Anyone interested in expanding their modeling skillset for the benefit of species conservation.
Summary and Objectives:
This course will provide participants with the skills needed to build species distribution models (SDMs) and habitat suitability maps that have become increasingly important for conservation planning. Topics explored include processing species occurrence and environmental datasets, model building, prediction and evaluation, and development of habitat suitability maps. Participants will learn to apply these skills in R statistical programming language through hands-on exercises. Three modeling methods will be covered: logistic regression, generalized linear mixed models and Random Forests.
Upon completion of this course, participants will be able to:
- Use R for spatial analyses and predictive modeling.
- Collate and process environmental predictors (raster layers).
- Apply logistic regression, generalized linear mixed models and Random Forest methods to build predictive models.
- Evaluate and validate model fit and performance.
- Develop and interpret habitat suitability maps based on model output.
- Fit kernel density estimates to estimate home range.
- Generate pseudo-absences for presence only data
Statistical Modeling for Conservation (CSP4210) and GIS Intermediate (CSP7204), or the equivalent experience (per Course Leader approval). Participants must also have a working knowledge of R Statistical programming language.