Community structure of ants in Brunei rain forest

Thesis

This thesis examines the patterns in the ground and low-vegetation ant community of primary lowland dipterocarp forest in Brunei, South East Asia.

The first experimental chapter establishes the importance of ants as the numerically dominant taxon of the forest floor litter arthropod community, how many species were found in a hectare of forest, and the probable minimum asymptotic number of species for each of four sampling methods. The amount of small fragments of litter is found to vary across the plot, increasing in the valley, and best explains differences in species richness of samples. Species richness is highest in the valley. Faunal, rather than microhabitat factors, are found to be the best predictors of certain litter species' abundance.

Chapter three shows that colony dispersion is not a good method with which to investigate competition in this habitat at present. The ground dwelling ant community are more often clumped than over-dispersed, but patterns are impossible to interpret in the light of ecological theory without knowing the biology and population dynamics of the species well. I show by reanalysis that two major flaws in a similar study by Levings and Franks (1982) confound their claim that overdispersion of ant colonies occurs in Panama. This result has implications for many other similar studies of spatial dispersion as evidence of competition.

Chapter four examines whether there is a predictable dominance hierarchy in the bait responsive species foraging in the litter and on tree trunks. No evidence of such a hierarchy was found, although significant temporal patterns unexpected from diel cycles imply avoidance of some species by others. Differences in recruitment ability were found between tree and ground foraging species and these correlated with speed to locate and persistence at baits. A method is offered to test whether ant species are solitary foragers.

Chapter five shows that there is no clear relationship between niche overlaps and association, i.e. avoidance of colonies or foragers at baits. Low temporal overlap and high food overlap were found, but did not appear to be related.

Authors: Levy, R

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford