Introduction to SDMs: theory and practice in R

Bob Muscarella

\(~\)

Introduction

This course was initially developed as a 3-day workshop at Sapienza University in June, 2021. Students have various options to work through at their own pace, following different specific lessons based on their prior knowledge and goals.

Learning objectives: In general, by the end of this course, you will be able to:

• Know the basic theory and concepts behind SDMs / ENMs
• Design, build and evaluate SDMs / ENMs using automated R scripts
• Understand the strengths and limitations of SDMs / ENMs for different purposes
• Use SDMs / ENMs to describe, predict, and project species distributions in space and time

\(~\)

Before the course

Reading

Please read these papers before the course begins as they provide important background information. We will discuss them on Day 1 of the course.

• Elith, J., & Leathwick, J. R. (2009). Species distribution models: ecological explanation and prediction across space and time. Annual Review of Ecology, Evolution, and Systematics 40, 677-697.
• Elith, J., & Graham, C. H. (2009) Do they? How do they? WHY do they differ? On finding reasons for differing performances of species distribution models. Ecography 32(1), 66-77.
• Merow et al. (2014) What do we gain from simplicity versus complexity in species distribution models? Ecography, 37, 1267–1281.

R Exercises

To prepare for the course, we will use some tutorials produced by the excellent Data Carpentry organization. Follow through the exercises below to refresh your general R skills and get you started with geospatial data and analyses. You are free to skip around in these materials based on your prior knowledge but I encourage you to follow through everything.

1. Data Carpentry: Introduction to R and RStudio
2. Data Carpentry: Introduction to Geospatial Concepts
3. Optional: Introduction to Geospatial Raster and Vector Data with R. To complete this exercise, you will first need to follow the setup instructions.

Note: If you are very new to R and want a more detailed introduction, I recommend: Data Carpentry: Data Analysis and Visualization in R for Ecologists. Note that going through this entire exercise will take about one full day.

\(~\)

Day 1: Introduction: data acquisition and cleaning

Morning session (Theory)

• Overview and introductions
• Lecture 1: The what, how, why of SDMs/ENMs
• Discussion of pre-course readings
  • What did you find most interesting / confusing?
  • What seems to be the most challenging aspect of building species distribution model?

Afternoon session (Practical)

Obtaining and cleaning data

• Exercise 1: Species data
• Exercise 2: Environment data

\(~\)

Day 2: Models: algorithms and evaluations

Morning session (Theory)

• Lecture 2: Model Evaluation (guest lecture by Jamie Kass). Video recording here

Afternoon session (Practical)

• Discussion of background (psuedo-absence) point selection. Video recording here
• Exercise 3: Overview of algorithms
• Exercise 4: Model evaluation

\(~\)

Day 3: Applications: possibilities and precautions

Morning session (Theory)

• Lecture 3: Applications: Potentials and pitfalls. Video recording here

Afternoon session (Practical)

• Live walk-through ENMeval vignette: Video recording here
• Continue exercises, apply knowledge to individual projects