DNA barcoding reveals genetic divergence and phylogenetic relationships of Scymnus species (Coleoptera: Coccinellidae) from Uzbekistan

Abstract

Ladybird beetles of the genus Scymnus (Coleoptera: Coccinellidae) are key predators in agroecosystems, but species-level identification remains challenging due to high morphological similarity among closely related taxa. Molecular data for Central Asian populations are scarce, limiting regional representation in global barcode libraries. Here, we analyzed mitochondrial cytochrome c oxidase subunit I (COI) sequences to investigate phylogenetic relationships and genetic divergence of four Scymnus species from Uzbekistan (S. rubromaculatus, S. frontalis, S. subvillosus, and S. apetzi). A total of 32 mitochondrial COI sequences (503 bp) were analyzed, comprising 31 sequences representing four species of Scymnus (16 newly generated from Uzbekistan and 15 recovered from GenBank) and one GenBank sequence of Axinoscymnus cardilobus used as the outgroup. Phylogenetic reconstruction using Maximum Likelihood and Bayesian Inference recovered distinct COI lineage clusters corresponding to the four initial morphology-based taxonomic assignments. The mean intraspecific K2P divergence was 0.00051 in S. rubromaculatus, 0.00362 in S. frontalis, 0.00811 in S. apetzi, and 0.03446 in S. subvillosus. The four newly generated S. subvillosus sequences from Uzbekistan were identical, whereas the elevated intraspecific divergence in the complete dataset was associated with publicly available records. The interspecific distances ranged from 0.11383 to 0.17778. Although the maximum intraspecific distance (0.09085) remained below the minimum interspecific distance (0.11383), the resulting separation should be regarded descriptive because the highly divergent public S. subvillosus records could not be independently taxonomically verified. These findings support the utility of COI to discriminate among the examined sequence groups, while highlighting the need for voucher-based morphological verification and broader multilocus sampling.