Small Mammals

Surveying Small Mammals with Humane Traps

Researcher - Tim van Berkel (Lead Scientist)

Aims

The aim of the small mammal trapping was to get an indication of the small mammal community assemblage and how this would differ for various habitats. Initially we tried to estimate density and habitat association but due to the small trapping effort this proved to be unsuccessful. Instead, relative abundance was calculated. We further wanted to see how trap types influenced trapping success for different species.

Methods

Traps were planned to be active for four weeks. Because two brown spiny rats (Maxomys rajah) were killed after the first seven trapping days by ants and termite infestations and time constraints it was decided not to continue trapping after the eighth trapping day, giving a total trapping effort of 664 trapping nights. Trapping was carried out from the 2-9 of January 2010.

Fifty four locally made Kasmin cage traps (3x3.5x9") and 38 Sherman foldable aluminium traps (6x6x19”) were placed along two parallel trap lines spaced 50m apart.

Traps were distributed along 23 stations, spaced twenty meters apart. Each station consisted of four traps, which were placed in a 10 meter radius. Traps were baited with peanut butter or a durian paste, which was refreshed every other day. As it is estimated that over 83% of Bornean mammals is at least partly arboreal (Mabberly, 1983) 18 traps were placed in trees and woody climbers up to a height 2m above the ground to capture arboreal species. To enhance trapping success traps were placed in places that are most likely to be used by small mammals. These are generally: along the edge and on top of logs; along paths across leaf litter or soil; near apparently active burrows and holes and in between the buttresses of large trees. All were placed in areas of relatively dense vegetation.

Traps were checked each day at dusk and dawn and rebaited where needed. Traps were traceable by means of a raffia string running from a central tree to each trap, something which increased operating speed considerably.

Captured animals were identified to species level and moved to self-made handling cones to allow body measurements to be taken. These included body mass, length from nose tip to tail base, tail length, left hind foot length, greatest ear length, gender, age class and ectoparasites.

Camera Traps

Terrestrial Vertebrate Camera Trapping

Researcher - Tim van Berkel (Lead Scientist)

Aims

Camera traps were used to survey the diversity and abundance of the terrestrial mammal and bird communities in the study area. We further tested how different camera types would influence capture rate.

Methods

Camera trapping was conducted from mid-December 2010 till the beginning of February 2011. Twenty five digital camera traps (14x Cuddeback Capture, 9x Bushnell Trophy Cam, 1x Bolyguard SG560) were used during the expedition. The Cuddeback and Bolyguard cameras took photos and used an incandescent flash when light conditions required, while Bushnell cameras used infrared light instead. The Bushnells were set to record infrared and thus grey scale videos ranging from 20 to 60 seconds. Eighteen cameras were spaced at a 500m interval along 2 parallel trails of 4km length which were cut to facilitate the VES surveys and which themselves were spaced 500m apart (Map 1). These were 12 Cuddeback, the Bolyguard and five Bushnell cameras. The remaining seven cameras were placed randomly in suislidele locations, two of which were placed in the upper canopy for three and four days each.

The total trapping effort equalled 546 trapping days (camera traps x days), giving a mean active period of 22 days per camera, although camera activity ranged from 3-43 days. Each terrestrial camera was mounted on a tree approximately 30cm of the ground along animal trails to optimize capture rate and were operational 24h/day as long as it was functional. Cameras were checked approximately every two weeks to change memory cards and batteries. In some cases cameras ran out of power before they were checked. In those cases the last picture or video taken is used as the last time the camera was active.

Photo and video analysis

Animals were identified with the help of illustrations and descriptions of (Payne et al., 1985) and were updated to current and revised names and species. As both chevrotain species Tragulus napu and T. kanchil, both muntjacs Muntjacus atherodes and M. muntjak and both Maxomys rajah and M. surifer are morphologically similar and therefore difficult to distinguish, each couple is grouped and identified to genus level following (Azlan and Lading, 2006) to facilitate analysis. Photos and or videos taken from a single camera of the same species taken within 40 minutes of each other were considered to be the same individuals and were considered to be a single trapping event unless individual recognition proved possible.

The data from the two arboreal traps were not used in the analysis.

Herpetofauna

Studying the Reptiles and Amphibians Around Camp Foyle

Researchers - Dale Mortiboys and Holli Kilburn

Aims

The aims of the survey were to esslidelish which herpetofauna species are present in the study area.

Methods

The success of a study aimed at sampling the biodiversity of an area is heavily dependent on the choice of sampling methods. Traditionally, herpetofauna have been sampled through active collecting techniques, searching for the animals in environments where researchers expected them to be found. For example, Vanzolini and Papavero (1967) and Corn and Bury (1990) recommended detailed searches under fallen trunks and stones, in bromeliads, within the leaf litter, at the base of tree trunks, and at stream edges. Although this approach can provide a representative sample of the species present in a given area, it is clearly biased by the ability, experience and effort of the researcher and does not provide standardized estimates of abundance., Passive sampling techniques on the contrary, are effective at reducing researcher bias, but they generally focus on one subgroup of the total fauna (e.g., pitfall trapping is directed at leaf litter taxa and is unsuislidele for sampling large, agile, or arboreal species).

Different groups of amphibians and reptiles were sampled with varying efficiency by different methods following (Crump and Scott, 1994; Jaeger and Inger, 1994; Doan, 2003). Frogs were sampled according to Heyer et al. (1994) and Olson et al. (1997).

Pitfalls

Pitfall traps with drift-fences are one of the most common herpetofauna sampling techniques (Heyer et al. 1994) and (Wilson et al. 1996).

Two pit-fall trap systems were used for 30 days. Each pit-fall trap system consisted of 7.8 litre buckets (25 cm depth and 20cm diameter) buried in the ground and set approximately 5m apart, with soft plastic drift fences with a height of 50 cm extended between them. In the first system the four buckets were put out in a line with the fence running through the middle of each bucket, or in the second system the buckets were laid out in a Y shape. One bucket was placed at the crossing while the other three were placed at the end of each fencing extension.

Due to availability it was not possible to acquire larger buckets, which are preferred over smaller ones as it will keep the larger species from escaping. Larger buckets are often considered to have a greater chance of success in capturing larger animals, e.g., large reptiles (Crosswhite et al., 1999; Cechin & Martins, 2000). Pitfall trapping is most effective at capturing leaf litter species (e.g. Bury & Corn, 1987; Greenberg et al., 1994; Enge, 2001), as well as some insectivorous arboreal species that descend to the leaf litter to forage (Crosswhite et al., 1999).

Pitfalls were checked once every morning and all frogs, reptiles and non-target species such as small mammals and invertebrates found in them were measured and released.

VES

Standard nocturnal visual encounter surveys (VES, Heyer et al. 1994) using LED head torches were usually conducted between 1800-2200 hrs. We carried out 20 transects of 4 hours duration each, totalling a sampling effort of 80 hours.

Visual searching (in and on vegetation, between rocks along the stream, on the forest floor) was the main focus of these survey nights. A steady pace (av. 750m/h) was maintained. At four points along each transect there was an intense search of wet areas (usually streams/bogs/rivers) for 15 minutes to increase effectiveness for amphibian searches. VES has been found to yield an equal number of species as other active methods whilst recording more unique species (Pearman et al., 1995; Adams et al., 1998; Rocha et al., 2004).

Diurnal VES were carried out during sunny periods of the day between 0900-1200 or 1400-1600. This is the period when skinks are most active.

Amphibians and reptiles were measured with dial callipers to the nearest 0.1 mm snout to ventral length (SVL) and tail lengths separately),

Campbell and Christman (1982) and Bury and Raphael (1983) have suggested that active sampling and pitfall traps provide an effective combination for the study of herpetofauna communities.

Preliminary Results

A total of 37 reptiles and 30 frogs were encountered during the various surveys. Twenty four of the 30 frogs have never been recorded in the area and for some this means a significant increase in geographic range. Appendices 2 and 3 give a comprehensive list of the species encountered.

Inverterbrates

Studying the Inveretbrates of the Rainforest

Researcher - Rusty Goodchild (Entemologist)

Aims

The aim of the invertebrate sampling was to gain an accurate idea of what invertebrate species were present in this area as well as seeing if any new undescribed species were present.

Knowing in advance there would be significant quantities of invertebrate organisms only certain groups were being surveyed. At the request of Dr Merlijn Jocke the arachnid orders thelyphonida and amblypygi, the arachnid sub-order cyphophthalmi from the order opiliones, and the insect families cerambycidae and elateridae were included. These groups will further be referred to as the BINCO 5. Based on my own knowledge and experience aranae (spiders) were also included.

We investigate how different species diversity and community composition are compared to northern Borneo sites. This will be particularly important to judge the value of these regions for biodiversity conservation, e.g. in the context of the proposed and ratified, trinational Heart of Borneo conservation area.

We sampled data on the species composition of nocturnal Lepidoptera and leaf litter ants from representative habitats.

Canopy

Canopy Access - Epiphytes and Herpetofauna Surveys
Researcher - Holli Kilburn

Aims

Very little high arboreal survey work has been conducted in Kalimantan Borneo. Reptiles and amphibians are very adapslidele animals with frogs especially occupying a very wide range of ecological niches from below ground to high canopy and within a large variety of habitats. In order to gain a more complete insight into the full range of species found in the area it was important to try and expand the research to encompass as much varied habitat as possible. The purpose of the high arboreal work was to give a more three dimensional insight into the ecology of the area, observe as wide a range of ecological niches as possible and also to potentially expand the knowledge of individual species ecological niches.

The same seven trees mentioned above were also surveyed for their vascular epiphyte populations. A log of the epiphyte species was compiled and the relative abundance of the epiphyte population was assessed using the DAFOR scale.

Hybrid Gibbon

Searching for Proof of Hybrod Gibbons in the Area

Researcher - Lara Rogers (Assistant Lead Scientist)

Aims

A sslidele hybrid population of the parent species Hylobates muelleri and H. albibarbis is known to be resident in the region East of the study area. This species is estimated to be in existence for over 10.000 years. We aimed to investigate the Joloi River banks for presence of the hybrid. Occurrence here could extend the known geographic range of the hybrid from an estimated 3,500 km² to approximately 10,000 km².The western boundary of its geographical range is now believed to be the Sungai Busang.

Methods

Five suislidele research sites spread up from the Mohot and Joloi Rivers were chosen. Two groups of three observers were stationed at two suislidele listening posts either side of the rivers by 05:00h, following the methods of Brockelman and Ali, 1987; Brockelman and Srikosmatara, 1993 and Cheyne et al, 2007.

Gibbon pairs sing duets most mornings for the purposes of maintaining pair-bonds and defining each pair’s territory (Cowlishaw, 1992). Gibbon calls are species specific, differences between Hylobates albibarbis and the hybrid H. albibarbis x H. muelleri are easily distinguished (Mather 1992, Chivers pers. comm.). Observers will record the gibbon species calling (H. albibarbis or H. albibarbis x H. muelleri), direction of each calling group heard, the time and estimated distance of the call to enable triangulation of group locations and the start and end time of each calling bout. This was repeated for two consecutive days to ensure that the majority groups within hearing distance were heard singing at least once.

Results

Several family groups of H. albibarbis were recorded to be present in the area along the main base camp along the Mohot River. The hybrid gibbon at this research site was not recorded. A single group of hybrid gibbons was heard calling at the last research site of the Joloi River, on the eastern side and right at the navigable end of the river. This group was heard calling amidst several groups of H. albibarbis.