Bird species prediction modeling: A comparison of multiple modeling techniques at Francis Marion National Forest in South Carolina

There have been many studies that try to determine what environmental variables are important to bird species for habitat selection. This is commonly determined using various forms of species distribution modeling. The goal of this research was to determine whether readily available models derived from the South Carolina GAP Analysis Program (GAP) data are sufficient for predicting bird species distributions at Francis Marion National Forest in coastal South Carolina or if more advanced techniques using data collected in situ provide better estimates of species patterns. Specifically, I compared GAP species predictions based on observed land use and land cover associations with distributions created using maximum entropy (Maxent) to determine if there are significant differences in the areas that bird species are predicted to be present.;GAP models were created using 2001 NLCD land use land cover data for the entire state. Maxent models were created using presence point counts for bird species found within the National Forest and ancillary environmental data. Models were created for multiple bird species to assess which variables were important for each species individually and to facilitate comparisons of bird guild effects. This research highlights the prediction differences between the two modeling techniques and what environmental factors are important to bird species for habitat selection.;This research found that there are differences between the GAP and Maxent predictions for the bird species selected in this study. There were large differences for several species between the two modeling techniques. I believe that the Maxent model provides more useful results and has the potential to provide more informed predictions. Additionally the Maxent model results produced more information that was usable for comparisons between bird migratory status and diet patterns. This information was used for comparisons between these groupings that allowed for generalizations to be made which could aid in conservation efforts.