Neuroptera: Osmylidae)

Org Divers Evol (2016) 16:225–232 DOI 10.1007/s13127-015-0242-6

ORIGINAL ARTICLE

The steps towards an inconspicuous vein fusion documented in Stenosmylinae forewings (Neuroptera: Osmylidae)

Guillaume Cousin1 & Olivier Béthoux1

Received: 16 July 2015 /Accepted: 22 October 2015 /Published online: 7 November 2015 # Gesellschaft für Biologische Systematik 2015

Abstract Based on respective numbers of branches in fore- identify homologous veins across the Insecta were essentially and hind wing, and on the morphology of fossil species, it has based on extant forms (Redtenbacher 1886), but fossil insects been assumed that a fusion of the posterior branch of the were also considered soon after (Brongniart 1893). Despite posterior media (MP2) with the anterior cubitus (CuA) was abundant accounts since that time and continuous improve- acquired convergently in fore- and hind wings in several fam- ment in our knowledge of fossil and extant forms, topographic ilies of Neuroptera (Insecta), including the Osmylidae. The homology conjectures (THCs) on wing venation are still de- corresponding topographic homology conjectures (THCs) bated, not only across orders but at finer levels as well. Indeed, had to assume inconspicuous vein fusion, but without inter- wing venation evolved considerably in the course of insect mediate condition being reported. Based on inter-specific, history, as documented by several examples on the complexity intra-specific and intra-individual variations observed in fore- reached in groups such as dragon- and damselflies (Bechly wings of a selection of Stenosmylinae species (Osmylidae), 1996) and crickets and katydids (Béthoux 2012a), among we herein document a complete transformation series ranging others. from the condition ‘MP2 and CuA approximating’ to ‘MP2 Taking into consideration both extant and fossil species is a and CuA fully fused, without distinct origin of MP2’,with well-grounded approach to address THCs, as transformations intermediate conditions in which the origin of MP2 has a are more likely to be rightly conjectured if few intermediate cross-vein-like appearance. The relevance of series of conspe- forms are missing (Remane 1952). Investigating intra- cific specimens to investigate problematic THCs is individual variation is a complementary option which has emphasized. been rarely exploited. Recent accounts have shown the impor- tant potential this approach has: the possibility for a vein to Keywords Insecta . Conjectures of topographic homology . translocate onto a neighbouring one was first evidenced, Wing venation . Intra-specific variability thanks to data partly derived from series of conspecific individuals (praying mantises; Béthoux and Wieland 2009); rare variants of some particular extant crickets proved decisive for favouring one THC over another and for identifying the fossil Introduction relatives (Béthoux 2012b); and differences exhibited by forewings of a single fossil specimen allowed the solving of a Wing venation is a character system which has received con- long-lasting issue of the forewing venation of cockroaches siderable attention from entomologists. Early attempts to (Guo et al. 2013). The large Neuroptera group is not free of debate on wing * Olivier Béthoux venation THCs. Recently, Shi et al. (2012)haveproposedthat [email protected] a fusion of the posterior branch of the posterior media (MP2) with the anterior cubitus (CuA) occurs in both wing pairs of Myrmeleontidae and Ascalaphidae (antlions, owlflies and 1 Centre de Recherches sur la Paléobiodiversité et les Paléoenvironnements (CR2P, UMR 7207), Sorbonne Universités - their kin; possibly in Nemopteridae—spoonwings; see MNHN, CNRS, UPMC-Paris6, Paris, France Aspöck et al. 1980; Shi et al. 2015). These authors also argued 226 G. Cousin, O. Béthoux that the same fusion occurs in hind wings of other taxa using Adobe Illustrator CS6 (Adobe Systems, San Jose, CA, (Osmylidae, Nymphidae), in which the fusion (if assumed USA). Photographs were taken using a digital camera Canon present) cannot be traced due to an inconspicuous origin of EOS 5D Mark III, coupled to a Canon 50-mm macro lens or to MP2 and pectinate fusion of CuA with MP2 (i.e. the succes- a Canon MP-E 65-mm macro lens (all Canon, Tokyo, Japan), sive, pectinate, emergence of branches of a vein fused with both equipped with polarizing filters. The resulting photo- another, the former lacking a distinct main stem). Shi et al. graphs were optimized using Adobe Photoshop CS6 (Adobe (2012) proposed translocation as the possible process by Systems, San Jose, CA, USA). which these fusions took place (if so). The wing venation used for reference for THC and the In the following, based on three species of Stenosmylinae corresponding colour coding follow Shi et al. (2012). (Osmylidae) in which forewing intra-specific variability was Measurements of the MP2+CuA connection (or lack thereof) documented, we provide a clear-cut case of a MP2+CuA fu- were carried out using photographs of wings mounted on sion taking place and becoming inconspicuous. The surveyed slides and, for unprepared material, under a stereomicroscope species are Oedosmylus brevis New 1989, Stenolysmus equipped with a graticule. A positive figure represents the extraneus (Walker 1853)andStenosmylus stenopterus occurrence of a fusion (then the figure accounts for its length), MacLachlan 1867, in addition to few specimens undeter- while a negative figure represents the lack of the fusion (then mined at the species level. the figure accounts for the minimum width of the area between the corresponding veins). These measurements are pro- vided using the following template: ‘min/average/max; stan- Material and methods dard deviation’. All wings preserving this area were measured. Wing length was measured based on mounted specimens Among specimens composing our survey sample, we selected only. Numbers of branches were estimated based on all avail- those whose wings were in pristine condition for being pre- able specimens. Some wings were damaged in particular areas pared, as follows: wings were cut off and mounted in white and therefore could not be used for the corresponding, partic- Euparal (Asco Laboratories, Manchester, UK); left wings ular measurements. These involve only a few cases and there- were mounted on their dorsal side, and right wings on the fore are not indicated. ventral side. As far as possible, wings were mounted unfolded. Specimens referred to as ‘QM’ and ‘AM’ are housed at the Queensland Museum (Brisbane, Australia) and the Australian Comparative analysis Museum (Sydney, Australia), respectively. Unprepared specimens from QM and AM were manipulated in ethanol, and All species we surveyed share the following traits: cross-veins pinned and dried, respectively. We surveyed eight specimens in the area between the anterior wing margin and ScP simple of O. brevis [all QM specimens; four prepared; five unpre- (with rare exception); a single cross-vein in the area between pared, three of them with partly damaged forewings; all males; ScP and RA, near the wing base; ScP and RA fuse in the last all from the same collecting event: Echo Pt. Lamington, QLD; fifth of the wing; RP posteriorly pectinate, with ‘main’ branches January–March 1992; G. Monteith (interception trap)], five further branched; and numbers of branches of MA, MP1, MP2, specimens of Stenol. extraneus [all QM specimens; three pre- CuA and CuP (as herein conjectured) similar in both wing pairs. pared; two unprepared, both with partly damaged forewings; including two males, three females; all from the same Oedosmylus brevis collecting event: Hornsby Heights, NSW; 17.iv.1979; K.J. Lambkin (at light)] and four individuals of Stenos. stenopterus Wing pairs subequal; fore- and hind wing about 20.5 and (all AM specimens; one prepared, which collecting data are 17.9 mm long, respectively; apex tapered; RP with about 11 main Timor Rock, Warrumbungle Range, NSW, mv lamp./26 branches, 30 terminal branches. Forewing (Fig. 1a,b,d–g): MA March 1971/D. K. McAlpine). A specimen of an undeter- fused with R short before the arising of RP+MA or opposite to it; mined Stenosmylinae species and four specimens of RP and MA diverging less than 1.0 mm distal to the fusion; MP1 Stenosmylus spp. were also surveyed (all AM; collecting data andMP2divergingatwingmid-length;MPandMP1aligned, of the former: Tubrabucca Barrington tops, NSW, 8 January 1956/R. Dobson; for the latter, various collecting data). Draft drawings were produced with the aid of a microscope equipped with a camera lucida (Zeiss SteREO Discovery V8 Fig. 1 Wing venation in Oedosmylus brevis New, 1989. a, b, d–g stereomicroscope equipped with a pair of W-PL ×10/23 eye Forewings. c Hind wing. a–c Wing overviews. d–g Details of the base pieces, a Plan Apo S ×1.0 FWD objective; all Zeiss, Jena, of MP2 with respect to CuA (asterisk indicates the base of MP2; dashes delimit the CuA area). a, f Specimen QM IWC OB 5, left forewing. b, d, e Germany). Transmitted light was obtained from a VisiLED Specimen QM IWC OB 4, left forewing. g Specimen QM IWC OB 3, left ACT basis (Schott, Stafford, UK). Drawings were finalized forewing Inconspicuous vein fusion in Stenosmylinae (Neuroptera) 227 228 G. Cousin, O. Béthoux

MP2 diverging obliquely; MP2 distinct from, shortly connected Fig. 2 Wing venation in Stenolysmus extraneus (Walker, 1853). a, b, d– with (average condition) or fused for a short distance with CuA g Forewings. c Hind wing. a–c Wing overviews. d–g Details of the base (−1.4 mm/−0.1 mm/0.8 mm; 0.5 mm); one forewing has the first of MP2 with respect to CuA (asterisk indicates the base of MP2; dashes delimit the CuA area). a, f Specimen QM IWC OB 8, left forewing. b, d, e fork of CuA, the point of connection of MP2 with CuA and the Specimen QM IWC OB 7, right forewing. g Specimen QM IWC OB 6, point of divergence of MP2 from MP2+CuA all superimposed left forewing (i.e. this organization is characterized by a point where two stems merge and three stems diverge); MP2 and CuA first branched distal to their divergence point; MA forked very distally, MP1 RP with distal branches further branched (about 11 main with an average of 4.5 branches (ranging from 4 to 5; posteriorly branches, 20 terminal branches). Forewing (Fig. 4a, c): MA pectinate), MP2 with an average of 6.9 branches (ranging from 6 fused with R short before the arising of RP+MA or opposite to 9; overall posteriorly pectinate), CuA with an average of 4.7 to it; RP and MA diverging almost immediately after the fu- branches (ranging from 3 to 6; overall anteriorly pectinate), CuP sion; no clear origin of MP2 as herein identified (see with an average of 16.9 branches (ranging from 14 to 19; poste- BDiscussion^); MP2 and CuA posteriorly pectinate; MA riorly pectinate). Hind wing (Fig. 1c): no clear origin of MP2 as forked very distally (2 or 3 branches), MP1 with 7 branches, herein identified (see BDiscussion^). MP2 with 3–4branches,CuAwith2–3 branches, CuP with 18 Stenolysmus extraneus branches (posteriorly pectinate). Hind wing (Fig. 4d): no clear origin of MP2 as herein identified (see BDiscussion^). Both wing pairs subequal; fore- and hind wing about 22.2 and Stenosmylus 20.1 mm long, respectively, oval, slightly pointed apex; RP sp. with about 10–12 main branches, with 27 to 33 terminal branches; nearly all vein branches between MA and CuA Forewing: no clear origin of MP2 as herein identified (see B ^ (included) terminated by a fork near wing margin. Forewing Discussion ); MP2 and CuA posteriorly pectinate except (Fig. 2a, b, d–g): MA fused with R short before the arising of for one specimen (implying a pectinate fusion of CuA with RP+MA or opposite to it; RP and MA diverging less than MP2, as compared with the corresponding number of 2.0 mm distal to the fusion; MP1 and MP2 diverging at wing branches in hind wing). Hind wing: no clear origin of MP2 B ^ mid-length; MP and MP1 aligned; MP2 diverging obliquely, as herein identified (see Discussion ). with an orientation distinctive from that of cross-veins; MP2 distinct from (Fig. 2a, f), or fused with CuA (Fig. 2b, d, e, g) for a long distance (−0.24 mm/1.4 mm/2.7 mm; 0.9 mm); Discussion MP2 and CuA first branched distal to their divergence point (i.e. no pectinate fusion of CuA with MP2); MA forked very Among the Osmylidae, the Kempyninae and Spilosmylinae distally, MP1 with an average of 4.2 branches (ranging from 4 have the MP fork located in a basal position, with MP2 dis- to 5; dichotomously branched), MP2 with an average of 7.8 tinct from CuA (New 1983a, 1988; among others). This is also branches (ranging from 6 to 10; overall posteriorly pectinate), the condition exhibited by fossil species currently assigned to CuA with an average of 4.4 branches (ranging from 4 to 5; various subfamilies within Osmylidae (Carpenter 1943;Ren dichotomously branched), CuP with an average of 18.3 and Yin 2003;Wangetal.2010;Makarkinetal.2014;among branches (ranging from 13 to 22; posteriorly pectinate). others). Members of the Porisminae and Eidoporisminae have Hind wing: no clear origin of MP2 as herein identified (see a fork of MP located distally, as in the species scrutinized BDiscussion^). herein, but there is no evidence of a connection of MP2 with CuA (Brongniart 1893;New1983b; etc.). Therefore, the con- Stenosmylinae gen. and sp. undet. dition ‘MP2 and CuA distinct’ is likely plesiomorphic for Stenosmylinae. Specimen composed of head, prothorax and forewing at- Based on respective numbers of branches in fore- and hind tached to it (Fig. 3a); forewings about 22 mm long, 7 mm wings, Shi et al. (2012) have hypothesized a fusion of MP2+ wide; MP2 distinct from CuA in the right forewing CuA in hind wings of various Osmylidae (among other cases (Fig. 3c), fused with it for 3.7 mm in the left forewing within Neuroptera). The corresponding THCs had to assume a (Fig. 3b). fully inconspicuous vein fusion of MP2 with CuA. Our survey of Stenosmylinae species provides direct evidence of such a Stenosmylus stenopterus fusion taking place and becoming inconspicuous, in forewings. We show that a fusion of MP2 with CuA occasionally Both wing pairs subequal; fore- and hind wing 19.1 and occurs in forewings of O. brevis,inwhichMP2andCuAcan 16.9 cm long, respectively; oval, with a slightly pointed apex; easily be distinguished (Fig. 1). With few exceptions, this Inconspicuous vein fusion in Stenosmylinae (Neuroptera) 229 230 G. Cousin, O. Béthoux

Fig. 3 Wing venation in an undetermined Stenosmylinae specimen (AM K.300571). a Habitus. b, c Details of the bases of MP2 with respect to CuA, as located in a. b Left forewing, reversed. c Right forewing

fusion is the rule in Stenol. extraneus, in which it is long, with (Fig. 2a)tomoderatelylong(Fig.2b)inStenol. extraneus. an origin of MP2 identifiable from neighbouring cross-veins Such a transition, resulting from a displacement of the MP1/ thanks to its obliquity (Fig. 2). The occurrence of the fusion is MP2 fork towards the wing base, is not as abrupt as it might also supported by intra-individual variations observed in an look. Firstly, the two conditions ‘MP2+CuA fusion absent’ undetermined Stenosmylinae specimen, in which one fore- and ‘MP2+CuA fusion long’ were observed in a single wing has distinct MP2 and CuA stems, the other a long individual (Fig. 3). Nearly the same ‘amount’ of displace- MP2+CuA fusion (Fig. 3). Finally, Stenos. stenopterus fore- ment of the MP1/MP2 fork would produce a full fusion wings exhibit no origin of MP2 we could distinguish (Fig. 4). of MP2 with CuA. Secondly, the transition from ‘MP2+ Yet, we argue that this species possesses the MP2+CuA fu- CuA fusion long’ to ‘MP2+CuA fusion full’ could have sion (concurring with New 1986) because of (A) the assumed ultimately resulted from a translocation, a transformation close phylogenetic affinities of the species with O. brevis and type whose occurrence is now well documented in insects Stenol. extraneus (in particular the latter; Kimmins 1940)and (Béthoux 2007, 2012b; Béthoux and Wieland 2009;Guo because of (B) the overall wing venation similarities between et al. 2013; Cui et al. 2015). By definition, this transfor- Stenol. extraneus and Stenosmylus spp., viz. (1) the position mation alters the length of the corresponding stems. For of the first fork of MP1, (2) the position of the first fork of example, in the specimen illustrated in fig. 17 in Béthoux MP2+CuA (for Stenos. stenopterus, as assumed in Fig. 4) (2012b), the right forewing as a short CuPaα (fig. 17A, and (3) the end of CuP. Given these aspects, assuming a G) while it is translocated onto M+CuA in the left fore- full fusion of MP2 with CuA is the only option to fit MP2 wing (fig. 17B, H). As a consequence, it is significantly and CuA between MP1 and CuP in forewings of Stenos. longer (as it originates at the wing base), and CuPa has stenopterus. It must be noted here that venation patterns at no distinct stem. the forewing base are identical in Stenol. extraneus and The transformation series (sensu Hennig 1966;i.e.an Stenos. stenopterus. As a consequence, there is no basal ordered series of character states of phylogenetic rele- ‘coalescence’, or lack thereof, that could be advocated to vance—see Grant and Kluge 2004) we document shows discard our hypothesis. that an inconspicuous fusion of MP2 with CuA was pro- Under this THC MP2 in Stenos. stenopterus is very long, gressively acquired in forewings of Stenosmylinae. The compared to that in Stenol. extraneus. Concomitantly, MP has resulting pattern bears striking similarities with that of no distinct stem in Stenos. stenopterus, while it is long hind wings (compare Fig. 4a, b). It suggests that the Inconspicuous vein fusion in Stenosmylinae (Neuroptera) 231

Fig. 4 Wing venation in Stenosmylus stenopterus MacLachlan, 1867 (AM K.84656). a, c Left forewing. b Left hind wing. a, b Wing overviews. c Details of the area between MP1 and CuP, without origin of MP2 distinct from cross-veins same steps could have produced the latter. Therefore, our Shi et al. (2012) had also to assume the occurrence of a observations provide independent support to the THCs pectinate fusion. This assumption also is supported by the case proposed by Shi et al. (2012) for hind wings of of Stenos. stenopterus, in which terminal branches of MP2 Osmylidae. The case we present indeed is an additional and CuA can be distinguished in hind wings (Fig. 4b) but case of convergent acquisition of the MP2+CuA fusion not in forewings (Fig. 4a), in which CuA emits successive among Neuroptera. branches from MP2+CuA. 232 G. Cousin, O. Béthoux

Conclusion species, and data on postabdominal morphology. Arthropod Structure & Development. Grant, T., & Kluge, A. G. (2004). Transformation series as an ideographic Consideration given to inter-specific, intra-specific and intra- character concept. Cladistics, 20(1), 23–31. individual variations allowed us to independently support Guo, Y., Béthoux, O., Gu, J., & Ren, D. (2013). Wing venation homol- THCs proposed by Shi et al. (2012). The transformation series ogies in Pennsylvanian ‘cockroachoids’ (Insecta) clarified thanks to herein documented in Stenosmylinae is a clear-cut case of a a remarkable specimen from the Pennsylvanian of Ningxia (China). – – vein fusion taking place and becoming fully inconspicuous. Journal of Systematic Palaeontology, 11(1 2), 41 46. Hennig, W. (1966). Phylogenetic systematics. Place published: Such cases have been assumed in various groups, such as in University of Illinois Press. Caelifera (grasshoppers), in which CuA is assumed to be in- Kimmins, D. E. (1940). A revision of the osmylid subfamilies conspicuously fused with CuP, while it has an oblique, distinct Stenosmylinae and Kalosmylinae. Novitates Zoologicae, 42(1), origin in close relatives (Sharov 1968; Béthoux and Nel 165–201. 2001). Based on the current case, we emphasize on the rele- MacLachlan, R. (1867). New genera and species, &c., of neuropterous insects; and a revision of Mr. F. Walker's British Museum Catalogue vance of series of conspecific individuals, ideally fossil and of Neuroptera, part ii. (1853), as far as the end of the genus extant, to address problematic THCs. Myrmeleon. Zoological Journal of the Linnean Society, 9(9–10), 230–244. Acknowledgments We are grateful to Christine L. Lambkin (Queens- Makarkin, V. N., Yang, Q., & Ren, D. (2014). A new basal osmylid land Museum) and David R. Britton Derek J. Smith (Australian Museum) neuropteran insect from the Middle Jurassic of China linking for loaning the specimens they were/are in charge of. V. Makarkin for his Osmylidae to the Permian-Triassic Archeosmylidae. Acta comments on an earlier version of this contribution. Palaeontologica Polonica, 59(1), 209–214. New, T. R. (1983a). A revision of the Australian Osmylidae: Kempyninae (Insecta: Neuroptera). Australian Journal of Zoology, 31(3), 393– 420. References New, T. R. (1983b). Revision of the osmylid subfamilies Porisminae and Eidoporisminae. Australian Journal of Zoology, 31(5), 763–770. Aspöck, H., Aspöck, U., & Hölzel, H. (1980). Die Neuropteren Europas: New, T. R. (1986). A new Australian genus of Stenosmylinae eine zusammenfassende Darstellung der Systematik, Ökologie und (Neuroptera: Osmylidae). Systematic Entomology, 11(4), 447–452. Chorologie der Neuropteroidea (Megaloptera, Raphidioptera, New, T. R. (1988). Osmylidae (Insecta: Neuroptera) from Irian Jaya. Planipennia) Europas. Krefeld: Geocke and Evers. Invertebrate Taxonomy, 2(6), 805–839. Bechly, G. (1996). Morphologische Untersuchungen am Flügelgeäder New, T. R. (1989). The genus Oedosmylus Krüger (Insecta: Neuroptera: der rezenten Libellen und deren Stammgruppenvertreter (Insecta; Osmylidae). Invertebrate Taxonomy, 3(2), 135–148. Pterygota; Odonata) unter besonderer Berücksichtigung der Redtenbacher, J. (1886). Vergleichende Studien über das Flügelgeäder Phylogenetischen Systematik und des Grundplanes der Odonata [re- der Insekten. Annalen des Kaiserlich-Königlichen vised edition including appendix in English]. Petalura, 2,1–402. Naturhistorischen Hofmuseums,1,153–232+pl. 9–20. Béthoux, O. (2007). Cladotypic taxonomy applied: titanopterans are or- Remane, A. (1952). Die Grundlagen des natürlichen Systems, der – thopterans. Arthropod Systematics & Phylogeny, 65(2), 135 156. vergleichenden Anatomie und der Phylogenetic.Leipzig: — Béthoux, O. (2012a). Grylloptera a unique origin of the stridulatory file Akademische Verlagsgesellschaft. in katydids, crickets, and their kin (Archaeorthoptera). Arthropod Ren, D., & Yin, J. (2003). New 'osmylid-like' fossil Neuroptera from the Systematics & Phylogeny, 70(1), 43–68. Middle Jurassic of Inner Mongolia, China. Journal of New York Béthoux, O. (2012b). King crickets, raspy crickets & weta, their wings, Entomological Society, 111(1), 1–11. their fossil relatives. Journal of Orthoptera Research, 21(2), 179– Sharov, A. G. (1968). Filogeniya orthopteroidnykh nasekomykh. Trudy 225. Paleontologicheskogo instituta, Akademiya Nauk SSSR, 118,1–216. Béthoux, O., & Nel, A. (2001). Venation pattern of Orthoptera. Journal of Orthoptera Research, 10(2), 195–198. Shi, C., Béthoux, O., Shih, C., & Ren, D. (2012). Guyiling jianboni gen. Béthoux, O., & Wieland, F. (2009). Evidence for Carboniferous origin of et sp.n., an antlion-like lacewing, illuminating homologies and trans- the order Mantodea (Insecta: Dictyoptera) gained from forewing formations in Neuroptera wing venation. Systematic Entomology, – morphology. Zoological Journal of the Linnean Society, 156(1), 37(3), 617 631. 79–113. Shi, C., Winterton, S. L., & Ren, D. (2015). Phylogeny of split- Brongniart, C. (1893). Recherches pour servir à l'histoire des insectes footed lacewings (Neuroptera, Nymphidae), with descriptions fossiles des temps primaires précédées d'une étude sur la nervation of new Cretaceous fossil species from China. Cladistics, 31(5), – des ailes des insectes. Bulletin de la Société d'Industrie Minérale de 455 490. Saint-Etienne (3rd series). 7: 124–615. Walker, F. (1853). Catalogue of the specimens of neuropterous insects in Carpenter, F. M. (1943). Osmylidae of the Florissant shales, Colorado the collection of the British Museum. Part I. –(Sialidae– (Insecta–Neuroptera). American Journal of Science, 241(12), 753– Nemopterides). London: Trustees of the British Museum. 760. Wang, Y., Liu, Z., Ren, D., & Shih, C. (2010). A new genus of Cui, Y., Béthoux, O., Klass, K.-D. & Ren, D. (2015). The Jurassic Protosmylinae from the Middle Jurassic of China (Neuroptera: Bajanzhargalanidae (Insecta: Grylloblattida?): new genera and Osmylidae). Zootaxa, 2480,45–53.