Abstract Reversing ongoing biodiversity loss requires effectively managing anthropogenic threats. Although conservation actions can mitigate these threats, prioritizing and targeting such actions remain challenging because species respond differently depending on the nature of the threats and their biological traits. We developed an optimization framework that explicitly considers the abatement of individual threats and species‐specific responses to conservation outcomes. We applied the approach in the Peruvian Amazon and selected sites where a shift from conventional agriculture to agroforestry could reduce negative impacts on Red List bird species. We measured species recovery benefit from agroforestry interventions by building on the Species Threat Abatement and Restoration (STAR) metric and using data on the Red List status of bird species, the proportion of species’ ranges impacted by agriculture that is targeted for a conservation intervention, and the expected severity of the threat from agriculture for the species. Interventions selected with our approach achieved 43% of the total potential species recovery benefit when all sites were selected compared with the 28% achieved with a traditional return‐on‐investment‐based approach (∼55% improvement) and the 14% achieved with a traditional species‐richness‐based approach (∼201% improvement) for the same target budget. Furthermore, focusing on species expected to experience the greatest impacts, a common indicator in conservation planning, resulted in substantially less site coverage than the traditional return on investment (∼41–60%) across budgets, indicating the importance of selecting fewer but more valuable sites. These results demonstrate that modeling species‐specific responses to the abatement of heterogeneous threats can improve efficiency in conservation planning for biodiversity but that sites targeted for intervention and outcomes vary depending on whether all species or only the species expected to experience the greatest impacts are prioritized.
Guerrero‐Pineda et al. (Sat,) studied this question.