At present, little is known about the roosting ecology of the Bechstein's bat (Myotis bechsteinii); by learning more about what factors influence where this species forms their maternity roosts we will be making a significant step in understanding their life history. With a low population density, low rates of reproduction and apparent specific habitat requirements, this species is particularly vulnerable to habitat loss. In 1995 only 1,500 individuals were thought to be present in Britain (Harris et al., 1995). Providing advice on how to manage land more effectively for rarer species is an important aspect of the UK Species Action Plan; this study will therefore help in providing information for the conservation of Bechstein's in the UK and also the rest of Europe. Both breeding and non-breeding females form maternity colonies from the end of May until the beginning of July to help in communally caring for the young (Kerth et al., 2001). Roost switching is common among maternity roosts and appears to occur every few days for about three months with the main group splitting off into non-associated sub-groups throughout this time (Kerth & Konig, 1996). This is an important yet highly vulnerable period for the bats as the survival of their young is dependant on them choosing near optimum conditions in which to roost and forage. Radio-tracking individual bats has located a number of maternity roosts in southern England. In a one year MPhil project I shall be going to the known maternity roosts in Sussex, Dorset and Hampshire to look at:

1. Physical properties of the roost trees and the cavities in which the bats use.
2. The habitat immediately surrounding a 10m radius of the roost tree.
3. The properties and habitats surrounding known non-roost trees.

Both tree and surrounding habitat may be important in providing a thermal buffer around the roost, creating near optimum conditions in the cavity itself. Surrounding foliage and the level of canopy cover may also be important in providing protection from avian predators or shelter from the elements (i.e. wind, rain and the sun) (Sedgeley & O'Donnell, 1999). After the data have been collected the properties of each site will be compared for any similarities or differences. These results can then be used to predict why some sites make more attractive maternity roosts than others. Whilst providing better protection for the roost sites we have already, this information should enable us to provide future management recommendations allowing more favourable maternity roost sites to be created.