Acceptance sampling for cost-effective surveillance of emerald ash borer in urban environments

Abstract

We develop an acceptance sampling approach for surveillance of the emerald ash borer (EAB), a harmful forest pest, in Winnipeg, Canada. We compare sampling strategies computed with two different management objectives. The first objective maximizes the expected area with detected infestations and the second objective minimizes the expected number of undetected infested trees in sites that were not inspected or where inspection did not find an infestation. The choice of the management objective influences the survey strategy: achieving the first objective involves selecting sites with high infestation rates proximal to the infested area, whereas the second objective requires inspecting sites with both high infestation rates and high host densities. Adding uncertainty prescribes inspecting a larger area with lower sampling rates and extending the surveys to farther distances from the infested locations. If a decision maker wants to minimize the worst-case damage from failed detections, the optimal strategy is to survey more sites with high host densities at farther distances, where EAB arrivals could cause significant damage if not detected quickly. Accounting for the uncertainty addresses possible variation in infestation rates and helps develop a more diversified survey strategy. The approach is generalizable and can support survey programmes for new pest incursions.