Hot topics in species distribution modelling

Guest lecture by Dr Alberto Jiménez-Valverde (Museo Nacional de Ciencias Naturales, Madrid, Spain)

Species distribution modelling (SDM), i.e., estimating species distributions by using presence-absence data and a set of ecogeographical predictors, has become a standard approach in biodiversity research. By mostly taking up already existing techniques from other research areas, the SDM field has grown exponentially since the 90s, an expansion that has neither been rooted nor accompanied by the deep theoretical thinking necessary.

In this talk it is my intention to discuss and ponder what, in my opinion, are key issues in distribution modelling, although surprisingly not all of these topics are central in the current mainstream literature. First, I will present the favourability function and I will explain the advantages of favourability over probability of presence or suitability. Second, I will discuss the problem in the selection of the geographic extent of analysis and its relationship with the niche vs. habitat vs. distribution debate. Third, I will talk about one of the most important sources of uncertainty in SDM, the selection of the predictors. Finally, I will discuss my latest ideas on models´ evaluation.

Publisert 4. juni 2014 11:15 - Sist endret 11. apr. 2022 14:22