Sympatric lineage divergence in cryptic Neotropical sweat bees (Hymenoptera: Halictidae: <Emphasis Type="Italic">Lasioglossum</Emphasis>)

Given ongoing biodiversity decline, an important concern is that a large fraction of species diversity is not yet documented. Correct delimitation of species remains a challenge, especially for small and morphologically uniform groups such as sweat bees (Halictidae). Here, we applied an integrative taxonomic approach to study diversity within the Neotropical sweat bee subgenus Dialictus (genus Lasioglossum). We used four statistical methods to delimit species based on cytochrome oxidase subunit I gene sequences: Automatic Barcode Gap Discovery (ABGD), two variants of the General Mixed Yule Coalescent (single-threshold (stGMYC) and Bayesian (bGMYC)) and the Refined Single Linkage analysis (RESL). We detected eight principal molecular operational taxonomic units (mOTUs). Subsequently, these lineages were evaluated using ten nuclear microsatellite loci and morphological and ecological analyses. Most mOTUs could be differentiated using microsatellites and morphology (82 % identified correctly), further supporting the status of mOTUs as independent biological units. For the two most widespread mOTUs, we analysed intra-lineage geographic variation using microsatellites but did not detect additional substructuring. We further tested if the lineages showed predictable patterns of co-occurrence and habitat preferences. While we did not find any evidence of preferential association or disassociation between taxa, we detected a slight positive effect of high crop cover favouring the abundance of some lineages. We show that integrated approaches using statistical analysis of DNA barcodes jointly with additional data can provide robust and objective means of delimiting species in morphologically difficult groups.