Developing rigorous ecological models is a fundamental goal of ecologists in order to forecast biotic responses to environmental change. A limitation of many models is they are a-mechanistic and lack integration of behavior, which is integral to animal biology. Woodland salamanders (family Plethodontidae) are small, lungless animals known to be sensitive to climate, and often abundant across a wide range of forest conditions. In this dissertation, I reviewed plant climbing among plethodontid salamanders and discuss hypotheses for this commonly observed, but underappreciated, behavior. I integrated biophysical and agent-based models (ABM) to examine how climbing behavior could affect the sensitivity of Plethodontid salamander activity time to climate by using a temperature differential to stimulate plant climbing, allowing salamanders to alter their evaporative water loss rates. I used observations of Plethodontids in the southern Appalachian mountains to evaluate the model predictions. Finally, I used three approaches a biophysical agent-based model, a field experiment, and spatially replicated surveys - to estimate the potential effects of Rhododendron and its management on terrestrial salamander activity time, survival, and abundance. The model predicted that climbing increased salamander activity time with greater soil temperature relative to air temperature and increased time since rain. Model predictions were confirmed by field observations. The applied biophysical model predicted a negligible effect of Rhododendron on salamander activity time, and capture-recapture analyses showed no short-term effects of Rhododendron management on salamander survival. Estimated abundances from comparative surveys indicated little to no effect of Rhododendron on salamander abundance across a rainfall gradient. By using multiple approaches, our results suggest that Rhododendron has a negligible effect on salamander performance or abundance, and cutting Rhododendron and burning the forest floor had no short-term effect on apparent survival. Including compensatory behaviors, like climbing, in models is important because they have the potential to moderate the effects of broader environmental changes - such as climate change - on animal performance and abundance.