Forecasting Buffelgrass (Cenchrus ciliaris) Distributions In Southern Arizona Under Multiple Climate Change Scenarios

Abstract

Buffelgrass (Cenchrus ciliaris) is an invasive grass that can alter fire regimes, reduce local biodiversity, and convert complex arid ecosystems into buffelgrass dominated grasslands. As buffelgrass populations continue to grow, it will be important to be able to predict which areas are most susceptible to future buffelgrass invasion. This study attempts to provide some insight into this problem by creating a model to predict changes in the extent of potential buffelgrass habitat under different climate change scenarios between 2020 and 2100. Maximum entropy modelling was conducted using known occurrences of buffelgrass in the Santa Catalina mountains of Southern Arizona in combination with 19 bioclimatic variables from WorldClim to create a baseline model, which was then applied to future climatic conditions under the Canadian Earth Systems Model 5 (CanESM5) for three different climate change scenarios. The maximum entropy method produced an accurate model with an area under curve (AUC) value of 0.9913 and in validation trails it was able to accurately predict the presence of buffelgrass with 91.37% accuracy. When applied to future climatic conditions, the model showed a 280% increase in potential buffelgrass habitat under light and moderate climate change scenarios, and a 501% increase under a more severe scenario. Considering this potential for buffelgrass to spread, it may be essential for land managers to aggressively combat buffelgrass introductions to prevent it from being able to spread further and continue to damage ecosystems, as well as emphasize the importance of minimizing the impacts of climate change.