Notizen: Summary 1. The ant fauna of 3 islands of the Central Seychelles was studied by means of pitfall traps. The number of species and their abundances were investigated in relation to differences in island size, habitats, and microclimate conditions. In addition the effects of these factors on niche breadth and niche overlap were studied. 2. Nineteen and species were collected that belonged to 14 genera in 4 subfamilies (Table 2). Different numbers of species were found on different islands (Table 1). There are no significant correlations between the numbers of species, sizes of islands, and the available resources. The distribution of congeneric species agrees with the expected increase in ‘relative diversity’ on smaller islands (Fig. 3). 3. Species diversity and species distribution depend to some extent on habitat differences. This was evaluated by phytosociologic methods and microclimatic measurements. The composition of the ant fauna on different islands is more constant in undisturbed, autochthonous habitats than in those habitats that are strongly influenced by man. 4. Specific niche breadth and relative abundance of ant species showed a significantly positive correlation. Within comparable habitats, average niche breadth and niche overlap are greatest on the smallest island. Because the relative density also increases on this island, one cannot speak of a lower ecologic saturation compared with the larger islands.

Notizen: Summary 1. Twenty-five species of web-building spiders (Araneae) were studied by examining standard study areas on 5 islands of the Central Seychelles group that differ in size and degree of isolation. The number of individuals and of species on each island was correlated to the area, altitude, degree of isolation, and resource availability of the respective island. The effect of these factors on niche breadth and niche overlap was also examined. 2. In respect to resource availability the structural diversity of each study area was determined by means of the Shannon-Weaver formula. Environmental conditions were broken down into 4 categories according to cover, soil, water, vegetational strata, and number of plant species. In addition to this quantitative method different habitats were defined according to existing plant associations (Table 2). Both methods produced similar results. 3. As the islands increased in size, both the spectrum of resources and the number of habitats also increased. Islands of lasrger surface revealed a higher mean as well as a greater range of structural diversity (Fig. 2). Associated with the increased spectrum of resources there was a significant increase of spider species. 4. The number S of web-building spider species increases with the size A of the island according to the Wilson formula S=cA z (Fig. 3). The present study revealed the coefficient z to be remarkably low. 5. The degree of isolation of an island, calculated according to macArthur and Wilson (1967), did not explain the differing numbers of spider species on the different islands. 6. Niche breadth and niche overlap were calculated in terms of numbers of habitats and of ranges of structural diversity. The most abundant spider species had the greatest niche breadth. The average niche breadth became smaller with the increase of island size. Specific niche breadths of spiders on different islands did not vary significantly between islands (Table 6). But in all cases there was a reduced niche overlap in the more diverse spider communities (Tables 5 and 8). 7. The total density of the spider community was highest in the structurally most diversified habitats. The total density of web spiders diminished with the decreasing size of the island. The correlation between total density and number of species was significantly positive. 8. The low value of the z coefficient in the species-area relationship is discussed. The possibility of a low ecologic saturation on smaller islands is considered. Ways in which species diversity may increase are examined. Analysis of resource availability proved to be more efficient for the solution of ecologic problems than the simple examination of the species-area relationship.