A new genus and species of Homoiopteridae from the Upper Carboniferous of the Intra-Sudetic Basin, Czech Republic (Insecta: Palaeodictyoptera)

New palaeodictyopterid Paraostrava stanislavi gen. n., sp. n. is described from the Upper Carboniferous (Duckmantian) deposits of the Jan Šverma Mine in northern Bohemia (Czech Republic). The new taxon based on hindwing venation is attributed to Homoiopteridae and compared with the other homoiopterid and heolid genera within Homoiopteroidea. Due to the poor state preservation of Boltopruvostia robusta, we consider this taxon as Palaeodictyoptera: Homoiopteridae of uncertain position and restore the well defined genus Ostrava Kukalová, 1960 (type species Ostrava nigra Kukalová, 1960). Some uncertainties in the current state of knowledge on the taxonomy and phylogeny of the Homoiopteridae are pointed out. The characters matrix used to separate the genera of Homoiopteroidea is included.


INTRODUCTION
The fossil insect order Palaeodictyoptera Goldberg, 1854 within superorder Dictyoneuridea Handlirsch, 1906 (Palaeodictyopteroida sensu Bechly, 1996) represents a widely diverse group of Palaeozoic insects ranging from the Lower Carboniferous (Namurian A) through Permian up to the Lower Triassic (Middle "Buntsandstein" by single record of Thuringopteryx gimmi Kuhn, 1937) with a peak of abundance in the Upper Carboniferous (Bechly, 1997;Sinitshenkova, 2002).Palaeodictyopteran insects with a distribution mainly restricted to the tropical belt of the northern hemisphere are commonly found in Euroamerican Upper Palaeozoic deposits, which indicate their probable thermophily (Sinitshenkova, 2002).The taxon Palaeodictyoptera is considered to be paraphyletic after Bechly (1996).Because of this state, it is necessary to revise the described species in the perspective of a future phylogenetic study of this group.But many species are based only on more or less complete wings.Thus, such analysis may be very difficult to achieve.Further orders of Palaeodictyopteroida like Megasecoptera, Diaphanopterodea and Permothemistida differ clearly in wing venation patterns and are therefore likely to be monophyletic groups (Grimaldi, 2001).We follow the wing venation terminology of Kukalová-Peck (1991).The systematics of Palaeodictyoptera is mainly based on the work of Sinitshenkova (2002), partly adapted from Riek (1976).Both systems are critically reviewed and discussed on the current position of Homoiopteridae.
The Intra-Sudetic Basin is a NW-SE extending structure situated in the northern part of former Bohemia (Czech Republic) (see Fig. 1).The major part of this basin lies in the territory of Poland while a narrow prominence is located in Žaclé surroundings and Broumov in the Czech Republic.The basin is filled with Lower Car-boniferous to Triassic continental sediments (Spudil, 1999).Generally, the overlying deposits of Šverma mine group seams are of fluvial origin.Zoopalaeontological and phytopalaeontological records were reviewed by Tásler et al. (1979).
Genus Paraostrava gen.n.Description.Wing membrane probably hyaline; wing covered with a dense irregular network, except in the more poorly preserved regions, i.e. the basal parts of areas between ScP and R, R and M, and MA and MP; extreme apex of convex ScA visible at its point of fusion with anterior wing margin, 87.9 mm from wing apex; concave ScP straight, reaching anterior wing margin very distally, 20.4 mm close to wing apex; numerous small simple cross-veins in area between C and ScP; basal part of area between ScP and R poorly preserved but with some visible cross-veins, distal part of same area with several short cross-veins; RA simple and straight, ending on anterior margin 8.6 mm from wing apex; RA and RP area with numerous long sigmoidal cross-veins; base of RP 48.4 mm from wing apex, 25.7 mm distal to division fork: M divides into MA and MP; first fork of RP 12.9 mm distal to its base, both two basal forks of RP divided into three branches; RP reaching wing apex; MA simple, slightly curved, ending on posterior wing margin 17.8 mm from wing apex; MP divided dichotomously into seven branches; MP area very broad reaching the posterior wing margin from 20.4 to 47.3 mm of wing apex; a rather strong oblique cross-vein between M and R, emerging from M at the point of division of M into MA and MP; at the same point, another strong oblique crossvein between CuA and M; CuA slightly curved, with two short but strong branches ending on posterior wing margin; fork of Cu into CuA and CuP poorly preserved, probably 95 mm from wing apex; CuP slightly curved, more or less parallel to CuA and forked into two branches, first 51.4 mm and second 55.0 mm from posterior wing margin; five anal veins preserved, forming a large anal area (hind wing).Discussion.The current state of our knowledge on higher systematics and division of Palaeodictyopteroida sensu Bechly (1996) is rather complicated and partly confusing.There is no real phylogenetic analysis of this group that we can simply follow.In this case we trace both currently supported systems with additional comments.
First, it is not possible to follow the key of the "Eupalaeodictyoptera" proposed by Riek (1976: 230).For example, this author characterized Homoiopteroidea sensu Riek (1976) by the following characters: "CuA branched"; "MA simple"; "ScP extending almost to wing apex".But the CuA is simple in several taxa currently included in the Homoiopteridae (Scepasma gigas Handlirsch, 1911, and Mazonopterum wolfforum Kukalová-Peck & Richardson, 1983) (Kukalová-Peck & Richardson, 1983).Sinitshenkova (2002: Fig. 138) proposed a new phylogeny and system of the superorder Dictyoneurida Handlirsch, 1906 (Palaeodictyopteroida sensu Bechly, 1996).This system is not based on a strict use of the cladistic method.At least in the most basal dichotomy, the clade Dictyoneurina is supported by the alleged synapomorphy "wing wide basally" and its sister clade Frankenholziina is supported by the alleged opposite state of character "wing base narrow", also considered as a synapomorphy.The indication of an outgroup to polarize the characters would be suitable.Thus, Sinitshenkova's classification cannot be considered based on the cladistic method, even if it uses the cladistic terminology.
First, Paraostrava differs from the Miracopteron Novokshonov, 1993(Miracopteridae Novokshonov, 1993 of order uncertain but with "similarities" with the Homoiopteridae, Miracopteron mirabile Novokshonov, 1993, based on the base of a hind wing) at least in its position of the M fork clearly distal in Paraostrava (Novokshonov, 1993).
Adolarryia bairdi Kukalová-Peck & Richardson, 1983 is based on a nymph, difficult to compare with an adult wing.Nevertheless, it has a CuA divided into long branches, unlike Paraostrava.Monsteropterum moravicum Kukalová-Peck, J. 1972 was attributed to Palaeodictyoptera with an uncertain systematic position by Carpenter (1992) and then relocated within Megasecoptera after the presence of the tergal projections by Kukalová-Peck (pers. comm.).It is based on a body with wing bases that cannot be compared with Paraostrava.Further three taxa were also considered by Carpenter (1992) as Palaeodictyoptera with uncertain position, although Brauckmann (1991a) listed them in Homoiopteridae (see also Brauckmann & Becker, 1992: 139).Amousus mazonus Handlirsch, 1911 and Ametretus laevis Handlirsch, 1911 are both based on hind wing bases, with forked CuA, unlike Paraostrava.Mammia alutacea Handlirsch, 1906 is based on a very incomplete wing fragment, but differing from Paraostrava in its MA strongly curved basally.
Mazothairos enormis Kukalová-Peck & Richardson, 1983 is based on the tergum of a single segment of a thorax.Of course, it is not possible to compare it with the other species of the family.Turneropterum turneri Kukalová-Peck & Richardson, 1983 is based on two wing bases.Nevertheless, it differs from Paraostrava in its CuA with three branches.Larryia osterbergi Kukalová-Peck & Richardson, 1983 has the base of RP in a very distal position, far from that of MA, and has MA distally forked and CuA divided into three main branches, unlike Paraostrava.Mazonopterum wolfforum Kukalová-Peck & Richardson, 1983 has no long sigmoidal secondary veins in the area between RA and RP, a long distance between bases of MA and of RP, even if that of RP is in a distal position, unlike Paraostrava.Mazonopterum shares with Paraostrava the simple MA and CuA.
Parathesoneura Sinitshenkova, 1977 (with two species P. carpenteri andP. anfractuosa, in Sharov &Sinitshenkova, 1977: 52-54) shares with Paraostrava the almost simple MA and CuA, and the presence of long sigmoidal veinlets in the area between RA and RP.However, they have a very long basal part of RP, between its base and first branch, and the base of MA distal to that of RP, unlike Paraostrava.
Note that there is a strong similarity between Parathesoneura and Archaemegaptilus Meunier, 1908 (type genus of the Archaemegaptilidae Handlirsch, 1919), viz.they share a very long basal part to RP, the same organization of MA, MP and CuA, and the presence of long sigmoidal veinlets in RA -RP area (Kukalová, 1969: Fig. 46).There are very few visible differences between the two taxa, in the number of branches of MP.Sinitshenkova (2002: 117) included the Archaemegaptilidae in the Eugereonoidea Handlirsch, 1906(sensu Sinitshenkova, 2002).She characterized this last super-family by (1) their wings long and narrow, and (2) the "archedictyon" lost.The first character is not obvious in the type specimen of Archaemegaptilus kiefferi Meunier, 1908 because the basal half of its wing is not preserved.The second character is convergently present in the Homoiopteroidea.Thus, a phylogenetic revision of the position of the Archaemegaptilidae would be necessary to clarify this situation.
Boltopruvostia Strand, 1929 is based on the type species B. robusta (Pruvost, 1919), a poorly known taxon, based on two wing bases attached to the thorax.Another species, B. nigra (Kukalová, 1960) is based on a nearly complete fore wing, originally attributed to the genus Ostrava Kukalová, 1960, later synonymized with Boltopruvostia (Kukalová, 1969).In Boltopruvostia nigra, the fork of the radius is distinctly basal to that of the median vein, unlike its distinctly distal position in B. robusta (Kukalová, 1960: Figs 1-2).This difference is not sufficient for a generic separation of Boltopruvostia because Brauckmann (1991b) demonstrated on rich material that a rather important variability occurs in the relative positions of these vein bases in Homoioptera vorhallensis.There are no visible differences between Boltopruvostia robusta and Paraostrava in their comparable preserved parts of wing venation, viz.bases of RP, fork of M, and bases of CuA and CuP (Pruvost, 1919: Pls 1-2).But nothing is known concerning the distal two-thirds of the wing venation of Boltopruvostia robusta.Thus, it is not possible to compare it accurately between Paraostrava and Boltopruvostia nigra.
Furthermore, Paraostrava and Boltopruvostia nigra differ in the relative widths of the area between RA and RP: about twice broader than that between RP and MA in Paraostrava, instead of being of the same width in Boltopruvostia nigra.This character is sufficient for a generic separation between these two taxa.But it is unknown in Boltopruvostia robusta.
As it is not possible to accurately attribute Paraostrava or Boltopruvostia nigra to Boltopruvostia, we choose better to restore the combination of genus Ostrava Kukalová, 1960 with Ostrava nigra Kukalová, 1960, to erect a new genus and species for Paraostrava stanislavi, and to consider Boltopruvostia robusta as Palaeodictyoptera: Homoiopteridae of uncertain position.

Fig. 1 .
Fig. 1.A -Geographical map of Permo-Carboniferous limnic basins of the Bohemian Massif and position of the Intra-Sudetic Basin.B -Stratigraphic column of the Intra-Sudetic Basin division with proximate time level of the presented fossil (gray box).Abbreviations: EL -Early Langsettian, LL -Late Langsettian, ED -Early Duckmantian, LD -Late Duckmantian.