Ancient incomplete lineage sorting of Hyles and Rhodafra (Lepidoptera: Sphingidae)

Anna K Hundsdoerfer,Ian J Kitching

doi:10.1007/s13127-020-00445-0

Anna K Hundsdoerfer, Ian J Kitching

Open Access

https://doi.org/10.1007/s13127-020-00445-0

Copy DOI

Abstract

The hawkmoth genus Rhodafra comprises two African species with unclear relationships, as their wing patterns are markedly different, with one species closely resembling species of a related genus, Hyles. The present paper aims to investigate the monophyly and phylogenetic position of Rhodafra in relation to Hyles and other genera of the subtribe Choerocampina (Sphingidae: Macroglossinae: Macroglossini) using mitochondrial and nuclear sequence data from more species and individuals than have hitherto been studied. As no fresh tissue of Rhodafra was available, ancient-DNA methodology was applied. All data corroborate the genus as monophyletic and that a similar wing pattern is not a good indicator of close phylogenetic relationship in this group of moths. Phylogenetic trees based on mitochondrial data agree in placing Rhodafra within Hyles. In contrast, analysis of nuclear EF1alpha sequences produces a topology in which Rhodafra is placed as the sister clade to Hyles. Although multispecies coalescent analyses suggest a polytomy between Rhodafra, Hyles lineata and the remaining Hyles, total evidence analyses corroborate Rhodafra as sister to Hyles. This relationship is interpreted as the favoured topology. For a more robust result, the question should be re-examined using genomic approaches.

Highlights

The family Sphingidae comprises four subfamilies, Langiinae, Smerinthinae, Sphinginae and Macroglossinae (Kitching and Rougerie et al 2018)
Datasets were compiled for three different approaches and all three were executed in two mt variants (COI + t-RNA-Leu + COII, and COI barcode only; overview in Table S2): (1) comparative gene trees not necessarily with the same taxonomic coverage, followed by two approaches with identical choices of OTUs for both EF1a and mt gene fragments; (2) concatenated alignments for total evidence (TE) analyses; and (3) multispecies coalescent (MSC) analyses with input from single gene trees
By including the first DNA sequence data from R. opheltes, our study has provided the first molecular phylogenetic confirmation of the monophyly of the genus, as well as demonstrating the reciprocal monophyly of the two species of Rhodafra, despite their divergent wing patterns

Summary

Introduction

The family Sphingidae (hawkmoths) comprises four subfamilies, Langiinae, Smerinthinae, Sphinginae and Macroglossinae (Kitching and Rougerie et al 2018). On the basis of modifications of this structure, Rothschild and Jordan (1903) divided the genera. Rothschild and Jordan’s (1903) subdivision of the Choerocampina has been accepted for over a century without change, recent phylogenetic analyses using DNA sequence data Kawahara et al 2009) suggest that a different subdivision will be recognized once all the genera have been analysed together. One of the genera whose phylogenetic placement is currently unclear is Rhodafra. It comprises two species, Rhodafra marshalli Rothschild & Jordan 1903, occurring from Kenya to Zambia and Zimbabwe, and R. opheltes (Cramer, 1780) from South Africa, which differ markedly in wing pattern

Methods

Results

Conclusion