Project in the picture

Incorporating microclimate into species distribution models

Anthropogenic activities have led to a worldwide decline in biodiversity. The current rate of species extinction is estimated to be 100-1000 times faster in comparison to eras without influential human activities. Some researchers even warn us that the earth is heading towards its sixth mass extinction event. Besides the loss of natural habitats and the spread of invasive alien species, climate change has been identified as one of the most important human-induced global drivers of biodiversity loss. To quantify and visualize the effects of climate change on biodiversity, species distribution models (SDMs) are a commonly used technique. SDMs link species occurrence data with corresponding climate and/or soil predictors. The modeled relationship can then be used to associate future climate scenarios with suitable areas for those species.


Within ecological studies, macroclimatic predictors (e.g. WorldClim, CHELSA) with a coarse spatial resolution (more than 1 square kilometer) are often used. As this macroclimate is interpolated from standardized weather stations in short, open grasslands (1.5 to 2 meters aboveground), it only represents average ambient air conditions as if forests did not exist. These climatic grids indeed ignore many climate-forcing processes that operate near the ground surface (i.e. the microclimate), yet these impact the habitat conditions for plant species significantly.


We will use and further expand the SoilTemp database ( containing in situ microclimatic measurements. Then we will link the microclimate with biophysical and physiographic predictor variables through machine-learning techniques. Thereafter we will use the microclimate map in order to predict the effects of climate change on forest plant species.


The project is coordinated by the division of Forest, Nature and Landscape (KU Leuven) and Fornalab (Ghent University), Belgium, by Stef Haesen (stef.haesen[AT], Pieter De Frenne and Koenraad Van Meerbeek, in collaboration with the FLEUR network.