ROOST SITE SELECTION AND ROOSTING ECOLOGY OF FOREST DWELLING BATS IN SOUTHERN BRITISH-COLUMBIA

We used radiotelemetry to examine the roost-site preferences of four s pecies of tree-roosting bats (Eptesicus fuscus, Lasionycteris noctivag ans, Myotis evotis, and M. volans) in southern British Columbia, Canad a, by radio-tracking bats to their day roosts. We found a total of 21 roost trees: 14 roosts were beneath loose bark, 5 were in cavities exc avated by woodpeckers, and 2 were in natural cavities. Entrance height increased with tree height, but roost entrances tended to be situated below the level of the canopy. Of the 22 tree and site variables exam ined, only 3 significantly discriminated between roost trees and avail able trees: tree height, distance to the nearest available tree, and p ercent canopy closure. Bats preferred tall trees associated with low p ercent canopy closure and a short distance to the nearest available tr ee. Bats roosted in western white pine, and to a lesser extent pondero sa pine and western larch;in intermediate stages of decay more often t han would be expected at random. Bats switched roosts frequently. The distance between subsequent roost trees was short, suggesting a degree of fidelity to a particular group of trees or area of forest. The num ber of days of rain during the roosting period significantly influence d the number of days spent in a particular roost, and thus ambient con ditions may restrict the frequency with which bats can switch roosts.