Allocating surveillance resources to reduce ecological invasions: maximizing detections and information about the threat

Allocating resources to detect invasive pests, diseases, and pathogens on exposure pathways requires a trade-off between the need to detect as many contaminated items as possible and the need to acquire knowledge about contamination rates. We develop a model and an algorithm that provide guidance for the allocation of inspection resources across multiple dynamic pathways in cases where not every item can be inspected. The model uses a null hypothesis that the contamination rate of a pathway is above a specified level: a risk cutoff. Pathways with a risk above the cutoff are fully inspected, and those with a risk below the cutoff level are monitored at a rate that would detect a change of the risk to being above the cutoff level with high probability. We base our decision on the 95% upper confidence limit for the contamination rate. We demonstrate via simulations and a data set that focusing inspection resources on specific pathways can result in substantially more effective intervention, and that the reduction in overall effectiveness of monitoring low-risk pathways need not be substantial. Use of the model demands the selection of the risk cutoff, and this limit can be set according to projected consequences.