The Ecology & Management of The Grey-Headed Flying-Fox

The ecology and management of the grey-headed flying-fox Pteropus poliocephalus
Billie Roberts
BSc (Hons) A thesis submitted in fulfilment of the requirements of the degree of Doctor of Philosophy Griffith School of Environment Griffith University September 2012
Photo Vivien Jones
This work is dedicated to the researchers, wildlife carers and community members trying to conserve unpopular threatened species. May you be inspired to learn more about these species and work harder to prevent their demise.
Louise Saunders
Abstract
Effective conservation and management of many highly mobile animal species, including flying-foxes of the genus Pteropus, are constrained by lack of knowledge of their ecology, especially of movement patterns over large spatial scales. This study deals with the conservation ecology of the grey-headed flying-fox (Pteropus poliocephalus), a highly mobile colonial-roosting mammal endemic to eastern Australia. Like many flying-foxes in the Asia-Pacific, this species frequently comes into conflict with humans for raiding commercial fruit crops, establishing roost sites in urban areas, and because it is a vector for some diseases of humans and their commensal animals. For these reasons, P. poliocephalus has been the subject of numerous control attempts, mostly conducted at the local scale (killing of individuals at feeding sites and roosts; attempts to destroy or relocate roosts). These attempts have rarely been successful in reducing conflict between humans and flying-foxes over the long term. Partly as a result of these actions and partly because of a loss of habitat, P. poliocephalus is considered a threatened species in some jurisdictions. The conservation and management of P. poliocephalus is therefore a difficult problem. The broad objectives of this thesis are to provide new information on the ecology of P. poliocephalus, particularly in relation to distribution, movement patterns and roosting behaviour, in the hope that this information will better inform the management and conservation of the species. Historical changes in the distribution of P. poliocephalus and another Australian flying-fox (P. alecto), whose range overlaps with P. poliocephalus, were assessed from locality records (dating from 1843 to 2007) obtained from a wide range of sources, filtered for reliability and spatial accuracy. Analysis of these records did not support previous claims that the distributions of these species had shifted in a manner driven by climate change. First, neither the northern or southern range limits of P. poliocephalus (Mackay, Queensland and Melbourne, Victoria respectively) have changed over the past century. Second, while the southern range limit of P. alecto extended southward by over 1000 km (more than 10 degrees latitude) during the twentieth century (from Rockhampton, Queensland to Sydney, New South Wales), the rate of expansion was much greater than observed changes in climate. The range of P. alecto expanded southward at about 100 km/ decade, compared with the 1026 km/ decade rate of isotherm change. Analyses of historical weather data show that the species consequently moved into colder regions than it had previously occupied.
Abstract
ii To quantify the movement patterns of P. poliocephalus, fourteen adult male P. poliocephalus were satellite tracked for 240 weeks (mean 25 weeks). Collectively, these individuals utilised 77 roost sites in an area spanning over 100,000 km2. The study showed that movement patterns of flying-foxes varied greatly between individuals and within individuals over time. The tracked flying-foxes exhibited a range of movement patterns, from relatively sedentary to frequent long distance travel (>300 km) between multiple roosts. Migrating flying-foxes were capable of rapid sustained long-distance flight, including one movement of 500 km within 48 hours. Seasonal movements were consistent with facultative latitudinal migration in part of the population. Flying-foxes shifted roost sites frequently: 64% of roost visits lasted <5 consecutive days, although a small proportion of roost stays were of >2 months duration (maximum stay at a roost was 133 days). Data from the 14 satellite tracked flying-foxes were then used to examine the feeding movements of P. poliocephalus and to analyse the landscape characteristics of their feeding and roosting sites. Tracked individuals of P. poliocephalus used a wide range of habitats from highly urbanised areas to substantial tracts of intact forest. The landscape-scale habitat characteristics within 2 km surrounding roosting and feeding sites differed statistically. Pteropus poliocephalus tended to select roost sites in landscapes containing a higher proportion of built and cleared habitats than expected from their occurrence in the environment. By contrast, they showed some avoidance of built areas for feeding. The distances moved from roosts to feeding sites were typically less than 20 km (mean 12 km), although there was a large amount of variation within and between individuals (the range of distances moved from roosts to feeding sites was 2 - 56 km). The final data chapter in this thesis reviews the outcomes of 10 attempts by public authorities to relocate roosts of flying-foxes, with a detailed examination of one long-term relocation attempt at Maclean in north-east New South Wales. Such relocations are often attempted when flying-foxes roost in or near human settlements. However the success (or otherwise) of this approach to managing conflict between humans and flying-foxes has not previously been systematically assessed. The review found that, while some relocation actions have succeeded in moving flying-foxes from their original roost sites, in most cases the effect has been temporary, and ongoing programs of dispersal have been required to prevent reestablishment of the original roosts. When disturbances were used to disperse flying-foxes from roosts, many flying-foxes did not join pre-existing roosts, instead establishing one or
Abstract
iii more new roosts, usually close to the original site. These new roosts were rarely at predetermined target sites (i.e. likely roost sites identified by proponents of the relocation); and frequently the locations of new roost sites were no more acceptable to the broader community than the original roost site. Although seldom systematically recorded, the costs of relocation could be high; for the two instances where details were available the costs were thousands at Maclean, and millions at a site in Melbourne. The results of this research have important implications for the management of P. poliocephalus. Data from the satellite tracking study show that flying-foxes are highly mobile, with individuals conducting frequent, often long-distance movements among roosts. Therefore, it is not surprising that past attempts at resolving conflicts between humans and flying-foxes, which have involved local-scale actions such as culling and disturbances at particular problem roosts, have rarely succeeded over the long-term. In addition, the review of relocation attempts has shown that the quantum of costs and resources required for relocations are beyond the means of most small rural and regional communities. This study and related research show that P. poliocephalus displays a preference for establishing roost sites within built environments. While the factors driving this preference are not well understood, it seems likely that the location of flying-fox roosts within human settlements will continue to be an issue of community conflict into the future. The management of this conflict is complex, involving consideration both of human welfare and the ecology of a highly mobile, threatened species. Therefore, robust approaches to conservation and management of flying-foxes need to be informed by a sound understanding of the species movement patterns and habitat preferences. The information provided in this thesis should therefore contribute to better conservation and management outcomes for P. poliocephalus and flying-foxes in general.
Abstract
Statement of Originality
I, Billie Roberts, certify that this work has not previously been submitted for a degree or diploma in any university. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made in the thesis itself.
Billie Roberts September 2012
Statement of Originality
vi
Contents
Abstract ...................................................................................................................................... i Statement of Originality .......................................................................................................... v List of Tables............................................................................................................................. x List of Figures ......................................................................................................................... xii List of Plates ............................................................................................................................ xv List of Appendices ................................................................................................................. xvi Acknowledgments................................................................................................................. xvii Research permits and animal ethics ..................................................................................... xx Publications produced from this research .......................................................................... xxi Chapter 1 General Introduction .......................................................................................... 1 1.1 1.2 1.3 1.3.1 1.3.2 1.3.3 1.3.4 1.3.5 1.4 1.5 Animal movement in dynamic landscapes ............................................................ 1 Managing long-distance movements ..................................................................... 2 Ecology of Pteropus a highly mobile genus....................................................... 3 Status and distribution ........................................................................................... 3 Roosting behaviour and habitat requirements ....................................................... 4 Feeding behaviour ................................................................................................. 6 Movement capability ............................................................................................. 7 Management of Pteropus ...................................................................................... 9 Australian flying-foxes ........................................................................................ 12 Aims and structure of this thesis ......................................................................... 12
Contents
vii Chapter 2 Latitudinal range shifts in Australian flying-foxes: a re-evaluation ............ 15 2.1 2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.3 2.3.1 2.3.2 2.3.3 2.3.4 2.4 2.4.1 2.4.2 2.5 Introduction ......................................................................................................... 15 Methods ............................................................................................................... 17 The species and their distributions ...................................................................... 17 Construction of distributional database ............................................................... 18 Distributional changes in Pteropus alecto and P. poliocephalus ........................ 19 Analysis of climate within range of P. alecto ..................................................... 20 Trends in the abundance of flying-foxes within specific roosts .......................... 20 Results ................................................................................................................. 21 Number and distribution of observations ............................................................ 21 Distributional changes in Pteropus alecto and P. poliocephalus ........................ 22 Climate dynamics at P. alectos changing southern limits.................................. 26 Abundance trends in the Currie Park roost.......................................................... 27 Discussion............................................................................................................ 27 Species-specific range dynamics and their possible causes ................................ 27 Interactions between the species ......................................................................... 31 Conclusions ......................................................................................................... 32
Chapter 3 Long-distance and frequent movements of the flying-fox Pteropus poliocephalus: implications for management .............................................................. 34 3.1 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.3.1 3.3.2 Introduction ......................................................................................................... 34 Methods ............................................................................................................... 35 Study species ....................................................................................................... 35 Study area and capture sites ................................................................................ 36 Satellite telemetry ................................................................................................ 37 Data manipulation and analysis ........................................................................... 38 Results ................................................................................................................. 40 Long-term and long-distance ranging behaviours ............................................... 40 Quantitative patterns of movement and roost site use ......................................... 45
Contents
viii 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.5 Discussion............................................................................................................ 52 Patterns of movement and roost site use ............................................................. 52 Comparison with other flying-foxes .................................................................... 53 Caveats to this study ............................................................................................ 54 Management implications ................................................................................... 55 Conclusion ........................................................................................................... 56
Chapter 4 Feeding movements and landscape use by P. poliocephalus ......................... 58 4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.3 4.3.1 4.3.2 4.3.3 4.4 4.4.1 4.4.2 4.4.3 4.4.4 Introduction ......................................................................................................... 58 Methods ............................................................................................................... 60 Capture and satellite telemetry ............................................................................ 60 Feeding sites ........................................................................................................ 61 Roost sites............................................................................................................ 62 Landscape measurements .................................................................................... 63 Data analysis ........................................................................................................ 64 Results ................................................................................................................. 65 Habitat characteristics of feeding and roost sites ................................................ 67 Movement between roosts and feeding sites ....................................................... 69 Factors influencing distance moved .................................................................... 72 Discussion............................................................................................................ 76 Roost sites............................................................................................................ 76 Feeding sites ........................................................................................................ 78 Comparison between feeding and roosting sites ................................................. 79 Factors influencing feeding movements .............................................................. 79
Chapter 5 The outcomes and costs of relocating flying-fox roosts: insights from the case of Maclean .............................................................................................................. 83 5.1 5.2 5.2.1 Introduction ......................................................................................................... 83 Study region and its flying-foxes ........................................................................ 84 Flying-fox roost sites in the Lower Clarence region ........................................... 84
Contents
ix 5.1.2 5.3 5.3.1 5.3.2 5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.5 5.5.1 5.5.2 5.5.3 5.4 Maclean flying-fox roost relocation .................................................................... 86 Methods ............................................................................................................... 89 Response of Maclean flying-foxes to relocation: survey methods ...................... 89 Determining financial costs and disturbance effort ............................................. 89 Results ................................................................................................................. 90 Disturbance method ............................................................................................. 90 Disturbance of flying-foxes at the Maclean roost ............................................... 90 Cost of the relocation........................................................................................... 92 Assessment of flying-fox roost sites used since the relocation ........................... 93 Discussion............................................................................................................ 94 Effect of disturbances on site use by flying-foxes............................................... 94 Cost-effectiveness of relocation attempts ............................................................ 97 Managing flying-fox relocations in the future..................................................... 98 Conclusion ......................................................................................................... 100
Chapter 6 General Discussion .......................................................................................... 103 6.1 6.2 6.3 6.4 Key findings of each data chapter ..................................................................... 103 Patterns of roost usage and mobility by flying-foxes ........................................ 107 Ecology and management of flying-foxes in human dominated environments 110 Research directions ............................................................................................ 116
References ............................................................................................................................. 118 Appendices ............................................................................................................................ 135
Contents
List of Tables
Table 2.1 Correlations (Spearman rank, N = 8 time-periods) between parameters describing the latitudinal distribution of Pteropus alecto and P. poliocephalus and time using point and grid records.. ......................................................................................... 23 Table 2.2 Differences in four climatic variables at four latitudinal locations representing the lower quartile and most southern limit of grid records of Pteropus alecto at two time-periods, 192150 and 200107. ................................................................... 26 Table 3.1 Net displacement moved by flying-foxes among roosts in time-periods of 130 weeks duration. ..................................................................................................... 48 Table 3.2 Cumulative displacement moved by flying-foxes among roosts in time-periods of 130 weeks duration. ............................................................................................ 49 Table 4.1 A comparison of landscape attributes within 2 km radii of roosts (R), feeding sites (F) and random points in the landscape (L) at two different sampling units; (i) all feeding records and (ii) independent feeding records................................. 69 Table 4.2 Correlations (Pearson) between the feeding distances moved and roost population size (log transformed), feeding visit duration (days) and feeding forest patch size (ha) using five different units as replicates; all feeding records, independent feeding records, individuals (with more than 10 feeding records) and within two individual flying-foxes which had large (>20) numbers of feeding records together with high variation in feeding distances (Bat 5 and Bat 10). ................. 73 Table 4.3 Correlations (Pearson) between the feeding distances moved and four landscape attributes within 2 km radii of roosts at two different sampling units; all feeding records (N=173) and independent feeding records (N=112)................................ 75 Table 5.1 Estimated costs of the relocation of flying-foxes from the Maclean Rainforest Reserve and the Melbourne Royal Botanical Gardens.. ....................................... 93 Table 5.2 Summary of known documented attempts to relocate Australian flying-fox roosts using non-lethal methods, during 1990 to 2009. .................................................. 96
List of Tables
xi Table A2.1 Summary of the key records of P. poliocephalus in its northern range (north of latitude 26oS). ..................................................................................................... 139 Table A2.2 List of recent (2008) records of P. poliocephalus in the northern part of its distribution.......................................................................................................... 140
List of Tables
xii
List of Figures
Figure 2.1 Locality (point) records for (a) Pteropus alecto (N = 870) and (b) P. poliocephalus (N = 2,506) in coastal eastern Australia from 1843 to 2007. ........ 22 Figure 2.2 Box plots showing the latitudinal distribution of Pteropus alecto (upper; grey) and P. poliocephalus (lower; white) in eastern Australia between 18432007 using (a) point records and (b) 50 x 50 km grid data. .......................................... 24 Figure 2.3 Percentage of point records of flying-foxes in different latitudinal bands within eastern Australia that were Pteropus poliocephalus at four time-periods; 1843 1950, 19511990, 19912000 and 20012007. ................................................... 25 Figure 2.4 Population estimates of Pteropus alecto and P. poliocephalus in the Currie Park roost, Lismore, during two time-periods, January 1988 July 1990 and July 1998 April 2004.......................................................................................................... 27 Figure 3.1 Roosts used by 14 satellite-tracked Pteropus poliocephalus between October 2007 and April 2009. ............................................................................................ 42 Figure 3.2 North-south movement patterns of 13 satellite-tracked Pteropus poliocephalus from their points of capture. ................................................................................. 43 Figure 3.3 Weekly movement patterns of 3 satellite-tracked Pteropus poliocephalus from their points of capture. .......................................................................................... 44 Figure 3.4 Frequency (%) distributions of the duration of roost visits and the distances between consecutively-recorded roosts. ............................................................... 47 Figure 3.5 Frequency distribution of net and cumulative displacements by flying-foxes among roosts in different time-periods................................................................. 50 Figure 3.6 Number of roosts used by flying-foxes in relation to cumulative displacement during different time-periods................................................................................ 51
List of Figures
xiii Figure 4.1 Map of the 45 different roost sites (triangles) associated with the 110 discrete feeding sites (circles) in use by 14 individuals of P. poliocephalus between October 2007 and April 2009. . ............................................................................ 66 Figure 4.2 Boxplots summarising the distribution (median, quartiles, range) of elevation (m) and the percentage of forest, built and open land within 2 km of roosts (R), feeding sites (F) and random points in the landscape (L) at two sampling units. 68 Figure 4.3 Frequency distribution (%) of the feeding distances moved by the 14 P. poliocephalus; (a) all 173 feeding records and (b) 112 independent feeding records .................................................................................................................. 70 Figure 4.4 Feeding distance (km) moved by each of the 14 P. poliocephalus; (a) all records and (b) independent records. ................................................................................ 71 Figure 4.5 Frequency distribution of the duration of feeding visits by 14 Pteropus poliocephalus (N = 110 visits to different sites). ................................................. 72 Figure 4.6 Relationships between the feeding distances moved (km between roost and feeding sites) and feeding visit duration (days) for cases where there was a relatively strong (P < 0.10) correlation; (a) all feeding records (N = 173), (b) individual flying-foxes (N = 9), and (c) Bat 5 (N = 24). ...................................... 74 Figure 4.7 Mean monthly feeding distances (+SD) of 14 Pteropus poliocephalus. .............. 75 Figure 5.1 All known flying-fox roost sites in the Lower Clarence region that were occupied during the period of licensed disturbances (April 1999 to December 2007) ....... 85 Figure 5.2 Roost habitat occupied by flying-foxes in the Maclean township ....................... 87 Figure 5.3 Documented disturbance effort (person-hours) required to disperse flying-foxes from the Maclean Rainforest Reserve during the period of licensed disturbances (April 1999 to December 2007). ......................................................................... 91 Figure A2.1 Distribution of the grey-headed flying-fox (Pteropus poliocephalus) in eastern Australia ............................................................................................................. 138
List of Figures
xiv Figure A2.2 Photo of a museum specimen stated to be a grey-headed flying-fox (P. poliocephalus) collected by A.J. Boyd in 1895 from the Herbert River in Qld . 138 Figure A10.3 Distribution of the flying-foxes in Australia .............................................. 152
List of Figures
xv
List of Plates
Plate 1. Pteropus poliocephalus hanging at roost; and with a satellite transmitter attached ............................................................................................................................ xxii Plate 2. Urban flying-fox roosts in the Botanical Gardens of Sydney; and Woodend, Ipswich .............................................................................................................. xxiii Plate 3. Author holding Pteropus poliocephalus while detangled from a mist net; and collaring an anaesthetised bat in the field.......................................................... xxiv Plate 4. Plate 5. Plate 6. The North Stradebroke Island and Fraser Island capture sites. .......................... xxv The Stafford and Canungra capture sites. ......................................................... xxvi Disturbances of flying-foxes at Maclean; and a volunteer creating noise........ xxvii
List of Plates
xvi
List of Appendices
Appendix 1 Data sources, time-period of records and number of records of Pteropus alecto and P. poliocephalus. ......................................................................................... 135 Appendix 2 A re-evaluation of the northern distributional limit of the grey-headed flyingfox, Pteropus poliocephalus ............................................................................... 137 Appendix 3 Grid records for Pteropus alecto (N = 222) in coastal eastern Australia over four time-periods; 18431950, 19511990, 19912000 and 20002007. ................. 142 Appendix 4 The distribution of Pteropus poliocephalus (N = 368) in coastal eastern Australia in 50 x 50 km grids (grid records) over four time periods; 18431950, 1951 1990, 19912000 and 20002007. ..................................................................... 143 Appendix 5 Orientation of solar powered satellite transmitters on flying-foxes. ................. 144 Appendix 6 Data summary: individuals characteristics, distances moved and use of roost sites over several time-periods, for 14 satellite-tracked Pteropus poliocephalus. A ............................................................................................................................ 145 Appendix 7 Frequency distribution of the weekly net displacement of all 13 individuals that had 12 or more weeks of data. ............................................................................ 148 Appendix 8 Data summary: individuals characteristics, number of fixes and the feeding distances moved by 14 Pteropus poliocephalus at two different sampling units; all records and independent records. .................................................................. 149 Appendix 9 Correlations (Pearson) between landscape attributes of roosts, feeding sites and random points in the landscape at two different sampling units; all feeding records and independent feeding records ........................................................... 150 Appendix 10 Food for thought: Flying-foxes and bush regeneration ................................... 151
List of Appendices
xvii
Acknowledgments
Supervisors and mentors First and foremost I would like to thank my supervisors Professor Carla Catterall, Dr Peggy Eby and Dr John Kanowski for their constant encouragement, advice and friendship over the past six-plus years. I am deeply grateful for your intellectual input and forbearance through the duration of this PhD, without which this project would not have ever come to fruition. You were a truly wonderful supervisory team thank you. Thank you also to Dr Les Hall for his generosity of information, photos and for encouraging me to initially commence research on flying-foxes. He also opened up several opportunities for me at the beginning of my career including unforgettable field trips to Cairns (during a cyclone) and exploring the bat caves of Borneo. Field work and technical support I would like to mention Tim Holmes and Craig Walker from EPH for logistical support during the satellite-tracking component. John Kennedy and Allan Jeffery (OEH) also for supporting my research for three-plus years while I worked from Grafton NPWS office. I am indebted to Dr Andrew Breed for field-work opportunities near Cairns, trapping and anesthesia advice and general bat chats. A huge thank you also to Craig Smith for providing initial advice on telemetry work, collar design, and the management and organisation of satellite data. Thanks to Guan Oon and Matthew Lecointe from Collecte Localisation Satellites (CLS) for additional Argos advice. Many people assisted in the field with the trapping and collaring of flying-foxes. Thanks to Debbie Melville, Clare Hourigan, Tim Holmes, Annette Scanlon, Guy Bottroff, Trish Wimberley, Aki Nakamura, Drew Cook, Tahli Cook, Brian Roberts, Karen Roberts, Ivan Thrash and Linda Behrendorff. Debbie Melville also assisted greatly with the captive trials held at the Moggill Koala Hospital. Peter Theilmann and Allan McKinnon of Qld Department of Environment and Heritage Protection (DEHP) also provided invaluable technical help with anaesthetising animals and assisting with the original GPS collars (unfortunately these units were not viable for use on flying-foxes).
Acknowledgments
xviii Thank you to Dr Kerryn Parry-Jones and Dr Anja Divljan for additional advice on collar design, trapping and anesthetics. Anja also verified and provided advice on the identification of museum specimens. Murray Ellis and Michael Bedward, both of the NSW Office of Environment and Heritage (OEH), assisted with the ArcView script used in analyses on Chapter 2. I am also grateful to the Lower Clarence flying-fox steering committee for their advice and expertise related to the management of the Maclean flying-fox roost (Chapter 5); Peter Baumann (NSW Department of Lands), Ken Fox Laurie (Birrigan-Gargle Local Aboriginal Land Council), John Kennedy (OEH), Ian Tiley (Clarence Valley Council), Peter Wrightson (Valley Watch Inc.), Helen Tyas Tunggal (Valley Watch Inc.) and Dr Karl Vernes (University of New England). Supplementary data I am extremely appreciative to all data sources and collectors of flying-fox records for the analysis of range shifts in Chapter 2 including Dr Chris Tidemann, Dr Hugh Spencer, Dr Justin Welbergen, Dr Kerryn Parry Jones, Dr Stephan Klose, Dr Patrina Birt, Dr Les Hall, Tim Pearson, David Lowe, Greg ONeill and Carol Palmer. Many of these records were from the Australia Bird and Bat Banding Scheme (ABBBS) and information was kindly coordinated by David Drynan. Various national and international museums provided additional records (see list of names in Appendix 1). Greg Richards provided personal records and records from Francis Ratcliffe. Peggy Eby provided numerous unpublished information on sightings of Pteropus alecto in NSW and abundance data from Lismore over 20 years. I would also like to acknowledge the many people who contributed to the production of Chapter 5 by providing information related to the relocation of the Maclean flying-fox roosts and/ or usage of roost sites in the surrounding area including Gill Bennett, Carole West, Jennie Clowes, Bill Sansom, Mick Forester (Clarence Valley Council), Nigel Greenup (previously of Clarence Valley Council), and Mark Williams (formerly OEH). Thanks also to Joe Adair and Scott Sullivan (DEHP), Carol Palmer, Eddie Webber and John McCarthy (Northern Territory Government), staff at Billabong Sanctuary and Charters Towers Regional Council for providing information on relocation case studies elsewhere.
Acknowledgments
xix Greg ONeill and Mary Starky (both DEHP) provided roost location data for various sites in south-east Queensland and verified potential new roost sites used by tagged bats and the presence and abundance of known roosts thank you for your assistance. Significant others Thanks for the support and friendship of Dr Adam Smith, all the wonderful ladies of Wildlife SOS, and the crew from carpool. These friendships have kept me laughing over the past several years. Vivien Jones for the use of her beautiful photographs and Louise Saunders for the illustrations throughout the thesis. A huge thanks to my parents for helping in any way they could throughout the project and for always believing that you should work with what you love (even if it makes you unpopular and sends you broke). Last, but not least thank you to my husband for supporting me over the many years and my beautiful children Tahli, Clay and Cedar I love you all very much. Funding This study was funded by Griffith University and Queensland Department of Environment and Heritage Protection. Valley Watch Inc also contributed money to the research for Chapter 5. New South Wales National Parks and Wildlife Service (Grafton) provided additional inkind support including use of vehicles, printing and office space.
Acknowledgments
xx
Research permits and animal ethics

Three permits were required to observe, monitor and satellite track flying-foxes as part of this project including: Organisation Griffith University Animal Ethics Committee Permit applied for: Ethical Clearance to catch, release and tag bats in Australia (including during times of relocation attempts) Permit Number AES/ 17/ 06/AEC Date approved Nov 06
NSW National Parks Scientific Licence for the Purpose of S12240 and Wildlife Service Science, Education or Conservation Qld Department of Natural Resources Scientific Research and Educational Purposes Permit
March 07
WISP04268207 Feb 07
Research permits and animal ethics
xxi
Publications produced from this research

Included in Chapters 2, 3 and 5 of this thesis are published papers that are co-authored with other researchers. I was principally responsible for conducting all the research reported within these papers, and wrote, drafted and revised all papers for publication. However, each of these papers was also written with the support of my supervisors; Professor Carla Catterall (Griffith University), Dr Peggy Eby and Dr John Kanowksi. All of whom provided intellectual input, statistical guidance, and critical feedback on each manuscript. Therefore, they have been included as co-authors on each in recognition of their contribution. In Chapter 5, Gillian Bennett provided historical information and population counts for Maclean and therefore was also added as a co-author. The bibliographic details for these papers are: Chapter 2: Roberts B.J., Catterall C.C., Eby P. and Kanowski J.K. (2012) Latitudinal range shifts in Australian flying-foxes: a re-evaluation. Austral Ecology, 37: 1222. Chapter 3: Roberts B.J., Catterall C.C., Eby P. and Kanowski J.K. (2012) Long-distance and frequent movements of the flying-fox Pteropus poliocephalus: implications for management. PloS ONE, 7(8): e42532. doi:10.1371/journal.pone.0042532. Chapter 5: Roberts B.J., Eby P., Catterall C.C., Kanowski J.K. and Bennett G. (2012) The outcomes and costs of relocating flying-fox camps: insights from the case of Maclean, Australia, pp. 277-287 in The Biology and Conservation of Australasian Bats, edited by B. Law, P. Eby, D. Lunney and L. Lumsden. Royal Zoological Society of NSW, Mosman, NSW, Australia. Appendix 2: Roberts B.J., Catterall C.C., Kanowski J.K. and Eby P. (2008) A re-evaluation of the northern distributional limit of the Grey-headed Flying-fox, Pteropus poliocephalus. Australasian Bat Society Newsletter, 31: 16-19. Appropriate acknowledgements of those who contributed to the research but did not qualify as authors are included in each published paper and in the thesis acknowledgments (p. xvii). An additional research note that I co-authored has also been included in the Appendix. The bibliographic details for this are: Appendix 10: den Exter K., Roberts B.J., Underwood A. and Martin L. (2011) Food for thought: Flying-foxes and bush regeneration. Big Scrub Landcare Newsletter, Lismore.
Publications produced from this research
xxii
Plate 1. Pteropus poliocephalus hanging at roost (above, B. Roberts); and with a satellite transmitter attached (below, DEHP).
Plates
xxiii
Plate 2. Urban flying-fox roosts in the Botanical Gardens of Sydney (above); and Woodend, Ipswich (below). Photos Steve Parish collection.
Plates
xxiv
Plate 3. Author holding Pteropus poliocephalus while detangled from a mist net (above); and collaring an anaesthetised bat in the field (below). Photos DEHP taken in 2008.
Plates
xxv
Plate 4. The North Stradbroke Island (above) and Fraser Island (below) capture sites. Image Google Earth.
Plates
xxvi
Plate 5. The Stafford (above) and Canungra (below) capture sites. Image Google Earth.
Plates
xxvii
Plate 6. Disturbances of flying-foxes at Maclean in 1999 (above); and a volunteer creating noise (below). Photos G. Bennett.
Plates
Chapter 1 General Introduction

1.1 Animal movement in dynamic landscapes
Ecology has been defined as the scientific study of the distribution and abundance of organisms (Andrewartha 1961), or more precisely as the scientific study of the interactions that determine the distribution and abundance of organisms (Krebs 2001). While few ecological studies are trivial, given the multitude of factors that may interact to influence patterns of distribution and abundance, studies of highly mobile animals are particularly demanding given variations in the factors that influence habitat suitability across large spatial scales. Mobile animals exhibit a range of responses to variations in resource availability and other environmental factors (Gauthreaux 1982; Travis and Dytham 1999). In Australia, highly mobile animal taxa (birds and bats) have evolved large-scale, often complex movement patterns in response to the largely irregular and ephemeral patterns of resource availability characteristic of much of the continent (Chan 2001; Kingsford and Norman 2002; Traill et al. 2010). Many species exhibit a combination of migratory, nomadic and sedentary behaviour, depending on the spatial and temporal availability of resources (Eby 1996; Chan 2001). For the purposes of this thesis, migration is defined as the periodic (usually bi-annual) movement to and from spatially disjunct seasonal ranges (Chan 2001; Roshier and Reid 2003; Mueller and Fagan 2008). Migratory species or populations usually inhabit landscapes with regular seasonal fluctuations in environmental conditions; migration is a widespread behaviour in many vertebrate and flying invertebrate taxa (Mueller and Fagan 2008). Sedentary or resident animals are those that remain within the same locality (as relevant to the scale of movement exhibited by a species) throughout the year or otherwise defined time period (Chan 2001; Roshier and Reid 2003; Mueller and Fagan 2008). Landscapes with little spatial or temporal variation in resource availability typically facilitate resident behaviour (Mueller and Fagan 2008). Nomadism, in contrast, refers to the behaviour of animals that are neither resident or migratory, but instead move across landscapes in routes that do not repeat faithfully each season or year (Mueller and Fagan 2008). Nomadic movements occur when resources are unpredictable in both space and time (Mueller and Fagan 2008).
2 Furthermore, many species cannot be clearly categorised as sedentary, migratory or nomadic (Roshier et al. 2008). For example, while almost 40% of landbirds breeding in Australia undertake migratory movements, a large proportion of these species include both migratory and resident populations (Chan 2001). There is increasing evidence of variation in movement patterns within species and amongst apparently ecologically-similar species occupying common landscapes (e.g. water birds, Kingsford and Norman 2002; Roshier et al. 2008). One of the current challenges for animal ecology - and therefore conservation and management - is the development of a better understanding of the range of movements undertaken by highly mobile species, particularly those that occupy landscapes where resources vary widely in time and space, and determining how these movement patterns are influenced by resource availability and behavioural factors. 1.2 Managing long-distance movements
Many vertebrates, including as noted above many species that occur in Australia, have the capacity to undertake long-distance movements, particularly those animals that can fly. For example, many species of birds are known to undertake journeys of thousands to tens of thousands of kilometres (e.g. Berthold 2001; Kingsford and Norman 2002; Gill et al. 2009). Many bats also undertake long-distance movements, although it is less common for bats to travel distances of more than a thousand kilometres (Fleming and Eby 2003). Birds, bats and other vertebrates that undertake large-scale movements do so to increase their access to mates, food and other resources, and to track favourable environmental conditions; the particular factors driving such movements may differ between species, as well as between sexes and age classes within species (Berthold 2001; Fleming and Eby 2003; Newton 2007). The capacity for long-distance movement enables some species to access widely spaced and temporally variable food resources, while roosting in habitats and areas which differ from those used for feeding. Often, such species roost in colonies. Coloniality is common in birds, where about 13% of species roost in colonies and exploit spatially separated feeding sites (Rolland et al. 1998). For example, Northern Gannetts (Morus bassanus) roost in large breeding colonies on cliffs and rocky islands off the coasts of Great Britain and Ireland, while foraging up to 180 km from their colonies to obtain daily food resources (Montevecchi et al. 2002).
3 Most megachiropteran bat species (family Pteropodidae) are also colonial, roosting together during the day in groups of varying sizes and flying out to feeding sites at night (Pierson and Rainey 1992; Mickleburgh et al. 1992). In Africa, the straw-coloured fruit bat (Eidolon helvum) travels distances of more than 15 km from its colonial roosts (which can number into the millions) to feed on locally abundant fruit (Richter and Cumming 2006). A capacity for long-distance movement increases the resources and habitats available to bats and other mobile species at the cost of increased energy expenditure, vulnerability to predation during movements and time allocated to finding food in new habitats (Kunz 1982; Karasov 1992). Establishing patterns of movements of highly mobile, colonially roosting species is crucial to understanding their ecology, life history and behaviour, and a prerequisite for their effective conservation and management. For example, species that move long-distances may be particularly sensitive to habitat destruction, not only because of the loss of resources at a local scale, but also because it increases the distances between remaining resources (Price et al. 1999; Warren et al. 2001; Wilcove and Wikelski 2008). For a successful conservation program, the species entire range (i.e. feeding and roosting locations) and stopover areas need to be considered. However, large-scale and frequent movements make the tracking of individuals and populations difficult. Therefore, knowledge of the key resources that require protection throughout a species geographical range is often limited. Conservation of these species is further complicated when movements occur across areas managed by different governments or agencies (e.g. Martin et al. 2007; Fleming and Eby 2003; Epstein et al. 2009). 1.3 Ecology of Pteropus a highly mobile genus
1.3.1 Status and distribution
Flying-foxes (Old World fruit bats of the genus Pteropus) are large-bodied bats. With 61 species, Pteropus is the largest of the 42 genera represented in the Family Pteropodidae. The majority (around 63%) of species of Pteropus are highly colonial, often roosting in large groups numbering into the hundreds or thousands (Pierson and Rainey 1992; Mickleburgh et al. 1992; IUCN 2012). Some flying-foxes are considered to be highly mobile, moving daily between spatially separated feeding and roosting sites (in the order of tens of kilometres apart), and over longer time scales (weeks and months) moving frequently among different roosts (sometimes hundreds of kilometres apart) (Pierson and Rainey 1992; Mickleburgh et al. 1992; Breed et al. 2010a).
4 Flying-foxes inhabit the tropics and subtropics of Asia (including the Indian subcontinent), Australia, the Indonesian archipelago, islands off East Africa (but not mainland Africa), and a number of islands in both the Indian and Pacific Oceans (Rainey and Pierson 1992). Around 85% of flying-fox species are confined to a single island or group of islands, whereas seven species occur across large continental areas (four in Asia and three in Australia) (Pierson and Rainey 1992; Mickleburgh et al. 1992). Information on the movement capabilities of many Pteropus within oceanic islands is unknown, but it has been suggested that they are less mobile than species that occur on large continents (Pierson and Rainey 1992; Mickleburgh et al. 1992). The distributions of Pteropus species may be affected by recent changes in habitat and climate (Mickleburgh et al. 1992; Hughes 2003). However, knowledge of the distributions of flying-foxes is often based on limited information, for example specimens from museum collections which for many species are incomplete and may not reflect actual distributions (Mickleburgh et al. 2002). Many Pteropus have undergone decreases in population size in recent times due to one or more of the following factors: overhunting, habitat loss, disturbance at roost sites and persecution at feeding sites (Mickleburgh et al. 1992; IUCN 2012). Currently, half of the worlds Pteropus species are classified as threatened by the IUCN (five are critically endangered, eight endangered and 18 vulnerable; IUCN 2012). Another nine species lack the basic data required to be able to assess their status (IUCN 2012). An additional four species are considered extinct (IUCN 2012).
1.3.2 Roosting behaviour and habitat requirements
Roost sites play a key role in the ecology of flying-foxes as they provide resting habitat and sites for social interactions (Kunz and Fenton 2003). Activities which may occur at these roosts include courtship and mating, mother-infant interactions, territorial defence, grooming and resting (Pierson and Rainey 1992; Kunz and Lumsden 2003). Flying-foxes spend most of the day within their roost sites, generally between the hours of dawn and dusk (Kunz 1982), although on occasions, individuals may return to the roost before dawn. For example, in Australia, adult male P. poliocephalus are known to return to the roost within a few hours of departure to defend territories during the mating season (Welbergen 2011). The genus Pteropus typically roost openly on the branches of emergent trees or within forest canopies (Pierson and Rainey 1992; Mickleburgh et al. 1992). Only a few exceptions to tree
5 roosting have been reported: Pteropus alecto has been found roosting in a natural limestone tower in northern Australia (Stager and Hall 1983) and the extinct P. subniger of the Mascarene Islands roosted in caves, rock clefts and tree hollows (Cheke and Dahl 1981). In large flying-fox colonies, roost sites may include a number of adjacent trees (Ratcliffe 1931; Pierson and Rainey 1992). Roost site fidelity is generally high among communally roosting Pteropus (Mickleburgh et al. 1992), and some roost sites have documented histories of several decades. For example, roosting sites of P. vampyrus in the Philippines that first were reported in the early 1900s were still in use a century later (Mickleburgh et al. 1992; Brooke et al. 2000). Roosting sites of P. giganteus in India and of P. tonganus in American Samoa have been in use for at least 60 years (McCann 1934; Brooke et al. 2000), and several roost sites of P. poliocephalus in Australia have been used repeatedly for more than 80 years (Lunney and Moon 1997; West 2002). From the limited information available it appears that Pteropus species roost in a wide variety of landscape and vegetation types within their ranges (Tidemann et al. 1999; Brooke et al. 2000; Mildenstein et al. 2005). Most information on the roost habitat of flying-foxes comes from Australia, where colonies form near food sources and individuals migrate to follow temporal shifts in fruit and flower resources (e.g. Eby 1991a; Richards 1990; Birt 2005). Pteropus alecto and P. poliocephalus from Australia roost in lowland coastal areas in rainforest, mangroves and other dense riparian vegetation (e.g. Loughland 1998; Tidemann 1999). In Malaysia and Indonesia, P. vampyrus has been recorded in mangrove and lowland forests (Heideman and Heaney 1992). Little is known about roosting behaviour of Pteropus species on islands except that colony size and the characteristics of roosting sites may vary among species greatly on different islands and are influenced by the amount of hunting pressure (Wilson and Graham 1992; Brooke et al. 2000; Wiles and Brooke 2009). For example, in American Samoa, roosts of P. tonganus are located in trees on cliff faces above the ocean or on steep mountainsides, locations that are either inaccessible to people or in protected areas where hunting is not allowed (Brooke et al. 2000). The roosting ecology of solitary flying-fox species is essentially unknown because they are often cryptic and hard to distinguish within their roost vegetation (Mickleburgh et al. 1992; Pierson and Rainey 1992). The specific criteria that Pteropus species use to select roost sites are poorly understood (Kunz and Lumsden 2003). However, some habitat characteristics such as canopy structure, surrounding physical features and location of feeding habitat in relation to the roost appear to
6 be important (Kunz and Lumsden 2003). For example, the presence of waterways and nearby food resources appear to be major factors influencing roost selection in Australian Pteropus (Richards 1990; Loughland 1998; Tidemann et al. 1999; Eby 2002b; Roberts 2005). Canopy height and height of emergent roost trees were also found to be important variables explaining roosting sites of P. poliocephalus and P. alecto in Australia (Tidemann et al. 1999; Roberts 2005). A review of Pteropus worldwide by Pierson and Rainey (1992) suggested that protection from strong winds and access to updraft for flight were also important roost features. Species of Pteropus show a range of tolerances to the alteration of their habitat by humans. In the Philippines P. vampyrus has been recorded roosting in commercial orchards (Heideman and Heaney 1992) and Thailands P. lylei roosts in trees within the grounds of Buddhist temples (Brown 2007). All four species of Australian flying-foxes (Birt et al. 1998; Hall and Richard 2000) and P. giganteus in southern India roost in both rural and urban areas (Molur et al. 2008). By contrast, some species such as P. hypomelanus from the Philippines show a strong preference for undisturbed forest types away from human activities (Mildenstein et al. 2005). It has been suggested that Australian flying-foxes in recent decades have been more frequently roosting and feeding in urban areas, due to both urban encroachment on traditional roosts and increased availability of urban food resources (Birt et al. 1998; Markus 2001; Parris and Hazell 2005). By contrast, on the Pacific islands, where there is active hunting, bats are extremely wary of people, roost in inaccessible locations, and abandon sites when disturbed by humans (Wilson and Graham 1992).
1.3.3 Feeding behaviour
A review of flying-fox ecology conducted in the early 1990s reported that baseline feeding information was available for around half the species within the genus Pteropus (Pierson and Rainey 1992), and little has changed since that time (Mickleburgh et al. 2002; Mildenstein 2002). The species that have been studied feed mainly on fruits and flowers (and sometimes leaves) from a diverse range of plants (>92 genera) (Marshall 1985; Mickleburgh et al. 1992; Pierson and Rainey 1992; Eby and Law 2008). Many Pteropus species rely heavily on figs in their diet; figs often bear abundant fruit for extended periods, and trees in an area often fruit asynchronously (e.g. Mickleburgh et al. 1992; Dumont 2003). Most other plant species used by feeding flying-foxes are more synchronous and seasonal in their production of fruit and flowers (Mickleburgh et al. 1992). Across their ranges, flying-foxes have also been successful
7 at exploiting a number of introduced plants and cultivated fruit as food resources (Marshall 1985; Wiles and Fujita 1992). Pteropus species contribute to the pollination and seed dispersal of more than 289 commercially and ecologically important plant species, including tropical hardwood trees (Fujita and Tuttle 1991; Utzurrum 1995; Eby 1996; Hodgkison et al. 2003; Southerton et al. 2004). Most Pteropus are almost exclusively nocturnal, leaving their roosts at dusk to feed and typically not returning until dawn. Diurnal feeding activity may occur under unusual circumstances such as in the aftermath of cyclones, e.g. P. tonganus in Samoa (Pierson et al. 1996) and P. niger on Mauritius (Cheke and Dahl 1981), or as a more regular feature in the behaviour of some island taxa, e.g. P. samoensis in Samoa (Thomson et al. 1998) and P. melanotus on Christmas Island (Tidemann 1985). Pteropus tend to feed on conspicuous, clumped and locally abundant food resources (Marshall 1985), and will often form noisy feeding groups, with intense displays of competition (Pierson and Rainey 1990). Australian flying-foxes can remain within a single feeding area for several weeks, in which long-term feeding territories may be established (McWilliam 1986; Eby 1996). Large aggregations of Australian flying-foxes at roost sites may be associated with an abundance of food such as flower pulses from nectar-rich Eucalyptus trees (Parry-Jones and Augee 1992; Eby 1991a; Eby 1996).
1.3.4 Movement capability
The amount of available information about flying-fox movements has increased in the last 20 years (e.g. Eby 1991a; Spencer et al. 1999; Epstein et al. 2009; Breed et al. 2010a). Daily, flying-foxes move between roosts and feeding sites. Some species typically roost singly close to food resources (e.g. P. samoensis in the South Pacific, Banack 1998; Brooke et al. 2000), whereas others may roost in large colonies and move distances of tens of kilometres to feed (e.g. Australian Pteropus, McWilliam 1986; Eby 1991b). Henceforth, these daily movements will be referred to as feeding movements. The feeding areas used by flying-foxes are almost always separated from their roosting areas (Pierson and Rainey 1992; Mickleburgh et al. 1992). On large land masses, flying-foxes have been recorded moving one-way nightly distances up to 50 km from roosts to feeding sites (Marshall 1985; Eby 1991a; Pierson and Rainey 1992). Flying-foxes on islands are reported
8 as feeding more locally (Pierson and Rainey 1992). However, island flying-foxes are underrepresented in the literature and P. tonganus in America Samoa has been recorded travelling up to 47 km between feeding and roosting sites (Banack and Grant 2002). Over a period of days and weeks, individual flying-foxes may move between different roost sites, sometimes tens to hundreds of kilometres apart (e.g. Pierson and Rainey 1992; Eby 1996; Spencer et al. 1999). However, the extent of movements between roosts is unclear as previous telemetry studies have typically only examined movements on weekly or fortnightly scales (Eby 1996; Tidemann and Nelson 2004). For example, two individual P. poliocephalus showed frequent interchange among several roost sites tens of kilometres apart over timeperiods of seven or more days (Tidemann and Nelson 2004). Henceforth, these movements among roosts will be referred to as inter-roost movements. Over longer periods, some individuals may migrate hundreds to thousands of kilometres (Eby 1991a; Epstein et al. 2009; Breed et al. 2010a). Henceforth, net displacement (the shortest path between an individuals first and final roost sites in a time-period) of over 300 km in a single month will be referred to as long-distance movements. Cumulative displacement (the sum of all recorded distances between consecutive roost sites) of up to 2000 km among roost sites in a single year have been recorded for an Australian flying-fox (P. poliocephalus, Tidemann and Nelson 2004). Two separate studies of P. vampyrus and P. alecto in south-east Asia showed cumulative movements of hundreds of kilometres within a year, including across international boundaries (Epstein et al. 2009; Breed et al. 2010a). Despite these recent contributions to the understanding of flying-foxes movement capabilities, movement patterns remain poorly known for most species, particularly across broad scales. This is mainly due to the difficulty of monitoring larger-scale movements. Most movement studies have used techniques of banding individuals or radio-tracking (e.g. Nelson 1965; Eby 1991a; Markus and Hall 2004; Mildenstein et al. 2005). The limitation of these methods is that they require both initial capture and the recapture/ resighting of specific individuals or to be located using antenna and receiver; a difficult task as flying-foxes often roost in dense vegetation or inaccessible locations, and their whereabouts are often unpredictable and vary across long distances, between seasons and years (Nelson 1965; Tidemann and Nelson 2011). Despite these limitations, Nelson (1965) used dye sprayed from a light aircraft to mark around 2,000 P. poliocephalus, which revealed two individuals moving 100 km between roosts over times of three months in eastern Australia; based on
9 sightings reported by members of the public. Radio-tracking is labour intensive, but works well for local-scale studies (say, distances <50 km) (Kenward 2001). Eby (1991a and 1996), however, overcame some of these limitations by tracking P. poliocephalus from light aircraft, which revealed net displacements of over 610 km in 32 days as well as cumulative movements of up to 1800 km over 6 months for some individuals. Since the 1990s, satellite telemetry has revealed the scale and complexity of the movements of many highly mobile animals (e.g. Cohn 1999; Shimazaki et al. 2004; Higuchi et al. 2005). Over that time, satellite transmitters have become increasingly lightweight and therefore more suitable for the study of bird and bat species (Tidemann and Nelson 2004; Wikelski et al. 2007). Satellite telemetry allows locations to be recorded remotely at relatively high spatial precision over large areas, repeatedly over a long time-period. However, this technology remains relatively expensive to purchase and deploy (Thomas et al. 2011). Since 2000, several studies have used satellite telemetry to monitor the movements of flying-foxes (i.e. Tidemann and Nelson 2004; Richter and Cumming 2008; Epstein et al. 2009; Breed et al. 2010a). The first satellite study by Tidemann and Nelson (2004) involved two individual P. poliocephalus, which both moved cumulative distances >2000 km and traversed four degrees of latitude (approx. 440 km) in Australia during one year. Epstein et al. (2009) satellite tracked seven adult male P. vampyrus and showed that these flying-foxes were highly mobile, travelling cumulative distances of hundreds of kilometres between roosting sites within a year and occupied home ranges that extended beyond Malaysia to include Indonesia and Thailand. Breed et al. (2010) used satellite telemetry on nine flying-foxes of three species (P. alecto, P. vampyrus and P. neohibernicus) from six different locations in Australia, Papua New Guinea, Indonesia, and Timor-Leste, which were recorded for a median 120 days (range 47342) days with a median total cumulative distance travelled of 393 km (range 793,011 km) per individual. Richter and Cumming (2008) tracked four African straw-coloured fruit bats (Eidolon helvum) and demonstrated feeding movements as far as 59 km from the roost, net displacements of 370 km per night and cumulative distances of up to 2,518 km in 149 days.
1.3.5 Management of Pteropus
Flying-foxes across their range regularly come into conflict with humans for three main reasons. First, their feeding behaviour leads them into conflict with domestic and commercial fruit growers (Mickleburgh et al. 1992; Waples 2002). In areas where fruit is grown on a commercial scale and where fruit trees have replaced native forest, such as in parts of
10 Australia, India, Malaysia and Indonesia, flying-foxes have been the target of large-scale eradication campaigns (Mickleburgh et al. 1992; Hall and Richards 2000; Epstein et al. 2009). For example, in Malaysia, unrestricted culling of flying-foxes (P. vampyrus and P. hympomelanus) has been permitted as they are considered to be agriculture pests in fruit orchards (Epstein et al. 2009). In India, all five species of flying-fox present have been classed as vermin due to the damage they can cause to fruit crops, enabling their unrestricted capture and killing (Singaravelan et al. 2009). Damage is caused when flying-foxes bite into fruits, or scratch their surfaces, making the fruit unmarketable. The level of fruit crop damage varies considerably among locations and years (Mickleburgh et al. 1992; Eby 1995). In Australia, cases of extensive use of commercial crops by flying-foxes have been linked with times when the availability of native food sources is low (Tidemann and Nelson 1987; DECCW 2009). Second, flying-foxes are reservoirs of a number of zoonotic diseases which can infect livestock and humans, including newly described zoonotic viral diseases (e.g. Hendra and Nipah virus: Fraser et al. 1996; Halpin et al. 2000; Johara et al. 2001). These viruses can cause significant morbidity and mortality in humans and domestic animals such as pigs and horses (Field and Kung 2011). Since two viruses (Lyssavirus and Hendra virus) were discovered in Australian flying-foxes in the mid-1990s (Fraser et al. 1996; Halpin et al. 2000) there has been increased research into viruses associated with bats worldwide (Breed et al. 2010b). In Australia, the discovery of these bat associated viruses has also raised a negative profile of flying-foxes in the media (West 2002; Lunney and Moon 2011), with some commentators and politicians advocating that flying-foxes be reduced in number, either by shooting them or by moving them away from particular roost sites (West 2002; Hyne 2010; Lunney and Moon 2011). Pteropus in Thailand have been killed due to the perceived threats of retroviral disease transmission to humans; in addition to their persecution due to superstition (Brown 2007). Third, the proximity of flying-fox roosts to human settlements can cause problems because of the noise and smell associated with roosts (Mickleburgh et al. 1992; Hall and Richards 2000; Eby and Lunney 2002a). This is particularly an issue for species that are strongly colonial (Mickleburgh et al. 1992). Some pre-existing roosts have become surrounded by urban development, while in other cases flying-foxes have set up new roosts in existing urban or suburban areas. For example, in Indonesia and Australia, planted trees such as in botanical
11 gardens in cities have provided new roosting opportunities for species such as P. vampyrus and P. poliocephalus (Mickleburgh et al. 1992; Richards 2002). In India and Pakistan, P. gigantus roosts in large colonies of hundreds to thousands of individuals in large trees in rural and urban areas, close to agricultural fields, ponds and by the sides of roads (Molur et al. 2008). In Australia, the level of conflict that arises has been related to the distance between flying-fox roosts and areas of human activity, and to the number of individuals in the roost (Eby and Lunney 2002b). Not all roosts near human habitation are sources of conflict and there are a number of examples of roosts in eastern Australia that are being managed successfully in urban settings (e.g. Pallin 2000; DECCW 2009). Historically, people have used destructive management methods aimed at individuals or their roosts to manage conflicts. Methods have included the total removal of roost trees and draining of swamps inhabited by flying-foxes, shooting animals within roosts and while feeding, surrounding orchards with high-voltage electric grids, capturing animals at orchards and killing with chloroform, spraying animals with diesel, and the use of explosives and fire within roosts (e.g. Dolbeer et al. 1988; Mickleburgh et al. 1992; Tanton 1999; West 2002; Hall 2002). The aim of these actions has been to control or manage populations by either reducing population numbers or inducing the flying-foxes to move elsewhere. Following European settlement of Australia from the 1800s, shooting flying-foxes within their roosts was a common management practice to prevent damage to fruit crops, but this practice has been largely prohibited or restricted since the early 1990s (Ratcliffe 1931; Eby and Lunney 2002a). Despite widespread use of these lethal techniques, flying-foxes have often continued to use roosts and feeding areas, perpetuating human-bat conflicts. Consequently there is a growing interest in the use of non-destructive methods of managing conflicts between flying-foxes and humans (Mickleburgh et al. 1992; Mickleburgh et al. 2002; IUCN 2012). In Australia, nondestructive techniques used to reduce crop damage from flying-foxes include the deployment of bright lights, scare guns, smoke, loud noise and bat distress calls (Eby and Lunney 2002a). Netting is effective in eliminating crop damage; however, its use is restricted because it is expensive to erect and maintain (Waples 2002). At roost sites, a variety of techniques have been used to harass the animals until they shift elsewhere, including the use of helicopters, bird-scare guns, smoke and loud continuous noise (Mickleburgh et al. 1992; Vardon et al. 1997; Tidemann 2002).
12 1.4 Australian flying-foxes
There are four Australian species of flying-fox: the little red flying-fox P. scapulatus, black flying-fox P. alecto, spectacled flying-fox P. conspicillatus and the grey-headed flying-fox P. poliocephalus. These species are all strongly colonial and roost in large aggregations which at times can exceed 100,000 individuals (Eby 1991a; Churchill 1998; Tidemann et al. 1999). In areas where their ranges overlap, these species may share daytime roost sites, although species are often stratified within the vegetation. For example, where P. scapulatus and P. alecto co-occur, P. scapulatus typically roosts lower in the vegetation, with P. alecto higher in the trees and at the roost periphery (Birt and Markus 1999; Roberts 2005). Roost fidelity is strong, with some of the roost locations described in the 1920s (Ratcliffe 1931) remaining in use in 2011 (pers. obs). Where flying-fox colonies occur near human habitation, residents have sometimes advocated for the flying-foxes to be relocated elsewhere (Birt et al. 1998; Hall and Richards 2000; Eby and Lunney 2002a). However, there has been frequent debate about the principle of relocation as well as the success of relocation methods in particular circumstances (Hall 2002; Tidemann 2002; West 2002). Two of the Australian flying-fox species, P. alecto and P. poliocephalus, are similar in roost selection, diet and movement patterns (McWilliam 1986; Eby 1996; Tidemann 1999; Hall and Richards 2000; Markus and Hall 2004). These species also show substantial range overlap in eastern Australia, although they also have considerable parts of their ranges that are discrete. Both these species have been reported as undergoing distributional changes during the 20th century (Hughes 2003; Markus et al. 2008; Tidemann et al. 2008), with the southern range limit of P. alecto extending southward (Eby et al. 1999; Tidemann 1999; Markus et al. 2008), and both northern and southern limits of P. poliocephalus also being reported as shifting southward (Tidemann 1999; van der Ree et al. 2006; Tidemann et al. 2008). However, some of these inferences have been contested (van der Ree et al. 2006; DECCW 2009). Pteropus poliocephalus has been listed both nationally and internationally as Vulnerable resulting from the combined effects of habitat loss, agricultural control efforts, presumed competition with P. alecto and disturbance at roosts (Duncan et al. 1999; IUCN 2012). 1.5 Aims and structure of this thesis
The overarching aim of this thesis is to provide new information on the ecology of the greyheaded flying-fox (Pteropus poliocephalus), particularly its movement patterns. This study
13 uses satellite telemetry to examine the feeding movements and habitat use of P. poliocephalus across broad spatial scales. It tracks animals at a combination of a much finer temporal resolution (i.e. every few days) and over a longer time period (i.e., >6 months) than most previous studies. The methodology and results of this project are also relevant to the conservation and management of other wide-ranging flying-fox species, both within Australia and elsewhere. More specifically, this thesis aims to:
quantitatively re-assess the temporal dynamics of distributions of P. poliocephalus and its congener P. alecto in eastern Australia over 150 years, using historical locality records from a wide range of sources;
provide spatial and temporal information on inter-roost movements of P. poliocephalus across its range, using satellite telemetry;
quantitatively compare the selection of roosting and feeding habitat by P. poliocephalus, and describe its pattern of movement between feeding and roosting sites; and
examine the consequences of attempts to relocate flying-fox roosts across a range of sites in eastern and northern Australia and consider the effectiveness of such actions in resolving humanbat conflicts.
The thesis is divided into six chapters: four that address the four objectives listed above, this Introduction and a General Discussion. Chapters 25 are structured as manuscripts for publication. Chapter 2 quantitatively re-evaluates the historical and present (from 1843 to 2007) distributions of P. poliocephalus and P. alecto in eastern Australia and considers the evidence that global warming has been the driver of distributional change. Chapter 3 quantifies the pattern of inter-roost movement by P. poliocephalus at different spatial and temporal scales, using satellite telemetry of 14 individuals for up to nine months and an area spanning over 100,000 km2, and discusses the implications of the findings for management. Chapter 4 quantifies the frequency distribution of feeding distances travelled during nightly feeding movements by P. poliocephalus, using satellite telemetry of 14 individuals. Chapter 4
14 also examines the landscape-scale factors associated with selection of both roost and feeding habitats. Chapter 5 provides a quantitative summary of both the activities and financial costs involved in a series of public actions to relocate a flying-fox roost in the township of Maclean, eastern Australia. This Chapter also reviews the occurrence and outcomes of other relocation attempts, and considers the implications for relocation as a management tool to resolve conflict between humans and flying-foxes. Chapter 6 summarises the major findings of the previous chapters and discusses the implications for current knowledge of flying-fox ecology and the effectiveness of different approaches to their management.
Louise Saunders
15
Chapter 2 Latitudinal range shifts in Australian flyingfoxes: a re-evaluation

2.1 Introduction
The study of the geographic distributions of species and the reasons why species occur where they do are fundamental themes in ecology (Krebs 2001). There is evidence of recent changes in the distributions of many species (Walther et al. 2002; Parmesan and Yohe 2003; Hickling et al. 2006), including plants (Grabherr et al. 1994; Sturm et al. 2001), birds (Thomas and Lennon 1999; Hockey and Midgley 2009) and butterflies (Parmesan 1996; Parmesan et al. 1999). Distributional change may occur in a variety of ways: for example, away from or into areas used intensively by humans (Catterall et al. 1998), or contraction from, or expansion into, fragmented habitat (Fahrig 2003). A commonly observed distributional shift in recent decades consists of range expansion towards higher latitudes or higher elevations (Walther et al. 2002; Parmesan and Yohe 2003; Root et al. 2003). Because climate change is predicted to shift species distributions towards higher latitudes and elevations, the emerging impacts of climate change have increasingly been held responsible for these observed distributional shifts (Walther et al. 2002; Hughes 2003; Parmesan and Yohe 2003; Root et al. 2003). These concerns have stimulated renewed interest in the influence of climatic variables on the range limits of species. However, there are risks associated with attributing observed changes in species distributions to climate change. First, given there is only scanty data on the distribution of most species in lightly settled regions, such as most of regional Australia, there is a risk that putative range shifts may be prematurely inferred from limited data. A second, related risk, is that observed changes in species distributions may be attributed to climate change without adequate scrutiny of alternative factors. Climate change is only one of many potential factors which may cause changes in the distributions of species. Of course, climate is a fundamental characteristic of habitat (Andrewartha and Birch 1954), but recent changes in temperature are relatively modest (for most areas of Australia, an increase in mean temperatures of <1oC over the past century: CSIRO 2007) compared with seasonal or daily variations in temperature tolerated by most species. Other factors which may drive shifts in species ranges include various types of
Chapter 2 Latitudinal range shifts in Australian flying-foxes
16 habitat modification, such as changes in resource availability, habitat loss and fragmentation; biotic interactions (e.g. competition, predation) amongst species; hunting and other forms of exploitation of populations by humans, and conversely the relocation of species to new areas by humans (Bell and Austin 1985; Towns and Daugherty 1994; Madsen and Fox 1995; Davis et al. 1998; Warren et al. 2001; Fahrig 2003; Root et al. 2003; Parmesan et al. 2005; Hockey and Midgley 2009). These factors often interact to drive shifts in species distributions (Brown et al. 1996; Hughes 2000; Gaston 2003). Consequently, attributing changes in species distributions to climate change needs to done cautiously: first, there needs to be robust evidence that a change in distribution has occurred; and secondly, attribution of any observed shift to climate change should also be based on sufficiently thorough analysis, and due consideration given to potential alternative hypotheses. Demonstrating that distributional shifts have occurred or are occurring requires statistically robust quantitative comparisons of historical and present data. Although there are good examples of such studies in the northern hemisphere (Walther et al. 2002; Parmesan and Yohe 2003; Root et al. 2003), there are few well-documented examples of latitudinal shifts for Australian terrestrial species. It has been suggested that the southern range limit of the black flying-fox (Pteropus alecto) in Australia has extended southward during the past century (Eby et al. 1999; Tidemann 1999; Hall and Richards 2000; Markus et al. 2008), and that both the northern and southern range limits of the grey-headed flying-fox (P. poliocephalus) have also shifted southwards (Tidemann 1999; van der Ree et al. 2006; Tidemann et al. 2008). These changes in distribution were initially attributed to climate change by Tidemann (1999). Although untested, this idea has been propagated by subsequent authors (Tidemann and Nelson 2004; van der Ree et al. 2006; Welbergen et al. 2007; Markus et al. 2008; Tidemann et al. 2008), cited as evidence supporting climate-change impacts (Hughes 2003), and repeatedly quoted in public statements as a local exemplar of such impacts. In the present study I undertake a quantitative analysis of historical and present distributions of P. alecto and P. poliocephalus in eastern Australia. The results are used to evaluate the robustness of previously published inferences concerning distributional shifts in these species. The data are further analysed to evaluate the contention that global climate change (in particular, changes in temperatures and rainfall) has been a driver of these shifts.
17 2.2 Methods
2.2.1 The species and their distributions
Pteropus alecto is widespread in coastal sub-tropical and tropical northern and eastern Australia (Hall and Richards 2000; Churchill 2008). It is also found in parts of southern New Guinea, and the Indonesian islands of Sumba, Lombok and Celebes (Markus et al. 2008). In northern Australia P. alecto has been found as far as 250 km inland (Thompson 1991), but typically occurs in coastal areas (Hall and Richards 2000). Since the 1930s P. alecto is reported as expanding its southern range limit from Maryborough (25.5oS) to Sydney (33.9oS), and this southward shift has been described as being particularly rapid since 1990 (Tidemann 1999; Markus et al. 2008; DECCW 2009). Pteropus poliocephalus is endemic to coastal eastern Australia, at mainly subtropical and temperate latitudes (Tidemann 1999; Tidemann et al. 2008). The current distribution of P. poliocephalus is reported as extending from around Maryborough (25.5oS) to Melbourne (37.8oS) (Tidemann 1999; Tidemann et al. 2008). The inland boundary is difficult to define but typically this species occurs from the coast to at least the Great Escarpment throughout the species range and as far inland as the tablelands and western slopes of the Great Diving Range in northern New South Wales and southern Queensland (Tidemann et al. 2008; DECCW 2009). It has been suggested that the northern range limit of P. poliocephalus has contracted by up to 750 km over the past 100 years (Tidemann 1999; Tidemann et al. 2008). In regions where the distributions of the two species overlap, they share daytime roosts and night time foraging habitat. Roost sites (roosts) typically comprise between hundreds and thousands of individuals and occur in dense, riparian or wetland forests including rainforests, Melaleuca forests, and mangroves (Eby 1991a; Tidemann 1999; Hall and Richards 2000; Roberts 2005). Roosts can be used on a seasonal or continuous basis, and the numbers of bats within a roost can fluctuate widely, associated with variation in surrounding flower and fruit resources (Eby 1991a; Tidemann 1999; Hall and Richards 2000). Both species are very mobile. Long-distance movements of more than 600 km in P. alecto (A. Breed pers. comm.) and more than 800 km in P. poliocephalus (Tidemann and Nelson 2004) have been recorded during a time-period of one month. Within a typical nights foraging, species commonly travel up to 20 km from the roost (Eby 1991a; Palmer and Woinarski 1999; Tidemann 1999; Markus and Hall 2004), often across cleared and fragmented landscapes.
18
2.2.2 Construction of distributional database
All available locality information for P. alecto and P. poliocephalus in eastern Australia were collated, from the time of early European exploration to December 2007. The study area extended from the northern to the southern extremities of the mainland, and 250 km inland to include the western slopes of the Great Diving Range. The resulting database contained information on the species (P. alecto or P. poliocephalus), date, latitude, longitude, name or description of locality, source of the record, observed behaviour (roosting, foraging or unspecified), and any information on numbers of individuals present (if available). Locality records were obtained from the following sources: the Australian Bird and Bat Banding Scheme; state and international museums; state government wildlife atlases and the Australian National Wildlife Collection; published and unpublished records from reputable sources (scientific papers, environmental reports, research notes); and researchers field surveys (Appendix 1). All localities were converted to decimal degrees. The total number of records compiled in the study (raw data) was 15,085 (P. alecto 3,442, P. poliocephalus 11,643: Appendix 1). The data were then filtered as follows. Records were deleted if they lacked information on location (latitude and longitude or location description), year or species. Repeated sightings at the same location in the same year were collapsed into one record per location per year per species. Records considered to be vagrants, defined as roost locations with less than 10 individuals were also deleted (N = 22). However, most records did not include information on the number of individuals present and may include vagrant sightings. A randomly selected subset (10%) of museum specimens, together with all museum specimens collected at the two species northern and southern range boundaries, were re-examined either by the museum curator or B. Roberts to determine whether the species had been correctly identified. Some erroneous early records at or near range limits were consequently removed (see Appendix 2). But recent museum records were found to be reliable. All non-museum records were also screened for reliability. For example, data compiled in state government wildlife atlases contained information from a range of sources including published literature, environmental consultants reports and other grey literature, unpublished field surveys and incidental observations. To ensure that the analysed data were reliable, records were only retained if an observer was named, and if this person was
19 considered to have the expertise needed to identify flying-foxes to species level (e.g. longterm naturalists, researchers, wildlife carers and experienced government staff); observers levels of expertise were assessed through consultation with experts familiar with the experience of individual collectors. To screen for spatial accuracy, the data were imported into ArcView (Version 3.3) and overlaid with geographical landmarks and bioregional boundaries to identify localities that were obviously outliers. Records (N = 26) that were obviously erroneous (i.e. in the ocean or that were >500 km outside likely range boundaries) were removed. At the conclusion of the screening process, 3,376 records (P. alecto 870, P. poliocephalus 2,506: Appendix 1) remained and this dataset is referred to as point records. For analyses, the point records were grouped in two ways, into either eight time-periods: 18431920, 19211950, 1951 1960, 19611970, 19711980, 19811990, 19912000 or 20012007, and four time-periods: 18431950, 19511990, 19912000 and 20012007. Observation effort varied spatially and temporally. To reduce consequent biases in the data, point records were converted to presence / absence records within 50 x 50 km grid cells. The scale was chosen to encompass the maximum known nightly foraging distances of P. alecto and P. poliocephalus (Eby 1991a; Tidemann 1999; Markus and Hall 2004). A grid lattice covering the study area was generated in ArcView (Version 3.3), with grid cells defined by their centre points. Each cell of the lattice was interrogated for presence of point records of each species within each of the eight time-periods, yielding 222 grid records for P. alecto and 368 for P. poliocephalus.
2.2.3 Distributional changes in Pteropus alecto and P. poliocephalus
Box plots were constructed from both the point records and grid records to describe the latitudinal distribution of each species across time. Whether the distribution of each species had expanded or contracted in a north-south distance over time was assessed by calculating Spearman rank correlation coefficients between time-periods (ranked) using each of five distributional variables: the median, minimum, maximum, and upper and lower 25% quartiles of the species latitudinal occurrence. To investigate changes through time in the proportion of each species across eastern Australia, the percentages of point records attributed to P. poliocephalus in six uniform
20 latitudinal bands (1014, 1519, 2024, 2529, 3034, 3539oS) were graphically compared among the four time-periods.
2.2.4 Analysis of climate within range of P. alecto
The influence of climate on the southern limit of P. alecto was investigated using temperature and rainfall data obtained from Australian Bureau of Meteorology (BOM 2010). I selected weather stations corresponding to each of four spatio-temporal options: the latitudes of lower quartile and most southern limit of grid records in both the 19211950 and 20012007 timeperiods. Suitable weather stations were sought on the basis of the following criteria: located within one degree of latitude (100 km) of the central point of the relevant grid cell; elevation < 20 m; and located within 30 km of the coast. Only one weather station at each of the four latitudinal locations met all these criteria: 19211950 lower quartile St Lawrence (22.35oS, 149.54oE); 19211950 southern limit Maryborough (25.52oS, 152.72oE); 20012007 lower quartile Archerfield Airport (Brisbane) (27.57oS, 153.01oE); and 20012007 southern limit Sydney Airport (33.86oS, 151.21oE). Climatic variables considered were: annual mean minimum and maximum temperatures, annual rainfall and mean annual number of days < 20oC (frost potential). A complete set of data was available across all four weather stations from 1939 to 1949, and data from this period were therefore used to indicate climate in 1921 1950 (weather data were incomplete at many stations prior to 1939 and in 1950).
2.2.5 Trends in the abundance of flying-foxes within specific roosts
Data on species abundances within specific flying-fox roosts were compiled from available roost counts at locations within the latitudinal zone in which distributional change had been suggested in the literature (25oS34oS), and the time-period of transition as revealed through analysis in the present study (19502007). Only one roost, Currie Park (28.8oS, 153.3oE), had reliable long-term abundance data for both P. alecto and P. poliocephalus that were collected using the same counting technique (exit counts) in the same months of the year (April and July), with the earliest count year being 1988. Exit counts involve a team of observers stationed around the roost who count the numbers of flying-foxes as they depart at dusk to feed (Eby et al. 1999). Data for Currie Park were compiled from published and unpublished surveys over two time-periods: 19881990 (Eby and Palmer 1991; Eby unpublished); and, 1998-2004 (Eby 2002a; Eby 2003; Eby 2004). Data were collected approximately monthly in 1988-1990 and during April and July in 19982004. The abundances of both species at Currie
21 Park were compared between 19881990 (N = 3) and 1998-2004 (N = 6) using MannWhitney U-tests. Abundance in each year of data was the number of each species recorded in exit counts for either April or July, or the average of these months if both were available. 2.3 Results
2.3.1 Number and distribution of observations
The earliest point record in the database was of a P. poliocephalus collected at Cape Upstart (19.8oS, 147.8oE) in 1843 (Dobson 1878; Anderson 1912). Over 41% of records of P. alecto and 27% of records of P. poliocephalus were from daytime roost sites (Appendix 1); only one record of P. alecto and five records of P. poliocephalus were at foraging locations; the remainder may have been from either roosting or foraging locations. The 870 point records for P. alecto were spread from 10.8oS (the northern tip of Cape York) to 33.9oS (Sydney), while 2,506 point records for P. poliocephalus ranged from 19.8oS (near Bowen) to 38.9oS (southernmost Victoria) (Figure 2.1). Records for both species were concentrated around latitudes 27.5oS and 34oS, corresponding with the human population centres of Brisbane and Sydney. The number of records increased markedly in the 19912000 and 20012007 time-periods and there was a substantial increase in records of P. alecto in a restricted latitudinal zone between 27.4oS (Brisbane, Qld) and 28.4oS (Murwillumbah, NSW) during 20012007 (Figure 2.1a). This was associated with a series of targeted surveys of flying-foxes during that time to document winter concentrations of P. poliocephalus in the region (e.g. Eby 2002a), which also recorded P. alecto. This effect was eliminated by the conversion of data to grid records (Figure 2.1b.).
22
Figure 2.1 Locality (point) records for (a) Pteropus alecto (N = 870) and (b) P. poliocephalus (N = 2,506) in coastal eastern Australia from 1843 to 2007. Eastern records were limited to within 250 km from the coastline. Place names: Mk = Mackay, R = Rockhampton, Mb = Maryborough, B = Brisbane, S = Sydney and M = Melbourne.
2.3.2 Distributional changes in Pteropus alecto and P. poliocephalus
There was a significant correlation between time and the most southerly record of P. alecto from both point and grid records (Table 2.1). The southern limit of known records of P. alecto expanded southward between 1912 and 2007 by 1,168 km or approximately 10.5o (Figure 2.2; Appendix 3). Prior to 1920, P. alecto was not recorded further south than 23oS (near Rockhampton) but by 2007 the species had been recorded at 34oS (in Sydney). Both the upper and lower quartiles of grid records for P. alecto also shifted southward over time (Figure 2.2) in a statistically significant manner (Table 2.1). However, there was no significant shift in its northernmost limit (Table 2.1).
23
Table 2.1 Correlations (Spearman rank, N = 8 time-periods) between parameters describing the latitudinal distribution of Pteropus alecto and P. poliocephalus and time using point and grid records (see Figure 2.2). * P < 0.05; ** P < 0.01.
P. alecto Point records Most northern record Upper quartile Median Lower quartile Most southern record -0.26 0.74* 0.52 0.74* 1.00** Grid records -0.41 0.77* 0.25 0.81* 0.99**
P. poliocephalus Point records 0.00 0.31 -0.40 -0.45 0.33 Grid records -0.07 0.02 -0.33 0.08 0.49
There were no significant changes in either the northern or southern range limits or distributional quartiles of P. poliocephalus (Table 2.1; Figure 2.2; Appendix 4). There were records in every decade between 1880s and 2000s of P. poliocephalus from the southernmost latitudes of continental Australia. At the northern limit, records of P. poliocephalus south of 23oS (Rockhampton) were common and consistent through time, and records of P. poliocephalus north of 23oS were rare both in the earlier historical periods (18431950) and in recent decades (19902007). Only two historical records occurred north of 23oS. Each was collected in the late 1800s during scientific expeditions. Pteropus poliocephalus was not recorded again north of 23oS until the 1990s from a single sighting of large numbers at Eungella in 1991 together with repeat sightings between 2005 and 2007 from the nearby Finch Hatton roost (both west of Mackay; 20.9oS). Close scrutiny of historical records in the far north of the species range (north of 20.9oS) revealed that most were either questionable or incorrect (see Appendix 2).
24
(a) Point records
(b) Grid records
Figure 2.2 Box plots showing the latitudinal distribution of Pteropus alecto (upper; grey) and P. poliocephalus (lower; white) in eastern Australia between 1843 2007 using (a) point records and (b) 50 x 50 km grid data. Unless otherwise stated the year label shows the mid-year of the time-period. Numbers above bars are sample sizes (number of records or number of grids).
Vagrant records were re-introduced into the dataset to assess their impact on results. The only effect on latitudinal distribution was a southward shift of the most southern record of P. alecto in the 19811990 time-period by two degrees (from 28.4oS to 30.6oS), resulting from a single observation; this did not affect the statistical relationships. Including vagrant sightings (N = 12) had no effect on the latitudinal distribution of P. poliocephalus.
25 In the zone of P. alectos southward expansion (2529oS), the percent of records that were P. poliocephalus declined significantly, from 92% (of 13 records) prior to 1950 to 51% (of 619 records) by the early 2000s (Figure 2.3; test of association chi squared = 8.682, P = 0.003, df = 1).
Figure 2.3 Percentage of point records of flying-foxes in different latitudinal bands within eastern Australia that were Pteropus poliocephalus at four time-periods; 18431950, 19511990, 19912000 and 20012007.
26
2.3.3 Climate dynamics at P. alectos changing southern limits
The mean annual maximum and minimum temperatures at locations representing P. alectos southern range limit in 192150 (southernmost record and lower quartile) increased by 0 2.0oC between 193949 and 200107 (Table 2.2). However the 200007 temperatures at the new locations representing its southern limits (in 200007) were substantially cooler than 193949 temperatures at its 192150 southern limits (effective decreases of 0.74.3oC in mean annual maximum and minimum temperatures, Table 2.2). Therefore, in spite of an overall pattern of warming over time at all locations, P. alecto experienced temperature decreases in the southern part of its distribution as a result of its range expansion. The limited available frost data (annual days < 2oC) indicate that P. alectos lower distributional quartile in 200007 includes frost-prone locations. Annual rainfall was generally lower in 200007 than in 193949, but did not vary consistently or meaningfully with latitude (Table 2.2).
Table 2.2 Differences in four climatic variables at four latitudinal locations representing the lower quartile and most southern limit of grid records of Pteropus alecto at two time-periods, 192150 and 200107. Climatic data from 1939 to 1949 represent the earlier period due to incomplete records pre-1939. Effective difference is the estimated temperature difference actually experienced by P. alecto as a consequence of changed distributional limits; with the relevant before and after temperatures highlighted in bold.
Annual max Annual min temperature oC temperature oC Latitudinal locations Lower quartile 1921-1950 2001-2007
Mean no. days <2oC 1939 49 na 7.5 2001 07
Annual rainfall (mm) 1939 49 2001 07
Approx. o S
1939 49
2001 07
1939 49
2001 07
22 28
28.5 25.5
29.0 26.8 -1.7
16.8 13.4
18.2 14.5 -2.3
0.0 2.9 -
870 920
683 803 -67
Effective difference Most southern record 1921-1950 2001-2007
26 34
27.3 21.6
27.3 23.0 -4.3
15.3 12.6
15.9 14.6 -0.7
na 0.7
2.0 0.0 -
1059 854
987 919 -140
Effective difference
The annual daily maximum or minimum air temperature per year, averaged over all years for either 20012007 or 19391949. Data not available. St Lawrence, Queensland (22.35S, 149.54E) Archerfield Airport, Brisbane, Queensland (27.57S, 153.01E) Maryborough, Queensland (25.52S; 152.72E) Sydney Airport (33.86S; 151.21E)
27
2.3.4 Abundance trends in the Currie Park roost
The abundance of P. alecto during April and July increased significantly in the Currie Park roost between 19881990 and 19982004 (U = 0.0, P = 0.02: Figure 2.4). Numbers of P. poliocephalus in 19881990 generally exceeded those during 19982004 (Figure 2.4), but there was no significant difference in either the abundance of P. poliocephalus (U = 4.5, P = 0.24) or the total number of flying-foxes (U = 5.0, P = 0.30).
Figure 2.4 Population estimates of Pteropus alecto and P. poliocephalus in the Currie Park roost, Lismore, during two time-periods, January 1988 July 1990 and July 1998 April 2004. Data are from roost exit counts in specific months which varied among years. The x-axis tick marks show December in each year. Approximately 1,000 P. alecto were first recorded in the Currie Park roost in January 1985 (L. Hall unpublished).
2.4
Discussion
2.4.1 Species-specific range dynamics and their possible causes
This study has detected a clear range shift to more southern latitudes by P. alecto, despite sampling variability and potential biases inherent in the use of data compiled from a variety of sources which include volunteer observers. Because P. alecto and P. poliocephalus commonly occur together and are morphologically and ecologically similar, the same biases (in particular, risk of misidentification) would presumably apply to both species. Therefore, the species differences in geographical range dynamics as revealed by the present study are very unlikely to be an artefact of sampling. From the late 1800s to 2007, P. alecto expanded its southern range boundary southward by 123 km per decade, on average. This rate is broadly similar to the poleward range-limit movement reported for some butterflies (up to 193 km/decade, Parmesan et al. 1999; Crozier 2004), but generally higher than observed for birds
28 (920 km/decade, Thomas and Lennon 1999; Hickling et al. 2006). In cases where the southward movement of P. alecto into new areas has been observed at particular sites, there have been initial sightings of tens to hundreds of individuals, which were then followed by sometimes-rapid increases of up to tens of thousands (although the number varies greatly), such as documented here for Currie Park. It is likely that some time may elapse from the time P. alecto first arrives in an area until it builds up to sufficient numbers to be noticed by observers, because P. alecto typically roosts with P. poliocephalus, and the two species are morphologically similar on casual observation. There are two major environmental changes that have previously been used to explain the shifts in flying-fox distribution, namely climate change and habitat modification. First, the poleward movement of P. alecto is superficially consistent with climate change scenarios (Hughes 2003). Tidemann (1999) suggested that increases in average rainfall and temperatures in eastern Australia have favoured the southward shift in P. alectos southern range limit, with a reduction in number of days with frosts being of particular importance. Neither the rainfall patterns reported here nor the more widely observed declining temporal rainfall trend in eastern Australia (CSIRO 2007) support a rainfall hypothesis. Australias continental average temperature has increased by approximately 0.9oC since 1910 and 0.1 0.2oC per decade since 1950 (Collins 2000; Hughes 2003; CSIRO 2007), including areas where P. alectos range has expanded. In general, a temperature increase of 0.10.2oC per decade would correspond with a latitudinal shift in isotherms of approximately 1026 km/decade (Hughes 2000). In contrast, P. alectos southward expansion rate has been much faster: about 100 km/decade. Furthermore, the analyses show that P. alecto has progressively established roosts in new areas that are considerably cooler than experienced in its previous historical distribution. While a general trend to warmer temperatures across Australia may have led to a reduction in frost frequency in some locations (Stone et al. 1996; CSIRO 2007), the occurrence of frost is strongly influenced by local topography, and P. alectos range expansion has been confined to the coast where frost potential is low and has not changed greatly with time. The data presented here do not support the hypothesis of climate change as a primary driver of P. alectos southward range expansion. The second environmental factor suggested to explain shifts in flying-fox distribution is habitat change. A large proportion of eastern Australian forests and woodlands have been
29 converted to agriculture and urban development during this studys time-frame, particularly affecting coastal lowlands and sub-coastal river basins (Catterall and Kingston 1993; NLWRA 2001; DECC 2006). Coastal and sub-coastal lowland rainforests, eucalypt and melaleuca communities, and heathlands are important to flying-foxes as foraging and roosting habitat (Eby 1995; Hall and Richards 2000; Roberts 2005). However, it would be expected that losses of these habitats would disadvantage both flying-fox species. There is evidence to suggest that P. alecto is not only expanding into urban areas but also persisting and increasing in abundance within those areas (Eby and Palmer 1991; Eby et al. 1999; Eby 2004; Roberts 2005). However, it is also likely that the population density of P. poliocephalus is also increasing in some urban areas (Hall 2002; van der Ree et al. 2006). Therefore, current evidence does not suggest a mechanism whereby habitat change could have caused the combined range expansion of P. alecto and relative range-stability in P. poliocephalus. There may be various alternative (but unpublished) hypotheses to climate change and habitat loss that could explain range shifts in Australian flying-foxes (see section 2.1 Introduction). However, only a select few of these might explain the southward range shift of P. alecto, but the range-stability of the ecologically similar P. poliocephalus. Alternative, but speculative and untested, explanations that might explain why it is mainly P. alecto that has been affected are evaluated below. First, one could speculate that P. alecto may be a recent colonist to Australia and hence may still be expanding its range. However, this hypothesis alone does not explain why the rate of expansion has increased in the last few decades. Second, flying-foxes have been shot to protect commercial fruit crops, both legally (under permit issued by state wildlife management agencies) and illegally (Hall and Richards 2000; DECCW 2009). The exact level of shooting across eastern Australia is unknown (DECCW 2009); but, if shooting pressure was greater in the north (regional Queensland), then it is possible that P. alecto has responded by moving southward to more secure locations. However, if this was the case it would also require the northern boundary of P. alecto to be contracting southward and there is no evidence of this occurring. Third, P. alecto may be responding to changing patterns of food availability. Several authors have suggested that recent changes in weather may be influencing the fruiting and flowering patterns of plants important to the diet of flying-foxes (Hughes 2003; DECCW 2009; Hudson et al. 2010; Keatley and Hudson 2012). Interruptions to nectar flow from preferred diet species have been documented in the range of P. poliocephalus and may drive the movement of individuals to new locations outside the species current range (P. Eby pers comm. 2012). It is possible that P. alecto may have also
30 experienced periods of food shortages across its range in north-eastern Australia, causing animals to move south in search of new feeding opportunities and then persisting in these new locations. Lastly, P. alecto is more sensitive to high temperatures than P. poliocephalus (Welbergen et al. 2007) and some may speculate that this is affecting the distribution of P. alecto. However, extreme temperatures are unlikely to be causing the southern range expansion of P. alecto as temperatures > 42oC cause mortality (Welbergen et al. 2007). Hot days are predicted to increase in intensity and frequency in south eastern Australia (Alexandra and Arblaster 2009) and do not explain why P. alecto are moving into areas likely to be experiencing an increase in the number of hot days. Despite previous repeated claims in the literature (Tidemann 1999; van der Ree et al. 2006; Tidemann et al. 2008), there is no evidence for a progressive contraction of the northern range limit of P. poliocephalus. Field surveys and satellite telemetry since 2007 also confirm that this species has continued to regularly use the far northern part of its distribution (Appendix 2). This negates suggestions that putative changes in the distribution of P. poliocephalus provide support for the impacts of climate change on Australian wildlife (Tidemann 1999; Hughes 2003). Furthermore, such a change in the northern range boundary would not necessarily be expected if high temperatures are not a limiting factor. Between 1921 and 1990, the northern range limit of P. poliocephalus appears to have temporarily shifted southward (by 5o latitude; Figure 2.2). This apparent contraction is consistent with the observation that many species show temporary spatial fluctuations in distributions at their range limits (Ayres and Lombardero 2000; Gaston 2003), or may simply be an artefact of low survey effort in this region during 19211990. The abundance of P. poliocephalus near its southern range limit in urban Melbourne has recently increased, accompanied by a change in its temporal pattern of occurrence from inconsistent to continuous, in spite of disturbance actions to relocate a large roost from the Royal Botanical Gardens (van der Ree et al. 2006). The southern limit of P. poliocephalus has previously been described as having recently extended southwards into urban Melbourne due to the influences of climate change, increased food availability within the urban landscape and loss of habitat elsewhere (Parris and Hazell 2005; Williams et al. 2006; van der Ree et al. 2006). However, that view has been contested (DECCW 2009). Our results do not support the inference of a southerly range extension for this species, as it was first recorded in the Melbourne area in 1869 and was consistently present in every following time-period (Fig. 2).
31 Nonetheless, a combination of increased food in urban Melbourne and habitat reduction elsewhere may have supported an increase in abundance. In the north, P. alecto records attenuate towards the Cape York Peninsula, even though this species range extends into New Guinea and Indonesia. In this zone of attenuation (between 1619oS), a third species, the spectacled flying-fox Pteropus conspicillatus occurs. These two species are ecologically and morphologically similar and their genetic status is unresolved (Webb and Tidemann 1995), and in this region they typically have not been recorded at the same roost sites (Churchill 2008). The apparent attenuation of Pteropus alectos northward range has increased over time, which may indicate a change in underlying distribution, or perhaps reflects the limited survey effort conducted in remote northern Australia, or alternatively a decrease in observers tendencies to identify ambiguous individuals as P. alecto rather than P. conspicillatus.
2.4.2 Interactions between the species
A further possible factor affecting species range shifts is that of competition between ecologically similar species. Expansions to the southern limit of P. alecto have increased the area in which its distribution overlaps with that of P. poliocephalus. The mean weight of P. alecto is 674 g and that of P. poliocephalus is 782 g (Churchill 2008). In their zone of sympatry the two species are very similar in reproductive biology and seasonality (Eby 1995; Markus et al. 2008; Tidemann et al. 2008), diet and feeding ecology (McWilliam 1986; Eby 1991a; Parry-Jones and Augee 1992; Eby 1995; Markus and Hall 2004), and roosting habitat (Tidemann 1999; Hall and Richards 2000). They are often found at the same roost sites (Hall and Richards 2000), although they commonly partition space within the roost (Nelson 1965; Welbergen 2005; Roberts 2005). The species appear to differ in their regularity of long distance movements, even though both are capable of rapid long-distance migrations. Distances of more than 1,000 km in a year are not uncommon for P. poliocephalus (Eby 1996; Tidemann and Nelson 2004; Chapter 3), whereas P. alecto appears to make long trips less frequently (Palmer and Woinarski 1999; Breed et al. 2010a). Further, where mixed-species roosts occur in the zone of sympatry, P. alecto tends to maintain small populations throughout the year, including times when P. poliocephalus have migrated elsewhere (P. Eby pers. obs.). These different movement
32 patterns may be driven by unknown ecological or behavioural differences, which may reduce the potential for competition. Nevertheless, given their known similarities, it is likely that these species compete for some resources (Eby et al. 1999). For example, if P. alecto is a superior competitor, it could displace P. poliocephalus from food or roost resources (Tidemann 1999) as it moves southwards. Alternatively, P. alecto may be the inferior competitor and its recent expansion may be in response to past declines in the population of P. poliocephalus (for example, those caused by hunting or habitat loss). However, there is no current evidence that P. alecto or P. poliocephalus use agonistic behaviours to compete directly for resources (DECCW 2009), but indirect forms of competition may also occur. The abundance data presented here suggest that P. poliocephalus may have declined within the area of distributional change since the arrival of P. alecto, although these results are not conclusive. Limited long-term data on the abundance of flying-foxes, and naturally high temporal variability in roost populations associated with the mobility and seasonal northsouth migrations of P. poliocephalus (Eby 1991a), make it difficult to verify long-term change. To clarify these issues requires further information, including better monitoring data at a range of sites in the geographic zone where P. alecto has expanded, and more information on the comparative ecology of the two species. 2.5 Conclusions
An important implication of this studys results is that more careful quantitative analyses are needed to support inferences concerning austral species range shifts in response to climate change. For example, poleward extensions have recently been claimed in the distributions of Australian bird species, and these have been interpreted as consistent with effects of climate change (Olsen et al. 2003; Chambers et al. 2005; Olsen 2007), yet these range extensions have not been confirmed by statistical analyses. While it is indeed possible that climateinduced latitudinal range shifts are occurring in a number of Australian species, it is important that distributional data are analysed systematically to verify such shifts before inferences are made. This studys data analyses do not support the previous claim that recent climatic patterns have driven changes in the geographical distributions of P. alecto and P. poliocephalus. However,
33 alternative processes of habitat change and interspecific interaction have also been insufficient to explain the two species observed contrasting patterns of temporal change, at least based on current knowledge. Further research is required to understand the causes of the recent rapid distributional change in P. alecto and the implications for its potential interaction with P. poliocephalus.
34
Chapter 3 Long-distance and frequent movements of the flying-fox Pteropus poliocephalus: implications for management
3.1 Introduction
Many flying vertebrates can travel relatively long distances in a few days (Berthold 2001; Kingsford and Norman 2002; Fleming and Eby 2003; Richter and Cumming 2008; Gill et al. 2009). For some species, long-distance movement provides access to widely spaced and temporally variable food resources, as well as the opportunity to use different areas for roosting and for feeding. Frequent long-distance movements create particular management and conservation challenges for the species concerned, because impacts or management actions in one place may have outcomes elsewhere in the species range. The fruit-bat or flying-fox family (Pteropodidae) contains many highly mobile species capable of strong flight. Flying-foxes feed at night on nectar and/or fruit, and roost during the day, often in large aggregations (Pierson and Rainey 1992). Their patterns of movement and roost usage are, in most cases, poorly understood (Pierson and Rainey 1992; Mickleburgh et al. 2002), even though a number of flying-fox species are the subject of conservation or management programs (Eby and Lunney 2002a; Mickleburgh et al. 2002). Radio-tracking and limited satellite telemetry studies have revealed complex patterns of movement of flyingfoxes between roosts, including long-distance movements in some species (Eby 1991; Eby 1996; Tidemann and Nelson 2004; Richter and Cumming 2008; Epstein et al. 2009; Breed et al. 2010a). The grey-headed flying-fox Pteropus poliocephalus is a large bat endemic to eastern Australia (average weight 782 g) (Churchill 2008). Radio-tracking and satellite telemetry studies of this species have documented a range of movement patterns from long-distance nomadic movements to short ranging behaviours (Eby 1991a; Eby 1996; Fleming and Eby 2003; Tidemann and Nelson 2004). Although predominantly feeding on native blossom and fruits, P. poliocephalus also raids commercial fruit crops (Ratcliffe 1931; Tidemann et al. 1997). The species has been implicated as a vector of several emerging zoonotic viruses of livestock and humans e.g. Hendra virus and Menangle virus (Fraser et al. 1996; Halpin et al. 2000; Philbey et al. 2008). It often roosts in urban settings, causing residents to object to its noise
Chapter 3 Long-distance movements of P. poliocephalus
35 and smell, and become anxious about perceived risks of disease (Hall and Richards 2000; Hall 2002). For decades, various social or economic interest groups have undertaken or vigorously advocated for management actions aimed at reducing local populations of P. poliocephalus, either through direct culling or by inducing animals to move elsewhere, through dispersal or relocation of roosting colonies (Ratcliffe 1931; Waples 2002; Hyne 2010; Chapter 5). These practices are based on the assumption that individuals frequent a particular locality or region, and can be controlled by actions undertaken at that scale. Such practices are unlikely to reduce local numbers if the species is highly mobile, unless the entire population is dramatically reduced. Such a reduction would not only require substantial effort, but would also run contrary to international and national conservation priorities (Mickleburgh et al. 1992; Duncan et al. 1999; Dickman and Fleming 2002). Effective management of flying-foxes requires a good understanding of the frequency and extent of movements among roost sites. This study quantifies the patterns of movement by P. poliocephalus among roost sites at a range of temporal and spatial scales. Data collected from 14 satellite-tracked individuals over a period of nine months and an area of 137,600 km2 were used to address the following questions: (1) What patterns of movement (in terms of distances and directions) are exhibited by P. poliocephalus at time scales of days, weeks, and months? (2) How much do these patterns vary among individuals? (3) What is the relationship between individuals movement patterns and their use of different roost sites? (4) What are the implications of these findings for management?
3.2
Methods
3.2.1 Study species
Pteropus poliocephalus is endemic to coastal eastern Australia, with a distribution extending from central Queensland (21oS) to Melbourne, Victoria (38oS) (Chapter 2). The movement patterns of P. poliocephalus vary widely and include sedentary behaviours associated with areas of high floristic diversity (Eby 1991; Eby 1996; Fleming and Eby 2003), regular
36 migrations that occur between seasonal, annually reliable nectar patches (Ratcliffe 1931; Eby 1991a; Eby 1996), nomadic movements between eruptive, rich flowering events (Eby 1996; Tidemann and Nelson 2004), and combinations of these. The degree to which any particular part of the population is sedentary, migratory or nomadic is unquantified, but only a very small proportion is believed to be sedentary. This belief is based on: (1) only a small number of roosts are continuously occupied by P. poliocephalus throughout their range; and (2) within these roosts only a small number of animals persist at times when the roost population is low. Little is known about the frequency of movement between roost sites, although telemetry studies have shown movements of individuals between roosts on weekly and fortnightly scales (Eby 1996; Tidemann and Nelson 2004). Daytime roost sites of this species often comprise thousands of individuals and occur in rainforest, riparian, mangrove or wetland forests (Ratcliffe 1931; Eby 1991a; Tidemann 1991). Some roosts are known to have been used for decades (Ratcliffe 1931; Tidemann 1991). The numbers of bats within a roost can fluctuate widely over time, associated primarily with variation in the availability of flower and fruit resources (Eby 1991a; Parry-Jones and Augee 1992; Eby and Lunney 2002c). Large aggregations are associated with flower pulses from nectar-rich species, primarily eucalypts (Eby 1991a; Parry-Jones and Augee 1992; Eby 1996), and are believed to support high numbers of migratory or nomadic animals. At night, P. poliocephalus typically feed on blossoms and fleshy fruits of trees within 20 km of their roosts, although distances between roosts and feeding sites may be as much as 50 km; the species utilises remnant forest, patches of vegetation on cleared land and urbanised areas (Eby 1991a; Tidemann 1999).
3.2.2 Study area and capture sites
Study animals were captured in October 2007 and June 2008 at four roost sites in south-east Queensland (SEQ): Canungra (28.0oS, 153.2oE), Stafford (27.4oS, 153.0oE), Fraser Island (25.4oS, 153.1oE) and North Stradbroke Island (Dunwich) (27.5oS, 153.4oE). Both the Canungra and Stafford roosts are occupied most months of the year and have a long history of use by P. poliocephalus, including use as maternity colonies, although numbers fluctuate within and between years (Birt 2005, and data held by government agencies, P. Eby and the author). At the time of trapping (October 2007), both roosts had approximately 10,000 bats present. Fraser and North Stradbroke Islands are irregularly used roosts, typically occupied in the autumn and winter months of certain years in response to prolific flowering of nectar-rich
37 trees. At the time of trapping (June 2008), both these roosts contained >50,000 P. poliocephalus. The study attempted to target the part of the population that was migratory or nomadic by capturing flying-foxes in: (1) roost sites used most of the year, but which had a large number of animals present at the time of capture in 2007, and therefore a high proportion of mobile individuals (Canungra and Safford, N = 4 individuals), and (2) seasonally used roost sites with large numbers of animals in 2008, where all animals were expected to be migratory or nomadic (Fraser Island and Dunwich, N = 10 individuals).
3.2.3 Satellite telemetry
Fourteen adult male P. poliocephalus were fitted with satellite transmitters and released at the capture site: in October 2007, two bats from Canungra and two from Stafford, and in June 2008, five bats from Fraser Island and five from North Stradbroke Island. Flying-foxes were captured using mist-nets (mesh size 31 mm) suspended between 11 m tall poles, set in the early morning and late evening in the flight path of animals leaving or entering the roost site. Inhalation anaesthesia (Isoflurane, Laser Animal Health Pty Limited) was used to sedate animals for attachment of the transmitter and collar, following the protocol described by Jonsson et al. (2004). Males were selected to minimise the impacts on bats of carrying the transmitters: males are larger than females and are not subject to stresses of pregnancy and lactation. Sixty-three male P. poliocephalus were caught and all were aged, weighed and assessed for general body condition. Only large reproductively mature males (weight 615845 g, forearm length 155171 mm) were fitted with solar-powered satellite transmitters (12 g PTT-100, Microwave Telemetry, Columbia, USA; (see also Smith et al. 2011). Total package weighed approximately 22 g (<3.6% of the body mass of the smallest individual); transmitters were fitted around the neck using a soft leather collar lined with 3 mm neoprene; oriented crosswise with the solar array facing upward in roosting animals to reduce recharge time, improve power, and enhance accuracy (see Appendix 5). Transmitters were programmed to a duty cycle of 10 hours on, alternating with 48 hours off, which provided sufficient off time to recharge the battery and provided regular location data including positional fixes during both day and night. The minimum time between two consecutive roost fixes was therefore 48 hours (with one recorded exception). The theoretical maximum number of fixes per 10-hour transmission cycle in the study area was 14, with accuracy and success dependent on the level of battery power. This duty cycle enables analysis of movements both between day roosts and
38 during nocturnal feeding, although only the former are considered here. Data were obtained via the Argos Global Data Collection and Location Service System (Collecte Localisation Satellites, France). Each data record (fix) included the latitude and longitude, presumed accuracy of the location (in seven classes), UTC time and date, and battery voltage.
3.2.4 Data manipulation and analysis
The raw transmission data from all 14 tagged flying-foxes comprised 4,356 location records, of varying presumed accuracy (see below). Records for roosting sites were extracted using fixes that were either: (1) obtained during daylight hours (between sunrise and sunset, data from the Geoscience Australia website adjusted for geographical location); or (2) cases where bats had returned to known roosts before sunrise. From the resulting 2,616 roost records, I removed the 14% which fell within Argos accuracy classes A, B and Z, with no upper limit to their error margins, yielding 2,254 fixes for further analysis. An average of 4.5 fixes were recorded per 10 hour on cycle (range 111, SD 1.12.4). These were mapped in ArcView 3.3, and manually scrutinised by overlaying geographical landmarks and known roost locations (Eby 2007), and data held by ecologists (P. Eby and B. Roberts) and management agencies. When large parts of the on duty cycle fell during daylight hours, roost locations were visible as clusters of fixes, generally within about a 2 km radius. These were reduced to one roost fix per bat per day as follows. Fixes that were within 2 km of a known roost location (N = 48) were adjusted to match its precise location. Away from known roost sites, potential unknown roosts (N = 29) were recorded if there were at least two daytime records less than 2 km apart, and the likely exact roost site was then sought by examination of aerial photographs, field inspection or discussions with local naturalists, wildlife carers and biologists. When such roosts could not be ground-verified, the fix with the best accuracy rating was used; three unverified roosts were clearly visited by two or more tagged bats, whereas the other 17 involved only one tagged individual. In nine of these cases the fix-clusters suggested that a single bat was using a temporary day roost away from established colonial roosts. All functionally impossible fixes were also removed by this process (163 records, i.e. 7%, including those in the ocean far from the coast or vast areas of treeless land). The resulting working data set comprised 2,091 roost records with Argos presumed accuracy other than A, B and Z (<1.5-5 km), 1,045 of which had a presumed accuracy class of 3 or 2 (<0.5 km). Each roost site was given an identifying name.
39 The number and duration of visits by each individual to each roost (counting revisits) were calculated from data on roost visits, defined by the dates of first and last consecutive fixes from a particular individual recorded at a particular roost site. The visit duration was the elapsed number of days at that roost (including some durations of one day only). During the 48 hour off duty cycle, there was no transmission, which caused regular data gaps of 23 days, and hence the roost visit durations included these gaps. In 20 cases, there were transmission gaps of >5 days within sequences of consecutive records at the same roost; these were considered unreliable data and were removed prior to analysis. Revisits to previously used roosts were counted separately; these occurred if an individual had meanwhile been recorded at other roost sites. I documented 161 separate roost visits by the 14 flying-foxes, from which the frequency distribution of roost visit durations was obtained. All roost sites used by more than one animal were further examined to see if there was synchrony in their arrival and departure dates. I also calculated the maximum step sizes (distances between consecutive useable fixes at different roosts) across each individuals full tracking duration, and the frequency distribution of step sizes pooled across all individuals but omitting 22 cases in which consecutive roosts were separated by transmission gaps of >5 days. To provide a time-standardised estimate of short-term inter-roost movements, I additionally calculated minimum, medium and maximum step sizes for each individual, and the step size frequency distribution across all individuals, using only those cases where consecutive different recorded roosts were two days apart (the minimum time provided by the transmitter duty cycle). I used sampling units of timed movements to analyse distances travelled and numbers of roost sites used during fixed time-periods of approximately 1, 5, 10, 20 and 30 weeks. To calculate movements for approximate 1-week periods, each individuals roost records were given a repeating number sequence between 1 and 7 to flag the day of each successive week between the dates of capture and last usable position fix. Particular weeks were selected for analyses if there was a valid roost record for day 1 together with either day 6, 7 or 8 plus at least one record in between. On 8% of all occasions there was no record for days 6, 7 or 8, and the next closest record was used (day 5 or 9). There were no records between the first and last day in 5% of these cases that were removed from the analysis, providing a sample of 333 weekly records, of which 87% had at least two records in between first and last days. To calculate movements for five weeks, records were collapsed into as many periods of five consecutive weeks as possible using the criterion that each consecutive week had sufficient
40 data (as defined above). Weeks without sufficient data were omitted, creating gaps between usable sequences. This process was repeated independently for time-periods of 10, 20 and 30 weeks. Three measurements were calculated for each timed movement period: net displacement (the shortest path between an individuals first and final roost sites in a time-period); cumulative displacement moved (the sum of all recorded distances between consecutive roost sites); and the number of different roosts used. The latter two measurements are minimum estimates due to possible data gaps during the transmitter off period or when accurate fixes were not obtained. I analysed the frequency distributions of the net and cumulative displacements, and the latters relationship with the number of roost sites used, across different time-periods. I calculated each individuals maximum displacement, first as the maximum distance between any pairwise combination from all roosts visited during any 10-week period recorded for that individual, and second across each individuals full tracking duration. The maximum step size (as defined previously) within a 10-week time span was further calculated using all 10-week time-periods available for each individual. Information on the regional flowering patterns of tree species that are important in the diet of P. poliocephalus (Eby 1996; Eby 2008) between October 2007 and April 2009 was obtained by maintaining regular contact with apiarists, as well as through direct observations by the author and discussions with naturalists and ecologists. 3.3 Results
3.3.1 Long-term and long-distance ranging behaviours
The 14 male P. poliocephalus were satellite-tracked for a range of 13277 days (average 175 days) during which roost location data were received for 8154 days per individual (average 78 days; Appendix 6). The total number of roost fixes received with an Argos presumed accuracy of <1.5 km ranged from 9324 per individual (average 149: Appendix 6); half the fixes had a presumed accuracy <0.5 km (averaging four useable fixes every two days; range 06 fixes). The recorded roost locations of all individuals spanned a north-south distance of 1,075 km and an east-west distance of 128 km (Figure 3.1). Twelve of the 14 tagged individuals had moved from the point of capture within three weeks (Figure 3.2). Distances and directions
41 travelled varied greatly, within and between regions (Figures 3.2 and 3.3). Five individuals travelled overall cumulative displacement distances >1,000 km (Appendix 6). The three individuals with the longest tracking durations (254277 days) had maximum overall displacement distances of 93, 547, and 855 km, whereas their respective net displacements from first to last roost fix were 33, 312, and 378 km, and the corresponding cumulative displacements were 489, 1652, and 1562 km (Figure 3.2, Appendix 6). Four individuals returned to their points of capture at some time during the study (Figure 3.2, Appendix 6). Two individuals moved a step size during two days (the minimum time between consecutive day roost fixes) as large as 166 km, but if night feeding records were also considered, five individuals had 2-day step sizes of >200 km, including one of 500 km (Appendix 6). One animal moved 130 km within 4.8 hours (calculated using night-time fixes), which is equivalent to an average flight speed of 27 km/hr.
42
Figure 3.1 Roosts used by 14 satellite-tracked Pteropus poliocephalus between October 2007 and April 2009. A. Location of the 77 roosts used during the study. B. Map of roosts showing the number of tracked individuals known to have used each roost, excluding the capture date (open circles = roosts used by one individual only). Triangles represent the roosts where individuals were captured.
43
Figure 3.2 North-south movement patterns of 13 satellite-tracked Pteropus poliocephalus from their points of capture. Only individuals with 12 or more weeks of data are presented. Legend shows the location and year of capture. Long-distance movements (net displacement >300 km per month) are highlighted in bold. East-west movements are not shown.
44
Figure 3.3 Weekly movement patterns of 3 satellite-tracked Pteropus poliocephalus from their points of capture. Triangles show the points of capture. Map shows the individuals identification numbers (see Appendix 6), their roost locations at each weeks start and the numbers of weeks spent at each roost (with number of separate visits if >1 in brackets); squares represents end points. For clarity, other roosts visited within-weeks are not shown. Summaries at bottom show number of transmission weeks, cumulative displacement and maximum displacement.
Individuals that achieved larger maximum displacements did so by making occasional rapid long-distance movements rather than through progressive sequences of smaller steps (Figures 3.2 and 3.3, Appendix 6). Overall, the flying-foxes tended to remain at the north of their range (2428o S) during winter (June-August), and moved to higher latitudes after September (Figure 3.2). Five individuals undertook a total of nine long-distance movements (net displacement >300 km during one month; Figure 3.2), three travelled northward (from
45 JanuaryMarch and in July), and five travelled southward (from SeptemberDecember and in March). Eight of the nine long-distance movements were to areas where there were reports of mass flowering of known food trees (including Eucalyptus tereticornis, E. siderophloia, E. pilularis, Corymbia citriodora, C. intermedia, Melaleuca quinquenervia: all Myrtaceae).
3.3.2 Quantitative patterns of movement and roost site use
A total of 77 roost sites were recorded (Figure 3.1). Of these, 48 (62%) had been documented prior to the present study. The remainder were mostly in remote locations to the north and west of the study area, and were used during spring. These could be previously undocumented roosts or may represent individuals or small transient groups near feeding areas. Nineteen roosts were used by two or more tagged individuals, and there were 25 occasions when two tagged individuals coincided in their roost location, although generally without synchrony of arrival or departure. Revisits to roost sites were common but variable, with one individual revisiting the same roost on eight different occasions during an elapsed time of 35 weeks. The average time between successive revisits across all individuals was 22 days (SD = 22, range 490, N = 61 pairs of visits). The recorded visit duration at any particular roost site was typically short, with 64% being for <5 days; however, the frequency distribution was skewed (Figure 3.4a). About 26% of roost visits involved >10 consecutive days (Figure 3.4a), but these visits accounted for 78% of the total 1,480 roost days. One flying-fox remained at its point of capture for 133 consecutive days. Extended periods at one roost were often followed by times of frequent movement among roost sites which were either in close proximity or involved progressive stops during longer north-south movements. Between consecutive recorded roosts, the flying-foxes travelled modal step sizes of 2150 km in elapsed times of 25 days; the mean step size in 25 days was 61 km (range 3576 km) and the mean 2-day step size was 45 km (range 3166 km) (Figure 3.4b,c). Within one-week periods, both the net and cumulative displacements were mostly of zero km (Table 3.1, Table 3.2, Figure 3.5), largely due to occasional long stays by individuals at particular roosts (see above). In 6% of the 214 cases of zero net weekly movement, the cumulative displacement exceeded the net displacement indicating that these individuals had made side-trips to other roosts during the week; there may have been some additional undetected side-trips during the 48-hr transmitter off period. All individuals showed broadly
46 similar distributions of weekly net displacement; most displacements were zero but some were >50 km and 10 of 13 flying-foxes tracked for >12 weeks had some weekly movements >100 km (Appendix 7).
47
Figure 3.4 Frequency (%) distributions of the duration of roost visits and the distances between consecutively-recorded roosts. Data from 14 satellite-tracked Pteropus poliocephalus. A: Roost visit durations (the elapsed number of consecutive days at a roost, with no data gaps >5 days). Revisits were counted separately. N=161 visits . Median roost visit duration = 2 days; mean = 9 days. B: Overall step sizes (distances between consecutive useable fixes at different roosts) across each individuals full tracking duration, after removal of elapsed durations of >5 days; N=148 steps. Median = 39 km; mean = 61 km. C: 2-day step sizes across standard 2-day time spans (the minimum possible under the transmitter duty cycle). N=64 steps. Median = 32 km; mean = 45
48 The maximum value for net displacement did not vary greatly among the five time-periods (range 576776 km), however, the medians, means and modes of both net and cumulative displacement increased greatly from one to 10 weeks whereas from 10 to 30 weeks this rate of increase was maintained only in cumulative displacement, whilst the net displacement increased more slowly (Table 3.1, Table 3.2, Figure 3.5). For example, at one week, the median and mode were all zero for both net and cumulative displacement, compared with median net and cumulative displacements of 147 km and 260 km respectively at 10 weeks, increasing to 158 km and 821 km at 30 weeks (Table 3.1, Table 3.2). This pattern suggests that a maximum latitudinal seasonal displacement had generally been reached within less than 30 weeks, although there were continuing regional and local movements at a scale of tens of km within a few weeks by all flying-foxes (see also Figure 3.2). Among the 13 individuals with >12 weeks of data, the range of maximum displacements was 35679 km at 10 weeks and 63855 km over the whole study (Appendix 6). A lack of zero net distances in histograms for longer time-periods (Figure 3.5) indicates low medium-term fidelity to individual roost sites over periods of several months.
Table 3.1 Net displacement moved by flying-foxes among roosts in time-periods of 130 weeks duration.
Time-period (weeks) 1 5 10 20 30
N 333 64 31 13 5
N = 0 km 214 14 3 0 0
Median km 0 48 147 90 158
Mean km (SD) 29 (68) 93 (123) 172 (173) 208 (259) 259 (206)
Range km 0576 0626 0662 27776 94576
Net displacement is the distance between first and last roosts during each time-period. N: number of timed movements in each time-period (see Methods 3.2) SD: standard deviation
49
Table 3.2 Cumulative displacement moved by flying-foxes among roosts in time-periods of 130 weeks duration.
Time-period (weeks) 1 5 10 20 30
N 333 64 31 13 5
N = 0 km 201 9 2 0 0
Median km 0 111 260 618 821
Mean km (SD) 33 (70) 161 (166) 316 (236) 584 (381) 798 (217)
Range km 0576 0690 0821 1201121 4891059
Cumulative displacement is the sum of all sequential distances between consecutive recorded roost sites during each time-period. N: number of timed movements in each time-period (see text) SD: standard deviation
50
Figure 3.5 Frequency distribution of net and cumulative displacements by flying-foxes among roosts in different time-periods. A: 1 week, B: 5 weeks, C: 10 weeks, D: 20 weeks, E: 30 weeks. Net displacement is the distance separating roost sites used by an individual at the start and at the end of a time-period. Cumulative displacement is the sum of all recorded distances between consecutive roost sites for each time-period. Hatched areas show the number of zero distances in a time-period. Sample sizes as for Table 3.1 and Table 3.2.
For cases where flying-foxes shifted roosts at least once during a time-period, the number of different recorded roost sites increased logarithmically with the cumulative displacement within each time-period up to 20 weeks; for 30 weeks where N=5 there was no significant relationship (Figure 3.6). The logarithmic relationships confirm that, while individuals used more roosts as they progressively travelled further, the longest travel distances were achieved
51 by increasing the step size between roosts. Between time-periods the relationship was linear: as elapsed time increased, flying-foxes increased the number of different roosts visited in proportion to the distance travelled, adding about one additional roost for each accumulated 100 km travelled, which on average occurred every 34 weeks although there was very large variation (Figure 3.6, Appendix 6).
Figure 3.6 Number of roosts used by flying-foxes in relation to cumulative displacement during different time-periods. A: 1 week (N = 132), B: 5 weeks (N = 55), C: 10 weeks (N = 29), D: 20 weeks (N = 13), E: 30 weeks (N = 5). X shows the median of each datase t. F graphs the median values of each dataset. Cases where flying-foxes did not shift roosts during a time-period were excluded. Fitted lines show significant regressions ( P<0.05), with either log or linear selected as the best fit based on Pearsons r values. Note that the results in A were unaffected by removal of the isolated point at displacement of 600 km. .
52 3.4 Discussion
3.4.1 Patterns of movement and roost site use
This study has tracked the broad-scale movement patterns of P. poliocephalus at a finer temporal resolution than has been previously possible, due to improvements in lightweight satellite telemetry. Our findings agree with previous work that Australian flying-foxes are highly mobile, capable of frequent movement among roost sites in a regional area as well as long-range movements (Eby 1991a; Spencer et al. 1991; Eby 1996; Tidemann and Nelson 2004). The statistical quantification of the species movement patterns shows that (i) individuals move among roost sites far more frequently than previously assumed; (ii) individuals are capable of rapid sustained long-distance flight and (iii) movements vary considerably between individuals and within individuals over time. These results are consistent with previous findings that P. poliocephalus exists as a single, genetically panmictic population across its range (Spencer et al. 1991; Webb and Tidemann 1996; Chan 2007). Previously suggested characterisations of movements by P. poliocephalus include annual north-south migration, annual return to the place of birth, and dispersal into small groups during winter (Ratcliffe 1931; Nelson 1965; Tidemann 1999; Tidemann and Nelson 2004). Our results support the proposal that P. poliocephalus exhibits a range of movement patterns, from residency to regional and long-distance nomadism, combined with facultative latitudinal migration (Eby 1996; Fleming and Eby 2003). Our data revealed some seasonally correlated long distance movements of several hundred kilometres, consistent with latitudinal migration, although only for a subset of individuals. Like their avian counterparts, Australias flyingfoxes display movement patterns that respond to the largely irregular and ephemeral resource shifts characteristic of the continent (Eby 1991a; Chan 2001; Breed et al. 2010a). Indeed, it is becoming increasingly apparent that patterns of large-scale movement in many animal species, particularly in the southern hemisphere, are more variable and complex than the regular north-south movements considered typical among long-distance migrants of the northern hemisphere (Berthold 2001; Chan 2001; Kingsford and Norman 2002). Our observation that most long-distance movements in this study were to locations where there was intensive flowering of known food species is consistent with the findings of previous studies (Eby 1991a; Eby 1996; Eby and Lunney 2002c). The high variability
53 between individuals and overall complexity of movement patterns evident in our data are consistent with the view that changes in the distribution of feeding opportunities primarily drives migration in this species (Eby 1991a; Eby 1996; Eby and Lunney 2002c). The plant species associated with long-distance movements in this study flower seasonally, however, these flowering patterns are complicated by very high inter-annual variability and spatial heterogeneity (Eby and Lunney 2002c; Eby and Law 2008). Local and regional movements among roosts by P. poliocephalus (at scales of tens of kilometres) would arguably be most likely to occur in situations where a diversity of forest types allows year-round access to food resources within a single region (McWilliam 1986; Eby 1991a; Eby 1996; Markus 2001). The results of this study support this proposition. Subtropical eastern Australia at about 25o30oS contains a high diversity of vegetation types, including high abundances of many different species of nectar-rich trees and shrubs, which flower in all months of the year (Keith 2004; Birtchnell and Gibson 2008). This is also the region where the highest population densities of P. poliocephalus occur (Eby and Lunney 2002c) and where the greatest concentration of roosts and local movements were observed in this study. There was regular interchange of individual P. poliocephalus among roost sites within a region, with individuals often remaining less than five days at a roost. Even irregularly used roosts remained occupied for much longer periods of time than they were used by individual animals. Short-duration stays may be a response to local shifts in food availability, or may assist individuals to obtain information about food resources, or improve social (e.g. mating) opportunities. Additionally, some roosts were clearly stopovers for animals undertaking longer movements. However, extended stays of months at a single roost were also observed. Individual flying-foxes have been reported roosting at a single site for periods up to seven months, in response to a reliable food source nearby (Nelson 1965; Eby 1991a; Eby 1996). In general, proximal cues for flying-fox movement and the causes of variation in their movement patterns remain poorly understood.
3.4.2 Comparison with other flying-foxes
Detailed information on the movements of the three other Australian flying-foxes are limited, although similar seasonal movements that track the changing availability of food resources and a range of other movements, including long-distance flights, have been recorded
54 (Tidemann et al. 1999; Sinclair et al. 1996). For example, there are reports of net movements of 486 km in 90 weeks for P. scapulatus (Sinclair et al. 1996) and 249 km in 49 weeks for P. alecto (Breed et al. 2010a). The broad-scale movement patterns of P. conspicillatus are unknown, although individuals of this species are known to fly up to 50 km in a single night to feed, hence the species is likely to have a capacity for longer-distance movements (Fox 2011). Pteropus scapulatus and P. alecto differ in their regularity of movements: P. scapulatus displays a primarily nomadic lifestyle, frequently moving long-distances between patchily distributed, ephemeral nectar resources (Palmer and Woinarksi 1999; Hall and Richards 2000; Birt 2005); whereas P. alecto appears to make long trips less often and some individuals can adopt a primarily sedentary lifestyle in areas of high floristic diversity (Markus 2001; Breed et al. 2010a). Most flying-foxes within the Asia-Pacific region have not been well studied or systematically monitored, however species from large land continents are believed to be wide-ranging, using multiple roosts across their range (Mickleburgh et al. 1992; Pierson and Rainey 1992; Epstein et al. 2009). In Asia, P. vampyrus has been reported moving 363 km in four days (Epstein et al. 2009), while Eidolon helvum in Africa has been reported moving 370 km in one night (Richter and Cumming 2008). Residency in these species and the frequency of movement between roosts has not been investigated. Flying-foxes within the Asia-Pacific region are often viewed as pests of commercial fruit crops or nuisances to the tourist trade (e.g. P. giganteus, P. hypomelanus, P. tonganus) (Mickleburgh et al. 1992; Wilson and Graham 1992; Singaravelan et al. 2009). Throughout this region, destructive management methods aimed at individuals or their habitat are used to manage conflicts (e.g. culling at orchards and roosts and/ or total habitat removal) (Mickleburgh et al. 1992; Singaravelan et al. 2009). In these and other widely-ranging species, mitigating conflict is extremely complex, but an understanding of species movements is vital to sound management.
3.4.3 Caveats to this study
All the data obtained in this study were from adult male P. poliocephalus; potential differences between sexes may warrant examination. In previous studies of P. poliocephalus, females have displayed the same general movement patterns as males (Eby 1991a; Eby 1996). However, lactating females of the ecologically similar black flying-fox (P. alecto) have been reported to travel greater distances between roosts and foraging sites than males (Palmer and Woinarski 1999).
55 Another caveat is that error estimates of location fixes in satellite-tracked flying-foxes can be considerably larger than those indicated by Argos (typically around 2.5 km for location class 0, and mainly up to 2 km for classes 13) (McKeown and Westcott 2011). I addressed this issue in three ways: first, by using a collar design which enhanced signal strength and therefore accuracy (see Methods 3.2.3 and Appendix 5); second, by using the clusters of within-day repeat fixes to identify roost locations; and third, by cross-referencing fixes with independent information on the locations of known roosts. In any case, a somewhat higher level of location error would not affect our substantive conclusions about movement patterns at the spatial scales studied here. Furthermore, our conclusion that individuals move frequently between roosts is conservative, as some roost shifts may have remained undetected during the 48-hour transmitter off periods.
3.4.4 Management implications
Pteropus poliocephalus and other flying-fox species inhabiting Australia (P. alecto, P. conspicillatus, P. scapulatus) are the target of community concerns in relation to public health, amenity and impacts on agriculture. Since European settlement of Australia, people have taken or advocated various actions aimed at reducing flying-fox numbers or moving problem roost sites (Ratcliffe 1931; Lunney and Moon 1997; Eby and Lunney 2002b). Until 1986 in New South Wales and 1992 in Queensland, all flying-fox species were listed in government statutes as agricultural pests in need of control (Thiriet 2010). Public calls for culling have re-emerged in recent years following the identification of a new and highly virulent zoonotic disease, Hendra virus, which may be transmitted to horses by flying-foxes (Halpin et al. 2000; Hyne 2010). However, since the 1990s, the conservation significance of flying-foxes in Australia has been increasingly recognised. Pteropus poliocephalus is listed as Vulnerable by national and state governments and under the international IUCN Red List (Eby and Lunney 2002a; Lunney et al. 2008). Protection and listing have occurred in response to large declines in abundance resulting from the combined effects of habitat destruction, agricultural control efforts (e.g. killing of individuals at orchards), competition with P. alecto and disturbance at roosts, and have been spurred by recognition of the flying-foxes important ecosystem functions as pollinators and seed dispersers of many tree species (Duncan et al. 1999; Dickman and Fleming 2002).
56 Despite such legislative protection, the use of lethal control measures to protect fruit crops and the harassment of flying-foxes at day roosts have continued, both legally (under permit issued by state wildlife management agencies) and illegally. These attempts at relocation and local control have ignored the capacity of flying-foxes to conduct frequent long-distance movements. This study has provided new data which support previous suggestions that culling cannot be a practical tool for reducing the impacts of flying-foxes, whether these comprise crop damage or disease risk (Eby and Lunney 2002a; Breed et al. 2010c). Culling of individuals can achieve only short-term population reduction at a local scale, since flyingfoxes can easily and rapidly re-occupy culled areas, and are especially likely to do so if there is nearby flowering of nectar-rich trees. Indeed, during more than a century of culling flyingfoxes in Australia and a longer-term reduction in their overall population size, the incidence of damage to fruit crops appeared unchanged (Ratcliffe 1931; Tidemann et al. 1997; Rogers 2002). High mobility means that the effects of local culling will be diluted across the entire population of the species. For culling to be effective, it would need to be mounted on a sufficiently large scale to reduce the species entire population size. Such an approach would be contrary to the conservation priorities discussed earlier. Public demands for roost relocations have also increased over the past two decades, concurrently with rapid increases in urbanisation in the coastal subtropics (Hyne 2010). The results of this study further suggest that roost relocation is unlikely to have the desired goal of permanently shifting a local flying-fox population from an area. As in the case of culling actions, even if relocation activities do induce flying-foxes to move away in the short term, empty roosts are likely to be reoccupied during subsequent months or years or, alternatively, new roost sites may be established nearby. Indeed, an analysis of 10 case studies concluded that actions to prevent roosts from being re-established typically need to be repeated many times within and between years (Chapter 5). For example, 12 years of repeated noise disturbances did not prevent flying-foxes from regularly returning in large numbers to a roost in one rural township, while there was a concurrent establishment of additional new urban roost sites in the area (West 2002; Chapter 5). 3.5 Conclusion
Quantitative analysis of the movement patterns of P. poliocephalus leads to the conclusion that control actions (i.e. killing of individuals at feeding sites and roosts; attempts to destroy or relocate roosts) are likely to have only temporary effects on local populations. These data
57 explain why both culling and relocation attempts have had limited success in resolving conflicts between humans and flying-foxes in Australia, despite both approaches having been utilised since European settlement. Developing alternative methods to manage these conflicts in a manner informed by a better understanding of species movement and roost usage patterns remains an important challenge. The data obtained in this study have improved our understanding of the frequency and complexity of movements of P. poliocephalus, and enabled clear conclusions to be drawn regarding the efficacy of current management practices. However, larger samples of individuals, preferably varying in age, sex and reproductive status, are required to provide a robust understanding of movement patterns. Further technological developments which enable lighter weight and lower cost satellite telemetry should facilitate research which extends the quantitative foundations provided by the present study to more comprehensive models of flying-fox movement patterns and their drivers. In the Asia-Pacific, the inadequate understanding of key aspects of flying-fox ecology carries a high risk that poor management decisions will be made, as has been the case for even for the relatively well-studied P. poliocephalus. However, it is reasonable to assume that for any winged vertebrate species that is shown to be a wide-ranging user of multiple colonial roosts, the same management message will apply. That is, that both culling animals at particular individual roost sites, and attempting to relocate these roosts through disturbance actions, are likely to have limited long-term success in permanently removing them from the region.
58
Chapter 4 Feeding movements and landscape use by P. poliocephalus

4.1 Introduction
Understanding how species utilise landscapes is an important part of understanding the ecology of a species; this basic ecological information allows managers to identify areas and habitats that are particularly important for conservation of a species (Price et al. 1999; Berthold 2001; Fleming and Eby 2003). For highly mobile, colonially roosting species of flying vertebrates, identifying areas important for conservation is often complicated, because the habitats used for roosting and feeding are often spatially separated and may be chosen using different criteria (Berthold 2001; Kingsford and Norman 2002; Fleming and Eby 2003). Such species may also utilise distinct areas separated by hundreds or thousands of kilometres in the course of seasonal or annual cycles, as their feeding and roosting locations track the distribution of resources in the landscape (e.g. Roshier et al. 2001; Kingsford and Norman 2002). Highly mobile species can often tolerate some loss or alteration of habitat across their range because they are able to traverse areas providing few resources to reach feeding or roosting habitat (Law and Dickman 1998; Catterall et al. 1998; Fahrig 2003). Australian flying-foxes (Pteropodidae) provide a good example of these issues. They have complex movement patterns because they feed on irregular and ephemeral resources (i.e. fruit and nectar) and form large communal day roosts often located many kilometres from feeding areas (Eby 1996; Markus and Hall 2004; Birt 2005). However, the specific criteria that flyingfoxes use to select feeding and roosting sites and the factors predicting the distance animals will travel to feed are not well understood (Eby 1996; Kunz and Lumsden 2003; McDonaldMadden et al. 2005). For wide-ranging species, the landscape-scale habitat characteristics play an important role in their movement and site selection decisions; however, this also has not been an area of investigation for the highly mobile flying-fox. Previous research into Australian flying-fox movements has utilised direct observations (McWilliam 1986; McDonald-Madden et al. 2005), banding techniques (Tidemann 1999), radio-telemetry (McWilliams 1986; Spencer et al. 1991; Eby 1996; Palmer and Woinarski 1999; Palmer et al. 2000; Markus and Hall 2004; Birt 2005) and, to a limited extent, satellite telemetry (Tidemann and Nelson 2004; Epstein et al. 2009; Breed et al. 2010a) to understand
59 the feeding behaviours of Australian flying-foxes. These methods have provided good baseline information, but with the exception of satellite telemetry are limited in their capacity to track the movements of individuals moving long-distances and therefore the habitats used by flying-foxes across broad spatial scales (see General Introduction). For example, radiotelemetry allows continuous, spatially accurate data to be collected throughout the night. However, the method is constrained to a relatively small study area, is biased toward relatively accessible habitat (as animals are generally tracked from the ground) and provides no information on study animals once they move out of the study region. The recent availability of GIS data on land cover for Australian states, combined with developments in lightweight satellite-telemetry, provide an opportunity to quantitatively investigate the landscape scale characteristics of feeding and roosting sites of Australian flying-foxes across broad spatial scales. Satellite technology allows animals to be monitored through time regardless of their migratory movements, across all types of landscapes (unrestricted by accessibility) and across large spatial scales. Satellite telemetry has some limitations, including discontinuous data collection with limited control of timing, limited fine-scale spatial accuracy (McKeown and Westcott 2011) and lack of direct information on behaviour, as data are collected remotely. Despite these limitations, satellite telemetry has the potential to fill a number of knowledge gaps about feeding and landscape use of highly mobile species. This study uses satellite telemetry to improve knowledge of the broad scale feeding ecology and habitat use of Pteropus. Specifically this study quantifies the feeding movements of P. poliocephalus, examines the landscape characteristics of their feeding and roosting sites and examines factors affecting the distance moved by individuals from roosts to feeding sites. Data from 14 satellite-tracked individuals collected over a period of two years and an area of 150,000 km2 are used to address the following questions: (1) What are the landscape-scale habitat characteristics of roosts and feeding sites and how do they differ? (2) Do flying-foxes select roosts and feeding sites on the basis of landscape-scale habitat characteristics? (3) What distances are travelled between roosts and feeding sites?
60 (4) What are the roles of specific factors such as roost population size, duration of visits to feeding sites, size of forest patches, time of year, elevation and land cover in determining the feeding distances moved? 4.2 Methods
4.2.1 Capture and satellite telemetry
Fourteen adult male P. poliocephalus were fitted with satellite transmitters and released at four capture sites in subtropical south-east Queensland, Australia; in October 2007 and June 2008 as outlined in Chapter 3. The four capture points were located at Canungra (28.0oS, 153.2oE), Stafford (27.4oS, 153.0oE), Fraser Island (25.4oS, 153.1oE) and North Stradbroke Island (Dunwich) (27.5oS, 153.4oE). Study animals were sampled from the predominantly migratory or nomadic portion of the population (see section 3.2) by targeting: (1) roost sites used most of the year, but which had a large number of animals present at the time of capture in 2007, and therefore a high proportion of mobile individuals (Canungra and Safford, N = 4 individuals), and (2) seasonally used roost sites with large numbers of animals in 2008, where all animals were expected to be migratory or nomadic (Fraser Island and Dunwich, N = 10 individuals). The transmitters were programmed to a duty cycle of 10 hrs on alternating with 48 hrs off. The raw data from the full transmission sequences of all 14 tagged flying-foxes comprised 4,356 location records, of varying presumed accuracy. Records for feeding sites were extracted using fixes that were obtained during the night, excluding times at dawn and dusk when flying-foxes are known to move between roost and feeding areas (the period one hour after sunset and before sunrise, data from the Geosience Australia website adjusted for geographical location). Cases where bats were recorded within 2 km of the day roost (which could have been due to low-accuracy records of night-roosting individuals) or cases where bats had returned to known day-roost locations before sunrise were excluded. From the resulting 1,913 feeding records, the 15% which fell within Argos accuracy classes A, B and Z were removed as they had no upper limit to their error margins, yielding 1,619 fixes for further analysis. An error in location accuracy of 2 km was assumed for all remaining fixes (LC>0) (McKeown and Westcott 2011).
61 An average of 3.1 night fixes of accuracy <2 km was recorded per 10 hour on cycle (range 19, SD 1.4). These were mapped in ArcView 3.3, and manually scrutinised by overlaying geographical landmarks and known roost locations (Eby 2007; and data held by the author, P. Eby and management agencies).
4.2.2 Feeding sites
To exclude possible cases where the satellite fixes reflected larger-scale or migratory movements and to ensure that the roost location was known or reasonably assumed, records were only included in analyses if they consisted of sufficiently accurate position fixes at the same roost site on two consecutive days together with at least two sufficiently accurate night (feeding) fixes at the same place during the intervening night. Given the remote nature of satellite telemetry, it was not possible to confirm the behaviour of the animal at the time of individual night fixes (e.g. feeding or in flight). Feeding fixes were identified where two night fixes separated in time (by at least one hour) were located within an area of 2 km radius and more than 2 km from the roost. Each of these clusters of fixes is referred to as a flying-fox feeding site and may have consisted of an individual feeding tree or a cluster of trees within that 2 km radius. The point locality of the feeding site was taken from the most accurate fix in the cluster (or in cases where there was more than one fix with the same accuracy, one was randomly chosen). Radio-telemetry studies have shown that flying-foxes can use more than one feeding site per night, separated by several kilometres (Eby 1996; Birt 2005). In this study, the potential for identifying visits to different feeding sites within one night was restricted by uncontrolled technical considerations including the configuration of satellites and the level of charge in the battery. In circumstances where feeding sites were identified at two or more distinct time intervals and spatial locations on a given night, the site containing the most accurate fix was selected for analysis. Where there was more than one site with the same level of accuracy, one was selected randomly. Fixes received from the same study individual at the same feeding site were treated as independent feeding sites if: 1) the feeding site was accessed from different roosting locations; 2) the feeding site was accessed from the same roost after an interval during which it had used a different feeding site; or 3) the feeding fixes were >7 days apart (an interval considered sufficiently long, given the movement patterns of flying-foxes (Chapter 3), to
62 constitute independent decisions by individuals about where to feed). Each resulting record, of a visit to feeding site in relation to a particular roost that was used on both the preceding and subsequent days, is termed a feeding record. There were 173 such feeding records (termed all feeding records) for which there was information on date and time, geographical coordinates, and accuracy (Argos Location Class) of the fix. Some individuals were recorded at the same feeding site over several consecutive weeks. To examine any effect of non-independence of data collected from the same animal, consecutive records at the same feeding site from the same roost were also condensed into single records. This reduced the sample size to 112 independent feeding records, which were analysed separately from the dataset comprising all 173 feeding records. From all 173 records, 110 were spatially distinct feeding sites (>2 km apart). This sample size differs slightly from the 112 independent feeding records as it does not incorporate the roost site used to access a given feeding site. Durations of visits by each individual to each of these 110 different feeding sites were calculated from data on feeding visits, defined by the dates of the first and last fixes in consecutive duty cycles from a particular individual at a particular feeding site (including some occasions when an individual used more than one feeding site per night). The visit duration was the elapsed number of days at that feeding site (including durations of one day only). Gaps in data collection of 24 days occurred due to the 48-hour off segment of the duty cycle and dates when only day (roost) fixes were obtained. When a sequence of consecutive records at the same feeding site spanned these gaps, they were included in calculations of visit duration. Revisits to previously used feeding sites were counted separately; these occurred if an individual had meanwhile been recorded at other roost sites or for any time period had exclusively fed at a separate feeding site.
4.2.3 Roost sites
The 173 feeding records were associated with 45 different roost sites. This discrepancy occurred because: (1) multiple animals used the same roosting site; and (2) individuals used the same roosting site through time and to access different feeding areas. Roost sites were fixes that were either: (1) obtained during daylight hours (between sunrise and sunset, data from the Geosience Australia website adjusted for geographical location); or (2) cases where bats had returned to known roosts (Eby 2007; and data held by the author, P. Eby and management agencies) before sunrise. For each of the 173 feeding records, the name (if
63 known) and co-ordinates of the roost, the date when the individual arrived at and departed from each roost (see also Chapter 3), and the roost population sizes, were recorded. Roost population size data were compiled from static counts collected at most roosts on the third Friday of every month from September 2007 until the end of 2009 (unpublished data held by authors and the Queensland Environment Department). These roost population data were estimates made by researchers, wildlife carers and experienced government staff. Roost population data were available for 108 of the 173 feeding records. Time lags between roost estimates and feeding records ranged from 0 to 19 days.
4.2.4 Landscape measurements
To determine factors associated with the selection of roost and feeding sites, landscape attributes of those sites were compared with 100 stratified random points drawn from areas in which feeding and roosting were recorded, as described below. The 110 different feeding sites and 45 associated roosts were mapped in Google Earth Pro and a 20 km circle was drawn around each. This distance corresponds with the typical known nightly feeding distance of P. poliocephalus (Eby 1991a; Tidemann 1999); data from this study. Seven discrete clusters of sites were identified and a perimeter line was drawn around the outer boundary of all 20 km circles within each cluster (Figure 4.1). The number of random points selected within each defined region was determined by the proportion of feeding sites contained by that region (e.g. a region with 23 of the 173 weekly feeding sites was allocated 13% of the 100 random points; rounded off to the nearest unit). Two feeding sites and two roost sites were >100 km outside these seven discrete regions and were omitted from some analyses. Four landscape attributes were measured at each of the three types of sites: feeding sites, roost sites and random sites. Landscape attributes were: (i) elevation (at the point co-ordinates of the site); and the percentages of three land cover categories; (ii) forest; (iii) built; and (iv) open within a 2 km radius (total area = 12.6 km2) of a site. For feeding sites, the size (ha) of the largest forest patch that fell in total or part within the 2 km radius was also measured. Land cover measurements were taken from high-resolution aerial and satellite imagery in Google Earth Pro at an altitude of 4 km or less (accessed NovDec 2011). The minimum size of any measured polygon was 5.0 ha. In cases when there were small amounts of a particular land cover scattered within a 2 km radius but the total amount was >5 ha a percentage was estimated. Forest refers to the amount of land that could be observed in the Google Earth Pro
64 imagery as continuous tracts of trees (crown cover >20%). Forest included both disturbed and intact stands of woody vegetation, regrowth and plantations (whether of native and exotic species). Built cover comprised built structures (e.g. human settlements ranging from suburban houses and gardens to highly urbanised areas) and impervious surfaces such as roads and industrial areas. Remaining types of land cover consisted mainly of agricultural land with scattered trees and bodies of water and were placed in the open category.
4.2.5 Data analysis
To investigate intercorrelations among the four landscape attributes (land cover and elevation) at the 2 km scale, Pearson correlation matrices were calculated. The measured values of the four landscape attributes were compared among each of the three different types of sites using: (1) all 173 feeding records; and (2) the 112 independent feeding sites. Whether there was evidence for active selection for landscape attributes of either roost or feeding sites was tested by comparing their landscape values with those of the 100 random points, using Mann Whitney U-tests. Differences in landscape values between roost and feeding sites were tested with the Wilcoxon Matched-Pairs Signed-Rank test. Boxplots were constructed from both all 173 feeding records and 112 independent feeding records to describe the distribution of each of the four landscape measurements. Two roost and two feeding sites that fell outside the defined regions were excluded from these analyses. For each of the 173 feeding records, the distance between the point locality of the feeding site and the centre of the associated roost site (feeding distance) was measured using the shortest straight path in Google Earth Pro. Quantitative summaries of frequency distributions of the feeding distances were made using either all feeding records or the independent feeding records as replicates. Analysis of variance (ANOVA) was used to test if the mean feeding distances differed between individuals, month and seasons. There were four seasons; summer (DecemberFebruary), autumn (MarchMay), winter (JuneAugust) and spring (September November). Only individuals or months with more than 10 records were included in any of the feeding distance analyses. Pearson correlations were used to test for significance of relationships between the feeding distances of flying-foxes and the following attributes: roost landscape characteristics (at 2 km radii), roost population size (log transformed), feeding visit duration and feeding forest patch size. These analyses were conducted at four different levels of the dataset, which vary in breadth of focus, potential redundancy and units of replication: (1) all feeding records, (2) independent feeding records, (3) mean values for nine different
65 individuals with >10 feeding records, and (4) for each of two individual flying-foxes (Bats 5 and 10) which had large (>20) numbers of feeding records together with high variation in feeding distances. 4.3 Results
The 14 male P. poliocephalus used in this study were satellite-tracked from tagging to transmitter failure for a range of 13277 days. The 173 feeding records were unevenly distributed across study animals and ranged from 124 per individual (average 12, SD 6.6; see Appendix 8). Approximately 97% of all feeding records had an Argos accuracy of <1.5 km (LC 1, 2 and 3) and 80% had an accuracy of <0.5 km (LC 2 and 3). The recorded unique roost sites (N = 45) and discrete feeding sites (N=110) of study animals spanned a north-south distance of 1,065 km and an east-west distance of 138 km (Figure 4.1). Most records were concentrated broadly in three localities: around latitudes 24.5oS (14% of records), 25.5oS (22%), and 28oS (44%) (Figure 4.1).
66
240 Gladstone 250
100
Bundaberg
200 260 300
Latitude (degrees south)
270 400
Brisbane
500 290 600 300
Coffs Harbour
700
310
Port Macquarie
800
320 900
Newcastle
330
Sydney
1000
1520
Longitude (degrees east)

Figure 4.1 Map of the 45 different roost sites (triangles) associated with the 110 discrete feeding sites (circles) in use by 14 individuals of P. poliocephalus between October 2007 and April 2009. The east west line in the middle of map shows the border between New South Wales and Queensland. Polygons outline the seven discrete regions where random points were selected.
Kilometers
280
67
4.3.1 Habitat characteristics of feeding and roost sites
The study animals tended to roost and feed in landscapes characterised by high percentages of forested land and low percentages of built land, although the percentages of forest and built varied greatly, from <5% to >95% (Figure 4.2). Of the 45 roosts used by study animals, 40% were in areas where no built land occurred within 2 km of the site, whereas 20% were located in substantially built environments (>35% of the surrounding land) and 4% were embedded in highly urban (>45% built) landscapes. Also 69% of 112 independent feeding sites had no measurable built land cover within 2 km, compared with 7% that were in substantially built landscapes. The patterns of difference in landscape characteristics between roost and feeding sites remained similar whether based on all or just the subset of independent records (Table 4.1, Figure 4.2). Roost sites had significantly less forest and more built land cover than feeding sites (Table 4.1, Figure 4.2). The elevation of roosting sites were significantly less than feeding sites (roost sites: median elevation 110 m; mean 101 m, SD 121; feeding sites: median elevation 99 m, mean 144 m, SD 138) (Table 4.1, Figure 4.2). Of the 171 records, 74% of roost sites were located at elevations <120 m, compared with only 58% of feeding sites, and 67% of 171 paired roost-feeding cases showed animals moving to higher elevations to feed. Compared with random points, roosts were located at sites with less surrounding forest cover, and more built and open land (Table 4.1, Figure 4.2). In contrast, feeding sites were located in areas with less surrounding built cover, but with similar amounts of surrounding forest cover as random points in the landscape (Figure 4.2, Table 4.1). Neither the elevations of roosts nor feeding sites differed significantly from those in the random sample, due to the high variation of landscape data. When the inter-relationships of these landscape variables were examined there were a number of significant relationships (Appendix 9). However, most r values were not especially high (r < 0.5; i.e. r2 < 0.25) other than strong negative relationships between percentage open land and forest when feeding sites were considered (Appendix 9). Therefore, all variables were retained in the analyses.
68
(a) All records

100 80
% Forest
97
73
82
50 40
% Built
60 40 20 0 R F L
30 20 10 0 R
512
F
575
L
790
100 80 60 40 20 0 R F L
Elevation (m)
500 400 300 200 100 0 R

97
% Open
F
73
L
82
(b) Independent records

100 80
% Forest
50 40
% Built
60 40 20 0 R F L
30 20 10 0 R
500 400
512
F
575
L
790
100 80
Elevation (m)
% Open
60 40 20 0 R F L
300 200 100 0 R F L
Figure 4.2 Boxplots summarising the distribution (median, quartiles, range) of elevation (m) and the percentage of forest, built and open land within 2 km of roosts (R), feeding sites (F) and random points in the landscape (L) at two sampling units; (a) all records (R=171, F=171, L=100) and (b) independent records (R=110, F=110, L=100). Maximum values written above bars. Also see Table 4.1.
69
Table 4.1 A comparison of landscape attributes within 2 km radii of roosts (R), feeding sites (F) and random points in the landscape (L) at two different sampling units; (i) all feeding records and (ii) independent feeding records. Two roost sites and two feeding sites that fell outside the defined regions were excluded from this analysis. Table shows P values.
R vs. L1 All feeding records % Forest % Built % Open Elevation Independent feeding records % Forest % Built % Open Elevation
1 2
F vs. L1 0.61 <0.001 0.37 0.14 0.25 <0.001 0.85 0.28
R vs. F2 <0.001 <0.001 0.68 <0.001 <0.001 <0.001 0.03 0.006
Direction F=L>R R>L>F R>L=F F>R=L* F=L>R R=L>F R>F=L F>R=L*
0.01 0.04 0.05 0.27

4
0.01 0.13 0.05 0.45
Mann-Whitney U- tests Wilcoxon Signed-Rank test 3 All records; Roost = 171, Feeding = 171, Landscape = 100 4 Independent records; Roost = 110, Feeding = 110, Landscape = 100 * Based on median values (see Figure 4.2). Note that medians of the roost and feeding did not differ significantly from the landscape medians because of the high variation in the latter.
4.3.2 Movement between roosts and feeding sites
The feeding distances (i.e. distance from roost to feeding site) moved were typically (85%) less than 20 km (range 256 km) and were similar between all 173 feeding records and the 112 independent feeding records (Figure 4.3, Figure 4.4). However, there was a large amount of variation in the feeding distance moved, both within and between individuals (Figure 4.4). ANOVA tests showed statistically significant variation in mean feeding distance among the nine individuals with 10 feeding records when using all records (ANOVA F=3.085, P=0.003), but mean feeding distance was not significant among the five individuals with 10 independent records (ANOVA F=1.913, P=0.121) (Figure 4.4). LSD tests showed that Bats 2 and 4 differed significantly from Bats 9 and 1114; although no clear pattern was seen when all records were considered.
70
40 30 20 10 0
(a) All records
Frequency (%)
<5
6-10
11-15
16-20
21-25
26-30
31-56
Feeding distance (km) 40 30 20
Frequency (%)
10
0 <5 6-10 11-15 16-20 21-25 26-30 31-56
Feeding distance (km)

Figure 4.3 Frequency distribution (%) of the feeding distances moved by the 14 P. poliocephalus; (a) all 173 feeding records (mean 12.3, median 9.8 km, SD 9.3 km); and (b) 112 independent feeding records (mean 13.4, median 10.9 km, SD 10.1 km).
Overall, the duration of time over which individuals revisited any particular feeding site was typically short, with 50% of feeding visits <5 days (Figure 4.5). However, the distribution of feeding visit durations was skewed, and 19% of visits were for periods >25 days (and up to 84 days, Figure 4.5). On 12 occasions, two different roosts were used by an individual bat to access the same feeding site.
71
60
(a) All records

Feeding distance (km) 50 40 30 20 10 0 0 1 2* 3 4* 5* 6 8* 9 10* 11* 12* 13* 14*
7 Bat
60

Feeding distance (km) 50 40 30 20 10 0 0 1 2 3 4* 5* 6 8* 9 10* 11 12 13 14*
7 Bat
Figure 4.4 Feeding distance (km) moved by each of the 14 P. poliocephalus; (a) all records and (b) independent records. Asterisks represent individuals with more than 10 feeding records that were included in analyses. Each point is the distance (km) separating the roost and feeding site. Solid lines show the mean for each data set. Bats ordered in rank-order of the mean feeding distances (all records). See also Appendix 8.
Using the 173 feeding records, study animals were recorded using more than one feeding site in a night on 54 occasions. In 13 of these, 35 feeding sites were used per night. The mean distance between successive pairs of within-night feeding sites was 7 km (SD 5 km, range 2 21 km, N=54. Five fixes were received from one animal (Bat 8, 80188) as it migrated
72 overnight between roosts located 210 km apart, taking an indirect path, travelling at least 246 km, and making at least one feeding stop during the trip.
50 40
Frequency (%)
30 20 10 0 1-5 6-10 10-15 Days 16-20 21-25 26-84
Figure 4.5 Frequency distribution of the duration of feeding visits by 14 Pteropus poliocephalus (N = 110 visits to different sites). A feeding visit is the time from the first to last consecutive satellite fix for any individual at a particular site, with no data gaps of >5 days. Revisits were counted separately. Mean and median durations were 12 and 6 days, respectively.
4.3.3 Factors influencing distance moved
Feeding distance was significantly negatively correlated with the duration of feeding visits when all 173 feeding records were analysed, (r=-0.224, P=0.003) (Table 4.2, Figure 4.6). This relationship was no longer significant when using the subset of independent records (r=0.097, P=0.309, Table 4.2). However, across the nine individuals with sufficient data (i.e. with more than 10 feeding records) there was a negative correlation between feeding distance and the duration of feeding visits (r=-0.638, P=0.065) and a negative trend was also seen for Bat 5 (r=-0.368, P=0.077) (Table 4.2, Figure 4.6).
73
Table 4.2
Correlations (Pearson) between the feeding distances moved and roost population size (log transformed), feeding visit duration (days) and feeding forest patch size (ha) using five different units as replicates; all feeding records, independent feeding records, individuals (with more than 10 feeding records) and within two individual flying-foxes which had large (>20) numbers of feeding records together with high variation in feeding distances (Bat 5 and Bat 10). The expected row refers to the projected correlation before conducting anaysises i.e. the relationship expected to be positively (+) or negatively (-) correlated.
N Expected All records1 Independent records Individuals Bat 53 Bat 10

4 2
Log roost population +,-
Feeding visit duration (days) -0.224* -0.097 -0.638

x
Feeding forest patch (ha) + 0.085 0.114 0.110 0.096 0.057
173 112 9 24 21
0.216 -0.124 0.470 -0.071 0.288
-0.368x -0.180
* P < 0.05; x Strong trend P <0.1 1 N = 173 for all correlations except roost population size, where N = 108 2 N = 112 for all correlations except roost population size, where N = 67 3 N = 24 for all correlations except roost population size, where N = 22 4 N = 21 for all correlations except roost population size, where N = 11
74
60
(a) All records

50
40
30 20 10 0 0 25 20 40 60 80
(b) Individual flying-foxes
20 15 10 5 0
0
40 35
10
20
30
40
(c) Bat 5
30 25 20 15 10 5 0 0 10 20 30 40
Feeding visit duration (days)

Figure 4.6 Relationships between the feeding distances moved (km between roost and feeding sites) and feeding visit duration (days) for cases where there was a relatively strong (P < 0.10) correlation; (a) all feeding records (N = 173), (b) individual flying-foxes (N = 9), and (c) Bat 5 (N = 24). Line shows the linear regression of the one relationship where P < 0.05 (see also Table 4.2).
75 Feeding distance did not vary with month (Figure 4.7, ANOVA F=0.673, P=0.695) or season (ANOVA F=1.194, P=0.314). Neither elevation nor type of land cover surrounding roosts were significantly associated with the distances that flying-foxes travelled to feed (Table 4.3).
Table 4.3 Correlations (Pearson) between the feeding distances moved and four landscape attributes within 2 km radii of roosts at two different sampling units; all feeding records (N=173) and independent feeding records (N=112).
Variable name All feeding records % Forest % Built % Open Elevation
0.07 0.02 0.06 0.03
0.38 0.80 0.42 0.68
Independent feeding records % Forest % Built % Open Elevation 0.05 0.03 0.09 0.06 0.62 0.75 0.35 0.57
30
Mean feeding distance (km)
22 16 10 15 7 4 5 1 26
25
25 20 15
26
23
10
5 0 Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov (6) (7) (4) (4) (2) (1) (5) (10) (9) (9) (11) (10)
Figure 4.7 Mean monthly feeding distances (+SD) of 14 Pteropus poliocephalus. Numbers above bars are sample sizes. Number of animals transmitting each month are in parentheses beneath the y axis.
76 4.4 Discussion
4.4.1 Roost sites
This study is the first to examine feeding distances and landscape use of flying-foxes using satellite telemetry. This technology allowed a more comprehensive assessment of the movements of animals than has been possible, due to the capacity to track movements in remote or otherwise logistically difficult areas. While individual roosts were located in landscapes that ranged from highly urbanised areas (97% built) to substantial tracts of intact forest (100% forest), the majority were surrounded by mosaics of open, built and forested lands. Significant differences between the proportions of landscape attributes in the areas surrounding roosts and random points indicate that the animals were selective in the landscapes they used for roosting. At roost sites P. poliocephalus selected landscapes with less forest cover and more open land than random, and these relationships were consistent in analyses of all feeding records and independent records. This trend might be associated with preference for roosting on level land or gentle slopes. Several previous studies have found the majority of P. poliocephalus roost sites in eastern Australia were on flat ground or that at least part of the roost site contained some level topography (<5o incline) (Eby 2002b; Peacock 2004; Roberts 2005). Clearing rates for urbanisation, agriculture and grazing have not been uniform across the east coast of Australia, with the steeper lands on hills and ranges retaining a higher proportion of their natural vegetation cover, whereas much of the vegetation along the flatter and more fertile coastal lowlands (< 100 m) has been cleared and fragmented (Catterall and Kingston 1993; Sattler and Williams 1999; Keith 2004). When all feeding records were included in the analysis, flying-foxes used built areas for roosting at a significantly greater frequency than their occurrence in the environment; however, this preference was no longer significant when only independent records were included in analyses. Roberts (2005) found roost sites of flying-foxes (P. poliocephalus and P. alecto) in south-east Queensland were located in areas surrounded by a significantly greater proportion of urban development than was the case from random points. Roberts (2005) also found that roost sites used regularly by flying-foxes were surrounded, on average, by twice the amount of urban land, than those used less regularly. The weaker trend in the
77 present study may be explained by the use of satellite telemetry which provides better information on remote and inaccessible roost sites away from urban areas and reinforces the benefit of satellite telemetry in documenting movements across all landscapes. Flying-foxes have been increasingly reported roosting in urban locations (e.g. Birt et al. 1998; Hall and Richards 2000; Hall 2002; Markus and Hall 2004; van der Ree et al. 2006). There is evidence that in some urban areas the numbers of roost sites are increasing (Birt et al. 1998; Hall 2002; Markus and Hall 2004; Roberts 2005) and that these sites may be used regularly by flying-foxes (Birt et al. 1998; Roberts 2005; van der Ree et al. 2006; Williams et al. 2006; this chapter). Nonetheless, this study has also shown that flying-foxes across their range will often roost outside urban areas. There is no evidence to suggest flying-foxes are abandoning rural locations to roost in urban areas, although, when present urban land appears to be a factor influencing roost placement at the local scale. Flying-foxes use of urban areas may have recently been over-emphasised because roost sites near built areas attract controversy (Eby and Lunney 2002a; Chapter 5), or because urban roosts are more readily documented than roosts in remote or inaccessible areas. A number of studies have examined the characteristics of roost sites used by Australian flying-foxes (e.g. Richards 1990; Loughland 1998; Palmer and Woinarski 1999; Tidemann et al. 1999; Peacock 2004), although fewer have examined the landscape scale attributes of such sites (e.g. Roberts 2005). This studys findings are consistent with previous observations that P. poliocephalus uses habitats with a wide range of tree cover characteristics for roosting (Tidemann 1999; Eby 2002b; Peacock 2004; Roberts 2005). The roost sites of P. poliocephalus are located within landscape types ranging from small remnant forest patches surrounded by open agricultural land to large tracts of continuous forest. In eastern Australia, P. poliocephalus appears to require forest patches at least 1 ha in size for roosting colonially in large groups (Eby 2002b; Roberts 2005). This study did not find that elevation played a role in the selection of roost sites, however, flying-foxes did move to higher elevations to feed. Previous studies found a preference for roosts of Australian flying-foxes at low elevations (McWilliams 1986; Richards 1990; Tidemann et al. 1999; Eby 2002b; Peacock 2004; Roberts 2005). Again, it is possible that this difference reflects the capacity for satellite telemetry data to document movements across all landscapes. Roosts at high elevations are more likely to be found in areas easily accessible and roost locations are less likely to be recorded than that at lower elevations. In addition, it is
78 possible that the study did not have the statistical power to determine elevation preferences because of the high variability of elevation in the landscape data.
4.4.2 Feeding sites
Pteropus poliocephalus has previously been reported to use a variety of treed landscapes for feeding within its range (e.g. McWilliam 1986; Eby 1996; Parry-Jones and Augee 2001; McDonald-Madden et al. 2005), however, this is the first study where landscape characteristics of feeding sites have been systematically assessed. This is also the first study to look at preferences in feeding landscapes for flying-foxes. Consistent with previous suggestions (e.g. Parry-Jones and Augee 1991; Eby 1996; Markus and Hall 2004), this study found that the feeding sites of P. poliocephalus included areas of natural forest, disturbed forest, open areas with scattered trees and vegetation within the urban matrix. Individual feeding sites were surrounded by landscapes that ranged from highly urbanised areas (82% built) to intact forest (100% forest). However, the majority comprised mosaics of landscape types. Pteropus poliocephalus was less selective of feeding landscapes than roosting landscapes. The proportions of forested and open land in the areas surrounding feeding sites did not differ from those surrounding random points. Nor was there evidence that P. poliocephalus selected feeding sites on the basis of elevation. However, there was a strong and significant negative relationship with the proportion of built land, suggesting the animals avoided built areas when feeding. This result was consistent in analyses of all records and independent records. However, the flying-foxes tracked in this study showed no preference for feeding in large continuous tracts of forest. In urban areas, scattered native and exotic trees, small forest patches and larger remnants retained within the urban matrix provide feeding sites. Open areas were usually agriculturally altered landscapes containing scattered trees and thin strips of riparian vegetation. Regardless of the statistical preferences, the flying-foxes moved freely across modified and disturbed environments, to access individual productive trees in all habitats. The selection against built areas may reflect a lower density of productive trees there, although, further research would be needed to quantify likely food availability between forest, rural and urban areas. Potential differences in flying-fox habitat use between different regions, differing in the extent of urbanisation and forest cover, also warrant further investigation.
79
4.4.3 Comparison between feeding and roosting sites
There are alternative, but untested explanations why flying-foxes may prefer to roost in urban areas. First, some authors have concluded that an increased food supply provided by increase in garden and streetscape plantings of fruit and nectar-bearing trees could explain an increase in the occurrence of flying-foxes in urban areas (Parry-Jones and Augee 2001; Markus and Hall 2004; Birt 2005; McDonald-Madden et al. 2005; van der Ree et al. 2006; Williams et al. 2006). However, the present study has shown that when tracked across all landscape types, P. poliocephalus was primarily feeding away from built environments. This may be because food availability may actually be lower in urban areas, as discussed above. Second, the establishment of flying-fox roosts in urban areas may be driven by factors other than the availability of food. Since flying-foxes can travel tens of kilometres each night they do not need to locate roosts immediately adjacent to preferred feeding locations (also see Eby 1991a; Spencer et al. 1991; Law and Dickman 1998; Chapter 3). Urban areas may contain specific habitat features which make small urban forest remnants attractive as flying-fox roost sites. For example, highly illuminated features of urban areas such as roads and other infrastructure may make night time navigation by flying-foxes easier (Birt et al. 1998; Hall and Richards 2000). Alternatively, the persecution of flying-foxes by humans in rural areas (Lunney and Moon 1997; Hall and Richards 2000) may have led to the establishment of some roosts in human settlements. Other possible, speculative explanations for the apparent selection of built-up areas for roost sites include microclimate moderation, i.e. the urban heat island effect (Parris and Hazell 2005), and the reduced likelihood of predation by some predators (e.g. raptors, varanids) in urban areas (Churchill 2008). Nevertheless, as flying is energetically expensive, reasonable proximity of roosts to food resources would be expected (as indeed was demonstrated in this study).
4.4.4 Factors influencing feeding movements
The extensive variation in the feeding distances of P. poliocephalus found in this study (256 km) is consistent with results of studies using both banding (Tidemann 1999; Tidemann and Nelson 2011) and radio-telemetry techniques (e.g. McWilliam 1986; Eby 1991a; Spencer et al. 1991; Eby 1996). Through the recovery of dead banded P. poliocephalus in urban Sydney, Tidemann (1999) concluded that 76% of flying-foxes foraged within 20 km of the roost, although the feeding range extending to 45 km. Spencer et al. (1991) reported roost-feeding site distances of 25 km for one flying-fox in regional Lismore and up to 17 km for two flyingChapter 4 Feeding movements and landscape use by P. poliocephalus
80 foxes in Sydney. Two additional radio-telemetry studies in Lismore reported direct distances between day roosts and feeding locations of up to 10 km (McWilliam 1986) and from 2.6 km to 37 km (Eby 1991b, 1996). When an animals entire flight path is tracked, flying-foxes can potentially travel considerably longer distances than their straight line path; with commuting distances of up to or greater than 50 km one way (Eby 1991a). Similar feeding distances have been found in other species of flying-fox. Based on radiotracking data, studies have found that P. alecto in northern and eastern Australia generally roosted within 15 km of feeding sites (range 0.129.3 km, McWilliam 1986; Palmer and Woinarski 1999; Palmer et al. 2000; Markus and Hall 2004); P. scapulatus in eastern Australia fed between 1 and 29 km from the roost (McWilliam 1986; Birt 2005); feeding distances for P. vampyrus and Acerodon jubatus in the Philippines ranged from 0.44 to 12.6 km (mean 5.0 km, Mildenstein et al. 2005) and P. tonganus in American Samoa travelled a mean distances of 22.8 km (Banack and Grant 2002). This study adds to the body of information about longer-range feeding, showing that 15% of feeding movements were greater than 20 km from the roost (7% were >30km) and on three occasions two individuals travelled >50 km one way to feed. The limited fine-scale accuracy of satellite telemetry (see Methods 4.2; McKeown and Westcott 2011) precluded the identification of feeding sites <2 km from the roost. While this had the potential to skew results toward longer distance movements, there were too few examples of nocturnal movements < 2 km (N=7) to alter the results. Various factors examined in this study including season, roost population size, size of feeding forest patch and landscape characteristics surrounding roosts did not explain the variation which occurred in feeding distances. However, flying-foxes that travelled shorter distances also remained for longer periods at particular roost site - feeding site combinations. Flyingfoxes can minimise commuting costs by feeding close to the roost and hence it would be expected that individuals would use feeding sites near to roosts (or vice versa) for as long as possible. However, where the availability of food varies across the landscape (in particular, where food sources of relatively high nutritional and energetic value are located at comparatively long distances from roosts), the costs of flying extra distances for food may be small relative to the quality and quantity of resources that can be harvested (see also Karasov 1992). Where possible, it could be expected that flying-foxes should shift to new roost sites located close to productive feeding sites.
81 Two radio-telemetry studies of P. poliocephalus and P. alecto have suggested that food type may influence feeding distance, with flying-foxes traveling greater distances to use particular energy-rich nectar resources (Eby 1996; Palmer and Woinarski 1999). In the eastern Australian subtropics, Eby (1995; 1996) observed P. poliocephalus moving consistently greater maximum distances (up to 50 km) when feeding on blossom from eucalypts compared with fruit (up to 29.5 km). Flying-foxes feeding on blossom also used more feeding trees per night, and distances between feeding trees were greater (up to 20 km between feeding sites for blossom, compared with 8.3 km for fruit). A study in northern Australia found that the feeding distances of P. alecto changed seasonally, with individuals travelling further during the dry season when feeding on Eucalyptus species compared with the wet season when feeding on locally abundant rainforest fruit (Palmer 1997; Palmer and Woinarski 1999). The feeding movements of P. poliocephalus varied considerably both among and within individuals. Similar wide variation in the distance moved by individuals to feed has been reported by other studies of this (e.g. Eby 1996) and other Australian flying-foxes (P. alecto Markus and Hall 2004; P. scapulatus Birt 2005). A study of P. poliocephalus in the eastern Australia subtropics found a negative relationship between the feeding distances moved and the individuals fidelity to a roost site (Eby 1996), consistent with this studys findings. Resident or sedentary flying-foxes persisted at the same roost throughout most of the year including periods when the roost population was low (Eby 1995; Eby 1996; Fleming and Eby 2003). Eby (1995; 1996) found that resident flying-foxes generally fed closer to the roost than the non-resident, migratory flying-foxes. Resident flying-foxes also used fewer feeding sites per night, rarely more than one, flew significantly shorter total distances per night and returned to specific feeding areas over longer periods (many weeks). The present study was not designed to identify the social status of individuals. In the present study, only one individual persisted at a particular roost for an extended period (Bat 5 spent 83% of days at one roost; Chapter 3). The remainder of the flying-foxes tracked in this study moved frequently among roost sites (Chapter 3). The typical duration of feeding visits for P. poliocephalus tracked in the present study was generally less than that reported for other studies. Half of the visits to feeding sites were for less than one week (mean 12 nights, median 6 nights), although 19% were for periods of >25 nights (up to 84 consecutive nights). Previous studies have repeatedly reported that flyingfoxes can maintain fidelity to feeding sites for an average of one month, although with
82 considerable variation (range 5-65 days) (McWilliam 1986; Spencer et al. 1991; Eby 1996; Palmer et al. 2000; Markus and Hall 2004). In the rural subtropical area of Lismore the maximum duration at feeding sites was 29, 37 and 23 nights for P. poliocephalus, P. alecto and P. scapulatus (McWilliam 1986). The relatively short feeding durations found in this study may be explained in part by the migratory status of the study animals. Eby (1996) found that the duration at feeding sites of P. poliocephalus differed with the migration status of the individual, with resident animals remaining at individual fig trees Ficus spp. for more than 4 weeks (range 18-46 nights, mean 34.9, N = 8) and migratory animals never staying at one feeding site for longer than 18 nights (range 5 18, mean 11.6, N = 8). The present study was designed to target migratory individuals (see Methods 4.2.1) and our results are similar to those of migratory animals found by Eby (1996). The short duration stays at feeding sites were also consistent with the frequent movements among roosts (see Chapter 3). Establishing how flying-foxes and other mobile species move across the landscape to feed, and establishing how they respond to different patterns of habitat alteration is an important requirement for effective conservation. This study has shown the value of using satellite telemetry to assess the habitat use of highly mobile species such as flying-foxes, as it allows a comprehensive examination of the all the landscapes used by individuals across their entire range including remote and inaccessible areas. It has been previously suggested that flying foxes may be tolerant to the urbanisation of their habitat because their mobility allows them to move freely across modified landscapes and to use resources within the urban matrix (Law and Dickman 1998; Birt et al. 1999; Markus and Hall 2004; Williams et al. 2006; Chapter 3). Our study confirms that P. poliocephalus respond positively to urbanisation around their roosts, but they do appear to be sensitive to urbanisation in terms of their selection of feeding habitat (i.e. they select for forest or open habitats, at least within the study area). Therefore, conservation of P. poliocephalus is likely to benefit from the retention of forest patches within 20 km of roost sites. In addition, effective conservation of P. poliocephalus involves a landscape-based approach, which considers the different feeding and roosting preferences of the species and maintains connectivity between these habitats across the large geographical range of the species.
83
Chapter 5 The outcomes and costs of relocating flyingfox roosts: insights from the case of Maclean
5.1 Introduction
The intentional movement of animals or populations from one location to another has become a popular tool to manage wildlife, both for conservation and to resolve humananimal conflicts (Griffith et al. 1989; Wolf et al. 1996; Fischer and Lindenmayer 2000; McLachlan et al. 2007). Today, relocations are still extensively used in the conservation of rare or threatened species; however, relocations are also increasingly being employed in wildlife management in attempts to remove individuals or populations of problem animals from locations where they come into conflict with humans (Fischer and Lindenmayer 2000; Massei et al. 2010). Relocations of problem animals are likely to continue to increase in frequency, as current trends of human population growth and landscape development suggest that the frequency of human-animal conflicts is likely to continue to grow into the future (Massei et al. 2010). Relocation of animals to resolve humananimal conflicts typically involves common or abundant species, and range from the relocation of a single problem animal to relocation of large groups (Massei et al. 2010). Traditionally, problem animals were managed by lethal methods such as shooting, culling and poisoning (Massei et al. 2010). However, growing public opposition to lethal methods in some developed countries has seen a shift toward relocations, which are perceived as a more humane and benign management method (Jones and Nealson 2003; Massei et al. 2010). Despite the perceived humaneness of relocations, the fate of relocated animals is rarely monitored and the overall effectiveness of these actions in resolving conflicts remains questionable (Fischer and Lindenmayer 2000; Massei et al. 2010). For example, relocated animals may create conflict in the area to which they are moved. In other cases, the problem may persist in the original area because, for example, the high mobility and homing behaviour of the relocated species results in relocated animals returning to the original area or because newly arrived individuals become problem animals (Fischer and Lindenmayer 2000; Massei et al. 2010). In eastern Australia, the relocation of roosts of flying-foxes (Pteropus spp.) is regularly proposed by some members of the community, typically in cases where these bats have
Chapter 5 The outcomes and costs of relocating flying-fox roosts
84 established colonies close to residential areas or when human development occurs too close to established roost sites (Birt et al. 1998; Hall and Richards 2000). The costs of relocating flying-fox roosts can be considerable (West 2002; Thiriet 2005; Roberts 2006; Nelson 2008a) and there is ongoing debate around the long-term success of such projects (Hall 2002; Tidemann 2002; West 2002). However, very little effort has been allocated to monitoring the activities involved in previous relocation attempts, or their costs or outcomes, despite their well-established and increasing use in Australia (Hall 2002; Tidemann 2002; West 2002; Lunney 2012). This paper examines the consequences of attempts to relocate a flying-fox roost at Maclean in north-east New South Wales (NSW). Based on the results, I discuss the utility of relocation as a management tool to resolve conflict between humans and flyingfoxes. 5.2 Study region and its flying-foxes
5.2.1 Flying-fox roost sites in the Lower Clarence region
The Lower Clarence region in north-eastern NSW covers an area of approximately 1,500 km2. Floodplains in the region have been extensively cleared for cane growing and cattle grazing, however, there are still some small areas of remnant rainforest and other types of native vegetation on the floodplains, and extensive areas of sclerophyll forests in the surrounding region (Figure 5.1). By the end of the 20th century the human population of the region was around 17,500, many of whom lived in settlements along the Clarence River.
85
Figure 5.1 All known flying-fox roost sites in the Lower Clarence region that were occupied during the period of licensed disturbances (April 1999 to December 2007). Yellow circles = historical sites used prior to the disturbances. Triangles = new sites that were occupied after the disturbance (red triangles continuously occupied sites and blue triangles were temporary sites generally on mangrove islands). AS = Ashby; BO = Bolorobo Island; IL = Iluka; LA = Lawrence (exact location unknown); MG = Maclean gully (350m from MRR); MRR = Maclean Rainforest Reserve; SL = Sleeper Island; TH = Thorny Island; UL = Ulgundahi Island; WA = Warregah Island.; WH = Whyna island; YA = Yamba; and YG = Yaegl Nature Reserve.
Flying-foxes were recorded in the region from 1885 (Tanton 1999; West 2002). The region is in the centre of the geographical range of the grey-headed flying-fox Pteropus poliocephalus, suggesting longer-term occupation (i.e. much longer than historical records). The first quantitative records of the occupancy and abundance of roosts commenced with a census of grey-headed flying-foxes undertaken by the Australasian Bat Society in July 1998. Since 1998, there have been regular broad-scale systematic surveys of the usage of roosts across the Clarence region (Eby et al. 1999; Eby unpublished data; Roberts 2006; Roberts unpublished data). Until 1994, the grey-headed flying-fox was the main occupant of roosts in the Lower Clarence region, with sporadic influxes of the little red flying-fox P. scapulatus (Eby et al.
86 1999; Tanton 1999; West 2002). By 2009, both grey-headed and black P. alecto flying-foxes frequently occurred together in roosts. According to historical records (Tanton 1999; West 2002), three roost sites have been repeatedly occupied over time: Maclean Rainforest Reserve (MRR), which is described in detail below; Yaegl Nature Reserve (located 2.8 km north-east of MRR), which is occupied during late summer and autumn of most years; and Angourie Road (14.8 km east of MRR), which is also occupied most years, but not continuously (Figure 5.1). Flying-foxes have also been recorded using many other sites in the region as roosts, but such sites appear to have been used temporarily or irregularly (Lunney and Moon 1997; Tanton 1999; B. Roberts pers. obs.). In the Lower Clarence, only two locations have been occupied year round: MRR in the absence of disturbances and, since 2004, a roost in the township of Iluka. These year-round roost sites are located in dense riparian rainforest or mangroves (Tanton 1999; Roberts 2006).
5.1.2 Maclean flying-fox roost relocation
MRR is a small (one hectare) patch of remnant subtropical rainforest located on the southwest periphery of the Maclean township (29.4643oS, 153.2042oE; Figure 5.2). Flying-foxes regularly roosted in MRR from at least the early 1890s to 1999. The number of flying-foxes using this site has fluctuated considerably over time and according to newspaper reports has occasionally exceeded 100,000 individuals (Tanton 1999; West 2002). Historical records show that since the early 1890s flying-foxes using this roost have been repeatedly disturbed by humans, initially to control numbers, and later in attempts to relocate them, so as to reduce vegetation damage and impacts upon the neighbouring community (Lunney and Moon 1997; Tanton 1999; West 2002). There are numerous reports of private and government sponsored hunts to destroy or disperse the roosting animals using shooting, fires and explosives (West 2002). However, flying-foxes continued to return to this site despite these disturbances. In the early 1990s, as a result of the legal protection of flying-foxes, these disturbances ceased and animals continuously occupied MRR without further harassment until 1999.
87
Figure 5.2 Roost habitat occupied by flying-foxes in the Maclean township. The red outline shows the original site used from at least the early 1890s to 1999 (MRR); the yellow outline shows the lower part of the Maclean gully (MG) occupied continuously since 2007; the green outline shows an additional area occupied at times of maximum population size after 1999 (upper Maclean gully and vegetation adjacent to Maclean High School). Arrows show residences impacted; further residential development has also been approved for the cleared areas around the Maclean gully.
Regardless of the presence of flying-foxes in MRR, the rainforest remnant and the surrounding land were set aside for public use by the Municipal Council of Maclean in 1889 (West 2002). As the Maclean township grew, several community facilities were constructed on the land including a cemetery, showground and, in the early 1960s, the local high school. The initial school buildings were positioned 80 m from MRR, but as the human population of Maclean grew, additional classrooms and other education facilities were constructed closer to the reserve, including construction of classrooms within 10 m of the flying-fox roost in 1996 (West 2002). In 1994 and 1996 there were significant influxes of flying-foxes, primarily little red flying-foxes, into the site (West 2002). This situation prompted increased pressure from
88 the school community and nearby residents for the removal of the bats, due to concerns about the odour, noise, faeces and urine associated with the roost, and the perceived threat of disease transfer from the flying-foxes to the local community (Tanton 1999; West 2002). The roosting flying-foxes also caused damage to parts of the canopy in the small patch of remnant rainforest. Other members of the community, including some residents, conservation groups, and welfare organisations, considered the site important for the local flying-fox population and argued that the roost should be protected. There was public discussion of a variety of management options to reduce the conflict, including relocating either the school or the flying-fox colony. By 1998 the NSW government responded to the ongoing conflict by forming a working party to discuss and implement a draft action plan (West 2002). The working party consisted of representatives from local and state government (including the Department of Education and Training (DET), the National Parks and Wildlife Service (NPWS), and Department of Land and Property Management Authority), the Maclean High School, and other sectors of the community (including the Maclean Parents and Citizens Association). The working party decided that the flying-fox colony should be subjected to a controlled disturbance regime, which aimed to reduce bat numbers at MRR and to induce them to move to the nearby Yaegl Nature Reserve. Throughout the process there was considerable political involvement (West 2002). Repeated, but irregular, use by flying-foxes (generally between February and June) had previously been reported at Yaegl. The species that frequented the site was generally unknown because of its inaccessibility. Recent observations suggest the Yaegl roost is primarily used by little red flying-foxes, although black and grey-headed flying-foxes are also known to have used the site. The relocation efforts broadly followed advice from a bat expert (relocation proposal by Dr C. Tidemann included in Tanton 1999). However, this was a controversial decision, and other bat ecologists questioned whether it would be an effective long-term solution (West 2002). The relocation activities at MRR, using loud noise, commenced in 1999 and were repeated in subsequent years, on an as-needs basis. By early 2000, the area of disturbance needed to be expanded to include nearby residential areas (which flying-foxes had by then begun to use). Dispersals ceased after 2007, due to a Federal government requirement for a new environmental assessment after the local Clarence Valley Council became a joint applicant for approvals (with DET). An approval to continue relocation of flying-foxes from parts of Maclean between August and October until 2015 was in place at the time of finalising this thesis (2012).
89 5.3 Methods
5.3.1 Response of Maclean flying-foxes to relocation: survey methods
Data on flying-fox occupancy and abundance within roosts across the Lower Clarence region over the period of April 1999 to December 2007 were compiled from a survey of the literature (Eby et al. 1999; Tanton 1999; Tidemann 2002; West 2002; Tidemann 2003; Roberts 2006), and monthly roost site surveys conducted from September 2007 to December 2009 as part of a broader research project (Roberts unpublished data). Information relevant to the relocation of flying-foxes from MRR was obtained in 2006 from the three involved stakeholders (the NSW Department of Environment, Climate Change and Water (DECCW), DET, and the Clarence Valley Council) through applications made under the NSW Freedom of Information Act 1982 (FOI). I obtained additional information about the location of historically and currently used roosts in the Lower Clarence region, patterns of flying-fox occupancy and abundance, and details of the attempts to relocate flying-foxes from MRR from the following sources: field notes of biologists and naturalists (P. Eby, B. Roberts, M. Williams, J. Kennedy); records of interested, long-term residents (G. Bennett, C. West, P. Wrightson); and interviews with council staff (B. Sansom, N. Greenup and M. Forester) and persons living near MRR (J. Storock, J. Clowes, H. Naylor).
5.3.2 Determining financial costs and disturbance effort
Costs associated with the relocation attempts were obtained from involved stakeholders (DECCW, DET and the Clarence Valley Council) through Freedom of Information (FOI) requests to the NSW State government. Costs were allocated to one of several categories including consultant fees and wages, plans of management, logistics of the dispersal, research and acquisition of alternative habitat. Actual costs associated with some aspects of the disturbance were difficult to obtain and it is likely that some components have not been included in the total cost. The effort (person-hours) required to disperse flying-foxes from Maclean was summarised from information obtained under FOI, conversations with council staff (N. Greenup and M. Forester), authors personal observations and published articles (Tidemann 2002, 2003). Effort was calculated on a monthly basis, using the number of days
90 on which dispersal efforts were known to occur, multiplied by the number of people involved and the total disturbance time per day. 5.4 Results
5.4.1 Disturbance method
The standard method used to disturb flying-foxes at MRR consisted of three or four people working around the roosts perimeter to generate loud, continuous noise. At the time of the initial relocation in April 1999, noise was generated for 30 minutes at dawn and dusk (Tidemann 2002, 2003). Subsequent disturbances lasted for up to two hours per day (typically split into two periods: morning before 9 am and afternoon after 2 pm). The noise was generated using stock-whips, car horns, metal drums, gongs, starting pistols, firecrackers, whistles and small, unmuffled two-stroke motors such as chainsaw and lawnmower engines. These disturbances were observed to cause an immediate response from the flying-foxes, with the majority of the animals taking to the sky, vocalising and circling around the roost site for prolonged periods of time, ranging from 220 minutes. Typically, all flying-foxes left the MRR after 214 days of disturbance activity. The human effort required to remove the animals appeared to be positively related to the number of flying-foxes in the roost, and the length of time that flying-foxes had been allowed to persist at the site prior to being disturbed, although the data does not exist to assess this systematically. Numbers of flying-foxes present at the start of each disturbance period varied, but were typically between 1,000 and 20,000.
5.4.2 Disturbance of flying-foxes at the Maclean roost
During the period of licensed disturbances (April 1999 to December 2007) there were 23 separate documented attempts by flying-foxes to re-establish a roost at MRR (Figure 5.3). For the 12 months after the first disturbance, there were monthly re-occupation attempts by flying-foxes. From 2000 to 2007, attempts by flying-foxes to re-establish the roost commonly occurred in September/October, during the start of the birthing season. In general, when flying-foxes attempted to return to MRR their numbers built up to 1,0002,000 individuals over a few days. If further disturbances did not commence immediately, their numbers typically continued to increase rapidly.
91
12 0
Disturbance effort (person-hours)
10 0 8 0 6 0 4 0 2 0
0
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
Figure 5.3 Documented disturbance effort (person-hours) required to disperse flying-foxes from the Maclean Rainforest Reserve during the period of licensed disturbances (April 1999 to December 2007). Note that the data do not include any unauthorised disturbances conducted by residents of Maclean. Data from Tidemann (2003), Clarence Valley Council and authors.
After each disturbance, flying-foxes roosted in scattered groups in trees within the high school grounds and the immediate surrounds, and made regular attempts to return to MRR either overnight or once the noise had abated. In most cases, a large proportion of the colony had moved 350 m northeast from MRR into vegetation around a nearby electricity substation and extending into residents backyards (lower parts of the Maclean gully; Figure 5.2) (West 2002; Tidemann 2003; B. Roberts pers. obs.). Flying-foxes typically remained in this area for several months, although residents often harassed the animals in an attempt to induce them to move on (B. Roberts, pers. obs.). There were no observations of flying-foxes moving from the MRR to the proposed replacement roost site at Yaegl Nature Reserve nor was there any evidence of an immediate increase in the population of Yaegl at the time of any of the relocations. After 1999, the frequency of attempts by flying-foxes to re-establish a colony at the MRR progressively declined, although flying-foxes still returned to the site 10 years after the initial
92 relocation. Between 2007 and 2009, the bats roosted continuously in the Maclean gully despite frequent unauthorised attempts by local residents to move them. The population size was typically 2,0007,000, and occasionally reached more than 20,000, at which times the roost area expanded 550 m further up the Maclean gully, affecting additional residents (Figure 5.2). By 2009, flying-foxes were roosting in an area substantially larger than the predisturbance roost (i.e. MRR only). At maximum population size, flying-foxes roosted in MRR, both the upper and lower parts of the Maclean gully and spill over into areas around the Maclean High School (Figure 5.2).
5.4.3 Cost of the relocation
Relocation attempts at Maclean between April 1999 and December 2006 comprised over 640 person-hours of effort and cost a total of $396,000, including $300,000 towards the purchase of land for an alternative roost site (Table 5.1; Figure 5.3). The actual total cost of relocations was difficult to obtain due to the lack of records, the time that had elapsed since the initial relocation, and difficulties with estimating the cost of participation by government representatives. Other costs that have not been included in Table 5.1, but that would have significantly contributed to the total include: the costs of attendance (time, travel and accommodation) for government representatives at several years of community meetings; wages and administration costs for the various government bodies involved in regulating the relocation; the cost of vaccinating (against Lyssavirus) wildlife carers, veterinarians and government staff who monitored the welfare of the animals during the disturbance (a regulatory condition for the relocation attempt); and legal costs incurred when a conservation group (North Coast Environment Council) took the licence holder (DET) to court to prevent disturbances during the maternity season. Works also took place in the late 1990s to reduce the flying-fox impact on Maclean High School (including covered walkways, air-conditioning and double glazing windows). The cost of these was at least $360,000, although this is not a cost of the relocation but rather one of impact mitigation.
93
Table 5.1 Estimated costs of the relocation of flying-foxes from the Maclean Rainforest Reserve and the Melbourne Royal Botanical Gardens. Several additional components of the Maclean costs are not included due to lack of records (see text). Cost for the Melbourne Royal Botanical Gardens derived from S. Toop (pers. comm. 2006) and Department of Sustainability and Environment (2005).
Category Consultant fees and wages
Description Wages for main investigator, assistants and government staff that assisted with the dispersal e.g. Maclean, Tanton (1999) and Melbourne, Department of Sustainability and Environment (2005) Equipment hire or purchase, materials, vehicles, contract labour e.g. radio-tracking movements, mapping alternative roost sites. Cost to purchase alternative habitat (Yaegl Nature Reserve*) and/ or enhance alternative habitat
Maclean Rainforest Melbourne Royal Reserve Botanical Gardens $51,000 Between $100,000 and $200,000
Plans of Management
$20,000
$1,700,000
Logistics of the Dispersal Research projects
$25,000 nil
$250,000 $300,000
Alternative habitat
$300,000
$600,000 Between $2,950,000 - $3,050,000
$396,000
* The Yaegl Nature Reserve was purchased by NPWS in 2001 (with Commonwealth Government assistance) because of reports of increasing use by flying-foxes; and, due to the ecological significance of the dominant Melaleuca swamp forest (an endangered ecological community).
5.4.4 Assessment of flying-fox roost sites used since the relocation
After the initial disturbance of the Maclean flying-fox roost in 1999, at least 12 sites were used as roost sites by flying-foxes across the Lower Clarence region (Figure 5.1). Five had been used as roosts prior to the 1999 disturbance (Ulgundahi Is., Angourie, Yaegl Nature Reserve, Ashby and Lawrence) and seven appear to be new sites that were only used after the disturbance (Maclean gully, Whyna Is., Sleeper Is., Thorny Is., Bolorobo Is., Iluka, Warregah Is.). Six of these new roost sites (all except the Maclean gully) are situated in small mangrove patches or islands in which tree cover has only recently (last 15 years) developed or redeveloped to the extent where it would provide sufficient roost habitat for the establishment of a flying-fox roost (see Roberts 2005 for roost habitat descriptions). Five were temporary roosts used by flying-foxes for weeks or months and then abandoned. In 2004, a new roost was established within the Iluka township, 16 km from MRR, and this site was then continuously occupied by flying-foxes until 2010 (when the present paper was finalised). Use
94 of temporary roosts in the Lower Clarence largely ceased after the establishment of the Iluka roost. Since the Iluka site is close to residential areas, affected residents subsequently began lobbying governments to disperse the animals from this new location (Roberts 2006). 5.5 Discussion
5.5.1 Effect of disturbances on site use by flying-foxes
Has the relocation of the Maclean flying-fox roost been successful? The Maclean example has been termed a success by some researchers (Tidemann 2003; Nelson 2008a, b) and by residents at Maclean and elsewhere who argue in favour of relocating roosts. Flying-foxes have indeed failed to maintain a continuous presence in MRR since 1999. However, they have continued with attempts to re-occupy this historically used roost site (at times for prolonged periods in numbers exceeding 20,000) often prompting conflict with the local school community. Furthermore, flying-foxes are now roosting year-round only 350 m away in the Maclean gully and have also established a new roost in an urban setting 16 km away at Iluka, both of which have resulted in additional conflict with residents. That is, the relocation is unlikely to be considered a success by the broader community or government authorities charged with managing the conflict, who now have to deal with a new set of complaints from Iluka and Maclean residents, while managing the continued attempts by flying-foxes to resume their original Maclean roost. Seen in this light, the Maclean disturbance program, rather than resolving the problem, appears to have merely succeeded in moving the problem elsewhere at considerable and ongoing cost to the local community, and expanding it so that an increasing number of people are affected. Attempts to relocate flying-fox roosts using non-lethal methods have become frequent in recent years (Table 5.2). Many other relocation attempts have resulted in qualitatively similar outcomes to those observed in the present study of relocation at the Maclean roost. Some have succeeded in moving flying-foxes from their original roost site, however in most cases the effect has been temporary, and ongoing programs of dispersal have been required after the flying-foxes made regular attempts to return, while others have simply been unsuccessful in dispersing the bats (Table 5.2). Examination of 10 previous relocation attempts also demonstrates that dispersal actions are typically unsuccessful in the resolution of human-animal conflict. In 80% of cases, conflict
95 was still being reported either at the original site or within the broader community after the relocation had concluded (Table 5.2). There were four main outcomes of these relocations that limited their ability to resolve conflicts. First, initial relocation actions often only shifted flying-foxes into the immediately surrounding area. In eight of the 10 relocations, flyingfoxes continued to roost within 500 m of the original site after initial dispersal attempts (Table 5.2). In most cases, such distances were not sufficient to reduce conflict within the community. Second, animals did not simply join pre-existing roost sites after the relocation action and a number of new roost sites were created within the region. From the seven cases that monitored the emergence of new roosts, five of the relocations resulted in the creation of new permanently-used roosts nearby, and the number of new sites ranged between 1 and 7 sites (median 2.0 roosts, Table 5.2). These new sites typically proved to be at least as - or more - controversial than the initial location. Third, it was not possible to pre-determine the location of new roost sites that established after a relocation attempt. In the three relocation attempts where destination sites were pre-identified, flying-foxes did not permanently shift their roosting activities into these locations (Table 5.2). Lastly, the number of animals that remained at a site post-relocation was often equal to or greater than the original population size. In six of the 10 relocations, the maximum population size was unchanged or larger than before the relocation commenced (Table 5.2). More generally, flying-foxes are very mobile animals, and the availability of food resources in the local area is an important influence on patterns of abundance in flying-fox roosts (Eby 1991a; Parry-Jones and Augee 1992), therefore it is not surprising that disturbance actions have rarely had lasting long-term effects on how flying-foxes use roost habitat.
96
Table 5.2
Summary of known documented attempts to relocate Australian flying-fox roosts using non-lethal methods, during 1990 to 2009.
Location Batchelor, NT
Population No. of years No. of Dist. (m) Use of preAcceptable new No. of new Maximum Conflict Spp at initial Method# disturbance disturbance moved from identified site location? (duration roosts population in reported during Source+ * action 19902009 actions initial site (Distance) of action) created 2008-09 200809? B 200 25,000 40,000 28,000 20,000 BNS LNS HLNPOW V NS 2 3 9 2 6 10 4 2 3+ 10+ 2 23 9+ 7 <400 <500 200 300 350 <300 <500 none identified temporarily (400 m) none identified none identified no (3 km) none identified no (8 km) yes (3 months) yes (9 months) no no no no yes (6 months) 1 unknown but at least 2 unknown 2 7 unknown 2 2,000 nil 40,000 nil >20,000 >200,000 nil no at related site yes at related site yes yes no 1,2 3,4,5 6,7 8,9 10 11,12,13 14,15,16
Boyne Island, Qld BR Charters Towers, Qld BR
Dallis Park, NSW BG Maclean, NSW Mataranka, NT Royal Botanic Gardens, Melbourne, Vic Royal Botanic Gardens, Sydney, NSW Singleton, NSW Townsville, Qld BG R
BR >200,000 BHLNOSW G 30,000 NS
3,000
LNPOW
10
<900
none identified
no
20,000
yes
11,16,17 3,16,18, 19 6, 20
GR BR
500 35,000
LNUW BNS
2 3
3 2-5
max 900 400
none identified none identified
no no
0 unknown
14,000 20,000
yes yes
* G = grey-headed flying-fox; B = black flying-fox; R = little red flying-fox # B = birdfrite; H = helicopter; L = lights; N = noise; P = physical deterrent; O = odour; S = smoke; U = ultrasonic sound; V = vegetation destruction; W = water. + 1 Phillips et al. 2007, 2 J. McCarthy (Northern Territory Government, pers comm. 2010), 3 Roberts 2006, 4 Queensland Parks and Wildlife Service 2002, 5 J. Adair (Department of Environment and Resource Management, pers. comm. 2010), 6 S. Sullivan (Department of Environment and Resource Management, pers. comm. 2010), 7 Charters Towers Regional Council (pers. comm. 2010), 8 Welbergen 2005, 9 Roberts 2008, 10 This study, 11 Vardon et al. 1997, 12 Carol Palmer (Northern Territory Government, pers. comm. 2010), 13 Eddie Webber (Northern Territory Government, pers. comm. 2010), 14 Tanton 1999, 15 Toop 2004, 16 van der Ree and North 2009, 17 Richards 2002, 18 Singleton City Council 2008, 19 Stevenson 2004, 20 Billabong Sanctuary (pers. comm. 2010).
97 For example, the dispersal of a roost from the Melbourne Royal Botanical Gardens eventually resulted in flying-foxes establishing two new roosts in unexpected locations (Yarra Bend Park and Geelong, 5 and 65 km respectively from Melbourne), rather than at a target site (Horseshoe Bend, 8 km away) identified in the relocation plan (Toop 2004; Department of Sustainability and Environment 2005). Flying-foxes returned almost monthly during the first six months of disturbances at Melbourne, however between 2004 and 2009 flying-foxes made only one attempt to return (R. van der Ree pers. comm. 2010). There have been three separate attempts to move roosting flying-foxes from Burdekin Park, Singleton (NSW) using spotlights and reflective material, water from fire hoses and sprinkler systems, and loud noise, with no success (Roberts 2006; Fletcher 2010). At Dallis Park (Murwillumbah, NSW) the habitat of a roost site was destroyed in 2004 to disperse and prevent re-establishment attempts by flying-foxes. The Murwillumbah area has been extensively cleared for agriculture and the flying-foxes utilised the nearest available patch of dense tall forest. Once the vegetation at Dallis Park had regrown to a suitable height three years later, the flying-foxes attempted to reestablish the original roost (Roberts 2008). At present, knowledge of the movement patterns of flying-foxes and the factors influencing the establishment and persistence of their roosts is insufficient to accurately predict where flying-foxes will move once relocated from a particular roost. For example, prior to disturbances of the grey-headed and black flying-foxes that roosted at MRR, it was suggested that they could be shifted to nearby Yaegl Nature Reserve (Tanton 1999; Tidemann 2002, 2003). However, this did not occur. Instead, Yaegl has been primarily used for short periods of time during the autumn and winter months by nomadic groups of little red flying-foxes. Relocations also have the potential to shift flying-fox roosts to nearby, possibly more controversial sites. In eastern Australia, flying-fox roosts occur in a variety of habitats from continuous forest to small remnant forest patches (Eby 2002b; Roberts 2005), but there is emerging evidence that there is a tendency for roosts to be situated in urban environments (Birt et al. 1998; Hall 2002; Roberts 2005; Chapter 4). Therefore, further relocation attempts in Maclean or Iluka may result in a shift to other urban areas in the region.
5.5.2 Cost-effectiveness of relocation attempts
An additional factor that requires consideration when assessing the success of a relocation attempt is the cost of dispersal. Cost is relevant because in most situations there may be a
98 range of alternative management actions to reduce conflict other than dispersal, such as subsidising double-glazing of windows and the air-conditioning of rooms to reduce impacts of noise and smell (see Roberts 2006). In some situations it may be possible to manage roost vegetation to encourage flying-foxes to roost further from areas of human activity (Coffs Harbour City Council 2007). Unlike dispersal, these mitigation measures have a relatively certain outcome. The issues of alternative approaches to the problem, their costs, and their social acceptability can be very complex. However, to date neither the alternatives to dispersal nor the long-term activities required for relocation have been fully costed, either at Maclean or elsewhere. The present paper is the first time where some attempt has been made to quantify the longterm cost of dispersing flying-foxes from their roost sites. The cost of relocating flying-foxes from Maclean so far has exceeded $400,000 by an unknown quantity (and still counting, as efforts are planned to continue) (Table 5.1). By comparison, Singleton City Council has spent approximately $117,000 on attempts to relocate flying-foxes from Burdekin Park, and estimated that another $320,000 over a three-year period would be needed (A. Fletcher pers. comm. 2006; Fletcher 2010). In Melbourne, thousands of person-hours of effort and approximately $3 million were needed (including associated research and purchase of additional habitat) (Table 5.1). The benefits of the Melbourne relocation in reducing conflict with the general community and protecting heritage trees could perhaps be considered to outweigh the financial cost. However, these resources are beyond the means of most small rural and regional communities.
5.5.3 Managing flying-fox relocations in the future
Relocation continues to be viewed as an attractive solution to problems arising from flyingfox roosts in urban areas. For example, between 2006 and 2009, proposals were made to State and/ or Commonwealth government to relocate eight flying-fox roosts in NSW, Queensland and the Northern Territory. However, it is important to determine the magnitude of the perceived problem before exploring potential management options, including relocation. For example, if noise, smell and faeces from a roost affect only a small number of residents, then more local-scale mitigation options such as creating buffers between houses and roosting flying-foxes or constructing sound barriers may be more effective solutions than attempted wholesale relocation of a roost (see Roberts 2006 for review of further management options and their estimated costs).
99 In many cases, public education campaigns can reduce antipathy towards flying-foxes and reduce the social or political imperative to do something about flying-fox roosts. For example, managers of some urban roosts (e.g. Bellingen, Coffs Harbour, Wingham Brush and Ku-ring-gai (Gordon) in NSW, and Woodend in Ipswich, Queensland), have acted to alleviate the concerns of nearby residents through strategies such as community-based roost revegetation programs, coupled with minor habitat modification around the roosts periphery, education days, and the promotion of tourism to roost sites (Pallin 2000; Smith 2002; Coffs Harbour City Council 2007; Hall 2006). Similar approaches have been used to successfully manage residents concerns around six flying-fox roosts in suburban Brisbane, Queensland, that were considered potential sources of major conflict (Hall 2002, 2006). Many of the conflicts between humans and flying-fox roosts may be attributed to poor planning and inappropriate development near established roost sites (West 2002; Smith 2002; Eby 2002b). Creating public open space buffers around established roost sites, aligned with more sympathetic developments, could minimise future conflict, particularly in new residential areas. This is mainly an issue for local government, although there may also be a role for State and/ or Commonwealth planning policies to guide development of areas adjoining flying-fox habitat, given that some flying-foxes species are classified as vulnerable to extinction under State and/or Commonwealth legislation. In cases where relocation is considered a preferred management option, the objectives of relocation and of what might constitute success need to be more clearly defined. In particular, the extent of responsibility of the proponent undertaking the relocation to the broader community (e.g. ensuring that any replacement roost is not a source of conflict) needs to be explicitly identified. The length of commitment to relocation also needs to be clearly understood by proponents, given that flying-foxes show high fidelity to traditionally-used roost sites (Ratcliffe 1931; Nelson 1965; Eby 1995; Richards 1995; Tidemann 1999; Tidemann et al. 1999). The continued attempts by flying-foxes to re-establish the Maclean roost may be related to the role of the site as a maternity roost. As flying-foxes can live for more than 15 years in the wild (Martin and McIlwee 2002; Divljan et al. 2006), attempts to re-establish the MRR roost may continue for another few years (if sites are occupied on the basis of individual memory), or indefinitely (if sites are occupied on the basis of habitat attributes or cultural transmission). Such factors need to be considered and addressed in decisions to disperse or relocate flying-fox roosts.
100 Future relocation attempts also need to be accompanied by an adequate monitoring program, to record the actions taken and their costs, and also to determine the short- and long-term outcomes of the disturbance. Monitoring of the outcomes could include both tracking the individual movements of affected animals (for example, with satellite- or radio-telemetry) over the first 12 months, and regularly monitoring of both the original site (i.e. species present, their abundance, breeding status) and other sites in the region. Without such monitoring, there is a significant risk that attempts at relocation will continue to be represented by proponents as successful, when in fact they have simply shifted the problem to other places or to the future, rather than solved it. In NSW, there is a flying-fox management policy to guide those wishing to relocate problematic roost sites (Department of Environment and Climate Change 2007). Through that policy there is potential to collect documentation to evaluate the success of relocations and improve future management actions. 5.4 Conclusion
The resolution of conflicts between humans and flying-foxes is important to the conservation and management of flying-foxes in Australia. The use of disturbance to induce roost relocation is currently commonly proposed as a management tool to reduce conflicts between humans and flying-foxes. However, such relocation attempts have largely been carried out in an ad hoc fashion and have lacked systematic documentation, costing and monitoring. Further, most relocations have had limited success in moving the flying-foxes to new sites, in some cases these new sites have been in unanticipated and undesirable locations, and relocation attempts may be costly. The location of flying-fox roosts in urban areas is likely to continue to be an issue of community conflict and conservation concern in the future. A better understanding of flying-fox relocations will significantly assist organisations responsible for managing flying-fox roosts and help identify long-term management solutions that are both ecologically sound and acceptable to the entire community. The results for flying-fox relocations also match the problems that have previously been raised in the literature for other relocated species, and for relocation actions in general. That is, relocations that aimed to solve human-animal conflict generally failed (Fischer and Lindenmayer 2000; Massei et al. 2010). The main problems frequently limiting the success of these human-animal conflict relocations were extensive movements of relocated animals and homing behaviour (Fischer and Lindenmayer 2000; Massei et al. 2010). In addition, if the
101 animals did shift locations, some animals created conflict at new sites that they occupied
(Massei
et al. 2010). Very few studies have reported the costs of relocations, established
criteria for judging the success or failure of their actions or conducted adequate monitoring after the relocation (Fischer and Lindenmayer 2000; Massei et al. 2010). Before considering the relocation of problem animals, wildlife agencies and the general public should be aware that: (1) there is little evidence that group relocations result in the resolution of the conflict; and (2) the cost of the relocation program can be high and might require sustained funding to prevent the reoccurrence of the problem.
103
Chapter 6 General Discussion

In this chapter, I outline the main findings of the previous chapters of the thesis. I discuss how the work has contributed to knowledge of flying-fox ecology and the implications of the findings for conservation and management. I then discuss a series of questions that arise from the work, particularly questions in relation to patterns of roost use and mobility. Finally, I conclude this chapter with suggestions for the direction of further research. 6.1 Key findings of each data chapter
Historical shifts in the ranges of two Australian flying-fox species and their relationship with climate change. In Chapter 2, a quantitative analysis of historical and present distributions of two species of flying-fox (Pteropus alecto and P. poliocephalus) in eastern Australia was undertaken. The results were used to evaluate evidence for (i) distributional shifts in these species and (ii) global warming being a driver of any such shifts. Recent range shifts towards higher latitudes have been reported for many animals and plants in the northern hemisphere and are commonly attributed to changes in climate (Walther et al. 2002; Parmesan and Yohe 2003; Root et al. 2003). Relatively little is known about such changes in the southern hemisphere, although it has been suggested that latitudinal distributions of the flying-foxes P. alecto and P. poliocephalus have extended polewards during the twentieth century in response to climate change in eastern Australia (Tidemann 1999; Hughes 2003). However, historical changes in these species distributions have not been examined systematically. In this study, I obtained historical locality records from a wide range of sources (including banding and museum records, government wildlife databases and unpublished records), and filtered them for reliability and spatial accuracy. The latitudinal distribution of each species was compared between eight time-periods (18431920, 19211950, five 10-year intervals between 1950 and 2000, and 20012007), using analyses of both point data (P. alecto 870 records, P. poliocephalus 2,506) and presence/absence data within 50 x 50 km grid cells. The results did not support the hypothesis that either species range is shifting in a manner driven primarily by climate change. In the case of P. poliocephalus, neither its northern nor southern range limits (Mackay, Queensland and Melbourne, Victoria respectively) changed
104 over time. In contrast, there was good evidence that P. alectos range extended southward by over 1,100 km (approximately 10.5 degrees latitude) during the twentieth century (from near Rockhampton, Queensland to Sydney, New South Wales). Within this zone of southward expansion (25-29oS), records of P. alecto increased from 8% of flying-fox records prior to 1950 to 49% in the early 2000s; and local count data showed that its abundance increased from several hundred to >10,000 individuals at specific roost sites as range expansion progressed. However, the rate at which its range has expanded southward (an average of 100 km/decade) is much faster than the rate of isotherm change over the same period (10-26 km/decade). Consequently, the species expanded its range southward into colder regions than it had previously occupied. The more general implication of this analysis is that climate change should not be uncritically inferred as a primary driver of species range shifts without careful quantitative analyses. Frequency and scale of movement among roosts and its implications for management. Chapter 3 investigated four main questions in relation to the inter-roost movements of P. poliocephalus: (1) What patterns of movement (in terms of distances and directions) are exhibited over increasing time scales? (2) How much do these patterns vary among individuals? (3) What is the relationship between individuals movement patterns and their use of different roost sites? (4) What are the implications of these findings for management? Fourteen adult males were tracked for 240 weeks (mean 25 weeks). Collectively, these individuals utilised 77 roost sites in an area spanning 1,075 km by 128 km. Movement patterns varied greatly between individuals, with some travelling long distances. Five individuals travelled cumulative distances >1,000 km over the study period. Five individuals showed net displacements >300 km during one month, including one movement of 500 km within 48 hours. Seasonal movements were consistent with facultative latitudinal migration in part of the population. Flying-foxes shifted roost sites frequently: 64% of roost visits lasted <5 consecutive days, although some individuals remained at one roost for several months. Modal 2-day distances between consecutive roosts were 2150 km (mean 45 km, range 3166 km). Of 13 individuals tracked for >12 weeks, 10 moved >100 km in one or more weeks. Median cumulative displacement distances over 1, 10 and 30 weeks were 0 km, 260 km and 821 km, respectively. On average, over increasing time-periods, one additional roost site was visited for each additional 100 km travelled.
105 The high mobility and frequent movements of these flying-foxes mean that the currentlyadvocated management approaches of culling animals and relocating them from roost sites have a high risk of failure. These methods are unlikely to permanently reduce local abundance of flying-foxes, because individuals have a high capacity to recolonise roosts in the short to medium term. Alternatives to local control measures, e.g. wholesale culling of flying-fox populations, run contrary to conservation objectives. More generally, the results of this chapter show that mitigation of conflict between humans and wide-ranging species such as P. poliocephalus is extremely complex and sound management of such conflicts must be built on a good understanding of species movement patterns. Feeding movements and selection of feeding and roosting habitats. Chapter 4 investigated four main questions in relation to the feeding movement and habitat selection of P. poliocephalus: (1) What are the landscape-scale habitat characteristics of roosts and feeding sites and how do they differ? (2) Do flying-foxes select roosts and feeding sites on the basis of landscape-scale habitat characteristics? (3) What distances are travelled between roosts and feeding sites? (4) What are the roles of specific factors such as roost population size, duration of visits to feeding sites, size of forest patches, time of year, elevation and land cover in determining the feeding distance moved? Understanding how flying vertebrates move across the landscape to feed and how they respond to different patterns of habitat alteration enables managers to conserve important locations, and identify areas that meet species annual life cycle needs (Price et al. 1999; Berthold 2001; Fleming and Eby 2003). Satellite telemetry of 14 adult male P. poliocephalus utilising 45 roost sites and 173 feeding sites were used to quantify their feeding movement and landscape use. The study found that P. poliocephalus selected roost sites in landscapes containing a higher proportion of built and open habitats than expected from their occurrence in the environment. By contrast, they were less selective of feeding landscapes, although showed some avoidance of built areas for feeding. The distances moved from roost to feeding sites were typically (85% of feeding movements) less than 20 km. However, there was a large amount of variation in the feeding distance moved, both within and between individuals (range 256 km; mean 12.3, SD 9.3 km). The
106 visit duration spent at feeding sites was typically short: 50% of feeding visits lasted <5 days, although some individuals remained at feeding sites for up to 84 consecutive days (range 184; mean 12 days; SD 14 days). Of the variables assessed in this study only the number of consecutive days during which individuals visited a particular feeding site was significantly correlated with feeding distances. Animals that remained at feeding sites for longer periods fed closer to roosts than individuals with short visit durations. Protection of essential habitat for P. poliocephalus is a primary conservation goal for this threatened species (DECC 2009). However, its habitat requirements are complex, and this complexity has been largely attributed to variability in the distribution and flowering and fruiting characteristics of diet species (e.g. Eby 1991a; Eby and Law 2008). The importance of conserving paired feeding and roosting habitats has also been acknowledged (DECC 2009). This work adds to the discussion of habitat requirements, and in particular the need to conserve associated feeding and roosting sites that vary in landscape characteristics and which may be separated by several kilometres. In addition, the finding that P. poliocephalus select roosting habitats near built and cleared areas emphasises the importance of developing improved methods for managing controversial roosts as part of conservation strategies for this species. Roost relocation efforts and their success as a management strategy. Chapter 5 examined the consequences and financial costs of a coordinated, governmentsponsored attempt to relocate a flying-fox roost at Maclean in north-east New South Wales (NSW). The Chapter also reviewed the occurrence and outcomes of other relocation attempts in Australia, and discussed the utility of relocation as a management tool to resolve conflict between humans and flying-foxes. Managing flying-fox roost sites in urban areas is an increasing challenge for government agencies along the east coast of Australia (Birt et al. 1998; Hall and Richards 2000; Eby and Lunney 2002b). In the township of Maclean, flying-foxes had used a roost as a maternity site for over 100 years. Due to a range of community concerns, flying-foxes were repeatedly induced to move from the Maclean roost during the period 19992007, by subjecting animals using the roost to periodic episodes of continuous loud noise.
107 Records compiled in this chapter show that the total cost of the Maclean relocation attempt was > $400,000 (if include cost of alternative land purchase), including 640 person-hours of effort. Flying-foxes made 23 attempts to return to the original roost location during the nine years of disturbance, although the frequency of attempts declined over time. Twelve other sites were used during this time as temporary roosts, including seven sites not previously occupied. In 2004, flying-foxes established a new continuously-occupied roost in the township of Iluka, 16 km north east of Maclean. Residents near the Iluka roost were by 2005 intensively lobbying governments to disperse the animals from this new location. Both the Iluka and Maclean roost sites were still in use in 2011. The outcome after nearly a decade of dispersal attempts at Maclean was that flying-foxes continued to return periodically to the original site, and there were more roost sites established in the region, over a wider area than previously known from historical records, and the number of affected residents experiencing conflict had increased. The location of flying-fox roosts near human settlements is likely to continue to be an issue of community conflict and conservation concern in the future. The experience at Maclean raises questions of how, and at what spatial and temporal scales, the success of relocation attempts should be determined. 6.2 Patterns of roost usage and mobility by flying-foxes: questions arising
from this work Why does Pteropus poliocephalus move frequently among roosts? The results of research conducted for this thesis, together with previously-published work by other researchers, show that the movement patterns of P. poliocephalus are far more complex and variable then that typically seen in north-south migrants in the northern hemisphere. Individual P. poliocephalus are highly mobile and exhibit a range of short- and long-distance movements over a range of temporal scales (Eby 1991a; Tidemann and Nelson 2004; Chapter 3). For example, it is common for some P. poliocephalus to regularly move among a number of roosts within a region or from one region to another over distances of several hundred kilometres (Eby 1996; Spencer et al. 1999; Tidemann and Nelson 2004; Chapter 3). Previous authors have argued that food availability is a major driver of the movement patterns of P. poliocephalus, as well as influencing the number of individuals within particular roosts at any given time (Eby 1991a; Parry-Jones and Augee 1992; Eby 1996). However, a clear
108 understanding of how the movement patterns of flying-foxes relates to the availability of food, particularly within a region, is still needed. If the drivers of flying-fox movement could be better understood, ecologists would be able to better predict patterns of roost occupation and inter-roost movements, contributing greatly to the management of flying-foxes in contentious locations. The available evidence on drivers of flying-fox movement patterns is summarised as follows. McWilliam (1986) radio-tracked three species of flying-fox (P. poliocephalus, P. alecto and P. scapulatus) in subtropical eastern Australia. The results suggest that the movements between roost sites at a local scale were correlated with shifts in available food resources, and some individuals of P. alecto moved between neighbouring roost sites in order to reduce commuting distances to new food sources. Eby (1996) examined temporal and spatial patterns in resource availability for P. poliocephalus, related these trends to patterns of animal abundance and movement, and concluded that the movement among roosts of migratory individuals was closely associated with patterns of food availability. Abandonment of feeding areas could not consistently be attributed to resource depletion, but migratory individuals generally shifted roosts to areas of mass eucalyptus flowering (Eby 1991a, Eby 1996). Markus (2000) found that some radio-tracked individuals of P. alecto in urban Brisbane regularly moved between two or three roost sites and that each shift was to a roost closer to the current feeding area. Parry-Jones and Augee (1992) found that the patterns of occupancy of a P. poliocephalus roost in the eastern subtropics were correlated with fluctuations in food supply, and that local abundances of blossom of tree species such as Angophora floribunda and Corymbia maculata (syn. Eucalyptus maculata) were associated with large colony numbers of >80,000 flying-foxes. Not all individual flying-foxes move regularly between roosts. Some show a high degree of fidelity to one roost, a behaviour which may also be associated with the availability of food. Both Eby (1996) and Markus (2000) found that both P. poliocephalus and P. alecto had a proportion of the population that were resident individuals which rarely shifted roost sites. They argued that this was made possible because the bats were roosting in locations where a high diversity of food plants provided a continuous food source, in particular, areas whose food supply included the availability of fruits such as Ficus. By using satellite telemetry, this study was able to provide additional information on the frequency of movements among roosts by migratory P. poliocephalus (Chapter 3). Capture
109 sites targeted in this study were those that were either seasonal roosts or had a high number of flying-foxes present to ensure that most of the individuals present were likely to be migratory (see Chapter 3 and 4, Methods). No study animals displayed the roosting traits of residents in that none remained at a single roost for longer than 133 consecutive days. The study showed that at times P. poliocephalus moved frequently between roost sites across a broad landscape, but there was also a high amount of exchange between roosts in a local region. Individuals typically spent less than five days at roosts, although there were some extended stays at one site followed by frequent short-stays at a number of roosts. All long-distance movements (net displacement >300 km in one month) were to areas of intensive flowering of known food tree species (all Myrtaceae): Eucalyptus tereticornis, E. siderophloia, E. pilularis, Corymbia citriodora, C. intermedia or Melaleuca quinquenervia. In Chapter 3, I also argue that short stays at closely positioned roosts occurred in local areas that provided continuous food. The duration of visits at feeding sites was typically short; half the visits were for around five days (Chapter 4), consistent with the short duration stays at roosts (Chapter 3). In some cases, individuals spent extended periods at one roost and used the same roost to access several feeding sites. Although not quantitatively analysed, inspection of the data suggest that only on rare occasions would individuals use different roosts to access the same feeding site, and these movements were usually associated with a reduction in feeding distances. This observation together with the significant negative correlation between individuals feeding distance from a roost site and the duration of its stay at feeding sites (Chapter 4) suggests that these flying-foxes selected roost sites as a consequence of their need to move among suitable feeding areas, while minimising flight distances from each roost. The energetic costs of flight constrain feeding distances (e.g. Eby 1991a; Eby 1996). In most locations, the food resources available to flying-foxes within nominal feeding distances of roosts fluctuate with time, and flying-foxes must move between roosts to access new food locations (see also McWilliam 1986; Markus and Hall 2004). The capacity to move long-distances over a period of days allows the species to utilise the ephemeral flower and fruit resources characteristic of Australia, across broad spatial scales (Eby 1991a; Eby 1996; Chapter 3). However, the mechanisms used by P. poliocephalus to track food over large areas or the proximal cues for flying-fox movement and causes of variation between individuals remain poorly understood.
110 6.3 Ecology and management of flying-foxes in human dominated
environments What is the behaviour of Pteropus poliocephalus in relation to built environments? In this discussion the term built environment is used with reference to human-made structures that provide the setting for human activities, particularly settlements, ranging in scale from rural residential (scattered houses on large blocks), villages (clusters of 100999 people), towns (clusters of 10009999 people) to urban areas (clusters of >10,000 people) (Source: Australia Bureau of Statics 2006). In eastern Australia, flying-foxes commonly roost in proximity to such built environments. Reports have suggested that, in recent times, there has been: (1) an increase in the number of roosts in villages, towns and urban areas (Birt et al. 1998; Hall 2002; Markus and Hall 2004; Roberts 2005); (2) an increase in the number of individuals within urban roosts, particularly in the southern part of the range of P. poliocephalus (van der Ree et al. 2006); and (3) an increase in the frequency of use of urban roosts, e.g. seasonally used roosts becoming continuously used (Birt et al. 1998; van der Ree et al. 2006; Williams et al. 2006). The presence of flying-fox roosts in or on the edge of built environments has been attributed both to the encroachment of human development on traditional roost sites (Birt et al. 1998; Hall 2002) and to a spatial shift of flying-fox populations into human settlements (Birt et al. 1998; Hall 2002; Parris and Hazell 2005; van der Ree et al. 2006; Williams et al. 2006). Roosts in close proximity to human settlements often generate conflict, with local communities becoming concerned about issues of noise, smell and disease transmission from bats to livestock and to humans (e.g. Hall and Richards 2000; Chapters 1 and 5). The increased presence of flying-foxes within human settlements also leads to a presumption that flying-fox populations are increasing, and hence that management actions should be undertaken to control their numbers, either by culling or by disturbance to roost sites (e.g. Eby and Lunney 2002a; Chapter 3). A better understanding of how and why flying-foxes use built environments is required to help resolve conflicts that occur between flying-foxes and humans in these locations. Explanations for increases in the occurrence of flying-foxes in urban areas have generally cited recent increases in the availability and quality of food in these locations, combined with habitat loss elsewhere in the species range. For example, a study of P. poliocephalus by
111 Parry-Jones and Augee (2001) stated that the introduction of fruiting and flowering plants (native and exotic) to the Sydney region has provided an increased variety of foods for flyingfoxes, permitting the continuous occupation of urban roosts. van der Ree et al. (2006) showed a significant increase in the total number of P. poliocephalus in urban Melbourne since the 1980s and a change from seasonal to permanent occupancy of a roost. The authors suggested that increased food availability within the urban landscape and loss of habitat elsewhere contributed, together with climate change, to the permanent establishment and growth of this colony. Similarly, Parris and Hazell (2005) argued that a combination of an increase in temperatures in Melbourne (primarily as a result of an urban heat island effect) and an increase in the supply of food (due to the establishment of non-indigenous plant species) had created more favourable conditions for flying-foxes. In addition, Williams et al. (2006) demonstrated that plantings of favoured food plants in Melbourne have increased the temporal availability of food for P. poliocephalus in that city: prior to European settlement, only 13 species recorded in the range-wide diet of P. poliocephalus grew in the Melbourne area, but by 2006, an additional 87 diet species and over 300,000 individual trees had been planted on Melbournes streets. Increases in the number of roosts recorded in other Australian cities (e.g. Brisbane: Markus and Hall 2004; Roberts 2005) have also been attributed to an increase in food supply (Markus and Hall 2004). However, it may be overly simplistic to conclude that: (1) changes in habitat usage by flyingfoxes have only occurred in urban locations; or (2) the overall population of flying-foxes has uniformly increased in urban areas. First, there is some evidence that roosts outside urban areas have also exhibited an increased frequency of occupation. A shift from seasonal to continuous use has been reported for roosts at Susan Island and Bellingen, both in small rural towns in north-east New South Wales (Lunney and Moon 1997; Smith 2002; J. Thomas pers. comm.). Second, at least two long-established urban roosts have declined in population size concurrently with an increase in the number of local sites: Indooroopilly Island in Brisbane (Hall 2002) and Gordon in Sydney (M. Beck, Ku-ring-gai Bat Conservation Society unpublished). Hall and Richards (2000) stated that various large roost sites (>200,000) observed in south-east Queensland in the 1970s no longer existed by the 1990s, having instead been replaced by a number of smaller roosts in urban locations. They concluded that this change from a few large to many smaller roosts has led to the misconception that flyingfoxes have actually increased in numbers in recent times. However, long-term data on the abundance of P. poliocephalus are limited at all but a few roost sites (Chapter 2). The lack of
112 long-term monitoring, and the naturally high temporal variability in roost populations associated with the high mobility of P. poliocephalus (Eby 1991a; Chapter 3), make it difficult to verify long-term change in abundance at any roost, or across particular roost types (e.g. urban or rural). The possible factors driving changes in the occupancy of roosts in rural residential areas, villages and towns are less clear than those in urban landscapes as the limited areas associated with these settlements provide much less scope for human structures to influence ambient temperature or for the abundance of diet plants to be substantially enhanced due to gardens or streetscapes plantings. Preferences for roosts within large cities, as well as within villages and towns, may perhaps be explained by factors such as protection from shooting and roost disturbances or assistance with navigation from illuminated features (Birt et al.1998; Hall and Richards 2000). Alternatively, the reduced likelihood of some predators (e.g. raptors, varanids) in built areas may explain the prevalence of flying-fox roosts. It is also possible that correlations between selection factors for human settlements and flying-fox roosts, i.e. level land near waterways, contribute to the selection of roosts in residents and urban areas (Roberts 2005). Although flying-foxes use many roost sites in urban areas, they also continue to use a large number of roosts outside urban locations. This study examined roost and feeding preferences across all landscapes, including a range of built environments. Of the 45 roosts used by study animals, 40% were in remote areas where no built land occurred within 2 km of the site and 50% were in areas where <5% of the surrounds contained built land (rural residential or villages). By contrast 20% were located in substantially built environments where >35% of the surrounds contained built land (towns) and 4% were embedded in urban landscapes (Chapter 4). It is important to distinguish between the following two processes: (1) an increased use overtime of urban areas for either roosting or feeding; and (2) a preference for feeding or roosting in built environments. Results of the present study weakly support the notion that P. poliocephalus selects roosts near built areas. However, this studys data also show that P. poliocephalus avoids feeding in built areas. It remains plausible that flying-foxes use of these built areas for feeding has increased in recent decades. But the built environment remains the less-preferred feeding habitat in spatial terms. The mobility of flying-foxes allows them to roost within human settlements and traverse open or modified areas to find trees that bear nectar and fruit (Law and Dickman 1998; McDonaldChapter 6 General Discussion
113 Madden et al. 2005). Human settlements may be providing alternative benefits to flying-foxes associated with the local characteristics of roost sites. However, further work is required to understand both this trend and the specific local on-ground characteristics by which P. poliocephalus select roost sites across their geographical range. Further quantification of feeding resources and their availability within both built areas, and open areas, compared with forests is also needed. What leads to the success or failure of roost relocation attempts? In Australia, the relocation of flying-fox roost sites has been frequently attempted to resolve humanbat conflict in urban and rural areas. Conflict arises when bats establish roosts close to human activities or when human development encroaches on established roost sites (Birt et al. 1998; Hall and Richards 2000; Chapter 5). Roost relocations refer to the intentional movement of entire colonies from one location to another, although exactly where flyingfoxes may move to when relocated cannot be pre-determined. A number of different techniques have been used to harass flying-foxes until they shift elsewhere, including continuous loud noise, bird-scare guns, helicopters and light aircraft, spraying with water, smoke and habitat destruction (Vardon et al. 1997; Tanton 1999; Tidemann 2002; Chapter 5). Similar techniques are also used at flying-fox roosts worldwide to manage conflicts (Mickleburgh et al. 1992). However, few relocation activities have been systematically monitored and the success or otherwise of such projects have been debated (Hall 2002; Tidemann 2002; West 2002). A better understanding of the costs involved and long-term outcomes of these activities is required. The few studies that have examined relocations have all been conducted within Australia, however these studies have only monitored short-term outcomes. They are summarised as follows. Vardon et al. (1997) reported on the success of a relocation attempt of more than 200,000 P. scapulatus from a roost at Mataranka, in northern Australia, during 1994-1995. Several methods were used including bird-scare guns and cattle crackers fired from shotguns, harassment by light aircraft (up to 12 passes per day), water jets and smoke. All methods were unsuccessful in moving flying-foxes more than 300 m from their original location and for periods of more than two days. Tidemann (2002, 2003) provided information on the dispersal of a P. poliocephalus roost in Maclean, northern NSW during 19992002. The study found that during this time there were 16 reoccupation attempts by flying-foxes to the original site, but the frequency of these attempts declined with time. However, the author noted that subChapter 6 General Discussion
114 optimal monitoring of dispersal actions made it impossible to determine where animals moved to and additional monitoring over longer time periods was required to improve the success of relocation attempts. A review of the same dispersal attempt at Maclean by West (2002) stated that animals did not move to predetermined locations as anticipated and that financial costs of the operation were high and ongoing. Phillips et al. (2007) reported on the relocation of a small roost of 200 P. alecto in Batchelor, North Territory over five months in 2004. Flying-foxes were dispersed using a combination of smoke, non-lethal plastic shot and noise. Initial actions fragmented the population to new locations within the township and follow-up dispersal actions at both the original sites and the new roost locations were required. After five months the flying-foxes had shifted approximately 400 m to the towns periphery but continued efforts were required to ensure that they did not re-establish in town. The author argued that the best longer-term solution to stop flying-foxes from trying to reestablish in town would be to remove the roost vegetation. Collectively, these studies indicate that severe roost disturbance may induce flying-fox roosts to shift short distances for short periods of time, but the longer-term success of these actions has often been limited, when flying-foxes return to their original roost sites. In the Melbourne Royal Botanical Gardens after six months of disturbance action the flyingfoxes left and began roosting at two separate locations, Yarra Bend Park and Geelong, 5 and 65 km respectively from Melbourne, but did not move to the target site identified in the relocation plan (Toop 2004; Department of Sustainability and Environment 2005; Chapter 5). Following negotiations with the land managers at these two new locations, the flying-foxes presence there was accepted. Since then, only one recorded attempt has been made by flyingfoxes up to the time of writing (2012) to re-establish the original site (Chapter 5). The action required thousands of person-hours of effort and approximately $3 million including associated research and purchase of additional habitat at one of the new sites (Department of Sustainability and Environment 2005; Chapter 5). To be consistent with the goals of minimising human-bat conflict, and with conservation and animal welfare objectives, a definition for success of relocations should include at least the following: (1) that conflict has been reduced within the broader community (not just the original site); and (2) there was minimal impact on the flying-foxes, in terms of animal welfare, population mortality or lowering reproduction. In Chapter 5, I examined the outcomes of 10 separate relocation attempts in Australia, during 19902009. This is the first
115 review of the long-term outcomes and costs of roost relocation attempts. The study found that relocations, regardless of the methods used, have had limited success in permanently resolving conflicts between humans and flying-foxes, and were costly. At Melbourne the relocation reduced the intense, continual conflict with the general community, however at the two new roost sites periodic noise disturbances were required to alter the extent of the roost to prevent the flying-foxes encroaching concerned neighbours (Department of Sustainability and Environment 2011a). In terms of welfare, Yarra Bend Park did not provide the same protection from high temperatures as was provided by Melbourne Botanical Gardens and thousands of P. poliocephalus died due to extreme heat events during 2004-2010 (Department of Sustainability and Environment 2011a and b). Given the high mobility of flying-foxes: a high frequency of movements between roosts, the large number of roosts used by individuals, the rate of return to specific roost sites and a capacity for long-distance movement (Eby 1991a; Eby 1996; Tidemann and Nelson 2004; Chapter 3), it seems very unlikely that relocation attempts could have a high probability of success in resolving human-bat conflicts over the long term. In terms of animal welfare, little is known about the impact of relocations on flying-foxes, and a better understanding is required. Garnett et al. (1998) reported that roost dispersals can result in high mortality, particular if conducted during the breeding season as it can result in the death of dependent juveniles. As past relocation attempts have failed to provide long-term solutions and have at times worsened conflict between humans and flying-foxes (e.g. by shifting flying-fox roosts to equally controversial areas), there is now an imperative to find new approaches to resolving such conflicts that are based on a sound understanding of the ecology of flying-foxes. Such approaches, and any future attempts at relocation, should be accompanied by an adequate long-term monitoring program, to allow evaluation of the success of the action. The success of relocation attempts or other approaches to managing human-flying-fox conflicts need to be considered over longer time frames than has so far been the case (years rather than weeks or months). Lastly, as P. poliocephalus is one intermixing population and during a single year individuals can visit a number of roost sites within the jurisdiction of multiple governments, agencies and landholders, a range-wide or national approach (rather than local or state) is needed for their effective management. Management of flying-foxes is further complicated because they are
116 locally unwanted in most of the locations where they roost and feed. A consistent approach to the management of flying-foxes is required to ensure that management actions in one region do not have adverse impacts in another (e.g. increase numbers at other contentious sites). However, in Australia, there is not a uniform approach to the management of flying-foxes and their roost sites, and the threatened species legislation guiding management approaches varies between states (Eby and Lunney 2002a; Thiriet 2010). 6.4 Research directions
This thesis has addressed some key questions related to the ecology of Pteropus poliocephalus, and considered the implications of these results for management. It has provided new knowledge about the species distribution, foraging ecology and roost management. However, there are still unresolved issues related to each of these topics that require further information to improve management and conservation. Below I propose a series of potential future research directions under three key themes. The list is not intended to be comprehensive and is not in order of importance. Species distribution Clarify the western boundary of the range of P. poliocephalus and its patterns of occupation in inland areas. Investigate the location, composition and temporal usage patterns of P. poliocephalus roosts in the north of the range (north of 24oS). Obtain long-term data on the abundance of P. poliocephalus within that part of its range where P. alecto has expanded in historical times, together with more information on the ecology of each species, to increase understanding of the interactions between the two species where their ranges overlap. Determine the drivers of the southern range extension of P. alecto (this study found that climate change was not a sufficient explanation; alternative hypotheses such as habitat loss have yet to be investigated). Movement and feeding ecology Determine whether different sex and age classes of P. poliocephalus exhibit different movement patterns (this study was restricted to adult male flying-foxes, but advances in
117 satellite telemetry should mean that it will be possible to track smaller animals in the future). Conduct additional research examining factors such as food type and migratory status influence on the feeding distances moved by P. poliocephalus. Identify the mechanisms used by P. poliocephalus to track the availability of food resources over large areas/ spatial scales. Develop methods for monitoring short-term (i.e. weeks and months) trends in nectar supply at a landscape scale to understand the frequency of movements of P. poliocephalus among roosts in a local region. Roost management Determine the criteria involved in roost selection by P. poliocephalus, particularly at a local scale, and investigate whether these criteria vary between landscapes and regions. Examine the impacts of roost relocations on animal welfare, mortality and reproduction (at present, monitoring of relocation attempts has been largely restricted to the abundance of flying-foxes at old and new roost sites). Obtain long-term data on the abundance of flying-foxes at a range of urban and non-urban (rural, native forest) roost sites, to determine whether flying-foxes are increasing in abundance in urban areas at the expense of non-urban areas. Conduct research on social attitudes to flying-foxes and what actions could be most effective in mitigating conflict with humans. Examine the role politics and the media play in determining management outcomes.
118
References
Alexandra L.V. and Arblaster J.M. (2009) Assessing trends in observed and modelled climate extremes over Australia in relation to future projections. International Journal of Climatology, 29: 417-435. Andersen K. (1912) Catalogue of the Chiroptera in the collections of the British Museum. Vol I. Megachiroptera. British Museum of Natural History, London. Andrewartha H.G. (1961) Introduction to the Study of Animal Populations. University of Chicago Press, Chicago. Andrewartha H. and Birch L. (1954) The Distribution and Abundance of Animals. University of Chicago Press, Chicago. Ayres M.P. and Lombardero M.J. (2000) Assessing the consequences of climate change for forest herbivore and pathogens. Science of the Total Environment, 262: 263286. Banack S.A. (1998) Diet selection and resource use by flying foxes (Genus Pteropus). Ecology, 79: 194967. Banack S.A. and Grant G.S. (2002) Spatial and temporal movement patterns of the flying fox, Pteropus tonganus, in American Samoa. Journal Wildlife Management, 66: 115463. Bartholomew G.A., Leitner P. and Nelson J.E. (1964) Body temperature, oxygen consumption, and heart rate in three species of Australian flying-foxes. Physiological Zoology, 37: 179198.
Bell D.V. and Austin L.W. (1985) The game-fishing season and its effects on overwintering wildfowl. Biological Conservation, 33: 6580.
Berthold P. (2001) Bird migration. A general survey, 2nd edn. Oxford University Press, UK. Birt P. (2005) Mutualistic interactions between the nectar-feeding little red flying-fox Pteropus scapulatus (Chiroptera: Pteropodidae) and eucalypts: habitat utilisation and pollination. Doctoral dissertation, University of Queensland, Brisbane. Birt P. and Markus N. (1999) Notes on the temporary displacement of Pteropus alecto and P. poliocephalus by P. scapulatus within a day-time campsite. Australian Mammalogy, 21: 107110. Birt P., Markus N., Collins L. and Hall L.S. (1998) Urban flying-foxes. Nature Australia, 26: 5459. Birtchnell M.J. and Gibson M. (2008) Flowering ecology of honey-producing flora in SouthEast Australia. Rural Industries Research and Development Corporation, Australia. BOM (2010) Australian Bureau of Meteorology, climate data online. Accessed July 2010. Available at http://www.bom.gov.au/
References
119 Breed A.C., Field H.E., Smith C.S., Edmonston J. and Meers J. (2010a) Bats without borders: long-distance movements and implications for disease risk management. EcoHealth, 8: 20412. Epub ahead of print http://www.springerlink.com/content/j0t5u3567t56117u/fulltext.pdf Breed A.C., Yu M., Barr J.A., Crameri G., Thalmann C.M., and Wang L.F. (2010b) Prevalence of henipavirus and rubulavirus antibodies in pteropid bats, Papua New Guinea. Emerging Infectious Diseases, 16: 1997-1999. Breed A., Field H., Perkins N., Eby P., Cunningham A. and Prowse S. (2010c) Re: Flyingfoxes carrying Hendra virus in Queensland pose a potential problem in other states. Australian Vet Journal, 88: 8 N24. Brooke A.P., Solek C. and Tualaulelei A. (2000) Roosting behavior of colonial and solitary flying foxes in American Samoa (Chiroptera: Pteropodidae). Biotropica, 32: 338350. Brown J.H., Stevens G.C. and Kaufman D.M. (1996) The geographic range: size, shape, boundaries and internal structure. Annual Review of Ecology and Systematics, 27: 597 623. Brown D.T. (2007) Biodiversity, bats and Buddhism: Protecting Pteropus in Thailand. The first international south-east Asian Bat Conference, Phuket, Thailand. Catterall C.P., Kingston M.B., Park K. and Sewell S (1998) Deforestation, urbanization and seasonality: interacting effects on a regional bird assemblage Biological Conservation, 84: 65-81. Catterall C.P. and Kingston M. (1993) Remnant bushland of south-east Queensland in the 1990s: Its distribution, loss, ecological consequences and future prospects. Griffith University and Brisbane City Council, Brisbane, Institute of Applied Environmental Research. Chambers L.E., Hughes L. and Weston M.A. (2005) Climate Change and its impact on Australias Avifauna. Emu, 105: 120. Chan J. (2007) Genetic estimates of dispersal and the implications for conservation management of grey-headed flying-fox (Pteropus poliocephalus), B.Sc. (Hons) Thesis, University of New South Wales. Chan K. (2001) Partial migration in Australian landbirds: a review. Emu, 101: 20812292. Cheke A.S. and Dahl J.F. (1981) The status of bats on western Indian Ocean Islands with special reference to Pteropus. Mammalia, 45: 205238. Churchill S. (1998) Australian bats. Reed New Holland, Sydney NSW. Churchill S. (2008) Australian bats. 2nd edn. Jacana Books, Crows Nest. Clancy G. (2001) The fauna of Susan Island, Grafton. A report to Susan Island Trust and National Parks and Wildlife Service, Grafton.
References
120 Coffs Harbour City Council (2007) Coffs Creek flying-fox roost strategy. Coffs City Council, NSW. Cohn J.P. (1999) Tracking wildlife. Bioscience, 49: 1217. Collett R. (1887) On a collection of mammals from central and northern Queensland. Zoologische Jahrbucher, 829940. Collins D. (2000) Annual temperature summary Australia records warmest decade. Climate Change Newsletter, 12: 6. Crozier L. (2004) Warmer winters drive butterfly range expansion by increasing survivorship. Ecology, 85: 231241. CSIRO (2007) Climate change in Australia. Technical report 2007. Bureau of Meteorology. Davis A.J., Jenkinson L.S., Lawton J.H., Shorrocks, B. and Wood S. (1998) Making mistakes when predicting shifts in species range in responses to global warming. Nature, 391: 783 786. DECC (2006) New South Wales State of the Environment 2006. Department of Environment and Conservation NSW, Sydney. Available at http://www.environment.nsw.gov.au/soe/soe2006/index.htm DECCW (2009) Draft national recovery plan for the grey-headed flying-fox Pteropus poliocephalus. Prepared by Dr Peggy Eby. Department of Environment, Climate Change and Water NSW, Sydney. Available at http://www.environment.nsw.gov.au/resources/threatenedspecies/08214dnrpflyingfox.pdf Department of Environment and Climate Change NSW (2007) Flying-fox camp management policy. Department of Environment and Climate Change NSW, Sydney. Department of Sustainability and Environment (2005) Flying-fox campsite management plan. Yarra Bend Park. Available at http://www.dse.vic.gov.au/plants-and-animals/flying-foxeshome-page/flying-foxes-at-yarra-bend-park Department of Sustainability and Environment (2011a) Standard operating procedure: nudging of grey-headed flying-foxes at Yarra Bend Park. Department of Sustainability and Environment. Department of Sustainability and Environment (2011b) Protocols for management of greyheaded flying-foxes at Yarra Bend Park during heat stress events. Department of Sustainability and Environment (unpublished). Dickman C. and Fleming M. (2002) Pest, or Passenger Pigeon? The New South Wales Scientific Committees assessment of the status of the grey-headed flying-fox, pp. 2028 in Managing the grey-headed flying-fox as a threatened species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman.
References
121 Divljan A., Parry-Jones K. and Wardle G.M. (2006) Age determination in the grey-headed flying-fox. Journal Wildlife Management, 70: 607611. Dobson G.E. (1878) Catalogue of the Chiroptera in the collection of the British Museum. British Museum of Natural History, London. Dolbeer R.R., Fielder L.R and Rasheed H. (1988) Management of fruit bat and rat populations in the Maldive Islands, Indian Ocean, pp. 112118 in Proceedings of the Thirteenth Vertebrate Pest Conference, edited by A.C. Crabb and R.E. Marsh. University of California, Davis. Dumont E.R. (2003) Bats and fruit, pp. 156208 in Ecology of bats, edited by T.H. Kunz and M.B. Fenton. University of Chicago Press, Chicago. Duncan A., Baker G.B. and Montgomery N. (1999) The Action Plan for Australian Bats. Environment Australia, Canberra. Available from: http://www.environment.gov.au/biodiversity/threatened/publications/action/bats/ Eby P. (1991a) Seasonal movements of grey-headed flying-foxes, Pteropus poliocephalus (Chiroptera: Pteropodidae), from two maternity roosts in northern New South Wales. Wildlife Research, 18: 547559. Eby P. (1991b) Finger-winged night workers: managing forests to conserve the role of greyheaded flying foxes as pollinators and seed dispersers, pp 91100 in Conservation of Australias forest fauna, edited by D. Lunney D. Royal Zoological Society of New South Wales, Mosman. Eby P. (1995) Biology and management of Pteropus in New South Wales. Species report. Report to NSW National Parks and Wildlife Service, Hurstville. Eby P. (1996) Interactions between the grey-headed flying-fox Pteropus poliocephalus (Chiroptera: Pteropodidae) and its diet plants seasonal movements and seed dispersal. Doctoral dissertation, University of New England, Armidale. Eby P. (2002a) National count of grey-headed flying-foxes July 27 and 28, 2002. Report to Environment Australia, Queensland Parks and Wildlife Service, NSW National Parks and Wildlife Service, Victoria Dept Natural Resources and Environment. Eby P. (2002b) Using New South Wales planning instruments to improve conservation and management of Grey-headed Flying-fox roosts, pp 240250 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of New South Wales, Mosman. Eby P. (2003) National count of grey-headed flying-foxes April 12 and 13, 2003. Report to Environment Australia, Queensland Parks and Wildlife Service, NSW National Parks and Wildlife Service, and Victoria Department Natural Resources and Environment. Eby P. (2004) National count of grey-headed flying-foxes April 3 and 4, 2004. Report to the Department of Environment and Heritage, Canberra.
References
122 Eby P. (2007) Flying-fox roosts of the grey-headed flying-fox. DECC, Hurstville, NSW: Department of Environment and Conservation. Eby P. and Law B (2008) Ranking the feeding habitat of grey-headed flying foxes for conservation management. Canberra: Department of Environment, Heritage, Water and the Arts. Eby P. and Lunney D. (eds) (2002a) Managing the grey-headed flying-fox as a threatened species in NSW. Royal Zoological Society of NSW, Mosman, NSW, 286 pp. Eby P. and Lunney D. (2002b) Managing the grey-headed flying-fox Pteropus poliocephalus as a threatened species in NSW: a context for debate, pp. 115 in Managing the greyheaded flying-fox as a threatened species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman, NSW. Eby P. and Lunney D. (2002c) Managing the grey-headed flying-fox Pteropus poliocephalus as a threatened species in NSW: adjusting to a long-term vision, pp. 273284 in Managing the grey-headed flying-fox as a threatened species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Eby P. and Palmer C. (1991) Flying-foxes in rainforest remnants in northern New South Wales, pp. 4856 in Rainforest remnants, edited by S. Phillips. NSW National Parks and Wildlife Service, Lismore. Eby P., Richards G., Collins L. and Parry-Jones K. (1999) The distribution, abundance and vulnerability to population reduction of a nomadic nectarivore, the grey-headed flying-fox Pteropus poliocephalus in New South Wales, during a period of resource concentration. Australian Zoologist, 31: 240253. Epstein J.H., Olival K.J., Pulliam J., Smith C., Westrum J., Hughes T., et al. (2009) Pteropus vampyrus, a hunted migratory species with a multinational home-range and a need for regional management. Journal of Applied Ecology, 46: 9911002. Fahrig L. (2003) Effects of habitat fragmentation on biodiversity. Annual Reviews in Ecology and Systematics, 34: 487515. Field H. and Kung N. (2011) Henipaviruses unanswered questions of lethal zoonoses. Current Opinion in Virology, 1: 658661. Fischer J. and Lindenmayer D.B. (2000) An assessment of the published results of animal relocations. Biological Conservation, 96: 111. Fleming T.H. and Eby P. (2003) Ecology of bat migration, pp. 156208 in Ecology of bats, edited by T.H. Kunz and M.B. Fenton. University of Chicago Press, Chicago. Fletcher A. (2010) Burdekin Park Flying-foxes feasibility of relocation project. Meeting of Singleton Council on 12 April 2010. Available at http://www.singleton.nsw.gov.au /agenda/Open/2010/CM_12042010_AGN_AT_WEB.HTM
References
123 Fox, S. (2011) The Spectacled Flying-fox review of past and present knowledge, pp. 136145 in The Biology and Conservation of Australasian Bats, edited by B. Law, P. Eby, D. Lunney and L. Lumsden. Royal Zoological Society of NSW, Mosman, NSW, Australia. Fraser G.C., Hooper P.T., Lunt R.A., Gould A.R., Gleeson L.J., Hyatt A.D., et al. (1996) Encephalitis caused by a Lyssavirus in fruit bats in Australia. Emerging Infectious Diseases, 2: 327331. Fujita M.S. and Tuttle M.D. (1991) Flying foxes (Chiroptera, Pteropodidae) threatened animals of key ecological and economic importance. Conservation Biology, 5, 455463. Garnett S., Whybird O. and Spencer H. (1998) The conservation status of the spectacled flying-fox, Pteropus conspicillatus, in Australia. Australian Zoologist, 31: 3854. Gaston K.J. (2003) The Structure and Dynamics of Geographic Ranges. Oxford University Press, New York. Gauthreaux S.A. (1982) The ecology and evolution of avian migration systems, pp. 93168 in Avian biology. Vol. VI, edited by D.S. Farner. Academic Press. Gill R.E., Tibbitts T.L, Douglas D.C., Handel1 C.M., Mulcahy D.M., Gottschalck J.M., et al. (2009) Extreme endurance flights by landbirds crossing the Pacific Ocean: ecological corridor rather than barrier? Proceedings of the Royal Society B, 276: 447457. Grabherr G., Gottfried M. and Pauli H. (1994) Climate effects on mountain plants. Nature, 369: 448. Griffith B., Scott J.M., Carpenter J.W. and Reed, C. (1989) Translocation as a species conservation tool: status and strategy. Science, 245: 477480. Hall L.S. (2002) Management of flying-fox camps: what have we learnt in the last twenty five years? pp. 215224 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Hall L.S. (2006) Flying-foxes as a pest species. Queensland pest animal symposium, Conference proceedings. Queensland Government, Toowoomba. Hall L.S. and Richards G.C. (2000) Flying-foxes, fruit and blossom bats of Australia. University of New South Wales Press, Sydney, 135 pp. Halpin K., Young P., Field H. and Mackenzie J. (2000) Isolation of Hendra virus from pteropid bats: a natural reservoir of Hendra virus. Journal of General Virology, 81: 1927 1932. Heideman P.D. and Heaney L.R. (1992) Pteropus vampyrus lanensis species account, pp. 141142 in Old World Fruit Bats: an Action Plan for Their Conservation, edited by S.P. Mickleburgh, A.M. Hutson, P.A. Racey. IUCN Press, Gland Switzerland.
References
124 Hickling R., Roy D.B., Hill J.K., Fox R and Thomas C.D. (2006) The distributions of a wide range of taxonomic groups are expanding polewards. Global Change Biology, 12: 450 455. Higuchi H., et al.. (2005) Migration of Honey-buzzards Pernis apivorus based on satellite tracking. Ornithological Science, 4: 109115. Hockey P. and Midgley G. (2009) Avian range changes and climate change: a cautionary tale from the Cape Peninsula. Ostrich, 80: 2934. Hodgkison R., Balding S.T., Zubaid A. and Kunz T.H. (2003) Fruit bats (Chiroptera : Pteropodidae) as seed dispersers and pollinators in a lowland Malaysian rain forest. Biotropica, 35: 491502. Hudson I.L., Keatley M. R. and Kang I. (2010) Wavelet characterization of eucalypt flowering and the influence of climate. Environmental and Ecological Statistics, Published first online 27 June, 2010. DOI: 10.1007/s10651-010-0149-5. Hughes L. (2000) Biological consequences of global warming: Is the signal already apparent? Trends in Ecology & Evolution, 15: 5661. Hughes L. (2003) Climate change and Australia: Trends, projections and impacts. Austral Ecology, 28: 423443. Hyne R. (2010) Flying-foxes carrying Hendra virus in Queensland pose a potential problem in other states. Australian Vet Journal., 88: N2021. IUCN (2012) IUCN Red List of Threatened Species. Version 2012.1. <www.iucnredlist.org>. Downloaded on 26 June 2012. Johara M., Field H., Rashdi A., Morrissy C., vander Heide B., Rota P., Adzhar A., White J., Daniels P., Jamaluddin A., Ksiazek T. (2001) Serological evidence of infection with Nipah virus in bats (order Chiroptera) in Peninsular Malaysia. Emerging Infectious Diseases, 7: 439441. Jones D. N. and Nealson T. (2003). Management of aggressive Australian magpies by translocation. Wildlife Research, 30: 167177. Jonsson N.N., Johnston S.D., Field H., de Jong C. and Smith C. (2004) Field anaesthesia of three Australian species of flying-fox. Veterinary Record, 154: 664. Karasov W.H. (1992) Daily energy expenditure and the cost of activity in mammals. American Zooogist, 32: 23848. Keatley M.R., Hudson I.L. (2012) Detecting change in an Australian flowering record: Comparisons of linear regression and cumulative sum analysis change point analysis. Austral Ecology, 37: 825-835. Keith D.A. (2004) Ocean shores to desert dunes. The native vegetation of New South Wales and the ACT, Department of Environment and Conservation, Hurtsville.
References
125 Keith D.A. and Scott J. (2005) Native vegetation of coastal floodplains a diagnosis of the major plant communities in New South Wales. Pacific Conservation Biology. 11: 81104. Kenward R.E. (2001) A manual for wildlife radio-tracking. Academic Press, London. Kingsford R.T. and Norman F.I. (2002) Australian waterbirds-products of the continents ecology. Emu, 102: 4769. Krebs C.J. (2001) Ecology: The experimental analysis of distribution and abundance. (5th edition). Benjamin-Cummings Publishing Co., USA. Kunz T.H. (1982) Roosting ecology, pp. 155 in Ecology of bats, edited by T.H. Kunz. Plenum Press, New York. Kunz T.H. and Fenton M.B. (2003) Bat ecology. University of Chicago Press, Chicago. 779 pp. Kunz T.H. and Lumsden L.F. (2003) Ecology of cavity and foliage roosting bats, pp. 156208 in Ecology of bats, edited by T.H. Kunz and M.B. Fenton. University of Chicago Press, Chicago. Law B.S. and Dickman C.R. (1998) The use of habitat mosaics by terrestrial vertebrate fauna: implications for conservation and management. Biodiversity and Conservation, 7: 323 333. Loughland R.A. (1998) Mangal roost selection by the flying fox Pteropus alecto (Megachiroptera: Pteropodidae). Marine and Freshwater Research, 49: 351352. Lunney D. (2012) Ethics and Australian mammalogy: reflections on 15 years (19912006) on an Animal Ethics Committee. Australian Mammalogy, 34: 117. Lunney D. and Leary T. (1985) An ecological interpretation of the European impact on the Bega District (NSW) from historical records. III. Consequences for native fauna. National Parks and Wildlife Service, Sydney. Lunney D. and Moon C. (1997) Flying-foxes and their camps in the rainforest remnants of north-east NSW, pp. 247277 in Australias Ever-Changing Forests III, edited by J. Dargavel. Centre for Resource and Environmental Studies, Australian National University, Canberra. Lunney D. and Moon C. (2011) Blind to bats. Traditional prejudices and todays bad press render bats invisible to public consciousness, pp. 44-63 in The biology and conservation of Australian bats, edited by B. Law, P. Eby, D. Lunney and L. Lumsden. Royal Zoological Society of NSW, Mosman. Lunney D., Richards G. and Dickman C. (2008) Pteropus poliocephalus, in IUCN 2012. IUCN Red List of Threatened Species. Version 2012.2. <www.iucnredlist.org>. Downloaded on 06 June 2012.
References
126 Madsen J. and Fox A.D. (1995) Impacts of hunting disturbance on waterbirds a review. Wildlife Biology, 1: 193207. Markus N. (2001) Ecology and behaviour of the black flying-fox Pteropus alecto in an urban environment. Doctoral dissertation, University of Queensland. Markus N. and Hall L. (2004) Foraging behaviour of the black flying-fox (Pteropus alecto) in the urban landscape of Brisbane, Queensland. Wildlife Research, 31: 345355. Markus N., Palmer C. and Hall, L.S. (2008) Black Flying-fox Pteropus alecto, pp 435-436 in The Mammals of Australia, edited by S. van Dyck and R. Strahan. New Holland Publishers, Sydney. Marshall A.G. (1985) Old World phytophagous bats (Megachiroptera) and their food plants: a survey. Zoological Journal of the Linnean Society, 83: 35169. Martin L. and A.P. McIlwee (2002) The reproductive biology and intrinsic capacity for increase of the grey-headed flying-fox Pteropus poliocephalus (Megachiroptera), and the implications of culling, pp. 91108 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Martin T.G., Cahdes I., Arcese P., Marra P.P., Possingham H.P. and Norris D.R. (2007) Optimal conservation of migratory species. PloS ONE, 2: e751. Massei G., Quy R.J., Gurney J. and Cowan D.P. (2010) Can translocations be used to mitigate human-wildlife conflicts? Wildlife Research, 37: 428-439. McCann C. (1934) Notes on the flying-fox (Pteropus giganteus Brunn). Journal Bombay Natural History Society, 41: 804818. McDonald-Madden E., Schreiber E.S.G., Forsyth D.M., Choquenot D. and Clancy T.F. (2005) Factors affecting the grey-headed flying fox (Pteropus poliocephalus: Pteropidae) foraging in the Melbourne metropolitan area, Australia. Austral Ecology, 30: 600608. McKeown A. and Westcott D.A. (2011) Assessing the accuracy of small satellite transmitters on free-living flying-foxes. Austral Ecology, 37: 295301. McLachlan J.S., Hellmann J.J. and Schwartz M.W. (2007) A framework for debate of assisted migration in an era of climate change. Conservation Biology, 21: 297302. McWilliam A.N. (1986) The feeding ecology of Pteropus in north-eastern New South Wales, Australia. Myotis, 2324: 201208. Mickleburgh S.P., Hutson A.M. and Racey P.A. (eds) (1992) Old World Fruit Bats. An action plan for their conservation. IUCN/SSC Chiroptera Specialist Group. IUCN, Gland, Switzerland. Mickleburgh S.P., Hutson A.M. and Racey P.A. (2002) A review of the global conservation status of bats. Oryx, 36: 1834.
References
127 Mildenstein T.L. (2002) Habitat selection of large flying foxes using radio telemetry: Targeting conservation efforts in Subic Bay, Philippines. M.S. Thesis, University of Montana. Mildenstein T.J., Stier S.C., Nuevo-Diego C.E. and Mills L.S. (2005) Habitat selection of endangered and endemic large flying-foxes in Subic Bay, Philippines. Biological Conservation, 126: 93102. Molur S., Srinivasulu C., Bates P. and Francis C. (2008) Pteropus giganteus. IUCN 2012. IUCN Red List of Threatened Species. Version 2012.1. <www.iucnredlist.org>. Downloaded on 19 July 2012. Montevecchi W.A., Cairns D.K. and Myers R.A. (2002) Predation on marine-phase Atlantic salmon by gannets in the northwest Atlantic. Canadian Journal of Fisheries and Aquatic Sciences, 59: 602612. Mueller T. and Fagan W.F. (2008) Search and navigation in dynamic environments from individual behaviors to population distributions. Oikos, 117: 654-664. Nelson J.E. (1963) The biology of the flying-fox (Genus Pteropus) in south eastern Queensland. Doctoral dissertation, University of Queensland, Brisbane. Nelson J.E. (1964) Vocal communication in Australian flying-foxes (Pteropodidae: Megachiroptera). Zeitschrift fur Tierpsychologie, 21: 857870. Nelson J.E. (1965) Movements of Australian flying-foxes (Pteropodidae: Megachiroptera). Australian Journal of Zoology, 13: 5373. Nelson J.E. (2008a) Report for the Department of Environment and Climate Change (DECC) regarding management of flying-foxes in Burdekin Park, Singleton. Singleton City Council, Singleton. Nelson J.E. (2008b) Options for the management of flying-foxes in Maclean. Department of Environment and Climate Change, Coffs Harbour. Newton I. (2007) The migration ecology of birds. Academic Press. 976 pages. NLWRA (2001) Australian Native Vegetation Assessment 2001. National Land and Water Resources Audit, Canberra. Olsen P. (2007) The state of Australias birds 2007. Birds in a changing climate. Wingspan, 14, Supplement. Olsen P., Weston M., Cunningham R. and Silcocks A. (2003) The state of Australias birds 2003. Wingspan, 13, Supplement. Pallin N. (2000) Ku-ring-gai Flying-fox Reserve: habitat restoration project, 15 years on. Ecological Management and Restoration, 1: 1020.
References
128 Palmer C. (1997) Ecology of the black flying fox Pteropus alecto in the seasonal tropics of the Northern Territory: resource tracking in a land scape mosaic and role in seed dispersal. M.Sc. Thesis, Northern Territory University, Darwin. Palmer C. and Woinarski J.C.Z. (1999) Seasonal roosts and foraging movements of the black flying-fox (Pteropus alecto) in the Northern Territory: resource tracking in a landscape mozaic. Wildlife Research, 26: 823838. Palmer C., Price O. and Bach C. (2000) Foraging ecology of the black flying fox (Pteropus alecto) in the seasonal tropics of the Northern Territory, Australian Wildlife Research, 27: 169178. Parmesan C. (1996) Climate and species range. Nature, 382: 765766. Parmesan C.P. and Yohe G. (2003) A globally coherent fingerprint of climate change across natural systems. Nature, 421: 3742. Parmesan C., Gaines S., Gonzales L., Kaufman D.M., Kingsolver J., Peterson A.T. and Sagarin R. (2005) Empirical perspectives on species borders: from traditional biogeography to global change. Oikos, 108: 5875. Parmesan C., Ryrholm N., Stefanescu C., Hill J.K., Thomas C.D., Descimon H., Huntley B., Kaila L., Kullberg J., Tammaru T., Tennent W.J., Thomas J.A. and Warren M. (1999) Poleward shifts in geographical ranges of butterfly species associated with regional warming. Nature, 399: 579583. Parris K.M. and Hazell D.L. (2005) Biotic effects of climate change in urban environments: the case of the Grey-headed Flying-fox (Pteropus poliocephalus) in Melbourne. Biological Conservation, 124: 267276. Parry-Jones K.A. and Augee M. (1991) Food selection by grey-headed flying-foxes (Pteropus poliocephalus) occupying a summer colony site near Gosford, NSW. Wildlife Research, 18: 111124. Parry-Jones K.A. and Augee M. (1992) Movements of grey-headed flying-foxes (Pteropus poliocephalus) to and from a colony site on the central coast of New South Wales. Wildlife Research, 19: 331340. Parry-Jones K.A. and Augee M.L. (2001) Factors affecting the occupation of a colony site in Sydney, New South Wales by the grey-headed flying-foxes, Pteropus poliocephalus (Pteropodidae). Austral Ecology, 26: 4755. Peacock L. (2004) Roost preference of the grey-headed flying-fox. Honours thesis, Faculty of Veterinary Science. University of Sydney, NSW. Philbey A.W., Kirkland P.D., Ross A.D., Field H.E., Srivastava M., Davis R.J., et al. (2008) Infection with Menangle virus in flying-foxes (Pteropus spp.) in Australia. Australian Vet Journal, 86: 449454.
References
129 Phillips P., Hauser P. and Letnic M. (2007) Displacement of Black Flying-foxes Pteropus alecto from Batchelor, Northern Territory. Australian Zoologist, 34: 119124. Pierson E.D. and Rainey W.E. (1992) The biology of flying-foxes of the genus Pteropus: A review, pp. 117 in Pacific Island flying-foxes: Proceedings of an International Conservation Conference, edited by D.E. Wilson and G.L. Graham. United States Department of the Interior, Fish and Wildlife Service, Washington. Pierson E.D., Elmqvist T., Rainey W.E. and Cox P.A. (1996) Effects of tropical cyclonic storms on flying fox populations on the South Pacific islands of Samoa. Conservation Biology, 10: 438451. Price O.F.,Woinarski J.C.Z. and Robinson D. (1999) Very large area requirements for frugivorous birds in monsoon rainforests of the Northern Territory, Australia. Biological Conservation, 91: 169180. Queensland Parks and Wildlife Service (2002) Guidelines for the management of flying-foxes in Calliope Shire. Memorandum of Agreement. Queensland Parks and Wildlife Service and Calliope Shire Council. Rainey W.E. and Pierson E.D. (1992) Distribution of the Pacific Island flying foxes, pp. 117 in Pacific Island flying-foxes: Proceedings of an International Conservation Conference, edited by D.E. Wilson and G.L. Graham. United States Department of the Interior, Fish and Wildlife Service, Washington. Ratcliffe F.N. (1931) The flying-fox (Pteropus) in Australia. Bulletin of the Council for Scientific and Industrial Research, 53: 180. Ratcliffe F.N. (1932) Notes on the fruit bats (Pteropus spp.) of Australia. Journal of Animal Ecology, 1: 3257. Ratcliffe F.N. (1938) Flying-fox and drifting sand: the adventures of a biologist in Australia. Chatto and Windus, London. Rhodes M. and Hall L.S. (1997) Bats of Fraser Island. Australian Zoologist, 30: 346350. Richards G.C. (1990) The Spectacled Flying Fox, Pteropus conspicillatus (Chiroptera: Pteropididae), in north Queensland. 1. Roost sites and distribution patterns. Australian Mammalogy, 13: 122. Richards G.C. (1995) A review of ecological interactions of fruit bats in Australian ecosystems. Symposia of the Zoological Society of London, 67: 7996. Richards G.C. (2002) The development of strategies for management of the flying fox colony at the Royal Botanic Gardens, Sydney, pp. 196201 in Managing the Grey-headed Flyingfox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman, NSW. Richter H.V. and Cumming G.S. (2006) Food availability and annual migration of the strawcolored fruit bat (Eidolon helvum). Journal of Zoology, 268: 3544.
References
130 Richter H.V. and Cumming G.S. (2008) First application of satellite telemetry to track African straw-coloured fruit bat migration. Journal of Zoology, 275: 172176. Roberts B.J. (2005) Habitat characteristics of flying-fox roosts in south-east Queensland. B.Sc. (Hons) thesis, Griffith University, Brisbane. Roberts B.J. (2006) Management of urban flying-fox roosts: issues of relevance to roosts in the Lower Clarence, NSW. Valley Watch Inc., Maclean. Available at http://www.clarencevalleywatch.com/currentprojectst12.htm Roberts B.J. (2008) Assessment of habitat of the grey-headed flying-fox (Pteropus poliocephalus) on Lots 14 (DP808228) and 6 (DP863080) Dunbible, NSW and the Tweed Shire Council Reserve, Dallis Park, NSW. A report prepared for the Department of Environment, Water, Heritage and the Arts. Rogers J. (2002) The economic and social implications of flying-fox predation on the north coast of NSW. pp. 5862 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Rolland C., Danchin E. and De Fraipont M. (1998) The evolution of coloniality in birds in relation to food, habitat, predation, and life-history traits: a comparative analysis. American Naturalist, 151: 514529. Root T.L., Price J.T., Hall KR., Schneider S.H., Rosenzweig C. and Pounds J.A. (2003) Fingerprints of global warming on wild animals and plants. Nature, 421: 5760. Roshier D. A. and Reid J.R.W. (2003) On animal distributions in dynamic landscapes. Ecography, 26: 539544. Roshier D.A., Robertson A.I., Kingsford R.T. and Green D.G. (2001) Continental-scale interactions with temporary resources may explain the paradox of large populations of desert waterbirds in Australia. Landscape Ecology, 16: 547556. Roshier D., Asmus M., and Klaassen M. (2008) What drives long-distance movements in the nomadic Grey Teal (Anas gracilis) in Australia? Ibis, 150: 474484. Sattler P. and Williams R. (eds) (1999) The Conservation Status of Queenslands Bioregional Ecosystems, Environmental Protection Agency, Brisbane. Shimazak H., et al. (2004) Network analysis of potential migration routes applied to identification of important stopover sites for Oriental White Storks (Ciconia boyciana). Ecological Research, 19: 683698. Sinclair E.A, Webb N.J., Marchant A.D. and Tidemann C.R. (1996) Genetic variation in the little red flying -fox Pteropus scapulatus (Megachiroptera: Pteropodidae): implications for management. Biological Conservation, 76: 4550.
References
131 Singaravelan N., Marimuthu G., Racey P.A. (2009) Do fruit bats deserve to be listed as vermin in the Indian Wildlife (Protection) & Amended Acts? A critical review. Oryx, 43: 608613. Singleton City Council (2008) Relocation of the grey-headed flying-fox from Burdekin Park. Referral to the Department of Environment, Water, Heritage and the Arts. Smith C.S., Epstein J.H., Breed A.C., Plowright R.K., Olival K.J., et al. (2011) Satellite telemetry and long-range bat movements. PLoS ONE, 6: e14696. doi:10.1371/journal.pone.0014696. Smith M. (2002) Management of roost sites of the grey-headed flying-fox Pteropus poliocephalus on the North Coast of NSW: a National Parks and Wildlife Service perspective, pp. 202214 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Southerton S.G., Birt P., Porter J. and Ford H.A. (2004) Review of gene movement by bats and birds and its potential significance for eucalypt plantation forestry. Australian Forestry, 67: 4453. Spencer H.J., Palmer C. and Parry-Jones K. (1991) Movements of fruit-bats in eastern Australia, determined by using radio-tracking. Wildlife Research, 18: 463468. Stager K.E. and Hall L.S. (1983) A cave-roosting colony of the black flying fox (Pteropus alecto) in Queensland, Australia. Journal of Mammal, 64: 523524. Stevenson L. (2004) Attempted relocation at Singleton, NSW great expectations, unexpected outcomes. Managing Flying-fox roosts from Melbourne to Mataranka. Australasian Bat Society, Ipswich, Qld. Stone R., Nicholls N. and Hammer G. (1996) Frost in northeast Australia: Trends and influences of phases of the southern oscillation. Journal of Climate, 9: 1896909. Strahan R. (1994) The Australian Museum Complete Book of Australian Mammals. Angus & Robertson Publishers, Sydney. Sturm M., Racine C. and Tape K. (2001) Increasing shrub abundance in the Arctic. Nature, 411: 546547. Tanton M.T. (1999) Maclean Rainforest Reserve Precinct Plan. Report for the Maclean Flying-fox Working Party. Thiriet D. (2005) The relocation of flying-fox colonies in Queensland. Environment and Planning Law Journal, 22: 231239. Thiriet D. (2010) Flying-fox Conservation Laws, Policies and Practices in Australia a case study in conserving unpopular species. The Australasian Journal of Natural Resources Law and Policy, 13: 161194.
References
132 Thomas C.D. and Lennon J.J. (1999) Birds extend their ranges northwards. Nature, 399: 213. Thomas B., Holland J.D. and Minot E.O. (2011) Wildlife tracking technology options and cost considerations. Wildlife Research, 38: 653663. Thompson B.G. (1991) A Field Guide to the Bats of the Northern Territory. Conservation Commission of the Northern Territory, Darwin. Thomson S.W., Brooke A.P. and Speakman S.K. (1998) Diurnal activity in the Samoan flying fox, Pteropus samoensis. Philosophical Transactions of the Royal Society of London, 353: 15951606. Tidemann C.R. (1985) A study of the status, habitat requirements and management of the two species of bats on Christmas Island (Indian Ocean). Australian National Parks and Wildlife Service. 78 pp. Tidemann C.R. (1999) Biology and management of the grey-headed flying-fox, Pteropus poliocephalus. Acta Chiropterologica, 1: 151164. Tidemann C.R. (2002) Sustainable Management of the Grey-headed Flying-fox Pteropus poliocephalus, pp 122127 in Managing the Grey-headed Flying-fox as a Threatened Species in NSW, edited by P.Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Tidemann C.R. (2003) Displacement of a flying-fox roost using sound. Ecological Management & Restoration, 4: 224226. Tidemann C.R. and Nelson J. E. (1987) Flying foxes (Chiroptera: Pteropodidae) and bananas: some interactions. Australian Mammology, 10: 133136 . Tidemann C.R. and Nelson J. (2004) Long-distance movements of the grey-headed flying fox (Pteropus poliocephalus). Journal Zoology, 263: 141146. Tidemann C.R. and Nelson J. (2011) Life expectancy, causes of death and movements of the grey-headed flying-fox (Pteropus poliocephalus) inferred from banding. Acta Chiropterologica, 13: 419429. Tidemann C.R., Kelson S.L. and Jamieson G. (1997) Flying-fox damage to orchard fruit in Australia incidence, extent and economic impact. Australian Biologist, 10: 179186. Tidemann C.R., Vardon M.J., Loughland R.A. and Brocklehurst P.A. (1999) Dry season roosts of flying-foxes (Pteropus spp.) in Kakadu World Heritage Area, north Australia. Journal of Zoology, 247: 155163. Tidemann C.R., Eby P., Parry-Jones K.A. and Nelson J.E. (2008) Grey-headed Flying-fox Pteropus poliocephalus, pp 444-445 in The Mammals of Australia, edited by S. van Dyck and R. Strahan. New Holland Publishers, Sydney.
References
133 Traill L.W., Bradshaw C.J.A and Brook B.W. (2010) Satellite telemetry and seasonal movements of Magpie Geese (Anseranas semipalmata) in tropical northern Australia. Emu, 110: 160-164. Travis J.M.J. and Dytham C. (1999) Habitat persistence, habitat availability and the evolution of dispersal. Proceedings of Royal Society of London B, 266: 723728. Toop S. (2004) Relocating Melbournes flying-foxes an overview of practices and processes. Managing Flying-fox roosts from Melbourne to Mataranka. Australasian Bat Society, Ipswich, Qld. Towns D.R. and Daugherty C.H. (1994) Patterns of range contractions and extinctions in the New Zealand herpofauna following human colonisation. NZ Journal of Zoology, 21: 325 339. Utzurrum, R.C.B. (1995) Feeding ecology of Philippine fruit bats: patterns of resource use and seed dispersal. Symposia of the Zoological Society of London, 67: 6377. van der Ree R. and North J.M. (2009) Public Environment Report. Proposed relocation of a roost of Grey-headed Flying-foxes (Pteropus poliocephalus) from the Royal Botanical Gardens Sydney. Botanical Gardens Trust, Sydney. van der Ree R., McDonnell M.J., Temby I. D., Nelson J. and Whittingham E. (2006) The establishment and dynamics of a recently established urban camp of Pteropus poliocephalus outside their geographic range. Journal of Zoology, 268: 177185. Vardon M.J., Simpson B.K., Sherwel1 D. and Tidemann C.R. (1997) Flying-foxes and tourists: a conservation dilemma in the Northern Territory. Australian Zoologist, 30: 310. Walther G., Post E., Convey P., Menzel A., Parmesan C., Beebee T.J.C., Fromentin J., Hoegh-Guldberg O. and Bairlein F. (2002) Ecological responses to recent climate change. Nature, 416: 389395. Waples K. (2002) Review of the NPWS policy on the mitigation of commercial crop damage by flying-foxes, pp. 3946 in Managing the grey-headed flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Warren M.S., Hill J.K., Asher T.J., Fox R., Huntley B., Roy D.B., Telfer M.G., Jeffcoate S., Harding P., Jeffcoate G., Willis S.G., Greatorex-Davies J.N., Moss D. and Thomas C.D. (2001) Rapid responses of British butterflies to opposing forces of climate and habitat change. Nature, 414: 6569. Webb N. and Tidemann C. (1995) Hybridisation between black (Pteropus alecto) and greyheaded (P. poliocephalus) flying-foxes (Megachiroptera: Pteropodidae). Australian Mammalogy, 18: 1926. Webb N.J. and Tidemann C.R. (1996) Mobility of Australian flying-foxes. Pteropus spp (Megachiroptera): evidence from genetic variation. Proceedings of the Royal Society of London B, 263: 497502.
References
134 Welbergen J.A. (2005) The social organisation of the grey-headed flying-fox (Pteropus poliocephalus). Doctoral dissertation, Department of Zoology. University of Cambridge, Cambridge, UK. Welbergen J.A. (2011) Fit females and fat polygynous males: seasonal body mass changes in the grey-headed flying fox, Oecologia, 165: 629637. Welbergen J.A., Klose S.M., Markus N. and Eby P. (2007) Climate change and the effects of temperature extremes on Australian flying-foxes. Proceedings of the Royal Society B, 275: 419425. West C. (2002) Contemporary issues in managing flying-fox roosts, pp. 176195 in Managing the grey-headed flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman. Wikelski M. et al. (2007) Going wild: what a global small-animal tracking system could do for experimental biologist. The Journal of Experimental Biology, 210: 181186. Wilcove, D.S. and Wikelski, M. (2008) Going, Going, gone: is animal migration disappearing. PloS Biology, 6: 13611364. Wiles G.J. and Brooke A.P. (2009) Conservation threats to bats in the Pacic Islands and insular Southeast Asia, pp.405459 in Island Bats: Evolution, Ecology, and Conservation, edited by T.H. Fleming and P.A. Racey. Chicago University Press, Chicago. Wiles G.J. and Fujita M.S. (1992) Food plants and economic importance of flying foxes on Pacific islands. pp. 2435 in Pacific island flying foxes: Proceedings of an international conservation conference, edited by D.E. Wilson and G.L. Graham. US Fish and Wildlife Service Biological Report 90(23). US Department of the Interior, Fish and Wildlife Service, Washington DC. Williams N.S., McDonnell M.J., Phelan G.K., Keim L.D. and van dee Ree R. (2006) Range expansion due to urbanization: Increased food resources attract Grey-headed Flying-foxes (Pteropus poliocephalus) to Melbourne. Austral Ecology, 31: 190198. Wilson D.E. and Graham G.L (eds) (1992) Pacific Island flying-foxes: Proceedings of an International Conservation Conference. United States Department of the Interior, Fish and Wildlife Service, Washington. Wolf C.M., Griffith B., Reed C. and Temple S.A. (1996) Avian and mammalian translocations: update and reanalysis of 1987 survey data. Conservation Biology, 10: 11421152.
References
135
Appendices Appendix 1
Data sources, time-period of records and number of records of Pteropus alecto and P. poliocephalus. Point records are those retained after screening for incomplete, unreliable, duplicate or misidentified information. Museum sources held specimen records; atlas sources held biological surveys, incidentals and literature records; and all other sources were based on biological surveys.
P. alecto Source Period Raw total Screened point records Total Roost Other Raw total P. poliocephalus Screened point records Total Roost Other
Atlas of New South Wales Wildlife (New South Wales 18632007 Department of Environment, Climate Change & Water) Atlas of Victorian Wildlife (Victorian Department of Sustainability & Environment) 18842005
110
76
76
1964
1212
1212
1340
154
30
124
Australia Bird & Bat 19852007 1491 Banding Scheme (ABBBS) Australian Museum 18792005 109 47 235 43 133 13 545 12 116
18 40 17 9 43 52 4 283 5 1
15 1 0 2 33 2 0 283 0 1
3 39 17 7 10 50 4 0 5 0
5831 176 41 117 155 125 7 1058 121 3
186 95 17 12 136 33 2 429 41 1
98 1 0 1 112 1 1 429 0 0
88 94 17 11 24 32 1 0 41 1
Australian National Wildlife 19611995 Collection (CSIRO) International museums Published & unpublished literature Queensland Museum South Australian Museum Surveys & incidentals Victorian Museum Western Australian Museum 18431991 18782001 19052000 19051930 19752007 18691996 1960
WildNet Fauna Data (Queensland Department of 19242006 Environment & Resource Management) TOTAL
588
322
21
301
705
188
15
173
3442
870
358
512
11643 2506
688
1818
ABBBS collectors who made records available were C. Tidemann, H. Spencer, J. Welbergen, K Parry-Jones, S. Klose and C. Palmer. The large reduction in the number of records for the Australian Bird and Bat Banding Scheme was due to multiple individuals being banded at the same location during the same year. International museums that held specimens were Smithsonian Institute National Museum of Natural History (Washington DC, USA), The Natural History Museum (London, UK), Field Museum (Chicago, USA), American
Appendices
136
Museum of Natural History (New York, USA), Zoologisches Museum Berlin (Berlin, Germany), Natural History Museum (Oslo, Norway), Museum of Comparative Zoology Harvard University (Cambridge, USA), Los Angeles County Museum (Los Angeles, USA), Royal Ontario Museum (Toronto), University of Washington Burke Museum of Natural History and Culture (Seattle, USA) and University of Kansas Natural History Museum (Lawrence, USA). References: Andersen 1912, Bartholomew et al. 1964, Birt 2005, Churchill 1998, Clancy 2001, Collett 1887, Dobson 1878, Eby 2002a, Eby 2003, Eby 2004, Eby & Palmer 1991, Eby et al. 1999, Hall 2002, Hall & Richards 2000, Lunney & Leary 1985, Nelson 1963, Nelson 1964, Nelson 1965, Pallin 2000, Ratcliffe 1931, Ratcliffe 1932, Ratcliffe 1938, Rhodes & Hall 1997, Strahan 1994, Tidemann 1999. Researcher and survey years (records not in any other data source) P. Eby (19942006), G. Richards (19611990), L. Hall (19851995), P. Birt (19962005), B. Roberts (20052007), G. ONeill (20002003), T. Pearson (20062007) and D. Lowe (20052007).
Appendices
137
Appendix 2
A re-evaluation of the northern distributional limit of the grey-headed flying-fox, Pteropus poliocephalus (Published in the Australasian Bat Society Newsletter, November 2008)
Authors: Billie J. Roberts, Carla P. Catterall, John Kanowski and Peggy Eby
Flying-foxes (family Pteropodidae) forage at night, roost during the day in large communal camps and routinely move large distances (100s of kms) in response to changing food availability. The grey-headed flying-fox (Pteropus poliocephalus) is a nationally vulnerable species endemic to coastal eastern Australia. Pteropus poliocephalus is a highly mobile animal that regularly moves between campsites across an expansive range. It has been frequently suggested that the distribution of this species has changed considerably during the past century. Several articles state that the northern limit of P. poliocephalus has contracted by over 750 km since the 1900s and that the current breeding range does not extend north of Maryborough in south east Queensland (Tidemann 1999, Tidemann and Nelson 2004, Tidemann et al. 2008). This contraction in the northern part of P. poliocephalus range has been variously cited as evidence for the emerging impacts of climate change (Tidemann 1999, Hughes 2003, Tidemann and Nelson 2004, Welbergen et al. 2007), and attributed to increased competition with the southward-expanding P. alecto (Tidemann 1999, Hall and Richards 2000, Eby 2006, van Dyck and Strahan 2008). However, these changes to the distribution of P. poliocephalus have not been examined systematically, and the extent of change is unclear. Here we re-assess the nature and extent of historical change to the northern limit of P. poliocephalus and assess whether the results support a conclusion that its range has contracted to the north. Locality records near the northern limit of P. poliocephalus distribution were collated from museum records, wildlife atlases, ABBBS and literature records. Records were scrutinized for accuracy and filtered to include only those coastal from the escarpment and north of 26oS (see Figure A4.1). As records within this part of the species range are limited, possibly due to large areas being unpopulated and inaccessible, the present distribution was obtained from field surveys including a current satellite telemetry study initiated in south-east Queensland.
Appendices
138 The results suggest that any change in the northern limit of P. poliocephalus is questionable. Past records show that there have been consistent observations of P. poliocephalus using the mid-northern extent of their range, from Mackay to Bundaberg (Table A4.1). Close scrutiny of records in the far north of the species range (north of Mackay) revealed that these are either questionable or incorrect. For example, the 1895 museum specimen collected on the Herbert River was found to be misidentified (Figure A4.2), and this record had contributed greatly to the impression that P. poliocephalus once occurred further north than at present.
Figure A2.1 Distribution of the grey-headed flying-fox (Pteropus poliocephalus) in eastern Australia (Map Wikipedia). The study area in the northern extent has been highlighted.
Figure A2.2 Photo of a museum specimen stated to be a grey-headed flying-fox (P. poliocephalus) collected by A.J. Boyd in 1895 from the Herbert River in Qld. The specimen is not a P. poliocephalus due to the lack of any fur on the lower legs. From the size and markings on the face this specimen is likely to be a spectacled flying-fox (P. conspicillatus).
Appendices
139
Table A2.1 Summary of the key records of P. poliocephalus in its northern range (north of latitude 26 S).
o
SOURCE Australian Museum Dobson 1878 WILDNET WILDNET WILDNET QLD Museum Collett 1887 Ratcliffe 1932 Ratcliffe 1932 SA Museum EPA survey Ratcliffe 1931 Nelson 1963 Ratcliffe 1932 ABBBS Ratcliffe 1932 L Hall survey L Hall survey WILDNET Birt 2005 WILDNET WILDNET WILDNET WILDNET WILDNET WILDNET Ratcliffe 1931 WILDNET Ratcliffe 1938 Birt 2005 EPA survey Nelson 1965
a
YEAR
a
GENERAL LOCALITY Herbert River Cape Upstart, Bowen Whitsunday Springs Resort, nr Bowen Proposed resort site, Midge Point Seaforth, nr Mackay Eungella Mackay Byfield Barmoya, nth Gladstone Fitzroy River, Fitzroy Valley Rockhampton region Rockhampton Rockhampton Raglan Targinie, 20km nth Gladstone Yarwun, nr Gladstone Wiggins Island Gladstone Gladstone 10km NW Gladstone CBD Turkey Beach Eurimbula National Park 1770-Agnes Water area Eurimbula beach-Agnes Waters area Deepwater National Park Deepwater National Park Miriam Vale Baffle Creek Granite Creek, general area Watalgan Bundaberg Harriot Island, Bundaberg Bundaberg
LAT -18.5167 -19.7654 -19.9512 -20.6561 -20.8984 -21.0800 -21.1443 -22.8205 -23.1195 -23.3330 -23.3651 -23.3755 -23.3755 -23.6621 -23.7167 -23.8174 -23.8226 -23.8439 -23.8439 -24.0843 -24.1544 -24.1984 -24.2068 -24.2818 -24.3151 -24.3734 -24.4597 -24.5740 -24.6367 -24.8734 -24.8734 -24.8830
LONG 146.3000 147.8252 148.1178 148.7067 148.9594 148.2900 149.1886 150.6514 150.5580 150.4167 150.4511 150.5124 150.5124 150.8176 151.1000 151.1259 151.2111 151.1747 151.1747 151.6499 151.7715 151.8927 151.8761 151.9177 151.9427 151.9594 151.9488 151.5370 152.0041 152.3469 152.3469 152.2670
1895 1843 1990 1990 1882 1991 1882 1929 1929 1930 1990 1929 1960 1929 2002 1929 1990 1995 2002 2000 1997 1982 1975 1998 1975 1983 1930 1995 1930 1998 2003 1961
b* b**
This museum specimen was found to be misidentified (see Figure A4.2). The specimen is not a grey-headed flying-fox (P. poliocephalus) due to the lack of any fur on the lower legs. This species is more likely a spectacled flying-fox (P. conspicillatus). b Record questionable as collector and detail of the observation not available. Both records also from an unpublished consultant report related to the development of a resort. * Sinclair Knight & Partners (1989) Whitsunday Springs Resort Initial Advice Statement on Environmental Effects. Unpublished report to Roach Investments Pty Ltd. ** Gutteridge Haskins and Davey (1990). Saros Resort Impact Assessment Study Report. Unpublished report to Pioneer Shire Council for Whitsunday Gold Pty Ltd.
Appendices
140 Recent field surveys and satellite telemetry have confirmed that P. poliocephalus continue to use this far northern part of its distribution (Table A4.2). In October 2008, 500 P. poliocephalus were observed in Finch Hatton (1 hr west of Mackay) including males and females with young. There are additional data by the landholders to suggest that P. poliocephalus have used this site annually since 1990 (D. Lowe pers obs). Recent satellite telemetry also shows P. poliocephalus using camps as far north as Turkey Beach (in the Gladstone region) between July and October 2008. At the time of publication (31/10/08) four of these tagged P. poliocephalus remained north of latitude 26oS. This confirms that the records in the northern limit are not just occasional records and that P. poliocephalus continued to use the full extent of its historical distribution as recently as 2008. To help ensure the conservation of northern populations, there is need for further investigation into the location, composition and temporal usage patterns of flying-fox camps in the Mackay region.
Table A2.2 List of recent (2008) records of P. poliocephalus in the northern part of its distribution. GENERAL LOCALITY
LAT
LONG
COMMENT
Five hundred P. poliocephalus incl. females with young Finch Hatton, observed 29/10/08. Additional reports of P. poliocephalus -21.1984 148.6011 nr Mackay occupying this camp annually since 1990 typically between July and May; numbers fluctuate. One tagged P. poliocephalus used this camp for 4 days in Turkey Beach -24.0843 151.6499 July. Also reports of large numbers of P. poliocephalus in April 2007 and 2008. Two tagged P. poliocephalus camped in this area during Miriam Vale -24.3734 151.9594 August and September, and stayed for 31 and 35 days. Two tagged P. poliocephalus used this colony briefly in July Agnes Water -24.1984 151.8927 and August, and stayed for 2 and 7 days. Five tagged P. poliocephalus used this camp between July and Baffle Creek -24.4597 151.9488 September. Number of days at this camp ranged from 2 to 50. Bundaberg P. poliocephalus regularly occupies this camp. Fifty P. -24.8734 152.3469 poliocephalus present on 16th October. Records of up to 500 P. poliocephalus using this camp annually since 1998. Tens of thousands of breeding P. poliocephalus observed on 15th October. Five tagged P. poliocephalus have used this -25.6388 152.3289 camp since the beginning of July and the numbers of days at the camp has varied between 2 and 119 days. Four of these tagged bats currently (as of 31/10/08) using the camp.
Woocoo
To summarise, there is no evidence of a change in the northern distributional limit of P. poliocephalus. However, it is unknown if this species frequency of use of its northern range has changed over time. Analyses of past and present patterns of occurrence of P.
Appendices
141 poliocephalus throughout its latitudinal extent are under way, and the findings will be published elsewhere.
References Eby P. (2006). Draft National Recovery Plan for the Grey-headed Flying-fox Pteropus poliocephalus. NSW Department of Environment and Conservation, Sydney. Hall L.S. and Richards G.C. (2000). Flying-foxes, Fruit and Blossom Bats of Australia. University of New South Wales Press, Sydney, 135 pp. Hughes L. (2003). Climate change and Australia: Trends, projections and impacts. Austral Ecology 28: 423443. Tidemann C.R. (1999). Biology and management of the Grey-headed Flying-fox, Pteropus poliocephalus. Acta Chiropterologica 1: 151164. Tidemann C.R. and Nelson J.E. (2004). Long-distance movements of the grey-headed flyingfox (Pteropus poliocephalus). J. Zool. Lond. 263: 141146. van Dyck S. and Strahan R. (eds) (2008). The Mammals of Australia. New Holland Publishers, Sydney, pp 444445. Welbergen J.A., Klose S.M., Markus N. and Eby P. (2007). Climate change and the effects of temperature extremes on Australian flying-foxes, Proceedings of the Royal Society B, 275: 419425.
Appendices
142
Appendix 3
Grid records for Pteropus alecto (N = 222) in coastal eastern Australia over four time-periods; 18431950, 19511990, 19912000 and 20002007. Each grey cell is 50 x 50 km. 1843 1950 1951 1990
-15o
500 km
-20o
-25o -30o
-35o
1991 2000
2001 2007
-15o
-20o
-25o -30o
-35o
Appendices
143
Appendix 4
The distribution of Pteropus poliocephalus (N = 368) in coastal eastern Australia in 50 x 50 km grids (grid records) over four time periods; 18431950, 19511990, 19912000 and 20002007.
18431950 -15o
19511990
500 km
-20o
-25o -30o
-35o
19912000 -15o
20012007
-20o
-25o -30o
-35o
Appendices
144
Appendix 5
Orientation of solar powered satellite transmitters on flying-foxes.
A: Collar design 1, a 12 g solar powered PTT with the bottom of the unit mounted to the collar and the solar array facing down when the animal was roosting (Photo T. Holmes). This orientation did not allow sufficient recharging of the unit and therefore number and accuracy of fixes were poor. B: Collar design 2, a 12 g solar powered PTT attached to the collar on its side with the solar array orientated towards the sky with an angled antenna (Photo G. Bottroff). This orientation reduced recharge time and improved power and accuracy and was the primary design used in this study.
Appendices
145
Appendix 6
Data summary: individuals characteristics, distances moved and use of roost sites over several time -periods, for 14 satellite-tracked Pteropus poliocephalus. All distances are between day roost sites, except for the longest day-night step, which can be between day roost and night feeding sites
Flying-fox number Capture site Weight (g) Forearm length (mm) Date of deployment End date Transmission days1 No. of days with location data2 Useable roost fixes3 Net overall displacement km4 Max displacement km/days5 Cumulative overall displacement km6 Max. step size km in all data/days7 Median step size km in 2 days/N 8 Min. step size km in 2 days/N 9 Max. step size km in 2 days/N 10 Longest day-night step km in 2 days11 N for 1-week periods Net displacement 78188 Stafford 660 159
18/10/07 31/10/07
78189 Stafford 660 159

17/10/07 28/1/08
78190 Canungra 739 172

16/10/07 18/4/08
78191 Canungra 696 164

16/10/07 11/6/08
78186B Dunwich 845 164

20/6/08 9/12/08
81072 Fraser 827 171

30/6/08 14/3/09
81073 Dunwich 815 170

22/6/08 26/12/08
81074 Dunwich 825 165

22/6/08 18/11/08
80187 Dunwich 822 168

22/6/08 27/11/08
80188 Fraser 740 162

30/6/08 9/2/09
80189 Fraser 789 160

30/6/08 1/4/09
80190 Fraser 798 162

1/7/08 7/12/08
80191 Dunwich 615 156

21/6/08 28/3/09
80192 Fraser 764 166

30/6/08 24/9/08
Mean 175 78 149 207
SD 75 38 102 190
13 8 9 183 183/5 195 125/2 125/1 125/1 125/1 125 2 98
101 41 75 183 195/39 241 110/5 30/1 30/1 30/1 44 14 21
182 54 45 158 377/137 702 190/5 230 25 26
235 106 324 0 454/53 1025 193/5 51/4 22/4 155/4 205 34 32
169 55 50 284 576/5 1121 576/5 500 19 60
254 94 122 378 855/119 1562 349/17 85/2 36/2 134/2 259 29 53
184 99 219 739 739/94 1102 335/5 23/3 13/3 146/3 336 26 37
146 76 120 278 327/95 840 165/5 13/10 12*/10 96/10 175 21 31
155 82 202 63 63/19 172 42/2 15/3 15/3 42/2 42 22 5
219 87 128 119 329/209 763 199/3 31/7 18*/7 166/7 166 29 24
271 125 255 312 547/181 1652 166/2 40/9 29/9 166/9 166 39 39
156 82 161 78 234/64 980 126/3 45/9 3*/9 103/9 123 22 27
277 154 322 33 93/173 489 49/2 10/14 4/14 49/14 49 40 12
84 35 59 94 94/2 194 94/2 94/1 94/1 94/1 79 11 17
788
491
47 33 109 179
36 38 48 126
34
23
Appendices
146
km/week12 No. roost sites used 2.0 1.0 1.3 1.4 1.0 1.4 1.4 1.8 1.4 1.4 1.5 in 1 week13 Cumulative km in 1 98 21 26 32 64 52 42 39 8 26 42 week14 N for 5-week 2 5 6 4 6 5 4 4 6 8 periods Net displacement 74 51 59 259 149 150 117 16 89 105 km in 5 weeks12 No. roost sites used 2.5 2.2 2.7 2.5 2.7 3.0 4.0 2.3 2.8 3.3 in 5 weeks13 Cumulative km in 5 110 137 153 280 210 220 194 43 117 206.5 weeks14 N for 10-week 1 2 3 2 3 2 2 2 3 4 periods Net displacement 195 18 114 466 301 372 127 32 169 186 km in 10 weeks12 No. roost sites used 4.0 2.5 4.3 3.5 4.0 5.0 6.5 3.0 4.7 5.0 in 10 weeks13 Max displacement 195 35 454 576 679 627 208 63 199 300 km in 10 weeks15 Max. step size km 110 35 193 576 349 335 165 42 199 166 in 10 weeks16 Cumulative km in 216 125 306 561 419 537 388 86 234 413 10 weeks14 N for 20-week 1 1 1 1 1 1 1 1 2 periods Net displacement 27 120 333 82 742 240 64 90 165 km in 20 weeks12 No. roost sites used 3.0 2.0 6.0 6.0 9.0 12.0 5.0 10.0 8.5 in 20 weeks13 Cumulative km in 135 120 1121 436 1074 776 172 618 826 20 weeks14 N for 30-week 1 1 1 1 periods Net displacement 107 576 158 357.00 km in 30 weeks12 No. roost sites used 10.0 10.0 12.0 10.0 in 30 weeks13 Cumulative km in 919 821 703 1059 30 weeks14 1 The total number of days from the date of deployment until the unit ceased functioning or until bat activity ceased (end date).
1.8 47 4 96 3.5 251 2 147 6.0 232 126 503 1 56 9.0 1005 -
2.0 12 8 21 2.9 61 4 40 4.3 54 46 122 2 41 6.0 245 1 36 10.0 489
1.4 18 2 38 3.0 83 1 72 5.0 94 94 165 -
1.5 38
0.3 23
94 2.9 159
66 0.5 74
172 4.4 286 187 313
136 1.1 226 154 167
178 7.0 593
209 3.0 390
247 10.4 798
219 0.9 217
Appendices
147
2 3
The number of days throughout the transmission period that bats were detected during daylight hours (between sunrise and sunset). The total number of fixes used for analysis; this included only securitised records with accuracy <2 km and located at a feasible roost site. 4 The distance in km between the roost at which the individual was captured and its last recorded roost, during the lifespan of the transmitter. 5 The furthest distance between any two recorded roosts during the lifespan of the transmitter and the no. of days between those fixes. 6 The sum of all sequential distances between consecutive recorded roost sites throughout the study. 7 The maximum inter-roost distance across all consecutive pairs of different recorded roosts during the lifespan of the transmitter, and the no. of days between those fixes; (total N = 167 steps across all individuals). 8 The median inter-roost distance across all consecutive pairs of different recorded roosts that were separated by 2 days, and the no. of such steps for each individual. 9 The minimum inter-roost distance across all consecutive pairs of different recorded roosts that were separated by 2 days, and the no. of such steps for each individual; asterisked records show movements involving temporary roosts. 10 The maximum inter-roost distance across all consecutive pairs of recorded roosts that were separated by 2 days, and the no. of such steps for each individual. 11 The maximum distance between consecutive useable fixes, recorded as either night feeding locations or day roost sites, which were separated by 2 days. 12 The distance in km between the first and last roost sites during five time-periods; one week, five weeks, 10 weeks, 20 weeks and 30 weeks; for each time-period this measurement is the average across N (as shown) timed movements for each individual. 13 The average number of different recorded roost sites during five time-periods; one week, five weeks, 10 weeks, 20 weeks and 30 weeks; for each time-period this measurement is the average across N (as shown) timed movements for each individual. 14 The sum of all sequential distances between consecutive recorded roost sites during five time-periods; one week, five weeks, 10 weeks, 20 weeks and 30 weeks; for each time-period this measurement is the average across N (as shown) timed movements for each individual. 15 The furthest distance between any two recorded roosts during time-periods of 10 weeks; this measurement is the maximum across N (as shown) 10-week movements for each individual. 16 The maximum inter-roost distance across all consecutive pairs of different recorded roosts during time-periods of 10 weeks; this measurement is the maximum across N (as shown) 10-week movements for each individual.
Appendices
148
Appendix 7
Frequency distribution of the weekly net displacement of all 13 individuals that had 12 or more weeks of data.
10 8 20
A 78189
B 78190
25 20 15
C 78191
Frequency
15
6 10 4 2 0 15 5
10 5 0
D 78186B
20
E 81073
15
F 81074
15
Frequency
10 10 5 5
10
0 20
0 30
G 80187
H 80191
20
I 81072
25 20 15 10 15
15
Frequency
10
10
5 5 0 25 0
0 20
J 80188
15 20 15 10 10
K 80189
15
L 80190
Frequency
10
5 5 5 0 0 0-5 5-50 51- 101- 201- 301- 401- 501100 200 300 400 500 600 0-5 5-50 51- 101- 201- 301- 401- 501100 200 300 400 500 600
0 10
M 80192
Net displacement (km)
Frequency
0 0-5 5-50 51- 101- 201- 301- 401- 501100 200 300 400 500 600
Appendices
149
Appendix 8
Data summary: individuals characteristics, number of fixes and the feeding distances moved by 14 Pteropus poliocephalus at two different sampling units; all records and independent records. Also see Appendix 5.
Bat ID # Feeding distance rank-order Date of deployment Transmission days1 Total feeding fixes2 Mean feeding visit duration (days)/ N N for all records4 Median feeding distance (km) Mean feeding distance (km) Standard error Max. feeding distance (km) Min. feeding distance (km) Mean roost population size x10 / N Mean feeding forest patch x103 (ha) N for independent feeding records Median feeding distance (km) Mean feeding distance (km) Standard error Max. feeding distance (km) Min. feeding distance (km) Mean roost population size x10 / N Mean feeding forest patch x103 (ha)
1 2 3 3 5 3 3
78188 1 13 10 1/1 1 5 5 5 5 20/1 10.0 1 5 5 5 5 20/1 10.0
78189 6 101 56 5/9 9 10 11 17 4 20/1 7.2 9 10 11 4.2 17 4 20/1 7.2
78190 7 182 29 2/6 6 10 13 32 3 4/5 6.9 6 10 13 32 3 4/5 6.9
78191 5 235 139 13/12 24 7 10 1.4 30 5 4/22 5.1 11 12 13 4.1 31 2 6/9 5.3
78186B 13 20/6/08 169 105 17/7 10 17 18 2.3 36 8 28/2 6.9 5 14 18 36 8 28/2 9.5
81072 12 30/6/08 254 140 15/5 14 12 15 2.6 32 3 64/8 10.0 8 11 14 30 3 62/5 10.0
81073 11 22/6/08 184 192 17/8 15 9 14 2.1 28 5 8/15 6.5 9 9 13 28 5 10/9 6.0
81074 9 22/6/08 146 147 7/6 7 14 13 21 6 5/1 4.0 6 14 13 21 6 5/1 4.6
80187 2 22/6/08 155 169 36/4 14 4 6 1.9 24 2 9/14 9.4 5 8 12 31 2 15/5 9.8
80188 8 30/6/08 219 150 8/12 17 12 13 2.0 31 2 50/6 8.5 13 12 13 4.8 31 2 50/6 8.7
80189 10 30/6/08 271 228 11/14 21 11 14 2.2 50 5 60/11 8.4 14 11 15 4.1 50 4.7 56/8 7.7
80190 14 1/7/08 156 169 6/10 12 14 20 5.4 56 2 100/4 8.6 10 14 21 2.2 56 2 100/3 8.5
80191 4 21/6/08 277 308 15/14 22 9 9 0.9 15 2 8/17 7.1 14 9 9 7.1 15 2 9/11 6.9
80192 3 30/6/08 84 71 17/2 1 7 7 7 7 50/1 10.0 1 7 7 7 7 50/1 10.0
Mean
SD
18/10/07 17/10/07 16/10/07 16/10/07
175 128 12 12 10 12 27 4 30 8 8 10 13 28 4 29 8
75 66 9 7 4 4 14 2 29 2 4 3 4 14 2 27 2
4 5
The total number of days from the date of deployment until the unit ceased functioning or until bat activity ceased (end date). The total number of feeding fixes throughout the transmission period that bats were detected during night time hours (between sunset and sunrise). Average duration of visits by each individual to feeding sites, defined by the dates of the first and last fixes in consecutive duty cycles from a particular individual at a particular feeding site (includes time when flying-foxes used more than one feeding site per night); this measurement is the average across N (as shown) visits for each individual. Total N = 110 visits to different sites. The number of feeding records per individual bat when all 173 records were included. The number of feeding records per individual bat when only the 112 independent records included .
Appendices
150
Appendix 9
Correlations (Pearson) between landscape attributes of roosts, feeding sites and random points in the landscape at two different sampling units; all feeding records and independent feeding records
Elevation All feeding records Roost sites Elevation % Forest % Built % Open Feeding sites Elevation % Forest % Built % Open Random points Elevation % Forest % Built % Open Independent feeding records2 Roost sites Elevation % Forest % Built % Open Feeding sites Elevation % Forest % Built % Open
1
% Forest
% Built
% Open
1.00 0.41** -0.42** -0.06 1.00 0.21** -0.11 -0.13 1.00 0.29** -0.28** -0.12
1.00 -0.66** -0.55**
1.00 -0.26*
1.00
1.00 -0.16* -0.90**
1.00 -0.21**
1.00
1.00 -0.46** -0.77**
1.00 -0.21*
1.00
1.00 0.39** -0.43** -0.12 1.00 0.14 -0.10 -0.22*
1.00 -0.64** -0.42**
1.00 -0.34*
1.00
1.00 -0.23** -0.88**
1.00 -0.20*
1.00
* P < 0.05; **P < 0.01 1 All records; Roost = 173, Feeding = 173, Landscape = 100 2 Independent records; Roost = 112, Feeding = 112, Landscape = 100
Appendices
151
Appendix 10
Food for thought: Flying-foxes and bush regeneration (Published in Big Scrub Landcare Newsletter, February 2011)
Authors: Dr Kristin den Exter, Billie Roberts, Angus Underwood, Dr Len Martin
Introduction
Flying-foxes are animals of extraordinary ecological and economic importance throughout forests of the tropics playing an essential role as forest pollinators and seed dispersers (Fujita and Tuttle 1991). Flying-foxes are considered a keystone species, that is a species whose effect is large, and one that is disproportionately large relative to its abundance (Power et al. 1996). Flying-foxes are the original bush regenerator with enviable characteristics such as a big wingspan, and the ability to fly and forage over large distances thereby pollinating and spreading the seed of numerous species, including rainforest species (e.g. Eby 1995). Three species of mega bats (Family: Pteropodidae) otherwise known as flying-foxes occur in the Big Scrub; the Grey-headed flying-fox (Pteropus poliocephalus), the Black flying-fox (Pteropus alecto); and the Little Red flying-fox (Pteropus scapulatus) (Figure A10.1). Whilst the Black flying-fox has expanded its range poleward (DECCW 2009) and the Little Red is highly nomadic often occurring in great numbers, the Grey-headed flying-foxes numbers are in decline and this species is listed as vulnerable under the NSW Threatened Species Act 1995 and the Commonwealth Environment Protection and Biodiversity Conservation Act 1999. The principle reason for this population decline is habitat loss. Habitat loss leading to restricted food supplies and a lack of suitable roosting sites (DECCW 2009). Clearly there is a role for human-assisted bush regeneration to help address this problem; however this may not always be a simple task.
Appendices
152
Figure A10.3
Distribution of the flying-foxes in Australia (Source: Roberts and Hall pers comm .
2010)
The camps of flying-foxes play a number of roles in the life histories of the animals and there is a growing awareness of the need to actively manage these sites to ensure the longevity of the vegetation and its suitability as roosting habitat, and to reduce conflict between flyingfoxes and humans. Although there are a number of sites that have a history of active management, these experiences have not been compiled and little information is available to guide others. A recent forum held in July 2010 in Brisbane was convened to discuss the management and restoration of flying-fox camps. The forum brought together over 80 people with an interested in camp management, focusing on maintenance and regeneration of camp vegetation. To build upon this recent forum, a panel was convened at this years Big Scrub Rainforest Day to discuss bush regeneration in flying-fox camps within the Big Scrub region. The aim of this paper is to take these discussions even further. Specifically we examine how to conduct bush regeneration underneath flying-foxes, particularly camps in small rainforest remnants, to reduce disturbance to the species and aid restoration of the vegetation communities.
Flying-fox feeding and roosting habitat
Flying-foxes commonly use forests and woodlands east of the escarpment in NSW and Qld as feeding habitat (Hall and Richards 2000). Current diet lists for the flying-foxes contains over
Appendices
153 100 species of trees and vines (e.g. Eby 1995). Approximately half of the plants in their diet are fleshy fruits from sub-tropical and warm temperate rainforest. The balance consists of blossoms from a range of vegetation communities dominated by Eucalypts, Melaleucas and Banksias. Flying-foxes also use a range of introduced plants most notably commercially grown fruits. Flying-foxes travel up to 50 km per night to search for food, although typically this distance is less than 20 km (Eby 1991a; Tidemann 1999). Flying-foxes accommodate shifts in the volume and locations of food by migrating, often over large distances (Tidemann and Nelson 2004). Flying-foxes roost during the day in large communal roosts or camps. Numbers can range from a few hundred to over 50,000 (Ratcliffe 1931; Eby et al. 1999), and are driven by the amount of food resources in the surrounding area (Ratcliffe 1931; Parry-Jones and Augee 2001). Flying-foxes exhibit strong fidelity to traditional campsites, with some roosts being occupied for >100 years. Although flying-foxes display a degree of flexibility in habitat choice, there are some key camp elements that are required (Ratcliffe 1931; Nelson 1965; Hall & Richards 2000; ParryJones and Augee 2001; Eby 2002b; Peacock 2004; Roberts 2005; Snoyman 2008) for example:
Camps need to be positioned within nightly commuting distance (generally, <20 km) of sufficient food resources.
Camps are situated in the coastal lowlands. Camps are commonly located in closed forest, Melaleuca swamps or stands of Casuarina and are generally found near rivers or creeks.
Camps occur in vegetation ranging from continuous forest to remnants as small as 1 ha Typically canopy height of at least 5 m. The microclimate characteristics of camps, such as temperature and humidity of sites, appears to be important.
There are eight known camps used by flying-foxes within the Big Scrub region. When examining the characteristics of these camps (Table A10.1) it is notable that most occur in small remnant rainforest patches. The largest continuously occupied camp is Rotary Park at 12.8 ha and the smallest is Lumley Park at 2.2 ha. These small patches present challenges to bush regenerators, managers and flying-foxes alike.
Appendices
154
Table A10.1: Attributes of known flying-fox camps in the Big Scrub, Northern NSW (Source: Roberts 2007) Boat Harbour Nature Reserve H 1940 R summer 2006 bk, gh 0 10,500 2,000 no 16.7 12 100 river rural SE 1020, with emergents to 25 m rainforest spp eucalyptus spp mixed 7080 na na Wilson Nature Reserve N 2003 R (twice) summer (Jan) 2004 bk, gh 0 20,000 na no 25.9 10 30 small creek urban (medium density) SSW 1020 rainforest spp na rainforest spp 6070 na na Maguires Creek unk c. 1990 R unk mid 1990s bk, gh unk unk 2,0003,000 no 4.1 100 60 creek residential (low density) F 1020 rainforest spp rainforest spp rainforest spp 7080 na na
Rotary Park Type 1 Year established Recent usage 2 Season occupied Year last occupied Species 3 Minimum number Maximum number Typical number Known maternity site Patch size (ha) Elevation (m) Distance to nearest drainage (m) Drainage type Surrounding area Aspect 4 Typical tree height (m) Dominant tree type Dominant roost tree Understorey type % canopy cover Height roosting bks (m) Height roosting gh (m) N c. 2005 Con all year current bks, gh 2,000 30,000 5,00010,000 yes 12.8 10 adj small creek urban (high density) SW 2030 rainforest spp rainforest spp na 6070 815 1525
Currie Park H at least 1960s Con all year Sept 2005 bks, gh, lr 0 80,000 40,00050,000 yes 6.2 3 25 river urban (low density) W 12 privet blackbean weedy 2030 na na
Booyong T >100 years Ann not mid-winter 2005 bk, gh 0 14,000 6,0007,000 yes 11.3 15 60 river rural F 2025 rainforest spp rainforest spp na 7080 15+ 1015
Lumley H at least 1950s Con all year current bk, gh 0 32,000 6,0007,000 yes 2.2 120 20 creek urban (high density) F 17 rainforest spp rainforest spp rainforest spp 3040 1015 812
Davis Scrub H 1970s R spring/ summer 1992 bk, gh 0 unk unk no 13.5 172 1,000 small creek rural F 2030 rainforest spp rainforest spp rainforest spp 8090 na na
Appendices
155
1 Type N = New site established in the last decade; T = Traditional camp site. Camp used regularly for as long as we have records >100 years. Also often used as a maternity site; H = Historical site (c. 50 yr history). 2 Usage Con = Continuously occupied (year-round occupancy), Ann = Annual (the site had been occupied in 80% of years, but not continuously), Ir = Irregular (occupancy in 2080% of years), R = Rare (occupancy has been recorder in less than 20% of years 3 Species bks = Black Flying-fox (Pteropus alecto); gh = Grey-headed Flying-fox (P. poliocephalus); lr = Little Red Flying-fox (P. scapulatus) 4 Aspect F = flat land (no aspect)
Appendices
156
Bush regeneration in Big Scrub Flying-fox Roosts
A number of issues require consideration when undertaking bush regeneration in flying-fox roost sites, for example the timing of works, the paradox of weeds as habitat and weeds as threats to vegetation community structure and health. Patch size is a major consideration when determining the appropriate course of action. Many bush regenerators are of the view that bush regeneration needs to occur in continuously occupied camps because of the increased weed infestation that invariably occurs with flying-fox occupation. The effects of continuous occupation of flying-foxes in very small patches are well known. For instance roosting in small patches causes damage to the canopy, letting in extra light, and there is also thought to be an increase in available nutrients. This disturbance leads to favourable conditions for the invasion by many weeds species, some of which are introduced by flyingfoxes. In Rotary Park (a newly established camp) regenerators have documented the invasion of a new suite of weeds possibly brought in flying-foxes including Giant Devils Fig, seedling of seeded bananas, Cockspur, Guava, Cocos Palm, Alexandra Palms etc (Rosemary Josephs pers comm. 2010). At Lumley Park, flying-foxes are estimated to have had a significant impact on canopy health in approximately half the site. The extra light, nutrient and weeds make restoration of this already small patch, more difficult. However, the lower and mid layers of the forest have recovered well despite continuous use by flying-foxes with a substantial revegetation effort (Darren Bailey, pers comm. 2010). Addressing the influx of new weeds may mean that greater resources are required for weed control than just at maintenance level (Heidi Lunn, pers comm. 2010). In terms of flying-fox habitat, weeds may prove allies where core vegetation is already under pressure from roosting flying-foxes. Flying-foxes will roost in weed species, as they show no selection preferences for tree species (Roberts 2005). For example in Rotary Park, flyingfoxes have been observed roosting in mature Large Leaved Privets (Rosemary Josephs pers comm. 2010). Lower stratum weeds even morning glory, are known to be used during heat events. There are numerous examples of places where understory weeds have been removed and the habitat made unsuitable for Flying-foxes.
Appendices
157 If weeds are left in core remnant vegetation then the viability and ongoing succession of the community is likely to be affected. From the bush regeneration perspective, the long-term sustainability of the vegetation patch can only be achieved through weed control of the understory weeds. So how do we accommodate dual goals of restoring Big Scrub remnants with sustainable flying-fox roosts? Can we maintain and create flying-fox habitat without the help of weeds? Is it a matter of allowing (some) weeds to stay in some places but not others? The questions are complex and often hotly contested, with or without the added complexity of flying-foxes. When discussing this amongst practitioners it is clear that many now consider it unwise to be advocating a sudden and total removal of all weed species unless the site warrants it. There are many reasons for the staged, gradual and strategic removal of weed species, flying-fox habitat requirements are just another consideration to take into account for some sites. Opportunities arise over the longer time frame when continuously occupied camps are abandoned for a short period of time. It is at these times that bush regeneration work that could not be done under occupation can be acted on. The key is to have the resources to be able to act at that time. This may currently be the case for Lismores Currie Park. In larger patches it appears that neither the weed nor canopy health issues are quite as critical. For example, at Booyong flying-foxes impact on the site is considered to be minimal, due to the large remnant size, small flying-fox population and the fact the core roosting area shifts through time. Flying-foxes are thought to still bring in Cocos Palm, Devils Fig and White Passionflower to this site but according to bush regenerators the level is manageable (Darren Bailey pers comm. 2010). So it seems logical that if the issues are lessened in the larger remnants that we look to increase the patch size of our smaller sites, but the question is how, especially when many of these remnants have confined boundaries. The topic of how to increase the size of the smaller remnants provided for an interesting discussion between the panel and the audience at Big Scrub Day including actively planting non-rainforest species, for example fast growing Eucalypts, in areas within existing reserve boundaries but away from core remnant vegetation, and even allowing the weeds to take their course, thereby providing the kind of vegetation structure required for flying-fox roosts (i.e. emergent trees about a dense understorey). This issue becomes more complex when considering most of the Big Scrub remnants are Endangered Ecological Communities and
Appendices
158 contain a range of threatened plant species. Whatever the management approach is it must consideration the conservation of the range of species that are threatened in the area. Whether the goal is to manage human flying-fox conflicts or to increase vegetation resilience to flying-fox roosting, there is consensus that we need to look at expanding the current sites. Some bush regenerators estimate that, whether the goal is to manage human flying-fox conflicts or to increase vegetation resilience to flying-fox roosting, there is consensus that we need to look at expanding the current sites. Some bush regenerators estimate that in the Big Scrub region it may only take between 1020 years, depending on the site, maintenance regimes, and species selection, for plantings to be utilised by flying-foxes. Large scale planting of rainforest trees has already been shown to provide habitat for flying-foxes at Fernleigh. Planting of mixed species rainforest trees along a riparian strip has been ongoing at a site managed by Mark Dunphy since 1991, which has created an area that has been used as a roost site for up to 500 Black flying-foxes (Angus Underwood, pers. obs. 2009). The flyingfoxes are regularly present at the site, which is around 100 m wide, roosting in both remnant trees at the top of a gully, as well as the surrounding emergent planted trees including Cudgerie, Hoop Pine and Blue Quandong.
Recommendations
When considering expanding roost habitat it would also be useful to know more about which tree species are more resilient to flying-fox roosting, or indeed may respond to flying-fox presence. It is critical to any such endeavours that the resources currently allocated for core vegetation work must not be re-directed, as this work is essential to the long term sustainability of the vegetation community. More resources are required to assist in the expansion of the vegetation structure required for flying-fox roosts, and we argue here that continuously occupied, maternity roosts in small remnants should be the priority. In our view the best practice management approach for bush regeneration under continuously occupied flying-fox roosts would be: 1) make an assessment of the impacts to flying-foxes of vegetation structure change that may occur as a result of bush regeneration before undertaking any works; and 2) taking a staged approach to any weed removal. As well as the staged removal of weeds, a number of other strategies can be employed to minimise the impact of bush regeneration on roosting flying-foxes, including the timing of works (preferably between
Appendices
159 MayJuly), eliminating the use of machinery and camp monitoring. Firstly, lets look at the timing of works. Working in the breeding season is an obvious issue that needs to be considered. Other than hoping and waiting for the site to be abandoned (discussed above) there are some other options: 1) Not to carry out works under active flying-fox maternity roosts during the breeding season; 2) Not to carry out works within a reasonable distance of active flying-fox maternity roosts during the breeding season; and 3) To continue only necessary works and to do this as quietly as possible with continual monitoring for signs of disturbance/distress. Whilst it might seem an obvious course of action to go with options 1 or 2 this may present problems for some practitioners working on core vegetation weed control within continuously occupied maternity roosts. As with many protocols it may not be that a one size fits all approach will work, so any guidelines need to allow for some flexibility and feedback with regards to site context. Critical breeding and maternity season (SeptMarch) is also prime weed control season. Canopy and roost height can also be variable, for example Booyong has a high canopy compared with many camps in coastal swamp forests, and disturbance levels can be lower under some higher roosting camps. So in some cases perhaps option three is the practical way to go? In any case, if work is required directly under a camp this should be carried out in a way to minimise the impacts on the camp. Work team should be small in size, noise level and time within camp should be kept to a minimum. Using powered or noisy machinery is not such a complex issue. Machinery, such as chainsaws, brushcutters and generators that are likely to disturb flying-foxes should not be used in and around occupied camps, particularly during the breeding times. Hand tools are preferred, but we re-iterate that an assessment be made of the impacts to flying-foxes of vegetation structure change before undertaking such activities, that a staged approach be taken to any weed removal and that no disturbance should be made to trees in which flying-foxes roost. Whatever the course of action taken, it is important to be able to monitor what is happening in the flying-foxes camp in terms of behaviour etc as well as what is happening with respect to canopy and weeds. It is important that bush regenerators working in flying-fox camps know what to look for, for example to know what constitutes normal behaviour and what constitutes disturbance or distress. This may involve bringing in ecologists to talk with bush
Appendices
160 regeneration teams. The recent initiative by Byron Shire Council is one example which involved flying-fox ecologist Billie Roberts presenting two free seminars on flying-fox ecology to both council staff and members of the public. In addition to this a set of guidelines have been developed for bush regenerators working in and around flying-fox camps which aims to minimise the impacts on both flying-fox camps and the OH&S risks on workers. Finally lets turn to what bush regenerators can do to minimise the OH&S risks. A site OH&S risk assessment should be undertaken to determine the appropriate level of personal protective equipment required. It is currently the practice of some to wear an outer layer of clothes (at least an overshirt) while working under flying-foxes. If workers find injured or sick flyingfoxes local wildlife carers should be contacted immediately. Workers should not touch or handle flying-foxes. If a worker is bitten or scratched wash the wound thoroughly with soap and water for at least five minutes and inform a doctor immediately. The authors welcome feedback and response to this article, as we see it, the conversation has just begun.
References Fujita M.S. and Tuttle M. D. (1991). Flying-foxes (Chiroptera: Pteropodidae): Threatened Animals of Key Ecological and Economic Importance. Conservation Biology. 5: 455463. Power M.E., Tilman D., Estes J.A., Menge B.A, Bond W.J., Mills L.S., Daily G.,Castilla JC., Lubchenco J. and Paine R.T. (1996). Challenges in the Quest for Keystones. Bioscience, 46, 8: 609620. Department of Environment, Climate Change & Water NSW (2009). Draft National Recovery Plan for the Grey-headed Flying-fox Pteropus poliocephalus. Prepared by Dr Peggy Eby. Department of Environment, Climate Change and Water, NSW, Sydney. Eby P. (1991). Seasonal movements of Grey-headed Flying-foxes, Pteropus poliocephalus (Chiroptera: Pteropodidae), from two maternity camps in northern New South Wales. Wildlife Research 18: 547559. Eby P. (1995). Biology and Management of Pteropus in New South Wales. Species Report. NSW National Parks and Wildlife Service, Hurstville. Eby P (2002). Using New South Wales planning instruments to improve conservation and management of Grey-headed Flying-fox camps, pp. 240250 in Managing the Grey-
Appendices
161 headed Flying-fox as a Threatened Species in NSW, edited by P. Eby and D. Lunney. Royal Zoological Society of NSW, Mosman, NSW. Eby P., Richards G., Collins L., and Parry-Jones K. (1999). The distribution, abundance and vulnerability to population reduction of a nomadic nectarivore, the Grey-headed flying-fox Pteropus poliocephalus in New South Wales, during a period of resource concentration. Australian Zoologist 31: 240253. Hall L.S. and Richards G. (2000). Flying-foxes: Fruit and Blossom Bats of Australia. UNSW Press, Sydney. Nelson J.E. (1965). Movements of Australian flying-foxes (Pteropodidae: Megachiroptera). Australian Journal of Zoology 13: 5373. Parry-Jones K.A. and Augee M.L. (2001). Factors affecting the occupation of a colony site in Sydney, New South Wales by the Grey-headed Flying-fox, Pteropus poliocephalus (Pteropodidae). Austral Ecology 26: 4755. Peacock L. (2004). Roost preference of the Grey-headed Flying-fox. BSc (Hons) thesis. Faculty of Veterinary Science, University of Sydney. Ratcliffe F.N. (1932). Notes on the fruit bats (Pteropus spp.) of Australia. Journal of Animal Ecology 1: 3257. Roberts B. (2005). Habitat characteristics of flying-fox camps in south-east Queensland. BSc (Hons) thesis. Griffith University, Brisbane. Snoyman S. (2008). Micro-climate Preferences of the Grey-headed Flying-fox, Pteropus poliocephalus (Chiroptera: Pteropodidae), within the Sydney Region. BSc (Hons) thesis. Macquarie University, Sydney. Tidemann C.R. (1999). Biology and management of the Grey-headed flying-fox, Pteropus poliocephalus. Acta Chiropterologica 1: 151164. Tidemann C.R. and Nelson J. (2004). Long-distance movements of the Grey-headed flyingfox (Pteropus poliocephalus). Journal of Zoology London 263: 141146.
Acknowledgements Thanks to Rosemary Josephs, Darren Bailey, Heidi Lunn for providing comments on those tricky questions of bush regeneration in flying-fox colonies, thanks to Peggy Eby, Billie Roberts and Liz Gould for the Flying-fox Forum and to Lib Ruytenberg (WIRES Northern Rivers) for her comments on flying-fox behaviour.
Appendices

The Ecology & Management of The Grey-Headed Flying-Fox

Uploaded by

Copyright:

Available Formats

You might also like

The Ecology & Management of The Grey-Headed Flying-Fox

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Ecology & Management of The Grey-Headed Flying-Fox

Uploaded by

Copyright:

Available Formats

The ecology and management of the grey-headed flying-fox Pteropus poliocephalus

Photo Vivien Jones

Billie Roberts September 2012

References ............................................................................................................................. 118 Appendices ............................................................................................................................ 135

Research permits and animal ethics

Research permits and animal ethics

Publications produced from this research

Publications produced from this research

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

1.3.1 Status and distribution

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

12 1.4 Australian flying-foxes

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 1 General Introduction

Chapter 2 Latitudinal range shifts in Australian flyingfoxes: a re-evaluation

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

2.2.1 The species and their distributions

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

2.3.1 Number and distribution of observations

Chapter 2 Latitudinal range shifts in Australian flying-foxes

2.3.2 Distributional changes in Pteropus alecto and P. poliocephalus

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

(a) Point records

(b) Grid records

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Mean no. days <2oC 1939 49 na 7.5 2001 07

Annual rainfall (mm) 1939 49 2001 07

29.0 26.8 -1.7

18.2 14.5 -2.3

683 803 -67

Effective difference Most southern record 1921-1950 2001-2007

27.3 23.0 -4.3

15.9 14.6 -0.7

987 919 -140

Chapter 2 Latitudinal range shifts in Australian flying-foxes

2.4.1 Species-specific range dynamics and their possible causes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 2 Latitudinal range shifts in Australian flying-foxes

Chapter 3 Long-distance movements of P. poliocephalus