RESUMO
Limited resources and taxonomic expertise in biodiversity surveys often lead to the application of the higher taxa approach (HTA),i.e., the identification of specimens to genus or higher taxonomic levels rather than to species. The reliability of the HTA varies significantly among studies, yet the factors underlying this variability have rarely been investigated. Bees are an ideal model taxon for testing the HTA because they are highly diverse, challenging to identify, and there is widespread interest in their role as native pollinators, driving demand for efficient diversity assessment tools. Using extensive bee data sets collected across three biomes and various habitats, we assessed the performance of the HTA in reflecting bee species richness and composition patterns at local scales, factors affecting this performance, and ways to improve it. The performance of the HTA varied considerably among biomes, taxonomic levels (genera and subfamilies), and diversity measures (species richness and composition); genus and subfamily richness accounted for 55-77% and 32-61% of the variation in species richness, respectively; genus and subfamily composition accounted for 28-87% and 26-80% of the variation in species composition, respectively. The number of species per higher taxon was a main factor influencing this performance (accounting for 63% of the variation), while the co-occurrence of taxonomically related species had no significant influence on the performance of the HTA. Further subdividing genera by body size contributed to the performance of the HTA and increased its accuracy in representation of compositional patterns by ~16%. Our results have several practical implications. The considerable variability found in the performance of the HTA in representing local-scale richness and composition patterns of bee species dictates caution in implementing this tool in bee surveys. When possible, an a priori evaluation of the expected performance of the HTA should be done, focusing on species distributions within higher taxonomic levels and the species: higher taxa ratio. Integrating morphological characteristics (such as body size) that consistently subdivide genera will improve the HTA's performance. Our results are likely applicable to the implementation of the HTA in other small-bodied and species-rich groups and may contribute to the cost-effectiveness of biodiversity surveys.
