Search

Remote islands harbour many endemic species and unique ecosystems. They are also some of the world's most human-impacted systems. It is essential to understand how island species and ecosystems behaved prior to major anthropogenic disruption as a basis for their conservation. This research aims to reconstruct the original, pre-colonial biodiversity of a remote oceanic island to understand the scale of past extinctions, vegetation changes and biodiversity knowledge gaps. We studied fossil remains from the North Atlantic island of Corvo (Azores), including pollen, charcoal, plant macrofossils, diatoms and geochemistry of wetland sediments from the central crater of the island, Caldeirão. A comprehensive list of current vascular plant species was compiled, along with a translation table comparing fossilized pollen to plant species and a framework for identifying extinctions and misclassifications. Pollen and macrofossils provide evidence for eight local extinctions from the island's flora and show that four species listed as 'introduced' are native. Up to 23 % of the pollen taxa represent extinct/misclassified species. Corvo's past environment was dynamic, shifting from glacial-era open vegetation to various Holocene forest communities, then almost completely deforested by fires, erosion and grazing following Portuguese colonisation. Historical human impacts explain high ecological turnover, several unrecorded extinctions and the present-day abundance of vegetation types like Sphagnum blanket mire. We use Corvo as a case study on how fossil inventories can address the Wallacean and Hookerian biodiversity knowledge gaps on remote islands. Accurate baselines allow stakeholders to make informed conservation decisions using limited financial and human resources, particularly on islands where profound anthropogenic disruption occurred before comprehensive ecological research. [Display omitted] • Human impacts transformed island vegetation and soils • Evidence for unrecorded extinctions and almost complete forest loss • Framework for determining native/introduced status from fossils • Key baseline information for habitat restoration [ABSTRACT FROM AUTHOR]

Protected areas (PAs) have been selected using either subjective or objective criteria applied to an extremely limited subset of biodiversity. Improved availability of species distribution data, better statistical tools to predict species distributions and algorithms to optimize spatial conservation planning allow many impediments to be overcome, particularly on small islands. This study analyses whether 219 species are adequately protected by PAs on Pico Island (the Azores, Portugal), and if they are as efficient as possible, maximizing species protection while minimizing costs. We performed distribution modelling of species’ potential distributions, proposed individual conservation targets (considering the context of each species in the archipelago and their current conservation status) to determine the efficiency of current PAs in meeting such targets and identify alternative or complementary areas relevant for conservation. Results showed that current PAs do not cover all taxa, leaving out important areas for conservation. We demonstrate that by using optimization algorithms it is possible to include most species groups in spatial conservation planning in the Azores with the current resources. With increasing availability of data and methods, this approach could be readily extended to other islands and regions with high endemism levels. [ABSTRACT FROM AUTHOR]