For the assessment of the habitat use of bats with consideration of a complexity of environmental features a largescale multivariate canonical correspondence analysis (CCA) was performed. Data were collected in April-September (1999 using mist-netting and bat-detectors (n ¼ 209 samplings) in Slovakia (n ¼ 160 sites). For the habitat description, 17 environmental variables that characterised the sampling site, tree-species composition and general character of surrounding forest were selected. In study sites, altogether 93% of Slovak chiropterofauna species were recorded but only 16 species (2466 individuals) were recorded in more than 5% of samplings. The most common and frequent species were Myotis mystacinus (frequency of occurrence ¼ 57%, species dominance ¼ 12%) and Myotis myotis (44%, 10%), followed by Barbastella barbastellus (32%, 6%), Plecotus auritus (32%, 6%), Nyctalus noctula (31%, 9%) and Myotis brandtii (31%, 7%). Of the tested environmental variables used in CCA analysis, 13 had significant influence on the species' presence. On the base of gradients in ordination analysis, there were two main groups of species. In the first ''true'' forest bats group the tree-dwelling and gleaning species Myotis bechsteinii, Myotis nattereri and P. auritus dominated. Their occurrence was connected mainly with old forests of natural tree-species composition. The second group comprised species, which foraged in more open space or along the forest edge (with the farmland) and avoided closed interior (particularly Eptesicus serotinus, N. noctula, Nyctalus leisleri, Pipistrellus pipistrellus). The presence of mountain species Eptesicus nilssonii and Vespertilio murinus was associated mainly with dominance of spruce. Suggested predictive modelling of species composition in forest bat assemblages in conditions of natural forests can help in nature conservation.

No abstract

No abstract