Modelling species niches using expert-based range maps

How well do models based on range maps curated by experts perform compared to models based on real occurrences?

Data resources used via GBIF : 4,769 species occurrences
Iberian worm lizard (Blanus cinereus)

Iberian worm lizard (Blanus cinereus), one of the species modelled in the study. Observed by Cesar Pollo licensed under CC BY-NC 4.0.

Species distribution models (SDMs) based on occurrence and environmental data are used extensively to model species’ niches and predict e.g. the spread of invasive species or responses to changing climates. Such models can only perform as well as the data that feeds them, and are thus vulnerable to biases in the underlying data.

This study considers IUCN-derived expert-based range maps as an alternative source of species distribution information, and evaluate the modelling performance of such data compared to models using empirical occurrence data. Although using the range maps as a base for SDM involves sampling pseudo-occurrences, the researcher finds that while comparing the predicted distributions of 85 species based on either GBIF-mediated occurrences or on expert-based range maps, the two approaches produce similar results.

The author concludes by arguing that expert-based maps may be used for SDM when true occurrence data is limited or known to be biased.


Fourcade Y (2016) Comparing species distributions modelled from occurrence data and from expert-based range maps. Implication for predicting range shifts with climate change. Ecological Informatics. Elsevier BV 36: 8–14. Available at: