Abstract:
A survey of ground-living spiders was conducted at Ngome State Forest, northern Natal. The area is situated on the southern African escarpment. The survey started in January 1992 and ended in January 1993. Five different habitat types were sampled with 180 pitfall traps. Pitfall traps were open continuously and were cleared once a month. Contents were sorted under a dissection microscope and all spiders were sorted into morphospecies, most of which were later identified. The habitat types that were compared were grass, open forest, dense forest, ecotone and pine. Grass, open forest and dense forest represented indigenous vegetation while pine represented exotic vegetation. Ecotone consisted of a mixture of indigenous forest plants and pine trees. Multivariate analysis of the data showed the different habitat types to support different ground-living spider communities. The composition of spider communities in pine differed most from all other habitat types. This result supports the hypothesis that the planting of pine trees largely affects the composition of ground-living spider communities. Spider diversity and richness indices indicated grass to be the most diverse habitat type, while pine was the least diverse habitat type (for ground-living spiders) . However, due to the large variation in spider diversity within habitat types, the differences in spider diversity between habitat types were not clearly defined and the results do not unambiguously support the hypothesis that exotic vegetation has lower ground-living spider diversity than indigenous vegetation. The habitat preferences and phenology of the 10 most abundantly trapped spider species were graphically represented and discussed as to make a contribution to these species's autecological relationships. The need for more surveys of this kind in order to identify larger scale ecological patterns is stressed.
