An integrative approach for species delimitation in the highly diversified stingless bee genus Trigona (Apoidea: Meliponini)

Applying DNA-barcoding for species identification to Neotropical stingless bees is challenging, but its usefulness has not yet been fully evaluated. We examined the efficiency of COI-barcoding to identify species within the genus Trigona, currently subdivided into 9 species-groups. To do so, we considered 99 newly collected (and morphologically identified) specimens across 9 Peruvian Departments and 1286 COI Trigona sequences retrieved from BOLD. To examine species partitioning, two distance-based (ABGD, ASAP) and one tree-based (mPTP) species delimitation methods were applied on the whole dataset. We also performed an ecological niche analysis to support distinction among some recognized allied species. COI-barcoding was proven to be reliable for 7 (out of 21) species of Trigona, also well-distinguished morphologically. In all other instances, we detected a morphological vs. molecular ID mismatch, mostly affecting sibling species (e.g. T. chanchamayoensis vs. T. muzoensis). Species delimitation methods were also somewhat incongruent, with ABGD and ASAP oversplitting some clusters/species. This was particularly evident within the ‘fulviventris’ group of species, where the actual distinction between T. fulviventris and T. guianae is blurred by slight overlapping of some morphological traits and distribution. However, the environmental clustering analysis also showed a lack of ecological distinction between these two species. All the above mentioned issues could be related to the absence of a reliable barcoding gap due to recent speciation processes, as well as for the incompleteness of the existing dichotomous morphological keys/databases. A “reverse taxonomy” and quantitative approaches (e.g. morphometry, chemical and genomic analyses) combined with the optimization of the existing tools for taxonomic identification of Meliponini are needed to resolve such issues.