One important challenge of community ecology research in the face of climate change is the prediction of ecological surprises from relatively slow changes in the environment. Theories of dynamical systems can predict critical values of environmental factors that lead to sudden shifts in the state and dynamical behavior of ecosystems: forests shifting to prairies, diversity loss in lakes shifting from clear to turbid water, equilibrium population abundance shifting to oscillations. This body of theory has led to the discovery of early warning signals. These are quantities that can be obtained from time series analysis and that aim to characterize the dynamics of ecosystems as they approach critical values in order to predict possible shifts. A different but related phenomenon are long transients, which occur if an ecosystem remains in an apparently stable state for a long time before it switches to other states on comparatively short time scales. Theory has identified several mechanisms that can lead to long transients in ecological systems, some related to critical values and regime shifts (above), others not. More recent studies have contributed a new perspective to the theme of sudden shifts by showing that not only the actual state of the environment but also the rate at which it changes is a key predictor of ecosystem response to change. These studies also show that knowledge of unstable ecosystem states is crucial for the prediction of ecosystem responses to environmental change. Rate-dependent ecosystem responses can occur away from critical points and stable ecological states. Combined, these theories force us to abolish assumptions of equilibrium and stable ecosystem states and to integrate the environment explicitly into models for predictions.This organized session will bring together key contributors from the different fields to this problem with the goal of revealing general predictions from individual studies conducted across model systems, from single populations to trophic interactions, and across ecological and evolutionary processes. Another goal is to make progress on key ecological concepts that can contribute to our understanding of rate-dependent responses and to their ecological interpretation. These include generalizing concepts of early warning signals and transient dynamics. Finally, we will initiate a broader discussion on the integration of rate-dependent responses to macroecological approaches into the study and management of species diversity under climate change.