Abstract
The surge of biodiversity data availability in recent decades has allowed researchers to ask questions on previously unthinkable scales, but knowledge gaps still remain. In this study, we aim to quantify potential gains to insect data on the Global Biodiversity Information Facility (GBIF) through further digitisation of natural history collections, assess to what degree this would fill biases in spatial and environmental record coverage, and deepen understanding of environmental bias with regard to climate rarity. Using mainland Afrotropical records for Catharsius Hope, 1837 (Coleoptera: Scarabaeidae), we compared inventory completeness of GBIF data to a dataset which combined these with records from a recent taxonomic revision. We analysed how this improved dataset reduced regional and environmental bias in the distribution of occurrence records using an approach that identifies well-surveyed spatial units of 100 × 100km as well as emerging techniques to classify rarity of climates. We found that the number of cells for which inventory completeness could be calculated, as well as coverage of climate types by 'well-sampled' cells, increased threefold when using the combined set compared to the GBIF set. Improvements to sampling in Central and Western Africa were particularly striking, and coverage of rare climates was similarly improved, as not a single well-sampled cell from the GBIF data alone occurred in the rarest climate types. These findings support existing literature that suggests data gaps on GBIF are still pervasive, especially for insects and in the tropics, and so, is not yet ready to serve as a standalone data source for all taxa. However, we show that natural history collections hold the necessary information to fill many of these gaps, and their further digitisation should be a priority.