SYSTEMATICS Family Affinity of the Genus Palaeopsychops Andersen with Description of a New Species from the Early Eocene of British Columbia, Canada (Neuroptera: Polystoechotidae) 1 2 VLADIMIR N. MAKARKIN AND S. BRUCE ARCHIBALD Ann. Entomol. Soc. Am. 96(3): 171Ð180 (2003) ABSTRACT Palaeopsychops dodgeorum sp. n. from the Early Eocene Okanagan Highlands of Quilchena, British Columbia, Canada is described. The systematic position of the genus Palaeopsychops Andersen, 2001 is discussed, interpreting this as most closely associated with Polystoechotidae. Osmylites protogaea (Hagen 1862) is considered as nomen nudum and an objective synonym of Osmylites excelsa (Oppenheim, 1888), syn. n. KEY WORDS Eocene, Polystoechotidae, Neuroptera, Okanagan Highlands, fossil insects THE FAMILY POLYSTOECHOTIDAE is now a relict group very likely belong to this family (Makarkin, personal represented by only four species belonging to three observation). genera: Polystoechotes Burmeister, 1839, Platysto- Late Cretaceous and Tertiary Polystoechotidae echotes Carpenter, 1940, and Fontecilla Nava´s, 1932, have hitherto been unknown, because the single de- which are restricted to the New World (Oswald 1998). scribed species (Polystoechotes piperatus Cockerell, Although fossils of this family are few and controver- 1908 from Late Eocene of Florissant) was considered sial, occurrences in the Mesozoic of insects referable to be a psychopsid (Carpenter 1943, MacLeod 1970). to Polystoechotidae indicate that this family may have However, examination of the holotype shows that a been more diverse and widespread. The Early Jurassic polystoechotid afÞnity is most probable, although this species Mesopolystoechus apicalis Martynov, 1937, de- species certainly doesnÕt belong to the genus Polysto- scribed from Tajikistan, Central Asia, has been con- echotes. Unfortunately, the sole specimen of this spe- sidered the single member of the family Mesopolys- cies is an incomplete and poorly preserved forewing, toechotidae Martynova, 1949. Lambkin (1988) and its systematic position is not clear. Another Ter- implied his opinion of possible synonymy of Me- tiary genus, Palaeopsychops Andersen, 2001, was re- sopolystoechotidae and Polystoechotidae when he re- cently described from the Late Paleocene/Early Eo- ferred to M. apicalis as a polystoechotid (p. 455) and cene of Denmark in the family Psychopsidae, “probable Polystoechotidae” (p. 457). He also re- however, we consider this genus as most closely as- viewed the other four fossil species from the Late sociated withPolystoechotidae(see “Determination Triassic to Late Jurassic probably belonging to Polys- of Family AfÞnity”). toechotidae: Lithosmylidia lineata Riek, 1955 (Late A large lacewing was recently Þgured along witha Triassic of Australia); Kasachstania fasciata PanÞlov, short description by Archibald and Mathewes (2000, 1980, Osmyliodea distincta PanÞlov, 1980, and Ptero- Figs. 6A, 20). The authors brießy discussed its possible calla superba PanÞlov, 1980 (all from the Late Jurassic family afÞnity, comparing it withOsmylidae and of Kazakhstan). Another two species were assigned to Polystoechotidae. In this paper we provide a detailed Mesopolystoechotidae by Whalley (1988) description of this specimen and discussion of its sys- [Megapolystoechus magnificus Tillyard, 1933 (Late tematic position. We interpret this as a new species of Triassic of England)] and Hong (1983) [Mesopolys- the genus Palaeopsychops. toechus wangyingziensis Hong, 1983 (Middle Jurassic of China)]. Apart from these occurrences, there are Materials and Methods undescribed species from the Late Jurassic of Kaza- khstan and Early Cretaceous of southern Siberia that The specimen is from the Early Eocene lacustrine shales of the “Coldwater beds” exposed along the banks of Quilchena Creek at Quilchena in south-cen- 1 Institute of Biology and Soil Sciences, Far Eastern Branchof the tral British Columbia, Canada. The Quilchena locality Russian Academy of Sciences, Vladivostok, 690022, Russia (e-mail: is the oldest of the “Okanagan Highlands” series of [email protected]). 2 Department of Organismic and Evolutionary Biology, Harvard fossiliferous Eocene shale and amber deposits in Brit- University, Museum of Comparative Zoology, 26 Oxford Street, Cam- ish Columbia and Washington State. This locality bridge, MA 02138 (e-mail: [email protected]). bears numerous fossil insect, Þsh, plant, and bird 0013-8746/03/0171Ð0180$04.00/0 ᭧ 2003 Entomological Society of America 172 ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA Vol. 96, no. 3 (feather, coprolite) remains. Although most shale sopolystoechus by Hong (1983), that is represented from Quilchena is easily worked, unfortunately the only by an incomplete forewing. particular nature of the matrix of this fossil (relatively In the drawing of the holotype of Palaeopsychops large-grained, consolidated to the fossil, and not ßak- latifasciatus (Fig. 10: Andersen 2001), the subcosta, ing away withneedles) makes preparation unusually anterior radial trace (“R1”), and anterior sectoral trace problematic. (“Rs”) all terminate together, however, the photo- Terminology of forewing venation and wing spaces graph (his Fig. 3) shows that whereas Sc and R1 do employed here mainly follows Oswald (1993a, 1998), probably fuse, Rs runs separately to the wing margin, in particular withregards to themedian vein: “MA” of as in our specimen. We Þnd it most probable that in authors (e.g., Carpenter 1940; MacLeod 1970) is the other species of Palaeopsychops the conÞguration of most proximal branchof Rs, and “MP”Õ is simply M. these veins is the same, and therefore, we assume that there is no “vena triplica ” in this genus (see Oswald 1993b: 41 for discussion of this character). Because the Taxonomy distal nygma (not mentioned in the text of Andersen (2001), but clearly seen in the photograph, Fig. 3) is situated in the neuropteran forewing between the Palaeopsychops Andersen 2001 most proximal two branches of Rs, it follows then that Palaeopsychops Andersen 2001: 422. AndersenÕs M1 is actually the most proximal branch of Diagnosis. Distinguished from other polystoechtid Rs, his M2 is M1, his Cu1 is M2 (Andersen 2001: Fig. 10), genera by the following combination of forewing his Cu2 is CuA, and his A1 is CuP (Andersen 2001: Fig. characters: (1) broad-subtriangular in shape [more or 9). less elongate in all other genera, except for Fontecilla]; (2) hind margin of wing not falcate [falcate in Fon- Palaeopsychops dodgeorum New Species tecilla]; (3) M2 strongly pectinately branched (shared by fossil genus Pterocalla) [few pectinate branches in Neuroptera incertae sedis: Archibald and Mathewes all other genera]; (4) outer gradate series of crossveins 2000: 1445, 1448, Figs. 6A, 20. in radial space well developed (shared by all extant Diagnosis. Distinguished from other species of ge- genera) [absent in fossil genera, except hindwings of nus by numerous, irregularly arranged crossveins in Mesopolystoechus apicalis and Osmyliodea distincta radial space (in all other species of Palaeopsychops (but see discussion)]; (5) transverse bands (fasciae) these crossveins form inner gradate series), somewhat distinct [absent in all extant genera and never de- more elongate shape of forewing. tected in any fossil genus]. Description. Forewing broad-subtriangular in Discussion. The genus Palaeopsychops was de- shape, length 45 mm. Trichosors indistinct, visible only scribed from the Late Paleocene/Early Eocene of at hind wing margin where margin well-preserved Denmark (“Mo-clay”), withfour species: P. latifascia- (see Fig. 5, not shown in Fig. 1). Distal nygma distinct, tus Andersen (type species), P. abruptus Andersen, P. appearing as small bulge in dilation between most angustifasciatus Andersen, and P. maculatus Andersen. proximal two branches of Rs (see Fig. 4). Costal space The latter species is possibly a synonym of one of the basally expanded, narrowed toward apex, withcom- other three, as the holotype is a hindwing, not forew- plete gradate series of crossveins parallel to anterior ing as in Andersen (2001) (the costal space not ex- wing margin. Subcostal veinlets always forked, some- panded basally as in forewings). Withthepresent times several times. Subcostal space rather wide along description of Palaeopsychops dodgeorum, the genus entire length, crossveins not preserved. Apical portion expands its distribution to NorthAmerica. of Sc not preserved (possibly obscured by matrix). R1 ϩ Palaeopsychops represents the single described Ter- (or Sc R1) entering margin apparently at wing apex, tiary genus in the family. A number of genera were withbrancheslong and forked. R 1 space somewhat erected from the Late Jurassic to Late Triassic (see narrower than subcostal space, with Ϸ10 crossveins. introduction) and none from the Cretaceous. Palaeo- Origin of Rs not preserved (possibly obscured by psychops is easily distinguished from all extant genera matrix), but clearly close to base of wing. Rs with (see “Diagnosis”). Of the Jurassic-Triassic genera only numerous (33Ð34 in number) pectinate branches, the genus Mesopolystoechus Martynov, 1937 is similar sometimes fusing withone another;many irregularly to Palaeopsychops, as seen from the venation of a spaced crossveins between them in central part of hindwing of Palaeopsychops maculatus (Andersen wing. Outer gradate series in radial space distinct, 2001: Fig. 12). The type species (M. apicalis) is rep- nearly parallel to posterior margin, continuing radio- resented by a well preserved,