Prescribed fire has increased as a forest management practice in southern Appalachia, but investigations into the effects of this silvicultural treatment on non-game wildlife inhabiting the region is limited. This study investigated the effects of prescribed fire on millipede and salamander populations. Seventeen sites within the Cherokee National Forest in east Tennessee that were treated once with prescribed fire between 1998 and 2002 were examined in the spring of 2002. Each burned plot was paired with an adjacent unburned plot. The number of individual millipedes and salamanders collected from burned plots was compared to the corresponding control plot. Millipedes declined in numbers following treatment with prescribed fire. The number of salamanders collected was insufficient to permit analyses comparing the number of salamanders found in burned and control sites. No difference between burned and control plots was detected in the species comprising the millipede communities.

Massachusetts sandplain heathlands are habitats of conservation concern, harboring many rare plant species and providing habitat for animals that depend on openlands. These heathlands are threatened by human development, shrub encroachment in the absence of disturbance, and potentially increasing soil nutrient levels. Sandplain heathlands are managed with prescribed fire, in order to maintain their open structure and maintain species diversity. In order to assess how past management was correlated with species change, I used a data set that spanned twenty years from three different heathlands in Massachusetts. I looked for correlations between management and species change. Correlations between species change and prescribed burning were very site, or microsite, specific, indicating that variables such as vegetation type and edaphic characteristics need to be taken into account before management is applied. Prescribed fire was also associated with an increase in ruderal species in one of the sites studied, indicating that there may be undesirable effects of prescribed fire in this system. Species diversity was negatively associated with shrub encroachment, reinforcing the importance of preventing shrubs from encroaching into these heathlands. I also found evidence that burning has not been a successful technique in preventing shrub encroachment in these sites. Lastly, the nitrophilic species Carex pensylvanica increased in all three sites, indicating that future studies should investigate the possibility that sandplain heathlands are currently experiencing nitrogen deposition beyond their critical loads.

In north Mississippi, I evaluated fuel loads in the late dormant season and after leaf fall in unburned areas and after a dormant season burn. Fuel loads beneath winged elm (Ulmus alata) were assessed to determine if this shade-tolerant species impacts fine fuel loads in the area immediately near its bole. I found leaf litter fuel loads are higher after leaf fall suggesting that burns conducted closer to leaf fall may have increased burn intensity. Burning reduced leaf litter fuels and exposed mineral soil which returned to pre-burn levels following leaf fall. Declines in duff layer fuels were not evident until after leaf fall. Leaf litter fuels underneath winged elms had higher mass and percentage of winged elm litter beneath them compared to areas away from them. These results will help determine appropriate time for restoring prescribed fire and the implications of encroachment by non-oak species into upland oak forests.

Pine savannas of the Southeastern United States are generally managed using prescribed fire, which creates the vegetation structure required by their associated fauna. However, little is known about the effects that size and spatial pattern of prescribed fire have on wildlife. During a 3 year radio-telemetry study, I monitored movement and hazard of adult bobwhites in a landscape managed with prescribed fires of varying size. I also monitored growth and survival of bobwhite chicks in response to fire size and spatial pattern. Adult movement was negatively related to percentage of home range burned. Hazard rates were negatively related to movement and percentage of home range burned. Chick growth was negatively related to movement and home range size. Chick survival was positively related to percentage of home range burned, and likelihood of brood success was negatively related to hatch date and positively related to number of eggs hatched.

Due largely to fire exclusion and land use changes, upland oak ecosystems in the central and eastern U.S. are shifting dominance from fire-tolerant oaks (Quercus spp.) to shade-tolerant, fire-sensitive species (mesophytes). This shift has been hypothesized to occur via a positive feedback loop termed mesophication, where mesophytes create shaded understory that limits oak growth and wetter fuels and soils, decreasing forest flammability. To determine how canopy water partitioning varies between oaks and mesophytes, I measured stemflow, throughfall, and surface soil moisture monthly over a 14-month period for overstory and midstory trees of oaks (Q. alba, Q. falcata) and hypothesized mesophytes (Carya tomentosa, Acer rubrum, Ulmus alata) in northern Mississippi. Overstory oaks partitioned 5.1% of rainwater into stemflow, while mesophytic species partitioned 7.2%, leading to 3.5% wetter soils under mesophytes. The hydrology of mesophyte canopies may reduce forest flammability and promote conditions favorable for mesophyte regeneration, ultimately compromising long-term oak regeneration.

In historically fire-dependent upland oak forests of the eastern U.S., anthropogenic fire exclusion is likely causing a hypothesized feedback loop between an increase in fire-sensitive species and self-promoting, fireree conditions at the detriment of oak regeneration. This study determined how shifts from oaks (Quercus stellata and Q. falcata) to fire-sensitive non-oaks (Carya spp., Liquidambar styraciflua, and Ulmus alata) affected flammability and related processes that consequently determine species composition. Using treatments of increasing non-oak leaf litter, experimental burns were conducted and flammability measured under field conditions, and a laboratory litter moisture desorption experiment was conducted. As litter composition shifted from oak-dominated to non-oak-dominated, flammability decreased (R2 = 0.59, P < 0.001) and moisture-holding capacity increased (R2=0.88, P<0.001). To prevent further shifts toward fireree conditions and loss of economically and ecologically valuable oaks, prescribed fire should be reintroduced soon while oak maintains overstory dominance and controls forest flammability.

Declining oak (Quercus spp.) dominance across the eastern U.S. is often attributed to fire exclusion and abundant white-tailed deer (Odocoileus virginianus). Fire restoration can negatively affect acorn germination and survival directly but also indirectly through impacts on seed predation. Similarly, masting events could influence oak regeneration success by altering deer behavior. To date, virtually no information exists assessing indirect effects of acorn consumers on oak regeneration. I developed two experiments to address this knowledge gap. In the first experiment, I determined if burning acorns affects predator removal rates. In the second experiment, I assessed the indirect effects of mast seeding on plant communities mediated by deer. Burning acorns decreased acorn removal rates. This could increase survival to spring for acorns that survived exposure to fire. Mast seeding increased local deer use, decreased the competitive advantage of local oak seedlings, but increased beta diversity in the understory.

Gopher tortoises are ecosystem engineers whose burrows provide habitat to >350 species. Prescribed fire is used to manage tortoise habitat, but fire timing is mostly restricted to the vegetative dormant season. Restricted fire timing in combination with white-tailed deer competition may negatively affect tortoises. To address these concerns, we quantified these species’ dietary overlap and conducted a field experiment to examine impacts of fire phenology on plants and animals. Although tortoises and deer consumed ~75% of the same plants, their diets were statistically dissimilar. Fire altered plant community composition and increased foliar crude protein and phosphorus while decreasing calcium. Deer detections were unaffected, but tortoises were detected more in fire treatment plots. We simultaneously monitored burrow and surface temperatures and found burrows provide thermal refuge. Our data suggests that fire timing affects plants in ways that can affect gopher tortoises, and burrows may mitigate some negative impacts of climate change.

As a result of decades of fire suppression, oaks (Quercus L.) and other disturbance-dependent tree species are experiencing widespread regeneration failure. Today, fire takes the form of relatively low to moderate intensity prescribed fire, used to restore fire adapted ecosystems, and wildfires which often vary in severity. I investigated long-term changes to forest structure and composition in response to repeated prescribed burning followed by an extended period of no fire. Burning reduced total basal area, midstory stem density and sapling stem density. However, the fire-free interval significantly increased sapling layer stem densities of oaks and competitor species. This research shows that repeated prescribed fire, followed by a fire-free interval, can allow oak seedlings to grow into sapling sized stems, but competitors also increase in density. I also investigated relationships between varying wildfire severity and stand structure, basal area, and tree recruitment. Both stem density and total basal area were significantly and negatively related to fire severity. Oak and pine recruitment was significantly and positively related to fire severity whereas competitors had no relationship. The positive relationships with fire severity and oak or pine sapling recruitment could have important implications for managers using prescribed fire or managing areas after wildfire.

The objective of this study is to determine whether changes in arthropod community structure in restored longleaf pine savannas corresponds to differences in vegetation structure often associated with burn frequency. Longleaf pine savannas are fire-maintained ecosystems characteristic of the southeastern United States and have experienced severe declines (around 97%) since European settlement. Changes in fire regime have been instrumental in the declines. Restoration of these ecosystems has involved reinstitution of periodic burnings to promote and maintain vegetative characteristics of the savannas. This study investigates trends in arthropod communities from areas heavily invaded by hardwood shrubs against those dominated by longleaf pines and associated vegetation. These data suggest that herb-dominated sites have higher overall diversity. While overall abundance differences were not found, significant differences have been detected at the order and family level, indicating that vegetation structure and periodic burning are important factors in maintaining arthropod communities characteristic of these savannas.