What do we know about the missing millions of Earths insect species and can we improve their collection: evidence from bark beetles?

Only 20% of the estimated five million species of insects on Earth are named despite over 240 years of taxonomy. Yet insects are poorly represented in protected area assessments, and insect declines are of concern globally. Here we explore how to increase the discovery of new species and understanding of this group through analysis of 10,097 tropical rainforest bark beetles (Scolytinae) from eight different ecological studies using beetles between 2000 and 2018 in the Australian Wet Tropics. Of the 107 species identified, 58 are undescribed: an increase of 35% on the 166 species known from Australia. As hypothesised, new species are significantly smaller, less abundant and less widespread than described species making them more extinction prone than named species. Rarefaction indicates doubling sampling would increase the number of species by 17. Flight Interception Traps (FIT) collected 84% of individuals and 98% of species confirming the effectiveness of a single sampling method for some beetles. Increased locations and collection from the canopy may sample further species rather than additional collecting methods. Scolytines are relatively well studied with a cadre of taxonomists at the forefront of using modern methods to resolve formerly intractable groups. These new species are more likely to be named than others in many other beetle groups where taxonomy has largely stalled. To increase species description rates and to avoid most species becoming extinct before being named, we call on taxonomists to use new character systems provided by DNA methods and to look at working with Artificial Intelligence tools. Significance StatementIn an era of rapid biodiversity loss, current conservation decisions for insects will continue to be based on a small and almost certainly biased sample of the worlds biota until more species are named. We demonstrate how large-scale sampling can dramatically increase the number of species discovered for one group of beetles and how these undescribed species are significantly smaller, less abundant and less widespread than named species. The identification and determination of undescribed species is rarely possible except when taxonomic expertise is available, as in the present study. Addressing the insect taxonomic bottleneck and increasing the rate of description will require the adoption of new and developing tools.