Prioritising the eradication of invasive species from island archipelagos with high reinvasion risk.
Eradicating invasive species from islands is a proven method for safeguarding threatened and endangered species from extinction. Island eradications can deliver lasting benefits, but require large up-front expenditure of limited conservation resources. The choice of islands must therefore be prioritised. Numerous tools have been developed to prioritise island eradications, but none fully account for the risk of those eradicated species later returning to the island: reinvasion. In this paper, we develop a prioritisation method for island eradications that accounts for the complexity of the reinvasion process. By merging spatially explicit metapopulation modelling with stochastic dynamic optimisation techniques, we construct a decision-support tool that optimises conservation outcomes in the presence of reinvasion risk. We applied this tool to two different case studies-rat (Rattus rattus) invasions in the Seaforth archipelago in New Zealand, and cane toad (Rhinella marina) invasions in the Dampier archipelago in Australia-to illustrate how state-dependent optimal policies can maximise expected conservation gains. In both case studies, incorporating reinvasion risk dramatically altered the optimal order of island eradications, and improved the potential conservation benefits. The increase in benefits was larger in Dampier than Seaforth (42% improvement versus 6%), as a consequence of both the characteristics of the invasive species, and the arrangement of the islands. Synthesis and applications. Our results illustrate the potential consequences of ignoring reinvasion risk. We recommend that reinvasion risk be explicitly included in any island eradication prioritisation involving an archipelago, particularly when some islands are close to the mainland.