Geometric morphometrics applied to the study of morphological variation in South American species of the genus Cormocephalus Newport, 1844 (Chilopoda, Scolopendromorpha, Scolopendridae), and description of a new species from Colombia

Some species of the genus Cormocephalus Newport, 1844 have been described based on a few specimens, which makes it difficult to quantify intra- and interspecific morphological variation. As a result, some diagnostic morphological characteristics may be ambiguous or have no real taxonomic value in species delimitation. In this context, geometric morphometrics emerges as a valuable tool capable of evaluating the variation in shape of a given taxon in a quantitative way, helping to clarify hypotheses about its morphological characteristics. In this study, geometric morphometrics was implemented to record shape variation in three external morphological structures (cephalic plate, forcipular coxosternite, and tergite 21) of three species: C. andinus (Kraepelin, 1903), C. brasiliensis Humbert & Saussure, 1870, and C. guildingii Newport, 1845, distributed in South America. The objective was to evaluate and describe the morphological differences between species within the genus. Additionally, ten out of the 13 South American species described within the genus were reviewed and used as comparative material in the description of a new species from Colombia. As a result, the geographic distribution of the examined species is expanded, and the list of species belonging to the guildingii-subgroup is updated. The following taxonomic changes are suggested: C. abundantis González-Sponga, 2000, and C. edithae González-Sponga, 2000, are proposed as junior synonyms of C. brasiliensis; C. amazonae (Chamberlin, 1914), as junior synonym of C. andinus; and C. facilis González-Sponga, 2000, C. glabratus González-Sponga, 2000, and C. maritme González-Sponga, 2000 as junior synonyms of C. guildingii. Regarding geometric morphometry, the shape of the three structures differed significantly among species: the analysis of canonical variates revealed low shape overlap. The cross-validation test demonstrated general discrimination rates of 81.25 % for the cephalic plate, 84.00 % for forcipular coxosternite, and 73.49 % for tergite 21. In terms of allometric percentage, only a low proportion of the shape variation can be attributed to size variation: 9.27 % for the cephalic plate, 6.18 % for forcipular coxosternite, and 3.89 % for tergite 21. The lower percentage rates found in C. guildingii may be associated with its broader geographic distribution compared to the other analyzed species, potentially resulting in greater morphological variation. Furthermore, the relatively smaller sample size may also have influenced these results. However, it can be concluded that morphometric analysis can provide enlightening insights and relatively robust results compared to other studies applied to similar taxa, highlighting the usefulness of morphometric analysis in the characterization of the group.