Pisces Text Logo
Pisces Text Logo2

Variations in the Ground Ant Communities of Kakamega Forest, Western Kenya: Potential Indicators of Forest Disturbance.
Anton Espira
A thesis submitted to the Board of the Faculty of Biological Sciences for the Degree of Doctor of Philosophy of the University of Oxford; Wolfson College, Trinity Term 2001

Price: CD-ROM: £38.00 + VAT/postage
           Download: £19.00 + VAT

The development of effective evaluation systems is an integral part of sustainable management strategies. The use of bioindicators is currently a common ecosystem evaluation procedure. This study investigates the utility of the ground ant fauna as a bioindicator of tropical rainforest disturbance and regeneration in Kakamega Forest, western Kenya.

This thesis is divided into three related sections. The first section (Chapters 3 and 4) examines the diversity of the ant assemblage in Kakamega forest in habitats of differing disturbance and management histories, and correlates this to floristic regeneration to evaluate its bioindicator potential. The second section (Chapters 5 and 7) examines various aspects of the ant community in the forest and relates these to habitat disturbance. The third section (Chapter 6) evaluates the effectiveness of two ground ant sampling techniques, Winkler sampling and pitfall trapping, in sampling the ground ant fauna and generating suitable data for use in ecosystem evaluation.

The overall diversity was found to be fairly high. 115 species of ants in 37 genera representing 8 subfamilies were collected. Results showed no significant difference in abundance and species richness of ants in the three forest habitats sampled. Numerically, the ant community was dominated by generalist Myrmicinae genera such as Tetramorium and Pheidole.

No clear relationships were observed between the ant fauna (diversity and abundance) and most measured floristic variables. However, ant species richness was positively correlated to tree species numbers and the quantity of wood debris in the habitats sampled.

No evidence for the existence of an ant mosaic was found. A number of species were found to be nocturnal and it is speculated that these developed their foraging patterns in response to competitive pressure from the more abundant myrmicine generalists. Investigations of the effect of swarm raids of the driver ant Dorylus nigricans on the ground arthropod community showed that most arthropod groups seemed unaffected, though spiders, millipedes and cockroaches showed a significant decline in numbers following a raid.

Winkler sampling proved the most effective ground ant sampling technique.

Pitfall trapping, however, was markedly superior in collecting large Ponerines and was the only system functionally capable of sampling litter-lacking habitats.

In conclusion (Chapter 8), an assessment of the utility of ant assemblages as bioindicators is questioned. Recommendations are proposed for the enhancement of ant bioindicator potential. The conservation status of Kakamega Forest is reviewed in the context of biodiversity preservation and resource utilisation. Recommendations are made with long-term economically feasible conservation of the forest in mind.

Back to main page