Abstract
Abstract
Sustainable wildlife harvest is crucial for ensuring ecological balance while maintaining healthy population offtake levels in the long term. Yet a substantial share of hunting is illegal and spatially undocumented, risking declines that pass unnoticed until recovery is no longer possible. Existing threat maps rarely distinguish illegal from legal harvest, and site-level poaching studies do not scale to global conservation planning. Here we develop a global predictive model of poaching probability for terrestrial mammals, integrating 1,168 georeferenced poaching occurrences with ecological and anthropogenic drivers. Hunted mammal richness and accessibility to cities were the strongest predictors, while economic effects varied with income group. Poaching probability peaked across Southeast Asia, sub-Saharan Africa, and parts of the Americas and Europe, with species richness dominant across most regions and accessibility dominant in remote areas. This work provides a scalable tool for prioritising enforcement, protected-area investment, and demand-reduction policy under the Global Biodiversity Framework.