The Role of Behavior in Influencing Headwater Salamander

Evolutionary theory predicts that animals have evolved to move in resposne to a suite of cues that maximize animal survival and growth while minimizing risk of injury or mortality.  When humans modify environments, new conditions can suddenly yield maladaptive behavioral responses to cues that were adaptive in undisturbed environments.  Changes in behavior can alter animal movement patterns impacting patterns of occupancy among habitats and fragment populations.  Our objectives were to determine which environmental features best predict salamander occupancy in an exurban region of the southern Appalachian Mountains and describe the behavioral mechanisms driving observed salamander distributions. Headwater streams generally have dense canopy cover that limits direct light penetration and regulates temperature and humidity known to influence salamander behavior.  Therefore, we predicted that canopy cover would be an important predicotr of stream salamander occupancy and behavior. A rapid, regional survey of stream salamander occupancy was conducted in the Little Tennessee River basin with the Coweeta Long-Term Ecological Research site.  Of a variety of factors and scales, upstream riparian forest and canopy cover were the two most important predictors of stream salamander occupancy.  Because deforestation increases light penetration to streams, we performed controlled lab and field experiments to determine if salamanders exhibited consistent behavioral responses to light under a range of environmental conditions.  Salamander larvae generally exhibited negative phototaxis that was exacerbated by the absence of refuge. Lastly, we tested whether negative phototaxis could negatively influence natural behaviors in a field environment by surveying homing behavior across canopy gaps.  All salamander life stages were unwilling to cross even small canopy gaps even in the absence of additional habitat alteration.  Overall, this work demonstrates the importance of considering the consequences of fine-scale individual behaviors in determining how broad-scale disturbance influences animal populations and suggests a method for long-term monitoring of animal populations over broad geographical areas.