Protected areas support more species than unprotected areas in Great Britain, but lose them equally rapidly

Protected areas are a key conservation tool, yet their effectiveness at maintaining biodiversity through time is rarely quantified. Here, we assess protected area effectiveness across sampled portions of Great Britain (primarily England) using regionalized (protected vs unprotected areas) Bayesian occupancy-detection models for 1238 invertebrate species at 1 km resolution, based on similar to 1 million occurrence records between 1990 and 2018. We quantified species richness, species trends, and compositional change (temporal beta diversity; decomposed into losses and gains). We report results overall, for two functional groups (pollinators and predators), and for rare and common species. Whilst we found that protected areas have 15 % more species on average than unprotected ones, declines in occupancy are of similar magnitude and species composition has changed 27 % across protected and unprotected areas, with losses dominating gains. Pollinators have suffered particularly severe declines. Still, protected areas are colonized by more locally-novel pollinator species than unprotected areas, suggesting that they might act as 'landing pads' for range-shifting pollinators. We find almost double the number of rare species in protected areas (although rare species trends are similar in protected and unprotected areas); whereas we uncover disproportionately steep declines for common species within protected areas. Our results highlight strong invertebrate reorganization and loss across both protected and unprotected areas. We therefore call for more effective protected areas, in combination with wider action, to bend the curve of biodiversity loss - where we provide a toolkit to quantify effectiveness. We must grasp the opportunity to effectively conserve biodiversity through time.