Topographic homology conjectures (= THCs) in fore- and hind wing venation of extant king crickets, raspy crickets and weta are re-evaluated. Based on the premises that topological homology is established so that the amount of transformation that has to be assumed to explain differences between patterns is minimized, a new set of THCs (= STHC), based on morphological data on the species cf. mexicanus de Saussure, 1859: 209, laudatum Johns, 1997, ornata Willemse, 1963, cf. bicornis Karny, 1929a, pinguipes? Rentz in Morton & Rentz, 1983, pinguipes Rentz in Morton & Rentz, 1983, an undetermined species, punctipennis Walker, 1869, and rufovaria Kirby, 1888, is elaborated. Among other transformations, the occurrence of a re-routing of MP along CuA (i.e., the basal M CuA stem splits into MA and MP CuA, instead of into MA MP and CuA) is demonstrated for pinguipes, punctipennis, and rufovaria. Concurrently, intra-specific variability in forewing venation is appreciated for laudatum, ornata, cf. bicornis, pinguipes, punctipennis, and rufovaria. A cladotypic-compliant nomenclatural scheme is elaborated based on the proposed THCs. The taxon Agryllacris nom.-dis.typ. nov. is defined based on the character state ‘in forewing, CuA CuPaα with two distal branches only’; the taxon Tagryllacris nom.-dis.-typ. nov. is defined based on the character state ‘in forewing, M CuA splits into MA and MP CuA’; the taxon Etagryllacris nom.-dis.-typ. nov. is defined based on the character state ‘in forewing, CuA CuPaα keep fused’; the taxon Metagryllacris nom. nov, dis. Zeuner, 1939, typ. nov. is defined based on the character state ‘in forewing, MA fused with R at wing base’. Based on new data, the species †perfecta Sharov, 1968 is identified as a member of the taxon Agryllacris, and the fossil species †megaptera Gorochov, 1987a, isimplicis Gorochov, 1987a, †elongata Gorochov, 1987a, †madygenioides Gorochov, 1987a, †perlonga Gorochov, 1987a, and † devexa Gorochov, 1987a are considered as junior synonyms of †perfecta. The nomenclatural treatment is extended to the unrelated taxon †Bintoniellidae nom. Handlirsch, 1938, dis. Sharov, 1968, typ. nov, defined based on the character state ‘in forewing, no distinct base of CuPaα diverging from CuPa (and fusing with CuA or M CuA)’. Belonging to this taxon, the species †primaria Sharov, 1968 is considered as a junior synonym † triassica Sharov, 1968:168.A consequence of the assignment of †perfecta to Agryllacris is that this group stems in the Triassic, at least, i.e., ca 170 million years earlier than previously assumed. As a result, the loss of the complex ‘tettigonid-gryllid-like’ (i.e., Grylloptera) stridulatory apparatus is no longer needed to account for the morphology of king crickets, raspy crickets and weta forewing. Finally, a refinement of the conce
Despite lack of consistent reconstruction of phylogenetic relationships among major ensiferan lineages (Legendre et al. 2010), it is generally admitted that king crickets, raspy crickets and weta diverged from a Mesozoic paraphyletic assemblage, namely the hagloideans (Gorochov 2001; Heads & Leuzinger 2011; Sharov 1968, 1971). Because hagloidean males possess the ‘tettigonid-gryllid-like’ stridulatory apparatus in forewings (including a file, that defines the taxon Grylloptera nom. Haeckel, 1896, dis.-typ. Béthoux, 2012), some assumed that king crickets, raspy crickets and weta experienced a loss of the corresponding complex structures (Gorochov 2001; Sharov 1968, 1971; among others). However several authors recognized the wing venation of these insects as plesiotypic, compared to that of other ensiferans (Karny 1929b), and assumed a primary lack of the Grylloptera stridulatory apparatus (Zeuner 1939). The assumption that king crickets, raspy crickets and weta diverged from a group that already possessed the Grylloptera stridulatory apparatus, is primarily supported by the lack of evident Mesozoic remains.
The wing venation of extant king crickets, raspy crickets and weta is sparsely documented. Apart from Karny (1930), most papers by this author include incomplete drawings of wing venation (e.g., Karny 1928b). Several other contributions focus on species that most likely represent the most derived members of the group (e.g., Karny 1928a, Ragge 1955). Illustrations and observations by Zeuner (1939) are minimal and will prove partly inaccurate. This lack of data jeopardizes the identification of fossil stem-relatives of king crickets, raspy crickets and weta.
In order to obtain reference data for a discussion on the origin of these insects, a survey of the holdings of the Australian National Insect Collection and CSIRO Ecosystem Sciences (Canberra, Australia) was carried out, complemented by a survey of material housed at the Queensland Museum (Brisbane, Australia), and of a few other institutions. Preliminary results suggested that relatives of king crickets, raspy crickets and weta could be identified among Early Mesozoic orthopterans. This finding prompted a broadening of the scope of this contribution to include data on intra-specific wing venation variability in extant species, in order to assist delimitation of related fossil species.
The paper is organized as follows: a comparative analysis is carried out, aiming to develop wing venation topological homologies in selected species (and including data on intra-specific variability). Based on the resulting conjectures, a character-state-based nomenclatural treatment (i.e., cladotypic) is proposed, including the revision of several fossil taxa relevant for the following section; finally, conjectures and nomenclature are used to elaborate on evolutionary aspects.
Material and methods
I use the cladotypic nomenclatural procedure I elaborated (2007b, c; 2010) and Lanham's species names (Dayrat et al. 2004, Lanham 1965) [see introduction and application of this approach in JOR by Béthoux & Herd (2009)]. In order to avoid a mixing of names erected under distinct procedures, vernacular versions of taxon names erected under the Linnaean procedure will be used (e.g., gryllacrididaeans instead of Gryllacrididae). The position of the species surveyed in this contribution under the traditional nomenclatural procedure is provided in Appendix 1. In addition, traditional genera are indicated in this section, between inverted commas (except in Appendix 2). Appendix 2 provides a nomenclatural treatment at the specific and generic levels compliant with the ICZN. Appendix 3 provides provisional compositions for the newly defined taxa.
Specimens referred to as ANIC IWC OB, QM, PIN, NHM, and MHNG are housed at the Australian National Insect Collection and CSIRO Ecosystem Sciences (Canberra, Australia), the Queensland Museum (Brisbane, Australia), the Palaeontological Institute of the Russian Academy of Sciences (Moscow, Russia), the Natural History Museum (London, UK), and the Museum d'histoire Naturelle (Geneva, Switzerland), respectively. Other institutions are mentioned in the text.
Extant species and specimens used for this contribution are as follows [following Bengston (1988) for open nomenclature]:
—The studied specimen was determined at the genus level. Since the type-species of the corresponding genus is mexicanus de Saussure, 1859: 209, I refer to the species as cf. mexicanus in the following.
—Specimen illustrated (ANIC pinned collection): ANIC IWC OB 26, ♀, Monte Verde, Powell Prop., ca 1580 m., Costa Rica, 23.iii.1974, determined by D.C.F. Rentz.
Species laudatum Johns, 1997 (genus “Transaevum Johns, 1997). —Determination of specimens of this species was complicated by the claim of the existence of two distinct and undescribed species in the genus by Johns (1997; balanced in Monteith & Field 2001). However, only one species has been formally described. Observation of the ‘type-series’ of the presumed second species housed at the Queensland Museum (material labeled ‘longihamatum’) revealed no distinctive features except a size comparatively smaller than material of laudatum, but nevertheless compatible with identity with this species. In addition, specimens of ‘longihamatum’ were collected from a locality where laudatum occurs, rendering the distinctiveness of ‘longihamatum’ unlikely according to CB. Monteith (pers. com. 2010). I presume that specimens housed at the ANIC and at the QM indicated as belonging to laudatum were determined by P.M. Johns; determination of two unassigned ANIC specimens by O. Béthoux was based on the fact that the set of ‘longihamatum’ specimens was composed of comparatively smaller specimens. All specimens were collected from Australia. Because of the large sample available, specimens belonging to the ANIC pinned collection with wings in resting position were excluded from the survey. The specimen of the paratype series of ‘longihamatum’ illustrated on Fig. 3H, Q was excluded from the measurement survey.
—Specimens prepared and illustrated (ANIC IWC OB 1–9 from ANIC ethanol collection; ANIC IWC OB 10 from ANIC pinned collection): ANIC IWC OB 5, ♀, 17.37S 145.34E, BS3 Massey Crk., 27.ii–26.iii.1996; ANIC IWC OB 6, ♀, 17.07S 145.38E, Mt. Haig Road, 10 kms. ENE of Tinaroo Dam Wall, Atherton Tableland, QLD, 03.ii. 1988; ANIC IWC OB 7, ♂, 17.02S 145.37E, Davies Crk., 15 km. from Kennedy Hwy., 21 km E. by S. of Mareeba, QLD, 01.iii.1988; ANIC IWC OB 8, ♀, 17.37S 145.34E, BS3 Massey Crk., QLD, 31.i–27.ii.1996; ANIC IWC OB 9, 3, 17.37S 145.34E, BS3 Massey Crk., QLD, 04.ii-06.iii.1995, determinedby O. Béthoux; ANICIWCOB 10, ♂, Kuranda, QLD, 27.ii.1981, determined by O. Béthoux; QM (unnumbered, ‘paratype series’ of ‘longihamatum’), ♂, Charmillin Ck. Xing, 950 m., Tully Falls Rd., QLD, 8.xii.l989-5.i.l990 (only left forewing prepared).
—Specimens used for measurements survey (in addition to the illustrated ones):
Females: ANIC pinned collection, 15.47S 145.17E, Moses Ck., 4 km NE of Mt. Finnigan, nr. Cooktown, QLD, 14.x. 1980 (paratype); ANIC pinned collection, Paluma, QLD, 17.x.1979 (paratype); ANIC pinned collection; Davies Ck. Rd., nr. Kuranda, QLD, 9.i.1979 (paratype); ANIC pinned collection; 16.50S 145.37E, 1.5 km S of Kuranda on Kennedy Hwy.‘Arona’, QLD, 10.xii.2003 (2 specimens); ANIC pinned collection; 17.25.43S 145.19.13E, Atherton Tableland, Mt. Hypipamee Nat. Pk., QLD, 1–12.ii.1998; ANIC ethanol collection, 17.05S 145.35E, Mt. Haig, 22 km NE by N of Atherton, QLD, 16.iii.1988 (2 specimens); 16.16S 145.02E, Forestry Camp Mt. Winsdor Tableland, NNW of Mt. Carbine, QLD, 23.1.1988; ANIC ethanol collection, 17.05S 145.35E, Mt Haig, 22 km NE by N of Atherton, QLD, 16.iii.1988; ANIC ethanol collection, 17.06S 145.36E, Mt. Haig, QLD, 4.ii–17.iii.1995; ANIC ethanol collection, 17.28S 145.29E, Longlands Gap, QLD, 5–27.ii.l996;ANICethanolcollection, 17.02S 145.37E, Davies Ck. (15 km from Kennedy Hwy.), 21 km E by S of Mareeba, QLD, 1.iii.1988 (2 specimens); QM ethanol collection, 16.26S 145.42E, Hughes Rd., Topaz, 4.xii. 1993, QLD (2 specimens, paratypes); QM ethanol collection, 16.05S 145.25E, upper roaring meg, 680m, 6 km W. of Capte tribulation, QLD, 10–22.xi.1993, (3 specimens, paratypes); QM pinned collection, Cardwell Range, Upper Broadwater Ck Valley, QLD, 17–21 .xii.1986, 700–80Om (only left forewing observed; paratype); QM pinned collection, Douglas Ck. Rd., Kirrama Range, QLD (only left forewing observed; paratype); QM pinned collection, Upper Broadwater Ck. Valley, Cardwell Range, QLD, 17–21 .xii.1986 (only right forewing observed; paratype); QM pinned collection, Head of Roots Ck, 12 km WNW Mossman, QLD, 28–29.xii.1990; QM pinned collection, Mt Lewis Rd., QLD.
Males: ANIC pinned collection, 32 km NW of Kennedy Hwy., QLD, 24.xii.1979 (paratype); ANIC pinnedcollection, 15.47S 145.17E, Moses Ck., 4 km NE of Mt. Finnigan, nr. Cooktown, QLD, 16.x.1980 (only left forewing observed; paratype); ANIC pinned collection; 18.59S 146.10E, Birthday Ck., 6 km. NW by W of Paluma, QLD, 25. ix.1980 (paratype); ANIC pinned collection, Kuranda, QLD, 27.ii.1981 (2 specimens; in one, only left forewing observed); ANIC pinned collection; 16.50S 145.37E, 1.5 km S of Kuranda on Kennedy Hwy. ‘Arona’, QLD, 10.xii.2003; ANIC ethanol collection, Mt. Lewis Rd., Julatten, QLD, 12.xi.1975 (paratype); ANIC ethanol collection, 17.02S 145.37E, Davies Ck. (15 km from Kennedy Hwy.), 22 km E by S of Mareeba, QLD, 24.iii.1988; ANIC ethanol collection, 17.05S 145.35E, Mt. Haig, 22 km NE by N of Atherton, QLD, 16.iii.1988; QM ethanol collection, 16.26S 145.42E, Hughes Rd., Topaz, QLD, 4.xii.1993 (paratype); QM ethanol collection, 1km S Cable Tower No7, Bellenden-Ker Range, QLD, 17–24.x.1981 (paratype); QM ethanol collection, upper roaring meg, 6 km W. Cape Tribulation, QLD, 8–9.xii.1993 (paratype); QM ethanol collection, upperlsley Ck., 9 km WSN of Edmonton, QLD, 29–30. xi.1993 (paratype); QM pinned collection, 17.16S 145.49E, Massey Range, 4 km W of Centre Bellenden Ker, 9–11.x. 1991 (holotype); QM pinned collection, 32 km NW of Kennedy, QLD, 24.ii.1979 (paratype); QM pinned collection, 17.14S 145.48E, Massey Range, 6 km NW of Centre Bellenden Ker, 11–12.xi.1991 (only right forewing observed; paratype); QM pinned collection, Charmillin Ck. Xing, Tully Falls Rd., QLD, 8.xii. 1989-5.i.1990 (2 specimens, both paratypes; only right forewings observed); QM pinned collection, head of Francis Ck, 12 km WNW Mossman, QLD, 30 xii.1989 (only right forewing observed); QM pinned collection, Mt Lewis Rd., 16km from Highway, QLD, 18.xii.1989-13.1.1990 (only left forewing observed).
—Determination of material of this species is straightforward as the species is very distinctive. All specimens were collected from Australia.
—Specimens prepared and illustrated (ANIC ethanol collection): ANIC IWC OB 15, ♂, 17.28S 145.29 E, Longlands Gap, QLD, 1–30.xi. 1995, determined by D. C.F. Rentz; ANICIWCOB 16, ♂, 17.35S 145.34E, BS3 Massey Crk., QLD, 3.x-2.xi.1995, determined by D.C.F. Rentz.
—Specimens used for measurements survey (in addition to illustrated specimens):
Females: ANIC, pinned collection, Atherton tableland, QLD (only right forewing observed); ANIC, pinned collection, no data, collected by JT Doyer (only right forewing observed); ANIC ethanol collection, 17.33S 145.32E, Mt Fisher, QLD, 4.ii–21.iii 1995; ANIC ethanol collection; 17.27S 145.29E, Hugh Nelson Rd. QLD, 1–30.xi.1995; ANIC ethanol collection, 17.15S 145.38E, Lake Barrine, Atherton Tableland, QLD, 6.xii. 1985; QM ethanol collection, 17.30S 145.36E Millaa Millaa Falls, 13–24.xi. 1994 (three specimens); QM ethanol collection, 16.26S 145.42E, Hughes Rd.,Topaz, QLD, 4.xii.1993; QM ethanol collection, 17.23S 145.28E, Upper Plath Rd., 9.xii.1995; QM ethanol collection, 17.17S 145.58E, Graham Range, 8–9.xii.1995; QM pinned collection, 17.26S 145.42E, Hughes Rd, Topaz, 6.xii.l993-25.ii.l994.
Males: ANIC pinned collection, Mt Haig, nr. Danbulla, QLD, 28.ix.1966; ANIC dry collection, Douglas Ck., lamb range, NE QLD, 12.x.1992 (only right forewing observed); ANIC ethanol collection, 17.37S 145.34E, Massey Ck., QLD, 2–20. xi.1995; QM ethanol collection, 17.28S 145.32E, Kenny Road, 25.xi.1994-10.1.1995; QM ethanol collection, 17.37S 145.46E, Palmerston NPE Margin, 10.xii.1995-7.ii.1996; QM ethanol collection, 17.27S 145.29E, Tower nr. the Crater NP, 25.xi.1994, 1230m; QM pinned collection, Bulurru, Topaz, 23.ix.1997 (only right forewing observed); QM pinned collection, 17.14S, 145.48E, Massey Range, 6 km NW of Centre Bellenden Ker, 11–12.x.1991 (only right forewing observed); QM pinned collection, 17.14S, 145.48E, Massey Range, 6 km NW of Centre Bellenden Ker, 11–12.x.1991.
—An issue similar to that faced with laudatum was raised by the claim that two species, similar but distinct from bicomis exist, but are undescribed yet (Johns 1997; balanced in Monteith & Field 2001). The selected material (illustrated and measured) was labeled ‘biconulata’, presumably by P.M. Johns. Apart from a comparatively smaller size, material of ‘biconulata’ does not significantly differ from that of bicomis, therefore I consider the existence of a ‘biconulata’ dubious (specimens of intermediate size exist). In any case, the survey is based on material from a single species, either bicomis, or ‘biconulata’. At worst the intraspecific variability will have been underestimated. All specimens were collected from Australia.
—Specimens prepared and illustrated (all ANIC ethanol collection): ANIC IWC OB 11 and ANIC IWC OB 12, ♀, 17.06.S 145.37E, GS2 Mt. Edith, QLD, 4.ii–17.iii.l995; ANIC IWC OB 13, ♀, 17.33S 145.32E, BS2 Mt. Fisher, QLD, 21.iii-5.iv.1995; ANIC IWC OB 14, ♀, 17.25S 145.29E, GS3 Hugh Nelson Rd., QLD, 5.iii–4.iv.l995.
—Specimens used for measurements survey (in addition to the illustrated ones):
Females: ANIC ethanol collection, Mt Belleden-Ker, Summit, 11.iv.1979, QLD (indicated as ‘paratype’); ANIC ethanol collection, 17.33S 145.32E, Mt. Fisher, QLD, 3.x.-2.xi.1995 (fourspecimens); ANIC ethanol collection, 17.37S 145.34E, Massey Crk., QLD, 2–30.V.1996 (fourspecimens); QM ethanol collection, Belleden Ker Range, Summit TV Stn., QLD, 25– 31.x. 1981 (indicated as ‘paratype’); QM ethanol collection, paratype; Belleden Ker Range, NQ, Cable Tower 3, 1054m, 1–7.xi.1981 (only right forewing observed);
Males: ANIC ethanol collection, 17.33S 145.32E, Mt. Fisher, QLD, 3.x.-2.xi.1995; QM ethanol collection Belleden Ker Range, Summit TV Stn., 25–31 .x. 1981 (indicated as ‘holotype’).
—The studied specimen was collected prior to investigations by Morton & Rentz (1983). It is housed in the ANIC pinned collection, together with pinguipes specimens. However based on its forewing venation, the specimen differs from representatives of species by its higher number of RP branches. Therefore I assume it could belong to a distinct species.
—Specimen prepared and illustrated: ANIC IWC OB 24, ANIC pinned collection, ♂, Stuart Hwy., 296 km S. of Tennant Crk., NT, Australia, 29.xi.1972.
—All specimens were collected from Australia and determined by D.C.F. Rentz. Only specimens from the ANIC ethanol collection were surveyed (i.e., pinned specimens were excluded), because this sample appeared sufficient. Two specimens were excluded from survey because of dubious determination. Collecting data of these specimens are as follows: 32.29S 124.13E, 59 km ESE of Balladonia Motel, WA, 29.1.1991.
—Specimens prepared and illustrated: ANIC IWC 17, ♂, ca 30.39S121.27E,7–11km N of Kalgoorlie„ WA, 17.xi.1978; ANIC IWC OB 18, ♀, ca 30.39S 121.27E, 7–11 km N of Kalgoorlie, WA, 17.xi.1978; ANIC IWC OB 19, ♀, 24.51S 133.HE, Stuart Hwy., 41 km N of Erlunda, NT, 4.xi.l988: ANIC IWC OB 20, ♀, 20.52S 130.16E, Tanami Desert, NT, 21 .iii.1988; ANICIWC OB 21, ♀, 23.17S 133.39E, 50 km. NNW of Alice Springs, NT, 15.xii.1989; ANIC IWC OB 22, ♂, ca 20 km E of Norseman, WA, 12–13.i.1986; ANIC IWC OB 23, ♂, 32.13S, 123.15E, 38 km. WNW of Balladonia Motel, WA, 22.xi.1980.
—Specimens used for measurements survey (in addition to the illustrated ones):
Females: 26.00S 131.25E, 26 km WSW of Milga Pk., NT, 18.i.1982 (paratype); 26.09S 130.35E, 56 km W of Amata, Musgrave Ranges, SA, 20–21.i.1982 (paratype); 25.12S 132.05E, Lasseter Hwy., 35 km ENE. of Curtin Springs, NT, 8.xi.1988; 25. HS 133.11E, Stuart Hwy., 4 km N of Erlunda, NT, 4.xi.1988; 20.52S 130.16E, Tanami Desert, NT, 28.iii.1988; 30.42S 121.28E, 4 km N by E of Kalgoorlie, WA, 17.ii.1978 (6 specimens); 25.11S 133.1E, Stuart Hwy., 4km N. of Erlunda, NT, 4.xi.1988; 25.21S 131.53E, Lasseter Hwy., 13km E by S of Curtin Springs, NT, 8.xi.1988 (2 specimens); ca. 30.39S 121.27E, 7–11 km N of Kalgoorlie, WA, 17.xi.1978 (4 specimens); 30.47S 121.27E, Kalgoorlie (airport), WA, 16.ii.1978 (4 specimens); 24.51S 133.11E, Stuart Hwy., lkm N of Erlunde, NT, 4.xi.1988; 28.47S 121.31E, 1 km S of Malcolm, WA, 19.ii.1978 (2 specimens).
Males: 26.00S 131.25E, 26 km WSW of Milga Pk., NT, 18.i.1982 (paratype); 26.00S 131.25E, 26 km WSW of Milga Pk., NT, 18.1.1982 (2 specimens; both paratypes); ca. 20 km E of Norseman, WA, 12–13.i.1986 (2 specimens); 1 km W of Sangsters Bore, Tanami Desert, NT, 20.iii.1985; 25.11S 133.11E, Stuart Hwy., 4 km N of Erlunda, NT, 27.x.1988; 20.52S 130.16E, Tanami Desert, NT, 18.iii.1988 (2 specimens); ca. 30.39S 121.27E, 7–11 km N of Kalgoorlie, WA 17.11.1978; 30.47S 121.27E, Kalgoorlie (airport), WA, 16.ii.1978 (4 specimens); 28.47S 121.31E, 1 km S. of Malcolm, WA, 19.ii.1978; ca. 30.39S 121.27E, 4 km N by E of Kalgoorlie, WA 17.11.1978.
Species ‘sp. undet.’.
—The corresponding specimen belongs to a set labeled as ‘Gryllacrididae Genus 3 Sp. 1’ in the ANIC pinned collection.
—Specimen illustrated: ANIC IWC OB 25, S, Yass, NSW., Australia, 13.xii. 1932.
—All surveyed specimens belong to the ANIC ethanol collection, were determined by D.C.F. Rentz, and were collected from Australia.
—Specimens prepared and illustrated: ANIC IWC OB 36, ♀, 11.45S, 142.35E, Heathland, QLD, 1–21.iii.1992; ANIW IWC OB 37, ♂, 15.50S, 145.20E, Gap Ck., 5 km. ESE of Mt. Finnigan, nr. Cooktown, QLD, 13–16.V.1981; ANIC IWC OB 38, ♂, 15.28S, 145.15E, 1km. w. of Cooktown, QLD, 12–13.V.1981. —Specimens used for measurements survey (in addition to the illustrated ones):
Females: 12.45S 143.17E, 8km E of Mt.Tozer, nr. Iron Range Nat. Pk., QLD, 8.vii.1986; 12.44S 143.16E, 6 km ENE of Mt. Tozer, nr. Iron Range Nat. Pk., QLD 1 .vii. 1986; Shiptons Flat, nr. Cooktown, QLD, 16–18.v.1981; 15.50S 145.20E, GapCk., 5 km ESE of Mt. Finnigan, nr. Cooktown, QLD, 13–16.V. 1981; 11.45S 142.35E, Heatlands, QLD, 18.ix.1992–21.x.1992.
Males: 12.43S 143.17E, 9 km ENE of Mt. Tozer, Iron Range Nat. Pk., QLD, 10.vii.1986; 12.44S 143.14E, 9 km ENE of Mt. Tozer, nr. Iron Range Nat. Pk., QLD, 28.vi.1986-4.vii.1986; 12.43S 143.17E, 9 km ENE of Mt. Tozer, Iron Range Nat. Pk., QLD, 10.vii.1986; 15.50S 145.20E, Gap Ck., 5 km ESE of Mt. Finnigan, nr. Cooktown, QLD., 13–16.V.1981; 15.04S 145.07E, Mt. Webb Nat. Pk., nr. Cooktown, QLD, 27–30. iv.1981; 17.00S 145.50E, Pine Ck. (nr. CSIRO Tower), 11 km SE by S of Cairns, QLD, 18.ii.1988; 17.53S 146.09E, Dunk Island (airstrip), QLD, 17.iv.1990.
—All specimens were collected from Australia, and all are housed at the ANIC (ethanol collection).
—Specimen prepared and illustrated: ANIC IWC OB 27, ♂, 10.27S, 105.40E, Christmas Island, Grants Well to Irvine, Hill Rd. track,, 27.iv. 1989, presumably determined by D.C.F Rentz. —Specimens used for measurements survey (in addition to the illustrated one; ♂ and ♀ treated together because of limited sample size):
Males: 10.27S 105.40E, Grants Well to Irvine, Hill Rd. track, Christmas Island, 27.iv.1989; 10.29S 105.38E, near Central Area, Workshop, Christmas Island, 17.iv.1989; 10.30S 105.55E, EW Parl track, Christmas Island, 28.iv.1989.
Female: 10.27S 105.40E, Grants Well to Irvine, Hill Rd. track, Christmas Island, 27.iv.1989.
Wings of illustrated extant specimens were cut off specimens preserved dry or under ethanol, softened as necessary, and mounted in white Euparal (Asco Laboratories, Manchester, UK). Left wings were mounted on their dorsal side, right wings on ventral side. As far as possible, wings were mounted unfolded.
Photographs reproduced on Figs 2–9, 11–19, 22, 23, 27, 28 were taken with a Canon EOS 450D digital camera coupled to a Canon 50 mm macro lens (and an extension tube as appropriate), and to a Canon MP-E 65 mm macro lens. Photographs reproduced on Figs 1, 21 were taken with a Nikon Coolpix C2500L. Other photographs were provided by various colleagues. For wings of extant material, transmitted light was obtained from a VisiLED ACT Basis for small wings, and from a regular light table for larger ones. Photographs were dust off manually using a stamp tool, and processed with Adobe Photoshop. Unless specified, photographs of fossil specimens were taken under dry conditions.
Except for drawings reproduced in Figs 1 and 24, drawings were prepared with the aid of a Zeiss SteREO Discovery V8 stereomicroscope equipped with a pair of W-PL 10×/23 eye pieces, a Plan Apo S 1.0x FWD objective, and a camera lucida. Due to the occurrence of structural folds in fore- and hind wings, and as a result of the compression of the three-dimensional structure that an insect wing is, creases occurred in many preparations of extant material, in particularly in hind wings. For example, ScP seems to be located posteriorly with respect to RA on Fig. 18, but is actually ‘creased’ below RA. In forewings, the membranous area posterior to AA veins was also problematic to mount unfolded, because it tends to fold even once cover glass has been put on. Drawings were corrected manually, and with the aid of Adobe Photoshop for hind wings vannuses (distortion transformation).
I use the conjecture of topological homology proposed by Béthoux & Nel (2001) for orthopterans, itself based on the serial insect wing venation pattern (Lameere 1922, 1923). Corresponding abbreviations and color-coding are as follows: CP, posterior Costa; ScA, anterior Subcosta; ScP, posteriorSubcosta; R, Radius; RA, anterior Radius; RP, posterior Radius; M, Media (green); MA, anterior Media; MP, posteriorMedia; CuA, anterior Cubitus (orange); CuP, posterior Cubitus; CuPa, anterior branch of CuP (purple); CuPb, posterior branch of CuP; CuPaα, anteriorbranch CuPa (blue); CuPaβ, posterior branch of CuPa (red); AA1, first anterior Analis. Correspondence with some other conjectures for orthopterans can be obtained from Béthoux & Nel (2002a: table 2) and on Fig. 10. The criticism expressed by Gorochov (2005) regarding Béthoux & Nel's (2001, 2002a) conjecture are addressed in Béthoux (2007a). Subsequent comments by Rasnitsyn (2007) are addressed in Béthoux (2008).
Other abbreviations are as follows: LFW, left forewing; LHW, left hind wing; RFW, right forewing; RHW, right hind wing. A vein (or vein sector) is said to be convex if located on an elevation, and concave if located in a depression. I refer to cross-veins aligned longitudinally to main veins, which are secondary structures, as ‘intercalary veins’.
Several transformation types are mentioned in the following and are better defined here. A translocation is the complete fusion of a vein (or one of its branches) with a surrounding vein (or branch), from the origin of the former, or from the wing base even. This transformation was documented in Pantcholmanvissiida nom. Béthoux 2007d, dis. Béthoux & Nel 2002b, typ. Béthoux 2007d (Béthoux 2007d) and in mantodeans (Béthoux & Wieland 2009). The term ‘abbreviation’ refers to the complete lack of forewing apex, without alteration of the basal area morphology, common among orthopterans. The term ‘pectinate fusion’ refers to a case involving the successive pectinate emergence of branches of a vein fused with another, the former lacking a distinct main stem (documented in neuropterans, see Shi et al. 2012).
Measurements of forewing length of unprepared material were made using a caliper. Although measures were made once only, a confidence interval can be estimated based on the largest difference observed between left and right forewings of a single individual. This difference amounts to 1.1 mm. This estimated error accounts for error in judging the positions of the measurement points, and the accuracy of the reading. It is considered constant (i.e., does not vary with wing size). Width of the area between M and CuPaα, opposite the free part of CuA (diverging from M + CuA, before its fusion with CuPaa), was measured under a microscope equipped with a graticule. Apositive figure represents a free CuA, while a negative one indicates (1) that a M + CuA + CuPaa fusion occurs (i.e., there is no free part of CuA: e.g., Fig. 8E, G), and (2) the length of this fusion. This approach was prompted by the fact that the series of states ranging from ‘long free part of CuA’ to ‘long M + CuA + CuPaa’ form a continuum (Fig. 13M–Q) better appreciated by a single measurement.
The rationale of such analysis is clarified in Béthoux (2012: 44). In summary, conjectures of topological homology (=THCs) are established so that the amount of transformation that has to be assumed to explain differences between patterns is minimized. Ideally, sets of THCs (=STHCs) can be compared based on the total amount of transformation assumed by each STHC However the methodology for this step is still little developed. In the following competing STHCs (see Fig. 10; ‘classical’ STHC including Fig. 10EH; alternative STHC including Fig. 10I–N) will be compared based on the occurrence (or lack thereof) of predicted intermediate patterns, and their consistence with respect to observed variants. This approach is grounded in that any given STHC can be challenged by a new, more performing one.
The wing venation of cf. mexicanus de Saussure, 1859: 209, laudatum Johns, 1997, ornata Willemse, 1963, cf. bicomis Karny, 1929a, pinguipes? Rentz in Morton & Rentz, 1983, pinguipes Rentz in Morton & Rentz, 1983, sp. undet., punctipennis Walker, 1869, and rufovaria Kirby, 1888, will be compared successively in the following. These species were selected based on their position as putative intermediates (i.e., exhibiting a unique pattern), reliable determination at the species level, and/or abundance of specimens available to the author. Variations observed in several aspects of wing venation are summarized in Appendix 4.
The fossil species †uralica Sharov, 1968 (Fig. 1) was selected to represent the presumed plesiotypic ‘oedischioid’ forewing morphology. As with most other Permian stem-orthopterans and Carboniferous relatives (Béthoux et al. 2012), †uralica possesses a common stem composed of M (green) and CuA (orange), a branched M (into MA and MP), a distinct free part of CuA, and CuP (purple) giving rise to CuPa (purple) and CuPb, a CuPa giving rise to CuPaα (blue) and CuPaβ (red), and a fusion of CuA with CuPaα resulting into a composite stem. Béthoux (2007d) suggested that CuA is simple in oedischoids and groups that derived from them. A simple CuA is conjectured for †uralica and in the following.
Species cf. mexicanus
A single specimen of cf. mexicanus was investigated (Fig. 2). Comparison with wing morphology of laudatum (Fig. 3) suggests that cf. mexicanus possesses abbreviated wings. In forewings, traits allowing this species to be distinguished from †uralica are: first branch of ScP reaching ScA; M + CuA and R coalescing near wing base (variable; Fig. 2G–H ); RA simple; MA simple; CuA + CuPaα with two branches (i.e., CuPaα simple; in the right forewing CuPaα is probably forked with the two resulting branches re-uniting into a single stem); CuA, CuPaα, CuPaβ, and AA1 and other anal veins reaching wing margin in a very distal position; intercalary veins occur in areas between CuPaβ and CuPb, CuPb and AA1, and between anal veins. Otherwise, as in †uralica, a distinct free part of CuA (emerging from M + CuA, before fusing with CuPaα) occurs. It must be indicated that it could easily be confused with surrounding cross-veins, which are comparatively strong. Minor variation is exhibited by this specimen: RP is 2- and 3-branched in the left and right forewing, respectively.
The hind wing morphology of cf. mexicanus poses an issue relevant to the whole comparative analysis of hind wing venation. Unlike Palaeozoic stem-orthopterans (see Sharov 1968, 1971: figs 8C, F, 11C), the free part of CuPaα fusing with CuA was not detected. Although the basal stem of CuP can be identified from the wing base (it gives rise to CuPaβ and CuPb, and is located anterior to AA1; Fig. 2I,J), the area between M + CuA and CuPaβ is filled with regular cross-vein-like structures (Fig. 2I,J). Three conjectures are equally likely: (1) CuPaα is fused with M + CuA from the wing base (i.e., CuPaα is translocated onto M + CuA; if so ‘M + CuA’ and ‘CuP’ on Fig. 21 should read ‘M + CuA + CuPaα’ and ‘CuPaβ + CuPb’, respectively); (2) CuPaα fuses with M + CuA at the narrowing of the area between M + CuA and CuP; (3) CuPaα was reduced to a cross-vein (i.e., is inconspicuous) but actually fuses with CuA distal to its divergence from M + CuA. Translocations have been documented in several cases (Béthoux 2007d, Béthoux et al. 2010, Béthoux & Wieland 2009), making the conjecture (1) possible. However, it will be demonstrated below that the free part of CuA (emerging from M + CuA, basal to its fusion with CuPaα) can turn into a rather inconspicuous cross-vein-like structure, rendering (3) equally plausible. This issue is left unresolved.
Apart from this aspect, traits allowing the hind wing of cf. mexicanus to be distinguished from that of Palaeozoic stem-orthopterans are: RA simple (possibly a consequence of abbreviation); lack of strong inflexion point of the wing outline opposite the end of CuPb; CuA + CuPaα fusing with CuPaβ (near posterior wing margin). Notice the occurrence of a single intercalary vein between MP and CuA + CuPaα + CuPaβ.
Wing venation of laudatum was investigated based on a larger sample. The usual forewing venation of the species is represented on Figs 3C,L, 4C (and see Appendix 4), and is characterized by: veinlet(s) from ScP reaching ScA; RA anteriorly pectinate; RP 2-branched; near wing base, occurrence of a convex intercalary vein between M + CuA and CuPa (covering hind wing base at rest); MA distinct from R; MP distinct from CuA / CuA + CuPaα; short free part of CuA; CuA and CuPaα diverging distally, both simple.
Various deviations from this pattern were observed. In a few cases no genuine ScP veinlets reaches ScA (compare Figs 3A,J, 4A to Figs 3B,K, 4B, same individual). Regarding this trait, some conditions are ambiguous even (see Fig. 3H,Q), with a structure that could equally be a ScP veinlet, or a strong cross-vein. The length of the free part of M (between the divergence point of CuA and the branching of MP) varies, although it is usually short. The length of the free part of CuA varies moderately (Table 1), with one specimen exhibiting a short M + CuA + CuPaα stem (Figs 3E,N, 4E). Among the whole set of surveyed material, two forewings exhibit a connection of RA with ScP (one of them reproduced on Fig. 3E,N); a single forewing exhibits a fusion of MA with R (Figs 3G,P, 4F); one forewing exhibits a pectinate fusion of RP with RA (i.e., anterior branch of RP translocated onto RA, with successive origin of RP branches; Fig. 3H,Q; and see the ornata case); one forewing lacks the usual free part of CuPaα (diverging from CuPa), indicative of a translocation of CuPaa onto M + CuA (Figs 31,R, 4G). Other cases observed in single forewings but not illustrated are: RP 3-branched; RP simple; MP branched (MA branch translocated onto MP?), with anterior branch vanishing; MA fused with RP at wing base.
Hind wings of laudatum (Fig. 5) differ from those of cf. mexicanus (and all other surveyed extant species) chiefly by the occurrence of a RP + MA fusion, of somewhat variable length (Fig. 5A,B,D,E; the hind wing reproduced on Fig. 5E is about 1.14 longer than that represented on Fig. 5A–B). Other notable traits are: occurrence of a single intercalary vein between MP and CuA + CuPaα + CuPaβ, and area between AA1 and the first AA2 vein filled with strongly sigmoidal cross-veins. RP is usually 2-branched, but the two hind wings of one of the 6 prepared specimens have a 3-branched RP (Fig. 5E).
This species proved to have forewings often exhibiting rare and unusual traits, to the point that a usual condition was impossible to delimit based on the available material. The most noticeable aspect regards the organization of RP. As observed on forewing represented on Fig. 6A,B,D,E,F,H, the stem of R emits several posterior stems earlier to the usual RA anterior branches, and sometimes additional posterior stem(s) distal to the first anterior ones. These successive posterior stems are interpreted as RP branches. In other words, it is assumed that a pectinate fusion of RP with RA occurs. An alternative conjecture would be that the first posterior branch is MA, but this option is easily discarded: (1) as in †uralica, cf. mexicanus, and laudatum, the stem of M near the origin of CuA is forked into two simple branches, therefore the anterior stem of M observed in ornata is likely the genuine MA, and (2) the left forewing of the specimen ANICIWCOB 16 (Fig. 6C,G) has a MA actually translocated onto R (demonstrated by the basal origin of this stem, and the simple free part of M, hence composed of MP alone), yet it possesses ‘successive posterior branches of R [indeed RP]’. The organization resulting from the pectinate fusion is variable: points of divergence of the successive RP branches, and the number of their own branches, vary widely. As a consequence apical branches are not always evident to assign either to RA or RP, and the range of RP branches is difficult to delimit (ca 4–6).
In this species it is nearly equally common that CuA diverges distal to the MA / MP fork (Fig. 6A,E; i.e., CuA is fused with MP for a short distance before diverging), basal to the MA / MP fork (as in species surveyed above), or that CuA is not distinguishable from surrounding cross-veins (Fig. 6D,H). The branching pattern of CuA + CuPaα is also variable: usually its two components diverge distally (Fig. 6A, E), but can keep fused (Fig. 6C,G; Appendix4). Intermediate conditions occur, either with CuA (Fig. 6D,H) or CuPaα vanishing. A rare and informative case was observed in the right forewing of the specimen ANIC IWC OB 15 (Fig. 6B,F). At the first glance, a ‘Vein-like portion’ (indicated as ° on Fig. 6B) could have been considered as homologous to the basal section of CuPaα, as usual, and as observed in the other wing of the same individual (°on Fig. 6A). However, this conjecture must be discarded: in this particular forewing the portion of CuPaα distal to the origin of CuA (indicated as * on Fig. 6B) diverges from CuPa, i.e., is not fused with CuA. As a consequence the structure indicated as °on Fig. 6B is a ‘phantom CuPaα’, i.e., a remnant of the usual course of CuPaα. This is possibly a case of ‘tracheal un-capture’ (see Béthoux 2012 for this transformation type).
Hind wings of ornata lack the pectinate fusion of RP with RA, and lack the RP + MA fusion observed in laudatum. The number of branches of RP varies between 3 and 4 (based on two hind wing pairs). A single intercalary vein occurs between MP and CuA + CuPaα + CuPaβ. A noticeable trait is the very basal origin of RP.
Species cf. bicornis
Illustrations of parts of the wing venation of bicornis are available from Karny (1929a: fig. 29). A more comprehensive treatment will prove necessary. Owing to their comparatively long ScA and short ScP, forewings of cf. bicornis might be abbreviated.
The usual organization of the connection of CuA with CuPaα is represented on Fig. 8C,I (and see Appendix 4). It involves a short CuA diverging from M near the point of divergence of MA and MP, and fusing with CuPaα. However, a M + CuA + CuPaα fusion (therefore without free part of CuA) was observed in 6 cases (4 of them composing 2 forewing pairs of single individuals; Fig. 8D–F,J–L). The standard deviation of the width of the ‘CuA area / length of M + CuA + CuPaα’ is higher than in other species (Table 1), indicating an unusual variability of this character in cf. bicornis. Notice that standard deviation falls down to 0.44 if forewings of the specimen ANIC IWC OB 12 (the one with the two longest MP + CuA + CuPaα fusions observed; Fig. 8E–F, K–L) is removed. Yet it is higher than in the other surveyed species.
In addition, the free part of CuA was observed to diverge distal to the MA / MP fork (i.e., CuA is briefly fused with MP) in 6 cases (Fig. 8A,G). Also, MP was observed to diverge from CuA + CuPaα (is re-routed along the free part of CuA) in 2 cases (Fig. 8B,H).
Except from the area involving the free part of CuA (or the lack thereof), the forewing venation of this species is comparatively stable. Veinlets of ScP rarely reach ScA (Appendix 4), RP is usually simple (20 cases), sometimes forked (13 cases; including 3 presumed cases of ‘RA + RP pectinate fusion’).
The hind wing morphology of cf. bicornis is unusual in that M + CuA and CuPa are very inconspicuous, and fuse near the wing base (Fig. 9). Although inconspicuous, a CuPaβ distinct from CuA + CuPaα was observed, as in species surveyed above (both veins fuse before reaching the wing margin). Another noticeable aspect is the occurrence of strong ScP branches. Such organization is common in forewings of orthopterans, but its occurrence in hind wings seems unique (at least among species considered herein).
The corresponding specimen is essential to the establishment of forewing venation homologies of the species surveyed thereafter. Two possible transformation series are summarized on Fig. 10. They concern topological homologies of the medio-cubital area in forewings [according to Béthoux & Nel's (2001) conjecture and nomenclature for the whole orthopterans; correspondence with the Zeuner's (1939), Sharov's (1968, 1971), and Ragge's (1955) conjectures, under which the structure indicated as * is MP, is indicated by annotations in gray; annotations in light gray follow Karny's (1928b) conjecture, under which the structure indicated as * is a cross-vein].
The plesiotypic condition observed in †uralica (see Fig. 1) is schematized on Fig. 10A, and two patterns observed in pinguipes and closely related species are schematized on Fig. 10B,C. The aim of this section is to interpret these two patterns according to the conjecture of topological homologies posited for †uralica. Fig. 10F represents a conjecture proposed by Sharov (1968, 1971: fig. 27.B; see also Karny 1928b: figs 2, 9), assuming a simple M (Sharov's ‘MA’), and a 3-branched CuA + CuPaa (Sharov's ‘MP + CuA1’). According to this author, his ‘MA’ is branched in †uralica, so the implied transformation is M becoming simple. Another needed transformation involves the branching pattern of his ‘MP + CuA1’ (here CuA + CuPaα), splitting into a simple ‘MP’ (here CuA) and a forked ‘CuA1’ (here CuPaα).
Fig. 10G represents the conjecture for the pattern reproduced on Fig. 10C according to Zeuner (1939: pl. 2, upper figure). This conjecture assumes a simple M (Zeuner's ‘MA’) and a 2-branched CuA + CuPaα (Zeuner's ‘MP + CuA’). Provided that Zeuner assumes a forewing venation ground plan of orthopterans with a branched ‘MA’ (Zeuner 1939: pi. 9, upper figure), this author assumes the same transformation as Sharov (1968, 1971), viz. M (or ‘MA’) becoming simple in pinguipes and closely related species.
Two patterns can bridge those represented on Fig. 10A and Fig. 10F. One of them is represented on Fig. 10E, and assumes that the branching pattern of CuA + CuPaα was modified first. The other one (not represented) assumes that M becomes simple first. The most critical aspects of the argument being made here is that neither of these two patterns has ever been observed in actual specimens.
To date, another transformation series has never been considered. Starting from Fig. 10A, it assumes that CuPaα becomes simple (Fig. 101). The corresponding pattern fits that observed in cf. mexicanus, laudatum, ornata, cf. bicornis, and was observed in the left forewing of the specimen ANIC IWC OB 24 (Fig. 11A,C,E,G). In the last case, this morphology is unusual for the species.
Then (Fig. 10J) MP is assumed to be re-routed together with the free part of CuA; the composite stem fuses with CuPaα; MP and CuA + CuPaa diverge distally; CuA and CuPaα diverge distally. Such a re-routing of MP was documented by Sharov (1968, 1971) in the Gigantitanidae (and see Béthoux 2007d: fig. 1D). This transformation was also documented in cf. bicomis (rare occurrences), and is therefore highly plausible.
This alternative transformation series supposes the occurrence of a common stem MP + CuA. Direct evidence for this common stem is provided by the right forewing of the specimen ANICIWC OB 24 (Fig. 11B,D,F,H). Exceptionally enough, this wing lacks the fusion of CuA (or MP + CuA) with CuPaα, present in all orthopterans since the Late Carboniferous (at least). Homologies in the right forewing are readily established based on the left one (compare Fig. 11E and F): the distal portions match, with a two-branched M, and simple CuA, CuPaα, and CuPaβ. Tracing MP and CuA backwards in the right forewing reveals that they form the common stem MP + CuA, as predicted.
This specimen indicates that the most plausible interpretation of the pattern represented on Fig. 10B is provided on Fig. 10J. From this step, interpretation of the pattern represented on Fig. 10C is straightforward: CuA and CuPaα keep being fused up to the posterior wing margin (Fig. 10 K).
Forewing THCs in this species are mainly established based on the previous case, but an additional specimen belonging to pinguipes supports the ‘MP + CuA’ conjecture (schematized on Fig. 10I–K). Forewings of this specimen are represented on Fig. 12. In both forewings M + CuA (presumably) separates into MA and MP + CuA, but MP diverges from MP + CuA and re-unites with MA, ‘re-forming’ the M stem observed in previous species. Another interesting aspect of this specimen regards CuA + CuPaα. This stem is easily identified in both wings as being located posterior to MP, and anterior to CuPaα (itself easily located based on the location of its point of origin, similar in both wings). The point is that CuA + CuPaα is forked in the left forewing (Fig. 12A,C) and simple in the right one (Fig. 12B,D). The former case in uncommon in this species and was observed in females only (Table 1). This specimen validates the transformation assumed to have occurred between patterns represented on Fig. 10B and Fig. 10C (namely, CuA + CuPaα becoming simple): apart from the brief MA — MP reunification, its left and right forewings are schematized in Fig. 10J and Fig. 10K, respectively.
The species could be surveyed based on a large sample (Appendix 4), a selection of which is represented in Figs 13–14. A trait that significantly varies is the connection of CuA with CuPaα. The usual organization is represented on Fig. 13G,J, with a short free part of CuA. A complete series ranging from a long free CuA (Fig. 13A,D,M) to a long M + CuA + CuPaα fusion (longest fusion represented on Fig. 131,L,Q) was documented in both males and females (Fig. 13; Appendix 4).
Unusual morphologies are documented in Fig. 14. Although the right forewing of the specimen ANIC IWC OB 20 possesses the usual CuPaα stem fusing with CuA (emerging from M + CuA; Fig. 13A,D,M),its left forewing has no such free CuPaα stem (Fig. 14A–B, K). Therefore it is assumed that CuPaα is translocated onto M + CuA. The left forewing of the specimen ANIC IWC OB 17 (Fig. 14D,F,L) has a MA fused with R opposite the fusion of MP and CuA with CuPaα. The fusion of MA with the radial system continues on along RP; at some point MA diverges from RP, shortly fuses with MP, and recovers its usual route. The right forewing of the same specimen (Fig. 14C, E) has the usual MA distinct from R and RP. Finally, the specimen ANICIWC OB 21 has both forewings with a branched CuA + CuPaα (Fig. 14G–J; already observed in the specimen ANIC IWC OB 19, Fig. 12A,C), and the left forewing has a short RP + MA fusion (Fig. 14H,J).
Instead of having a stem of R straight, as in species surveyed above, the hind wings of pinguipes possess a R stem posteriorly bent basal to the point of divergence of RP (Fig. 15). Another peculiar aspect regards M. A fusion of a ‘median element’ with RP is evident (Fig. 15C). Posterior to this ‘median element’ occurs the genuine CuA + CuPaα, identified based on its distal fusion with CuPaβ. Whether the ‘median element’ is MA, or M, will be demonstrated based on the next case. Suffices to notice that (1) the free part of the ‘median element’ is simple (genuine RP branches are located very distally from its origin), and that (2) two intercalary veins occur between the free part of the ‘median element’ and CuA + CuPaα + CuPaβ (unlike in previous species, in which a single intercalary vein occurs). Therefore two plausible options are to be considered: (a) a fusion of M with RP occurs (i.e., MA and MP stay fused); or (b) a fusion of RP with MA occurs, and MP runs together with CuA + CuPaα, as in forewings (but does not diverge distally).
With respect to previous species, a peculiar trait is the organization of RA branches: in both fore- and hind wings, they arise basal to the first RP fork, and run parallel and close to the main RA stem for some distance.
The corresponding single specimen was selected for differences observed among forewings, and for hind wing venation, allowing that observed in pinguipes to be interpreted. Both species are closely related, owing to the occurrence of a MP + CuA + CuPaα stem, of a simple CuA + CuPaα in forewings, and to the organization of RA branches in both wing pairs (viz. delimiting very narrow areas).
Although the right forewing possesses a usual free part of CuPaα fusing with M + CuA (Fig. 16A,D,G), this portion does not occur in the left forewing (Fig. 16B,E,H). It is assumed that CuPaα is translocated onto M + CuA from the wing base, a transformation already conjectured in a specimen of pinguipes (Fig. 14A–B,K).
Regarding hind wings, as in pinguipes, a fusion of a ‘median element’ with RP was observed (Fig.14C,F,I). In addition, as in pinguipes, two intercalary veins occur between the first posterior branch of this composite stem and CuA + CuPaα (itself identified based on its distal fusion with CuPaβ). However, unlike in pinguipes, two ‘median elements’ diverge posteriorly basal to genuine RP branches (themselves arising distally). These two ‘elements’ are likely to be MA and MP. Therefore MA and MP are located between RP and its branches, and an area with two intercalary veins. Applied to pinguipes, it implies that the simple ‘median element’ is an undivided M [i.e., the above-cited conjecture (a) is correct]. It is not excluded that the morphology observed in the specimens ANIC IWC OB 25 is unusual for the corresponding species.
The usual organization of M, CuA, and of branches of CuPa in forewings of punctipennis is represented on Fig. 17A,E, and is schematized on Fig. 10D. As aptly recognized by Karny (1910), Zeuner (1939), and Ragge (1955), MA (‘M’ of previous authors) is fused with R at the wing base (Fig. 10H,L–N). It is the most plausible conjecture for the occurrence of a single stem basal to the RA / RP fork. Although tracheae were uneasy to observe in material of punctipennis, the stem of ‘R’ is provided with two evident tracheae in forewings of rufovaria (Fig. 19E), viz. R and MA.
Several conjectures are possible forthe remaining medio-cubital elements. According to Ragge (1955: fig. 54), M is simple and no fusion of ‘MP’ (here CuA) occurs, and ‘Cu1a’ (here CuPaα) and ‘Cu1b’ (here CuPaβ) diverge distally. However, this option is balanced in the text (pp. 76–77), in which this author suggested that ‘MP’ (CuA) could be fused with ‘Cu1’ (CuPa). This option is schematized on Fig. 10H. An issue with this interpretation is that it can hardly account for the occurrence of the pattern observed in the left forewing of the specimen ANIC IWC OB 37 (Fig. 17B,F; schematized on Fig. 10L): it would necessitate a branched ‘Cu1b’ (here CuPaβ) (but this vein is simple in all orthopterans), or a translocation of ‘Cu1a’ (here CuPaα) onto ‘Cu1b’ (here CuPaβ). A more plausible option is based on one of the THC proposed for pinguipes and schematized on Fig. 10K), implying the occurrence of a common stem MP + CuA + CuPaα which splits into MP and a simple CuA + CuPaα. Applying this pattern to the left forewing of the specimen ANIC IWC OB 37 (Fig. 17B, F) leads to the interpretation schematized on Fig. 10L: in both cases (Fig. 10K, L), posterior to MA successively occur the simple MP, then the composite (and simple) CuA + CuPaα stem, and then CuPaβ. On Fig. 10L, if traced backwards, the course of these elements reveals a composite stem CuA + CuPaα + CuPaβ, and a more basal composite stem MP + CuA + CuPa (CuPa or CuPaα + CuPaβ). No evidence was found for a ‘composite anterior stem’ [i.e., composed of MP and other ‘elements’, as suggested by Ragge (1955)]. Therefore it is assumed that in the usual forewing morphology of punctipennis MP + CuA + CuPa / CuPaα + CuPaβ divides into MP and a simple CuA + CuPa / CuPaα + CuPaβ (Fig. 10M). Among specimens of punctipennis, a single forewing was observed with an undivided composite MP + CuA + CuPa / CuPaα + CuPaβ stem (Fig. 17D,H), but it is not uncommon to observe the first fork of MP + CuA + CuPa / CuPaα + CuPaβ in a distal position (Fig. 17C,G). One specimen was observed with MP diverging from MA (hence M; Appendix 4). Very probably CuPa does not actually forks into CuPaα and CuPaβ in the usual morphology.
Limited variation was observed in the number of RP branches (♂, 11 forewings with 2-branched RP, 7 with 3-branched RP; ♀, 4 forewings with 2-branched RP, 8 with 3-branched RP). The organization of ScA is peculiar in that branches of the main stem originate at the wing base, resulting in a seemingly simple ScA complemented by branches of another vein / vein sector anterior to it. However, all these branches clearly belong to ScA.
As for hind wings, a fusion of M with R (instead of with RP as in pinguipes) occurs near the base in punctipennis (Figs 171, 18). Then M diverges from R but usually re-fuses with RP, near the origin of the latter, for some distance (Fig. 18B–D). This second fusion is very variable, ranging from ‘fusion absent’ (Fig. 18A) to a full fusion of M with R and RP, with a distal free part of M only (Fig. 18E–F). The length of the M + RP fusion itself varies widely (Fig. 18B–D).
Wing venation of this species (Fig. 19) is very similar to that of punctipennis. It was investigated primarily because tracheation was well-discernible in the available material. In forewings the MA trachea running ‘inside’ the radial vein is evident (Fig. 19E). Interestingly R and MA tracheae are entwined, with MA taking an anterior position with respect to RP for some distance. Another interesting aspect regards MP + CuA + CuPa / CuPaα + CuPaβ. This structure, genuinely composite, is provided with a single trachea.
Summary and comment
The comparative analysis demonstrates that the ‘classical’ STHC (Fig. 10E–H), which basically assumes a simple M (starting from the pattern represented on Fig. 10F), does not fit the data as well as the alternative one (Fig. 10I–N). In particular, the pattern represented on Fig. 10E, which is needed to complete the ‘classical’ series, has never been actually observed. In addition the alternative STHC accommodates more satisfactorily the observed variants, such as those represented on Figs 11, 12, 17B,F. In particular the wing venation pattern represented on Fig. 17B,F (and schematized on Fig. 10L) cannot be satisfactorily explained if one assumes a simple M. In order to facilitate comparison with actual material, the favored THCs are summarized on Fig. 20 (including interpretation for †perfecta Sharov, 1968; see below).
An oblique and comparatively strong cross-vein-like structure was observed between R and M + CuA in various specimens (* on Figs 12G, 14K, 19E). It has never been observed in Palaeozoic orthopterans and therefore is supernumerary. Its nature is unknown.
Taxonomic and nomenclatural implications
Taxonomy and nomenclature of several fossil species and taxa only remotely related to extant king crickets, raspy crickets & weta are considered first, It can be mentioned here that material yielded by the Dzaylyaucho locality [Madygen Formation, Kyrgyzstan; Ladinian/ Carnian, late Middle to early Late Triassic, according to Dobruskina (1995)] was deformed during fossilization (Rasnitsyn 1982; Sharov 1968, 1971; Figs 21, 27C,I). Based on preliminary investigations on material from this locality (in prep.), a plastic elongation of 150% best explains the observed aspect ratios. Species involved by such deformation are †abscissa Gorochov, 1987a, †triassica Sharov, 1968: 168, †perfecta, and †orientalis Sharov, 1968 (and species considered as junior synonyms of these in the following).
Taxon Archaeorthoptera nom. Béthoux & Nel, 2002a,
dis.-typ. Béthoux, 2007d
Species †abscissa Gorochov, 1987a
Discussion.—This species is assigned by Gorochov (1987a) to his proparagryllacridinaean assemblage, to which belongs †perfecta, revised below. The species †abscissa is treated separately from members of the taxon Agryllacris (as defined below, and including †perfecta) due to the following. It is based on single, fragmentary, and only moderately well-preserved specimen (Fig. 22). Differences with †perfecta (and presumed closely related species) are substantial: according to Gorochov (1987a), it can be characterized by (1) small size, (2) ScP ending not far from wing apex, (3) RA 2-branched, (4) CuA + CuPaα 3-branched (instead of 2 in proparagryllacridinaean), (5) loss of veins ‘parallelization’. Regarding the character state (4), it must be mentioned that the wing base is not preserved, therefore it cannot be excluded that CuA + CuPaα was provided with more than 3 branches.
It is unclear why Gorochov (1987a) considers this species as closely related to †perfecta. Several hagloidaean species collected from the same locality possess a simple MA (Gorochov 1986: figs 16, 19–23, 25–27, 29, 32–33, 36–38, 43–44), and although males of this assemblage do have MA and MP strongly bent in their proximal part (unlike in †abscissa), females do not (Sharov 1968: fig. 20B, D; Gorochov 1986: figs 32–34; see also Jurassic material illustrated by Sharov 1968). Assignment of the holotype of †abscissa to a female of one of the hagloidean species reported from Madygen is consistent with the characters listed above: for example, the holotype of saussurei Gorochov 1986 (Gorochov 1986: fig. 21) has a width similar to that of †abscissa (about 9.0 mm in the former, 7.1 mm in the latter) (1), ScP ending close to wing apex (2), RA with 2 to 3 branches (3), CuA + CuPaα with more than 2 branches (4), and non-parallel main veins (5). In addition, although the comparatively poor preservation does not allow a conclusive statement, the holotype of †abscissa exhibits ‘ScP veins’ that can be interpreted as secondary intercalary veins instead (* on Fig. 22), a trait proper to stem-hagloideans (Béthoux & Nel 2002a, Béthoux et al. 2002). In conclusion, †abscissa is better considered as a female of one of the numerous hagloidean species already described from the same locality. It will not be considered further.
Taxon †Bintoniellidae nom. Handlirsch, 1938,
dis. Sharov, 1968, typ. nov.
Definition.—Species that evolved from the metapopulation lineage in which the character state ‘in forewing, no distinct base of CuPaα diverging from CuPα (and fusing with CuA or M + CuA)’, as exhibited by †triassica Sharov, 1968: 168 and †brodiei Handlirsch, 1938, has been acquired.
Cladotypes.—Specimen PIN 2069/2320 (holotype of †triassica; Fig. 23A–B; Dzaylyaucho; Madygen Formation, Kyrgyzstan; Ladinian/ Carnian, late Middle to early Late Triassic), and NHM I.6784 [collected from the same locality as the holotype of †brodiei and assigned by Whalley (1982) to this species; Fig. 24C–E; Binton, Warwickshire, UK; Lower Jurassic].
Discussion.—The antonym of the defining character state is ‘in forewing, CuPaα with a distinct base diverging from CuPa, distal to the wing base (and fusing with CuA or M + CuA)’. This state is most likely plesiomorphic as it is exhibited by the vast majority of Palaeozoic Archaeorthoptera and saltatorians (Béthoux et al. 2012; Béthoux & Nel 2003, 2005a, b; Sharov 1968, 1971; among others). The corresponding taxon name was erected by Handlirsch (1938) for the species †brodiei only, known from abundant material (Figs 24–25). Later on Sharov (1968, 1971) assigned the species †triassica to this taxon. According to this author “the differences [of bintoniellidaeans with respect to vitimiidaeans] are only that the base of CuA1 [CuPaa] disappeared […] and became a crossvein so that the stem of MP + CuA1 [CuA+ CuPaa] branches from M”. Rather than assuming a ‘cross-vein-like’ transformation of CuPaα, I assume that this stem is translocated onto M + CuA, i.e., is fused with M + CuA from the wing base. Whether Sharov (1968) should be granted for associating a single diagnostic character state with the taxon name ‘Bintoniellidae’ is an interesting case. The formulation of the defining character state I propose leaves both options open (viz. cross-vein-like transformation vs translocation). It does not assume any particular transformation, and instead focuses on the condition observed in †triassica and †brodiei. The statement by Sharov (1968, 1971) is both a plain description of the observed condition, and a hypothesis on the transformation responsible for this condition. Provided that his statement fits, in part, the chosen character state formulation, I propose to grant Sharov for recognizing the defining character state of the taxon. Under cladotypic nomenclature, this makes the name ‘Bintoniellidae’ preoccupied (i.e., it has been previously associated with a single character state — presumably derived).
According to the provisional delimitation of †triassica (see below), the second available specimen of this species exhibits the antonym of the defining character state (Fig. 23C–D). However, I assume that this case accounts for the rare occurrence of a plesiomorphic condition, fitting within the range of usual intra-specific variability (see below).
Species †triassica Sharov, 1968: 168
Discussion.—The forewing of the holotype of †triassica lacks a distinct origin of CuPaα from CuPa (i.e., it exhibits a M + CuA + CuPaα basal stem; Fig. 23A–B). Such configuration was documented by Gorochov (1987a) in the species †crassa Gorochov, 1987a, †oblonga Gorochov, 1987a, and †angusta Gorochov, 1987a. All were collected from the same locality as the holotype of †triassica, and each is documented after a single specimen. These specimens were available to Sharov, who probably assigned them to ‘his’ †triassica [17 forewing imprints are reported by Sharov (1968, 1971) but without indication of specimen numbers]. The additional species erected by Gorochov (1987a) differ from †triassica by a few traits. Revision of the corresponding material was not carried out, and proper data is lacking to allow Gorochov's (1987a) delimitation to be debated. On the other hand, grounds for Gorochov's (1987a) delimitation was never made explicit. I provisionally and informally consider these species as junior synonyms of †triassica.
The holotype of the species †primaria differs from material mentioned above by the occurrence of a distinct CuPaα stem diverging from CuPa distal to the wing base (a plesiomorphic condition; Fig. 23 C–D). The survey of extant material carried out above demonstrates that the rare occurrence of a plesiomorphic condition is a common case at the species level. Therefore, assuming that †triassica is recorded thanks to four specimens exhibiting the derived condition (viz. the holotypes of †triassica, ‘†crassa’, ‘†oblonga’, and ‘†angusta’), it is reasonable to assume that the condition exhibited by this specimen merely accounts for intra-specific variation in †triassica. After all, it can merely be considered as an atavism. Observations made in laudatum and pinguipes of rare occurrence of a M + CuA + CuPaα (Fig. 3I,R, and Fig. 14A,B, respectively) mirror the assumption of a rare occurrence of the ancestral condition in †triassica (inclusive of †primaria).
Sharov (1968: 168; and Sharov 1971: 172) indicates that, in addition to the distinct CuPaα base, †primaria differs from †triassica by a “curved CuA2 [CuPaβ]”. However no difference was noted in that respect (Fig. 23). Provided that no other traits allows †primaria to be distinguished from †triassica (chiefly because of the very fragmentary state of the holotype of the former), I consider the former as a junior synonym of the latter.
Taxon Agryllacris nom.-dis.-typ. nov.
Etymology.—Part of ‘Tagryllacris’ (see below).
Definition.—Species that evolved from the metapopulation lineage in which the character state ‘in forewing, CuA + CuPaα with two distal branches only’, as exhibited by laudatum Johns, 1997 and carli Griffini, 1911a, has been acquired.
Discussion.—The antonym of the defining character state is ‘CuA + CuPaα with more, or less, than two distal branches’. The putative plesiomorphic condition is ‘CuA + CuPaα with more than two distal branches’: it is exhibited by the vast majority of Palaeozoic Archaeorthoptera and saltatorians (Béthoux & Nel 2002b, 2005a, b; Sharov 1968, 1971; among others, and Fig. 1). The two distal branches of CuA + CuPaα are conjectured as being CuA, and CuPaα, respectively. If so, and if assuming a simple CuA in †uralica, the character state ‘CuPaα simple’ (as opposed to ‘CuPaα branched’) is a synonym of the defining character state.
The morphology of carli Griffini, 1911a was investigated based on photographs of the holotype (Fig. 26A,B). Another specimen housed at Institute of Zoology, Chinese Academy of Sciences (Beijing, PR. of China; specimen number IOZ(E) 1343118), determined as carli, was observed. It exhibits the defining character state.
The possible importance of the defining character state to the systematics of the gryllacrididaeans seems to have been ignored by previous authors, in connection with the improper identification of fossil species. It must be acknowledged that this condition was most likely acquired convergently in elcanid orthopterans, and in †brodiei (see below), at least. A preliminary composition of Agryllacris is provided in Appendix 3, but several cases require a detailed discussion.
The species †perfecta is herein assigned to Agryllacris because it exhibits the defining character state of this taxon (Figs 20A, B, 27, 28). This view contrasts with that of Sharov (1968: 82–84; and Sharov 1971: 83–84) who excludes direct relationships of this species (and other proparagryllacrididaeans) with gryllacrididaeans, and instead considers the species as ‘Ensifera incertae sedis’ [Gorochov (1995a) assigns the corresponding species to his paraphyletic Palaeozoic and Triassic assemblage ‘Oedischioidea’]. Sharov (1968, 1971) argues that the five-segmented tarsus provided with an arolium which he describes for †perfecta (fig. 33D; and see Fig. 29A) excludes relationships with the gryllacrididaeans, possessing a 4-segmented tarsus without a broad arolium. However photographic data indicate that an alternative interpretation of the corresponding (moderately well preserved) specimen is equally plausible (Fig. 29B). Firstly, it is not evident that the arolium is preserved (or broad). Secondly, Sharov's (1968, 1971) 4th tarsal segment is not evident. Additionally, even if one adopts Sharov's (1968, 1971) interpretation, the states assumed by this author are plesiomorphic at the level of Archaeorthoptera [see Béthoux (2009a); and see Sharov (1968, 1971: figs 11D, 12E), representing 5-segmented tarsi in stem-orthopterans]. In other words, the hypothesis that †perfecta is a stem-gryllacrididaean is not contradicted by Sharov's (1968, 1971) interpretation.
Based on fossil imprints only, Sharov (1968, 1971) argues that “the wings formed a sheath around the body at rest” in †perfecta, and that the species “resemble in this respect the Triassic Phasmatodea on one hand and the Gryllacrididae on the other; however, it is also unjustified to place them into any of these two groups because of the method of wing folding”. However, how wing folding in †perfecta and gryllacrididaeans is similar but dissimilar is as obscure as the way to determine wing folding at rest based on fossil imprints of mostly isolated wings. Sharov (1968, 1971) further argues that, in †perfecta, “the costa [ScA] has displaced the proximal branches of Sc [ScP], but it does no cross them, retaining the same position as in the Oedischiidae […]”. A wide range of variation of this character was documented among the surveyed extant species, in particular at the intra-specific level. And indeed the condition formulated by Sharov (1968, 1971) was observed in ornata and rufovaria, among others. What Sharov (1968, 1971) means by “displacement of the beginning of CuA [CuPa?] almost to the middle of CuP [CuPb]”, assumed to be absent in gryllacrididaeans but present in †perfecta according to him, is obscure to me.
In summary Sharov's (1968, 1971) argument does not allow the hypothesis that †perfecta is a stem-gryllacrididaean to be discarded. It must be noticed that the defining character state of Agryllacris has been subjected to iterative acquisitions (see below). However, assignment of †perfecta to this taxon is also supported by the very long CuPb, reaching the posterior wing margin distal to the second third of wing length, which is a putative apomorphy. I found no character that could reject the hypothesis that †perfecta belongs to Agryllacris.
Data on a single forewing of the Triassic species †giganteus Riek, 1955 is available (Riek 1955: fig. 31). This species has forked M and CuA + CuPaα, as in †perfecta. Similarities between the two species extend to the wing shape and size too. Based on this evidence I propose to assign †giganteus to Agryllacris. This proposition complies with Riek (1955), who considered the corresponding species as a gryllacrididaean.
The species †orientalis Sharov, 1968 is also assigned to Agryllacris because it exhibits the defining character state of this taxon (Sharov 1968, 1971:fig. 18C). This species, assigned to the genus ‘Madygenia’ by Sharov (1968, 1971), is considered as a ‘stem-tettigonid’ by this author, but this option is discarded by Gorochov (1995a, b), who considers it as closely related to †perfecta (i.e., as a stem-orthopteran). I concur because (1) in addition to the defining character state of Agryllacris, †orientalis also has a simple MA, and (2) ground for Sharov's (1968, 1971) hypothesis is virtually non-existent. This author (p. 47, p. 45, respectively) argues that “other characters of the venation show that Madygenia deviates from the line of evolution to the Tettigoniidae: RS [RP] has lost its pectinate form, MP + CuA1 [CuA + CuPaa] has retained only two long veins; the base of CuA [CuPa] has been lost in the hind wings, while CuA2 [CuPab] became fused with MP + CuA1 [CuA + CuPaa], forming a single vein”. However (1) Sharov (1968, 1971) illustrates tettigonids with a pectinate RP in forewing (fig. 19E,H); (2) CuA + CuPaα is two branched in forewings of †perfecta, †giganteus, and laudatum (among others), and Sharov (1968, 1971) illustrates a tettigonid with a CuA + CuPaα with more than 2 branches in forewing (fig. 19E); (3) above I discuss the case of CuPa in hind wings of surveyed extant species and demonstrate that it conforms to the statement made by Sharov (1968, 1971). Finally, Sharov (1968, 1971) advocates the occurrence of a secondary archedictyon. However the tettigonid archedictyon is made of very small cells, themselves composing larger cells delimited by stronger cross-veins, while cells in †orientalis are larger and do not compose larger cells. It suggests that these ‘archedictyons’ are not homologous.
Gorochov (1987a) erected a number of species assumed to be closely related to †orientalis, based on material collected from the same locality as †orientalis. These species are †extremalis Gorochov, 1987a, †ovalis Gorochov, 1987a, †longissima Gorochov, 1987a, and †grandis Gorochov, 1987a. As for †triassica-like species, each is documented after a single specimen, belonging to the set of specimens available to Sharov, and that this author probably assigned to †orientalis [39 imprints are reported by Sharov (1968, 1971) but without indication of specimen numbers]. The additional species differ from †orientalis by a few traits which could fit within an intraspecific range of variation. Revision of the corresponding material was not carried out, therefore I consider these species as junior synonyms of †orientalis on a provisional and informal ground.
A number of poorly documented fossil species have been considered at some point as gryllacrididaeans. Zeuner (1939), fol- lowed by Sharov (1968, 1971), considers scotica Zeuner, 1939 as a gryllacrididaean [and Gorochov (1995a) as a stenopelmatoidean]. However, the only available forewing of this species exhibits a ScA that seemingly crosses ScP branches (Zeuner 1939: pl. 72, fig. 5; Sharov 1968, 1971: fig. 27A; contra Zeuner 1939: pl. 46, fig. 5), a trait diagnostic of prophalangopsidaeans and absent in gryllacrididaeans (at least all those I surveyed; contra Sharov 1968, 1971). Also a CuA + CuPaα with more than 2 branches does not occur in gryllacrididaeans, but is a usual condition in prophalangopsidaeans. Indeed comparison with the forewing of the presumed female of obscura Walker, 1869 (Liu et al. 2009: fig. 1) strongly suggests that scotica is not a gryllacrididaean but a female of a prophalangpseidaean. The same applies to the species maculata Cockerell, 1908, considered by Zeuner (1939) as a gryllacrididaean (data on the species are minimal due to the very fragmentary state of the only available specimen).
I concur with Gorochov (1995a) and consider gryllacroides Zeuner, 1937 as a hagloidean incertae sedis. A revision of the material of ungeri Heer, 1849 would be needed before it could be conclusively assigned to Agryllacris, because the available illustrations (e.g., in Zeuner 1937) are insufficient. The assignment of †crassifemur Riek, 1956 to the gryllacrididaeans by Riek (1956) must also be discussed [the species is considered as a proparagryllacrididaean by Sharov (1968, 1971)]. The wing venation interpretation proposed by Riek (1956: fig. 2) is problematic: if one wants to make Riek's drawing fit a typical Agryllacris wing venation, ‘Sc’ should be the stem of R (Gorochov 1995a: fig. 222). However, ‘Sc’ / ScP and R are strongly concave, and convex, respectively, in all Archaeorthoptera, and can hardly be mistaken. Due to this inconsistency, and to the fragmentary state of the corresponding material, I provisionally exclude †crassifemur from Agryllacris.
Except for a single specimen (Fig. 24N–P), forewings of the species †brodiei exhibit the defining character state of Agryllacris (Figs 24, 25A). In addition, as in Agryllacris, MA and MP are both simple. However †brodiei (1) has a short ScA (it is comparatively longer in Agryllacris), (2) a constant width of the area between CuPb and AA1 (it is broadened distally in Agryllacris), (3) and CuPb reaching the posterior wing margin opposite wing mid-length (distal to the second third in Agryllacris). In addition †brodiei shares with †triassica (4) a CuPaα translocated onto M + CuA, and (5) a pterostigma-like structure in fore-and hind wings (Figs 23, 25; lacking in Agryllacris), indicating close relationships of these two species. Provided that †triassica has a multi-branched MA and CuA + CuPaα, the simple MA and the 2-branched CuA + CuPaα observed in †brodiei is considered as acquired independently from Agryllacris.
Species †perfecta Sharov, 1968
Figs 20A, B, 27–29
Discussion.—Among species listed by Gorochov (1987b) as belonging to the subfamily “Proparagryllacridinae’, the species listed above, in addition to †abscissa Gorochov, 1987, were all collected from the Dzaylyaucho locality. Each of the species erected by Gorochov (1987b) is documented based on a single specimen. Data collected from fossil and extant material will be used to reconsider Gorochov's (1987b) species delimitations.
Gorochov (1987b) assigned the species †megaptera Gorochov, 1987b to the monotypic genus ‘†Batkenella’ Gorochov, 1987b. This author does not provide a differential diagnosis of †megaptera, but of ‘†Batkenella’. According to this diagnosis, the species †megaptera can be differentiated from other proparagryllacridinaeans after (1) a broad forewing, (2) a broad area between anterior wing margin and ScP, (3) long RA/ RP and MA/ MP areas, and (4) a basal position of the point of divergence of MA and MP. However the width of †megaptera forewing (Fig. 28C,F) does not differ significantly from that of the specimen PIN 2240/4112 (Fig. 28B,E), paratype of †perfecta. Provided a possible width alteration of 150% due to a plastic deformation during fossilization, the width of the area between the anterior wing margin and ScP is to be considered as similar to that exhibited by the specimen PIN 2240/4132 (Fig. 27A,G), holotype of † perfecta. The RA / RP area actually is shorter in † megaptera than in the specimen PIN 2240/4112 (Fig. 28B,E).The length of the MA/ MP area relates to the point of divergence of MA and MP, supposedly comparatively basal. However the length of M (from the divergence of CuA to its first fork) represents 41.2% of the distance between the point of divergence of CuA and the point of origin of RP in † megaptera, while is represents 43.9% in the holotype of †perfecta (Fig. 27A, G). Such difference is to be considered as insignificant, as demonstrated by the variation of this trait documented in laudatum (Fig. 3). The forewing pair of the specimen PIN 2240/4113 [holotype of †madygenioides Gorochov, 1987a, below considered as a junior synonym of †perfecta] also demonstrates that the character is variable in this (set of) species: in this specimen the left forewing has a comparatively basal point of divergence of MA/ MP, but this point is located distal to the origin of RP in the right forewing.
According to Gorochov ( 1987b) the species †simplicis Gorochov, 1987b is characterized by (1) a short forewing, (2) a short ScA, (3) a broad area between anterior wing margin and ScP, and (4) CuP and AA branches forming an angle with the wing longitudinal axis. The forewing length of the holotype of †simplicis (Fig. 27G,M) represents about 83% of that of the holotype of †perfecta (Fig. 27A,G) (length of about 30.2 mm, and 36.6 mm, respectively). This ratio falls within the intra-specific variation observed in extant species (Appendix 4; not even taking sexual dimorphism and deformation into consideration). The vein ScA reaches the anterior wing margin about 11.4 mm distal to wing base in the holotype of †perfecta, 10.7 mm in †simplicis. Reported to the wing length, ScA is comparatively longerin †simplicis (about 35.4%) than in †perfecta (about 32.8%), contradicting Gorochov's (1987b) 2nd character (state). In †perfecta the width of the area between the anterior wing margin and ScP represents about 23.9% of wing width, about 23.8% in †simplicis. What Gorochov (1987b) refers to as a comparatively ‘broad area between anterior wing margin and ScP’ (3) is therefore unclear. Explanation is provided by Gorochov (1987b) regarding the character (4) (p. 25): “in G. perfecta CuP and A [AA] are located almost parallel to the wing longitudinal axis”. Therefore one must be able to distinguish veins forming ‘an undetermined angle with wing longitudinal axis’ from ‘veins almost parallel to wing longitudinal axis’ in order to diagnose †simplicis. Unfortunately this does not fall within the range of my capacities [and see below; in addition a wing elongated along its longitudinal axis (e.g., Figs 27B–C,F,H– I, L) will exhibit veins forming a comparatively lower angle with the longitudinal axis].
According to Gorochov (1987b) the species †elongata Gorochov, 1987b is characterized by (1) a narrow inter-radial (RA/ RP?) area, (2) the position of MP close to the branching point of CuA + CuPaα, (3) numerous branches of RA, (4) narrow area between anterior wing margin and ScA, (5) long ScA, and (6) orientation of CuP and AA branches intermediate between †perfecta and †simplicis. However, the width of the RA/ RP area represents about 12.3% of wing width in the holotype of perfecta (Fig. 27A,G), 12.9% in the holotype of †simplicis (Fig. 27DJ) (measurements taken at the mid-length of RP from its origin to its first fork). These ratios, independent of possible deformation, indicate that the character (1) is not species relevant. Variation in the proximity of MP and CuA near their point of divergence (from M, and from CuA + CuPaα, respectively; 2) was observed as intra-individual variation in the specimen ANIC IWC OB 26 (Fig. 2A–B,D–E; with an actual short fusion of MP with CuA). In addition, in the specimen ANICIWC OB 24, MP is distinct from CuA in the left forewing (Fig. 11A,C,E,G), but fuses with it for a short distance in the right forewing (Fig. 11B,D,F,H). This represents a variation as substantial as that observed in †simplicis as compared to †perfecta. The formulation of the character (3) (viz. ‘numerous branches of RA’) is enigmatic, because †simplicis has 2 branches of RA, and the holotype of †perfecta 3. The correct formulation was probably ‘fewer branches of RA’ rather than ‘numerous branches of RA’. However, a variation of 2 to 3 branches was observed within the forewing pair of a single specimen of cf. mexicanus (ANIC IWC OB 26; Fig. 2A–B,D–E), and similar range of variation (4 to 5) was documented in material of pinguipes (Fig. 14) (among others). The character state (4) is impossible to appreciate simply because the corresponding area is not preserved on the holotype (Gorochov 1987b: fig. 7; Fig. 27J). Similarly, the length of ScA (5) is difficult to appreciate in this context, because ScA forms a very narrow angle with the anterior wing margin near its end. Without proper preservation of the area between the anterior wing margin and ScA, the length of the latter is therefore impossible to measure with accuracy. It has been mentioned above that the orientation of CuP and AA branches with respect to the wing longitudinal axis was impossible to distinguish between †perfecta and †similis. That †elongata is supposed to exhibit an intermediate condition reinforces the opinion that this character is not useful for discriminating the different species recognized by Gorochov (1987b).
According to Gorochov (1987b) the species †madygenioides Gorochov, 1987b is characterized by (1) the location of the bases of RP, MP, and the branching point of CuA + CuPaα, as in †perfecta and †simplicis but different from that observed in †elongata, and (2) CuPaα fused with M + CuA (as a consequence, basal to the divergence of CuA from M + CuA). Apparently this diagnosis is only based on the left forewing of the holotype of †madygenioides (Gorochov 1987b: fig. 8; to be compared with Fig. 27C,I). It is quite unfortunate that the right forewing was not given full consideration by Gorochov (1987b) because it shows that the location of the base of MP is variable in this individual: indeed it is located distal to the origin of RP in this wing. Therefore the holotype of †madygenoides shows that the character (1 ) varies more widely within a single individual than among species as delimited by Gorochov (1987b). Regarding the fusion of CuPaα with CuA vs M + CuA, both states were documented in laudatum (Fig. 4), cf. bicornis (Fig. 8), and pinguipes (Fig. 13).
Gorochov (1987b) assigned the species †perlonga Gorochov, 1987b to the monotypic genus ‘†Dolichobatkenella’ Gorochov, 1987b. This author does not provide a differential diagnosis of †perlonga, but of ‘†Dolichobatkenella’. According to this diagnosis, †perionga can be differentiated from other proparagryllacridinaean taxa after (1) narrow and elongated forewing, and (2) CuPaα translocated onto CuPaβ. As mentioned above characters relating to wing aspect ratios (1) are better ignored for material provided by the Dzaylyaucho locality, because of plastic deformation that affected the material. Regarding the character (2), unusual translocations were documented in laudatum (translocation of the anterior branch of RP onto RA, see Fig. 3H,Q; translocation of CuPaα onto M + CuA, see Fig. 3I,R), ornata (in particular branches of RP onto RA, see Fig. 6), and pinguipes (translocation of CuPaα onto M + CuA, Fig. 14A,B, K). Therefore it cannot be excluded that †perlonga is merely a variant of †perfecta.
Gorochov (1987b) assigned the species †devexa Gorochov, 1987b to the monotypic genus ‘†Eubatkenella’ Gorochov, 1987b. This author does not provide a differential diagnosis of †devexa, but of ‘†Eubatkenella’. According to this diagnosis, †devexa can be differentiated from other proparagryllacridinaean taxa after (1) oblique branches of ScP, and (2) translocation of MP onto CuA + CuPaα. Regarding the character state (1), the corresponding area is so poorly preserved (Fig. 27K; if preserved at all) that the character cannot be properly appreciated. Regarding the character state (2), such rare translocations are considered trivial for delimiting orthopteran species, because they occur repeatedly as rare variations in extant species (see above).
Data on fossil and extant material indicate that characters used by Gorochov (1987b) to delimit a number of species from the same locality as †perfecta could not be observed on the actual specimens, fall within the range of intra-specific variability, and/or likely are the consequence of deformation that affected the fossil material. In summary the corresponding species are to be considered as junior synonyms of †perfecta.
At this step it must be noticed that Sharov (1968, 1971) reports “7 imprints of body with wings, 65 imprints of wings of males, 12 imprints of fore wings of females and one imprint of a hind wing from the same locality”. Unfortunately, apart from specimens used for his restoration (fig. 33), specimen numbers were not provided. However, the first part of PIN numbers refers to specific collections, and the numbers 2240 and 2555 indicate collections supervized by Sharov himself in 1964 and 1965, respectively (Shcherbakov 2008, and A.P. Rasnitysn, pers. com. to OB, 2011). Therefore all specimens used by Gorochov (1987b) to erect ‘ †perfecta-like’ species were available to Sharov in 1968, and there is little doubt that they belong to the set of specimens determined as †perfecta by himself. If so I concur with Sharov's (1968, 1971) species delimitation.
Taxon Tagryllacris nom.-dis.-typ. nov.
Etymology.—Part of ‘Etagryllacris’ (see below).
Definition.—Species that evolved from the metapopulation lineage in which the character state ‘in forewing, M + CuA splits into MA and MP + CuA’, as exhibited by pinguipes Rentz in Morton & Rentz, 1983 and magnifica Brunner von Wattenwyl, 1888, has been acquired.
Cladotypes.—Paratype of pinguipes Rentz in Morton & Rentz 1983 (ANIC ethanol collection; label reading: 26.00S 131.25E, 26km WSW. of Milga Park, H.S., N.T., 18.i. 1982), andholotype of magnifica Brunner von Wattenwyl, 1888 (housed at the MNHG; Fig. 26C,D).
Discussion.—The antonym of the defining character state is ‘in forewing, M + CuA splits into M and CuA’. This condition was documented in several species of the taxon Agryllacris (Figs 2–4, 6), and occurs in all Palaeozoic saltatorians (Béthoux & Nel 2002b, Carpenter 1992, Sharov 1968, Sharov 1971, among others). Therefore the defining character state of Etagryllacris likely is derived.
The same character state defines the Gigatitanidae nom. Sharov 1968, dis.-typ. Béthoux 2007d. The two taxa can be distinguished based on numerous apomorphies, on each side. It is not questionable that the same defining character state was acquired independently in Tagryllacris and Gigatitanidae (i.e., is a parallelism s.l. as defined below).
The morphology of magnifica was investigated based on photographs of the holotype (Fig. 26C,D), and photographs of ANIC (drawers containing) dry specimens of this species, and of closely related ones. The usual organization of the median and cubital systems in forewings conforms to that observed in the forewings of the holotype specimen, viz. exhibits a MP fused with the free part of CuA and with CuA + CuPaα (and then diverging), and a CuA + CuPaα being simple.
Taxon Etagryllacris nom.-dis.-typ. nov.
Etymology.—Part of ‘Metagryllacris’ (see below).
Definition. —Species that evolved from the metapopulation lineage in which the character state ‘in forewing, CuA + CuPaα keep fused’, as exhibited by pinguipes Rentz in Morton & Rentz, 1983 and magnifica Brunner von Wattenwyl, 1888, has been acquired.
Cladotypes.—Paratype of pinguipes Rentz in Morton & Rentz, 1983 (ANIC ethanol collection; label reading: 26.00S 131.25E, 26km WSW. of Milga Park, H. S., N.T., 18.i 1982), and holotype of magnifica Brunner von Wattenwyl, 1888 (housed at the MNHG; Fig. 26C,D).
Discussion.—A synonym of the defining character state is ‘CuA + CuPaα simple’. The antonym of the defining character state is ‘in forewing, CuA and CuPaα diverge distally’. Distal divergence of CuA and CuPaα was documented in other species of the taxon Agryllacris (Figs 2–4, 6), and in all Palaeozoic saltatorians (Béthoux & Nel 2002b, Carpenter 1992, Sharov 1968, Sharov 1971, among others). Therefore the defining character state of Etagryllacris likely is derived. A branched CuA + CuPaα occurs only rarely in pinguipes (Appendix 4).
Thespecies bicornis exhibits the defining characterstate of Etagryllacris (Fig. 8). However I propose to exclude this species from this taxon primarily because it lacks several character states shared by Tagryllacris species (Fig. 30). In other words, I consider the occurrence of a simple CuA + CuPaα in bicornis as the result of homoplasic evolution. Note that bicornis is not to be automatically assigned to Etagryllacris because it exhibits the defining character state of this taxon: Etagryllacris is the lineage in which the defining character as observed in pinguipes and punctipennis was acquired and, according to the proposed phylogenetic scheme, bicornis did not derive from this lineage.
Taxon Metagryllacris nom. nov., dis. Zeuner, 1939, typ. nov.
Etymology.—From ‘meta’, change, transition, in Greek, and ‘Gryllacris’, a taxon name commonly used under the Linnaean procedure; the name ‘Metagryllacris’ accounts for the adaptation of ‘Gryllacris’ under the cladotypic procedure.
Definition.— Species that evolved from the metapopulation lineage in which the character state ‘in forewing, MA fused with R at wing base’, as exhibited by punctipennis Walker 1869, and rufovaria Kirby, 1888, has been acquired.
Discussion. —The defining character state was first outlined by Karny (1910: 38; Karny's ‘M’ is MA; and see Karny 1930) as typical of his ‘Typus IV’ (also present in his ‘Typus V’, presumable derived from ‘Typus IV’). However he also listed “Radius sector [RP] less developed than in Type I, obviously reduced” in addition to the defining character state of Metagryllacris. Zeuner (1939: 60), acknowledging Karny, recognized the same character as the single trait proper to a single group, referred to as ‘B’ in his determination key (and including Karny's Types IV and V). Based on his examination of an undetermined species of the genus ‘Gryllacris’, Ragge (1955) also recognizes this trait as proper to (a subset of) gryllacridids. It turns out that Zeuner (1939) is the first author to associate the defining character state only with a single lineage (‘B’), and his therefore granted priority for the characterization of the group.
Adaptation of the taxon name ‘Gryllacris’, which is commonly used under the Linnaean procedure, could have appeared as a suitable option for this taxon. However several ‘Gryllacris’ species were assigned by Karny (1930) to ‘Types’ other than IV and V. In contrast this taxon name has a more restricted composition nowadays, excluding some species belonging to the taxon Metagryllacris (Eades et al. 2012). The taxon name ‘Gryllacris’ is therefore highly polysemic under the Linnaean procedure, and therefore was avoided. With this provision I propose to coin the new taxon name Metagryllacris.
Origin and evolution.—According to Dobruskina (1995; followed by Shcherbakov 2008), the Dzaylyaucho locality, from which the material of †perfecta and †orientalis (and ‘†orientalis-like’ species) has been unearthed, is Ladinian-Carnian in age (late Middle to early Late Triassic). Therefore the assignment of these species to the taxon Agryllacris implies that this lineage diverged ca 235 million years ago (mya) at least. This is in sharp contrast with the prevalent view (Heads & Leuzinger 2011), predicting a divergence of king crickets, raspy crickets & weta (forming a monophyletic group, or not) ca 65 mya (at best).
The new proposition has direct implications on scenarios regarding the lack of the Grylloptera stridulatory apparatus in extant king crickets, raspy crickets & weta. The earliest species showing such a stridulatory apparatus were collected from the same locality as †perfecta and †orientalis (Béthoux 2012; Gorochov, 1986; Sharov 1968, 1971). According to Béthoux (2012) the file, typical of this apparatus, was acquired once. The point is to determine whether Agryllacris belongs to Grylloptera (i.e., if Agryllacris derived from a lineage possessing the Grylloptera file), or not (if Agryllacris does not derive from a lineage possessing the Grylloptera file).
Different degrees of complexity in stridulatory apparatuses are documented in Grylloptera from Dzaylyaucho (Béthoux 2012; Gorochov 1986; Sharov 1968, 1971). In other words, by the time of existence of †perfecta and †orientalis, the Grylloptera stridulatory apparatus was undergoing its early evolution. Therefore it is unparsimonious to assume that †perfecta and †orientalis could have already gained and lost the corresponding organization. Indeed Gorochov (1995a, b) roots †perfecta and †orientalis in the ‘oedischioid wastebasket’, composed of stem-orthopterans that never gained the Grylloptera stridulatory apparatus. I concur with this view: indeed the ‘oedischioid’ †buttsi Béthoux & Beckemeyer, 2007 and †noblensis Béthoux & Beckemeyer, 2007, possessing forewings with a low number of MA and CuA + CuPaα branches, and lacking the Grylloptera file, could well form the Early Permian ‘oedischioid’ stock from which Agryllacris derived.
Particular aspects of wing venation evolution.—In forewing, species of the taxon Metagryllacris exhibit a ScA with a long posterior stem, often simple, and separated from other ScA branches by broad area filled with strong cross-veins (Figs 17, 19). This single stem cannot be mistaken with ScP because it is strongly convex (additionally the genuine ScP is evident). In contrast, in other species of the taxon Agryllacris, ScA is composed of a single main stem regularly emitting anterior branches, more or less all similar. Without proper phylogenetic scheme, the anterior area in forewing of Metagryllacris species could have been interpreted as composed of a simple ScA, and of a CP area anterior to it. The occurrence of a Costal system to which CP would belong is hypothesized by Kukalová-Peck ( 1991; and references therein) in her insect ground plan. However observations by this author have been questioned. In particular the presumably plesiomorphic occurrence of a CP area in stem-Amphiesmenoptera, observed by Kukalová-Peck & Willmann (1990), was later discarded by Willmann (1997). Elcanid orthopterans provide the only convincing support for a presumed CP area: in the most recent representatives (see Sharov 1968, 1971: fig. 14H), three veins occur anterior to RA, which could be interpreted as ScP (Sharov's ‘Sc’), ScA (Sharov's C'), and CP (not labelled by Sharov) (notice that Carpenter dealt with this case by hypothesizing a bifid ScP; Carpenter 1992: fig. 99.7). However observations in Metagryllacris match the interpretation by Sharov (1968, 1971) of the elcanid pattern, with the anterior area being filled with a ScA provided with a simple and long posterior stem, and with another anterior stem. In other words no hypothetical CP occurs. These observations suggest that the occurrence of a Costal system in insect wings might have to be reconsidered.
The actual course and organization of tracheae within wing veins have long been recognized as a potential source of inference for establishing conjectures of topological homologies (Comstock & Needham 1898,1899).Typically, theobservation of a pair of tracheae running in a single ‘vein’ is indicative of a composite stem, such as R and MA in rufovaria, each represented by a trachea within the ‘R vein’ (Fig. 19E). However, the same material demonstrates that the corollary, namely that a single trachea is indicative a single vein, is not supported: the eminently composite stem MP + CuA + CuPa / CuPaα + CuPaβ is provided with a single trachea (Fig. 19E) in the same species. It implies that vein fusion is probably ‘completed’ by tracheal fusion at some point.
Variation, variability, and iterative evolution.—Variation in the variability of wing venation has been seldom investigated (but see Ross 2012; and references therein). The available data on Agryllacris demonstrate that the range of variation varies among the surveyed species (Tab. 1). For example, although the number of branches of CuA + CuPaα was found to be very stable in laudatum and bicomis, it varies within a limited range in pinquipes and punctipennis, and more widely in ornata. The character ‘width of the area with the free part of CuA / length of the M + CuA + CuPaα common stem’ is comparatively stable, and variable, in laudatum and pinquipes, respectively (Appendix 4).
The connection (or the lack thereof) of MP with CuA and CuA + CuPaα is also a relevant case. A connection is consistently absent in laudatum and ornata, while species of the taxon Tagryllacris usually exhibit a fusion of MP with CuA and CuA + CuPaα (as a consequence, there is not free stem of M). However, the left forewing of the specimen ANICIWC OB 24 (Fig. 11 A,C,E,G), itself belonging to the taxon Tagryllacris, exhibits the condition observed in laudatum and ornata. In addition a peculiar intermediate condition was observed in a single pinguipes specimen: the actual fusion of MP with CuA and CuA + CuPaα occurs, but MP further re-unites with MA (Fig. 12), reforming the stem of M observed in the earlier lineages. The fusion of MP with CuA and CuA + CuPaα occurs more consistently in species of the taxon Metagryllacris, without intermediate condition being observed. These examples indicate that variability possibly varies according to the phylogenetic position of a given species. Note that a similar variation in variability has been documented in the fusion of a branch of RP with RA in mantodeans (Béthoux & Wieland 2009), with occurrence of intermediate conditions in variable species.
Based on these examples, variation in character states distribution can be encompasses by three cases (Fig. 31). First, within a species exhibiting an apomorphic condition, rare variants exhibit the condition usually present in successive stem-lineages. This is the case of CuA + CuPaα being forked in pinguipes and punctipennis (Appendix 4), for example. These are atavisms, viz. the rare occurrence of a plesiomorphic condition. Although this concept usually applies to species only remotely related to the corresponding stem-lineages, it seems appropriate to use it even in species closely related to these (Fig. 31, sp. 8 in case 1): it can be argued that atavism borders on intra-specific variability indeed. The connection between atavism and reversal (or taxic atavism) is evident (Hall 2003; Stiassny 1992): both are the consequence of the persistence of functional genetic modules responsible for plesiomorphic states. The term ‘atavism’ refers to low frequency of occurrence of the plesiomorphic state in a given species (Fig. 31, sp. 10 in case 2), while reversal is characterized by a high frequency of occurrence (Fig. 31, sp. 11 in case 2).
The complementary configuration, viz. the rare occurrence of an apomorphic condition in a species usually exhibiting the plesiomorphic one, is equally plausible. Examples were documented above, such as the pectiante fusion of RP with RA, usually observed in ornata, and found to occur rarelyin laudatum (Fig. 3H,Q; and see Béthoux & Wieland 2009: fig. 13C–D). I failed to find the proper counterpart to the term ‘atavism’ for such case, therefore I propose to coin the term ‘pronatism’ (‘pro-’ = before; ‘natus’ = born). It is proposed to apply this concept only to lineages which derived directly from the corresponding stem-lineage (such as spp. 2 and 6 in case 1 on Fig. 31; see below for sp. 3; referred to as direct relative in the following). As for atavism, it is assumed that pronatism is based on a genetic organization similar to that underlying the high-frequency condition (but see below for a broader sense). And as for atavism and reversal, an evident connection exists between pronatism and parallelism (sensu Hall 2003): assuming similar genetic ground, the former is characterized by low frequency, the latter high frequency. And as for atavism, pronatism borders on intra-specific variability.
At this step confusion between parallelism on one hand, and atavism, reversal, and pronatism on the other, is possible. Hall (2003: tab. 2) defines parallelism as follows: “a feature present in closely related organisms but not present continuously in all members of the lineage” and which development is based on “normally similar developmental pathways”. However, this definition applies to atavism and reversal as well, because they are the re-expression of a genetic organization inherited from the same stem-lineage (therefore arguably similar), and of course to pronatism as defined above. I propose to give a narrower sense to parallelism, restricted to the repeated acquisition of a derived condition at high frequency, due to similar genetic ground. Atavism and reversal apply to expression of a plesiomorphic conditions.
We are left with a final case involving (1) a derived condition occurring at high frequency in a lineage, and (2) the same derived condition occurring at low frequency within a lineage which is not a direct relative of the former (Fig. 31, case 3, spp. 3 and 4 on one hand, and 7 and 11 on the other; and sp. 3 as opposed to spp. 7–11 in case 1). The translocation of CuPaα onto M + CuA fits this case: it occurs at low frequency in laudatum (Fig. 31, R), pinguipes (Fig. 14A–B, K), and in one wing of the specimen ANICIWC OB 25 (Fig. 16B,E,H), but at high frequency in †brodiei (Fig. 24), which is only remotely related to species of the taxon Agryllacris. Assuming similar genetic ground, the observed rare occurrences do not qualify for pronatism: they do not predate the expression of the corresponding derived condition at high frequency. This case better fits with parallelism as outlined above, except for the occurrence frequency. I propose to refer to the low frequency occurrences as ‘cryptoparallelisms’.
At this stage I propose to coin the term ‘iteronastism’ (‘intero-’= repeated) to encompass the concepts of atavism, reversal, cryptoparallelism, pronatism, and parallelism altogether. All have in common the repeated occurrence of a condition (either plesiomorphic or apomorphic) due to a similar underlying genetic organization. Defined on this ground, an iteronatism differs from a convergence in that the latter is due to a different underlying genetic organization (Hall 2003; and references therein).
Finally, I propose to extend the concept of iteronatism beyond repeated appearances due to similar underlying genetic organization. Given an initial pattern (i.e., the orthopteran forewing venation pattern) and a transformation type commonly observed (i.e. vein translocation), one can predict iterative appearances of the very same derived condition (i.e. CuPaα diverging from M + CuA as a consequence of a translocation) in various lineages, regardless of the adaptative significance of the transformation. In other words, provided an initial morphological context, exhaustion of morphological character states (sensu Wagner 2000), due to a limited number of transformation types, is likely to occur. A similar genetic organization is not necessarily underlying these iterative appearances, but a propensity of the condition to appear can be advocated. This broader perspective applies to all cases of iteronatism detailed above, and to the actual cases investigated herein. The concepts elaborated above are summarized as follows:
Iteronatism: repeated occurrence of a condition of any polarity, and at any species-level frequency, due to a similar underlying genetic organization (s.str.) / to a shared propensity of the condition to appear, provided the ancestral context (s.l.).
Atavism: occurrence of a plesiomorphic condition at low species-level frequency, due to the persistence and re-expression of an ancestral genetic organization; can repeat.
Reversal: occurrence of a plesiomorphic condition at high species-level frequency, due to the persistence and re-expression of an ancestral genetic organization; can repeat.
Cryptoparallelism: repeated occurrence of an apomorphic condition at low species-level frequency, due to a similar underlying genetic organization (s.str.) / to a shared propensity of the condition to appear, provided the ancestral context (s.l.).
Pronatism: appearance of an apomorphic condition at low species-- level frequency, predating its expression at high frequency in a direct relative, due to a similar underlying genetic organization (s.sfr.) / to a shared propensity of the condition to appear, provided the ancestral context (s.l.); can repeat.
Parallelism: repeated occurrence of an apomorphic condition at high species-level frequency, due to a similar underlying genetic organization (s.sir.) / a shared propensity of the condition to appear, provided the ancestral context (s.l).
This contribution exemplifies the importance of giving full consideration to intra-specific variation for establishing conjectures of topological homology. Indeed within-species variants can serve as Remane's (1952) intermediates and fill a gap in a transformation series otherwise incomplete. In that respect within-specimen variation taking place in symmetrical and/or serially arranged organs, such as wings, is a pinnacle. The most illuminating encountered case regards the specimen ANIC IWC OB 24 (Fig. 11), which right forewing exhibits an extremely rare configuration providing logical evidence in favor of a new STHC.
There is an unfortunate tendency of palaeoentomologists to erect new species without giving proper consideration to intra-specific variation (Béthoux 2009b). The current contribution exemplifies this issue, with six species, erected by a single author, demonstrated to be synonyms of a junior one, thanks to data on extant material. Where this trend is dramatic is that demonstrating junior synonymy is much more labor-intensive than erection of doubtful species, and can be achieved only when appropriate extant material is available (e.g., Schneider 1977, 1978), and/or on the basis of exceptional sample of fossil material (e.g., Cui et al. 2011). The routine erection of new species based on any observed difference is fundamentally flawed in that it neglects the fact that biological evolution could not take place without intra-specific variability.
A consequence of the assignment of †perfecta to Agryllacris is that this group stems in the Triassic, at least, in other words ca 170 my earlier than previously assumed. A loss of the complex gryllopteran stridulatory apparatus is no longer needed to account for the morphology of king crickets, raspy crickets and weta forewing.
Only a subset of recognized transformations was used to develop a new nomenclatural scheme (Fig. 30). The surveyed material was insufficient to allow relative timing of acquisition of the corresponding character states, and to define corresponding lineages. It is anticipated that this comparative analysis will allow wing venation homologies to be assessed in additional species, and the longest branches in the proposed transformation series to be split according to the sequence of character states acquisition.
This contribution is dedicated to the memory of Marylène Aubry, whose requirement for accuracy, and support along the years, contributed to the completion of this work. I am grateful to Maurice Béthoux for his long-term support, in particular in the last two years. I am also grateful to three anonymous reviewers and members of the editorial board of JOR for their useful comments, in particular regarding clarity and language. This contribution was partly supported by a SYS-resource grant from the London Natural History Museum (UK; visit of the author in 2002). I thank members of the paleoentomological staff of the Palaeontological Institute of the Russian Academy of Sciences (Moscow, Russian Federation) for their assistance during two research visits (2002, 2008). In particular the author is grateful to D.E. Shcherbakov for providing photographs of the specimen PIN 2555/1445, and to A.P. Rasnitsyn for providing photographs of the specimen PIN 2240/4112, and of labels of various PIN specimens. I thank the Alexander von Humboldt Foundation for the position of postdoctoral research fellow I have been granted (2007–2009), during which data on †perfecta were collected. I am grateful to members of the ANIC (Canberra, Australia) for assistance and useful discussion during my stay at this institution (2010), supported by the Australian Government Department of Education, Employment and Workplace Relations (Endeavour Award Postdoctoral Fellowship). I am grateful to G. Monteith, C. Lambkin, and G. Thompson for assistance during my visit at the Queensland Museum (2010; Brisbane, Australia). I am grateful to C.-X. Liu (Chinese National Academy of Sciences, Beijing, China) for assistance with the collection she is in charge off, and to Y.-Y. Cui (Capital Normal University, Beijing, China) for photographs of several extant specimens (2012). I am grateful to J. Hollier and P. Schwendinger (Museum of Natural History, Geneva, Switzerland) for assistance with literature data, and photographs of holotypes housed in their institution. I am grateful to I. Kogan (Freiberg University, Freiberg, Germany), A. Hennings, and M. Koch (Senckenberg Research Institute and Nature Museum, Frankfurt, Germany) for assistance with translation of Russian and German literature. Completion of this contribution would have been impossible without the taxonomic expertise of D. Rentz, who determined most of the extant specimens investigated herein. I thank him and Barbara and Buck for enjoyable fieldwork and stay at, and in the vicinity, of Kuranda. Along a similar personal line, I thank my ‘2010 housemates’ for a lot, and ‘Au revoir Simone’ for ‘Still Night, Still Light’.
Names and classification of species surveyed in the comparative analysis and indicated on Fig. 30 according to the current traditional nomenclature (Eades et al. 2012), and approximate correspondence with current cladotypic nomenclature.
Nomenclatural treatment in accordance to the ICZN
Genus † Probintoniella Sharov, 1968
†Probintoniella Sharov, 1968, p. 168
†Oshiellana Gorochov, 1994, p. 54, syn. nov.
Discussion.—The type-species of the genus †Oshiellana Gorochov, 1994 is considered as a junior synonym of the type-species of the genus † Probintoniella Sharov, 1968, therefore the former is a junior synonym of the latter.
Discussion.—See main text for discussion on proposed synonymy.
Genus †Gryllacrimima Sharov, 1968
† Gryllacrimima Sharov, 1968, p. 185
†Batkenella Gorochov, 1987a, p. 25, syn. nov.
†Dolichobaikenella Gorochov, 1987a, p. 26, syn. nov.
†Eubatkenella Gorochov, 1987a, p. 26, syn. nov.
†Brevibatkenella Gorochov, 1987a, p. 26, syn. nov.
Discussion.— All type-species of listed genera are considered as junior synonyms of Gryllacrimima perfecta Sharov, 1968 (see below), therefore the corresponding genera are junior synonyms of Gryllacrimima Sharov, 1968.
Discussion.— See main text for discussion on proposed synonymies.
Provisional composition of the taxon Agryllacris nom.-dis.-typ. nov., mainly based on photographs available from Eades et al. (2012) (and few direct observations, and literature data) allowing conclusive assignment.
in various characters in forewings of laudatum Johns, 1997, ornata Willemse, 1963, bicomis Karny, 1929a, pinguipes Rentz in Morton & Rentz, 1983, punctipennis Walker, 1869, and rufovaria Kirby, 1888; °, °°,°°°, not applicable/observable in 1, 2, 3 cases, respectively; *, ***, ******, based on n-1, n-4, n-6, respectively.