THE FACTORS AFFECTING SPATIAL DISTRIBUTION OF BATS

General aims

To examine the factors that determine the spatial distribution of insectivorous bat species in two ecological communities in Buton Island, SE Sulawesi

Objectives

1) To investigate the spatial and temporal distribution patterns of insectivorous bat species within the forest interior

2) To establish the roosting and social ecology of the forest-interior species Kerivoula papillosa

3) To develop a sound library of calls of open-space insectivorous species, to allow acoustic surveys

4) To test validity of the allotonic frequency hypothesis

Introduction

Bats are the second most speciose mammalian order and the only mammals to have evolved true flight. This unique trait, togther with echolocation, means they are highly adapted to their ecological niche, but also highly constrained. Bats exhibit widely contrasting patterns of roosting and social ecology. Species may be monogamous, polygamous or promiscuous. Furthermore, while some species roost in large colonies in caves, other may occur in small groups or as solitary individuals. Our ongoing research programme in Buton principally aims to determine the correlates of species distribution. It is likely that variation in wing morphology and call design among different species limit their use to specific habitat types. For example, species, which live in the highly cluttered environment of the forest, must use high frequency calls to resolve prey items. These calls attenuate readily thereby minimising too many echoes from the surrounding environment. Such species have broad wings allowing manoeuvrable flight. By comparison, species living in open-spaces require higher amplitude calls of lower frequency, and tend to have narrow wings. In addition within habitats, differences in distributions may further reflect contrasting roosting requirements as well as subtle morphological variation.

Methods

Fieldwork was undertaken from June to August 2002 at the Kakenauwe Reserve (Labundo Bundo, Buton Island, SE Sulawesi) and the surrounding area. The Reserve consists of a 1 km^2 grid, with parallel trails approximately 100 metres apart. For Objective 1, all forest-interior bats were captured in harp-traps, which are particularly suited to trapping species which high frequency echolocation calls. Six harp traps were used. Traps were checked twice daily and moved every morning with the first of each set of six traps starting approximately 50 metres from the last. The checking of the traps coincided with the most active times for the bats themselves – dawn and dusk. Trapping started on the 10^th west trail, continued down the west trails to west 0 then back up the north trails on the grid to finish at north 10.

All bats were identified to species level, sexed, aged, and measured (forearm length (FA) and body mass). Females were checked for reproductive condition. The animals were left in the bags overnight in order for the collection of faecal samples and released the next day at the same trapping spot.

The spatial monitoring of species will provide a means of establishing whether populations are uniformly or unevenly distributed. This will help to establish if ‘hotspots’ exist which may be indicative of differences in roosting ecology. For example, species that require caves for roosting are expected to be more clustered in their distribution than species that roost in trees.

A more detailed study was also undertaken on the roosting ecology of one species, Kerivoula papillosa (objective 2). Eight specimens were fitted with 0.75g radio-transmitters (Holohil, Canada) and released at the point of capture. These were tracked using a telemetry receiver attached to a three-element antenna (Biotrack, UK). Roosts were identified by homing-in on the highest amplitude signals. All roost trees were measured and characteristics of the holes were recorded (height and size of cavity, dryness of cavity, aspect of cavity). In addition, for each roost tree, the surrounding canopy cover and tree density were recorded. To determine whether this species actively selects tree holes with particular characteristics, fifty trees with holes were identified using a random sampling technique, and the same measurements were taken. Statistical treatments will employ principal component and logistic regression analyses.

Insectivorous bat species, which belong to the ‘open-space’ and ‘edge-gap’ communities, are often able to avoid standing mist-nets but are also unable to be captured in harp-traps. To increase our understanding of the relative abundance of these species, and to allow future monitoring of habitat use, a library of calls of these species was developed (objective 3). Bat were caught using flick nets erected on bamboo poles and used over the river Umala Dongkula and around the Labundo Bundo village. Mist nets erected on tall bamboo poles were used to catch open space bats at the Paddi fields behind the Kakenauwe Reserve. Captured individuals were processed as described above and all bat were recorded on release using a Pettersson Ultrasound Detector D980. Each call was time-expanded and recorded on to a Sony WM D6C Walkman cassette recorder. The calls were analysed using the program BATSOUND. Once established, the library will allow students to undertake acoustic transects, recording the activity of bats in different habitat types.

The allotonic frequency hypothesis is a selective strategy that may confer an advantage to certain species of insectivorous bat ( Microchiroptera ) for the capturing of noctuid (eared) moths. The evolutionary arms race between bat and moth is one of the most spectacular examples of predator/prey coevolution. While these moths ears have evolved to filter irrelevant stimuli during the race – any stimulus that is not a bat call – differing species of bat have responded by evolving calls that appear to produce frequencies outside of the moths hearing range – typically between 20 and 50 kHz (Fullard, 1982). This is the basis of the allotonic frequency hypothesis. Sensitivity to a broad range of ultrasound frequencies has evolved in the hearing organs of those insects that have to avoid predation through the employment of drastic escape manoeuvres (Stumpner A, von Helversen, 2001). These defensive behaviours employed by noctuid moths have long been known to increase the moths survival by reducing their chance of being caught by up to 40% (Roeder, 1967; Acharya and Fenton, 1999). Thus, while hearing in noctuid moths has evolved as a defence against echolocating bats (Fullard,1987; Surlykke, 1988), it is also responsible for the evolvement of differing frequencies of bat call by acting as a major selection pressure on these bats (Fenton and Fullard, 1981).

The ability for frequency changes in calls to be selected and sustained is a major key to any insectivorous bats success and one that may have given these animals the leading edge in the arms race with the moth. The allotonic frequency hypothesis is therefore of particular interest, assuming that it is advantageous, and one that lies open for testing.

The projects aims are to obtain information about which species of insectivorous bat has successfully increased its access to moths over other species through the collection and analysis of bat faecal samples. Dietary analysis should show which species of bat favours moth predation. A comparison of these particular results can then be made with echolocation frequency to reveal which is the optimal frequency used in moth predation.

Previous studies have already revealed that bats which forage around the forest edge (edge gap bats) tend to have different call patterns with a higher frequency than open-space bats in order to successfully catch prey when at a higher density in denser habitats (Kingston et al. unpub). The same study also revealed the opposite for open space bats whose calls are used for long range detection for the detection of low density prey items. These lower call frequencies are thought to be used for the predation of large prey items and the open-space bats that use them are thought to have a higher wing loading ratio (Kingston et al. unpub), thus edge gap and internal bats with lower wing loading ratios that feed on smaller insects must be the species that are more vulnerable to forest disturbance.

Initial Results

To date over 432 individual bat specimens have been caught and processed. Forest bats comprised Hipposideros cervinus, H. cineraceus and H. diadema, together with Rhinolophus euryotis, R. philippinensis and R. celebensis. In addition, Kerivoula papillosa, K. hardwickei, Murina florium and Phoniscus jargorii were also caught.

Open-space bats caught were Hipposideros dinops, Tadarida sarasinorum, Pipistrellus javanicus/tenuis and Cheiromeles parvidens. The edge-gap bats caught were Emballonura nigrescens and Miniopterus australis.

Analysis of spatial distributions will be undertaken using GIS software. Early analysis of the roosting ecology of Kerivoula papillosa suggests selective roosting of some tree species and a highly unusual social system, with a skew toward males within the group.

The analysis of call structure and faecal pellets will be completed at a later date.

Reports

A final report will be prepared by Dr Steve Rossiter, Queen Mary & Westfield College, University of London by May 2003