Towards Reliable Mapping of Biosecurity Risk: Incorporating Uncertainty and Decision Makers’ Risk Aversion

Pest risk maps are an important source of decision support when devising strategies to minimize introductions of invasive organisms and mitigate their impacts. When possible management responses to an invader include costly or socially sensitive activities, decision makers tend to follow a more certain (i.e. risk-averse) course of action. We present a new mapping technique that assesses pest invasion risk from the perspective of a risk-averse decision maker. We demonstrate the approach by evaluating the likelihood that an invasive forest pest will be transported to one of the continental US states or Canadian provinces in infested fi rewood that may be carried by visitors to US federal campgrounds. We test the impact of the risk aversion assumption using distributions of plausible pest arrival scenarios generated with a geographically explicit model developed from data documenting camper travel across the study area. Next, we prioritize regions of high and low pest arrival risk via application of two stochastic ordering techniques that employ, respectively, fi rstand second-degree stochastic dominance rules, the latter of which incorporates the notion of risk aversion. We then identify regions in the study area where incorporating risk aversion changes a region’s pest risk value considerably. While both methods identifi ed similar areas of highest and lowest risk, they diff ered in how they demarcated moderaterisk areas. Each method provides a tractable way to incorporate decision-making preferences into fi nal risk estimates, and thus helps to better align these estimates with particular decision-making scenarios about an organism of concern. Overall, in corporation of risk aversion helps to refi ne the set of locations that could be confi dently targeted for costly inspections and outreach activities.