Landmark-based geometric morphometrics as a complementary tool for female identification of Euplatypus compositus and Euplatypus parallelus

Le résumé fourni par la source

Abstract. Chatpiyaphat K, Rizali A, Rahardjo BT, Wang J, Ye X, Sittichaya W, Johnson AJ, Hulcr J, Tarno H. 2026. Landmark-based geometric morphometrics as a complementary tool for female identification of Euplatypus compositus and Euplatypus parallelus. Biodiversitas 27 (4): d270439. https://doi.org/10.13057/biodiv/d270439. Identification of female Platypodinae ambrosia beetles is often difficult due to strong sexual dimorphism and limited diagnostic characters, particularly in Euplatypus compositus and Euplatypus parallelus, two closely related species in the southern United States with overlapping habitats and highly similar female morphology. Although the presence or absence of a pronotal mycangial pit is diagnostic, this structure can be obscured by specimen damage. We evaluated landmark-based geometric morphometrics (GM) as a complementary method for distinguishing females of the two species. A total of 46 specimens were analyzed for the pronotum (E. compositus n: 17; E. parallelus n: 29) and 43 for the elytra (E. compositus n: 16; E. parallelus n: 27). Multivariate regression revealed significant allometric scaling in pronotum shape (R²: 18.9%, p: 0.001), whereas elytral shape showed no detectable allometric effect (R²: 0.7%, p: 0.867). Cross-validated discriminant analysis (LOOCV in XYOM) indicated stronger species discrimination for the pronotum (84.8% accuracy; TAU: 67%) than for the elytra (76.7% accuracy; TAU: 50%), and repeated stratified k-fold cross-validation in R confirmed more robust and consistent discrimination for the pronotum but weaker and more variable performance for the elytra. Geometric morphometrics can detect subtle interspecific differences not apparent through visual inspection and may serve as a complementary identification approach when specimens are well preserved. However, this study demonstrates discriminatory potential under controlled conditions only. Because species identity is confounded with geographic origin and collection history, the observed shape differences, while consistent with species-level variation, cannot be attributed to species identity alone with full confidence. Validation using broader, more balanced sampling remains necessary.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.