A B S T R A C T Protected Areas (PAs) are core components of conservation strategies, but the networks they form are rarely assessed for their effectiveness over time. We tested different aspects of effectiveness of the British PA network in achieving long-term biodiversity outcomes, including species representativeness of initial location choices and network resilience (in terms of species persistence). Using 10 × 10 km cells, 'landscapes', with contrasting cover of protected areas managed

more »

... ifically for biodiversity conservation, we evaluated these aspects of effectiveness by analysing species distribution changes of over 2800 species of animals and plants from 1974 to 2014. Landscapes that contained PAs in 1974 had higher species representativeness than landscapes without PAs, but landscapes with low PA coverage (median). Many species distributions have declined since 1974, and the distributional trends of declining and priority species were similar (on average) in landscapes containing PAs and in the wider countryside, implying PA-containing landscapes were not resilient to landscape-scale pressures. Nonetheless, PAs did have a small positive impact over time on landscape-scale representation trends of declining species, and priority species. Regardless of PA coverage, topographically heterogeneous landscapes were more likely to retain priority species between 1974 and 2014, and less likely to be colonised by expanding species. Despite landscapes with low PA coverage disproportionately contributing to overall PA network representativeness, they are less resilient than landscapes with high PA coverage, which jeopardises their value in the long-term and will require landscapescale habitat conservation and restoration to address.