Introduction
Understanding species distributions is fundamental for conservation and management. Accurate maps of species distributions enables assessment of conservation priorities worldwide, based on diversity, threat or endemism, facilitating the development of targeted conservation strategies. Thus, to perform such prioritizations or gap analyses, the spatial accuracy of species ranges is crucial. In the last two decades, IUCN and similar species range maps have become a global currency for the assessment of species distributions across scales and subsequent conservation funding (Jetz et al. 2012, Jenkins et al. 2013, Brooks et al. 2019). Though these maps have been created for varying purposes by different groups, they are now used for global biodiversity analyses, and little has been done to test whether they are truly fit for this purpose.
Species range modeling is highly-sensitive to the data and methods employed. For instance, even basic means of trimming species IUCN ranges using habitat or elevation changes the RedList status of many species via corresponding distributional changes (Ocampo-Peñuela et al. 2016, Li et al. 2017). The reliability of such analyses is crucial not only to apportion adequate and appropriate conservation efforts, but to understand even basic biodiversity patterns, as errors could highlight the wrong areas and species, and with limited resources available this translates to failures to conserve areas and ecosystems critically needing protection.
Unfortunately, accurate and reliable data are rare for most taxonomic groups due to intense sampling biases (Hughes et al. in review ). As a consequence, IUCN and similar data underlie countless studies, including those on distribution (Buckley & Jetz 2008, Holt et al. 2013, Rissler & Smith 2010), climate change (Akcakaya et al. 2006, Sandel et al. 2011, Sekerciuglu et al. 2012), and especially conservation (Jenkins et al. 2013, La Saout et al. 2013, Jetz et al. 2014, Pouzols et al. 2014, Jenkins et al. 2015, Meyer et al. 2015, Runge et al. 2015, Ocampo-Peñuela et al. 2016, Betts et al. 2017, Li et al. 2017, Moran & Kanemoto 2017, Brooks et al. 2019, Mason et al. 2020). These include analyses at varying resolutions, though 10km2 is a common resolution since these data are made available for download and analysis (https://biodiversitymapping.org), yet many of these use the maps at a higher resolution than has been shown appropriate Guidance (https://www.conservationtraining.org/enrol/index.php?id=156) states that distribution data should be utilized in developing species range polygons, yet the adherence to these guidelines is not apparent. It is dangerous to use questionable data for real-world conservation, and researchers should carefully assess the limits of the data before applying them to real-world issues.
Here, we test the widespread assumption that the most up-to-date expert range maps (ERMs) (for mammals, odonata, amphibians, birds, and reptiles) provide consistent and standardized species range estimates, exploring whether these range boundaries are associated with administrative divisions. We also assess the proportion of species distribution points for each species falling within their designated IUCN ranges. We discuss and showcase alternative means of mapping diversity patterns, and demonstrate the tradeoff between different approaches, and the limits of applications for each. Understanding biases and inconsistencies helps inform appropriate use and ensures data are fit for purpose, and to make better use of available data to effectively inform research and management decisions based on a clearer understanding of species and richness patterns.