Molecular Phylogenetics of the Genus Ceratitis (Diptera: Tephritidae)
Total Page:16
File Type:pdf, Size:1020Kb
Molecular Phylogenetics and Evolution 38 (2006) 216–230 www.elsevier.com/locate/ympev Molecular phylogenetics of the genus Ceratitis (Diptera: Tephritidae) Norman B. Barr ¤, Bruce A. McPheron Department of Entomology, Pennsylvania State University, University Park, PA 16802, USA Received 29 March 2005; revised 3 October 2005; accepted 5 October 2005 Abstract The Afrotropical fruit Xy genus Ceratitis MacLeay is an economically important group that comprises over 89 species, subdivided into six subgenera. Cladistic analyses of morphological and host use characters have produced several phylogenetic hypotheses for the genus. Only monophyly of the subgenera Pardalaspis and Ceratitis (sensu stricto) and polyphyly of the subgenus Ceratalaspis are common to all of these phylogenies. In this study, the hypotheses developed from morphological and host use characters are tested using gene trees pro- duced from DNA sequence data of two mitochondrial genes (cytochrome oxidase I and NADH-dehydrogenase subunit 6) and a nuclear gene (period). Comparison of gene trees indicates the following relationships: the subgenus Pardalaspis is monophyletic, subsection A of the subgenus Pterandrus is monophyletic, the subgenus Pterandrus may be either paraphyletic or polyphyletic, the subgenus Ceratalaspis is polyphyletic, and the subgenus Ceratitis s. s. might not be monophyletic. In addition, the genera Ceratitis and Trirhithrum do not form reciprocally monophyletic clades in the gene trees. Although the data statistically reject monophyly for Trirhithrum under the Shimoda- ira–Hasegawa test, they do not reject monophyly of Ceratitis. 2005 Elsevier Inc. All rights reserved. Keywords: Ceratitis; Trirhithrum; Tephritidae; ND6; COI; period 1. Introduction cies, C. capitata (Wiedemann) (commonly known as the Mediterranean fruit Xy), is already an invasive species The genus Ceratitis MacLeay (Diptera: Tephritidae) with established populations throughout tropical, sub- comprises over 89 Afrotropical species of fruit Xy (De tropical, and mild temperate habitats worldwide (Vera Meyer, 2000a). Like other tephritids, Ceratitis biology is et al., 2002). associated with host plants that provide mating sites for Despite the economic importance of the genus, few adults and food for developing larvae. Although most studies have investigated the systematics of Ceratitis (De Ceratitis species feed on fruit, several have specialized on Meyer, 2000a, 2005). Ceratitis is one of nine genera in the the seedpods of monocotyledons or the cones of gymno- tribe Ceratitidini (subfamily Dacinae); although evidence sperms (De Meyer, 2000a, 2001a). The genus is also inter- supports monophyly of the tribe Ceratitidini and subfam- esting in that several of its species are extremely ily Dacinae (Han and McPheron, 2000; Korneyev, 2000; polyphagous, each feeding on hundreds of plant species Smith et al., 2002), systematic relationships among the from over 30 plant families. Due to their ecological depen- nine ceratitidine genera are not well understood. The dence on hosts and generalist (opportunist) feeding genus is subdivided into six subgenera: Ceratitis sensu behavior, many Ceratitis species are agricultural pests in stricto (8 spp.), Ceratalaspis Hancock (33 spp.), Pardala- Africa and considered potential invasive species. One spe- spis Bezzi (10 spp.), Pterandrus Bezzi (36 spp.), Hop- lolophomyia Bezzi (1 sp.), and Acropteromma Bezzi (1 sp.); the most recent taxonomic reviews of these subgenera are * Corresponding author. Present address: USDA APHIS PPQ CPHST Pest Detection Diagnostics and Management Laboratory, Edinburg, TX by De Meyer (1996, 1998, 2000b), and De Meyer and 78541-9398, USA. Copeland (2001). A taxonomic review of Pterandrus by E-mail address: [email protected] (N.B. Barr). De Meyer and Freidberg is in press. (Taxa that will be 1055-7903/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2005.10.013 N.B. Barr, B.A. McPheron / Molecular Phylogenetics and Evolution 38 (2006) 216–230 217 described formally therein, and included in our study, are 2. Materials and methods referred to by two letter codes: ‘ST,’ ‘CO,’ and ‘PE.’) De Meyer (2005) constructed Ceratitis phylogenies by 2.1. Taxon sampling cladistic analysis of morphological and host use charac- ters for 89 species. The results suggested that the subgen- A total of 48 species from the subfamily Dacinae were era Pardalaspis and Ceratitis s. s. are monophyletic, that included in this study. Thirty-two of these species were Cer- the subgenus Pterandrus might be monophyletic or para- atitis from the following subgenera: 4 Ceratitis s. s., 15 phyletic, that the subgenus Ceratalaspis is polyphyletic, Pterandrus, 3 Pardalaspis, 9 Ceratalaspis, and 1 Hop- that the monotypic subgenus Hoplolophomyia belongs in lolophomyia. Four additional ceratitidine genera were rep- Ceratalaspis, and that the monotypic subgenus resented: Capparimyia Bezzi (1 sp.), Carpophthoromyia Acropteromma may belong outside of the genus. Austen (1 sp.), Neoceratitis Hendel (1 sp.), and Trirhithrum To examine the hypotheses generated by the morpho- Bezzi (9 spp.). In addition, 2 ceratitidine species (collection logical analysis, we have constructed gene trees from 32 numbers 1195 and 2591) that have not yet been assigned to Ceratitis species representing Wve of the six subgenera. In a genus were included in the study. (Ceratitidini n sp. 1195 addition, species from four other ceratitidine genera are was reared from Lepidotrichilia volkensii and Ceratitidini n included to test the systematic relationships within the sp. 2591 was reared from Rubia cordifolia.) Two species tribe. The mitochondrial genome has many excellent (Dacus demmerezi (Bezzi) and Bactrocera oleae (Rossi)) characteristics that make it an ideal tool for the study of from the tribe Dacini were included as outgroups. Collec- insect phylogenetics (Simon et al., 1994) and a majority of tion information is listed in Tables 1 and 2 for each species tephritid molecular phylogenetic studies have used the used to construct phylogenies. With the exception of mitochondrial 12S ribosomal RNA and 16S ribosomal D. demmerezi (from Reunion), Neoceratitis cyanescens RNA genes (see Han and McPheron, 1997; Han et al., (Bezzi) (from Reunion), C. colae Silvestri (from Ghana), 2002; Han and Ro, 2005; McPheron and Han, 1997; and C. penicillata Bigot (from Nigeria) specimens in Table McPheron et al., 2000; Muraji and Nakahara, 2001; 1, and several C. anonae Graham (from Ghana), C. fasci- Smith et al., 2002, 2003). However, these markers are ventris (Bezzi) (from Mali), and C. rosa (from Reunion, some of the more slowly evolving mitochondrial genes, Malawi, and South Africa) specimens in Table 2, samples and studies using them often looked at taxonomic ques- were from Kenyan collections made by R. Copeland. tions deeper than the genus level (Han and McPheron, Voucher specimens of each collection were placed in the 2000; Han et al., 2002; Han and Ro, 2005). Protein Frost Entomological Museum (Pennsylvania State Univer- encoding genes such as cytochome oxidase I (COI) and sity) and the Royal Museum of Central Africa, Teruven, cytochrome oxidase II (COII) tend to evolve more quickly Belgium. To assess the eVect of intraspeciWc variation on the and have been useful for resolution of many taxonomic results, multiple (5–20) individuals from several Ceratitis problems (e.g., subfamily Dacinae and genus Bactrocera spp. were included for analysis of the ND6 gene (Table 2). Macquart; Smith et al., 2002, 2003). Although less fre- quently used, the NADH-dehydrogenase subunit genes 2.2. Laboratory methods (e.g., ND1, ND4L, and ND6) are typically more variable than CO genes (Simon et al., 1994). An additional advan- DNA was extracted from each individual Xy using either tage of protein coding genes over rRNA genes is that the total nucleic acid protocol of Han and McPheron sequence alignments are unambiguous (it is often diYcult (1997) or the DNeasy animal tissue protocol (Qiagen, to determine homology of nucleotide sites in ribosomal Valencia, CA). Extractions were performed on the head genes because of insertion and deletion mutations). and thorax of Xies using the Han and McPheron protocol; Therefore, assuming that many of the species in the Cera- wings and abdomens were removed prior to extraction and titis phylogeny are relatively young (estimated separation stored in 95% ethanol. Heads, heads plus thoraces, or entire of two million years between C. capitata and C. rosa Kar- Xies were used to extract DNA with the DNeasy method. sch (Torti et al., 1998)), two protein encoding genes were Tissue was incubated in tissue lysis buVer between three selected for analysis: COI and ND6. Because mitochon- and forty-eight hours and DNA was extracted following drial genes are linked, the inclusion of an independent the standard protocol for animal tissues. Samples were not marker in the study was highly desirable. Unfortunately, homogenized, and the extracted body parts (or body) were few primers capable of amplifying across Tephritidae or stored in ethanol with the rest of the body of the Xy. This Dacinae have been developed for nuclear genes that are non-destructive method was performed on both pinned and informative for relatively young taxa. One exception is alcohol preserved samples. the period gene where primers have been developed for its The polymerase chain reaction (PCR) was used to C3C5 region in Anastrepha Schiner (subfamily Trypenti- amplify two mitochondrial genes