In this application, we demonstrate how to spatially model biodiversity and the utility of Boosted Regression Trees (BRT) using methods and modified code from Elith et al. (2008). This application uses a combination of remotely sensed and GIS environmental layers to model floral diversity in the Bale Mountain, Ethiopia. The BRT model determines mathematical relationships between physical characteristics of a landscape (i.e., environmental layers) and unique species count data collected at geo-referenced field plots (i.e., response variable), to make spatially explicit predictions across the larger study site.
Modeling Spatial Biodiversity using Boosted Regression Trees PDF | DATA