Imprecisely georeferenced specimen data provide unique information on species' distributions and environmental tolerances: Don't let the perfect be the enemy of the good

AimConservation assessments frequently use occurrence records to estimate species geographic distributions and environmental tolerances. Typically, records with imprecise geolocality information are discarded before analysis because they cannot be matched confidently to environmental conditions. However, removing records can artificially truncate species environmental and geographic distributions. Here we evaluate the trade-offs between using versus discarding imprecise records when estimating species ranges and climatic tolerances. LocationNorth America. Time period970-2019 and 2061-2080. MethodsWe collated records from 44 species in the genus Asclepias (milkweeds). Records were designated "precise" if they could be matched confidently to environmental data, and "imprecise" if not. We compared estimates of extent of occurrence (EOO), climatic niche breadth, and exposure to climate change using precise records only, as well as precise plus imprecise records together. To estimate EOO, we conservatively assigned imprecise records to points within their area of likely collection that were closest to the centroid of precise records. Similarly, to estimate climatic tolerances and exposure to climate change, we matched imprecise records to climate values that were most similar to the mean across precise records. ResultsAcross all species, including imprecise records increased EOO by 85% (median value; range across species: 0-2011%). Univariate niche breadth in mean annual temperature and precipitation increased by 25% (0-353%) and 28% (0-292%), respectively, while multivariate niche volume increased by 175% (8-13909%). Adding imprecise records increased suitable area in the present and area that remained suitable in the future. Main conclusionsImprecise records provide novel information about species distributions and climatic niche tolerances. While the default practice of discarding imprecise records ensures that only accurate data are used, it dramatically reduces estimates of range size and overestimates exposure to climate change. The benefits of discarding imprecisely geolocated records must be balanced against the loss of information incurred by their elimination.