Adapting the concept of functionally dominant species for observational data

Conservation ecologists often rely on surrogate species to identify biodiversity hotspots due to the high cost of monitoring programs. While the keystone species approach is an appealing framework for that purpose, it has been criticized for its lack of a clear threshold to identify functionally important species and for its limited ability to handle observational data variability. Here, we propose a modified version of the functionally dominant species (FDS) framework using a bootstrapping random sampling method implemented with either strict or flexible parameters to identify species that disproportionately contribute to the increase or the decrease of biodiversity. We tested our approach on plant, bird, and fish communities of 37 lake-edge wetlands. We identified eight FDS using a 95% confidence interval, of which two displayed a positive contribution to diversity while six had a negative contribution. Using a 99% confidence interval, we found four FDS, all displaying a negative contribution to biodiversity. Most of the identified FDS had ecological or biological traits that support their disproportionate impact on biodiversity. By addressing the limitations of the keystone species framework and providing a statistical framework for analyzing observational data, our method represents a promising tool for conservation ecology.