National recovery plan for the Christmas Island Pipistrelle Pipistrellus murrayi
Prepared by Martin Schulz and Linda F. Lumsden
Department of the Environment and Heritage, 2004
ISBN 0 642 55012 3
Appendix two: Pipistrelle monitoring protocol
- 1). Vehicle-based driving detection
- 2). Stationary site detection
- Monitoring equipment recommended
This monitoring strategy has been adapted from Lumsden et al. (1999):
The most appropriate way to monitor the pipistrelle population is to employ similar techniques to those used in the 1998 survey (refer to Appendix One for stationary sampling methods), so that results can be compared. It is recommended that two strategies be used to investigate aspects of the Christmas Island Pipistrelle status: 1) decline in activity levels at known sites, and 2) contraction in distribution.
This strategy will provide distributional information, which currently is of importance given the westward range contraction of the species. The methodology for this strategy is outlined below.
- Provides information on distributional change.
- Much quicker and easier in providing useful information.
- Provides limited information on abundance levels at particular sites.
Vehicle-based Driving Detection Methodology
- Drive a standardised route encompassing all regions of the island and passing through sites of low, moderate and high bat activity (Figure 13). This route is to be driven at a speed of ≤ 40 kph with two people, a driver and a bat detector operator. The latter person will hold the detector out of the passenger window, at a constant angle, pointing forward in the direction the vehicle is travelling.
- Route: Start at the Parks Australia North office in the Settlement, down Murray Road, past Central Area Workshop along the road to The Dales, down Winifred Beach Track to the Circuit Track, U-turn, back to the road to The Dales, turn right, continue east to LB4, turn right, continue to Coconut Corner, along EW Baseline to NS Baseline, past the Airport, down the upper road to the Casino, back along the coast through the Settlement to the Parks Office (Figure 13).
- The censuses are only to be conducted in suitable weather conditions: no rain, wind speed
- They are to be conducted within a standardised 2.5 hour time period, commencing no earlier than 45 minutes after sunset and finishing no later than 3 hours 15 minutes after sunset. It is likely to take approximately 2.0 to 2.5 hours to complete the circuit; therefore it should commence as soon as possible following 45 minutes after official sunset.
- Equipment: See below.
- Recording Technique: Two potential methods depending on equipment availability and the ability of a GPS to provide continual accurate locational information (NB: this may be a problem due to rainforest canopy cover, the typically high vertical extent of the cloud and continuous cloud cover):
- The detector operator would record the output of the detector continually on to cassette tapes for the entire driving route. The operator would give the speedometer reading and location at all road junctions on to the tape, using the voice activation function on the detector. Whenever a sound thought to be a bat call was heard, the speedometer would also be recorded by the operator onto the cassette tape. It is important that the bat detector operator is familiar with the sound of pipistrelle calls heard over the external microphone of the Anabat detector. A familiarisation session should be run for any new operator prior to a monitoring sampling session. Where possible, the same bat detector operator should be used. It is recommended that the resulting tapes are sent for analysis to someone experienced in distinguishing between bat and insect sounds. Hence this sampling strategy would also require the small cost of analysing the recordings.
- The Anabat detector would be connected to the Anabat CF Storage Zcaim, with a compatible GPS also hooked up to provide locational information to be continually downloaded. With a limited amount of training, the files produced using this technique could be analysed on the island, as insect noise is largely excluded using this technique (G. Richards, pers. comm.).
This strategy will provide information on changes in abundance levels at known sites, which is important to monitor given the apparent decrease in abundance at some localities.
A selection of sites with varying activity levels from the 1998 survey (Lumsden et al. 1999) would be sampled. Use of the stationary detector method should have the advantage that the equipment can be set up and left all night and then retrieved the following day, minimising labour costs. Deployment of two or more units per sample night would reduce the number of sampling nights required. The following 10 sites are recommended for sampling.Site # Location Lat./Long. Activity level 4 Pipeline track NW of Jedda Cave, 0.4 km NW of Powerline Track 10o28'27" 105o38'16" Low 8 0.1 km down track to rehab area 22S, at gate 10o29'26" 105o36'57" Moderate 14 Circuit Track, 0.15 km E of Winifred Beach Track 10o29'38" 105o34'09" Moderate 16 Winifred Beach carpark 10o30'00" 105o32'58" Moderate 19 Track in front of Research Station 10o29'28" 105o38'46" Low 21 Winifred Beach Track, 0.1 km S of Dales Road 10o28'40" 105o34'25" High 28 At start of walking track to Hughs Dale waterfall, 20 m from the road. 10o28'32" 105o33'38" Moderate 34 ML 106 area, on track heading W, 0.6 km from Blowholes Road. 10o30'24" 105o38'35" High 44 Road to Dales, 0.5 km E of intersection below LB4 10o28'36" 105o36'15" High 70 Toms Ridge, 3.0 km NW of Dales Road 10o27'32" 105o33'36" Moderate
Equipment and Technique: As for Appendix One.
Analysis: Electronic files could be analysed on the island if someone was appropriately trained, or could be sent off-island to someone experienced in distinguishing between bat and insect sounds.
- Provide information on abundance levels at a number of known sites.
- A number of sites would need to be sampled to show patterns and as only one detector unit is currently available on the island only one site could be sampled a night (although see recommendations below). This limitation would mean that multiple nights would have to be devoted to pipistrelle monitoring within each time period.
- Provides information on distributional change with a large amount of effort (i.e. sampling nights).
Note: The Parks Australia North Anabat detector was found to be faulty in 1998 and needs to be returned to Titley Electronics for repairs, before the monitoring program commences. It appeared to be a problem with the speaker affecting the volume of the calls. Since there is considerable vehicle noise when undertaking vehicle detection, it is necessary to have the volume level working correctly. In addition new equipment would need to be purchased as outlined below.
- 2 Anabat II Detectors (one as a backup, Supplier: Titley Electronics) (for both monitoring strategies)
- 2 Cassette players and blank cassettes (Vehicle-based driving detection only)
- 2 Anabat CF Storage Zcaim, 2 128MB Compact Flash Memory Cards (for data storage) and Anabat software (Supplier: Titley Electronics) (definitely Stationary Sampling, preferably Vehicle-based driving detection)
- 1 USB Reader/Writer or a Parallel Reader/Writer (to download information from CF Zcaim to computer).
- Associated cables (connecting tape recorder to detector, CF Zcaim to detector; Supplier: Titley Electronics) (for both monitoring strategies)
- Garmin Etrex GPS (or similar common brand water-resistant hand-held GPS) (for both monitoring strategies).
- 2 Waterproof tool box (Stationary Sampling)
- 2 Folding Chairs (Stationary Sampling)
- Waterproof cover (Stationary Sampling) (Lumsden et al. 1999 used a laminated manilla folder attached to the lid of the tool box with Velcro, Plate 6 and 7).
To detect changes in status at particular sites (Stationary Sites) or distributional changes (Vehicle-based Driving Detection) it will require a minimum of two years of sampling but is recommended to be conducted on a yearly basis for the full five-year life of this Plan.
It is recommended that the:
- Stationary Sampling be undertaken one time per year per site in May-June to coincide with the 1998 survey, and
- Vehicle-based Detection Sampling be undertaken on two nights at four monthly intervals, with one sample period to coincide with the timing of the 1998 survey (May-June).