COS 89-3 - The Hyper-Envelope Modeling Interface (HEMI): A novel approach to habitat suitability modeling

Wednesday, August 8, 2012: 8:40 AM
Portland Blrm 255, Oregon Convention Center
Jim Graham, Natural Resource Ecology Laborary, Fort Collins, CO, Nick Young, Natural Resource Ecology Laboratory, Colorado State University, Ft. Collins, CO, Catherine Jarnevich, Fort Collins Science Center, U.S. Geological Survey, Fort Collins, CO, Greg Newman, Natural Resource Ecology Laborary, Colorado State University, Fort Collins, CO, Paul Evangelista, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO and Thomas J. Stohlgren, Natural Resource Ecology Laboratory, Fort Collins
Background/Question/Methods

Habitat suitability maps are commonly created by modeling a species’ environmental niche from existing occurrences and environmental characteristics.  Here, we introduce the Hyper-Envelope Modeling Interface (HEMI); providing a new method for creating habitat suitability models using Bezier surfaces to model a species niche in environmental space.  HEMI allows modeled surfaces to be visualized and edited in environmental space based on expert knowledge and does not require absence points for model development and evaluation. The modeled surfaces require relatively few parameters compared to similar modeling approaches and may produce models that better match ecological niche theory.

As a case study to introduce and demonstrate HEMI, we modeled the invasive species tamarisk (Tamarix spp.) in the western United States.  We compare results from HEMI with those from existing envelope and similar modeling approaches (including BioClim, BioMapper, and Maxent).  We used synthetic surfaces to create visualizations of the various models in environmental space and used a modified Area Under the Curve (AUC) statistic as a measure of model performance.

Results/Conclusions

We show that HEMI produced better AUC values, except for Maxent.  While Maxent produced slightly better results, HEMI created a model with only 10 parameters while Maxent produced a model with over 100.  Additionally, HEMI allowed visualization and editing of the model in environmental space to develop multiple simulated potential habitat scenarios.  The use of Bezier surfaces can provide simple models that match our expectations of biological niche models and, at least in some cases, out-perform more complex approaches.