A declining rate of recovery following disturbance has been proposed as an important early warning for impending tipping points in complex systems. Despite extensive theoretical and laboratory studies, this ‘critical slowing down' remains largely untested in the complex settings of real-world ecosystems. Here, we provide both observational and experimental support of critical slowing down along natural stress gradients in tidal marsh ecosystems. Time series of aerial images of European marsh development reveal a consistent lengthening of recovery time as inundation stress increases. We corroborate this finding with transplantation experiments in European and North American tidal marshes. In particular, our results emphasize the power of direct observational or experimental measures of recovery over indirect statistical signatures, such as spatial variance or autocorrelation. Our results indicate that the phenomenon of critical slowing down can provide a powerful tool to probe the resilience of natural ecosystems.

An increase in the frequency and intensity of environmental crises associated with accelerating human-induced global change is of substantial concern to policy makers. The potential impacts, especially on the poor, are exacerbated in an increasingly connected world that enables the emergence of crises that are coupled in time and space. We discuss two factors that can interact to contribute to such an increased concatenation of crises: (1) the increasing strength of global vs. local drivers of change, so that changes become increasingly synchronized; and (2) unprecedented potential for the propagation of crises, and an enhanced risk of management interventions in one region becoming drivers elsewhere, because of increased connectivity. We discuss the oil-food-financial crisis of 2007 to 2008 as an example of a concatenated crisis with origin and ultimate impacts in far removed parts of the globe. The potential for a future of concatenated shocks requires adaptations in science and governance including (a) an increased tolerance of uncertainty and surprise, (b) strengthening capacity for early detection and response to shocks, and (c) flexibility in response to enable adaptation and learning.

The question of what and how to measure ecological resilience has been troubling ecologists since Holling 1973's seminal paper in which he defined resilience as the ability of a system to withstand perturbations without shifting to a different state. This definition moved the focus from studying the local stability of a single attractor to which a system always converges, to the idea that a system may converge to different states when perturbed. These two concepts have later on led to the definitions of engineering (local stability) vs ecological (non-local stability) resilience metrics. While engineering resilience is associated to clear metrics, measuring ecological resilience has remained elusive. As a result, the two notions have been studied largely independently from one another and although several attempts have been devoted to mapping them together in some kind of coherent framework, the extent to which they overlap or complement each other in quantifying the resilience of a system is not yet fully understood. In this perspective, we focus on metrics that quantify resilience following Holling's definition based on the concept of a stability landscape. We explore the relationships between different engineering and ecological resilience metrics derived from bistable systems and show that, for low dimensional ecological models, the correlation between engineering and ecological resilience can be high. We also review current approaches for measuring resilience from models and data, and we outline challenges which, if answered, could help us make progress towards a more reliable quantification of resilience in practice.