Community dynamics in a changing world: Tipping, tracking, and early-warning signals
Monday, August 2, 2021
Link To Share This Presentation: https://cdmcd.co/w3QMa7
Frithjof Lutscher, Department of Mathematics and Statistics, University of Ottawa, Ottawa, ON, Canada, Ramesh Arumugam, Department of Biology, McGill University, QC, Canada and Frédéric Guichard, Department of Biology, McGill University, Montreal, QC, Canada
Department of Mathematics and Statistics, University of Ottawa Ottawa, ON, Canada
Background/Question/Methods Persistence of ecological communities under climate change and human activities depends on how fast the environment is changing and on how species respond to that change. Theories of alternate stable states have been used to predict regime shifts of ecosystems as equilibrium responses to sufficiently slow environmental change. While it is known that the actual rate of environmental change is a key factor affecting the response, most existing models of community dynamics are based on equilibrium analysis and can therefore not predict how transient behaviour and regime shifts depend on the rate of environmental change. Our work contributes to non-equilibrium theory that explicitly considers the influence of this rate of environmental change. We use a two-patch metacommunity model where habitat quality or connectivity change at a certain rate. Through numerical simulation, we study how regime shifts depend on the rate of environmental change and compare the outcome with a stability analysis in the corresponding constant environment. Results/Conclusions Our analysis reveals a novel phenomenon that we call “tracking unstable states”: as environmental conditions change, the metacommunity remains at equilibrium states that have become unstable due to environmental change. This tracking leads to not only a delayed response but to qualitatively different regime shifts, including local extinctions, that are not predicted by alternative stable state theory or steady-state analysis. Instead, we find that in stochastic simulations, early warning signals predict changes in community state away from bifurcation points reasonably well. Our work highlights the importance of including the rate of change explicitly into model formulation and of studying the resulting transient dynamics. A community may appear stable even though the underlying dynamics are unstable and the community is headed for extinction. Thus, this study reveals how the rate of environmental change reshapes community responses and predicts community persistence away from equilibrium states and also away from critical points.