manager to oversee the building of the facility; c) deciding on a location for the facility (within or external to the park) – the cleared area at the Pink House is recommended; d) design of the flight enclosures and working spaces, ensuring they are secure, predator-proof, have power and water, and meet temperature and humidity requirements; e) employment of two staff dedicated to maintaining the bats and running the facility; and f) establishing formal partnerships with an existing wildlife facility and bat biologists to facilitate rapid access to such expertise as required.

1.8 Costings The costings of the captive breeding program cannot be accurately determined until the location, the facility design and staffing requirements are decided upon. However, outlined below are rough estimates of the level of funding required. These estimates assume there is no requirement for land purchase, and that the long-term facility was built in the clearing at the Pink House. They are based on the facility operating for 10 years (and hence the costs are for the full 10 year period). Included in this costing is targeted research to determine the cause/s of the decline, as this is essential before captive animals can be released. Note that

in addition to what is costed below, additional funds will be needed to mitigate the cause/s of the decline once it is determined. However, it is not possible to estimate what this may cost until these factor/s are identified.

Item Cost $K

Emergency rescue phase Trap individuals to take into captivity and establish captive colony (4 bat biologists, 1 260 veterinarian to capture animals, establish temporary holding facility, acclimatise animals to captivity, assess health)

Long-term phase Project manager to oversee the construction of the facility (1 person for 3 months, plus on- 30 costs and operating) Construction of long-term captive breeding facility 800 Ongoing maintenance of captive breeding colony (2 keeping staff, periodic visits by bat 3,000 biologists and veterinarian, maintenance of food supply, plus operating costs and transport to and from the island, @ $300K/year for 10 years) Targeted research to determine cause of the decline 750 Total (to establish and operate the facility for 10 years) $4,930

Potential funding opportunities in addition to the Commonwealth Government should be investigated including Christmas Island Phosphates, philanthropic sponsorship from both overseas and Australia, corporate sponsorship and through tourism to the island.

1.9 Management and Monitoring Recommendations for the Remaining Wild Population While issues around the captive breeding program are being worked through it is important that a number of on-ground actions be undertaken as a matter of urgency. These will also assist in locating bats for the captive program. These include: • Continuing the regular detector sampling at the standard sites in the west of the island to document the changing status. • Continually monitor the only known communal roost using bat detector surveys and undertake weekly emergence counts to determine the number of individuals. • Locate the alternative roost used by this colony of bats by undertaking detector monitoring at potential roost trees nearby. • Regularly maintain the predator-proof barrier on the known roost. • Locate potential roosts in the western end of the island and set detectors under a sample of these. • Facilitate the attachment of predator-proof barriers on all potential roost trees within 200 m of the currently used tree. • Ensure the key roosting and foraging areas remain free of Yellow Crazy Ant supercolonies.

2. Introduction and Background

The Christmas Island Pipistrelle is endemic to Christmas Island, and is the only species of insectivorous bat on the island (Lumsden et al. 2007). Its distribution and abundance has declined dramatically in recent years. Surveys undertaken in the mid-1980s found the species to be widespread and common across the island (Tidemann 1985). However, by the mid- 1990s there had been a marked reduction in abundance and a westward range contraction (Lumsden and Cherry 1997, Lumsden et al. 1999). This decline has continued at a rapid rate and the species is now confined to the far west of the island, no longer occurring across most of its former range (Lumsden et al. 2007, DNP 2008). Long-term monitoring using ultrasonic bat detectors indicates this species has undergone a 99% decline in relative abundance since 1994 (Lumsden et al. 2007, DNP 2008, Christmas Island National Park [CINP] unpublished data).

This data suggests that, if the current rate of decline continues, this species could be extinct by 2009 (Fig. 1). As the only nocturnal insectivore on the island this could have serious consequences for the overall ecological balance, and the ecosystem service role the pipistrelle fulfils, in consuming large quantities of insects, would be lost. This decline is occurring extremely rapidly and the rate of decline is not slowing. Since 2006, CINP has undertaken monthly detector recordings at 12 sites in the west of the island, which includes the majority of the sites where this species can still be recorded. In January 2006, the mean number of passes per night at these sites was 168.3 ± 108.6. In September 2008 there were only 0.96 ± 0.83 passes/night – a decline of 99.4% in less than 3 years (Fig. 2). A reassessment of the population size was made in January 2009. While the precise number of individuals could not be determined, only four individuals were known to be alive and it is believe that the total population size could have been as low as 20 individuals.

100

90 80

on 70 i at 60

50 40 1994 popul of 30 % 20

10 0 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010

Fig. 1. The trend of decline in the Christmas Island Pipistrelle from 1994 to 2008. Data is from Lumsden and Cherry 1997, Lumsden et al. 1999, Corbett et al. 2003, DNP 2008, CINP unpublished data. Note that the species was already in decline in 1994. These results are based on repeated sampling using bat detectors at set sites across the island up to 2006. Since 2006 the detector monitoring has focused just on the west of the island in areas where the species still occurs. The 2007 and 2008 figures are based on the relationship between this data and that of 2006 at these same sites, for comparative purposes.

300

250 e t i s / t h g i 200 n es/ pass

r 150 o ct e t e d . 100 ean no M 50

0 l t l t r ry y ri y y r r ri y y r r e r y y g r v rch ne ul rch ne ul b e ar r u e ua ar Ma J gus be be Ma J gus be be b a A b /No n ru Ap Ju u o Ap Ju u to m m nu ru y/ m t b Ma A em ct Ma A em a b l te Ja pt O pt Oc ve ce J Ju p Oc Fe e e o e Fe e S S N D S

2006 2007 2008

Fig. 2. The decline in mean number of detector passes at 12 sites sampled regularly since 2006. In January 2006 there was a mean of 168.3 ± 108.6 passes/night at these sites. In September 2008 this had crashed to just 0.96 ± 0.83 passes/night – a decline of 99.4%, although sampling in October/November showed a slight increase. Data from CINP. Note: a ‘detector pass’ is one individual flying past the detector once.

Due to these rapid declines, the Christmas Island Pipistrelle was listed as Critically Endangered under the Environment Protection and Biodiversity Conservation Act 1999 in 2006. However, the cause of this rapid decline remains unknown. A number of potential threatening processes have been identified (Schulz and Lumsden 2004, Lumsden et al. 2007, DNP 2008). Extensive areas of habitat are available, with 75% of the island covered by primary or secondary rainforest. The pipistrelle is a generalist aerial insectivore and there appears to be an abundance of nocturnal flying insects. Predation or disturbance at roost sites has been considered one of the most likely threats to the survival of the species (Schulz and Lumsden 2004, Lumsden et al. 2007). This species roosts predominantly under exfoliating bark on dead trees, many of which are heavily decayed and currently collapsing (Lumsden et al. 2007, DNP 2008). The location of only one communal roost tree is known, however, it is likely that this colony also uses a second roost site. When population sizes were larger, females formed colonies of up to approximately 50 individuals, with males often roosting solitarily (Lumsden et al. 1999, 2007). The only known colony in 2009 contained just four individuals. A number of introduced species may be preying on, or disturbing bats from within their roosts, e.g. Common Wolf Snake Lycodon aulicus capucinus, Feral Cat Felis catus, Black Rat Rattus rattus and Giant Centipede Scolapendra morsitans. Giant Centipedes and Black Rats have been recorded commonly on roost-type trees. Although not considered the primary cause of decline, the dramatic increase in numbers of Yellow Crazy Ants Anoplolepis gracilipes is likely to have exacerbated the situation. It is also possible that some

form of disease may be contributing to the decline, and to test this, biological information was collected from 52 individuals trapped in 2005 (Lumsden et al. 2007). All appeared in good condition and the majority of females were breeding. All biological samples were normal, with the exception of low white blood cell counts, and possible regenerative anaemia. However, the significance of these findings are unclear, as it is not known if these parameters are typical for this species or represent ill-health.

Recommendations for future management of the Christmas Island Pipistrelle are provided in Lumsden et al. (2007) and DNP (2008). These reports identify the establishment of a captive breeding colony of Christmas Island Pipistrelles as an extremely high priority, in order to provide: 1) insurance against further decline in numbers; and 2) a source of animals to re- establish wild populations once the cause of decline has been identified and rectified (Lumsden et al. 2007). If the predictions indicating the species could be become extinct in 2009 are correct, it is obvious that a captive breeding colony needs to be established with great urgency. With so few individuals left (as of January 2009), it is imperative that this occurs by March-April 2009.

The proposed establishment of a captive breeding program for the Christmas Island Pipistrelle poses a number of challenges, due to unknowns regarding how many individuals remain, how this species will adapt to captivity and the remote location of Christmas Island resulting in logistical difficulties associated with transport and quarantine if an off-shore location was selected. This project was initiated to address these issues and to document the state of knowledge of captive husbandry for small insectivorous bats, summarise the biology of the Christmas Island Pipistrelle relevant to keeping them in captivity, and identify processes required to establish a captive breeding program.

Christmas Island National Park have been undertaking a wide range of management and monitoring activities related to the Christmas Island Pipistrelle over the last six years, including conducting extensive detector surveys throughout the island and locating the remaining foraging areas, monitoring roost sites, installing protective barriers around roost trees, locating potential roost sites, monitoring potential predators at roost trees, establishing artificial roost sites, protecting foraging habitat and controlling Yellow Crazy Ants. The detailed monitoring actions have been critical in documenting the rapid decline in recent years. To enable the available resources to be most effectively used to protect the remaining individuals and monitor the trends in the population, these activities will be reviewed to identify the highest priority actions.

The general objectives of this project are to: • as far as possible investigate and document the options, requirements and timelines for establishing a captive breeding program for the Christmas Island Pipistrelle; and • make recommendations for future in-situ management of the species.

The specific objectives of the project are to: • undertake a review of captive breeding programs for small insectivorous bats elsewhere in the world; • clarify the specific objectives of a captive breeding program and identify performance criteria for each objective; • identify potential wildlife receiving facilities, including determining their level of expertise and potential long-term commitment to a partnership for a captive breeding program and the feasibility of an on-island facility; • document the steps needed to establish a captive breeding program, including actions that would be required on Christmas Island and at the receiving facility; • investigate all quarantine requirements;

• document transport requirements and options of moving the animals from Christmas Island to the new facility if an off-shore location was recommended; • outline the research required to determine if there are sufficient animals in the wild population to commence a captive breeding program; and to obtain an estimate of the current population size. Make recommendations as to the optimal number of animals to be taken into captivity and the relative merits of taking all animals that can be caught versus trying to retain a ‘viable’ population on the island; • compile as much information as possible about the species to progress a captive breeding program, including information on the ecology of the species, and any other information useful for establishing husbandry protocols and ensuring sound management of genetic variability in a captive population; • propose a timeline and the steps required for the implementation of a captive breeding program, including referral under the EPBC Act as a matter of national environmental significance, meeting quarantine requirements (including export permits if the offshore option is selected), and investigation of a Cooperative Conservation Program with the selected facility. • make recommendations as to the preferred option for the location of the captive facility; • review the existing monitoring and management programs on the island and make recommendations for future in-situ monitoring procedures and management actions for the species.

As this report was nearing completion it became apparent that many aspects of the planning for the captive breeding program could not be resolved until a reassessment of the current size of the population was undertaken. As a result a visit was made to the island in January 2009 by LL with the aim of clarifying a range of issues and to have detailed discussions on the options as to how to proceed. The results from the field component undertaken during this visit are reported below.

3 Assessment of status and population size in January 2009

3.1 Objectives The status and population size of the Christmas Island Pipistrelle was reassessed during a 2 week visit by LL to the island in January 2009. It was undertaken to provide a clearer indication of the current situation and to assist in the assessment of the feasibility and likelihood of success of the captive breeding program. The specific aims of this trip were to: • estimate the size and status of the remaining population by undertaking observations at foraging and roosting areas; • determine how readily individuals could be caught in foraging areas; • investigate how readily the species responds to captivity (if any bats were caught), by retaining a small number of individuals in captivity for several days; • discuss in detail the programs and data collected by CINP to assist with plans for the captive breeding program and the review of the current management and monitoring actions; and • progress the planning for the captive breeding program.

The field components are summarised below. Other information is incorporated into the remainder of the report.

3.2 Attempts to trap bats in foraging areas Trapping was undertaken at the start of the Winifred Beach Track and on nearby tracks. Two techniques were used: harp traps and very fine monofilament mist nets (Ecotone, Poland). Harp traps were set along narrow tracks in the vicinity of the standard detector sampling sites L22, C03 and D03 (Fig. 3). Mistnets were set in two locations along the Winifred Beach Track, immediately either side of the gate (in the section near L22 where there is only a narrow gap in the vegetation, and to the north of the gate where there is a wider clearing along the road). Mist nets were monitored continually. Harp traps were checked throughout the night and closed at 0400 hrs to allow any breeding females that were trapped sufficient time before dawn to return to their young in the day roost. The intention was to hold any males or non-breeding females for several days to determine how readily they would accept captivity and being hand fed.

Trapping was undertaken on five nights, between 10 and 16 January 2009, with a total of 9 harp trap nights and 38 mist-net hours. In addition, MS undertook 15 harp trap nights over five consecutive nights between 4 and 8 June 2008. This trapping comprised 13 harp trap nights adjacent to the start of Winfred Beach Track monitoring site (L22), and two harp trap nights in the least overgrown section of track in the secondary regrowth to the west of this area, where bats were trapped successfully in 2005 (Lumsden et al. 2007). No pipistrelles were caught in either June 2008 or January 2009, despite suitable trapping conditions and bats recorded on the bat detectors at these sites during the period the traps/nets were open. In contrast, in 2005, a total of 52 individuals were caught in 33 harp trap nights in this area.

To assess the likelihood of trap success, LL and Greg Richards sat beside the traps and nets and observed the behaviour of the bats through night vision goggles (American Technologies Network Corp., ATN NVG-7). It appeared that at any one time only a single individual was present, and that although there were a number of passes on the detector for a night (i.e. a number of times when a bat flew past the detector), this represented only a small number of individuals – probably less than four. It was also apparent that the bats were able to detect

Fig. 3. A harp trap set along the Winifred Beach Track at site L22 in January 2009. both the traps and mist nets and avoid them. One individual was observed to fly up to a harp trap on multiple occasions and then fly off. One harp trap was set within 5 m, and hence the detection zone, of the detector at L22. On 12 January 2009, a total of 40 passes were recorded on this detector and so this individual (or individuals) is likely to have approached the trap up to 40 times and not been caught. It had been hoped that the very fine monofilament mist nets might have been more effective (Fig. 4). However, the bats also appeared to readily detect and avoid these nets. These traps and nets are highly effective in catching small bats elsewhere, and the traps have been successful in catching the Christmas Island Pipistrelle in the past (Lumsden and Cherry 1997, Lumsden et al. 1998, 2007). There are two possible reasons why no bats were caught in January 2009. It is estimated that harp traps often catch only 5-10% of the bats that approach the trap (Dobson 1999). Despite this it is possible to catch large numbers of bats in these traps (for example, 50-100 individuals are regularly caught in a night in a trap in surveys in Victoria, LL pers. obs.). When large numbers are caught it is apparent that there are even larger numbers of individuals flying in the area. However, with the current situation on Christmas Island, if only one or two individuals are foraging in this area, they might be behaving like the 90% of individuals in other areas that do not get caught. Alternatively, they might be especially proficient in detecting and avoiding the traps and nets. The forests on Christmas Island have a large number of orb-weaving spiders that produce strong, sticky webs the bats could get caught in if they did not detect and avoid them. So it is plausible they are very used to detecting fine, net type structures. They also appeared to learn where the traps and nets were and then continued to avoid them on subsequent nights. In situations where there are large numbers of bats there is frequently a turnover of individuals so that there are always new individuals to encounter the traps for the first time. However, on Christmas Island it is likely that it is the same few individuals returning to this foraging area every night and so once they detected the traps/nets they either avoided the area or just avoided the traps/nets within the area. The number of passes recorded on the

bat detector decreased markedly in the nights following trapping events. In the fortnight preceding the commencement of trapping there was an average of 24 passes per night on the detector at L22. In the four nights following the first night of trapping this declined to 3.5 passes per night. On one of these nights a detector set only 100 m away in the more open section of the track recorded 80 passes (Greg Richards pers. comm.), suggesting the bats had moved a short distance away to avoid the area where the traps/nets were set.

Fig. 4. A bat caught in a monofilament mist net. (Note this photo is of a similar sized bat, the Little Forest Bat Vespadelus vulturnus; Photo Chris Tzaros).

These trapping results indicate that the number of bats has dropped to such a low level that it is now very difficult to capture bats at their foraging sites. The start of the Winifred Beach Track has been a foraging hotspot for many years (at least since 1998) and, together with the small tracks through the adjacent secondary regrowth, has always been the most reliable location for trapping bats. As the species has declined, this area has remained the focal point for monitoring and catching. Extensive detector surveys across the island undertaken by CINP in recent years have not revealed any equivalent locations. In 2005, many individuals were trapped along small tracks through secondary regrowth adjacent to the start of the Winifred Beach Track. These tracks have now overgrown such that they no longer provide suitable foraging or catching sites. It is recommended that these tracks be reopened (see Management section) in an attempt to improve the foraging habitat in these areas and hence provide addition trapping locations. However, unless and until this is successful, the start of the Winifred Beach Track is the only location that there is the potential to trap bats in their foraging areas. It is apparent that to obtain sufficient bats to form a captive population, alternate capture strategies with be required, such as catching individuals at roosts and modifying the type of trapping that is undertaken in the foraging areas.

3.3 Observations at roosts Of the nine communal roosts located in 2005/06 (Hoye 2006; Lumsden et al. 2007), six have collapsed, two appear to have been deserted (see below), with only one currently occupied (Roost 565). Three exit counts were conducted at this roost to determine how many individuals were occupying it. These were undertaken on 6, 10 and 17 January 2009. On the

first two occasions, four individuals were observed to leave the roost. Most of the bats flew from the roost and immediately left the vicinity, resulting in just a single pass on the detector for each individual. Occasionally individuals would circle the roost once or twice before leaving. The four individuals exited the roost within a 4-5 minute period, commencing at approximately 1845 hrs. On 6 January, observations continued at the roost for a further 3 hours using night vision goggles. One individual returned to the vicinity of the roost at 1930 hrs and foraged in the area until 2000 hrs when a second individual joined it. These two individuals then commenced flying up to the roost and off again, as if checking the roost for safety or to see if other individuals were in the roost (this is a typical behavioural pattern of tree-hole roosting bats). These two bats entered the roost at 2015 hrs and two minutes later another two individuals flew within range and also entered the roost. These four bats remained in the roost for approximately one hour and then all four left the roost between 2115 and 2130 hrs. The interpretation of this behaviour was that these four individuals were lactating females that foraged for approximately 1.5 hrs after leaving the roost and then returned to the roost to suckle their young for an hour. In 2005, females gave birth to their young in mid-December (Lumsden et al. 2007) and so if a similar pattern was followed the 2008/09 season, it would be expected that the young would still be dependent in the roost in early-mid January. (More details are provided on the breeding biology later in the report). This is a typical pattern shown by other species of insectivorous bats where the young are left in the roost while the female forages. The female then returns regularly throughout the night (in roughly 3 hr intervals – 2 hrs of foraging away from the roost and 1 hr in the roost suckling the young) (Lumsden 2004).

If this assumption is correct, it suggests that there are four adult females present in this roost with four dependent young. In addition, it indicates that although very few individuals appear to remain, these individuals are successfully breeding.

CINP have set a bat detector at this roost regularly since November 2006, with data from 288 nights. This extensive dataset enables a detailed examination of the use of this roost. These data have been summarised into the number of detector passes recorded in 2-hr blocks throughout the night, revealing several interesting patterns. Firstly, there are high levels of activity around the roost throughout the night, with the number of passes typically greater during the night and especially prior to dawn, than on dusk. The observations at the roost help to explain this pattern. After exiting the roost the bats leave the area immediately, or circle the roost just once, and hence each individual is recording only one, or a few, passes. However, as the bats return to the roost they circle around the roost multiple times, either foraging or checking the roost before entering it. Therefore at these times, individual bats are recording a large number of passes. Secondly, during the breeding season it is not surprising that the bats are returning to the roost so that the females can suckle their young. However, outside the breeding season (i.e. in all months except December and January), it was expected that the bats would not return as frequently to their roosts during the night, remaining for greater periods in foraging areas. However, the pattern of high levels of activity near the roost occurred in all months of sampling. Thirdly, it is clear from these data that the bats are present at this roost on a high proportion of days. This is atypical for bats that roost under bark. Typically species that roost under bark or in tree hollows have a number of roost sites that they rotate between, often staying in a roost for only a single day before moving to another roost (Lewis 1995). They do this even when they have dependent young in the roost, carrying their young in flight to the new roost (Lumsden 2004). On examination of the 288 nights of data from Roost 565, it appears that bats were continually using this roost for considerable periods of time. They appeared to be present virtually every day of sampling (over 200 nights) between November 2006 and January 2008. For 28 days in February 2008 it appears they were not using this roost, as there were no calls on dusk, but that the roost they were using was close by, as there were still numerous calls throughout the night and on dawn. Unfortunately it is not known how long they remained at this alternate roost as a branch fell on the detector and there were additional equipment problems for many months,

which precluded sampling. When sampling resumed in September 2008 bats were again using this roost, however, during the October sampling period they were not, while in December they appeared to interchange between roosts. This pattern continued into January 2009 with bats present from 1-10 January, at an alternate roost for 2 days, possibly back for the next day and then away again for the next 3 days. This was confirmed by a roost emergence watch on 17 January when no bats left the roost during the observation period of 1830 to 2000 hrs, but bats were recorded on the detector in the vicinity from 2230 hrs onwards. The trigger for bats to move roosts is not known. Interestingly however, sometime between 10 and 15 January a termite trail formed up the roost tree which came close to the piece of bark where the bats roost (more detail is provided in the management section). This corresponded with when the bats shifted roost but it is not known if there was any link between the two events. This termite trail could have deterred the bats from using the roost, or alternatively allowed a potential predator to breach the barrier at the base of the tree and gain access to the roost.

It is unclear if the pattern of using the same roost tree for long periods of time is typical of this species or a response to the threat causing the decline. The radiotracking data collected in 1998 (Lumsden et al. 1999), 2005 (Lumsden et al. 2007) and 2006 (Hoye 2006) suggests that individuals do use multiple roosts with some evidence of switching between adjacent roosts. Is it possible that the remaining bats have learnt that Roost 565 is a safe tree (because of the predator-proof barrier around its base) and are therefore remaining in this tree for long periods. However, if this was the case why is this colony still declining? There were at least 54 individuals in this colony in December 2005 (at that time roosting in the adjacent Roost 13 which has since fallen). It is not possible from the detector data to determine the number of individuals using the roost. The number of detector passes is highly variable between nights, mostly likely due to the number of times individuals circle around the roost, rather than variation in the number of individuals actually using the roost. It does appear, however, that the number of passes may have decreased since 2006. If predation from roosts is impacting on the species, then it is possible that this occur when the bats are roosting at an alternate roost tree, which is not protected by a predator-proof barrier. There is a great urgency to find the alternate roost/s used by this colony and protect it/them.

One pattern is very clear from the detector data – it is possible to use detectors to determine if a tree is a roost tree. The high levels of activity throughout the night and prior to dawn, even when the bats are using a different roost nearby, indicates they continually check the main roost throughout the night. Therefore, placing detectors below a potential roost tree is a good technique for ascertaining if that tree is used. Calls need to be checked closely, however, to distinguish between a tree that is being used as a roost during the day and one being visited during the night. If there are calls on dusk it is likely the roost is occupied. However, a roost emergence watch would be required to confirm occupancy by bats, and to determine the number of individuals using the roost. A high number of calls in the half hour prior to dawn is not a good indication of whether the roost is going to be used that day, as all nights show high levels of activity during this period, irrespective of whether the roost is used that day or not.

The bark on Roost 565 is currently precariously attached (Fig. 5). This tree had considerably more bark on it when located in 2006 (K. Retallick pers. comm.), most of which has fallen off. It is conceivable that this bark – especially the upper piece which is where the bats roost, could fall off at any time. When it does fall off the bats will be forced to move to a new roost site. It is not know which of the available trees nearby will be selected, so it is important that all nearby potential roost trees are protected with a predator-proof barrier, so that irrespective of which tree they select, it is protected (see management section).

Fig. 5. The loose piece of bark on Roost 565 used as a roost site by Christmas Island Pipistrelles. The arrow indicates where the bats emerge.

The two communal roost trees of the nine that were located in 2005/06 (Hoye 2006; Lumsden et al. 2007), that appear to have been deserted, were re-examined to determine if they were still suitable as roosts and if they were currently occupied. These roosts (Roost 14 and 686) are close together and were used by the one colony of bats, in conjunction with Roost 21 that was 10 m from Roost 14. Roost 21 collapsed in May 2007. Thirty to 40 individuals used this cluster of roosts in 2005 (Lumsden et al. 2007). CINP detector data from these roosts indicate that Roost 14 was used throughout 2006 and 2007, until at least February 2008. A detector was set at both roosts on 7 January 2009 for two nights and a roost emergence watch was undertaken at Roost 14. No bats were observed leaving the roost and there were no calls recorded on the detectors at either roost. Based on the observations at the currently used roost tree, where bats are present almost every day and the detector at the base of the roost tree records bats in the vicinity even if not roosting in the tree, it appears that Roosts 14 and 686 are no longer being used and that this cluster of 30-40 bats has died or moved roosting areas. Although the top of the tree where Roost 14 was located, has collapsed sometime since it was first located (probably in 2006), the bark under which the bats were roosting is still present (Fig. 6).

Fig. 6. Roost tree 14 in 2005 (left) and 2009 (right). The top of the tree (above the fork) has fallen off, however the area of loose bark below this fork, where the bats were roosting in 2005, remains.

In 2006, seven artificial roosts were established to provide additional roosting opportunities due to the concern over the high rate of roost tree collapse. Further detail on these is provided in section 5.7. Five of six remaining boxes were checked in January 2009 (one box was destroyed in a tree fall), by using a spotlight to illuminate the internal cavities and checking with binoculars. No bats were present and there was no evidence that they had been used (e.g. staining or droppings on the walls).

3.4 Attempts to locate new roosting areas To supplement the work being undertaken by CINP aimed at locating new roost sites (see management section), several transects were walked using a bat detector during the night. Assuming bats roosting at other roosts are behaving the same as those at Roost 565, there should be considerable activity around the roost, especially during the pre-dawn period. On 10 January from 0430 – 0510 hrs (i.e. the time the bats are typically circling around the roost) a 2.3 km transect was walked along the Winifred Beach Track from the gate near the start of the track to past the Circuit Track. No bats were detected. On 12 January a 500 m transect was walked along the Dales Track at 0500 hrs. Again, no bats were detected. In addition, on 9 January a transect was walked along the Winifred Beach Track from the start at L22 to the Sydney Dale carpark and back between 1930 and 2030 hrs. No bats were detected.

3.5 Assessment of population size in January 2009 While it is not possible to accurately assess the remaining population size, the impression is that it is very low. The two counts of four individuals in Roost 565 suggests that this may be all the individuals that use this roost. Radiotracking work in 2005/06 indicated that individuals that roosted in Roost 565 (and the adjacent Roost 13), foraged at the start of the Winifred Beach Track (at L22). Therefore it is quite possible that the individuals that were observed at the foraging site along the Winifred Beach Track were the same individuals that roosted at Roost 565. The detector data from these two sites were compared and there was at least a partial negative correlation, i.e. the periods of time where there was the most activity around the roost there was the least activity at the foraging area. The impression at the foraging area, was that only a small number of individuals were present – maybe as low as 1-4 individuals

The extensive detector data collected throughout this area by CINP have not revealed any other regularly used foraging or roosting areas, although it can not be ruled out that some remain un-located. However, if there are, they are likely to be used by only a small numbers of individuals in a similar pattern to the known sites. Therefore, it is quite conceivable that there may be less than 20 individuals of this species remaining. This has obvious consequences for the urgency of establishing the captive breeding program.

4. Captive management program

4.1 Review of captive breeding programs for small insectivorous bats To assist with the planning for a captive management program for the Christmas Island Pipistrelle, we undertook a review of captive management and breeding programs of insectivorous bats from throughout the world. This review is based on both published and unpublished information, to document the current level of knowledge and experience on keeping small bats in captivity. Information on other species of pipistrelles was especially sought.

4.1.1 Bats in captivity A large number of bat species have been kept in captivity throughout the world and there is extensive experience with maintaining and breeding bats in captivity. However, it appears that only two species of bats, both flying-foxes, have been bred in captivity specifically as part of an endangered species recovery program: the Rodrigues Fruit Bat Pteropus rodricensis and Comoros Fruit Bat P. livingstonii (Racey and Entwistle 2003). A captive breeding program was established for the Rodrigues Fruit Bat in 1976, when the wild population in Mauritius fell to 75-80 individuals due to hunting, deforestation and loss of both foraging and roosting habitat due to cyclone damage (Carrol and Mace 1988, Mickleburgh et al. 1992, Racey and Entwistle 2003). Captive colonies of this species have been held at a number of zoos as a safeguard to protect it from future catastrophic events and extinction (Carrol et al. 1988). The captive program commenced with just 18 individuals in 1976, and in 2008 was estimated to be 685 individuals. During the same period the wild population has independently increased to approximately 1200 individuals (WAZA 2008) due to education programs to increase public awareness and reduce hunting pressure, and the protection and restoration of foraging and roosting habitat. Both captive and wild populations have retained remarkable genetic diversity considering the low number of founder individuals (O’Brien et al. 2007). While the original intention of the program was to return captive-bred individuals to the wild, this has not occurred, in part due to the ability to implement in-situ management actions to address the threatening processes.

While many species of microchiroteran bats (microbats) have been held and successfully bred in captivity, we have found no evidence of captive breeding programs specifically for conservation purposes to return viable populations to the wild. Only a few species have been considered for captive breeding recovery programs: the Indian endemic species Hipposideros durgadasi and Latidens salimali were recommended by Molur et al. (2002) for captive breeding programs, although the South Asian region zoos at the time were not considered to be at a standard that would enable them to successfully breed threatened bats. Despite the lack of effort to date, a review of threatened species captive breeding programs recommended that zoos could provide a greater conservation benefit by including more small animals, and commented that bats were suitable for cost-effective captive breeding programs (Balmford et al. 1996).

Despite the limited number of programs specifically for conservation purposes, many species of microbat have been kept in captivity for the purpose of rehabilitation, research or experimentation. As a result there is a wealth of experience on the captive care and husbandry of bats that can be used to guide captive breeding programs for the recovery of endangered species. There are a number of books, and book sections, specifically addressing the captive care of bats (e.g. Hall 1982, Hopkins 1990, Barnard 1995, Lollar and Schmidt-French 1998, Jackson 2003), which provide extensive information on housing, feeding, and diagnosis and treatment of ill-health or injuries, including medications. In the

USA, UK and Australia there are extensive programs to rehabilitate injured or orphaned microbats and these can provide a useful resource for developing captive maintenance and breeding protocols. For example, in the UK hundreds of injured or orphaned pipistrelles are taken into captivity by specialist bat carers each year: at one wildlife centre in the north-west of England a total of 748 pipistrelles were taken into care over a 10-year period (Kelly et al. 2008).

A number of species of pipistrelles have been successfully bred in captivity in the northern hemisphere. In Italy, a colony of Savi’s Pipistrelle P. savii and Kuhl’s Pipistrelle P. kuhlii have been maintained in captivity with females mating and giving birth to young and successfully rearing them to independence (Dondini and Vergari 1995). In the UK the Pipistrelle P. pipistrellus has been maintained for many years in captive breeding colonies (Kleiman 1969; Racey 1969; Racey and Kleiman 1970; Hughes et al. 1995). It appears, therefore, that species in this genus can be successfully maintained and bred in captivity.

New Zealand has a long history of maintaining bats in captivity, with varying success. A captive colony of the New Zealand Short-tailed Bat Mystacina tuberculata was first established at Wellington Zoo in 1981, but was unsuccessful, with all seven individuals dying from septicaemia (Heath et al. 1987; Blanchard 1992). Ear lesions resulting from a mite infestation were thought to be the primary cause, although lack of experience in bat management and limited flight space were considered contributing factors (Heath et al. 1987; Blanchard 1992). A second colony of Short-tailed Bats (five male and four female adults), was held at Wellington Zoo in 1992 (Blanchard 1998). Two males died the day after arrival, with a tentative diagnosis of gastro-intestinal upset (Blanchard 1998). Six pregnancies resulted in 3 live births, although no young survived beyond 5 months of age. One wild-caught male, however, survived for more than 6 years in captivity.

In the late 1990s an ambitious conservation program was undertaken by the New Zealand Department of Conservation to take a large number of Short-tailed Bats into captivity on Codfish Island while a rat eradication program was undertaken. A preliminary trial was conducted in 1996 to develop captive management protocols. A total of 36 bats were kept in captivity for 41 days in a flight enclosure set within the forest on the island (Sedgeley 1997). The bats responded well to captivity, quickly adapting to artificial food and roost sites, and remained in good condition throughout the captive period. At the end of the trial the bats were released, with radio tracking revealing they rapidly integrated back into the wild population (Sedgeley 1997). Following this trial, the rat eradication captive program was undertaken in 1999, whereby 399 bats were held in captivity for 12 weeks. These also remained in good condition with the majority (97%) increasing in weight. With the exception of one incident where 45 individuals died during a routine monitoring check, only nine individuals died in captivity, with 345 successfully released to the wild (Sedgeley and Anderson 2000).

In 2005, the New Zealand Department of Conservation attempted to establish a colony of Short-tailed Bats on a pest-free island (Kapiti Island) by translocating juveniles (Adams et al. 2005). Twenty-five pregnant females were taken into captivity for 10 weeks, until they gave birth and their young had learnt to fly and feed on insects. Video surveillance was used to monitor the bats to keep handling and disturbance to a minimum. Six adults and four pups died. The remaining adults were released at their original capture location, while the young were released on the predator-free island. Six months later a number of the released individuals were recaught, and found to have lesions on their ears. This was thought to be due to scabies resulting from inflammation caused by mites (Gartrell 2007), similar to the earlier problem in the zoo colony. These individuals were treated with antibiotics, however, only partial amputation of the ears solved the problem, rendering the bats unfit for the wild. Extensive captive maintenance protocols have been developed in New Zealand (Blanchard 1998; Sedgeley pers. comm.), and these can be used to assist in the development of a detailed protocol for the Christmas Island Pipistrelle.

The experience of keeping bats in captivity in Australia is summarised by Hall (1982), Hopkins (1990) and Jackson (2003). Many Australian zoos hold captive colonies of flying-foxes, but to our knowledge the only species of microbat currently held in a zoo is the large, carnivorous Ghost Bat Macroderma gigas which is held at Taronga Zoo in Sydney, Perth Zoo and Adelaide Zoo. Colonies of small insectivorous bats have previously been held at Taronga Zoo and at several universities. In the 1980s a captive colony of Gould’s Long-eared Bats Nyctophilus gouldi were housed in an outside enclosure at the Australian National University, Canberra for 4 years, with 33 females giving birth to young during this period (Phillips and Inwards 1985). The health of individuals was apparently good, with the only comment on mortality that ‘deaths were generally the result of drowning’. Hosken (1998) held a captive colony of eight female and five male Lesser Long-eared Bats Nyctophilus geoffroyi in outside enclosures at the University of Western Australia to experimentally investigate sperm fertility and skewed paternity. Individuals were trapped from the wild in January and February and mating commenced soon after in March. Seven of the eight females successfully mated and gave birth to young.

Considerable guidance can also be gained from experienced bat biologists who have maintained bats in captivity for many years. For example, one of us (LL) has kept a range of species in captivity for over 25 years, including two Eastern Freetail Bats that have been kept for 19 years, and was involved in the successful New Zealand program. There are also a large number of wildlife carers throughout Australia that have extensive experience in the maintenance of bats in captivity and have bred a number of species successfully in captivity.

4.1.2 Enclosures While smaller enclosures have been used to house bats in various studies, enclosure sizes have tended to increase over the years, corresponding with greater success in maintaining and breeding microbats. A number of examples where small microbats have been maintained successfully are provided in Table 1. Various species of European pipistrelles have been held successfully in flight enclosures 3 x 3 m or 4 x 4 m (Dondini and Vergari 1995, Hughes et al. 1995). In the USA, Lollar and Schmidt-French (1998) suggest the minimum flight area for pipistrelles of 2.7 x 3.7 m, with 3.7 x 3.7 m better and 7.3 x 7.3 m ideal, particularly if outdoors. L-shaped flight cages are used by some people to provide increased opportunity for bats to develop turning agility (Lollar and Schmidt-French 1998), while others install vegetation to provide obstacles to avoid (Kelly et al. 2008). A study in the UK investigated the success of releasing hand-reared pipistrelles that had been held in varying size enclosures (Kelly et al. 2008). Individuals that had had either limited pre-release flight training (20 minutes per night indoors), or more prolonged flight training but in a small enclosure (3 x 2 m) did not survive in the wild for more than a day or two, despite appearing to fly proficiently and strongly when they were released. In contrast, individuals kept for the three weeks in a large outdoor flight enclosure (7 x 4 x 2.3 m) with free flying insects to provide prey catching opportunities, successfully survived in the wild, at least for the two weeks they could be tracking. This indicates that although bats can be kept in smaller enclosures for short periods of time, larger enclosures are required for longer term housing, especially prior to release to give sustained flight practice.

In the New Zealand captive programs, an enclosure of 9.2 x 4.6 x 1.8 m in size was used for 36 individuals in the preliminary Codfish Island trial. For the larger program, four enclosures (approximately 14 x 5 x 2 m) were used, each accommodating up to 100 bats. The enclosures were constructed of untreated timber covered with chicken mesh and lined with shade cloth. Each enclosure had a double door entry / exit system, multiple roost boxes and feeding platforms, and acclimatisation cages (Sedgeley and Anderson 2000). Acclimatisation cages were used to house up to 20 bats, while they adapted to the captive food, before releasing them into the main enclosures. These were 2.0 x 0.6 x 1.0 m high and constructed of untreated

timber frames with solid plywood top and base and walls of shade cloth. Access was provided by a well fitted fully hinged door with an interior curtain hung in the doorway to prevent flighty bats escaping when the door was opened. A three stage process was used to introduce the bats to captivity. Bats were initially held in transfer boxes (0.4 x 0.3 x 0.2 m) lined with shade cloth and hessian (Sedgeley 1997). The next day the bats were transferred to the roost box in the acclimatisation cage if their weight was above a minimum level. On the third day, they were again weighted and transferred to a roost box in the free flight enclosure. Bats were given water and fed an abundance of mealworms each night (Sedgeley and Anderson 2000). Any bat that was underweight or in poor condition was held in a transfer box by itself and feed regularly until its condition improved.

Table 1. The size of flight enclosures used for successful captive colonies of small microbat species. The species listed here are 4-10 g in size, with the exception of the Short- tailed Bat which is 15 g.

Species Size of enclosure Location Reference (length x width x height)

Savi’s Pipistrelle P. savii 3 x 3 x 3 m Italy Dondini and Vergari 1995 and Kuhl’s Pipistrelle P. kuhlii Pipistrelle Pipistrellus 4 m x 4 m x 2 m UK Hughes et al. 1995 pipistrellus Pipistrelle Pipistrellus 6 x 2 x 3 m UK Eales et al. 1988 pipistrellus Pipistrelle Pipistrellus spp. 7 x 4 x 2.3 m UK Kelly et al. 2008 Pipistrellus sp. 2.7 x 3.7 m minimum USA Lollar and Schmidt-French 7.3 x 7.3 m optimal 1998 Lesser Long-eared Bat 5 x 2 x 2.2 m WA Hosken (1998) Nyctophilus geoffroyi Lesser Long-eared Bat 6 x 2 x 2.3 m Canberra Ellis et al. 1991 Nyctophilus geoffroyi Gould’s Long-eared Bats 5 x 2 m and 3 m Canberra Phillips and Inwards 1985 Nyctophilus gouldi Short-tailed Bat Mystacina 9.2 x 4.6 x 1.8 m NZ Sedgeley 1997 tuberculata Short-tailed Bat Mystacina 14 x 5 x 2 m NZ Sedgeley and Anderson tuberculata 2000

4.1.3 Feeding bats in captivity The standard diet for microbats in captivity is live mealworms, the larvae of the Mealworm Beetle Tenebrio molitor, supplemented with a mixture to provide the required levels of protein, vitamins and minerals (Hopkins 1990; Barnard 1995; Lollar and Schmidt-French 1998; Jackson 2003). Mealworms fed only on bran do not provided a balanced diet and individuals fed unsupplemented mealworms quickly loose fur and start to deteriorate (LL pers. obs.). While a range of supplements are used throughout the world, in Australia and New Zealand,

the most common is the Wombaroo Small Carnivore or Insectivore Rearing Mix (Wombaroo Food Products, Mt Barker, South Australia). Two Eastern Freetail Bats Mormopterus sp. have been successfully maintained in captivity for 19 years, fed exclusively on mealworms supplemented with Wombaroo mix (LL pers. obs.). Both individuals have remained healthy and in good condition throughout this time. In the USA, some rehabilitators use blended mixes that include various baby foods, incorporated with mealworms (Lollar and Schmidt- French 1998).

As the majority of insectivorous bats catch their prey in flight, they need to be taught to feed on mealworms. The human investment made during the first few days in captivity can determine the future of a captive bat (Hopkins 1990). Teaching insectivorous bats to feed in captivity is time-consuming, as initially all individuals need to be hand-fed. When presented with a mealworm, some individuals respond immediately and start eating, while others take a bit more encouragement. Bats are first presented with a decapitated mealworm, smearing the innards into their mouths or on their lips. After this they usually rapidly progress to taking whole mealworms (Fig. 7). Once the bats are readily taking mealworms while being hand-fed, they can be taught to feed themselves from a tray of mealworms which greatly reduces the time investment required (Lollar and Schmidt-French 1998). Mealworms need to be provided in smooth-sided dishes to ensure they do not escape. Species differ in the rate at which they learn to self feed. A group of 20 Pipistrelles P. pipistrellus taken into captivity in the UK were trained to self feed on mealworms within two days of captivity, after which time only a few individuals had to be hand-fed additional mealworms (Hughes et al. 1995). Feeding routines should occur in the same way and at the same time so that bats become conditioned to meal time (Wilson 1988). Quiet surroundings with a minimum of disturbance will also increase the chance of successful feeding (Wilson 1988).

Fig. 7. Hand-feeding a Lesser Long-eared Bat Nyctophilus geoffroyi on a mealworm.

Conditioning bats to accept an artificial diet may be assisted through the process of observational learning, with individuals more quickly learning to self feed by observing feeding conspecifics (Gaudet and Fenton 1984). Placing bats in small containers with ready access to

food facilitates this process (Baer and Holguin 1971, Sedgeley and Anderson 2000). Once they have learnt to self feed they can be transferred into larger flight cages. Bats typically increase in weight in captivity. Regular weighing of individuals is required to monitor weight loss or gain, which can be rectified by varying the amount of food provided (Sedgeley and Anderson 2000). The artificial mealworm diet can be supplemented by wild insects attracted into the enclosure by using a black light (Jackson 2003).

4.1.4 Breeding in captivity While there have been some failures in attempts to maintain and breed bats in captivity, there have also been many successes. Individuals need to be in good physical, physiological and psychological condition to breed in captivity (Wilson 1988). The main requirements are the provision of an appropriate social environment and the correct diet (Dondini and Vegari 1995). The greatest success in keeping and breeding bats in captivity is achieved if social organisation is comparable to that in the wild, as abnormal grouping can lead to physiological and behavioural changes, including reduced longevity and an increased incidence of reproductive failures (Rasweiler 1975). Aggressive interactions may interfere with the normal establishment of territories, the procurement of mates or the feeding and protection of the young (Rasweiler 1975). Despite these difficulties, successful mating of microbats in captivity has been recorded for a number of species, including two species of pipistelles P. subflavus (Mohos 1961) and Savi’s Pipistrelle P. savii (Dondini and Vegari 1995).

The typical reproductive pattern shown by many species of small microbats involves sperm storage, whereby the males become reproductively active over summer (or the wet season in the tropics) and mate with the females in autumn. The females store the sperm in their reproductive tract over winter when there are few insects available (the dry season in the tropics) and in spring as insect numbers increase, they ovulate and the pregnancy commences (Racey 1982). Gestation is typically about 3 months resulting in the young being born at the start of summer (start of the wet in the tropics) when there are abundant insects. Young commence flying about a month later (Altringham 1996).

Mating activity in bats may be indicated by aggression and territorial defence (Racey and Kleiman 1970). The mating system is promiscuous and individuals will mate multiple times during a season. Some males dominate and copulate with all available females (Orr 1954). As a result it is recommended that in captivity only a single male, or a small number of males, is housed with a group of females during the mating period, and that females are kept separate from males during pregnancy and lactation (Lollar and Schmidt-French 1998). Unnecessary handling and disturbance of female bats, particularly over the periods of sperm storage and pregnancy, may compromise reproductive success (Kleiman 1969, Dondini and Vegari 1995). However, this needs to be balanced against the need for regular monitoring to identify loss of condition or difficulties during pregnancy and allow intervention if required (Lollar and Schmidt-French 1998). After mating, females form a vaginal plug preventing further matings. The ejection of vaginal plugs can occur naturally, but can also be triggered by handling in captivity (Phillips and Inwards 1985). However, some females have successfully become pregnant and given birth despite being handled frequently (seven of eight female Lesser Long-eared Bats gave birth despite being handled twice a day during the sperm storage phase; Hosken 1998).

Pipistrelles (P. pipistrellus) in the northern hemisphere appear to readily breed in captivity. In the 1960s a captive colony was established by catching females after they had mated (Kleiman 1969). The majority of individuals gave birth and successfully reared their young in captivity (Racey and Kleiman 1970). Sexes were mixed during winter, but segregated during pregnancy and lactation over summer and spring as is the habit of this species in the wild. Hughes et al. (1995) took 19 adult female Pipistrelles into captivity during the late stage of pregnancy. Seventeen of the females gave birth to young. The young developed at the same

rate as their wild counterparts, despite being frequently handled and used in flight experiments investigating the ontogeny of flight. At the end of the several month investigation, all the adults and 16 of the 17 young had survived and were successfully released.

Females form maternity colonies where the young can cluster together while the females are away from the roost foraging. Female Pipistrelles generally retrieve their young for the first three days after birth, although the young detach from the nipple after one week and only attach while feeding after the third week (Kleiman 1969). Most young in the Kleiman (1969) study doubled their weight within 10 days, with eye slits appearing about day two and forearms reaching adult size after five weeks. In some cases, Pipistrelles appeared to suckle their young indiscriminately, with a number of females found to be suckling offspring that were not their own (Kleiman 1969). Eales et al. (1988) also observed five out of 11 captive juvenile Pipistrelles suckling from two different females, which suggests that shared nursing may be a feature of maternal behaviour in Pipistrelles. This is likely to increase survival rates of young in captivity.

4.1.5 Longevity Longevity for microbats in captivity has often been less than the lifespan expected in the wild, although captive Big Brown Bats Eptesicus fuscus have reportedly been successfully held for 15 years, which approaches wild longevity (Jackson 2003). Captive colonies of over 60 Mexican Freetail Bats have been successfully maintained in captivity, showing high longevity and reproductive success, with individuals surviving at least eight years and producing young that have survived four years to date and currently participate in mating activity (Lollar and Schmidt-French 1998). Some individuals have lived longer in captivity than expected in the wild, such as the two Eastern Freetailed Bats held in captivity for 19 years (LL pers. obs.). Some species appear not to cope with captive situations and die within weeks, such as the slit-faced bats of southern Africa (Fenton et al. 1983). These species appear highly strung and subject to stress. Fortunately, the Christmas Island Pipistrelle is expected to behave more like the earlier mentioned species and other species of pipistrelles, than the African slit- faced bats.

4.1.6 Conclusion Although various species bats have been held in captivity for many years, we can find no evidence of attempts to establish a captive breeding program as part of a recovery plan for an endangered species. Although separate efforts in New Zealand have attempted to both temporarily hold wild bats, and translocate offspring of temporarily captive wild-inseminated females, these have not been combined to establish a breeding colony of wild bats in captivity with the purpose of releasing them and/or their offspring to the wild. The lack of captive breeding for conservation purposes is probably due to a number of factors. Firstly, although many species of microbats around the world are listed as threatened, few have reached critically low population numbers (such as is apparent for the Christmas Island Pipistrelle), necessitating a captive breeding program. For most species in situ management activities are the highest priority, as there are identifiable habitat or predator control actions that would rectify the situation. Secondly, although there is extensive knowledge on maintaining bats in captivity, detailed captive breeding protocols are still being refined.

There are several challenges to successful maintenance and breeding. Firstly, the bats need to accept the artificial diet provided in captivity. Teaching bats to adapt their prey-recognition system to include mealworms in a bowl rather than insects in flight can at times be a difficult process, however, most species adapt readily. It is time consuming teaching bats to feed, and this may be a limiting factor in the number of bats able to be successfully introduced to a captive program at any one time. Secondly, reducing stress is required, by providing a large enclosure to enable sustained flight, and introducing wild flying insects to enable natural foraging behaviour and to provide behavioural enrichment. Thirdly, the appropriate social

groupings need to be provided to facilitate breeding, with mixed-sex colonies during the mating season and segregation of the sexes during pregnancy, birth and lactation. It is unclear whether in situ or ex situ captive housing is a determinant of success, with mixed results occurring from both situations. Enclosures situated within the natural environment however, are more likely to provide optimal conditions.

Increased knowledge of bat behaviour has led to a greater awareness of dietary and environmental requirements, resulting in increasingly sophisticated methods of housing, including enlarging enclosure sizes, and a corresponding increase in the occurrence of successfully held captive populations (e.g. Lollar and Schmidt-French 1998, Sedgeley and Anderson 2000). While most of the knowledge on captive maintenance and breeding is derived from bats held in captivity for research purposes or for rehabilitation, much can be learnt from these studies on the husbandry of bats in captivity.

The key points that can be learnt from this review, in relation to the captive breeding program for the Christmas Island Pipistrelle are: • many species of small microbats have been kept and bred successfully in captivity, including other species of pipistrelles; • there is a large amount of expertise available on the maintenance of bats in captivity; • the greatest success will be achieved by using people that have experience specifically in handling and maintaining bats in captivity; • the larger the enclosure the greater the chance of successful breeding; • there are standard dietary protocols that have been highly successful in maintaining bats in captivity (i.e. mealworms supplemented with Wombaroo mix); and • bats can be successfully kept for long periods of time in captivity.

4.2 Specific objectives of a captive breeding program The overall aim of the captive breeding program is ultimately to reduce the risk of the Christmas Island Pipistrelle going extinct. Bats will be maintained and bred in captivity while the cause/s of the decline in the wild is identified and rectified, enabling the release of captive- bred animals to successfully re-establish a wild, self-sustaining population. Due to the current very low number of individuals in the wild, the number of founder individuals is likely to be small and so it will be necessary to maintain the captive breeding program for 10 years to be able to build the numbers up sufficiently for a release program (for example, if only 10 individuals could be brought into captivity, at best the captive colony is likely to be only 50 individuals after 5 years which is too small for a release program). During this time actions will be required to identify and rectify the cause of the decline on the island, and animals can be experimentally released to assist this process. If the program was terminated earlier than 10 years, requiring the animals to be released into the wild they should be monitored closely to investigate if the threatening processes, even if not identified, have lessened.

The specific objectives and their associated performance criteria are:

Objective 1. To successfully trap sufficient Christmas Island Pipistrelles to form the basis of a successful captive breeding program. Performance criteria 1. As many as possible of the remaining animals left in the wild are trapped and brought into captivity, with females preferably representing more than 70% of individuals.

Objective 2. To develop captive husbandry techniques for the Christmas Island Pipistrelle to a sufficient standard that it is possible to maintain and breed animals in captivity. Performance criteria 2 A high proportion of individuals (at least 90%) survive in captivity for at least a year (preferably 3 years).

Objective 3. To successfully breed Christmas Island Pipistrelles in captivity such that the overall number of individuals held in captivity increases. Performance criteria 3. The majority of adult females (preferably greater than 75%) give birth to young each year, with these captive-born offspring themselves also successfully raising young.

Objective 4. To ensure captive-bred Christmas Island Pipistrelles behave in a manner similar to the wild, including forming appropriate social groups and in their ability to catch insects in flight. Performance criteria 4. Captive-born young successfully integrate into social groupings and feed on wild flying insects introduced into the enclosure.

Objective 5. To undertake genetic analysis of all individuals taken into captivity, and those born in captivity, to establish breeding aggregations to ensure maximum genetic diversity in captive-bred Christmas Island Pipistrelles. Performance criteria 5. High levels (relative to the wild population) of genetic diversity are maintained in the captive population.

Objective 6. To use the captive animals to assist in determining the cause of the decline, especially in relation to disease issues, by continually assessing the health of individuals and undertaking detailed post mortems of any individuals that die.

Performance criteria 6. Any disease or ill-health issues are identified and methods to control these in both the captive and wild populations are investigated.

Objective 7. To investigate the possibility of using the captive colony to experimentally test alternate theories of the cause of decline, without jeopardising the health of the animals. Performance criteria 7. Greater knowledge is obtained of threats to wild individuals and the cause of the decline.

Objective 8. To determine roosting box preferences by trialling a number of box designs in captivity that can then be used to supplement roost sites in the wild if required. Performance criteria 8. Roosting box preferences are determined and adapted for wild conditions.

Objective 9. To successfully release animals into the wild once the cause/s of the decline has been rectified, using an experimental, adaptive management approach. Performance criteria 9. Captive-bred animals are successfully released into the wild, and extensive monitoring is undertaken to determine their survival and breeding success.

Objective 10. To contribute to restoring the population in the wild to its pre-decline levels. Performance criteria 10. Monitoring programs reveal that Christmas Island Pipistrelle distribution and relative abundance levels are at least comparable to those recorded in 1994, and these levels are maintained for at least 10 years after the completion of the captive breeding program.

4.3 Potential receiving facilities Five possible location options have been investigated for the captive breeding program. Three categories of locations are represented: on Christmas Island (the on-shore option); south-east Asia (Singapore Zoological Gardens); and the Australian mainland, both in tropical (Territory Wildlife Park, Darwin) and temperate locations (Zoos Victoria and Taronga Zoo, Sydney). Discussions have been held with representatives from each of the existing organisations, which are outlined below.

4.3.1 Singapore Zoological Gardens The Singapore Zoological Gardens was selected as an option due to being geographically closer than the Australian mainland, and having a similar climate to Christmas Island. In 2007, Dani Best, Parks Australia North, Darwin, had initial discussions with Charlene Yeong, the Conservation and Research Officer at Singapore Zoological Gardens. We re-established contact in July 2008 to determine if they were still interested in being involved in the captive program, which they were. Over the next several months we had regular email contact with Charlene in relation to their quarantine requirements (see Quarantine section) and capacity to provide the captive breeding facilities required. The outcome of these discussions is outlined in the following section.

4.3.2 Territory Wildlife Park, Darwin The Territory Wildlife Park was recommended as it is involved in a number of captive breeding programs with other endangered species, and is in a similar tropical location and environment to Christmas Island. Early in 2008, Mick Jeffery, Parks Australia North and Dani Best had been in contact with Dion Wedd, Curator, Territory Wildlife Park to determine their level of interest in a captive program. They provided some initial indication of costs to establish and run a captive breeding colony. We contacted Dion more recently and they are still interested in the project, but advised that they are not a Quarantine Approved Premises (see below for more detail) and would require financial support to bring their facilities up to this standard. They have also recently lost their microbat skills base, but Dion felt they could rectify this if funds were available. Their biggest concern was that since the cause of the decline on Christmas Island was not known but that disease was a possible factor, that there was a risk of bringing this to the Australian mainland.

4.3.3 Zoos Victoria We have had discussions with Dr Graeme Gillespie, Director Wildlife Conservation and Science and Russel Traher, Curator, Healesville Sanctuary at Zoos Victoria, who have expressed interest in this project. If the captive colony was to be located in southern Australia an artificial microclimatic environment would need to be provided. Graeme advised that they regularly do this for a wide range of tropical species (for example the large butterfly house at Melbourne Zoo) and this was not a barrier to housing the colony in a temperate area. However, Graeme, recognised that there were considerable benefits in housing the colony on Christmas Island and indicated that Zoos Victoria would also be keen to provide specialist expertise and advice if the captive colony was established on the island. Graeme and Russel assisted in estimating the costs for establishing and staffing the breeding program outlined in this report.

4.3.4 Taronga Zoo, Sydney Dr David Middleton, former Senior Wildlife Veterinarian and founder of the Australian Wildlife Health Centre at Healesville Sanctuary, Zoos Victoria, and veterinarian consultant on the Lumsden et al. (2007) project, suggested including Taronga Zoo in the list of options. The Taronga Conservation Society Australia, who is responsible for the operation of Taronga Zoo, are involved in both in situ and ex situ projects to assist conservation recovery projects. In

David’s opinion, they are leaders in the field of captive breeding for conservation recovery programs, have extensive skills and interest in wildlife heath in wild populations and could assist in helping determine if the decline is due to a health issue. We contacted Dr Rebecca Spindler, Research Manager, who was interested in the problem. However, after further discussions she indicated that due to the quarantine issues (outlined below) and the risk of potentially bringing disease into Australia, that Taronga Zoo would be hesitant to house the captive colony and take responsibility for their ongoing captive care. They would however be willing to provide advice and help wherever they could.

4.3.5 Christmas Island As there are no existing suitable facilities, keeping staff or veterinarians on Christmas Island, a facility would need to be built and relevant keeping staff and veterinarians contracted if the program was to be located on island. Rather than trying to run this purely from the island, employing new people working in isolation, we believe the most efficient and effective way to establish and run the facility would be to form a formal partnership with an existing zoo. The zoo could then provide (with the required funding) the relevant expertise to help design and establish the facility, and provide the full time keeping staff and periodic visits by veterinary staff.

Consideration would need to be given as to where the captive breeding facility would be built on the island. Within, or close to, rainforest would provide the optimal microclimatic conditions, although there is a risk that if an environmental contaminant is contributing to the decline then this may still operate. However, this could be the case no matter where the facility was located on the island. Toxicology testing of insects would be required before wild insects were fed to the captive bats. The other considerations on the location are land tenure (i.e. within the National Park or external to it), being close to existing facilities to provide power and water, and ensuring the security of the facility. The Parks Australia Research Station in the middle of the island (the Pink House) would be an ideal location for the facility.

Consideration also needs to be given to how the project would be managed, both initially to establish it and on an on-going basis. A project manager would need to be appointed to oversee the construction and establishment of the facility. Advice would need to be sought from the partner zoo regarding staff and facilities. Options as to whether staff would be employed directly by the zoo or by CINP would need to be discussed.

If a captive breeding facility was to be established on the island, it would provide the opportunity for captive colonies to also be established for other critically threatened species, that occur on the island, especially many of the reptile species.

4.4 Quarantine issues A number of quarantine issues have been investigated to determine if it would be possible to take the animals to either Singapore or the Australian mainland. Quarantine issues for bringing live food for the bats to Christmas Island also needs consideration.

4.4.1 Singapore Zoo The Agri-food and Veterinary Authority (AVA) of Singapore, which is the government body that regulates imports and exports of animals, has documented guidelines on “Veterinary Conditions for the Importation of Zoological Animals and Birds” including Chiroptera (bats) (Appendix 1).

The conditions that are most relevant to establishing a captive colony of Christmas Island Pipistrelles in Singapore are: • the bats have been kept in captivity in the country of export for at least 6 months prior to export or since birth; • the animals shall be accompanied by a veterinary health certificate dated not more than seven (7) days prior to export and signed or endorsed by the competent Veterinary Authority of the country of export; • the bats are free from endo- and ectoparasites, and treated for them as well; • the bats are tested for and free of Nipah virus; • the bats are tested for and free of Lyssavirus; and • the bats have been examined and found to be healthy and free from any clinical signs of infectious or contagious disease at the time of export.

These conditions have a number of implications. To house the bats on Christmas Island for 6 months prior to export, would require the construction of temporary holding facilities and the employment of full time staff, in which case it may be more efficient to retain the bats on the island. A veterinarian would need to be part of the team involved in taking the bats into captivity and preparing them for export. This veterinarian should be an experienced wildlife veterinarian, ideally one that has expertise with small microbats.

The pipistrelles examined during 2005 had very few external parasites (only a small number of mites) and no obvious internal parasites (Lumsden et al. 2007). Assuming bats caught to be taken into captivity were in the same condition, they are likely to have a low parasite load. Due to this it may be preferable to reduce the amount of chemicals the individuals are exposed to. However, if the Singapore Zoo required them to be totally parasite free, then the most appropriate method would need to be discussed with veterinarian specialists. Various methods have been tried on other species in captivity, to varying success. Determining the external parasite load can be done on the island. However, to investigate if they have internal parasites, faecal remains would need to be sent to a parasite expert for examination (Dr Ian Bevridge, Melbourne University). This additional step in the process would need to be factored into the timeframe and the appropriate permits obtained.

Nipah virus is a recently emerged paramyxovirus that was first described in Malaysia and Singapore in 2000 after an outbreak of severe encephalitis in people with close contact to pigs (Chua et al. 2000). Bats, in particular flying-foxes, were subsequently found to be the natural host of this virus (Johara et al. 2001). Forty-six individuals of six species of microbats were also tested, of which one (3% of total), a House Bat Scotophilus kuhlii, had antibodies to the virus (Johara et al. 2001), indicating it had come in contact with, and had the potential to spread, the virus. The House Bat is in the same family as the Christmas Island Pipistrelle (i.e. Vespertilionidae) and so, although it is highly unlikely that Christmas Island Pipistrelles would have been exposed to, or carry, Nipah virus, there would be a requirement to test for it.

We contacted Dr Kim Halpin and Dr Deborah Middleton, from the Diagnosis, Surveillance and Response Group, at the Australian Animal Health Laboratory (AAHL), CSIRO, in Geelong Victoria, who have conducted research on Nipah virus (Middleton et al. 2007). AAHL is the only Australian facility that undertakes testing for Nipah virus. The most reliable way to determine if bats have been exposed to Nipah virus is to test blood serum for antibodies to the virus. This method is accepted by most quarantine organisations, including Biosecurity Australia. However the amount of blood that would be required will be problematic. The advice from Dr Kim Halpin, is that a minimum of 50-100 microlitres (ul) of serum would be required to undertake this test. To obtain this amount of serum, 250-500 ul of blood would need to be taken from each bat. During our 2005/06 study it was very difficult to take blood from these bats due to the small size of both the bats and their veins, despite this being undertaken by an experienced wildlife veterinarian (Lumsden et al. 2007). The only vein it was possible to get blood from was one in the tail membrane, and the amount that could be taken was approximately 50 ul of blood. This amount of blood would produce approximately 10 ul of serum which is only 10-20% of the required amount. In addition, it was only possible to obtain blood from the larger females, with the tail membrane vein too small to puncture to take blood from the smaller females and the majority of the males. It would therefore not be possible to undertake any tests on these individuals.

Singapore Zoo also has the requirement that all bats are tested for and free of Australian Bat Lyssavirus. Testing for the actual virus requires brain tissue and so this will not be possible. However, antibodies to lyssavirus can be tested from blood serum in a similar way to Nipah virus testing. To do this an equivalent quantity of blood would be required. This could be taken at a different time to the blood taken for Nipah virus testing – for example if the bats needed to be kept in captivity for 6 months prior to exportation, the blood samples could be taken at regular intervals. However, to obtain the amount of blood required for these two tests would mean that blood would have to be taken maybe 10-20 times.

It would not be possible to take this amount of blood from each bat without severely compromising the health of the individual (Dr David Middleton, pers. comm.). As a result it will not be possible to meet the quarantine requirements for importing the bats to Singapore. Therefore, unless Singapore Zoo and the Singapore AVA agreed to waive this requirement, Singapore Zoo would not be able to be considered as a possible option for the location of the captive breeding colony.

We outlined this situation to Charlene and received the following response:

“I have spoken to our vets on the methods of testing for Nipah and lyssavirus, and they had the same concerns as you do about the samples and volume of blood required for testing. The quarantine requirements are set by the Agri-food and Veterinary Authority, which is the government regulatory body for importation and exportation of animals. We’ll check with them on whether it will be possible to waive these requirements. I am also checking with my Zoology colleagues on the feasibility of having the captive colony here. One of my concerns, which was mentioned when Dani was here, is that we have experience and have been successful in managing some megachiropterans – Cynopterus brachyotis and Pteropus vampyrus – but not microchiropterans (microbats). We would require some training and advice from other zoological institutions to set up the colony of P. murrayi here, and I’m not sure if this would be the best thing to do for a species of such high conservation concern!”

On 10 September 2008 we received the following email from Charlene: ‘Unfortunately, it has been decided that we are focusing on local species and will not be able to provide an ex-situ breeding site for P. murrayi. However, if needed, we will be glad to offer advice as best as we can.’

In conclusion, therefore, Singapore Zoo can no longer be considered as the location for the captive breeding colony.

4.4.2 Australian Zoos/Wildlife Parks To bring the bats to the Australian mainland a number of assessments and permits would be required. The process would be to: • Obtain a permit from DEWHA – referral under the EPBC Act as a matter of national environmental significance. • Submit this and a proposal including all the available information to Biosecurity Australia who would then undertake an Import Risk Assessment (IRA). This would review all the available literature, assess the risk and develop risk management conditions. • Australian Quarantine and Inspection Service (AQIS) would then implement this risk management plan.

We contacted Clare Jones from Biosecurity Australia who undertakes the risk assessments for ‘Non-ruminants and Zoo Animals’. Clare advised that when conducting an IRA they look at a range of issues that would impact on the risk associated with importation of animals. These may include: • Risk of disease introduction. Information is required on the number of animals to be imported, the prevalence of certain diseases (primarily exotic diseases) in the source population. • The likelihood that, if a disease were introduced into Australia through importation of the bats, it spreads and establishes in Australia (in both bats and any other susceptible species). Assessing this would mainly involve considering the characteristics of each disease, whether susceptible Australian animals are likely to come into contact with the disease agent, and whether the Australian environment and animal populations are conducive to spread and establishment of disease. Information on the specific diseases which these bats may carry and how the bats will be transported and housed would be needed for the assessment. • The potential impacts of the introduction of exotic diseases through importation of the bats. This would include impacts on the environment, humans, animals, and other economic impacts.

Clare advised that IRAs typically take about 12 months to complete. She said the more information that is available, especially if it is published in peer reviewed journals, the quicker the process would be. Unfortunately, we do not have most of the information needed for an IRA for the Christmas Island Pipistrelle – e.g. what diseases they carry (if any) and, if disease is causing the decline of the species, there would be a high risk of bringing this into Australia. Due to the lack of knowledge, the process is likely to take longer and Biosecurity Australia are likely to require stricter quarantine conditions and additional testing of the individuals before they are brought to the mainland.

Clare advised that the IRA would likely require that all bats were tested for lyssavirus and Nipah virus. Therefore the quantity of blood would be the same as indicated above for Singapore Zoo. In addition, there might also now be a requirement to test for Hendra virus, which to date has only been recorded in flying-foxes, but has many similarities to Nipah virus. As Nipah virus has been recorded in microbats, there may be a requirement to test for Hendra virus as well, especially given the recent outbreaks of this virus in Queensland. However, as Hendra virus is endemic there may not be a requirement for testing. If there was, more blood would required from each individual – i.e. another amount the same as for each of the other two tests. If any individuals tested positive to antibodies for any of these three viruses, it is highly likely that none of them would be able to be brought into Australia.

If the cause of the decline in the pipistrelle is disease related it is most likely to be a factor other than Nipah or lyssavirus. Therefore there may be additional requirements to test for a range of other possible diseases, also requiring blood.

Biosecurity Australia currently have a 2 year backlog of high priority IRAs, so unless this project could be given a higher priority and aspects fast-tracked, it is possible that it could take 3 years before an assessment was completed to determine if it was possible to bring the animals into Australia. When we explained the urgency of the situation, Clare suggested that it may be possible to allow these bats into Australia sooner for an approved captive breeding program. Biosecurity Australia would need to thoroughly examine any available research, and depending on what they found, it may be possible to allow the bats in without the completion of a full risk assessment under certain (probably quite strict) quarantine conditions. Although there would be a preference for testing the animals for all potential viruses, an inability to test the bats for these viruses may not exclude the possibility of importing them into Australia under quarantine control, if other measures provided sufficient assurances that the risk of introducing disease is managed (Clare Jones, Biosecurity Australia pers. comm.). These measures could include importing the bats under a ‘closed system’ – that is, that they would remain under strict quarantine control until re-exported to Christmas Island. The bats would probably also require housing under specific conditions, such as double fencing of enclosures to ensure no contact with outside animals, staff would be required to undertake biosecurity measures such as disinfection after handling the animals, and no opportunity for animals outside of the enclosure to come into contact with material contaminated by the bats.

In conclusion, a Biosecurity Australia IRA assessment would need to be undertaken before animals could be imported into Australia, and this could take some time (possibly years). Until the IRA was completed, the conditions placed on importing the bats into the Australian mainland and whether it would be possible to meet these conditions, are unknowns. While it may be possible to waive the requirements for virus testing, this would not be confirmed until the assessment was completed.

If the captive colony was to be established at the Territory Wildlife Park, Darwin, in addition to the above requirements, it would need to be registered as an Quarantine Approved Premise (QAP) by AQIS, as all animals brought in from Christmas Island would need to go, at least initially, to a QAP. The Territory Wildlife Park is not a QAP (Dion Wedd, pers. comm.). We have attempted to determine what would be required for the Territory Wildlife Park to become accredited by AQIS however, based on advice from Clare Jones, Biosecurity Australia and from searching the AQIS website, there is not a single set of standard guidelines, and that each facility would need to be assessed based on their specific requirements. Clare advised that as this case is a little different to the typical QAPs it would require a more thorough analysis of potential risks in the housing, transport, movement and disposal of bats and associated waste. This process presumably would also take some time.

4.4.3 Christmas Island There will be a requirement to hold Christmas Island Pipistrelles in captivity on the island, whether this is as part of a long-term program, or temporarily before moving them to an off- shore location. As a result a suitable food source will be required. The standard food used for captive insectivorous bats is live mealworms supplemented with a suitable mix to provide the necessary proteins, vitamins and minerals (such as the Wombaroo Insectivore or Small Carnivore Mix). Large quantities of mealworms would need to be brought alive to the island from the mainland. Breeding colonies would then need to be established so that there was always a regular supply. It may be necessary to request a mainland supplier to grow small mealworms as the majority of mealworms commercially available are too large to be easily consumed by this species.

We contacted AQIS on Christmas Island to investigate the permit and quarantine requirements for bringing live mealworms into Christmas Island. There are no existing guidelines covering the importation of insect larvae to Christmas Island, and it may be necessary for Biosecurity Australia to undertake an Import Risk Assessment. We completed an Import Permit Application and emailed it to Hermana Boll, AQIS on 4 September 2008. This permit was issued on 9 March 2009, and so it will now be possible to take the necessary quantities of mealworms to the island. It is recommended that this diet is supplemented with wild insects, however, these would need to be tested for possible toxicants first (more detail is provided in the captive management plan section).

4.5 Transport requirements Due to the island’s remoteness there are a number of issues and unknowns associated with transporting animals to any of the potential off-shore facilities. These include the length of time the animals would need to be in transit and the resulting impact on their wellbeing. It is not known how the animals would respond to being transported, and there is the potential risk of fatalities during this process. The bats would need to be housed in specifically constructed temperature and humidity controlled, sealed containers and held within the main cabin of the aeroplane to avoid any pressurisation problems. As they are accustomed to high humidity on the island, the generally low humidity in planes could result in them drying out if they were not held in containers which controlled the humidity. These containers would also need to have an adequate oxygen supply.

The cheapest option to take animals to the Australian mainland would be on commercial flights through Perth. (There are currently flights from Christmas Island to Kuala Lumpur, however this would result in additional quarantine issues). This would likely require bats to be in the environmentally controlled containers for over 12 hours, including processing and flight times to reach Perth. A stop-off would be required in Perth to enable the animals to be checked, fed and watered, thus requiring additional quarantine procedures. An agreement would be needed with Perth Zoo to use their quarantine facilities. It may be necessary to stabilise the animals for several days before embarking on the next leg of the journey to Darwin, Sydney or Melbourne. On arrival they would need to go directly into a Quarantine Approved Premise. Unforeseen problems, such as the re-routing of the aircraft or enforced overstays at en-route airports would result in additional transit times.

Alternatively, it may be possible to organise a direct flight from Christmas Island to the city of the proposed facility, which would reduce the overall transit time and hence potentially reduce the impact on the bats. One possibility is the potential use of a military or Customs flight. However, arrangements for such an option would require high level negotiations between the relative Commonwealth departments, and may not be available at the time required.

As it is not known how the bats will respond to being transported long distances, it would be advisable to undertake the transportation of the whole colony in a number of events, trialling moving just a few individuals first. This would also spread the risk, by not subjecting all individuals to an activity with a high level of unknowns, i.e. if it was found that they did not cope with being transported, and worse case scenario, many individuals died in transit, the majority of the remaining population of the species could be wiped out in a single event.

Transport requirements to the Australian mainland have not been investigated in further detail at this stage, as the recommendation is that the captive colony is established on Christmas Island (see below). If at a later stage it was decided to take animals to the mainland, either to form the main colony, or as a second, back-up, colony, transport options will need to be investigated then in more detail at that time.

4.6 Advantages and disadvantages of the various options There are advantages, disadvantages and risks in each of the five potential locations for the captive breeding program, many of which have been discussed already. These are summarised in Appendix 2. The main consideration discussed below is whether an on-island or off-shore option is preferable.

4.6.1 Overall unknowns There are a large number of unknowns in establishing a captive colony, irrespective of where it is located. These include: • Are there enough animals remaining in the wild to establish a viable colony in captivity? • If there are, will it be possible to catch enough of them to bring them into captivity? • Will it be possible to maintain the animals in captivity – i.e. will they feed, survive etc? • Will they breed in captivity? • Will it be possible to determine the cause/s of the decline? • If the majority of the remaining wild animals are taken into captivity will this reduce the possibility of determining the factor/s causing the decline? • Will it be possible to rectify the cause of the decline once it is determined? • Will it be possible to establish captive-bred animals into the wild once the cause/s of the decline is controlled?

4.6.2 Advantages of taking the animals to an existing off-shore facility • An existing facility will have access to trained staff, pathology, veterinarians etc. Although these facilities will have extensive general knowledge on wildlife care and husbandry, specific information on microbat health is minimal everywhere. • Animals will be removed from the environment where the decline is occurring – since the cause/s of the decline is not known it can not be ruled out that environmental contaminants are involved and that these contaminants would continue to impact on captive animals if they remained on the island.

4.6.3 Risks and disadvantages of taking the animals off-shore • While we have unofficial advice that it may be possible to bring the bats to the Australian mainland without testing for viruses etc, until an IRA is undertaken this can not be confirmed. Therefore, there is the risk that the outcome of the IRA would prohibit the bats being brought to the mainland or include conditions that can not be met without jeopardising the health of the individuals. • By the time an IRA is completed (possibly up to 3 years, but could be fast tracked) it is likely that the species will be extinct or in such low numbers as to be virtually extinct. If animals were taken into temporary captivity on the island prior to the IRA being finalised, there would be the requirement for a longer-term ‘temporary’ facility to be built and staff employed, in which case it would be more efficient to retain the facility on the island. • It is unknown how well the bats will tolerate the flight to the mainland. It is possible that there may be fatalities during this phase. • Direct flights would be optimal, but if it was not possible to charter or organise a plane through another government department, there would need to be a stop over at Perth under quarantine conditions. The impact of this extended time in transit is also unknown. • Transportation of animals would need to be undertaken in batches, to reduce the risk of ‘having all eggs in the one basket’ due to unknowns associated with their ability to cope with being transported. This would increase the costs of transportation.

• There would be a requirement to construct transport containers with controlled temperature and humidity and an adequate air supply, to maintain suitable conditions for the bats and to meet Biosecurity Australia ‘closed environment’ quarantine conditions. • It is likely that the bats would have to be held in quarantine conditions for the entire time they were on the mainland (i.e. years), which would increase handling time and requirements. However, this would be required not just to ensure that no pathogens were brought from the island to the mainland, but also to prevent the risk that captive- bred individuals may bring pathogens from the mainland with them when reintroduced to the island. • If disease is contributing to the decline of the species then it is likely that this disease could be brought to the mainland and, even if strict quarantine conditions were adhered to, there would be a risk this could be spread to species on the mainland. If it became as prevalent on the mainland as it has been on the island (if this was what was causing the decline) it could have serious implications for the conservation of many of the mainland species. Since the form of any possible disease agent has not yet been identified, it is not possible to establish mitigation measures. • If some form of ill-health was affecting the population, the stress associated with the prolonged transportation may increase the pathological effect of this agent, further debilitating already compromised individuals. • If taken to southern Australia, temperature and humidity controlled environments would need to be built or existing enclosures modified, which would presumably be more expensive than if established in a tropical environment. In addition, it may be difficult or impossible to exactly replicate the island’s conditions, including day length, and as a result this may detrimentally affect breeding success. Even if taken to Darwin, conditions might not be the same as on the island. • Removing everything from island takes away the focus and ownership of the project for the local community etc.

4.6.4 Advantages of establishing the facility on Christmas Island • The establishment of the captive colony could commence as soon temporary enclosures were established and staff contracted, and these facilities used while long- term enclosures were being built. Since the wild population is declining so rapidly, the quicker animals can be taken into captivity the greater the chance of success in reducing the risk of the species going extinct. • Animals could be brought into captivity as available and appropriate – i.e. it may be possible to bring a few males in first to establish husbandry protocols before larger numbers of females were brought in. • Climatic conditions are going to be optimal for housing and to promote breeding. Day length and climate often determine breeding success in captivity. • It may be easier to work out the cause/s of the decline if the facility is on island – e.g. if a sick animal was located in the wild there would be experts on site to assess it immediately. • Artificial diets can be supplemented with natural food attracted to the cages with black lights – this will enable young born in captivity to learn how to catch insects in flight and become familiar with the natural food supply. • When it comes time to release animals this can be done experimentally with a few individuals released initially to monitor their survival. • It may be easier to obtain additional funding, e.g. from Christmas Island Phosphates if the facility was on the island.

4.6.5 Disadvantages of establishing the facility on Christmas Island • All the expertise would need to be brought onto the island, both with respect to the full time keeping staff and visits by veterinarian and bat biologist specialists. • If samples needed to be sent to the mainland for analysis quarantine requirements would apply and suitable ways of transporting the samples would need to be established (however, biological samples have been taken to the mainland previously; Lumsden et al. 2007). • A reliable supply of captive food (i.e. mealworms) would need to be established on the island. • Quarantine issues relating to bringing in mealworms may need to be resolved (a permit is currently being issued). • It would be more expensive to build the facility on the island than on the mainland due to transport costs (however if it was built in southern Australia the cost of building a climate controlled facility could be more expensive). • Animals would still be exposed to potential environmental contaminants (e.g. airborne poisons) which may have a deleterious impact on captive animals, if these were involved in causing the decline of the species in the wild.

In assessing all the advantages and disadvantages, there are far more advantages and fewer risks to establishing the captive colony on Christmas Island than taking the bats to a mainland facility. In summary, a captive colony could be established much quicker on the island, have a far greater chance of success with fewer risks to the health and survival of the animals.

It is therefore recommended that a captive breeding facility is established on Christmas Island and consideration given to incorporating captive breeding of the critically threatened reptiles. It is also recommended that a formal partnership is established with an existing zoo to provide the specialist advice and expertise required, especially veterinary and pathology staff and resources, and advice on enclosures, husbandry and health monitoring protocols. Zoos Victoria are keen to provide these services and we recommend that further discussions are undertaken with them to progress this possibility.

4.7 Legislative requirements A number of formal requirements need to be addressed before a captive breeding facility can be established on Christmas Island. These include:

• Referral under the Environment Protection and Biodiversity Conservation Act 1999 as a matter of national environmental significance. This report, in conjunction with earlier reports, could form the basis of the documentation required for this referral, including the rapid decline of the species, the urgent need for a captive breeding program to avert its imminent extinction, and the issues associated with establishing a captive breeding facility. • Investigation to determine if an Animal Ethics Committee approval is required to undertake this work. • A Christmas Island Territory permit application to take and keep a protected species will be required. Existing information will form the basis of this permit application. • An ‘Access to Biological Resources for Non-commercial Purposes’ permit from DEWHA. • Quarantine approval to import mealworms to Christmas Island. • The establishment of a Cooperative Conservation Program with the zoo selected to provide the specialist advice and expertise.

In addition to the above, if it was decided instead to use an existing facility on the mainland to house the captive colony, documentation would need to be compiled for Biosecurity Australia for the application for an Import Risk Assessment.

4.8 Biology and ecology of the Christmas Island Pipistrelle in the wild, relevant to a captive breeding program The known biology and ecology of the Christmas Island Pipistrelle is summarised based on information collected during field projects conducted over the past 20 years. Understanding how the species behaves in the wild will provide important information to facilitate the successful establishment and maintenance of a captive colony. Where specific data is not available, reference is made to information from other similar species.

4.8.1 Diet Only limited dietary information is available for the Christmas Island Pipistrelle in the wild (Table 2). In March and September 1984, Tidemann (1985) examined faecal pellets and stomach contents and found the remains of moths, beetles, diptera, micro-wasps and thrips, with moths and beetles dominating. In 1994, scats were collected from 10 individuals caught during the non-breeding season which revealed a range of insect orders, with moths, beetles and flying ants predominantly consumed, collectively constituting 99% of the diet (Lumsden and Cherry 1997). DNP (unpublished data) also recorded moths, beetles, bugs and flies, as well as bark lice in the diet of the pipistrelle.

Table 2. Summary of available dietary information for the Christmas Island Pipistrelle. 1984 data is from Tidemann (1985); 1994 is from Lumsden and Cherry (1997); and 2004 is from DNP unpublished data.

Prey type 1984 1994 2004

Moths (Lepidoptera) Present 51.5% Present Beetles (Coleoptera) Present 25.8% Present Flying ants (Hymenoptera) – 21.5% – Bugs (Hemiptera) – 1.1% Present Flies (Diptera) Present 0.1% Present Micro-wasps (Hymenoptera) Present 0 – Thrips (Thysanoptera) Present 0 – Bark lice (Psocoptera) – 0 Present

While it appears a wide range of small flying insects are taken, data on insect availability is required to determine if the bats are opportunistically taking insects within the manageable size range, or whether particular orders are being actively selected and whether this varies seasonally as insect numbers fluctuate. It is likely that all prey is taken in flight rather than any gleaning of insects off vegetation or the ground, as is the case with other species of pipistrelles. Microbats typically consume over half of their body weight in insects in a night.

Implications for captive colony: Moths and beetles are typical prey items for many bat species, so in this respect the Christmas Island Pipistrelle is similar to many other species. Therefore it is likely that they will respond to a captive diet in a similar, positive way. Although it is likely that in the wild all prey are taken in flight, similar species (such as the Little Forest Bat Vespadelus vulturnus) will learn to feed by taking mealworms out of a tray. Therefore, it is expected that it will be possible to train captive animals to self feed on mealworms (supplemented with Wombaroo Insectivore Mix). This will be their staple diet. However, it will be important to supplement this

with wild food, both to provide a balanced diet and to ensure that captive-bred young learn to forage on flying prey. As both moths and beetles are attracted to black light it will be possible to use a light to attract these into the enclosure to provide additional prey. Since bats typically consume over half of their body weight in insects in a night it is important that large quantities of food is provided.

4.8.2 Flight pattern The Christmas Island Pipistrelle is a small bat with a manoeuvrable flight pattern. Although it typically forages in small clearings and edge habitats, roost sites are located within primary rainforest where the vegetation is cluttered (Lumsden et al. 1999, 2007). Recent observations indicate its flight is so manoeuvrable that it is able to detect and avoid harp traps and mist nets.

Implications for captive colony: Due to the manoeuvrable flight pattern the enclosures do not need to be as large as they would if the species was a fast flying, less manoeuvrable species. It is likely that bats would be able to fly within an area 3.5 x 3.5 m (such as the proposed temporary fly-wire mesh tents – refer to the captive management section) and it may be possible to house small numbers of bats temporarily in these enclosures. However for longer term maintenance and to house larger numbers of bats to enable the social groupings needed for successful breeding, larger enclosures will be needed. Lollar and Schmidt-French (1998) suggest the minimum flight area for similar-sized species of pipistrelles in the USA is about 2.7 x 3.7 m, with 3.7 x 3.7 m being better and 7.3 m x 7.3 m being ideal. Enclosure sizes of 10 x 4 m is recommended for the long-term housing of the Christmas Island Pipistrelle.

4.8.3 Nightly activity patterns In 2005, individuals typically emerged from their roosts during the dusk period (13.3 ± 6.0 mins after sunset, range 0-31 min after sunset; Lumsden et al. 2007). However, in 1984, Tidemann (1985) reported Christmas Island Pipistrelles hawking insects along roads and ecotones during the late afternoon, 1 hour before sunset. No daytime foraging of the pipistrelle was observed in studies in 1994, 1998, 2004 – 2009 (Lumsden and Cherry 1997; Lumsden et al. 1999; CINP unpubl. data; LL pers. obs.), with bats first appearing during the dusk period. This led to the suggestion that a temporal shift in foraging behaviour had occurred, possibly due to predation risk from a diurnal predator, such as the recently self-introduced Nankeen Kestrel (Lumsden et al. 1999). Foraging by bats during daylight hours on islands elsewhere in the world has been attributed to a lack of diurnal avian predators (Speakman 1995). Emergence times from roosts were slightly earlier in 1998 than in 2005 (Lumsden et al. 1999, 2007). It is not known if the apparent difference in these emergence times represents a shift in behaviour due to predation, or is due to differences associated with the season, weather or reproductive condition of the females.

At foraging areas, the highest levels of activity are within the first two hours after sunset (Lumsden and Cherry 1997, Lumsden et al. 1999). After this post-dusk peak, activity levels are relatively consistent until a pre-dawn increase in activity. From observations undertaken in January 2009 and the detector data collected at the known roost, it is apparent that individuals spend considerable amounts of time during the night in flight near the roost. This occurs both during the breeding season when females return to suckle young, and outside of the breeding season. It is not known if at times they also use night roosts within their foraging areas.

Implications for captive colony: By observing the times bats typically emerge from their roosts while in captivity and hence in a protected environment, it may be possible to shed light on whether the risk of predation by diurnal birds influenced their behaviour, and hence may be a factor in their decline. Within the

captive enclosures a range of roosting opportunities will be required to enable individuals to choose whether to return to day roosts or select alternate night roosts.

4.8.4 Breeding biology The breeding biology of the Christmas Island Pipistrelle appears to follow a similar pattern to many other small microbats. There appears to be a single breeding season, with females producing a single young each year. Mating is likely to occur in early winter with males trapped in May-July having either fully or partially enlarged testes, indicating they were approaching peak reproductive condition, and that mating was imminent (Lumsden and Cherry 1997, Lumsden et al. 1999). Tidemann (1985) demonstrated sperm storage in female Christmas Island Pipistrelles, and a delay between mating and ovulation. This strategy is used by other species of insectivorous bats in environments where there are seasonal fluctuations in food supply (Racey 1982). Tidemann (1985) suggested that pregnancy is likely to commence during September with a single young born in December, at the start of the wet season when insect numbers are highest. This strategy was confirmed by Lumsden et al. (2007). In mid-late December 2005, females were either heavily pregnant with parturition imminent, or had recently given birth, as judged by the condition of the lactating nipples.

The majority of females caught during the 2005 breeding season were breeding, with 82% either pregnant or having recently given birth (Lumsden et al. 2007). Of the seven females that were not in breeding condition, one appeared to have bred in previous years, while the other six had not bred before. Some species of bats breed in the first year of their life, while others do not commence breeding until the second year (Barclay and Harder 2003). It is not possible to age bats once they are more than several months old, and so the age of these non-breeding females could not be determined. It is possible that these six females were first year individuals, although it is also possible that they were adults that were not breeding for some unknown reason. The proportion of Christmas Island Pipistrelle females breeding is consistent with other species in this family (Vespertilionidae): a review of the reproductive rate of a wide range of microbat species revealed that 85% ± 3% of females breed each year (Barclay et al. 2004).

Nothing is known of the development of young Christmas Island Pipistrelles. However, if they follow a similar pattern to other small microbats, young are likely to be approximately 25-30% of the mother’s body weight when born (Altringham 1996), and hence weigh approximately 1 g. Young bats grow rapidly and reach adult dimensions within approximately 4 weeks. During this time they are suckled exclusively on milk, feeding periodically throughout the day and night. Females have high energy demands at this time and need to feed on large quantities of food. In a captive colony of Pipistrelles in the UK young remained attached to the females bodies for the first few days, being first found separate from the mother on day 4 (Hughes et al. 1995). The eyes of these pups opened days 4-6, and downy fur started to appear days 6-8. Juveniles started taking mealworms from the age of 26 days. The young need to reach 90-95% of the adult skeletal size (weight 70%) before they can commence flying (Altringham 1996). Young start catching their own prey as soon as they commence flying but typically continue suckling from their mother for several more weeks. Young bats when they start flying sometimes crash land and the mothers will retrieve them from the ground (a floor surface in the enclosures will be required to enable this to happen). Although orphaned young learn to successfully fly and catch prey without input from their mother, it is likely that these skills are learnt or perfected more quickly if the young can learn from their mother.

While all the evidence from Christmas Island Pipistrelles caught throughout the year indicates there is a single synchronous breeding season, other species of tropical bats breed throughout the year producing multiple litters (Tuttle and Stevenson 1982). Some species that typically have a single young once a year, have been recorded breeding more frequently, or

alternatively producing twins in captivity where there is an abundant food supply (Lollar and Schmidt-French 1998). It is considered that seasonality of food sources is the most important factor determining reproductive patterns in tropical bats (Tuttle and Stevenson 1982).

Implications for captive colony: If the bats breed at the same rate as in the wild, the most optimistic possibility is that the majority of females in captivity will produce a single young each year. The survival rate of young in the wild is not known and so it can not be predicted what proportion of young are likely to survive to maturity. It is possible that females breed in the first year of their life (i.e. mating at approximately 6 months) although some might not breed until their second year. Therefore, the reproductive output is likely to be low and the captive population is likely to increase in size only slowly.

4.8.5 Longevity As there have been no long-term banding studies of this species, the longevity in the wild is not known. Longevity for small microbats in southern Australia is typically 7-8 years (Vespadelus spp.; L. Lumsden pers. obs.). Other species of the genus Pipistrellus have been recorded at 6-15 years (Tuttle and Stevenson 1982, Jackson 2003). Pipistrelles in England were recorded living up to 8 years in the wild, with an annual adult survival rate of 0.64, with a lower juvenile survival rate in their first year (Thompson 1987). The oldest individual recorded of an Australian species in the wild was a Southern Bent-wing Bat caught in Victoria at 21.5 years old (Lumsden and Gray 2001). Overseas some temperate species have been recorded living to 33 years (Arlettaz et al. 2002). Less information is available on the longevity of tropical species and it is possible their life span may be less. However, the tropical species Myotis nigricans in Panama has been recorded living to seven years (Wilson and Tyson 1970). Although there are some records of temperate species living to advanced years, it is expected that the Christmas Island Pipistrelle is more likely to live to approximately 8 years.

Implications for captive colony: If the captive colony is retained for 5 years it is possible that some of the original founding animals would still be alive at the time of release which would be of benefit for when releasing animals as they would be more wild savvy. However, the majority of animals being released would be captive-bred animals. If the captive colony is required for 10 years it is unlikely that any wild born animals will be alive at the end of the period.

4.8.6 Social organisation Little is known about the social organisation of the Christmas Island Pipistrelle. There is conflicting information regarding sex ratios. In June to mid-August 1988 Chris Tidemann (pers. comm.) shot 21 individuals, all of which were males. As a result Chris believed the females were entering torpor (a mild form of hibernation) over the dry season when prey availability was lower. In contrast in June-July 1994, a roughly equal sex ratio was found in the 22 individuals trapped (12 males, 10 females; Lumsden and Cherry 1997). A similar pattern was found in May-June 1998 where the sex ratio was also roughly equal (61 males : 65 females; Lumsden et al. 1999). In contrast, during the wet season of 2005 (December), 73% of the 52 individuals caught were females (Lumsden et al. 2007). It was not known, however, if this was an accurate reflection of the sex ratio of the remaining population or was an indication of different foraging locations used by males and females.

It appears that females, at least during the breeding season, form communal maternity roosts while males tend to roost in small numbers (Lumsden et al. 2007). Maternity roosts located in 2005 contained a mean of 29.4 ± 16.5 individuals (range 11 – 54 individuals). While the individuals being tracked that led to the location of these roosts were female, the sex of the other individuals using these roosts was not known. It is possible that these roosts contained predominantly, if not exclusively, females, similar to many other species of tree hole roosting

bat (Kunz and Lumsden 2003). Although it was not possible to determine accurately the number of individuals in roosts used by males, it was believed that these roosts supported either a single individual or just a small number of bats (Lumsden et al. 2007). During the non-breeding season in June 1998 females roosted in colonies of up to 47 individuals, although this may vary since one female roosted solitarily on one day. Males roosted solitarily or in small groups (6 individuals), although again this may be variable as one male roosted in a colony of 28 individuals (Lumsden et al. 1999). Observations of the remaining communal roost in January 2009 indicated only 4 individuals were using this roost. These four individuals returned regularly during the night, using a pattern typical of females returning to suckle their young.

Implications for captive colony: If the males and females segregate during the breeding season it will be necessary to have multiple enclosures to enable females to be kept separate from the males. Within enclosures it will be necessary to have sufficient roosting boxes to enable individuals to roost solitarily if they choose to. As the numbers of individuals in maternity colonies in the wild in the past were typically fairly large, it may be necessary to have some level of critical mass of bats for breeding to be successful, but how many individuals this may be is unknown. It appears that the four females in the only known remaining colony are breeding, however it is not possible to determine the likely survival of their young.

4.8.7 Roosting habitat During the non-breeding season, Christmas Island Pipistrelles have been recorded roosting in a range of situations including under exfoliating bark on dead trees, under loose dead fronds of palm and pandanus trees, in hollows of large live trees and under strangler fig arms (Lumsden et al. 1999). The majority of the maternity roosts located in 2005/06 were under exfoliating bark on dead Tristiropsis acutangula trees (Fig. 8; Hoye 2006; Lumsden et al. 2007). Only one of the nine roosts located in 2005/06 is still being used in 2008/09; most had collapsed or been deserted by 2007 (DNP 2008). In an attempt to provide additional roosting sites CINP established seven artificial roost boxes in Sydney Dale near existing roost sites (refer Fig. 14). These have been modified recently to make them more suitable (see management section below). There is no evidence that they have been used as roosts by bats.

Implications for captive colony: Roosting boxes that simulate loose bark on dead trees are most likely to be accepted by the bats as roosting sites in captivity. Internal cavity dimensions need to be large enough to house the entire colony within an enclosure as it is possible they may choose to roost communally. By trialling a number of different nesting box designs it may be possible to determine what is most likely to be used in the wild, if additional artificial roosts were required in the wild.

4.8.8 Roosting behaviour Individuals used multiple roosts within the period of time they were radio tracked in both the 1998 and 2005 (Lumsden et al. 1999, 2007). Shifting frequently between a number of roosts within a defined roost area is typical of many tree hole roosting bats (Kunz and Lumsden 2003). During the non-breeding season, some individuals moved to a new roost each day, while others remained in the same roost for more than 8 days (Lumsden et al. 1999). During the breeding season in 2005 the short period of time the transmitters remained attached to the individuals precluded any investigation of roost shifting behaviour. Observations in January 2009 and examination of the detector data from the remaining roost indicate that in recent years individuals have remained in the known communal roost for extended periods of time but in the last year appear to have alternated between two roosts – the other of which has not yet been located. It is not known why individuals shift roost sites but a number of theories

Fig. 8. A maternity roost located in 2005 under loose bark lifting off a dead Tristiropsis acutangula that was used by a colony of 32 Christmas Island Pipistrelles (Roost 14).

include: to avoid predation, to reduce parasite load build up, to select the appropriate microclimate, or to know multiple roosts in case ephemeral roosts collapse (Lewis 1995).

Implications for captive colony: Multiple roosting sites will need to be provided within enclosures to enable animals to move between roosts.

4.8.9 Morphometric data This species is one of the smallest bats in Australia. Females are typically larger than males (Lumsden et al. 1999, 2007). Some information is available on the size and weight of individuals at different times of the year based on the four studies on this species that have trapped individuals and taken morphometric data (Table 3). The body weights for individuals caught in December 2005 were significantly greater than the weights of individuals caught in May-June 1998 (Lumsden et al. 1999, 2007). It is not known if this represents an improvement in the condition of the bats between 1998 and 2005 or reflects seasonal changes in weight in response to insect availability. Christmas Island experienced a severe drought in 1997 and early 1998 which may have led to a reduction in availability of insects (Lumsden et al. 1999). However, the weights of the bats in the two dry season sampling periods (1998 and 1994) were similar, despite 1994 receiving average rainfall (Lumsden and Cherry 1997).

Information is available on the condition of individuals during the breeding season based on the data collected during the wet season in December 2005. This provides information on pregnant and lactating females (Lumsden et al. 2007). Pregnant females weighed considerably more than lactating or non-breeding females (Table 3). The heaviest pregnant

female trapped was 5.6 g when first caught on 14 December 2005 and a massive 6.2 g, with parturition considered to be imminent, when retrapped on 22 December 2005. If the non- breeding weight of this individual was 4.1 g (the mean for non-breeding individuals), this would mean that the weight of the young (and associated fluids etc) was 2.1 g.

There is no data available on the size of the young when they start flying as no trapping has been undertaken during the months when young become independent (likely to be January and February). However, their forearm length is likely to be similar to the adults, with their weight approximately 70% of the adult’s weight.

For species of bats that undergo torpor, especially in temperate environments, the body weight of individuals typically fluctuates markedly: considerable fat reserves are stored before entering the period of torpor, which are then gradually used over the following months. For example, in temperate areas, body weights in autumn can often be an additional 40% above those in spring (L. Lumsden pers. obs.). Fluctuations in body weight in tropical bats is likely to be less but it is possible that they follow a similar pattern, putting on fat reserves during the wet season and gradually loosing weight over the dry season. This would match the pattern observed with the Christmas Island Pipistrelles. Interestingly, individuals in captivity often follow a similar pattern to their wild counterparts, irrespective of the amount of food they are given (L. Lumsden, pers. obs.).

Table 3. The weight and forearm length of individuals caught during various seasons and years. All females in May-June 1998, June-July 1994 and 1984 were in non-breeding condition. The December 2005 data is taken from Lumsden et al. (2007), May-June 1998 from Lumsden et al. 1999) and June-July 1994 from Lumsden and Cherry (1997), and March and September 1984 from Tidemann (1985).

Date Sex Reproductive Forearm Weight n condition (mm) (g)

Dec 2005 Male All 31.1 ± 0.6 3.8 ± 0.3 14 Female All 31.6 ± 0.6 4.5 ± 0.5 38 Pregnant 31.2 ± 0.6 5.5 ± 0.3 6 Lactating 31.7 ± 0.6 4.3 ± 0.3 25 Non-breeding 31.4 ± 0.7 4.1 ± 0.2 7

May-June 1998 Male All 30.7 ± 0.7 3.4 ± 0.3 58 Female All 31.2 ± 0.7 3.6 ± 0.2 61

June-July 1994 Male All 31.5 ± 0.8 3.5 ± 0.3 12 Female All 31.5 ± 0.7 3.7 ± 0.4 10

March/Sept 1984 Male All 31.0 ± 0.7 15 Female All 31.5 ± 0.7 11

March 1984 Male All 3.2 ± 0.4 3 Female All 3.2 ± 0.1 6

Sept 1984 Male All 3.1 ± 0.2 12 Female All 3.2 ± 0.1 5

Implications for captive colony: The available data will provide a benchmark against which captive animals can be compared, including seasonal variation, breeding and non-breeding weights and the likely weights of young.

4.8.10 Health condition The most detailed information on the health status of the Christmas Island Pipistrelle is from the 2005/06 study. A summary is provided here and full details can be found in Lumsden et al. (2007). The health of all individuals trapped during this study were assessed by a wildlife veterinarian. Individuals were examined externally for obvious signs of ill-health (e.g. the presence of wounds, lesions or obvious discharges). External examinations of the 52 trapped individuals revealed no obvious external indication of ill-health. They all appeared in good physical condition, their fur looked healthy and they had high body weights (Table 3). In addition, the majority of females were breeding. Had the females been in poor condition or their health compromised, it could be expected that they would have forgone reproduction to increase their own chance of survival (Barclay et al. 2004).

Very few external parasites have been found on the Christmas Island Pipistrelle (Lumsden et al. 2007). The only ones found were a small number of mites on the wing membranes of 10 of the 52 individuals (19%) trapped. Microbats on mainland Australia typically have mites, ticks and bat flies (Jackson 2003). No ticks or bat flies were observed on the pipistrelles. Faeces were collected to determine if they had internal parasites. Examination of the faecal smears and floats revealed no evidence of internal parasites (Ian Beveridge, Melbourne University, pers. comm.).

Swabs were taken for viral and bacteriological testing from the external opening of the respiratory system, urogenital area and wing membrane. The swabs sent for bacterial analysis revealed no bacteria when examined microscopically and the bacterial culture produced only light growths of mixed skin flora. These results indicate that no significant bacterial pathogens were detected in these samples. Similarly, no viruses were detected in the swabs sent for viral isolation.

Blood samples were taken from a subset of individuals. The only vein that was large enough to obtain a sample was the lateral tail vein. In some smaller individuals (predominantly males), even this vein was too small to successfully obtain a sample. Blood was taken by piercing the vein with a sterile 30 gauge needle and the blood collected in a micro-pipette tube, from which a blood smear was taken. The blood smears showed the red blood cells to have mild to moderate polychromasia with occasional nucleated red cells. Polychromasia and nucleated red cells are associated with regenerative anaemia, and in domestic animals are usually an indicator of previous or chronic disease. However, similar sized bats from Victoria were also found to show a level of polychromasia, and this may be a normal feature of microbat blood rather than indicating previous or chronic disease (Philippa McLaren, Gribbles Veterinary Pathology Laboratory, pers. comm.).

The morphology of the white blood cells was normal and the number of platelets was adequate. However, the blood was considered to be leukopenic, i.e. lower than expected numbers of white blood cells were found. Leukopenia has been associated with a range of diseases such as infectious conditions and toxic insults. The white blood cell count was predominantly 1-2 x 109/L which is lower than other Australian native mammals which typically range from 2-15 x 109/L (Clark 2004). Victorian forest bats also had low white cell counts, although not quite as low as the Christmas Island Pipistrelles (2-5 x 109/L). Therefore, it appears that microbats may have lower white blood cell counts than other mammals, but that the pipistrelles are somewhat lower again. The significance of this finding however remains unclear, as it is not known if these levels are typical for this species, typical for island species

that are not exposed to high levels of disease factors, or if in fact it does indicate some form of ill-health.

No blood parasites were found. Blood parasites are considered a possible cause of the extinction of the two species of endemic rats on Christmas Island at the start of the 20th century (Pickering and Norris 1996).

Implications for captive colony: The available data provide a benchmark against which captive animals can be compared. As parasite loads are very low, treating for parasites is not considered necessary and would only be undertaken if the need arose.

4.9 Captive management plan This section provides a brief outline of the captive management requirements of the colony, in relation to housing, feeding and monitoring. A full captive management plan will need to be developed in future, however the aspects covered here are the ones considered most relevant to the planning for the program. The captive management requirements outlined below are based on the colony being located on Christmas Island. If the colony was to be on the Australian mainland, especially in a temperate region, the enclosures would need to be in a temperature, humidity and day-length controlled facility to simulate the conditions on Christmas Island.

Due to the critically low number of individuals remaining in the wild and the uncertainty around being able to catch sufficient individuals to start an effective captive breeding program, it is recommended that the program is undertaken in two phases: a) the initial emergency rescue and establishment of a temporary facility; and b) if sufficient animals are collected, the establishment and on-going maintenance of the longer term facility. While it will be important to commence the planning for the longer term facility now, as the rescue phase needs to be undertaken immediately while animals remain in the wild, this component can not wait until the planning, funding and building of the long-term facility is complete.

4.9.1 Housing Temporary facilities for the emergency rescue phase can be established very quickly. A room or building is required that can be made predator-proof, has access to power and water, is secure, and can be kept at a temperature and humidity similar to in the forest. It is proposed that the bunkhouse room at the Pink House (Parks Australia Research Station) is converted into a temporary pipistrelle holding area (Fig. 9). This would require removing the bunks and partitions, sealing any gaps and ensuring the room was predator-proof. Within this room fly- wire ‘tents’ would be used as the holding enclosures for the bats (Fig. 10). The type of enclosures proposed are 3.5 x 3.5 m in floor size and two of these would fit into this room. As it is anticipated that only low number of individuals are likely to be caught, this space would be sufficient in the short term to hold up to 15 individuals, with the capacity to keep males and females separate as required. Nesting opportunities and feeding trays would be established within the enclosures. A number of smaller enclosures could also be used if required (e.g. 1 x 0.5 m), however, individuals would not be able to fly in these. Shade cloth would need to be installed at the edge of the verandah so that late afternoon sun did not penetrate room housing the enclosures.

These enclosures would be large enough for individuals to fly, and are certainly adequate for short-term housing, however larger enclosures would be required for long-term housing, especially when the young were learning to fly. In addition they would not provide sufficient space to house colonies of females to simulate communal maternity roosting conditions, and allow them all to fly. Therefore if the decision was made at the end of the emergency rescue phase to proceed to the long-term phase, outside flight enclosures would need to be built. It would not be desirable to house the colony in these temporary enclosures for more than several months.

The long-term facility would consist of a number of aviary-style flight enclosures set within a forest clearing in a quiet, secure location. The Pink House grounds would be an ideal location. The small clearing around the building is surrounded by forest (Fig. 11). There is an abundance of insects in this area that would be suitable for attracting into the enclosure as supplementary food. The area is quiet and away from noise and industry, and has access to power and water. The flight enclosures could be set along the edge of the forest which would provide significant protection from wind and sun exposure, and would at least partially simulate being in the forest without having to be under the canopy. Setting the enclosures

Fig. 9. The Parks Australia Research Station (the Pink House) in which the temporary facility could be established.

Fig. 10. Fly-wire mesh shelter that could be used as temporary holding enclosures within a predator-proof room – this one is 3.5 m x 3.5 m at ground level and about 2 m high, with an enclosed floor.

Fig. 11. The edge of the forest at the Pink House clearing where flight enclosures for the long-term facility could be built. The edge of the forest would provide some protection from wind and sun exposure.

within the forest would provide a preferable microclimate, however, this would require some clearing of vegetation and there would be a greater risk of tree or branch fall onto the enclosures than if they were constructed at the forest edge. Quick growing plants could be established along the exposed length of the enclosures to provide greater protection from the elements, if required. It will be important to test the microclimate provided by various locations at the Pink House and compare these to the roosting area in the Sydney Dale. Therefore it is recommended that several weather stations are set for a few months near the known roost, the edge of the forest at the Pink House and within the forest there. This will inform the sighting of the enclosures and the requirement for additional shade.

The long-term enclosures will need to be designed in conjunction with zoo experts. Existing designs, such as those recently used for Orange-bellied Parrots at Healesville Sanctuary, Zoos Victoria, can be used as a basis. We recommend that six flight enclosures,10 x 4 x 2.5 m in size are constructed. As the number of individuals initially will be low until the captive population builds up, three enclosures could be constructed initially to spread the costs over several years. In conjunction with the flight enclosures there needs to be a separate room constructed to be used as a work area where animals can be examined by the veterinarian and where the bats food is prepared. This workspace area needs power and water. An additional room will be required to house the mealworm colony. All flight enclosures need to be accessible from this workspace so that animals being moved from the enclosures to the examination area remain within the enclosed facility.

It is recommended that the flight enclosures are constructed from a double mesh. The outer layer needs to be a sturdy wire mesh with the inner layer a softer, less abrasive material, such as polyethylene plastic mesh, or equivalent, which unlike wire mesh would not be abrasive to the bats feet, and would not be corroded by bat urine (Jackson 2003). An elevated roof will need to be built over the whole enclosed area to provide shade and protection against branch fall. A solid metal roof would provide the necessary level of protection, however it will require coating with a sound buffering material (e.g. several layers of shade cloth) to reduce the noise made by heavy rain on a tin roof. All entrances to enclosures will need to be double doors to prevent animals escaping. Within the enclosures, multiple roosting boxes will be established as well as multiple feeding platforms with food and water trays.

One of the most important requirements will be to ensure the enclosures are predator-proof, keeping out all potential predators such as centipedes, wolf snakes and rats. It will also need to be made crab proof so that Robber Crabs are not able to climb on it. Engineering advice may be required to determine the most effective method of predator proofing.

As it might be necessary to experiment with various social aggregations of the bats for optimal breeding opportunities, it will be important to either have a large number of enclosures, or the capacity to partition the larger enclosures into smaller spaces to house a smaller group of individuals. There will also need to be smaller containers to house sick individuals or to acclimatise new individuals brought to captivity.

4.9.2 Feeding requirements It is recommended that the captive bats are fed on a mixture of mealworms (supplemented with a Wombaroo mix) and free flying insects from the wild (as outlined earlier). Microbats typically eat large quantities of insects in the wild, consuming over half their body weight in a night. Captive bats will probably eat less as they are likely to fly for shorter periods, however large quantities will still be required. In New Zealand, to maintain the captive colony of 400 Short-tailed Bats, 80,000 mealworms were required each week (with a total of 2.5 million used during the 4 month program; Sedgeley and Anderson 2000). For a colony of 30 Christmas Island Pipistrelles, it is estimated that over 2000 mealworms will be required each week. While it may be possible to fly extra mealworms in from the mainland on a regular basis, it will be important to establish a large mealworm breeding colony to ensure there is always a ready supply of mealworms. It may be necessary to house the mealworm breeding colonies in an air conditioned environment. Mealworms are very susceptible to developing mould or fungal contamination in humid conditions which would be detrimental to the health of the bats (Lollar and Schmidt-French 1998). However, mealworm growth is slowed in cold conditions so a balance will need to be found. Once at the optimal size, mealworms can be stored in a fridge until they are ready to be used. For several days before feeding the mealworms to the bats, the mealworms should be placed in a high concentration of Wombaroo Insectivore Rearing or Small Carnivore mix to gut load them with the required vitamins and minerals. Additional supplements may need to be provided to the females during the breeding season as high levels of calcium are required for lactating females and this is critical for the ossification of the bones of the young (Barclay 1995).

The process for training bats to feed on mealworms, starting with hand feeding and then training them to self feed from trays, is described earlier. Feeding platforms will need to be installed (at approximately waist height) on which large trays containing mealworms would be placed just before dusk each night. These trays need to be large enough for the bats to be able to land on or crawl into, with smooth sides to prevent the mealworms from escaping. Bats can either fly to the wall immediately above the tray and hang from there to feed, or crawl into the tray to feed. Shallow water dishes (to reduce the risk of drowning) would also be on the feeding platforms. Multiple feeding platforms in each enclosure will allow individuals to feed separately if desired. If the mealworm trays are large enough, multiple individuals could

feed concurrently. Providing abundant supplies of food at the feeding stations is likely to reduce the chance of aggressive encounters between the bats while feeding. Remaining food should be retrieved and measured each morning to monitor overall food intake.

It will be important to supplement the artificial food source with wild insects attracted to black lights to ensure captive bred individuals learn to catch prey in flight and to promote regular exercise. However, it is typically not possible to ensure sufficient suitable wild food and so a regular supply of mealworms will be required. In addition, as the cause of the decline of the pipistrelle is not known, toxicological investigations of the wild insects should be conducted to ensure there are no environmental contaminants or pesticides carried by the island’s insects that would continue to affect the captive animals. As a precautionary approach it is recommended that the captive bats are not fed on wild insects until this testing is completed.

4.9.3 Monitoring protocol To monitor individuals’ health and reproductive condition, each bat will need to be regularly weighed and examined. However, a balance is required between adequate checking to allow signs of ill-health to be detected, and minimising handling to reduce stress which could interfere with breeding, such as increasing the rate of abortions or rejection of vaginal plugs. The efficacy of using remote control infrared video cameras in enclosures and/or nest boxes to observe behaviour should be investigated. Such systems have been used successfully in zoos, and in wild situations, for example within a maternity cave for the Southern Bent-wing Bat at Naracoorte in South Australia. A full monitoring protocol will need to be developed by the veterinarian, keepers and bat biologist and updated regularly as experience increases.

All captive bats will require marking in some form to enable individual recognition so that their health can be regularly assessed and genetically optimal breeding aggregations can be formed. Long-term marking techniques have not yet been tested on this species. Banding using bat bands supplied by the Australian Bird and Bat Banding Scheme would be the simplest form of marking. Banding can cause injuries to the forearms of some individuals (Baker et al. 2001) and hence if this technique was used, the condition of the bands would need to be closely monitored and bands removed if they started to cause irritation. Alternatively, it may be possible to use passive implantable transponders (PIT tags). These have been used successfully on larger bats. However, the smallest implantable tag currently available is 11 mm long, which may be too large for this small species. Fur clipping has been used for short term individual marking (Lumsden et al. 1998, 2007), however, this mark is only likely to remain visible for a couple of months.

It will be important to undertake a genetic analysis of each individual to determine if there are optimal breeding combinations to enable the maximum genetic diversity in the captive population. Genetic analysis can be undertaken from either hair follicles or a small amount of wing tissue. Both can be taken from bats with little impact to individuals. These samples are then sent to a genetics laboratory for analysis.

Individuals will require regular veterinary examinations. Viggers et al. (1993) reviewed the importance of disease in the failure of reintroduction programs for a range of endangered species. They argued that the planning of reintroduction programs should include an examination of the potential impacts of disease on extant populations and on animals targeted for release. A number of steps were outlined that may minimise the risk of disease affecting the success of a captive breeding and reintroduction program: • isolating animals in quarantine for the maximum incubation period of any potential disease so that developmental signs of the disease can be observed and treated where possible; • undertaking regular clinical examination by an experienced wildlife veterinarian; • examining faeces for the signs of internal parasites;

• undertaking haematology and serum biochemistry profiles to aid in the diagnosis of organ dysfunction and disease; • serological screening and examining microbial cultures for the detection of infectious diseases; and • a full post-mortem examination of any animals that die in captivity, including histopathology and microbial culture.

The wildlife veterinarian involved in the project will determine what tests should be undertaken (given the problems associated with the collection of blood) and how frequently such tests need to be done.

4.10 Obtaining bats for the captive colony There are a number of issues relating to when and how to take animals into captivity. Due to the critically low number of individuals currently remaining in the wild, it will be essential to establish the temporary facility immediately, prior to the building of the longer-term facilities.

4.10.1 Recommended number of individuals to be taken into captivity As there have been no conservation-orientated captive management programs undertaken for small microbats, there is little in the literature to guide the decision on the optimal number of individuals to take into captivity. For other faunal groups a founding colony of 50 individuals is often recommended for a captive breeding program (G. Gillespie, pers. comm.). However, if the current estimate of less than 20 individual Christmas Island Pipistrelles remaining in the wild is correct, in reality the number that can be brought into captivity will be determined by the number that can be caught. If further work reveals that there are more individuals in the wild than this number, it is recommended that up to 30 individuals are taken into captivity, with half to two-thirds of these being females.

Ideally, in addition to the captive colony, a ‘viable’ population should be retained in the wild, and intensive research be undertaken to investigate the cause of the species’ decline. If all animals were taken from the wild, it would reduce the options for determining the cause of the decline. If disease or ill-health are contributing to the decline then it will be possible to investigate this using the captive animals, and it may be possible to experimentally test some of the other theories on why the species’ is declining. However, as the number of individuals remaining in the population appear to be so low, all individuals that can be caught should be taken into captivity. While it will be more difficult to determine the cause of the decline if no individuals remain in the wild, some options remain and attempts should continue. In reality it will not be possible to catch every individual, so some will remain in the wild, but how long they will survive is not known.

4.10.2 Timing of taking animals into captivity The rapid decline of the population and the prediction that the species could go extinct within months dictates the urgency of the situation. If the current colony in the only known roost contains four breeding females with four dependent young (as assumed from observations in January 2009), ideally animals should be caught soon after the young have commenced flying, while all eight individuals are still alive. This therefore has to occur by March-April 2009. This timing would also have the least interruption of the breeding cycle as it would be after the young were independent but before mating, which occurs in June-July. Delaying the capture of individuals any longer than this is likely to result in fewer individuals available for capture and greater difficulty in catching them.

4.10.3 Techniques for catching animals to take into captivity Trapping sufficient individuals to establish the captive breeding colony will require a combination of approaches. The difficulty in trapping bats at their foraging areas in January 2009 indicates it is going to require considerable time, effort and ingenuity. Although not successful in January, continued efforts will be undertaken in known foraging areas using a large number of traps and mist nets, and modifications of the standard techniques (such as forming an enclosed mistnet flight tunnel where the entrance can be closed once the bat has flown down the tunnel). Additional detector work will be undertaken at other locations in an attempt to find more foraging areas where bats could be trapped. If bats are captured, these individuals can either be taken into captivity or fitted with a radio transmitter, with the aim of tracking them back to their roosts, where it may be possible to trap larger numbers of individuals. If this was to be done, an option for getting high above the ground (such as in a

helicopter) would need to be considered to increase the chances of finding the bats at their roosts.

Another focus for capturing animals will be on developing techniques for catching bats at roosts. New techniques will need to be developed with the aim of catching entire colonies. The most successful is likely to be a catching mechanism that covers all the exit points from the roost but with a shoot leading into a holding bag. As the bats try to leave the roost they would be confined within the trap with no option but to fall into the bag below. Similar types of traps have been used successfully at roost entrances to buildings and trees (Kunz and Kurta 1988). Such trapping would need to be undertaken extremely carefully due to the decayed nature of the roost trees and the flighty nature of the bats. Depending on how many colonies were found, a decision would need to be made as to whether to take all individuals from a single colony (which might survive well together in captivity due to their social bond, but might also be more closely related and hence have lower genetic diversity), versus trying to take small numbers from each colony found (assuming multiple colonies are found). Due to the small number of individuals remaining in the wild, it is likely that it will take some time to catch the required number of animals to establish the captive colony. Trapping at communal roosts is more likely to catch females than males. If these roosts exclusively contain females (which is likely during the breeding season but possibly more mixed outside this period), it will be necessary to trap males while they are foraging at night in traps and mist nets.

In an attempt to locate new roosts, extensive searches will be made throughout the known roosting areas to locate the typical roost trees used by females, i.e. loose sheets of bark on dead Tristiropsis acutangula trees. CINP have commenced this process and have so far identified 130 potential roost trees. Bat detectors are being placed at a subset of these in an attempt to locate new roosts. This work will be expanded as part of the initial capture phase by bringing a large number of detectors to the island.

4.11 Funding costs and opportunities The full cost of the captive breeding program cannot be determined in detail until the location, the facility design and size, and staffing requirements are decided. However, outlined below are rough estimates (Table 4). These include the cost of construction of enclosures and the ongoing maintenance of the colony, with estimates from Graeme Gillespie and Russel Traher from Zoos Victoria. These costings are based on the facility being required for 10 years, and hence the costs are for the full 10 year period. If the program is required to run for longer, additional funds will need to be sought.

While the main aim of this report is to investigate the feasibility of a captive breeding program, these costings include a component to identify the cause of the decline. The establishment of a captive colony can not be seen as a total solution to the problem – what it aims to do is to provide insurance against the imminent extinction of the Christmas Island Pipistrelle and gives more time to determine the cause/s of the decline and mitigate these cause/s, such that individuals can then be reintroduced to the wild. Therefore, in addition to the work outlined in this report, a long-term strategy will need to be put into place to direct the recovery of this species. This will include monitoring and management of any animals remaining in the wild, management orientated research to determine the cause of the decline, mitigation actions to reduce these threats, reintroduction of captive animals to the wild and extensive monitoring for a number of years to determine their survival, breeding success and the long-term prospects for improved conservation of the species. In the costings below, a component is included for targeted research to identify the cause/s of the decline. This is an estimate only and depends on how quickly a cause can be identified. Until the cause is determined it is not possible to estimate how much it will cost to mitigate it. However, additional funds for mitigation measures will be required in the future. Additional funds (not included in Table 4) will also be required for the reintroduction program and subsequent monitoring of the re-established population.

It is essential that the emergency rescue phase is undertaken immediately while there are still some animals alive in the wild. It is therefore recommended that the first phase is initiated prior to all aspects of the long-term facility are resolved. There will however need to be an in principal support for the long-term project and the capacity for a rapid decision on if the long- term facility was to proceed. It is recommended that after 10 weeks of intensive effort by a team of bat biologists attempting to trap bats to take into captivity and acclimatise them to captivity for the emergency rescue phase, an assessment is made as to whether to proceed. This decision would be based on how many individuals had been trapped by that stage and the prospects for trapping more individuals. If, after 10 weeks of intensive effort, no individuals have been caught, then attempts should cease. If a small number of individuals had been caught a decision would then need to be made as to whether there were sufficient to form the basis of a viable captive colony. If it was decided that there were not enough individuals and that the trial should cease, those individuals could then be released back into the wild.

Due to the low numbers of bats in the wild it will take an extended period of time to catch sufficient animals to take into captivity for the emergency rescue phase. The costings in Table 4 are based on four experienced bat biologists undertaking field work for 6 weeks. Once bats are caught a veterinarian would visit the island for 2 weeks to assess the health of the animals. A further 4 weeks is allocated for two people, an experienced bat biologist and a zoo keeper, to acclimatise the bats to captivity, and teach them to feed on mealworms. If a disease or health issue was identified, there would be a requirement for the veterinarian to remain longer. Transport, accommodation, living expenses and equipment costs have also been included.

Table 4. The estimated cost of establishing and maintaining a captive breeding program for the Christmas Island Pipistrelle on Christmas Island for ten years.

Item Cost $K

As there will be considerable organisational and management actions needed for the construction of the long-term facility, a project manager has been included to oversee the process.

The costs for building the facility are based on the costs for building the Orange-bellied Parrot facility at Taroona, NSW and Healesville Sanctuary (Zoos Victoria). This is based on six enclosures (10 x 4 x 2.5 m), plus a treatment room where bats can be examined, and a room to breed and maintain the mealworms. It also includes the drafting of the design specifications, and the site service requirements (electricity, plumbing). To spread the costs, three flight enclosures could be built initially (sufficient to hold two groups of breeding females with males kept separately), with additional enclosures built as the number of bats increased.

To run the captive breeding colony, two experienced keepers will be required to maintain and monitor the bats, and to breed and maintain the mealworms and collect the natural food components. The colony will need to be attended 365 days a year, and hence it cannot be undertaken by a single person. The second keeper will be needed to work on weekends, and when the first keeper is sick or on recreation leave. It is likely that once the protocols have been established and the bats fully acclimatised to captivity, but while the number of individuals in captivity is still low, that the second keeper could either work part-time or be involved in other projects. It may be possible for a person currently living on the island, provided they have appropriate animal husbandry expertise, to fill this second keeper role.

Advice from Graeme Gillespie (Director of Conservation at Zoos Victoria) is that for a number of reasons it would be preferable for Zoos Victoria (if they were the selected zoo) to staff and run the facility, rather than the staff being employed by CINP. This is because running captive breeding programs is their area of expertise and there are a range of support services that the staff would need to draw on from the zoo. They could also ensure professional standards and be able to undertake training more effectively. Furthermore, it would enhance Zoos Victoria role as a collaborator which also provides potential showcasing, educational interpretation, and potential fund raising opportunities through the zoo campuses. If the keeping staff were zoo employees it would be easier to rotate staff throughout the 10 year period as required. If

the second keeper was a long-term island resident, this would provide continuity of knowledge in the program.

It should be noted that for little additional cost (possibly an extra $100K), it would be possible to incorporate a captive breeding program for the critically threatened reptiles on the island.

It may be possible to obtain funding for the captive breeding program, in addition to that provided by the Commonwealth Government. Possible options are: • Christmas Island Phosphates have already made a commitment to undertake management actions to help improve the conservation status of the Christmas Island Pipistrelle and may contribute to a captive breeding program. • Bat Conservation International is an international organisation committed to bat conservation throughout the world and have links with many philanthropic organisations in USA. • IUCN Chiroptera Specialist Group could be contacted re any suggestions or potential assistance they could provide. • Philanthropic organisations in Australia – David Middleton has links to philanthropic organisations and could be approached to provide advice on possible options. • Zoos Victoria may be able to assist in raising funds if they were a formal collaborator. • Corporate sponsorship. • Additional funding may also be available as part of tourism to the island.

4.12 Discussion on the feasibility and likelihood of success of a captive breeding program for the Christmas Island Pipistrelle As there are a number of unknowns in relation to a captive breeding program for the Christmas Island Pipistrelle, the following points below are provided on the feasibility and likelihood of success of such a program, as part of a discussion with CINP, DEWHA Threatened Species and Communities Section, and in response to questions raised by DNP when commenting on the draft report. Although answering these questions in this format results in some repetition with earlier sections of this report, it has been presented this way to provide as much information as possible, in a single location, to assist in making an informed decision on the establishment of the captive breeding program.

When would the emergency rescue phase and temporary facilities need to be in place? Due to the low numbers of individuals remaining it is critical that this is undertaken by March- April 2009, timed to occur soon after the young are independent to increase the pool of bats available to be caught. Delaying the commencement of this phase, will lead to fewer bats remaining and it will be harder to catch them, thus reducing the likelihood of success and increasing the cost of the program.

How long would it take to construct the temporary facility (time between decision and completion)? If the fly-wire mesh tents recommended earlier were used, these could be bought in Australia and flown to the island quickly, and established immediately in the Pink House. Removing the furniture and predator-proofing the bunkhouse room would only take a day. The main requirement is to contract a team of bat specialists to catch the bats. In addition, an agreement would need to be made with the selected zoo to provide a veterinarian and keeper at short notice once bats were caught. As it could take several weeks before bats were caught, it would be more cost effective to bring the zoo staff over as required, rather than from the commencement of the project. A formal partnership or contract would need to be organised with the selected zoo. Ten weeks have been allocated for the emergency rescue phase – 6 weeks to locate and trap the bats and 4 weeks to acclimatise them to captivity and assess their health. It is anticipated that it should be possible to commence the trapping component of the emergency rescue within one month of a decision being made to undertake this rescue phase.

When would the permanent facilities need to be in place? While the fly-wire mesh tents suggested above are suitable as temporary enclosures, they are not large enough for optimal long-term housing – the bats would need larger enclosures for sustained flight, especially when the young are learning to fly, plus adequate space is required for housing colonies of females to simulate communal maternity roosting conditions. Ideally the temporary enclosures should not be used for more than 3-6 months. However, there are no concerns in using them for this length of time, while the permanent enclosures were built. The permanent enclosures would need to be established before the first breeding season – i.e. November – December.

How long would it take to construct the permanent facility, including any planning/building approvals (i.e. time between decision and completion)? Advice would be needed from Christmas Island as to the requirements and timelines for planning and building approvals. An architect would be required to draw up the design specifications of the enclosures, in conjunction with bat biologists, zoo staff and CINP. If the required materials were not available on Christmas Island they would need to be freighted in. Builders, electricians and plumbers would need to be contracted to build the facility and install power and water. The plans are likely to be reasonably straight forward and so it should be possible to construct the facility within several months after receiving the plans. There are

existing plans for enclosures, such as those used for Orange-bellied Parrots, that could be modified to suit the Christmas Island Pipistrelle, which would speed up the process.

When do staff need to be in place? The catching team for the emergency rescue phase will need to commence searching for bats and trapping by March-April 2009. Due to the significant concern by bat researchers in Australia and overseas for the conservation of this species, there are a large number of people that are willing to assist with this process at short notice. We have no doubt that we could assemble the required number of experienced people within weeks. A large amount of equipment would be needed to be taken to the island, however, as there are now additional flights to the island, the restrictions on the amount of air freight are likely to be eased. This will enable, the majority, if not all, of the equipment to be taken across at the same time as the people and hence reducing delays.

The greatest chance of success is if the trapping commences in March-April 2009. This years young will be independent by this stage and free flying, so there will be a larger number of individuals to potentially catch. If the catching phase was delayed, there is a risk that these young would have died (mortality rates for most species are typically higher for first year animals than for adults). In addition, it is critical that bats are caught and acclimatised to captivity well before the mating season commences in June, to maximise the likelihood that breeding is successful in the first year of the captive program. Six weeks could be needed to locate additional areas where bats can be caught and to catch as many individuals as possible. Three years ago when the numbers of bats were higher, this would have been easily achievable within 10 days. However, now the numbers have declined to such low levels, it is likely to take many weeks. It should be apparent after 6 weeks whether sufficient animals can be trapped to form the basis of the captive breeding program. At that stage a decision could be made whether to continue with the program and secure funds for the long- term program.

Once the long term program is approved, the project manager should be appointed as soon as possible, with the appropriate person already selected and on standby. This person would coordinate the planning and implementation components of establishing the long-term facility, with advice from the bat biologists and zoo staff. This will include overseeing the design and construction of the facility. A veterinarian and one keeper would need to be on standby to fly to the island as soon as bats were caught. If these were existing Zoos Victoria staff it would facilitate this occurring quickly. These staff would need to overlap with the bat specialist leading the effort to catch the bats, so that a full exchange of information was possible. The second keeper could be appointed as required. Once there was a decision to proceed with the long-term program it is envisaged that Zoos Victoria would appoint staff to fill the keeper role/s and the permanent zoo staff would return to Melbourne.

What is the likelihood that sufficient bats will be able to be caught to form a viable captive colony? It is apparent that the numbers are now critically low and, while not impossible, it will be a challenge to trap sufficient individuals to form a viable captive colony. When numbers were higher sufficient individuals could be caught by using harp traps. However, now that numbers are so low a range of approaches will be required, e.g. trapping at roosts and trapping in foraging areas using a mistnet tunnel (i.e. modifying nets to form an enclosed tunnel from which there is no escape once entered). It is believed that the mistnet tunnel would have a much higher chance of trap success than the standard mist net formation used previously in foraging areas, where the bats can fly up to and avoid the net by flying over or around it if they detected it. A similar principal is highly successful for catching very manoeuvrable bats that normally avoid nets within caves, i.e. completely block a cave passage with a mistnet and have a second net attached to the roof that can be unfurled after a bat has flown past. The bat is then confined between the two nets and its flight space can gradually be restricted

further until it is trapped in the net. The Christmas Island version of this would require constructing a mistnet tunnel by joining a number of nets together with one end, the sides and roof completely sealed and the opening could be closed with a net that could be quickly unfurled. In watching the behaviour of the pipistrelles in January 2009, they would fly up to the net and circle several times in front of it before flying off, so this would give sufficient time to close the entrance to the tunnel.

In addition to trapping in the foraging areas it will be necessary to trap individuals at the roost. Techniques will need to be developed and tested thoroughly on trees with similar characteristics before it is undertaken on the main roost tree. It is believed that it will be possible to develop a suitable method to do this. To trap at the roosts it will be critical to locate the alternative roost tree/s being used by the colony that uses Roost 565. By bringing four experienced bat biologists (plus a number of volunteers) to the island this gives the best chance of success, in conjunction with the work CINP staff are currently doing to locate potential roost trees. It is because the likelihood of catching large numbers of individuals is low, that we have recommended the two phase approach, i.e. to wait to make the decision on the long-term program until the emergency rescue phase is in progress and it can be determined if it will be successful. There will however, need to be in principal support from potential funding bodies prior to the rescue phase, so that if bats are trapped, the long-term phase can proceed immediately.

The above discussion on the likelihood of catching sufficient animals to commence a captive breeding program is based on this trapping being undertaken as soon as possible. As the numbers of bats are continuing to decline at a rapid rate, every month that there is a delay in commencing this process increases the risk of failure. The greatest chance of success would be if the program could start immediately (i.e. by March-April 2009).

What is the likelihood that the bats will acclimatise and survive in captivity? Most bats acclimatise to captivity very well, including other species of pipistrelles. Due to the extensive experience available in maintaining bats in captivity, we are confident that this will be successful. We are very confident that the bats will adapt to captivity, learn to feed, maintain weight etc. In general bats do not suffer from capture or handling stress and readily commence feeding on mealworms and can be maintained on this diet for long periods of time. Although the Christmas Island Pipistrelle has not previously been kept in captivity there is no evidence to suggest that it will behave differently from the many other species of small insectivorous bats, including other species of pipistrelles, that can be readily kept in captivity. This is assuming that the bats are healthy when brought into captivity (i.e. that the decline in the wild is not due to some form of ill-health or disease). If there are existing health problems when the animals are caught, the success rate will depend on how quickly this can be identified and rectified. All the individuals caught in 2005 were healthy and in good condition, and so the expectation is that this will also be the case in 2009. If there was however evidence of ill-health, individuals from different colonies (i.e. individuals that may not have roosted together in the wild) would initially be kept separate to avoid the risk of contaminating other individuals.

During normal trapping exercises, bats have a very low risk of being injured. For example in 30 years of trapping insectivorous bats, LL has trapped over 35,000 individuals with a death rate of well under 0.1% (mostly resulting from predation within the trap). No deaths have been attributed to capture stress. Of the 194 Christmas Island Pipistrelles trapped since 1994 only one died (0.5%): an individual that succumbed in the trap to Yellow Crazy Ants at a time prior to the threat of this species being recognised (Lumsden et al. 2007). The modified capture techniques that will be required to catch the bats now that numbers are so low, especially trapping at the roost, do have slightly higher risks involved. However, these risks are still considered to be low, and much lower than leaving the animals in the wild. To minimise these risks further, trapping in foraging areas will be attempted before trapping at roosts. Trapping

in foraging areas, using the mistnet tunnel approach, is considered to have virtually no risks of injuring the bats, and is as safe as the standard mistnetting technique, but will be more effective.

What is the likelihood that the colony will breed and form an effective captive population? If the animals are in good health (i.e. assuming that disease isn’t a factor in the decline), if the enclosures are large enough to enable the formation of the optimal social groupings, and if the temperature, humidity and day light are natural by having the animals in outside enclosures in, or close to, the forest, then there is a high chance of successful breeding and that an effective breeding population could be maintained. Depending on how small the founding population is, the genetic diversity will need to be considered. As the current genetic diversity in the wild is not known it is not possible to speculate on whether there will be sufficient genetic diversity to form a viable captive population. However, other species have maintained a high genetic diversity despite starting with a small founder population (e.g. the Rodrigues Fruit Bat captive population commenced with just 18 individuals and is now approximately 700 with a high level of genetic diversity, O’Brien et al. 2007; the Australian population of European Rabbits commenced with the introduction of just 25 individuals, Rolls 1969).

Many species of bats commence breeding in their first year and breed throughout their life, and the existing information available on the Christmas Island Pipistrelle suggests it also follows this pattern. Therefore no individuals will be too young or too old to breed, and every female would have the potential to breed each year. The sex ratio of the captive colony is not critical. Males mate with multiple females and so as long as there are at least a few males present there will be sufficient to mate with all the females. The number of females is more critical than the number of males. Given that the communal roosts are most likely to be exclusively, or predominantly female, there is a greater chance of catching females than males at roosts. Trapping in foraging areas using the mistnet tunnel method outlined above is likely to be critical for trapping males.

Is there sufficient knowledge about the feeding, housing and breeding of microbats in captivity to establish a captive breeding program? There is extensive knowledge and experience on the keeping of microbats in captivity, both in Australia and overseas, as summarised above in the review of captive breeding programs. Books and book sections have been written specifically on this topic (e.g. Hall 1982, Hopkins 1990, Barnard 1995, Lollar and Schmidt-French 1998, Jackson 2003). While there is little experience in keeping microbats in captivity in zoos in Australia, wildlife carers have extensive experience and this expertise can be drawn on for this program. For example, collectively wildlife carers in Australia would have cared for tens of thousands of individuals of microbats. We believe that there is currently sufficient information on the feeding and housing of microbats in captivity to successfully commence a captive program, assuming that people experienced with microbats are used to establish it. Zoo staff that are experienced with other faunal groups, but not with bats, should not be used without adequate training by experienced bat researchers and carers. Staff from Zoos Victoria do have experience with microbats, whereas many other zoos do not.

In addition to being successfully maintained in captivity, many species have successfully bred in captivity, with the key to success being providing suitable environmental conditions (i.e. large flight enclosures with roosting opportunities), physical conditions (i.e. providing a balanced diet using supplemented mealworms and wild insects) and social organisation (by simulating the seasonal colony composition found in the wild).

Will it be possible to provide appropriate habitat and protection from predators (e.g. in a tent)? The fly-wire mesh tents recommended for the temporary enclosures would provide adequate space and habitat for short-term housing. These enclosures would provide sufficient space for flight for the adults, although the larger, long-term enclosures would be required when they were forming breeding colonies and to facilitate the young learning to fly. However, for the short-term (i.e. up to 6 months) the tents are more than adequate. LL has kept bats for several months in these tents and then successfully released them to the wild. These tents will need to be established in a predator-proof room, as the tents alone could not ensure protection from all predators. In examining the bunkhouse room at the Pink House, we believe that it will be possible to adequately predator-proof this room by blocking the vents and any gaps, ensuring the doors fitted securely, installing a metal barrier (approximately 30 cm high) immediately in front or behind both doors that would prevent anything crawling in under the door, and establishing strict protocols for people entering the room to ensure the doors were always closed. There are a number of species that would need to be excluded. Robber Crabs could be prevented entry by the metal barrier and ensuring doors were always kept firmly closed. Black Rats could be prevented by ensuring all holes or gaps were blocked, and doors kept closed. Wolf Snakes could be excluded by ensuring there were no gaps under the doors. Giant Centipedes could be excluded by ensuring all gaps and vents were blocked. Even if a Wolf Snake or Giant Centipede did gain access to the room they would not be able to access the bats in the tents as these are totally sealed, and these animals do not have the capacity to chew through the mesh of the tent (unlike a rat for example). Ants would also be excluded by the fine mesh of the tents. During our 1994, 1998 and 2005 studies we lived in the bunkhouse room of the Pink House for a total of 12 weeks, and even without the additional predator-proofing work recommended above, we can not remember any occasion that we observed any of these potential predators within the bunkhouse room. We are therefore confident that with these predator-proofing measures, the added protection of the tents themselves to exclude small animals, and the daily vigilance of staff attending the bats, that these measures would ensure there was no risk from predators to the captive bats in these temporary enclosures.

Are there any animal ethics issues or considerations in relation to capturing bats and undertaking a captive breeding program? While consideration needs to be given to the welfare of any animal taken into captivity, if the capture process and the handling was undertaken by experienced bat biologists it is believed that the impact on the animals would be low. Compared to some other groups of animals (such as small birds), bats do not appear to be overly stressed by the capture process. For example, they will typically sit calmly in a hand, without biting and commence feeding if offered a mealworm immediately after capture, something that a highly stressed animal would not be expected to do. Our previous studies on the pipistrelle have all be subjected to, and approved, by an animal ethics committee. The committee was satisfied that welfare issues associated with the capture process were acceptable. The committee also approved the taking of individuals into captivity for a short period of time during January 2009 as a trial of the captive program. If the bats are maintained in captivity by trained keeping staff or wildlife carers that are experienced in the handling of microbats, and in suitable enclosures, it is believed that they can be maintained in captivity without significant welfare issues.

A greater welfare issue would result from leaving the bats in the wild, which based on the current trends in decline, will inevitably lead to the death of the remaining individuals. In captivity, at least, they have the chance to survive – in the wild they are doomed.

What will be the effect of capturing individuals for the captive breeding program on the ability of the remaining wild population to breed? As the numbers of pipistrelles is getting to such a critically low level and are rapidly declining in the wild, the intention of the captive breeding program is to trap as many of the remaining

animals as possible. If this program had commenced earlier while there were larger numbers present, the intention would have been to take only a subset of the wild animals for the captive program and to actively manage and monitor the remaining animals with the hope that the wild population remained viable. However, it is believed that the population in the wild is currently not viable and hence the highest priority is to take all the animals into captivity.

What is the likelihood that the threats in the wild will be identified, and that they will be able to be mitigated? Despite various projects over a number of years attempting to determine the cause/s of the decline, these have not been able to confirm any of the possible theories. Further targeted studies will be required. If most, or all of the remaining animals are taken into captivity, then it will be more difficult to identify cause/s than if there were bats remaining in the wild. If the decline is due to some form of disease, then there is a good chance that this could be determined from the captive animals. If it is some broad environmental factor (for example an airborne toxicant) then this could be determined without any bats left in the wild. If it is a factor specific to the bats in the wild, e.g. predation, it will be more difficult to determine. It is not possible at this stage to comment on the likelihood that the cause/s of the decline can be mitigated, as each potential cause would have a different chance of mitigation success.

How likely is it that bats raised in captivity will adapt to the natural environment when released? It will be necessary to maintain bats in captivity for many years to be able to build population numbers sufficiently for a release program. Therefore, most if not all of the individuals to be released to the wild will have been captive born. We believe there are several keys factors in the success of this program. Firstly, the rescue has to be commenced immediately so that the founder population is as large as possible. Secondly, while the temporary enclosures are suitable to house the bats initially so that the rescue can commence immediately, large flight enclosures will be required to house the bats for the long-term. Large enclosures will facilitate the necessary social groupings (e.g. enabling a group of females to form a maternity colony), will enable individuals to learn to catch prey in flight, and give sustained flight practice necessary for survival in the wild. A study on the post-release survival of hand-reared pipistrelle bats (Pipistrellus spp.) in the UK, found that extensive pre-release flight conditioning in a large flight enclosure was required for individuals to survive in the wild (Kelly et al. 2008). All individuals that had limited flight training in small spaces failed to survive when released. In contrast, individuals that were flown extensively in large flight enclosures that contained flying insects to provide prey catching opportunities, all survived for at least the 2 week period after release while they could be tracked. During this time individuals were catching insects, and so had learnt this skill without having to be ‘trained’ by other individuals. In the case of the Christmas Island Pipistrelle, the young will all be raised by their mothers, and so can learn from them, further enhancing this ability, rather than being hand-reared in the absence of adults as in the Kelly et al. (2008) study. Therefore, it is believed that if the bats are given a large space, and provided with flying insects, that they will have the sufficient flight and prey- catching skills to survive in the wild.

The other key resource the pipistrelles will require when being released to the wild is suitable roosting sites. To facilitate this, individuals will be acclimatised to nest boxes in captivity, and these boxes will be installed in the wild when the bats are released so that there are familiar roosts to use. This will have the added benefit of being able to predator proof these roosts by applying Tanglefoot to the pole and ropes.

Some marsupial reintroduction programs have failed due to the captive-bred animals not recognising the risk of potential predators (Serena 1995). For the Christmas Island Pipistrelle, the risk of predation is low while in flight due to the lack of nocturnal predators, therefore predation from the roosts will be the main concern. If they can be trained to use bat boxes this

will protect them from climbing, potential predators, however, consideration may also need to be given to training them to recognise the risk of predation before release.

An extensive monitoring program, using an adaptive management approach, will be required when the bats are released to the wild to assess their survival and potential risks. Once there are sufficient numbers in captivity it will be possible to experimentally release individuals, which may help to reveal the cause of the decline.

Would the extinction of the pipistrelle result in further consequences (e.g. does it provide an ecosystem service role that can not be replaced)? When the Christmas Island Pipistrelle was common and widespread, it was likely to have played a critical ecosystem services role in the consummation of large quantities of insects. It is the only vertebrate predator of small, flying, nocturnal insects on the island. The only other nocturnal predator is the Christmas Island Hawk-Owl Ninox natalis, which preys on medium to large insects such as crickets, large beetles and large moths, by snatching the prey off foliage (Hill and Lill 1998). As the pipistrelles take small insects (i.e. < 1 cm), and catch their prey in flight, there is probably little overlap in the prey taken by the two species. Bats typically consume over half their body weight in insects in a night. If we assume that prior to the decline there were in the order of 5000 pipistrelles on the island (it will never be possible to know exactly how many individuals once occurred, and this could be an underestimate), each taking half their body weight in insects in a night, these bats would have consumed over 3,500 kg of insects a year. The loss of these bats would have to have had, and continue to have, an impact on the insect balance on the island, contributing to ecosystem imbalance.

In summary, although the situation is dire and urgent action is required, it is believed that if undertaken immediately the emergency rescue phase could be successful. As this is relatively inexpensive compared to the long-term phase, it is recommended that funding is provided immediately for this component. This will then give much greater clarity on whether the whole program is likely to be a success, before extensive funding is committed.

5 Review of existing monitoring and management programs

This section reviews the Christmas Island Pipistrelle monitoring and management programs currently being conducted by CINP. The extensive monitoring work undertaken by CINP has been instrumental in our understanding of the changing status of the species. The Christmas Island Pipistrelle is the most intensively monitored species of bat in Australia. It is essential that this monitoring and management continues to both document the current status of the species and to reduce the risk of its extinction. While continuing to undertake these management actions is a high priority, especially prior to the establishment of the captive breeding program, these actions alone will not prevent the species going extinct. If they were not undertaken, however, the risk of the species going extinct would increase.

The following actions are prioritised so that the available resources can be directed to those that are likely to have the most impact given the current status of the species. All actions that have been undertaken in the past have been valuable and have provided important information, even if they are now not considered a high priority. The actions listed below are the ones considered most important and feasible with the current level of resources. If more resources were available, implementing targeted predator control programs for the most likely potential predators, such as the Giant Centipede and Black Rat, specifically within the known roosting area would be highly beneficial.

5.1 Stationary detector monitoring Objective: To assess changes in Christmas Island Pipistrelle relative abundance (as reflected in the number of bat passes recorded) at prime foraging areas established as monitoring sites. The regular sampling undertaken by CINP staff since 2004 has been critically important in our understanding of the continuing decline in the species. Current Programs: Since 2006, the stationary detector sampling has been concentrated in the west of the island where the species was still known to occur in 2006. Twelve sites that were established as part of the extensive Christmas Island Biodiversity Monitoring Programme (DNP 2008), have been regularly monitored. Anabat detectors were set for 7 nights at each site, typically each month (except during the wet season November 2006 to February 2007, and March to June 2008 due to equipment problems). The data from the standard 12 sites is summarised in Fig. 2. In 2008, 23 additional sites were sampled along the Winifred Beach Track and Dales/Martin Point Track, plus other sites around the lights of the Immigration Detention Centre, in an attempt to locate other foraging areas. Issues: • At times, regular sampling has been affected by the malfunctioning of the CINP-owned Anabat detectors and CFZcaims (Titley Electronics, Ballina). This was rectified by the purchase of three new Anabat SD1 bat detectors, and these have been working successfully since June 2008. • The disused tracks at some of the monitoring points have become very overgrown, especially in the secondary regrowth west of the start of the Winifred Beach Track (sites A03 and A04 in particular), reducing suitability for foraging. Recommendations: • Continue monitoring at the nine standard sites in the western section of the island for 7 nights, every month, using three new SD1 Anabat detectors provided by LL. As part of the regular monitoring, continue sampling sites A03, A04 and D03 after these tracks are reopened (see below) to determine if the pipistrelles return to this area. Since the three S sites in the

centre of the island have not recorded pipistrelles since mid-2007, discontinue sampling these sites. Priority: High. Risk Assessment: • If this regular monitoring was not undertaken, it would not be possible to document the current status and continuing decline of the species. This information is also critical in identifying the areas of activity to direct where trapping within foraging areas should be undertaken during the capture phase of the captive program.

5.2 Driving detector monitoring Objective: To document changes in the distribution of the Christmas Island Pipistrelle. Since the stationary detector monitoring is focused in the west of the island at known sites, this program aims to periodically re-sample other parts of the island, in case the bats have recolonised new areas. Current Programs: Driving detector monitoring has been undertaken sporadically in recent years. Monitoring has followed the technique outlined in the Christmas Island Pipistrelle Recovery Plan (Schulz and Lumsden 2004). Sampling has been undertaken either along the predetermined route outlined in Schulz and Lumsden (2004), or along all drivable roads and tracks across the island. The latter approach was taken by CINP staff in March and May 2008 and again in June 2008 (Tiernan 2008a). The data collected during the March and May surveys were inconclusive due to uncertainty in correctly distinguishing pipistrelle calls audibly from those of crickets/katydids. This was resolved in June 2008 during the visit by MS. During the June 2008 sampling a total of 66 person hours were spent over three nights surveying for bats across the whole island, driving every accessible track at ≤20 kph. No pipistrelles were recorded during this survey (CINP unpublished data). In July 2008, targeted driving and walking surveys were undertaken over three nights in the west of the island to focus on areas where the species has been recorded in recent years. A total of 52 person hours were expended and 91 km were covered by either foot or car (CINP unpublished data). Only two pipistrelle calls were recorded – one pass at the Sydney Dale carpark and one within 50 m of the Winifred Beach Track gate. This rate of two passes in 91 km is extremely alarming. In addition, CI Phosphate environmental staff conducted an island wide survey in September 2008 sampling the 84 stationary sites established by Lumsden et al. (1999). No pipistrelles were recorded (G. Richards, consultant, pers. comm.). Issues: • Low densities of foraging bats may be missed by this technique due the small sampling time spent at any one point. • Detectability and distribution may be influenced by the time of the year. Recommendations: • The results from the island wide sampling over recent years, including the 2008 sampling, all indicate that the species now only occurs in the far west of the island. Therefore undertaking further island wide surveys, is currently considered a low priority. Priority: Low. Risk Assessment: • The risk in not undertaking surveys throughout the island is that if there are small pockets of bats remaining elsewhere on the island they may be overlooked.

5.3 Monitoring occupancy and numbers of bats in remaining roosts Objective: To document the usage of known roosts by the Christmas Island Pipistrelle, by regularly monitoring how frequently the roost is occupied using a bat detector set below the

roost, and periodically undertaking a roost watch to determine how many individuals are present. This will assist in estimating overall numbers of individuals remaining in the population. Current Programs: Only one of nine communal roosts located in 2005 and 2006 (Hoye 2006; Lumsden et al. 2007) is currently occupied (Roost 565). The others have either collapsed (six of the nine have collapsed), or, despite still appearing to have suitable roosting sites under bark, detector recordings reveal they are no longer being used (two roosts). The remaining roost has been monitored regularly since November 2006, using a bat detector, with 288 nights of sampling undertaken. By examining the detector results and watching the bats at the roost in January 2009, it was apparent that on leaving the roost the bats exit quickly and move away from the area, however, throughout the night they return to the vicinity of the roost and circle in front of it numerous times resulting in a greater number of calls recorded than during the dusk period. Due to this extensive amount of activity it is quite clear when the roost is occupied from the detector data. It cannot however, give an accurate count of individuals using the roost, and to determine numbers, an exit watch is required. In January 2009, bat detectors were set at the two apparently abandoned roosts (Roost 14 and 686) to determine if they were being re-used. No calls were recorded. Issues: • Dusk emergence counts require the use of night vision equipment and staff to be trained in order for the counts to be consistent. Recommendations: • Continue the regular detector monitoring of Roost 565, keeping a detector at this site continually until the bark falls off and the site is deserted, or all animals are caught as part of the captive breeding program. Priority: High. • Undertake regular dusk emergence counts at Roost 565 to determine the number of individuals using the roost. This will be critically important during January – March 2009 to monitor if the roost continues to be used by just four individuals, and to assess if and when the four young of these females become independent, and if they survive. Weekly counts would be optimal. The observer is to be in place well before dark and remain until at least 15 minutes after the last bat has emerged, or the time when the last bat would typically emerge – e.g. in January be in place by 1815 hrs and not leave until 1915-1930 hrs. The observer needs to be quiet and still during this time with their eyes continually focused on the roost. To obtain accurate counts, night vision scopes or goggles are required. DNP (2008) used a night vision monocular, and if this worked satisfactorily it should continue to be used. However, binocular goggles are easier to use for the extended length of time that constant observation of the roost is required in low light levels (30-45 minutes). Priority: High. • Establish a catching tray underneath the roost to collect faecal pellets that drop out of the roost. These could then be used for further dietary analysis. Priority: Medium. Risk Assessment: • It is critically important to monitor the last know roost site to determine when the bats are using this roost and how many individuals are present. This information is essential to help determine the remaining population size and to determine where individuals are roosting when the catching phase of the captive program commences.

5.4 Predator-proof known roosts Objective: To continue the protection all known roosts of the Christmas Island Pipistrelle. Current Programs: In 2007, protective sleeves were fitted around the remaining roost trees and their adjoining trees and saplings. These sleeves comprise the blue tarp material used for crab fences, nailed to the tree with the top and bottom sealed with silicone (Fig. 12). Infra- red cameras were stationed at some of these trees for extended periods, to assess the

effectiveness of these barriers. This indicated that the sleeves significantly reduced the number of potential predators, such as Asian Wolf Snake, Black Rat and Giant Centipede, from accessing roost trees, and hence are an effective method to protect the roost. Issues: • These sleeves are only currently deployed at the known occupied communal roost tree (Roost 565), and the two standing previously used roost trees (Roosts 14 and 686). • The sleeves require checking and maintenance as required. Recommendations: • Regularly maintain the fitted sleeves on Roost 565 and the surrounding interconnecting trees to maximise predator deterrence. Every 6 months check and clean the sleeves, and repair and re-apply silicone as required. They are due for maintenance now. In mid-January a termite trail formed over the barrier on Roost 565, which would give traction to potential predators to climb the tree (Fig. 13). In addition these barriers no longer appear as smooth and shiny as they once were and hence may not provide as great a barrier as when first installed. A wolf spider climbed over the surface easily and hence it is possible that centipedes could do likewise. As Yellow Crazy Ants are in the area and supercolonies are not far away, consider applying Tanglefoot to the sleeve. As ants appear to be the only species able to cross these barriers when they are new, apply the Tanglefoot in the centre so that non- target species approaching the barrier from either below or above are not impacted. Although Roosts 14 and 686 are not currently used, it would be prudent to continue the maintenance of the protective sleeves on these trees as well in case individuals return to these roosts. Priority: High. Risk Assessment: • It is critical to protect the remaining known roost tree from potential predators.

Fig. 12. The protective sleeves that have been installed on roost trees and interconnecting trees.

Fig. 13. A termite trail that formed on the protective sleeve on Roost 565 between 10 and 15 January 2009.

5.5 Location of potential roost sites Objective: To locate potential roost trees within the Winifred Beach Track/Sydney Dales/Dales/Martin Point Track area to document the availability of suitable roost trees, determine locations at which to set bat detectors in an attempt to locate used roosts, and identify potential trees on which to install protective barriers. Current Programs: CINP has recently commenced systematic surveys across this area to identify potential roost trees. To date (January 2009), 150 ha have been searched and 130 potential roost trees (i.e. dead Tristiropsis acutangula trees with sheets of exfoliating bark) have been located. Detectors have been set at a number of potential roost trees to determine if they are being used as roosts. • Systematically search for similar dead roost trees with exfoliating sheets of bark in the Sydney Dale area. Using dusk watches determine whether these trees are used as roosts by the pipistrelle. If roosts are found, fit protective sleeves around the trunk of the roost tree and any trees or saplings touching the roost tree. Issues: • The lifespan of the only currently known roost tree is limited since the tree is dead and the roost is located under lifting bark that is peeling off and likely to fall in the near future (Fig. 5). This tree has already lost extensive amounts of bark since it was first located in 2006. When this roost is no longer available, the bats will need to shift to a new roost. All roosts are likely to be within 100-200 m of the existing roost. As it can not be predicted which roost they are likely to move to, all nearby potential roost trees should be protected. • Walking detectors into remote areas will necessitate the design and construction of more portable stands on which to set the detectors. Recommendations: • Complete the program of searching for potential roosts in the area outlined in discussions in January 2009. Priority: High.

• Set detectors under potential roost trees across this area, for a week at each site, commencing near to Roost 565 and radiating out. Four detectors are to be used for this work. As it is known that there is an alternative roost currently being used that is close to Roost 565, it is critically important that this is located a soon as possible. Priority: High. • If a roost tree is located, conduct a roost exit watch to determine the number of individuals using the roost, and fit a protective sleeve to the roost tree and interconnecting trees. Priority: High. • Facilitate and supervise the work to be undertaken by Christmas Island Phosphates who will attach protective sleeves to all potential roost trees within 200 m of Roost 565. Priority: High. Risk Assessment: • The risk in not locating and protecting alternate roost trees is that the individuals that roost in Roost 565 plus at least one other roost, are not protected from predators when they are not within their main roost tree and could be being killed when using the alternate (non- protected) roosts.

5.6 Monitoring potential predators using infra-red cameras Objective: To document potential predators accessing known or potential roost trees. Current Programs: Since 2006 up to four infra-red cameras (Fauna Focus FF120 camera setup, FaunaTech, Bairnsdale, Victoria) have been deployed by CINP staff on roost trees in the Sydney Dale area. These were typically left in place for a number of weeks or months at a time with the photographs downloaded on a weekly basis. DNP (2008) document the results from these cameras up until the end of 2006. Cameras were initially set on roost trees on a rotational basis. Early in 2008 it was suggested that these cameras be shifted to trees that appeared similar to typical roost type trees both within and outside the known roosting area in Sydney Dale. The aim of this was to investigate if, in the areas where the pipistrelle has disappeared, there are more potential predators climbing roost-type trees. In addition it will reveal if unprotected potential roost trees close to the main roost tree had more potential predators. Many potential predators have been observed on these trees (centipedes, rats, wolf snake) as well as crabs. Recommendations: • Continue to set one camera at a potential roost tree, close to the road to facilitate easy downloading, in the Sydney Dales area to monitor levels of potential predators. Priority: Medium. Risk Assessment: • This work is providing an on-going monitoring of potential predators at roost type trees, and may provide insights into changes in numbers of potential predators in the area. If it was discontinued it may reduce the amount of warning of changes in the area.

5.7 Artificial roosts Objective: Due to the concern over the rapidly collapsing roost trees, artificial roosts were established to provide additional roosting habitat, which could be kept predator-free. Current Programs: Seven artificial roosts were established in May 2006 near to known roosts in the Sydney Dale area, with two close to the currently used communal roost (Fig. 14). These artificial roosts are set on 6 m high steel poles that are guyed for stability. Each roost contains two boxes and four separate cavities (DNP 2008) designed to simulate the cavity space of the most commonly used roost type, i.e. loose bark on dead trees. The base of the roost poles and the guy ropes are coated with Tanglefoot to prevent potential predators climbing the poles (Fig. 14).

Issues: • It is not known if the current design of artificial bat roost is suitable for this species or if alternate designs would be more readily accepted. • When the boxes were built a platform was installed at the base of the landing pad below the entrance. Unfortunately this may have deterred bats from using the boxes as it would have restricted the area available for the bats to land and then climb into the box. In addition, it precluded checking if the boxes were being used by shining a spotlight into the chambers. As a result the boxes were not checked for the two years after they were installed. This problem was rectified in 2008 and the boxes can now be monitored by shining a strong light directly into the interior space of the box and checking with binoculars. Five of the remaining six boxes were checked in January 2009 and found to be empty, with no evidence of them being used. One box was destroyed in a tree fall. • The recent surveys of the Dales area for potential roost sites have revealed greater numbers of potential roosts than anticipated. Given the very low population numbers now present it is likely that there is an abundance of potential roosts available for use by the bats. It is now considered that it is a higher priority to protect potential roost trees with the protective sleeves, rather than erecting artificial roosts. If additional boxes were considered alternative designs could be trialled from either lifting bark or a material resembling lifting bark and fitted to standing dead trees to simulate natural roosts. Recommendations: • The Tanglefoot is replaced every 6 months on the remaining box poles, with the internal cavities checked with a spotlight and binoculars. Priority: Medium. Risk Assessment: • The risk of not maintaining and checking the boxes if that if the bats start using them they will go undetected. This is not considered a high risk however.

Fig. 14. Artificial roost box set on a 6 m pole, that is coated with Tanglefoot.

5.8 Protect and enhance foraging habitat Objective: To protect and enhance key foraging habitat of the Christmas Island Pipistrelle outside the Christmas Island National Park. Current Programs: In recent years there have been proposals to mine phosphate stockpiles in secondary regrowth adjacent to the Winifred Beach Track (Field 26). Some of these areas have been used as foraging areas by the pipistrelle. Submissions by CINP have resulted in these areas being left unmined. Issues: • If extensive secondary regrowth areas were cleared as part of a mining operation this would result in the loss of suitable foraging habitat. • Small clearings in the form of tracks through secondary regrowth, however, do provide suitable foraging habitat. In 2005, tracks were constructed through the secondary regrowth in Field 26j and these provided prime foraging areas that were used extensively by a large number of pipistrelles (for example, in December 2005, 43 individuals were trapped in this area; Fig. 15). These tracks have regrown such that they now no longer provide suitable forging habitat (Fig. 16). • Extensive and intensive detector surveys need to be undertaken as part of any mining impact assessment. Limited sampling of a site in one period may not truly reflect how important some secondary regrowth areas may be in terms of foraging habitat. This is because of the variation of foraging habitat usage by the bats between nights. Now that the number of individuals is very low, sampling over extended periods is required. • The impact of mining in areas adjacent to foraging locations needs to be considered, even if no bats were recorded exactly in the area to be mined. Recommendations: • Support the proposal that Greg Richards will recommend to Christmas Island Phosphates to re-open the tracks in the secondary regrowth to the west of Winifred Beach Track in an attempt to recreate prime foraging habitat as occurred in 2005 (Field 26j managed by Christmas Island Phosphates). The tracks that were established in 2005 should be re- cleared using a small machine, such that these tracks are only 2 m wide. In the process of promoting new foraging habitat, this will also provide additional sites that can be used to trap bats for the captive program. Priority: High. • Ensure data and information sharing with CI Phosphates, which can be facilitated by Greg Richards, the CI Phosphates consultant. Priority: High. Risk Assessment: • There is a high risk that if the area of secondary regrowth adjacent to the start of the Winifred Beach Track was completely cleared for mining that this would have a detrimental impact on the foraging habitat of the remaining individuals. Although the bats do not currently use this area extensively for foraging at present, if the area was extensively cleared it could impact not just the foraging activity in this area but also in the main foraging area along the adjacent Winifred Beach Track. • While total clearing of this area is considered highly detrimental, the clearing of small tracks within this secondary regrowth, specifically for the purpose of providing additional foraging habitat, is considered a highly beneficial and the advantages of undertaking this action significantly outweigh any potential risks. The original small tracks were installed in 2005 and within months the bats had taken advantage of these small clearings and were using them extensively as foraging areas. It is expected that if the tracks were re-cleared in 2009 the bats would again quickly commence using them (although due to the much reduced overall numbers of bats remaining now compared to in 2005, the number of individuals using it would also be reduced). In addition to providing additional foraging areas for the bats it would provide additional locations to potentially trap them for the captive program. These tracks were highly successful locations for trapping bats in 2005 (Lumsden et al. 2007) and would increase the chances of successful trapping in foraging areas. The sole intention of clearing

these tracks would be to provide foraging habitat for the pipistrelles. Once the young regrowth along these narrow tracks had been cleared (which could be done in several hours during the day while the bats were not in the area) there would be no subsequent vehicle or machinery traffic along these tracks. If there was any concern that there could be increased vehicular traffic into the area (which is highly unlikely), a barrier could be placed across the start of the track network to prevent all access. We believe there is no risk that the clearing of these tracks would be detrimental to the pipistrelles and it has the potential to be highly beneficial. The only risk is that nothing happens – i.e. that the work is undertaken and the bats do not use it in which case the situation is the status quo, with the only consequence that a small amount of 3-year old regrowth has been cleared.

Fig. 15. One of the cleared tracks in Field 26j that provided optimal foraging habitat and areas at which to trap bats in 2005.

Fig. 16. Overgrown tracks in Field 26j in 2009, which are no longer used for foraging.

5.9 Yellow Crazy Ant control Objective: To reduce potential impacts on the Christmas Island Pipistrelle by the Yellow Crazy Ant. Current Programs: Some of the key roost and foraging areas of the Christmas Island Pipistrelle occur within areas that have been, or are currently being, treated to reduce the high numbers of Yellow Crazy Ants that are present. This treatment involves hand-baiting areas with the insecticide Fipronil. Issues: • If supercolonies formed in the vicinity of roost sites it is highly likely that these roosts would be deserted. • In areas where supercolonies have formed and have subsequently been baited with Fipronil, insect prey is likely to be reduced, particularly any insects that have part of their life stage on the forest floor. Recommendations: • Regularly monitor the key roosting area in Sydney Dale to ensure it remains free of ant supercolonies. If any supercolonies start to form ensure that they are controlled immediately. Priority: High. • Seek advice from the Crazy Ant Scientific Advisory Panel (CASAP) in relation to impacts of Fipronil on non-target invertebrates. Subject to CASAP advice investigate lower impact baiting methods in and around key foraging areas where Yellow Crazy Ant numbers are high. Investigate methods that can be used to prevent supercolony formation. Priority: Medium. Risk Assessment: • If Yellow Crazy Ants were not controlled in the main roost area they would have the potential to cause the final demise of the population.

5.10 Feral Cat control trial Objective: To reduce potentially adverse impacts of key Feral Cat prey species that are likely to increase in numbers following the Feral Cat control trial. Current Programs: A Feral Cat control trial using toxic baits is being undertaken across the island by the WA Department of Environment and Conservation and the Victorian Department of Sustainability and Environment. Issues: While Feral Cats have been identified as a potential predator of the Christmas Island Pipistrelle, it is likely that other species such as the Black Rat are a much greater threat to this species. Consequently, unless there is a simultaneous rat control program, the removal of cats has the potential to lead to an increase in the rat population. Little is known about the predator-prey relationship between cats and rats on the island, and if cat numbers limit rat numbers. However, since 45% of the cats diet on Christmas Island is rats (Corbett et al. 2003), it would be wise to take the precautionary approach and assume they are having an impact on rat numbers. In addition, cats may be preying on other potential predators of the pipistrelle such as the Asian Wolf Snake.

Discussions were held with CINP and Michael Johnston (DSE, one of the researchers on the cat project) prior to the commencement of the trial in 2008 to investigate ways to mitigate any potential impacts on the pipistrelle. The west of the island (Fig. 17) was designated as an exclusion zone for the use of toxic baits. The first phase of this project has been completed and bait stations were established across the island. Bait stations were established within this exclusion zone, however toxic baits will not be used at these sites and they will be used as control and monitoring sites.

Recommendations: • Continue to follow the guidelines established whereby no toxic baits are employed within the exclusion zone in the west of the island. Priority: High. • Monitor the relative abundance of rats, by assessing footprints in the sand pads at both toxic and non-toxic bait stations. This will enable an investigation of the impact of the reduction in cat numbers on rat abundance. It will be essential to monitor these over a longer time period than the cats will be monitored as the rats will increase over a number of generations. Priority: Medium. Risk Assessment: • If reducing the abundance of cats on the island led to an increase in the abundance of rats this could add further pressure to the remaining population of bats and increase the risk of extinction.

Fig. 17. Exclusion zone for using toxic baits as part of the Feral Cat eradication trial. The circles roughly indicate the main roosting and foraging areas used in the last five years, and the dashed lined the extent of the agreed exclusion zone.

References Adams, L., Wyeth, J., Brady, P., Lloyd, B., Whiteford, T., Pickard, J., Hancock, H. and Ruffell, J. (2005). The capture and translocation of Southern Short-tailed Bats from Tararua Forest Park to the National Wildlife Centre, Mount Bruce and transfer of juveniles to Kapiti Island. Unpublished report. Altringham, J.D. (1996). Bats Biology and Behaviour. Oxford University Press, Oxford. Arlettaz, R., Christe, P. and Desfayes, M. (2002). 33 years, a new longevity record for a European bat. Mammalia 66, 441-442. Baer, G.M., and Holguin, G.M. (1971). Breeding Mexican Freetail Bats in captivity. The American Midland Naturalist 85, 515-517. Baker, G.B., Lumsden, L.F., Dettmann, E.B., Schedvin, N.K., Schulz, M., Watkins, D. and Jansen, L. (2001). The effect of forearm bands on insectivorous bats (Microchiroptera) in Australia. Wildlife Research 28, 229-237. Balmford, A., Mace, G.M. and Leader-Williams, N. (1996). Designing the Ark: setting priorities for captive breeding. Conservation Biology 1996, 719-27. Barclay, R.M.R. (1995). Does energy or calcium availability constrain reproduction by bats? Symposia of the Zoological Society London 67, 245-58. Barclay, R.M.R. and Harder, L.D. (2003). Life histories of bats: life in the slow lane. Pp. 209- 256. In: Kunz, T. H. and Fenton, M. B. (Eds.). Bat Ecology. The University of Chicago Press, Chicago. Barclay, R.M.R., Ulmer, J., MacKenzie, C.J.A., Thompson, M.S., Olson, L., McCool, J., Elvie Cropley, E. and Poll, G. (2004). Variation in the reproductive rate of bats. Canadian Journal of Zoology 82, 688–693. Barnard, S.M. (1995). Bats in Captivity. Wild Ones Animal Books, Springville, California. Blanchard, B. (1992). The New Zealand lesser short-tailed bat Mystacina tuberculata at Wellington Zoological Gardens. International Zoo Yearbook 31, 92-96. Blanchard, B. (1998). New Zealand Lesser Short-tailed Bat (Mystacina tuberculata). Advocacy and Captive Management Plan. Wellington Zoo. Unpublished report. Carroll, J.B., and Mace, G.M. (1988). Population management of the Rodrigues fruit bat Pteropus rodricensis in captivity. International Zoo Yearbook 27, 70-78. Chua, K.B., Bellini, W.J., Rota, P.A., Harcourt, B.H., Tamin, A., Lam, S.K., Kslazek, T.G., Rollin, P.E., Zaki, S.R., Shieh, W.-J., Goldsmith, C.S., Gubler, D.J., Roehrig, J.T., Eaton, B., Gould, A.R., Olson, J., Field, H., Daniels, P., Ling, A.E., Peters, C.J., Anderson, L.J. and Mahy, B.W.J. (2000). Nipah Virus: A recently emergent deadly paramyxovirus. Science 288, 1432-1435. Clark, P. (2004). Haematology of Australian Mammals. CSIRO Publishing, Melbourne. Corbett, L., Crome, F. and Richards, G. (2003). Fauna survey of mine lease applications and national park reference areas, Christmas Island, August 2002. Appendix G. In: CIP (ed.). Draft Environmental Impact Statement for the Proposed Christmas Island Phosphate Mines (9 Sites). Christmas Island Phosphates, Perth. Director of National Parks (DNP) (2008). Christmas Island Biodiversity Monitoring Program: December 2003 to April 2007. Report to Department of Finance and Deregulation, Canberra. Dobson K. (1999). Behavioural, flight and echolocation mechanisms of south-eastern bats (Microchiroptera) that enable avoidance of capture in harp traps. Honours thesis. Department of Zoology, Monash University, Clayton, Victoria. Dondini, G., and Vergari, S. (1995). Rearing and first reproduction of the Savi's pipistrelle Pipistrellus savii at Group of Study and Conservation Chiroptera, Florence. International Zoo Yearbook 34, 143-146.

Eales, L.A., Bullock, D.J. and Slater, P.J.B. (1988). Shared nursing in captive pipistrelles (Pipistrellus pipistrellus)? Journal of Zoology, London 216, 584-587. Ellis, W.A.H., Marples, T.G. and Phillips, W.R. (1991). The effects of a temperature- determined food supply on the annual activity cycle of the Lesser Long-eared Bat, Nyctophilus geoffroyi Leach, 1821 (Microchiroptera: Vespertilionidae). Australian Journal of Zoology 39, 263-271. Fenton, M.B., Gaudet, C.L. and Leonard, M.L. (1983). Feeding behaviour of the bats Nycteris grandis and Nycteris thebaica (Nycteridae) in captivity Journal of Zoology, London 200, 347-354. Gartrell, B.D. (2007). Dermatitis of the pinnae in Lesser Short-tailed bats, Mystacina tuberculata, translocated to Kapiti Island. Kokako 14(2), 25-31. Gaudet, C.L. and Fenton, M.B. (1984). Observational learning in three species of insectivorous bats (Chiroptera). Animal Behaviour 32, 385-388. Heath, A.C.G., Julian, A.F., Daniel, M.J. and Bishop, D.M. (1987). Mite infestation (Acari: Laelapidae) of New Zealand short-tailed bats, Mystacina tuberculata, in captivity. Journal of the Royal Society of New Zealand 17, 41-47. Hall, L.S. (1982). Management of Microchiroptera in captivity. Pp. 157-160. In: The Management of Australian Mammals in Captivity. Ed. D.D. Evans. Zoological Board of Victoria, Melbourne. Hill, F.A.R. and Lill, A. (1998). Diet and roost site characteristics of the Christmas Island Hawk-Owl Ninox natalis. Emu 98, 227-33. Hopkins, C.S. (1990). Carnivorous and insectivorous bats. Pp. 171-184. In: Care and Handling of Australian Native Animals: Emergency Care and Captive Management: Hand, S.J. (Ed.) Surrey Beatty & Sons P/L, Chipping Norton, NSW. Hosken, D.J. (1998). Sperm fertility and skewed paternity during sperm competition in the Australian long-eared bat Nyctophilus geoffroyi (Chiroptera: Vespertilionidae). Journal of Zoology, London 245, 93-100. Hoye, G. (2006). Radiotelemetry of Christmas Island Pipistrelle and subsequent recommendations for conservation. Fly By Night Bat Surveys, Sydney. Unpublished report to Parks Australia North, Christmas Island. Hughes, P.M., Rayner, J.M.V. and Jones, G. (1995). Ontogeny of 'true' flight and other aspects of growth in the bat Pipistrellus pipistrellus. Journal of Zoology, London 235, 291-318. Jackson, S. (2003). Australian Mammals: Biology and Captive Management. CSIRO Publishing, Melbourne. Johara, M., Field, H., Rashdi, A., Morrissy, C., van der Heide, B., Rota, P., Adzhar, A., White, J., Daniels, P., Jamaluddin, A. and Ksiazek, T. (2001). Nipah virus infection in bats (Order Chiroptera) in Peninsular Malaysia. Emerging Infectious Diseases 7, 439-41. Kelly, A., Goodwin, S., Grogan, A. and Mathews, F. (2008). Post-release survival of hand- reared pipistrelle bats (Pipistrellus spp). Animal Welfare 17, 375-382. Kleiman, D.G. (1969). Maternal care, growth rate, and development in the noctule (Nyctalus noctula), pipistrelle (Pipistrellus pipistrellus), and serotine (Eptesicus serotinus) bats. Journal of Zoology, London 157,187-211. Kunz, T.H. and Kurta, A. (1988). Capture methods and holding devices. In Ecological and Behavioral Methods for the Study of Bats (ed. T.H. Kunz), pp. 1-29. Washington, D.C.: Smithsonian Institution Press. Kunz, T.H. and Lumsden, L.F. (2003). Ecology of cavity and foliage roosting bats. Pp. 3-89. In: Kunz, T.H. and Fenton, M.B. (eds.). Bat Ecology. The University of Chicago Press, Chicago. Lewis, S.E. (1995). Roost fidelity of bats: a review. Journal of Mammalogy 76, 481-96. Lollar, A. and Schmidt-French, B. (1998). Captive Care and Medical Reference for the Rehabilitation of Insectivorous Bats. Bat World, Mineral Wells, Texas.

Lumsden L. F. (2004) The ecology and conservation of insectivorous bats in rural landscapes. PhD thesis. p. 285. Deakin University. Lumsden, L. and Cherry, K. (1997). Report on a preliminary investigation of the Christmas Island Pipistrelle Pipistrellus murrayi, in June – July 1994. Arthur Rylah Institute for Environmental Research, Heidelberg, Victoria. Lumsden, L. and Gray, P. (2001). Longevity record for a Southern Bent-wing Bat Miniopterus schreibersii bassanii. The Australasian Bat Society Newsletter 16, 43-44. Lumsden, L., Silins, J. and Schulz, M. (1999). Population dynamics and ecology of the Christmas Island Pipistrelle Pipistrellus murrayi on Christmas Island. Report for Parks Australia North – Christmas Island. Arthur Rylah Institute for Environmental Research, Heidelberg, Victoria. Lumsden, L., Schulz, M., Ashton, R. and Middleton, D. (2007). Investigation of threats to the Christmas Island Pipistrelle. A report to the Department of the Environment and Water Resources. Arthur Rylah Institute for Environmental Research, Department of Sustainability and Environment, Heidelberg, Victoria. Mickleburgh, S.P., Hutson, A.M. and Racey, P.A. (1992). Old World Fruit Bats. An Action Plan for their Conservation. International Union for Conservation of Nature and Natural Resources, Gland, Switzerland. Middleton, D.J., Morrissy, C.J., van der Heide, B.M., Russell, G.M., Braun, M.A., Westbury, H.A., Halpin, K. and Daniels, P.W. (2007). Experimental Nipah virus infection in pteropid bats (Pteropus poliocephalus). Journal of Comparative Pathology 136, 266- 272. Mohos, S.C. (1961). Bats as laboratory animals. The Anatomical Record 139, 369-378. Molur, S., Marimuthu, G., Srinivasulu, C., Mistry, S., Hutson, A.M., Bates, P.J.J., Walker, S., Padma Priya, K. and Binu Priya, A.R. editors. (2002). Status of South Asian Chiroptera: Conservation Assessment and Management Plan (C.A.M.P.) Workshop Report, 2002. Zoo Outreach Organisation, CBSG South Asia and WILD, Coimbatore, India, CD-Rom. O'Brien, J., McCracken, G.F., Say, L. and Hayden, T.J. (2007). Rodrigues fruit bats (Pteropus rodricensis, Megachiroptera: Pteropodidae) retain genetic diversity despite population declines and founder events. Conservation Genetics 8, 1073-1082. Orr, R.T. (1954). Natural history of the Pallid Bat, Antrozous pallidus (LeConte). Proceedings of the California Academy of Sciences 28, 165-246. Phillips, W.R. and Inwards, S.J. (1985). The annual activity and breeding cycles of Gould's Long-eared Bat, Nyctophilus gouldi (Microchiroptera: Vespertilionidae). Australian Journal of Zoology 33, 111-126. Pickering, J. and Norris, C.A. (1996). New evidence concerning the extinction of the endemic murid Rattus macleari Thomas 1887, from Christmas Island, Indian Ocean. Australian Mammalogy 19: 19-25. Racey, P.A. (1969). Diagnosis of pregnancy and experimental extension of gestation in the pipistrelle bat, Pipistrellus pipistrellus. Journal of Reproduction and Fertility 19, 465- 474. Racey, P.A. (1982). Ecology of bat reproduction. pp. 57-104. In: Ecology of Bats (ed T. H. Kunz). Plenum Press, New York. Racey, P.A. and Entwistle, A.C. (2003). Conservation ecology of bats. pp. 680-743. In: Bat Ecology (eds. T.H. Kunz and M.B. Fenton). The University of Chicago Press, Chicago. Racey, P.A. and Kleiman, D.G. (1970). Maintenance and breeding in captivity of some vespertilionid bats, with special reference to the noctule. International Zoo Yearbook 10, 65-70. Rasweiler, IV, J.J. (1975). Maintaining and breeding neotropical frugivorous, nectarivorous and pollenivorous bats. International Zoo Yearbook 15, 18-30.

Richards, G.C. (2008). Status of the Christmas Island Pipistrelle on mining leases and environs in the northwest of Christmas Island. Unpublished report for Christmas Island Phosphates P/L. Rolls, E.C. (1969). They All Ran Wild: the story of pests on the land in Australia. Angus and Robertson, Sydney. Schulz, M. and Lumsden, L.F. (2004). National Recovery Plan for the Christmas Island Pipistrelle Pipistrellus murrayi. Commonwealth of Australia, Canberra. Sedgeley, J. (1997). Captive management and reintroduction to the wild of Short-tailed Bats on Codfish Island in Winter 1996. Unpublished report to the Department of Conservation. Sedgeley, J. and Anderson, M. (2000). Capture and captive maintenance of Short-tailed Bats on Whenua Hou and monitoring of wild bats during the Kiore eradication programme winter 1998. Unpublished report to the Department of Conservation, Invercargill. Serena, M. (ed.) (1995). Reintroduction Biology of Australian and New Zealand Fauna. Surrey Beatty & Sons, Chipping Norton. Speakman, J.R. (1995). Chiropteran nocturnality. Symposia of the Zoological Society London 67: 187-201. Tidemann, C. (1985). A study of the status, habitat requirements and management of the two species of bats on Christmas Island (Indian Ocean). Report to Australian National Parks and Wildlife Service, Canberra. Thompson, M.J.A. (1987). Longevity and survival of female pipistrelle bats (Pipistrellus pipistrellus) on the Vale of York, England. Journal of Zoology, London 211, 209-214. Tuttle, M.D. and Stevenson, D. (1982). Growth and survival of bats. pp. 105-50. In: Ecology of Bats (ed T. H. Kunz). Plenum Press, New York. Viggers, K.L., Lindenmayer, D.B. and Spratt, D.M. (1993). The importance of disease in reintroduction programmes. Wildlife Research 20, 687-698. WAZA (World Association of Zoos and Aquariums), 2008, www.waza.org, accessed 2/07/2008. Wilson, D.E. (1988). Maintaining bats for captive studies. pp. 247-64. In: Ecological and Behavioral Methods for the Study of Bats. (ed. T. H. Kunz). Smithsonian Institution Press, Washington, D.C. Wilson, D.E. and Tyson, E.L. (1970). Longevity records for Artibeus jamaicensis and Myotis nigricans. Journal of Mammalogy 51, 203.

Appendix 1. The Agri-food and Veterinary Authority (AVA) of Singapore guidelines that would need to be met for the importation of Christmas Island Pipistrelles to Singapore. Three additional requirements specified by Singapore Zoo are included at the end.

REPUBLIC OF SINGAPORE

THE ANIMALS AND BIRDS ACT (CHAP.7)

VETERINARY CONDITIONS FOR THE IMPORTATION OF ZOOLOGICAL ANIMALS AND BIRDS(2/9) IN THE ORDERS CARNIVORA, CHIROPTERA, DERMOPTERA, EDENTATA, HYRACOIDEA, INSECTIVORA, LARGOMORPHA, MARSUPIALA AND RODENTIA(2/2)

I COUNTRY OF EXPORT Australia, New Zealand, United Kingdom (Great Britain and Northern Ireland) or Ireland. II PURPOSE Exhibition in Zoos. III IMPORT PERMIT Each animal shall be accompanied by a valid import permit issued by the Agri-food and Veterinary Authority (AVA) of Singapore.

IV VETERINARY CERTIFICATION Each consignment of animals shall be accompanied by a veterinary health certificate dated not more than seven (7) days prior to export and signed or endorsed by the competent Veterinary Authority of the country of export describing the age, breed, sex, colour, markings or other points of identification, and the name and address of the premises of origin of each animal, and certifying to the effect that: (i) the country has been free from Rabies for at least 3 years prior to export and no vaccination against the disease is permitted, (ii) the animals have been kept in captivity in the country of export for at least 6 months prior to export or since birth, (iii) the animals came from an area in which infectious and contagious animal diseases are under control, (iv) the animals have not been vaccinated with any vaccine within 30 days of export, (v) the animals have been examined and found to be healthy and free from any clinical signs of infectious or contagious disease at the time of export, (including freedom from infection with Hantaan virus/Korean haemorrhagic fever in the case of animals in the ORDER Rodentia).

V DECLARATION FROM MASTER/CAPTAIN OF SHIP/AIRCRAFT Each consignment of animals shall be accompanied by a signed declaration from the master/captain of the ship/aircraft in which the animals were carried stating that: (i) the animal(s) had been embarked or emplaned in one of the country of export specified above. (ii) no other animal was taken on board the ship/aircraft after it left the country of export.

(iii) the animal(s) had not been landed at any intermediate port outside the country of export. In the case of an aircraft, if the animal was landed, the animal had been conveyed in a crate and remained within the precincts of the airport at which the aircraft landed. (Prior permission from AVA is required should there be a change of aircraft carrying the animal(s)).

VI QUARANTINE Upon arrival, the animals shall be quarantined for a period of not less than 30 days at an approved quarantine area at the Zoo. The Zoo authority or owner shall report any sickness or death of the animal to the Head/Import & Export Division (Head/IED) and such animal may not be disposed off without his permission.

VII APPLICATION FOR IMPORT PERMIT The Zoo authority must apply for the import permit by submitting a written application to the Import & Export Division at least three(3) weeks prior to the date of arrival of the animals giving the following information: scientific and common names of the animal(s), number of heads, country of export, name and address of premises of origin, name and address of exporter/consignor, flight number, date of arrival, and CITES category of the animal(s), and ports of call of the aircraft/ship carrying the animals.

VIII NOTIFICATION OF ARRIVAL The Zoo authority shall contact the Quarantine Office as stated in the Import Permit at least one working day before the arrival of the animals.

IX VETERINARY INSPECTION On arrival at the port of disembarkation in Singapore, the animals and documents (Import Permit, Veterinary Health Certificate and Captain's Declaration) shall be presented to and examined by an authorised Veterinary Officer. If the consignment of animals is found to be healthy and documents are in order, the animals shall be taken straight to the approved quarantine area at the Zoo and shall not be released from quarantine without the permission of the Head/IED.

X PENALTY If any animal is not healthy or if any document is not in order, the animals shall be quarantined for a further 30 days or longer/returned or destroyed at the discretion of the Director-General of Agri-food and Veterinary Services. In addition the importer is subject to prosecution.

XI FEES (i) Import Permit ) (ii) Veterinary inspection and processing of documents ) see Fee Schedule: Attachment-2.

XII OTHER REQUIREMENTS/INFORMATION (i) Under The Endangered Species (Import and Export) Act, animal species listed in the Appendices of CITES (Convention On International Trade in Endangered Species of Fauna and Flora) must also be accompanied by a CITES import permit issued by AVA and an export/re-export certificate from the country of export/origin of the animal. Application for the CITES import permit should be made at least 14 days prior to the date of importation by submitting the completed application form to the Import & Export Division. For CITES Import Permit fee, please see Fee Schedule: Attachment-2.

(ii) A licence issued by AVA must be obtained to use any place for exhibition, sale or export of any animal or bird. (iii) Any expenditure incurred in the process of importation shall be borne by the zoo authority. (iv) Regulations and fees are subject to change without notice notwithstanding the issuance of a licence/permit by AVA (v) Order Carnivora includes: dogs, cats, jackals, foxes, wolves, bears, raccoons, coatis, pandas, otters, weasels, martens, polecats, badgers, skunks, mink, ratel, genets, civets, linsangs, mongooses, hyaenas, ocelots, pumas, cheetahs, lions, tigers, leopards. Order Chiroptera includes: bats, flying foxes. Order Dermoptera includes: flying lemurs. Order Edentata includes: anteaters, sloths, armadillos. Order Hyracoidea includes: hyraxes. Order Insectivora includes: solenodons, tenrecs, otter shrews, golden moles, hedgehogs, elephant shrews, shrews, moles, desmans. Order Largomorpha includes: pikas, rabbits, hares. Order Marsupiala includes: opossums, marsupial mice, dasyures, marsupial moles, marsupial anteaters, bandicoots, rat opossums, cuscuses, phalanges, koala, wombats, wallabies, kangaroos. Order Rodentia includes: gophers, squirrels, chipmunks, marmots, scaly-tailed squirrels, pocket mice, kangaroo rats, beavers, mountain beavers, springhaas, mice, rats, hamsters, lemmings, voles, gerbils, water rats, dormice, jumping mice, jerboas, muskrats, porcupines, cavies, guinea pigs, capybara, chinchillas, spiny rats, gundis, coypus.

ISSUED BY: APPLICATION/ENQUIRIES: HEAD/REGULATORY & HEALTH PLANNING DIVISION HEAD/IMPORT & EXPORT DIVISION AGRI-FOOD AND VETERINARY AUTHORITY OF SINGAPORE 5 MAXWELL ROAD #02-00 5 MAXWELL ROAD #03-00 TOWER BLOCK, MND COMPLEX TOWER BLOCK, MND COMPLEX SINGAPORE 069110 REPUBLIC OF SINGAPORE 069110 TELEPHONE: [65] 62270670 FAX: [65] 62206068 FAX: [65] 62276305 E-mail: [email protected] E-mail: [email protected] [email protected]

Additional conditions required by Singapore Zoo

In addition to the conditions stated in the AVA document, Singapore Zoo also need the bats to be:

• Free from endo- and ectoparasites, and treated for them as well

• Tested for and free of Nipah virus

• Tested for and free of Lyssavirus

Appendix 2. A summary of the advantages and disadvantages of the five possible locations for establishing a captive breeding facility for the Christmas Island Pipistrelle.

Issue Christmas Island Singapore Zoo Territory Wildlife Park Zoos Victoria Taronga Zoo Quarantine Not required Yes No – would need additional Yes Yes Approved Premises funding and time to meet (QAP) requirements Quarantine Not required for the bats There are existing Import Risk Assessment (IRA) Import Risk Assessment Import Risk Assessment documentation however would be guidelines for importing would need to be developed by would need to be would need to be required to bring in bats to Singapore Biosecurity Australia. developed by Biosecurity developed by Biosecurity mealworms QAP documentation would Australia Australia need to be completed Quarantine timelines AQIS on Christmas Island Nil – all documentation IRA could take up to 3 years (2 IRA could take up to 3 IRA could take up to 3 for documentation is currently processing currently in place yr backlog, 1 yr to complete) years (2 yr backlog, 1 yr to years (2 yr backlog, 1 yr to our permit for bringing in however it might be possible to complete) however it complete) however it might mealworms fast track this. might be possible to fast be possible to fast track Time required to complete QAP track this. this. status is unknown – no clear guidelines are available – assessment is on a case by case basis. Ability to meet Currently being Unable to meet The initial advice was that an As for TWP but already As for TWP but already quarantine determined for the quarantine requirement to IRA would require testing for have QAP have QAP requirements mealworms by AQIS on test for lyssavirus and lyssavirus and Nipah virus. Christmas Island Nipah virus due to the However this will not be quantity of blood possible due to quantity of required. blood required. Advice from Biosecurity Aust. is that it may be possible to waive this requirement if high biosecurity measures were in place for transit, housing, handling, etc. Would require obtaining QAP status.

Issue Christmas Island Singapore Zoo Territory Wildlife Park Zoos Victoria Taronga Zoo Impact of quarantine Nil Will be unable to meet Will be unable to meet virus As for TWP As for TWP requirements on virus testing testing requirements. Other bats requirements. Other requirements may include requirements include treating for parasites – the treating for parasites – impacts of which is unknown. the impacts of treating Until an IRA is conducted it is with the required not known if other treatments chemicals is unknown. will be required and what impacts these may have. Impact of Minimal – transporting Flight to Singapore could If using commercial flight the As for TWP As for TWP transportation on from capture point to be undertaken in a bats would have to go through bats facility would take less relatively short period of Perth, requiring a stop over in than an hour time (compared to quarantine to allow bats to feed mainland options) and drink. Bats would need to reducing the impact of be kept in the plane’s cabin in a transportation however container with high humidity suitable transport and temperature ensuring containers would still be sufficient oxygen for the trip. required. This would need to be sealed for quarantine purposes. Unknown how the bats will cope with being transported – there is the potential for fatalities. Removal of bats Potential that if left on the Would be removed from Would be removed from the Would be removed from Would be removed from from environment in island they might still be the existing environment. existing environment. the existing environment. the existing environment. which decline is subjected to whatever is occurring causing the decline Microclimate Optimal Generally similar, Generally similar, although may Would require artificial Would require artificial conditions although may be subtle be subtle differences temperature, humidity and temperature, humidity and differences day length controlled day length controlled environments environments

Issue Christmas Island Singapore Zoo Territory Wildlife Park Zoos Victoria Taronga Zoo Temporary holding Bats could go directly into Bats would need to be As for Singapore Zoo. As for Singapore Zoo. As for Singapore Zoo. facilities and staff the new facility when housed on the island for available, although possibly several months temporary housing would while husbandry is be required while the refined, animals adapt to long-term facility was captivity and all the being built. animals required are caught. Therefore temporary enclosures would need to be established and staff employed. Long-term facilities Would need to be built – May be possible to use Would need to be built – Would require special Would require special would need to obtain existing facilities. expertise at the zoo would facilities to be built or facilities to be built or input from zoo experts on facilitate this. existing ones modified. existing ones modified. design etc. Staff – availability of Specialist staff would Have general husbandry Have general husbandry Have extensive general Have extensive general specialists at long- need to be employed expertise and have expertise and captive breeding husbandry expertise, with husbandry expertise and term facility either directly or through colonies of flying-foxes for other species, but little direct small mammals, including some experience with a collaborating zoo. but no direct experience experience with microbats. husbandry and microbats. Have extensive with microbats. veterinarian experience expertise in captive with microbats. Have breeding programs for extensive expertise in other species and in captive breeding wildlife health. programs for other species and in wildlife health. Transport options Easy Direct flight would be Commercial flights would If a direct flight could be If a direct flight could be preferable. If the flight require a stop over in Perth in organised this would organised this would needed to go through quarantine, as there are no reduce the overall reduce the overall Malaysia, quarantine direct flights to Darwin. A direct travelling time and be less travelling time and be less requirements would need flight would be less of a risk to of a risk to the animals. of a risk to the animals. to be addressed. the animals and should be Multiple flights would be Multiple flights would be investigated. Multiple flights required, thus increasing required, thus increasing would be required. the cost. the cost.

Issue Christmas Island Singapore Zoo Territory Wildlife Park Zoos Victoria Taronga Zoo Stocking of colony Would be able to take May be possible to Due to costs and logistics As for TWP As for TWP animals in as they were undertake a couple of associated with transporting to caught. A few animals flights to bring animals to the mainland, it is likely that this could be collected first to Zoo. would be done as infrequently refine the husbandry as possible, increasing the risk techniques and others if anything went wrong, and brought in when possible reducing the ability to add to the and appropriate. colony where possible and appropriate. CI Community / High As Singapore is in the CIP would be less likely to As for TWP although As for TWP although CI Phosphate P/L general region and CIP assist with funding. CI probably worse probably worse involvement have head quarters there community would have less this option would provide involvement and ownership of greater links than the the project. Australian mainland. Commitment to Singapore Zoo have Still interested in the project but Detailed discussions have While Taronga were project notified us that they will a lot of issues would need to be been undertaken with interested in the project, not be able to provide an worked out. senior staff and they after a number of ex-situ captive breeding recognise that there are discussions they decided facility. many advantages to that they would be unable having the colony on the to take responsibility for island. They are keen to the captive colony but they provide the captive were willing to provide husbandry expertise for a advice and assistance. colony established on Christmas Island.