Surrogates for Resilience of Social–Ecological Systems

From its roots in ecology, resilience (Holling 1973) has more recently been applied to social-ecological systems, or SES. Theories of changing resilience explicitly address the persistence or breakdown of diverse states of complex systems (Gunderson and Holling 2002). These ideas have attracted interest from interdisciplinary research groups interested in change, conservation or restoration of SES (Berkes and others 2003; Scheffer and others 2003). From a practical standpoint, resilience theory provides a conceptual foundation for sustainable development (Folke and others 2002). The transition from the ory to practice, however, requires assessment or estimation of resilience (Carpenter and others 2001). So far, there is little experience with esti mating resilience of SES, and little understanding of the sensitivity of resilience measures to changes in SES. This shortage of practical field experience is a barrier to building understanding through empirical study of resilience in SES. Direct measurement of resilience is difficult be cause it requires measuring the thresholds or boundaries that separate alternate domains of dynamics for SES. The only sure way to detect a threshold in a complex system is to cross it (Carpenter 2003). Yet threshold-crossings do not occur very often. In the natural sciences, much understanding of thresholds has come from delib erate manipulations of ecosystems, or before/after studies of large disturbances (Turner and Dale