Molecular Phylogenetics and Evolution 127 (2018) 600–605 Contents lists available at ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Phylogenetics of moth-like butterflies (Papilionoidea: Hedylidae) based on a new 13-locus target capture probe set T ⁎ Akito Y. Kawaharaa,b,c, , Jesse W. Breinholta,d, Marianne Espelanda,e, Caroline Storera, David Plotkina,b, Kelly M. Dextera, Emmanuel F.A. Toussainta, Ryan A. St Laurenta,c, Gunnar Brehmf, Sergio Vargasg, Dimitri Forerog, Naomi E. Pierceh, David J. Lohmani,j,k a Florida Museum of Natural History, University of Florida, Gainesville, FL 32611, USA b Entomology and Nematology Department, University of Florida, Gainesville, FL 32611, USA c Department of Biology, University of Florida, Gainesville, FL 32611, USA d RAPiD Genomics, 747 SW 2nd Avenue, IMB#14, Gainesville, FL 32601, USA e Arthropoda Department, Zoological Research Museum Alexander Koenig, Adenauer Allee 160, Bonn 53113, Germany f Institute of Zoology and Evolutionary Biology with Phyletic Museum, Friedrich-Schiller-University Jena, Jena 07743, Germany g Laboratorio de Entomología, Departamento de Biología, Pontificia Universidad Javeriana, Bogotá, Colombia h Department of Organismic and Evolutionary Biology and Museum of Comparative Zoology, Harvard University, Cambridge, MA 02138, USA i Biology Department, City College of New York, New York, NY 10031, USA j Ph.D. Program in Biology, Graduate Center, City University of New York, New York, NY 10016, USA k Entomology Section, National Museum of the Philippines, Manila 1000, Philippines ARTICLE INFO ABSTRACT Keywords: The Neotropical moth-like butterflies (Hedylidae) are perhaps the most unusual butterfly family. In addition to Anchored hybrid enrichment being species-poor, this family is predominantly nocturnal and has anti-bat ultrasound hearing organs. Exon capture Evolutionary relationships among the 36 described species are largely unexplored. A new, target capture, an- Lepidoptera systematics chored hybrid enrichment probe set (‘BUTTERFLY2.0’) was developed to infer relationships of hedylids and Papilionoidea some of their butterfly relatives. The probe set includes 13 genes that have historically been used in butterfly Phylogenomics phylogenetics. Our dataset comprised of up to 10,898 aligned base pairs from 22 hedylid species and 19 out- groups. Eleven of the thirteen loci were successfully captured from all samples, and the remaining loci were captured from ≥94% of samples. The inferred phylogeny was consistent with recent molecular studies by placing Hedylidae sister to Hesperiidae, and the tree had robust support for 80% of nodes. Our results are also consistent with morphological studies, with Macrosoma tipulata as the sister species to all remaining hedylids, followed by M. semiermis sister to the remaining species in the genus. We tested the hypothesis that nocturnality evolved once from diurnality in Hedylidae, and demonstrate that the ancestral condition was likely diurnal, with a shift to nocturnality early in the diversification of this family. The BUTTERFLY2.0 probe set includes standard butterfly phylogenetics markers, captures sequences from decades-old museum specimens, and is a cost-effective technique to infer phylogenetic relationships of the butterfly tree of life. 1. Introduction moth-like features, such as clubless antennae and nocturnal adult ac- tivity, which led many authors to believe that they were not butterflies Day-flying butterflies have fascinated researchers and enthusiasts but belonging to the family Geometridae (Prout, 1910). Despite the for centuries. The 36 described species of moth-like butterflies similarity of these two families, Geometridae possess a pair of bat ul- (Hedylidae) have received little attention, mainly because they are trasound-detecting tympanal organs on the abdomen, whereas hedylids predominantly nocturnal and restricted to Central and South America have hearing organs on their wings (Yack and Fullard, 2000). Hedylids (Scoble, 1986, 1990b; Kawahara et al., 2018). Most prior studies have also have larvae that have more pairs of prolegs than geometrids focused on their behavior (Yack and Fullard, 2000), neuroethology (Scoble, 1986). Due to many similarities that hedylids share with other (Yack et al., 2007a) or vision (Yack et al., 2007b), and the family still Lepidoptera, the phylogenetic position of Hedylidae in the Lepidoptera lacks a published genome (Triant et al., 2018). Hedylids have several remained largely uncertain. Recent molecular phylogenetic studies ⁎ Corresponding author at: Florida Museum of Natural History, University of Florida, Gainesville, FL 32611, USA. E-mail address: kawahara@flmnh.ufl.edu (A.Y. Kawahara). https://doi.org/10.1016/j.ympev.2018.06.002 Received 27 March 2018; Received in revised form 25 May 2018; Accepted 2 June 2018 Available online 11 June 2018 1055-7903/ © 2018 Elsevier Inc. All rights reserved. A.Y. Kawahara et al. Molecular Phylogenetics and Evolution 127 (2018) 600–605 have provided strong support for the Hedylidae belonging in the Pa- Table 1 pilionoidea (butterflies), as the sister-group to Hesperiidae (Heikkilä The 13 loci of the BUTTERFLY2.0 probe set, with sequence lengths (bp) and et al., 2012; Espeland et al., 2018; Breinholt et al., 2018; Kawahara and summary statistics for the probe and flanking regions of each gene. Breinholt, 2014). Locus no. Gene name Probe (P) Flank (F) P + F A limited number of systematic studies have been conducted on Hedylidae. Most are morphological classifications that are not based on L1 Cytochrome c oxidase I (COI) 658 415 1073 a phylogenetic analysis. One of the first major taxonomic studies was by L2 Thiolase 1024 43 1067 L3 Carbamoyl-phosphate synthetase 2, 1569 426 1995 Prout (1932), who used wing pattern to divide the group into the aspartate transcarbamylase, and genera Hedyle Guenée, Lasiopates Warren, Macrosoma Hübner, Phelli- dihydroorotase (CAD) nodes Guenée, and Venodes Guenée. Scoble (1986) evaluated Prout’s L4 Catalase (CAT) 1300 173 1473 genera and examined their morphology in greater detail, including that L5 Dopa decarboxylase (DDC) 579 150 729 L6 Elongation factor 1 alpha (EF1-a) 1049 508 1557 of the genitalia and immature stages. Although Scoble agreed that some L7 Glyceraldehyde 3-phosphate 599 392 991 hedylids could be placed into subgroups based on wing pattern simi- dehydrogenase (GAPDH) larity, he determined that all species are morphologically similar en- L8 Hairy cell leukemia protein 1 (HCL) 633 209 842 ough to justify treating them as a single genus, Macrosoma. He postu- L9 Isocitrate dehydrogenase (IDH) 708 298 1006 lated that species in the family constitute a monophyletic group, L10 Malate dehydrogenase (MDH) 718 268 986 L11 Ribosomal protein S2 (RPS2) 472 191 663 although he did not conduct a formal phylogenetic analysis. Lourido L12 Ribosomal protein S5 (RPS5) 554 221 775 (2011) conducted a cladistic analysis of Hedylidae, in which she coded L13 Wingless 239 361 600 111 morphological characters from the 36 recognized species. Her work Total 9827 3930 13,757 proposed a new classification to revalidate Hedyle Guenée and Phelli- Average 756 302 1058 nodes Guenée that, together with Macrosoma Hübner, would comprise three hedylid genera. However, relationships among many species were not resolved due to low branch support and unresolved polytomies, and alignments used as the template for probe design. Probes of 120 bp only three outgroups were included. Furthermore, the study was not were placed across each template region with 2x tiling. To avoid over- formally published, and therefore the proposed taxonomic changes are capture of the COI gene, we only designed two probes, placed 144 bp not valid under Article 9.12 of the Code (ICZN, 1999). Here, we test the apart, that appeared to vary the least across the alignment. We sum- monophyly of Hedylidae, examine relationships among its species, and marize our probe and flanking region capture in R, using ggplot2 test the hypothesis that nocturnality evolved once from diurnality in the (Wickham, 2009)(Fig. 1). family, using a molecular dataset that was built from a newly-devel- oped anchored hybrid enrichment probe set. This target capture kit fi includes 13 loci that were carefully chosen to overlap with most ex- 2.2. Taxon sampling and species identi cation isting Lepidoptera molecular datasets. We tested the capture success of this probe set on 22 Hedylidae species and outgroups sampled from six We sampled 22 of the 36 described hedylid species, and included butterfly families and one moth family. multiple specimens per species when possible. We also included 19 outgroups, representing all butterfly families and one moth, resulting in fi 2. Methods a nal dataset of 48 specimens representing 41 species. Hedylid species were identified using the morphological descriptions of Scoble (1990a, 2.1. Probe design: the BUTTERFLY2.0 anchored hybrid enrichment kit b) and Lévêque (2007), and by comparison with original type material. A complete catalogue with images of all analyzed species and their type fi The BUTTERFLY2.0 probe set captures 13 gene regions including specimens is provided in Supplementary File 1. Initial identi cations those most commonly included in studies on butterfly phylogenetics were cross-checked using the COI barcode sequences from this dataset (e.g., Wahlberg et al., 2016; Mitter et al. 2011; Wahlberg et al., 2005; with specimens from the curated collection of the McGuire Center for Wahlberg and Wheat, 2008; Simonsen