Improving biodiversity protection through artificial intelligence

Over a million species face extinction, highlighting the urgent need for conservation policies. This paper introduces CAPTAIN, a reinforcement learning framework for spatial conservation prioritization. CAPTAIN consistently outperforms existing software using simulated and empirical data, optimizing the trade-off between cost and biodiversity protection across multiple metrics. Under budget constraints, it protects significantly more species than random or naive methods. Regular biodiversity monitoring, even with citizen science data, further enhances outcomes. AI offers significant potential for improving conservation in a resource-limited world.

Daniele Silvestro, Stefano Goria, Thomas Sterner, Alexandre Antonelli