Species discovery and dental ecometrics: good news, bad news and recommendations for the future

New mammal species have always been a source of major zoological intrigue, and whilst discoveries of large mammals are becoming less frequent, species like the saola and dingiso continue to enter the zoological literature. Modern communities are often assumed to be complete and are used for constructing community-level models. One example is ecometric modelling. Here, we estimate the number of large herbivorous mammals not yet described based on description curves. We also investigate the effect of species description on the stability and accuracy of previously established relationships between traits and climate. The contemporary description record of large herbivorous mammals is incomplete, with at least 83 species undescribed. Primates, artiodactyls and marsupials are estimated to contain the greatest undescribed diversity, with particular gaps in the Neotropic, Afrotropic and Indomalayan realms. We find that beyond similar to 40% completeness, there is a limited impact on trait-environment relationships of increased species description, but that there is a high mismatch between true and predicted climatic values for published models. Consequently, mammalian species that have not yet been discovered are unlikely to have a large impact on the accuracy of trait-environment models, but we suggest possible alterations to previous approaches that might improve the accuracy of future models. [GRAPHICS] .