Completing the web: identifying sampling bias and knowledge gaps within South African spider surveys (Arachnida, Araneae)

Species distribution datasets are fundamental for macroecological studies, although there is an overarching need to ensure that these datasets are representative of the entire community. Shortfalls, or knowledge gaps, within biodiversity datasets originate for a range of reasons, and can lead to incorrect conclusions or recommendations being drawn. Spatial scale influences the interpretations of diversity patterns and thus is an important aspect to consider. South Africa has a rich history of spider sampling and as such, it is possible to investigate the influence that scale, both spatial and taxonomic, has on the overall interpretations of how complete the spider knowledge base is in the country. To do this, we draw on curated natural history spider collections and determine how complete the spider assemblages are across twelve unique combinations of taxonomic and spatial scales. Overall, we received 121 605 usable records from seven collections, with spider records and diversity, being concentrated along the eastern and coastal regions of South Africa. We show that assemblage completeness increases with both increasing taxonomic and spatial scales, and as such, knowledge of the distribution of spider families at the biome level is largely complete. Moreover, we show that our fine-scale knowledge of spider assemblages in South Africa is relatively poor, yet we do identify, even at fine scales, assemblages in South Africa that can be considered complete. We identify under-sampled regions of the country, which in turn are congruent with the distribution of under-sampled regions found in other South African invertebrate groups. We show that the scaling of completeness can only be interpreted in one direction: as scale increases so does completeness. These findings will have important implications for spider research and conservation in South Africa, given that regions where completeness is highest correspond strongly to areas in South Africa with the highest threats to biodiversity.