The extinction risk of species is globally monitored by the IUCN Red List (RL). Despite the goal to update extinction risk assessments at least every 10 years, over 21,000 species (ca. 18% of those assessed) currently have outdated assessments. The resources needed to keep assessments up to date present a clear sustainability challenge. Furthermore, even a 10-year reassessment interval is too long to allow timely conservation interventions in many cases, and the application of RL criteria often suffer from inconsistencies across different groups. Several authors have proposed automated methods to conduct preliminary screenings to guide reassessment of species, considering species intrinsic vulnerability (inferred from their traits and other factors), historical trends, and satellite-derived imagery combined with macroecological modelling. So far, however, these approaches have remained largely academic exercises, and have rarely been incorporated in the official RL reassessment routines.
By bringing together RL practitioners and ecological modellers, we aim to develop an innovative, rapid and consistent framework for prioritizing RL assessments. We will build on recent approaches to derive input data for the RL assessment process, for example inferring population decline from satellite-borne estimates of habitat loss. We will automatize the identification of species likely to have changed their RL status. These species would be prioritised for re-assessment, to verify the validity of our predictions and update their extinction risk as appropriate. Prioritizing the reassessment of species will reduce assessors workload and deliver faster information for more rapid and effective conservation actions.
Streamlining the reassessment procedure by identifying reassessment priorities is currently only possible for most vertebrate species, for which a number of trait and spatial data are available. Other groups can, in principle, be assessed using citizen science data (Maes, et al. 2015, Zizka, et al. 2020) . However, these only allow for estimating species¿ extent of occurrence and the application of RL criterion B, an approach which is highly sensitive to spatial biases in sampling effort. As such, testing the strengths of alternative automated approaches can only be done comprehensively for groups that have received more research attention so far (typically terrestrial vertebrates). However, improving the cost-efficiency of reassessment for these groups can indirectly benefit lesser known groups in two ways. First, methodological improvements on the identification of reassessment priorities will benefit other groups as soon as data become available. Second, by freeing some of the resources currently invested in the reassessment of data-rich groups, it might be possible to increase funding availability for other groups.
As improved global knowledge of biodiversity decline unveils the magnitude of the current crisis, and the true pace of species loss, conservation action must build on a more comprehensive, representative, and rapidly updated monitoring of extinction risk. This work can set the basis of a technological revolution within the RL, that will take advantage of state-of-the-art knowledge and cutting-edge methodology to improve the efficiency, economic sustainability, and reduce existing taxonomic biases in the current RL reassessment process.