Understanding responses to perturbations: from whole communities to individual species
Ecological communities are frequently perturbed through both natural and human-driven impacts. In order to avoid the loss of biodiversity and of community functioning, it is crucial to understand and predict how communities might respond to future perturbations. We have developed theoretical approaches to learn how communities and their species might respond to perturbations from past data. By combining dynamical systems theory and empirical dynamic modeling, we have introduced an approach to detect which species in a community are the most sensitive to perturbations on abundances when population dynamics are out of equilibrium (Medeiros et al 2023 Ecology Letters). We have also proposed a framework to disentangle the impact of species correlated responses on whole-community sensitivity to perturbations under non-equilibrium dynamics (Medeiros et al 2023 Ecology). These data-driven approaches allow us to quantify responses to perturbations directly from time-series data and can complement traditional indicators of vulnerability to perturbations. In addition to these data-driven approaches, we have developed model-driven approaches (e.g., assuming Lotka-Volterra population dynamics) to understand responses to perturbations. We have shown that recovery from perturbations that affect species abundances and resistance to perturbations that affect model parameters can provide similar information about the stability of a community (Medeiros et al 2021 Journal of Animal Ecology). These results contribute to our understanding of the connections between the different dimensions of ecological stability.
Understanding changes in the species composition of ecological communities
A longstanding problem in community ecology is to understand whether some changes in species composition can be more likely and predictable than others. We have used a model-driven approach based on structural stability to show that species combinations that can tolerate a larger range of environmental conditions (i.e., larger range of model parameters) are more likely to be observed (Medeiros et al 2021 American Naturalist). These results provide an explanation of why certain combinations of interacting species are observed more often in nature. We have also merged the structural stability approach with ideas from statistical mechanics to perform accurate forecasts of which species are more likely to persist under uncertainty of environmental conditions and species composition (Saavedra et al 2020 Ecology Letters).
Understanding how coevolution operates in networks of interacting species
Although coevolution between pairs of interacting species has been studied for decades, only recently we have started to investigate how coevolution shapes species traits in networks with multiple interacting species. I explored this problem in my Master’s research by merging concepts from the Geographic Mosaic Theory of Coevolution and a trait evolution model. Using this model and information on several empirical mutualistic networks, we have shown that gene flow can have the unexpected effect of increasing trait matching between interacting species (Medeiros et al 2018 PNAS).