Putting insects on the map: near-global variation in sphingid moth richness along spatial and environmental gradients

Despite their vast diversity and vital ecological role, insects are notoriously underrepresented in biogeography and conservation, and key broad-scale ecological hypotheses about them remain untested-largely due to generally incomplete and very coarse spatial distribution knowledge. Integrating records from publications, field work and natural history collections, we used a mixture of species distribution models and expert estimates to provide geographic distributions and emergent richness patterns for all ca 1000 sphingid moth species found outside the Americas in high spatial detail. Total sphingid moth richness, the first for a higher insect group to be documented at this scale, shows distinct maxima in the wet tropics of Africa and the Oriental with notable decay toward Australasia. Using multivariate models controlling for spatial autocorrelation, we found that primary productivity is the dominant environmental variable associated with moth richness, while temperature, contrary to our predictions, is an unexpectedly weak predictor. This is in stark contrast to the importance we identify for temperature as a niche variable of individual species. Despite divergent life histories, both main sub-groups of moths exhibit these relationships. Tribal-level deconstruction of richness and climatic niche patterns indicate idiosyncratic effects of biogeographic history for some of the less species-rich tribes, which in some cases exhibit distinct richness peaks away from the tropics. The study confirms, for a diverse insect group, overall richness associations of remarkable similarity to those documented for vertebrates and highlights the significant within-taxon structure that underpins emergent macroecological patterns. Results do not, however, meet predictions from vertebratederived hypotheses on how thermoregulation affects the strength of temperature-richness effects. Our study thus broadens the taxonomic focus in this data-deficient discourse. Our procedures of processing incomplete, scattered distribution data are a template for application to other taxa and regions.