Back to index of mammal studies

Back to index of 2002 research reports 

MICROHABITAT USE BY THE TROPICAL FOREST RODENT, BUNOMYS ANDREWSI, ON BUTON ISLAND, SULAWESI. 

INTRODUCTION 

Research on the tropical forests of Buton island are still in their preliminary stages of development, but the huge diversity of species it harbours has already been established emerged. This diversity is reflected by the similarly rich number of small mammal species that reside in this haven.  However, the sheer numbers of animals found here also brings great interspecific competition, and has led to the divergence and subsequent specialization into every imaginable niche the ecosystem offers.  Specific species have become superbly adapted to their given environmentMy investigation is concerned with  A fundamental rule for the distribution of a species is that differences in abundance are often correlated with differences in habitat.  A species may perceive and respond to a habitat in two ways.  A fine-grained species uses subsets of a habitat mosaic in direct proportion to the abundance to those subsets and the population densities.  In contrast, a coarse-grained species will select certain habitats preferentially over others (Levins 1968). However, it has been suggested that habitat selection by a species may be carried out on either a microhabitat or macrohabitat scale, so where is the line drawn? Here the study will be conducted on a microhabitat scale to search for possible characteristics in the microhabitat influencing distribution. 

The habitat usage of small mammals such as B. andrewsi, might also be influenced by their population densities. For example, a species may show less marked microhabitat preferences at higher population densities.  Mark and recapture techniques can be used to estimate the size of the populations in the study areas and the influence on distribution considered.  

By comparing the distribution of B. andrewsi on the trapping grid to microhabitat data, any relationship between the distribution and microhabitat characteristics can be investigated. 

METHOD 

The study was conducted between June and August 2002.  The research was based in the Forest Reserves around the village of Labundobundo. Four different trapping grids were set up. Two were situated in the one kilometre square grid in the Kakenauwe Forest Reserve.  The third trapping site was in the ‘La Pago’ study grid in Lambusanga Forest reserve, and the fourth was at the more distant ‘Anoa’ grid in Lambusanga. 

Each of the trapping grids consisted of 49 trap stations, in a 7x7 formation, all situated 20m apart.  Each trap site contained two traps; one small and one large Elliot aluminium box trap.  The traps were baited with a standard small mammal bait mix of oats, peanut butter and mealworms and the trapping was carried out over a ten-night period. 

The traps were checked daily, each morning, and re-baited when necessary.  Basic morphometric data was taken from all animals trapped, and each animal was given a unique mark (either a microchip or fur-clip) before being released where it was caught. 

In addition to collecting trapping data, a microhabitat assessment was carried out at each trap station. A 5x5m quadrat was set up with the centre of the quadrat located at the mid-point between the two Elliot traps. A digital photo of canopy cover was taken from directly above this centre point. Over the entire 5x5m quadrat the number of trees over 15cm in circumference was recorded and their circumferences measured. In addition, the length and diameter of any logs (over 10cm in diameter) crossing the quadrat were recorded. The distance of both logs and trees from the trap was also recorded. A marked 1m pole was used to provide an index of vegetation density, sighting from each of the corners to the centre point at a height of 0.5m from ground level. The number of lianas crossing these lines from the corners to the centre point was also recorded.  Then using smaller 1x1m quadrats, situated in the 4 corners of the larger 5x5m quadrat, the percentage vegetation cover, percentage exposed rock and number of small woody stems were recorded.  

Over the whole 120x120m trapping grid, any trees over 1.5m in circumference were mapped.

RESULTS 

The trapping data collected consists of the catch rates, recapture rates and distribution of the captures at each of the 4 trapping grids. 

Table 1: B. andrewsi captures at each site over a ten night trapping period

A suitable mark-recapture model will be used to provide population estimates for each of the trapping grids.

REPORT

A dissertation entitled Microhabitat usage by the tropical forest rodent, Bunomys andrewsii will be produced by Lucy Barton, University of Southampton by April 2003

REFERENCES

Levins, R. (1968). Evolution in changing environments. Princeton University Press, Princeton, New Jersey, USA