Zoologica Scripta Molecular and morphological phylogeny of Diplazontinae (Hymenoptera, Ichneumonidae) SERAINA KLOPFSTEIN,DONALD L. J. QUICKE,CHRISTIAN KROPF &HOLGER FRICK Submitted: 22 September 2010 Klopfstein, S., Quicke, D. L. J., Kropf, C. & Frick, H. (2011) Molecular and morphological Accepted: 5 May 2011 phylogeny of Diplazontinae (Hymenoptera, Ichneumonidae). — Zoologica Scripta, 40, doi:10.1111/j.1463-6409.2011.00481.x 379–402. Parasitoid wasps are among the most species rich and at the same time most understudied of all metazoan taxa. To understand their diversification and test hypotheses about their evolu- tion, we need robust phylogenetic hypotheses. Here, we reconstruct the phylogeny of the subfamily Diplazontinae using four genes and 66 morphological characters both in separate analyses and in a total evidence approach. The resulting phylogeny is highly resolved, with most clades supported by multiple independent data partitions. It contains three highly sup- ported genus groups, for which we suggest morphological and behavioural synapomorphies. The placement of some of the genera, especially Xestopelta Dasch, is unexpected, but also sup- ported by morphology. Most of the genera are retrieved as monophyletic, with the exception of the morphologically diverse genus Syrphoctonus Fo¨rster. We split this genus into three gen- era, including Fossatyloides gen. n., to restore the phylogeny–classification link. Conflict between the morphological and the molecular topology was mostly resolved in favour of the molecular partition in the total evidence approach. We discuss reasons for this finding, and suggest strategies for future taxon and character sampling in Diplazontinae. Corresponding author: Seraina Klopfstein, Naturhistoriska Riksmuseet, Enheten fo¨r Entomologi, Box 50007, 104 05 Stockholm, Sweden. E-mail: [email protected] Seraina Klopfstein, Christian Kropf and Holger Frick, Natural History Museum (Invertebrates), Bernastrasse 15, CH-3005 Bern, Switzerland. E-mails: [email protected], holger.frick@ gmx.li Seraina Klopfstein and Holger Frick, Division of Community Ecology, Institute of Ecology and Evolu- tion, University of Bern, Baltzerstr. 7, 3012 Bern, Switzerland Donald L. J. Quicke, Division of Biology, Imperial College London, Silwood Park Campus, Ascot, Berkshire SL5 7PY, UK; Department of Entomology, Natural History Museum, London SW7 5BD, UK. E-mail: [email protected] Introduction ited attention, as is the case in most parasitoid groups The subfamily Diplazontinae represents one of the smaller, (Quicke 1997; Jones et al. 2009). The only fauna for which morphologically more homogeneous subfamilies of the spe- Diplazontinae have been thoroughly revised is the Nearctic cies-rich Ichneumonidae. As far as we are aware, these par- (Dasch 1964), and this morphological revision did not asitoids are associated exclusively with hoverfly hosts include any phylogenetic analysis, but only some informal (Diptera, Syrphidae), with a strong bias towards the aphido- comments about relationships between genera. Intra-gen- phagous species (Kamal 1926; Dusek et al. 1979; Rotheray eric relationships were also proposed by Fitton & Rotheray 1981, 1984, 1990; Fitton & Rotheray 1982; Ngamo Tinkeu (1982), but these authors usually did not present synapo- & Hance 1997). Their association with a closely related morphies for the suggested groupings. The only previous group of hosts makes them an ideal group to study the evo- phylogenetic analysis of the subfamily with reasonable gen- lution of host relations and specialization (Page & Charles- eric coverage included 42 species and was based on two ton 1998). Moreover, a peculiar mode of antennal courtship molecular markers (Klopfstein et al. 2010a). It found three has recently been reported from a species of this group highly supported genus groups which were informally called (Steiner et al. 2010). To study the evolution of such ecologi- Sussaba, Syrphoctonus and Diplazon genus groups. The place- cal or behavioural characters, robust phylogenetic hypothe- ment of some genera in these groups was unexpected, for ses of the group are needed. In Diplazontinae, both example, the highly derived Bioblapsis in the Syrphoctonus taxonomy and phylogeny have, to date, only received lim- genus group and Campocraspedon in the Diplazon genus ª 2011 The Authors d Zoologica Scripta ª 2011 The Norwegian Academy of Science and Letters, 40, 4, July 2011, pp 379–402 379 Phylogeny of Diplazontinae d S. Klopfstein et al. group. Two genera, Syrphoctonus Fo¨rster and Tymmophorus the related Ichneumoniformes; the latter was used as the Schmiedeknecht, were recovered as paraphyletic, with vary- functional outgroup for the whole dataset (Appendix 1). ing support. We here extend this study by including 66 morphological characters and two additional molecular Morphological characters for phylogenetic reconstruction markers, the mitochondrial NADH1 and the F2 copy of the As there is currently no morphological phylogeny of the nuclear elongation factor 1a (EF1a). The parallel use of group available, we newly scored 66 characters for most of molecular and morphological data allows us to obtain and the species included in the molecular analysis (Appendix 2). test morphological synapomorphies of higher level taxa and This number might seem low compared with the number provides a basis for exploring congruence and conflict of species analysed, but reflects the fact that we used a among these different sources of data (Patterson et al. 1993; dense taxon sampling at the species level, and that the Carpenter & Wheeler 1999; Wahlberg et al. 2005). Using Diplazontinae are comparatively uniform in terms of dis- congruence with the independently derived molecular data- crete morphological differences. Many closely related spe- set as a measure of performance (Miyamoto & Fitch 1995; cies only show small morphological differences, which Cunningham 1997), we identify the preferred weighting often involve only colour or sculpture characters, which strategy and strength for the morphological data partition amount to 11 and 12 of the 66 characters, respectively. To when analysed under the parsimony criterion. The extended reduce any bias that could stem from excluding characters taxon sampling now including 70 species and two additional that are considered unreliable at higher taxonomic levels, genera allows us to test the previous phylogenetic hypothe- we aimed to include every character that was parsimony sis and to assess the status of the unsupported genera. Addi- informative and different between any two species and that tionally, we include two genera which in the past have been could be scored unambiguously. The matrix contains both difficult to associate with other genera, that is, Xestopelta new characters (characters 3, 6, 7, 8, 32, 38, 44, 45, 47, Dasch and Schachticraspedon Diller. 52, 53, 54, 55, 56, 61, 62; see Appendix 2) and characters mentioned in taxonomic revisions or keys of the group Materials and methods (Beirne 1941; Dasch 1964; Townes 1971; Fitton & Roth- Taxon sampling eray 1982). Some character states are illustrated in Fig. 1. We included 85 individuals of 70 Diplazontinae species (73 The ninth sternite and ninth and tenth tergites of the male species for the molecular dataset) in our study, covering 13 have been studied in Diplazontinae by Beirne (1941) and of the 20 described genera (Appendix 1). Only species for Dasch (1964), who both mainly used them for species which we could obtain sequences from all four molecular delimitation. This character complex proved to be of great markers were included in the analyses, although for six spe- value for defining genus groups in the present study. The cies [Promethes sulcator (Gravenhorst), Sussaba dorsalis different states of this character complex are shown in (Holmgren), Sussaba flavipes (Lucas), Sussaba pulchella Fig. 2. Descriptions of characters, their states and consis- (Holmgren), Syrphoctonus fissorius (Gravenhorst) and Wold- tency index (CI) and retention index (RI) on the total evi- stedtius flavolineatus (Gravenhorst)], different specimens had dence topology are shown in Appendix 2, and the to be used to obtain sequences from the different genes. character matrix is provided as Supporting information. To provide as much information as possible, for some Unless mentioned otherwise, the characters were scored undescribed taxa, we use names that are not yet available; based on female specimens. Morphological terminology these are disclaimed from nomenclatural purposes accord- follows Townes (1969). ing to article 8.3 of the International Code of Zoological Nomenclature (1999). These new names are Diplazon wy- Molecular methods manni, Diplazon zetteli, Homotropus lucidus, Schachticraspedon Genomic DNA was extracted from whole specimens pre- kropfi, Sussaba roberti and Woldstedtius bauri. served in 80% ethanol and from dried specimens using the Monophyly of the Diplazontinae is supported by a num- Promega Wizard (Promega UK, Southampton, United ber of adult (Beirne 1941; Dasch 1964; Townes 1971; Wahl Kingdom) kit for blood and tissue extractions. Vouchers 1990; Wahl & Gauld 1998), larval (Wahl 1990; Wahl & and DNA samples are kept at the Natural History Gauld 1998) and biological characters (e.g. Dasch 1964; Museum in Berne, at the Swedish Malaise Trap Project Fitton & Rotheray 1982; Yu et al. 2005) and in molecular and in the Insect Museum of the University of Wyoming studies (Quicke et al. 2009). They form part of the informal (Appendix