Search

Small-scale distribution of ground beetles (Cole- optera, Carabidae) was examined, based on catches among 300 pitfall traps, in a coniferous forest in southern Finland. The sample from the whole snow-free season comprised 2405 individuals of twenty-two species. Each of the most numerous species (Pterostichus oblongopunctatus (F.), Calathus micropterus (Dft.), Leistus terminatus (Hellw. in Pz.), Notiophilus biguttatus (F.) and Cychrus caraboides (L.)) was non-randomly distributed and formed aggrega- tions within the site of c. 1.3 ha in area. Although the five species occurred in every microhabitat defined in the plot, four of them (N. biguttatus was an exception) were more abundantly found in certain microhabitat types than in the others. In addition to vegetation around the traps, the numbers of Formica ants in the traps correlated with the numbers of carabids caught, mostly negatively. The consid- erable variation in catches and species richness among single traps and among blocks of 16 traps tended to even out, when larger subsamples were taken from the total pool (blocks of 48 traps). The distribution patterns observed within the plot were compared to a reference data set from similar habitat in the same region. Species distributions among microhabitats were slightly different in the reference set and the predic- tive success was relatively poor, probably due to different scales of study in the two data sets. Variation in species distribution in the two spatial scales studied (within a habitat patch and among them) is suppos- edly due to different factors. Although no direct evidence is available from our study area, we suggest that active micro- habitat selection explains the small-scale distribution within the study plot, whereas dynamics of local populations, influenced by regional-scale differences in habitat composi- tion, are the most likely explanation for the distribution pat- terns among habitat patches.

The equilibrium theory of island biogeography of MacArthur and Wilson aimed at providing a unifying framework for studying insular ecology. However, the theory does not elaborate the concept of an 'island' in an ecological framework. I suggest that different types of islands be ecologically defined on the basis of processes relative to which some ecologically interesting entities on those 'islands' actually are isolated from their surround- ings. Four main island scales can be discerned, namely, the scales of (1) individuals, (2) population dynamics, (3) population differentiation, and (4) evolutionary divergence. I illustrate these different island types and their interfaces with examples mainly from northwestem Europe. The equilibrium hypothesis seems relevant primarily on the population dynamics scale. A research programme following such differentiation would include two phases: first, assessment of factors which make the environment studied 'insular', and second, construction of specific models for analysing ecological processes in those particular insular conditions.

Land birds breeding on forty-four islands (size range 0.5-582 ha) in the Vargskar archipelago of the kland Islands (SW Finland) were censused in three (or two) breeding seasons in 1976-80 by one-visit methods. The islands were divided into five size classes and expected population sizes of land birds were estimated on the basis of (1) habitat composition of the islands and (2) population densities of each species in equivalent habitats in the source area, Main Aland. Prevalence is defined as the ratio between observed and expected population sizes. Observed densities of most of the fifty-six species studied did not deviate from expected values. Eleven species were more abundant on the islands than expected, including dominant generalists and species favoured by edges and bushy, heterogeneous habitat structure. Fourteen species had smaller populations on the islands than expected, but only seven of them consistently so in all census years and all island size classes. Habitat micro-structure seems important for three low-prevalence species, nest predation is invoked in two cases, in one case poor wintering capability seems a plausible factor, while one case remains obscure. Rarities were observed in the island censuses about as frequently as on Main Aland. Interspecific competition seems a plausible explanation for the absence of one species from the archipelago. One species, Turdus pilaris, changed its prevalence pattern dramatically in the study area by disappearing from the archipelago after the exceptionally cold winter of 1978/ 79; recolonization has taken place very slowly. The only species possibly showing density compensation is the dominant generalist Fringilla coelebs. Its populations on small islands are about doubled compared with the expecta- tions. Total densities are also higher than expected on small islands, but in most cases F. coelebs contributes less than half to the increase. In addition, there are no puzzling absences from small islands, but the poverty of their bird com- munities seems to result from reduced habitat diversity. Therefore autecological population responses rather than density compensation seem a plausible explana- tion of excessive population sizes of single species such as F. coelebs. We con- clude that, in these northern conditions, it seems appropriate to think of island communities as assemblages of species colonizing the islands according to their autecological (especially habitat) requirements. Ubiquitous habitat require- ments and flexibility in habitat utilization seem important correlates of coloniza