Day-roost use by northern long-eared bat (Myotis septentrionalis) maternity colonies on the Fort Knox military reservation, Kentucky, USA, resulted in formation of non-random networks of roosts that exhibited a trend toward centralization. Centralization of day-roost networks was reflected in the social structure of colonies, which were characterized by dense associations among individuals within colonies. Social structure varied among colonies and appeared to be related to reproductive condition; colonies exhibited greater cohesiveness during parturition and lactation. Northern long-eared bat maternity colonies appeared to be exclusive, occupying distinct roosting areas with one or several areas receiving intense use.

Day-roost removal simulations suggested a linear relationship between colony fragmentation and roost loss, and that loss of >20% of roosts is required to initiate colony fragmentation. Experimental hierarchical removal of day-roosts yielded results consistent with simulations, as removal of the single most-central (primary) roost had no impact on colony fragmentation, whereas removal of 24% of less-central (secondary) roosts resulted in partial network fragmentation. Patterns of colony day-roost and space use were similar pre- and post-removal treatments. Day-roost removal did not alter the number of roosts used by individual bats, but distances moved between roosts were greater in the secondary roost-removal treatment group. Day-roost characteristics largely were consistent pre-post treatment for both treatment groups.

Historical data from an Indiana bat (Myotis sodalis) maternity colony revealed that this species also exhibits a non-random social assorting dynamic. Non-random social assortment resulted in a closely connected centralized network of day-roosts. Individuals within the Indiana bat maternity colony exhibited close social connections with colony members, but subgroups likely existed within the colony. Indiana bat day-roosting ecology appears flexible, as patterns of roost and space use differed substantially between years.

Development of specific, but tactical, management approaches for individual colonies of both northern long-eared and Indiana bats may be possible. Such approaches would allow land managers to manage for both northern long-eared bat habitat and other objectives. However, the nature of targeted management approaches employed likely will depend on the unique forest context and dynamic within which individual colonies reside.