Habitat modelling in GIMOLUS – webGIS-based e-learning modules using logistic regression to assess species-habitat relationships

Habitat modelling is a tool for regionalisation of biotic information by predicting the spatially explicit distribution of species on the basis of environmental properties. Since this recent method has a high relevance in questions of nature conservation and ecological research, there is a need to teach it in environmental studies. The multimedia GIMOLUS-environment provides all elements necessary to teach and learn the application of this method in a realistic setting. It’s constituting an innovation to environmental courses as it introduces the proceeding of a complete modelling procedure on practical examples offering a high interactivity level on low technical requirements. The internet-based learning unit ‘habitat modelling with logistic regression’ uses webGIS applications to illustrate the sampling procedure and the spatial extrapolation of habitat models. The learning unit comprises six learning modules. Spatially explicit habitat modelling is presented starting from sampling, through model evaluation, to predictions of the probability of species occurrence. Step by step, the complete model calculation as well as the evaluation are proceeded in the GIMOLUS-environment, as shown in an exemplary case study. The data sets used in the analyses of species-habitat-relationships are sampled in a virtual landscape.