The phylogenetic relationships of Morgan's Sphinx, Xanthopan morganii (Walker), the tribe Acherontiini, and allied long-tongued hawkmoths (Lepidoptera: Sphingidae, Sphinginae)

• Published in: Zoological journal of the Linnean Society 2002-08, Vol.135 (4), p.471-527
• Main Author: KITCHING, IAN J.
• Format: Article
• Language: English
• Subjects: Cocytius group; immature stages; implied weighting; morphology; Orchidaceae; phylogeny; pollination; Sphingidae; Sphinginae; successive approximations character weighting; total evidence; Xanthopan morganii
• ISSN: 0024-4082; 1096-3642
• DOI: 10.1046/j.1096-3642.2002.00021.x

A cladistic analysis is presented of the hawkmoths of the tribe Acherontiini, Morgan's Sphinx (Xanthopan morganii (Walker)), and related genera. The study aims to test the monophyly of tribe Acherontiini; the hypothesis that all taxa with extremely long probosces (some Acherontiini, Meganoton rubescens, Neococytius, Xanthopan) form a monophyletic group, or at least fall within a single reasonably compact clade; and, within this group, to determine whether Xanthopan is more closely related to Acherontiini or to Cocytius and Neococytius. The data set comprises 109 characters derived from adult and immature stage morphology, biology and behaviour. These data were analysed using equal weighting, successive approximations character weighting (SACW) and implied weighting. All weighting schemes agreed on the monophyly of Acherontiini and of a group of genera comprising Amphimoea, Cocytius and Neococytius (the Cocytius group). Several other generic and suprageneric clades were also consistently recovered. However, those hawkmoths with extremely long probosces were never recovered as a monophyletic group. The relationships of Xanthopan were also ambiguous. Equal weighting and SACW placed Xanthopan + Meganoton rubescens (Butler) as sister to the Cocytius group, while implied weighting placed Xanthopan as sister to Acherontiini. This latter relationship is based primarily on shared possession of a pilifer/palp hearing organ. Further analyses suggested the two components of this organ were not biologically independent. Downweighting this feature accordingly resulted in all weighting schemes converging on the topology found by equal weighting. Exclusion of the incomplete subset of immature stage data had no effect under implied weighting but equal weighting and SACW now recovered a Neotropical clade comprising Manduca and the Cocytius group, while Xanthopan was placed with M. rubescens and Panogena. Downweighting the pilifer/palp hearing organ under implied weighting again caused convergence with the equal weighting/SACW results. Thus, the relationships of Xanthopan remain equivocal and further data, particularly from the immature stages, will be required to elucidate its phylogenetic position further. © 2002 The Linnean Society of London. Zoological Journal of the Linnean Society, 2002, 135, 471–527.