Publications

Appendix one: Resurvey of known pipistrelle sites affected by yellow crazy ant supercolonies

It is recommended that sites known to contain important foraging and roosting areas in 1998, that were subsequently infested with Yellow Crazy Ant supercolonies, be resurveyed to investigate the impact of this introduced species on the pipistrelle.

A. Sites

Sites of high or moderate Christmas Island Pipistrelle activity and roost areas (from Lumsden et al. 1999) that were subsequently infested with Yellow Crazy Ant supercolonies, and those that were located nearby (i.e. within foraging range).

B. Sampling Period and Climatic Conditions

For comparative purposes a resurvey of these sites needs to be conducted between 10 May and 20 June. Nights with rain periods are to be avoided due to possible suppression of bat activity and to the resultant reduced sensitivity of the bat detector microphone.

C. Sampling Techniques

These techniques are identical to those used by Lumsden et al. (1999), except where indicated, so that results can be compared with the 1998 study.

Bat Sampling: In the 1998 study, bat calls were recorded on Anabat II bat detectors linked to tape recorders via delay switches. There have been significant advances in recent years in bat detector technology and it is recommended for future work on the Christmas Island Pipistrelle that Anabat detectors are used in conjunction with the new Anabat CF Storage Zcaim (Titley Electronics, Ballina, NSW). This system downloads calls directly to a memory card and avoids the previous problem of cassette tapes filling up with high frequency insect noise. All-night recordings are now possible, and units should be set to record from before dusk until after dawn. This will provide more detailed information of activity levels throughout the night and will also document any daytime flying activity. The number of bat passes between 1830 and 2100 hrs (Christmas Island Standard Time) will be used for comparison with the 1998 and 1994 studies. The detector and CF Zcaim are to be housed in a waterproof box with only the microphone exposed (e.g. refer to Plate 6 in Lumsden et al. 1999). To protect the units from Robber Crabs Birgus latro these units need to be placed above the ground, for example on folding chairs. The chair arms are used to position the box at an angle of 45° (Plate 7 in Lumsden et al. 1999). A waterproof cover extending out from the top of the box protects the microphone from rain. The sensitivity dial on the detectors are set at seven, and on this setting bats can be detected at a range of 15 to 20 m. The majority of sites are sampled for a single night, although repeat sampling is required if rain occurs during the sampling period.

Data analysis: Since the pipistrelle is the only bat on the island that echolocates, all bat echolocation calls recorded on the Anabat detector/CF Zcaim can be assigned to this species. However, insects on Christmas Island produce high frequency noise that sounds very similar to bat calls and is recorded by the detector. If recordings are made onto cassette tapes the calls should be examined using Anabat software to distinguish them. However, if a CF Zcaim is used, as recommended, insect noise is largely eliminated (G. Richards, pers. comm). Each time a bat flies past the detector, the CF Zcaim saves a computer file, with a unique date and time stamp. Where there is continuous activity, a file is automatically saved every 16 seconds. Feeding buzzes are identified by the rapid increase in pulse rate of the echolocation call. These buzzes indicate foraging activity as distinct from commuting bats, and are counted separately for each recording session.

To compare with earlier studies, activity levels are assessed by the number of passes over a 150-minute time period, commencing 45 minutes after sunset (i.e. 1830 to 2100 hrs). Activity levels are defined as high (> 1 pass per minute, i.e. > 150 passes in 150 min.), moderate (1 bat pass per 1 to 10 minutes, i.e. 15 to 150 passes in 150 min.) and low (< 1 bat pass per 10 min., i.e. < 15 passes in 150 min.).

Insect Sampling: Insect availability is to be measured using a light trap at each detector site. The light trap consists of a 20 litre plastic bucket with a cone fitted inside (Plate 9 in Lumsden et al. 1999). Attached to the cone is a fluorescent light with two tubes, a black light and a standard light. The cone projects the light upwards to sample the airspace in which the bats forage. At the base of the cone is a jar containing 70% alcohol, into which the attracted insects fall. The light is powered by a 12 V battery and a timer activates the unit to turn on at dusk and off at dawn. To keep out of reach of Robber Crabs the light trap is hung on a rope tied between two trees, or placed on a folding chair.

Captured insects are to be dried at 80oC for three hours and weighed to give a dry weight of available prey. Only insects < 15 mm in total body length are included so that the assessment is based on prey items within the likely size range taken by the pipistrelle.

The number of insect calls recorded on the bat detectors could be used as an additional measure of insect activity (G. Richards, pers. comm.), although only a component of the insect fauna would be recorded in this way.