For this module it is recommended that you take GEMSx006.1, GEMSx006.2 and GEMSx006.3 or GEMSx006.6, GEMSx006.7 and GEMSx006.8 beforehand or have some understanding of species distribution models and how and why we need them as well as knowledge of how to access species occurrence data and environmental predictors that can be used as an input for species distribution models. Knowledge of R or other geospatial and scientific data analysis software or language and specialized species distribution modeling software like CLIMEX, R based SDM packages or MAXENT are required.
Successful completion of this course will equip any student, researcher, practitioner, or extension worker with the ability to conduct sound and robust species distribution modelling. It starts from choosing the right SDM approach for the kind of biological information, occurrence dataset or environmental predictors to which we have access. Even though the focus will be on correlative SDMs, the discussion also will cover mechanistic SDMs. Throughout this practical course we will learn best practices that will help optimize developing, parametrizing and running species distribution models (SDMs). R data analysis software will be the major platform for most of the SDMs, however we will also learn how to run MaxEnt models for presence-only data.
This workshop is part of a 5-module series on species distribution modeling to provide researchers the ability to undertake a spatio-temporal accounting of biotic threats to crops, natural landscapes or the human society.
- Correlative vs. mechanistic models
- Environmental niche modeling
- Training and testing SDM models
- Validating SDM model results