Predictive ecology

Interactions within ecosystems are really complex, and we do not have a lot of data on them. We build models to predict the interactions between species in space and time. We then tie these predictions on network structure to functional properties of ecological networks.

These models are required to forecast the structure of novel communities that will emerge through global changes. They also require to develop new tools to merge data, and improved statistical models to fill in the gaps between the different datasets.

This research reveals an important aspect of the dissimilarity of ecological communities: the same species can interact in different ways. This opens a lot of stimulating questions. Do species and interactions respond to the same environmental cues and variables? Which are more informative? What is the spatial scale at which variations ceases to matter? How much of this variation is random?

We are particularly interested in the relationship between network structure and ecosystem functioning, but also work on landscape connectivity, and the transmission of both ideas and diseases. Our flagship project on this theme is to develop networks of habitats patches, to help populations move across highly fragmented areas.