An organism’s physiology interacts with microclimate to influence its fitness and behavior. Biophysical models can predict fitness within given climatic parameters, but are rarely consider behavior or environmental variability. Integrating biophysical models with Individual- or Agent-Based Models (IBM/ABM) allow for incorporation of geographic, individual, and behavioral differences. Southern Appalachia consists of a variable landscape with high salamander diversity. Many salamander species are known to facultatively climb plants, but little is known about this behavior. Though it is assumed salamanders limit climbing due to the risk of dehydration, under conditions common to Southern Appalachian forests, the risk may not be as great as previously thought. Under certain conditions, such as when soil temperatures are greater than air temperatures, climbing may lower salamander’s body temperatures and water loss rates. Based on this concept, I added an opportunity to climb vegetation to an established individual-based biophysical model. The model estimates salamander activity time when climbing and not in the presence and absence of Rhododendron. Rhododendron is a patchily distributed midstory canopy shrub in Southern Appalachia and is known to buffer climatic variation. I compared the simulated results with field data, specifically looking at seasonality and differences in climbing probability with and without Rhododendron.
Results/Conclusions
The model allowed virtual salamanders (based on Plethodon shermani) to climb plants when the soil temperature was higher than the surrounding air. This suggested climbing should be observed more frequently in the spring and fall. Field data somewhat supports this prediction with increased observations of salamanders climbing in the fall, but none observed climbing in the spring. This inconsistency suggests another factor (other than water loss) is likely influencing climbing behaviors. The model predicted increased climbing probability in areas without a midstory canopy, which is generally consistent with field observations, and likely reflects the greater variability of areas without a midstory canopy’s microclimate buffering capacity. Though the model predominantly estimated increased activity time under Rhododendron, climbing in areas without Rhododendron made up for or surpassed those estimates in activity, particularly in the fall. This supports the importance of including behavioral variability with biophysical models. Future models will include additional environmental variables and test alternative mechanisms for facultative arboreality in salamanders.