S. Mahunka szerk.: Folia Entomologica Hungarica 51. (Budapest, 1990)
I. Open dolomitic grassland ('open habitat') . (Seseli leucospermo-Festucetum pallentis.) This is the first stage of dolomite succession after the lichenmoss communities (Jakucs 1981). Festuca pallens and Seseli leucospermum are the dominant and characteristic plant species. The rock/grass ratio is about the same . II. Dolomitic steppe meadow 'steppe habitat' . ( Chrysopogono-Caricetum humilis. ) There are several steppe plant species in this phase, where Carex humilis and Chrysopogon gryllus are the dominant and characteristic species. This plant community is almost closed, the rock/grass ratio is about 5/95. III. Sessile-turkey oak forest ('forest habitat'). ( Quercetum petraea-cerris .) This forest type is mainly composed of Quercus cerris and Q. petraea. The presence of Fraxinus ornus is also pronounced. The trees are not too high ( 10-15 m), because of the shallow soil . METHODS For sampling we used altogether 324 pitfall traps, 108 for each of the 3 vegetation phases. Plastic jars were used with a mouth diameter of 9 cm. The traps contained ethylene-glycol. The project lasted for about six months, from 7th of Aptil to 23rd of October, 1988. We visited the traps in every fortnight, altogether fourteen times, and mean values of the 14 visits were used. Only beetle data were analyzed. We compared the beetle communities connected with the three different plant communities with the help of community structure parameters, like species richness, species abundance and species-abundance distributions. The Sorensen index was applied to measure the similarity between the habitat: C = 2*J / (A+B) , where A= number of species caught on site A, B= number of species caught on site B, J= number of joint occurrences. If C= 1, then the two communities are identical, if C=0, then totally different (Southwood 1978). The Shannon-Weaver ' s diversity index was evaluated in relation to different beetle communities. The diversity index is: H = - p(i) X In p (i) , where p(i) is the proportion of individuals in the i-th species (Southwood 1978) . Eguitability was calculated according to the formula: J = H / In S, where S is the number of species present. We applied agglomerative cluster analysis based on 154 species, which were collected more than one time . Squared Euclidean distance and single linkage grouping strategy was chosen. Multiple stepwise regression was also applied for the Coleoptera data. These statistical analyses were carried out by the SPSS /PC+ statistical program package (Norusis 1986). RESULTS 32 families, 139 genera and 233 species are represented by a total number of 23,837 captured specimens. We recorded 13,952 individuals belonging to 153 species in the 'forest habitat' , 7562 individuals of 129 species in the 'steppe-habitat ', and 2323 individuals of 56 species in the 'open habitat'. The species-abundance distributions of the three beetle communities are shown in Fig. 1. On double logarithmic scale these curves show linear relationships, but their slopes are different (Table 1). J 8
