Efficiently locating conservation boundaries: Searching for the Tasmanian devil facial tumour disease front

Conservation management actions and decisions are often defined by the location of ecological boundaries, for example, the present range of invasive or threatened species. The position of these boundaries can be cryptic, and managers must therefore directly sample sites, an expensive and time-consuming process. While accurate boundary location techniques have been considered by ecological theorists, the issue of cost-effective, or optimal boundary location has not. We propose a general framework for boundary location which incorporates both cost-efficiency and uncertainty. To illustrate its application, we use it to help locate an infectious disease front in the endangered Tasmanian devil population. The method ensures optimal spatial sampling by maximizing the expected information gained from each sample. When resources are limited, our method provides more accurate estimates of the boundary location than traditional sampling protocols. Using a formal decision theory sampling design encourages economically efficient actions, and provides defensible and transparent rationale for management actions.