Course of the Department of Physical Geography at Marburg University
Species distribution modelling (SDM) is a key competence for ecogeographical research and applied nature conservation. It allows researchers to estimate current distributions of species and to also predict their future distributions under climate change scenarios. However, SDM is a dynamic field characterized by rapid advancements and many uncertainties.
This course will go beyond the basic SDM knowledge and will delve into all the small and very specific challenges it presents. Participants will gain ability in generating virtual species, critically evaluating the selection of environmental predictors, identifying knowledge gaps in the field, and independently constructing and fine-tune SDMs. The main focus of the course will be on generating an experimental research setup in which we will tackle some of the uncertainties.
Intended learning outcomes
At the end of this course you should
- be familiar with virtual species and gained the ability to generate them.
- have explored the impact of selecting environmental predictors on SDM.
- be able to identified gaps in current knowledge within a specific field of research.
- be able to independently constructed your own SDMs.
- engage with various modeling strategies including tuning, variable selection, training, testing, and validation techniques.
- have designed, executed, and analyzed your own scientific experiments.
- have developed awareness regarding the capabilities and limitations of SDMs.
Course features
This course is intended as a blended learning module in our study program although the provided introductions, explanations and examples might be useful for self-study, too.
This course will take place in the classroom (F 14 | 00A19). The first session will take place on Tuesday 14.04.2026 at 09:15 am. Course material will be provided in the Ilias course environment (only accessible for members of the course who are logged-in into Ilias).
Course times
Tuesday 09:15 - 11:15.
Syllabus
| Session | Date | Topic | Content |
|---|---|---|---|
| Basics | |||
| 01 | 14.04.2026 | Basics | Welcome, R, spatial data in R, Git, Github |
| SDM Basics | |||
| 02 | 21.04.2026 | A short history of SDM | How did SDM evolve? Where are we now? |
| SDM workflow | |||
| 03 | 28.04.2026 | Train your first model | Overview, conceptualisation, data processing and model fitting for SDM |
| SDM evaluation metrics | |||
| 04 | 05.05.2026 | Model evaluation | What do we know about calculation and accuracy of evaluation metrics? |
| Exkurs: SDM with audio data | |||
| 05 | 12.05.2026 | ||
| Artificial landscapes | |||
| 06 | 19.05.2026 | Artificial landscapes | Create neutral landscape models |
| Virtual species | |||
| 07 | 26.05.2026 | Virtual species | Generate virtual species with different ecological niches |
| From idea to insight: Your personal research project | |||
| 08 | 02.06.2026 | Your research project | Select your research question and start to work on your project outline. |
| 09 | 09.06.2026 | Present your project outline | Present project outlines to peers for constructive feedback. |
| 10 | 16.06.2026 | Free working session | No input in this session. |
| 11 | 23.06.2026 | Free working session | No input in this session. Its just you and your project. Work on it and use the time to ask questions. |
| 12 | 30.06.2026 | Free working session | No input in this session. Its just you and your project. Work on it and use the time to ask questions. |
| 13 | 07.07.2026 | Free working session | No input in this session. Its just you and your project. Work on it and use the time to ask questions. |
| Synthesis |
| 14 | 14.07.2026 | Wrap up | Get and give feedback from your peers and instructors, tell us how you self-assess your skills, and happy holidays |
Deliverables
The course grading will be based on on your final SDM project.
Preparation and prerequisites
Knowledge of R and of handling spatial data is beneficial. Initial experience with species distribution modeling is helpful, for example from our basic species distribution modeling course. All software needed for this course is free and open source.
If you have no experience with R we highly recommend the base R course, which can be found here.
Team
