Merodon aureus

Vujić, Ante, Zorić, Ljiljana Šašić, Ačanski, Jelena, Likov, Laura, Radenković, Snežana, Djan, Mihajla, Milić, Dubravka, Šebić, Anja, Ranković, Milica & Khaghaninia, Samad, 2020, Hide-and-seek with hoverflies: Merodon aureus - a species, a complex or a subgroup?, Zoological Journal of the Linnean Society 190, pp. 974-1001 : 979

Merodon aureus complex

We assessed wing-shape differences among males and females of M. aureus and the new species M. calidus and M. ortus . Stepwise discriminant analysis showed that all pairs of species differed significantly in wing shape ( Table 2). Additionally, our DA correctly classified species, with an overall classification success of 96% for males and 98% for females. Among the 136 male specimens, six were misclassified: two M. aureus as M. calidus ; three M. calidus specimens (as M. aureus ) and only one specimen of M. ortus (as M. calidus ). Among our female specimens, only one of 66 was misclassified ( M. aureus as M. calidus ). A similar outcome was obtained using the Gaussian naïve Bayes’ classifier (94% of male and 98% of female specimens were correctly classified). Importantly, in both classifications, the M. aureus lectotype was correctly classified with posterior probability of 99%. Among male specimens, successive CVA generated two highly significant axes (CV1: Wilks’ Lambda = 0.147; χ2 = 241.446; P <0.01; CV2: Wilks’ Lambda = 0.667; χ2 = 50.9595; P <0.01). CV1 represents most of the wing-shape variation (88%) and describes wing-shape differences among M. calidus and the other two species, whereas CV2 (with 12% variation) separates males of M. ortus from the other two species ( Fig. 4A). The most obvious wing-shape differences were among M. calidus and the other two species, whereas M. aureus and M. ortus exhibited the most similar wings ( Fig. 4B). Females of M. aureus and M. calidus were separated along one highly significant CV axis (CV1: Wilks’ Lambda = 0.2196; χ2 = 83.3834; P <0.01).