Daniel Fortin

This paper introduces a framework using Hidden Markov Models with Step Selection Functions to infer behavioural states from movement and environmental data, showing improved accuracy in identifying ecologically meaningful states.

The paper evaluates using indicator species from co-occurrence networks to predict biodiversity, finding they can reconstruct species assemblages but are often outperformed by climate-based models.

This paper develops a network-based method to reconstruct biodiversity using indicator species, finding that latitudinal gradients affect species composition and interaction patterns, with higher complexity in tropical regions.