Informing spatial conservation prioritization with species’ traits

New Guinea, the most botanically diverse island on the planet, is the location for one of the boldest conservation initiatives. The Manokwari Declaration aims to achieve 70% conservation designation for the Bird's Head Peninsula. This is 40% higher than the 2022 Global Biodiversity Framework target. However, there is a lack of species occurrence data to support evidence of where biodiversity can be best protected. To address this, we integrated plant trait data from taxonomic descriptions in species occurrence models that can inform conservation planning. Inclusion of traits improved the performance of co-occurrence models of ∼800 plant species across the 100,000-km² landscape. Traits generally improved model performance, but not all traits contributed equally (e.g., leaf size and red flower color most improved accuracy of occurrence prediction). Likewise, trait-parameterized models tended to be most useful with rare species occurrence prediction, but this was inconsistent among traits. Under 70% protection, three-quarters of the areas selected conserved trait diversity. Critically, trait diversity also increased the chances that areas at high risk of deforestation were selected as conservation priorities. Overall, we found that plant species' traits, often key parameters of ecosystem function and resilience, improved spatial conservation planning.