The social organisation of the grey-headed flying-fox, P. poliocephalus, was studied in the Tweed Valley (28º 31' S, 153º 32' E) in the north-eastern corner of N.S.W., Australia. The only other study on the roosting behaviour of this species (Nelson, 1965) was conducted in this same area. The field work was conducted between December and July from December 2000 to July 2003.

The Tweed Valley (1303 km2) is the ancient remains of a shield volcano with Mt. Warning (1156 metres) at the volcanic core. The steep rim of the surrounding ranges, or caldera have offered catchments for the Tweed River whose short tributaries have carved out the caldera (Figure 1). The Tweed Valley supports an abundant variety of rainforests, eucalypt forests and coastal wetlands. The biogeographic region that includes the Tweed Shire is regarded internationally as a significant centre of biodiversity (Kingston et al., 1999).

Figure 1. Location of the study area and main study site in northern NSW, Australia.
Inset: Mt Warning Caldera with the locations of the three study colonies (red dots); centre top: Dallis Park colony; centre: Uki colony; bottom right: Brunswick Heads Colony. figure composed from the Landsat 7 Picture Mosaic of Australia produced by the Australian Greenhouse Office (AGO), © Commonwealth of Australia, 2004.

On December 4th, 1823, John Oxley was the first European to see the Tweed Valley where upon exploring 7 miles upstream of the Tweed River he entered into his journal:

"A deep rich valley clothed with magnificent trees, the beautiful uniformity of which was only interrupted by the turns and windings of the river, which here and there appeared like small lakes. The background was Mt. Warning. The view was altogether beautiful beyond description. The scenery here exceeded anything I have previously seen in Australia." (from Steele, 1972)

Flora
The Macley-MacPherson overlap described by Burbidge (1960) identifies a region including the Tweed Shire as the transition between two phytographic zones. As a consequence the Tweed Shire has the highest plant diversity in New South Wales within a 2000 km section of the Australian east coast (e.g. Williams et al., 2001). The region supports both tropical and temperate species, many of which are at the limits of their range (McDonald & Elsol, 1984). As a result of land clearance for agricultural and residential land uses many species have become rare, vulnerable and endangered (see Schenk & Wallace, 1996).

Fauna
In terms of numbers of species present, the Tweed Shire has one of the highest vertebrate biodiversities of any region in Australia: The region supports more species of bird, fish, amphibian, and mammals than Kakadu (Northern Territory, Australia), and similar numbers of reptiles (Williams et al., 2001). Only in the wet tropics are similar numbers of species found within these groups.

Climate
The climate in the Tweed Shire is classified as sub-tropical maritime (Table 1). It experiences mild winters with evening minimums just below 10°C and warm summers with afternoon average maxima of just below 30°C. The Tweed has the highest average rainfall in New South Wales with an annual average of 1,701 mm. February and March are usually the wettest months and September the driest.

In this area there were three study colonies. The Dallis Park colony (DP), the Brunswick Heads colony (BH) and the Uki colony (UKI). All three colonies were home to two Pteropus species (P. poliocephalus and P. alecto). They were average to large colonies in size and contained approximately 7,000 (UKI), 16,000 (BH) and 30,000 (DP) bats.
Dallis Park, the main study site, was the focus of the study between December and July from December 2000 to July 2003. During these time periods, between one and three people were working in the colony all day until sunset between 4-6 times per week.
Dallis Park lies near the centre of the caldera (Figure 1), in a rural area just south of Murwillumbah (Figure 2). Permission was obtained from the landowner, J. H. Williams & Sons, a large primary producer and builder supplier, to conduct research on this land for the duration of the study. There was no access to the public.

Figure 2. View of Mt Warning (centre) in the Tweed Valley. The Dallis Park colony is indicated with the ellipse in the foreground. (photograph by J. Welbergen)

Figure 3. Aerial view of the Dallis Park colony. (photograph by J. Welbergen)
The colony is on a 2 hectares strip of swampland covered by tree species such as Paperbark (Melaleuca spp.) and Eucalypts (Eucalyptus spp.) (Figure 3). Since the land had not been maintained for over a decade it was also home to noxious weeds such as camphor laurel (Cinnamomum camphora) and lantana (Lantana camara).

Working in a Pteropus colony (Figure 4) is not for the faint of heart as it is one of the more hostile environments for aspiring behavioural ecologists. The Dallis Park colony is located in a swamp rife with mosquitoes and poisonous snakes (eastern Brown Snake Pseudonaja textilis; Tiger Snake, Notechis scutatus; Coastal Taipan, Oxyuranus scutellatus), and is home to the giant Stinging tree, (Dendrocnide excelsa (Urticacea)). In summer it is hot and humid and autumn rains transform the swamp into a veritable lake. To add insult to injury, one must endure the ever-present drizzle of guano.

Figure 4. Inside view of the Dallis Park colony. (photograph by J. Welbergen)

In the colony an elaborate network of paths was cleared with an axe and machete. The paths were maintained regularly at night after the bats had left the colony. The paths facilitated movement through the colony during the day and this further reduced disturbance to the bats. At no point were the paths more than 20 metres apart and this allowed any trees that were occupied by bats to be viewed from at least two angles.

The bats were quickly habituated to our presence early during the field seasons as evidenced by the fact that we could move about in the colony without causing them to leave their roosting positions.
A hide was built (Figure 5) in the centre of DP to enable unobtrusive observations on behaviour and to function as a base to store equipment and from which to conduct the research. The hide provided an excellent view of up to approximately 1500 individuals and bats could easily be observed individually binoculars or a Hi-8 Sony Handicam ©.

Figure 5. Research station inside the DP colony. (photograph by J. Welbergen)

Capture, mark and resighting techniques
An important part of the study involved the capture, mark and resighting of individuals, and the development of a viable system for catching individuals was one of the main challenges of the study. For further information on the capture, mark and resighting techniques, please contact J. Welbergen directly.