A web-based system for public–private sector collaborative ecosystem management

 To preserve biodiversity over centuries, ecosystem management will need to be accepted and practiced by individuals from a broad spectrum of society's strata. Also, management decisions will need to be based on reliable judgments of the cause and effect relationships that govern an ecosystem's dynamics. This article describes an extant, web-based ecosystem management system (EMS) that allows (a) wide participation in ecosystem assessment and policy impact predictions, (b) convenient construction of probabilistic models of ecosystem processes through an influence diagram, and (c) automatic creation of ecosystem assessment reports. For illustration, the system is used to first model the cheetah population in Kenya, and then to assess the impact on this population of different management options. The influence diagram used herein extends standard influence diagram theory to allow representation of variables governed by stochastic differential equations, birth–death processes, and other nongaussian, continuous probability distributions. For many ecosystems, data sets on ecosystem health indicators can be incomplete, small, and contain unknown measurement errors. Some amount of knowledge of an ecosystem's dynamics however, may exist in the form of expert opinion derived from ecological theory. The proposed EMS uses a nonbayesian parameter estimation method, called consistency analysis that finds parameter estimates such that the fitted ecosystem model is as faithful as possible to both the available data and the collected body of expert opinion. For illustration, consistency analysis is used to estimate the cheetah viability influence diagram using all known cheetah surveys in the country of Kenya plus current understanding of factors such as habitat and prey availability that affect cheetah population dynamics.