Date of Graduation


Document Type


Degree Type



Davis College of Agriculture, Natural Resources and Design


Wood Science and Technology

Committee Chair

John W. Edwards.


In 2007 and 2008, I examined roost selection of forest-interior dwelling northern myotis (Myotis septentrionalis) maternity colonies in stands treated with prescribed fire (hereafter, fire) and in unburned (hereafter, control) stands on the Fernow Experimental Forest, West Virginia. Using radio telemetry, I tracked 36 female northern myotis to 69 roost trees; 25 in the fire treatment and 44 in the control treatment. Using logistic regression and an information-theoretic model selection approach, I determined that within the fire treatment, northern myotis maternity colonies were more likely to use cavity trees that were smaller in diameter, higher in crown class, and located in stands with lower basal area, gentler slopes, and higher percentage of fire-killed stems than random trees. Moreover, roosts often were surrounded by trees that were in the upper crown classes. In the control treatment, northern myotis were more likely to roost nearer the tops of larger diameter cavity trees in early stages of decay that were surrounded by decaying trees in the upper crown classes than random trees. Roost trees in the fire treatment were associated with larger overall canopy gaps than roost trees within the control treatment. Regardless of treatment, northern myotis maternity colonies roosted in black locust (Robinia pseudoacacia) in greater proportion than its availability.;At the Fernow Experimental Forest in West Virginia, I examined roost networks of northern myotis in forest stands subjected to prescribed fire and those that were not in 2007--2009. I determined that northern myotis did conform to fission-fusion models, forming social groups whose roost areas and roost networks overlapped to some extent. Roost networks largely resembled scale-free network models, 61% of which had single identifiable central node roost trees on which the remainder of the roost trees in the network centered. In control, i.e., unburned, treatments, central node roost trees were in early stages of decay and surrounded by higher basal area than other trees within the networks. In prescribed fire treatments, central node roost trees were smaller in diameter, lower in the forest canopy, and surrounded by lower basal area than other trees in the networks. Compared to central node roost trees in control treatments, those in prescribed fire treatments had higher amounts of bark coverage. My results indicate that northern myotis form social groups that roost in networks of trees that center on a central node roost tree that is selected based on surrounding basal area and decay stage. Forest disturbances such as prescribed fire may result in changes to roost networks through inputs in suitable roost trees by senescence and creation of forest canopy gaps.;Forest disturbances, particularly permanent land use conversions, may fragment, isolate, and concentrate summer populations of philopatric bats, potentially resulting in genetically definable subpopulations. Also, subpopulations may be distinguishable at large spatial scales, regardless of forest disturbances and other land use conversions. I examined genetic structuring of the forest-interior dwelling northern myotis at multiple spatial scales, from watershed to regional, in West Virginia and New York. I genotyped microsatellites at 18 nuclear loci from 174 individuals. In many cases, loci in subpopulations that I defined a priori based on spatial attributes departed from Hardy-Weinberg equilibrium; those loci that were in equilibrium were inconsistent at differentiating subpopulations based on spatial scale. My results indicate that some level of genetic structuring is evident among northern myotis, but this structuring could not be attributed solely to geographic location. It is likely that the ability to migrate relatively long distances, exploit small forest patches, and use networks of swarming sites located throughout the Appalachian Mountain range has allowed northern myotis to maintain high gene flow regardless of forest disturbances resulting in habitat patchiness at local and regional spatial scales.;Understanding the management impact, positive or negative, of prescribed fire on endangered Indiana myotis (Myotis sodalis) roosting habitat is of paramount interest to resource managers in the Appalachian Mountains and elsewhere. From 2001 to 2009, we examined roost selection of Indiana myotis in burned and unburned forests in Tucker County, West Virginia. Our research indicates that use of fire for forest management purposes, at minimum provoked no response from Indiana myotis in terms of roost tree selection, but rather created additional roost resources, suggesting prescribed fire can be an effective management tool for Indiana myotis. (Abstract shortened by UMI.).