Ecosystems are forecasted to face an increasing number of external pulses that are extreme in their intensity and duration. However, the theory of how ecosystems will respond to changes in external forcing (climate change, eutrophication, etc.) has focused primarily on changes in long-term equilibria in response to slowly changing conditions or how external pulses will affect the variation around a single equilibrium. In contrast, we know relatively little about what types of shorter perturbation are likely to force transitions to new self-reinforcing states or result in ecosystem responses that exhibit slow recovery to their former state. We applied pulses of different intensities and duration to spatially explicit models of systems with underlying thresholds. These thresholds can make the ecosystems prone to regime shifts to new self-reinforcing states at the patch and ecosystem scale. Specifically, we asked how spatial heterogeneity and connectivity affect ecosystems’ resistance (displace during a pulse), engineering resilience (recovery rate after a pulse), and ecological resilience (whether the systems failed to return to its previous state).
Results/Conclusions
In general, all three response metrics depended on both the intensity and duration of external pulses. Even external responses with an extreme intensity must be maintained for several time-steps before prolonged recovery (low engineering resilience) or a regime shift occurred (exceeded ecological resilience). Less heterogeneous systems showed universally low resistance and resilience to multiple types of pulses. Systems with moderate to high heterogeneity, but little spatial connectivity, have low ecosystem-scale engineering and ecological resilience (the recovery across all patches), but exhibit some strong local-scale resilience (ability of some patches to recover). Local- and ecosystem-scale resilience was maximized with moderate to high heterogeneity and moderate connectivity. In contrast, heterogeneous, but highly connected systems are more prone to large-scale regime shifts. Our results present testable hypotheses about how spatial properties affect the resistance and resilience to large-scale, extreme changes in driver variables. The results suggest that for many ecosystems, a loss of spatial connectivity and heterogeneity will increase the probability of unwanted regime shifts at multiple scales, or at least results in slower recovery times after perturbations.