Revision of the genus Uenoites Belousov & Kabak, 2016 (Coleoptera: Carabidae: Trechini): A quantitative framework for defining taxonomic structure. 1. The gregoryi species group

The first species group of the genus Uenoites Belousov & Kabak, 2016 (Carabidae: Trechini) is defined and revised. The previously misunderstood geographic position of the type locality of Uenoites gregoryi (Jeannel, 1937) is corrected. This species is the oldest taxon now classified in the genus Uenoites. It is redescribed and figured based on newly collected material from the type locality. Further data on its distribution and bionomics are also provided. Two subspecies of U. gregoryi and four closely related species are described as new to science: U. gregoryi robustus ssp. nov., U. gregoryi gege ssp. nov., U. angulosus sp. nov., U. microps sp. nov., U. setipes sp. nov., and U. oculatus sp. nov. All these species occur in a rather small area between the valleys of the Lancang (Mekong) and Chang Jiang (Yangtze) rivers, NW of Lijiang City, Yunnan, China. All these species are assigned to a newly established species group named after its first known species, the gregoryi species group. All species are morphologically described and provided with photographs of the habitus and male genitalia. Their morphometric, sexual, and geographic variation, as well as allometric scaling, are assessed. Vector lollipop and transformation grid plots are provided to better illustrate differences in the the external morphology and male genitalia. An identification key for all members of the gregoryi species group is provided. The distribution of all taxa is mapped. The combined use of linear (traditional) and geometric morphometrics, together with a range of exploratory classification methods (principal component analysis, hierarchical clustering, and clustering on PCA scores) and pairwise comparisons of morphological distances, proved effective in detecting gaps between operational taxonomic units and determining their taxonomic status within the group. Comparison of the ratios with raw measurements within the linear morphometrics approach showed that, albeit statistically less robust, ratios carried more biologically meaningful signal than raw measurements, since the former are more shape-dependent, whereas the latter are more size-dependent. Preliminary PCA decomposition of raw measurements drastically improved the results of hierarchical clustering, yet they still remained more size-dependent. The Ward.D2 and average-linkage clustering methods proved to be the best when balanced against each other, particularly in the case of unevenly sampled taxa. An algorithm is proposed to identify the most stable clusters in taxonomically ambiguous groups as potential taxa for further testing. It uses a data frame of character states and is based on quality criteria derived from independent lines of evidence. Geographic populations of U. microps sp. nov. and U. gregoryi were studied using the Mantel test and distance-based redundancy analysis. The former species, despite being the most widely distributed, showed only slight geographic variation, comparable to sexual dimorphism and elevation effects, whereas the latter showed a much higher proportion of geographic variation, supporting its split into three subspecies. At the confirmatory stage, supervised discriminant analyses correctly classified practically all specimens into the established taxa, including subspecies, although the obtained models were based on traditional morphometric characters alone. Species of the gregoryi group are mostly riparian, like other members of the Kozlovites lineage, and are associated with running water in the forest and lower alpine zones at middle and high elevations. Only a few taxa were collected in wet habitats of the brush and forest litter.