Use of forest strata by bats in temperate forests M. Plank 1,2 , K. Fiedler 1 & G. Reiter 2 1 Department of Animal Biodiversity, University of Vienna, Vienna, Austria 2 Austrian Coordination Centre for Bat Conservation and Research (KFFOE), Leonding, Austria Keywords acoustic monitoring; bat community; batcorder; Chiroptera; stratified habitat use. Correspondence Michael Plank, Dachsteinstraße 8, A-4614 Marchtrenk, Austria. Tel: +43 680 2083599 Email: Michael.plank@gmx.at Editor: Nigel Bennett Received 25 April 2011; revised 8 August 2011; accepted 18 August 2011 doi:10.1111/j.1469-7998.2011.00859.x Abstract Concurrent recordings of bat calls at the ground and canopy levels were compared to analyse the vertical stratification of habitat use in a broad-leaved deciduous forest in central Europe. The recording effort was 48 nights, by sampling 16 sites three times in 2010. Overall, 2170 call sequences were recorded; 40% of them at the canopy level. Sixteen bat species were identified, 13 of which occurred at ground level and 14 in the canopy. By fitting generalized linear mixed models for seven species and all calls sequences combined, the effects of stratum, time period, ambient temperature and forest structure on bat-calling activity were assessed. Four species preferred the ground level (Barbastella barbastellus, Eptesicus seroti- nus, Myotis daubentonii, Pipistrellus pygmaeus). M. bechsteinii was the only species which had a significant preference for the canopy level in the pregnancy, lactation and post-lactation period, whereas two further species (M. alcathoe, P. pipistrel- lus) showed a significant canopy preference in at least one period. Therefore, canopy sampling should be undertaken when aiming at monitoring such species for conservation purposes. Populations of certain target species will be underes- timated by ground recordings alone. Introduction Habitat selection by bats in temperate forests had been the focus of several studies (Kanuch et al., 2008; Napal et al., 2010). Because of its spatial extent, the forest interior is the most important foraging habitat, which also harbours the highest bat species diversity (Celuch & Kropil, 2008). Numer- ous studies on vertical stratification within forests are avail- able on bats from tropical ecosystems, stimulated by the complex vertical structure and species composition in rainfor- ests (e.g. Henry et al., 2004; Kalko et al., 2008; Pereira, Marques & Palmeirim, 2010). All these studies show differ- ences in the abundance and species composition between the understory level and the canopy stratum. Therefore, many bat species can be assigned to a specific stratum, whereas others forage through all forest strata. In temperate-zone forests (North-America: Bradshaw, 1996; Wunder & Carey, 1996; Kalcounis et al., 1999; Hayes & Gruver, 2000; Australia: Adams, Law & French, 2009; Europe: Collins & Jones, 2009; Zeus, 2010), the canopy is an important foraging habitat where overall activity is similar to the understory. In North- American and European forests, bat species composition appears to be similar between the strata, but there are signifi- cant differences in the activity level of certain species and species groups. However, most studies suffer from insufficient species discrimination (e.g. in the genus Myotis) or small sample sizes. The aim of the present study was to assess the stratification of bats in a forest region in central Europe by means of auto- mated acoustic monitoring, which is at least as effective as mist-netting in regard to surveying bat community structures (O’Farrell & Gannon, 1999; MacSwiney, Clarke & Racey, 2008). Of course, acoustic surveys are biased because some echolocation calls are more readily detected than others. This depends on the call intensity and the varying sensitivity on the microphones to certain frequencies. Comparative acoustic surveys can be valuable as biases in detectability of calls between strata are probably negligible. To assess the influence of the stratum on the foraging activity, it is necessary to include the most important forest parameters and the tem- perature in the calculated models. The forest structure is strongly linked to the bat species present (Ford et al., 2005) and temperature can be used as a predictor for insect activity (Dajoz, 2000). For some bats, like ground gleaning and aerial foraging species, a clear stratification was expected. In con- trast, the life habits of other species are so poorly known that habitat use cannot be predicted. We tested the following predictions: • A clear stratification for ground gleaning bats and open space species will occur because of their foraging behaviour, whereas bats that hunt close to the vegetation will use both strata equally. • Bat species that forage in both strata can change their prefer- ence according to their reproduction cycle because lactating Journal of Zoology Journal of Zoology. Print ISSN 0952-8369 154 Journal of Zoology 286 (2012) 154–162 © 2011 The Authors. Journal of Zoology © 2011 The Zoological Society of London