Optimizing search strategies for invasive pests: learn before you leap.

Published online
09 Feb 2011
Content type
Journal article
Journal title
Journal of Applied Ecology

Baxter, P. W. J. & Possingham, H. P.
Contact email(s)

Publication language
Australia & Queensland


Strategic searching for invasive pests presents a formidable challenge for conservation managers. Limited funding can necessitate choosing between surveying many sites cursorily, or focusing intensively on fewer sites. While existing knowledge may help to target more likely sites, e.g. with species distribution models (maps), this knowledge is not flawless and improving it also requires management investment. In a rare example of trading-off action against knowledge gain, we combine search coverage and accuracy, and its future improvement, within a single optimisation framework. More specifically we examine under which circumstances managers should adopt one of two search-and-control strategies (cursory or focused), and when they should divert funding to improving knowledge, making better predictive maps that benefit future searches. We use a family of Receiver Operating Characteristic curves to reflect the quality of maps that direct search efforts. We demonstrate our framework by linking these to a logistic model of invasive spread such as that for the red imported fire ant Solenopsis invicta in south-east Queensland, Australia. Cursory widespread searching is only optimal if the pest is already widespread or knowledge is poor, otherwise focused searching exploiting the map is preferable. For longer management timeframes, eradication is more likely if funds are initially devoted to improving knowledge, even if this results in a short-term explosion of the pest population. Synthesis and applications: By combining trade-offs between knowledge acquisition and utilization, managers can better focus - and justify - their spending to achieve optimal results in invasive control efforts. This framework can improve the efficiency of any ecological management that relies on predicting occurrence.

Key words