A Dynamical Systems Framework for Resilience in Ecology

Rising interest in the resilience of ecological systems has spawned diverse interpretations of the term's precise meaning. This paper classifies and explores definitions of resilience from the ecological literature using a dynamical systems framework. A model consisting of ordinary differential equations is assumed to represent the ecological system. The question " resilience of what to what? " posed by Carpenter et al. [2001] informs two broad categories of definitions, based on resilience to state variable perturbations and to parameter changes, respectively. Definitions of resilience to state variable perturbations include measures of basin size (relevant to one-time perturbations) and of basin steepness (relevant to repeated perturbations). Resilience to parameter changes can be quantified by viewing parameters as state variables, but could also take into account reversibility. The dynamical systems viewpoint yields the following key insights: (i) system properties that confer resistance to state variable perturbations can differ from those that promote recovery from them, (ii) recovery rates, oft deemed the purview of " engineering resilience, " do matter for ecological resilience, and (iii) resilience to repeated state variable perturbations correlates with resilience to parameter changes due to the critical slowing down phenomenon.