Containing over 900 species worldwide (Nowak 1994) and with tropical communities often comprising in excess of 50 species (Kingston et al in press 2001), the bats (Order Chiroptera) are among the most highly diverse of mammalian orders. The high diversity of these tropical communities raises important questions as to how the species of these differing yet diverse bat guilds interact, coexist, and to whether a form of resource partitioning or niche specialisation occurs.

In the Palaeotropics, first and most apparent partition within the Chiroptera occurs between the fruit eating Megachiroptera and the insectivorous Microchiroptera. Visual and olfactory senses are the primary modes of orientation and location of food within the Megachiroptera, with only one genus, Rousettus possessing any form of echolocation, used solely for orientation (Nowak 1994). The Microchiroptera however, have a more sophisticated form of echolocation, enabling them to orientate and locate prey within a previously under exploited night sky. Echolocating insectivorous bats which employ this mode of sensory awareness have various problems to overcome, dependant on their foraging habitat, with bats exploiting the open space habitats showing a greatly modified form of echolocation to those that inhabit the forest interior (Neuweiler, 1990-cited Kingston et al submitted). Open space bats are required to search for widely distributed prey items over a large 3-dimensional space (Kingston et al submitted). This necessitates long-range detection.

The physical properties of high frequency sound waves, and the bats ability to interpret the returning information (the echo) into a successful prey capture, is largely reliant on two things, the length of the pulse, and its attenuation through the air. These pulses are described as either frequency modulated (FM) or constant frequency (CF), although most microbats use a combination of the two, dependant on their current activity (Altringham 1999).

The open space bats first problem is to locate their insect prey over the large areas in which they forage. During this early ‘search’ phase the bats emit a more CF dominated pulse. This is good for the initial detection of prey, but with its long (typically 10-50ms (milliseconds)) call duration, a shorter FM element is required for distance and positional information during the approach to point of capture.

Another property of high frequency sound waves is in the correlation between a higher frequency and shorter wavelength. Due to this correlation it has been hypothesised by others (such as Pye 1993-cited Kingston et al 2000) that a smaller bat, emitting a higher frequency, would be better able to (and therefore specialise in), the detection of smaller prey items than one with a lower frequency. The lower frequency sound waves would stand a greater chance of ‘missing’ the target. A high frequency bat would also be good for the detection of larger prey items, but due to the increased attenuation of these high frequencies, the lower frequency call would gain a marked advantage over larger distances.

The other main factor affecting foraging behaviour is wing morphology, with the two major components of this being wing loading (WL) and aspect ratio (AR). Wing loading describes the relationship between the weight and wing area (mass x g/wing area) with a high WL equating to a fast efficient flier. Aspect ratio focuses on the wingspan to wing area (wingspan²/wing area), desribes the shape of the wing. A high aspect ratio confers energetic efficieny during flight, but limits manoeurability.

1. To determine the species composition of a tropical guild of open space insectivorous bats.
2. To establish to what extent resource partitioning occurs within a guild of aerial open space bats.
3. To determine whether resource partitioning is due to echolocation frequency, wing morphology, or a combination of both.

The primary open space study site was situated along the Umala Dongkula river (ref: 542083 to 552115, Lembar 2210-64, 1:50,000 Mataompana) where it focused on a relatively small section, approximately ½ km from the coastline (ref: 550115). This stretch of riverine habitat, influenced by tidal patterns, provided a continuous uncluttered aerial open space habitat for foraging bats. The surrounding vegetation varied from a more dense forest-type condition on the up-river littoral margins, to a more open swamp and isolated coconut-stand environment further towards the coast.

Sampling occurred between 9^th July to 20^th August. Capture nights were undertaken from 17:30 to 19:30, with mist nets of differing sizes, and of various combinations, employed perpendicular to the flow of the river. The two combinations deployed included a lower 9m (metre) net (approx. 2m above the water) set 1 metre from a higher 12m (at approx. 4m in height) configuration, and two 9m nets, stacked one above the other. The nets were placed at two marked locations, one slightly down river to the other, enabling direct comparisons between successive capture nights.

Some species, possessing an increased echolocating sensitivity (and therefore ability to detect the stationary net), demanded a modification to the capture method. This involved a method known as ‘flick’ netting. One shorter 5m net was used for this method, attached to the top of two 3m bamboo poles. These poles were then held by two people at an angle of approximately 45° and ‘flicked’ in a forwards and back sweep on the approach of a bat

Bats caught were immediately removed from the net and placed in small cloth bags for transport back to Labundobundo and the projects laboratory. Bats were held for 24hours to enable complete passage of food material throught the gut. Processing occurred the following morning, where ecomorphological data for each individual was ascertained. Species identification following Corbet and Hill, Payne and Francis and other source material (Kingston, 2001-personal communication), with sex and reproductive condition also noted. Measurements taken consisting of calliper measurements were length of forearm (to the nearest 0.1mm), and body weight measurements, using Pesola spring scales (to the nearest 0.1grams). Each individual was then tagged with a uniquely numbered metal ring, closed securely around the narrowest section of the forearm, and finally a dorsal view wing trace drawn of the fully extended right wing.

Bats were taken back to the point of capture the following afternoon (at approx. 17:00) where their echolocation calls were recorded during release. Recording was achieved through the combination of a S-25 Bat Detector, Ultra Sound Advice (USA) (enabling human detection of the high frequency calls), linked to a USA S-350 digital signal processor (which time expands the wavelength), and finally connected to a Sony WM-D6C Walkman cassette recorder. These calls will then be input into the software programme BatSound Pro, where start, end, and peak frequencies (kHz), as well as the length of pulse and interval between pulses (ms) will be determined.

Dietary analysis work firstly required the accumulation of an insect collection, firstly to provide information on the insects active during bat foraging activity, but also to act as a reference collection during the analysis of faecal samples. Insects were caught during times of mist net deployment, using a 12 volt powered Heath trap, set with an alcohol collection container at its base. Insect specimens shall be classified to order, with slides made of each species. Faecal pellets shall be analysed under a dissecting microscope for identifiable arthropodal remains.

Over 14 nights of both mist and flick netting a total of 121 bats were caught. This resulted from 303 mist netting hours and 71 flick net hours. Open space and edge and gap bats caught included 57 Tadarida sarasinorum, 33 Emballonura spp., 2 Cheiromeles parvidens, 2 Miniopterus spp., 1 Myotis sp. and 2 Pipistrellus sp.

Dietary studies, wing morphology and echolocation analysis will be completed in the U.K..

Report

A report entitled Resource partitioning in aerial open space bats;relationships between ecomorphology and diet was completed by Tony Wood, University of Reading in March 2002.  This dissertation received a First..