Marissa F. McBride, Mark A. Burgman, and Fiona M. Fidler. University of Melbourne
As a field dominated by complex processes and often only limited data, expert opinion represents an important source of knowledge and understanding in ecology. The importance of expert knowledge for informing environmental decision-making is increasingly being recognized, but little work has focused on how this information should be obtained and what degree of accuracy we might expect. Expert judgment in disciplines such as medicine and risk management has been shown to be subject to a range of context and cognitive biases, with the elicitation process playing an important role in mitigating these effects. We review the use expert opinion in the ecological literature and in management applications and compare to current best practices in other fields. Ecological knowledge is typically represented qualitatively or based on integration of past studies, and methods for incorporating uncertainties are often limited. We compare a number of alternative approaches for eliciting expert estimates of species abundances and life history parameters, using current field studies and retrospective scenarios to allow for evaluation. Results show that elicitation method significantly affects outcomes and that eliciting quantitative representations of uncertainty and weighted expert aggregation improves performance and reliability. These findings highlight the importance of using appropriate elicitation techniques and suggest the need for additional work on evaluating experts in ecological settings.