Revision of the Palaearctic species of Lysitermus Foerster (Hymenoptera, Braconidae, Hormiinae)

Abstract The three Palaearctic species of Lysitermus Foerster, 1863 (Braconidae, Hormiinae, Lysitermini) are revised. The type species is described for the first time together with both of the other species. Lysitermus suecicus (Hedqvist, 1957) is a new synonym of L. tritoma (Bouček, 1956), and L. longiventris (Tobias, 1976) of L. talitzkii (Tobias, 1971), stat. nov.


Introduction
The little-known Palaearctic and Afrotropical genus Lysitermus Foerster, 1863 (Braconidae, Hormiinae, Lysitermini) was described by Foerster in 1863 without a description of its type species, L. pallidus Foerster, 1863. The identity of the genus was unclear, which resulted in three new generic names for this taxon (Rogadinaspis Bouček, 1956;Paracedria Hedqvist, 1957;Prolysitermus Tobias, 1971). Hedqvist (1963) recognised his earlier mistake and synonymised the first two with Lysitermus Foerster. He also synonymised their type species (Rogadinaspis tritoma Bouček, 1956 and Paracedria suecicus Hedqvist, 1957) with L. pallidus. The type series of Lysitermus was examined by the first author in 1979, and it proved to be also congeneric with Prolysitermus Tobias (van Achterberg 1982) and Lysitermus sensu Tobias (1971Tobias ( , 1976 was renamed in Tritermus van Achterberg, 1982. In the generic revision of Afrotropical and West Palaearctic genera of Rogadinae (van Achterberg 1991) the author attempted to construct the first key to the European species based mainly on the type series and their original descriptions (but the latter was absent for L. pallidus). Four species were recognised by using the shape of the second metasomal suture (sinuate in L. tritoma;Figs 24,25), subparallel-sided metasoma (of L. longiventris (Tobias, 1976)) and the relative length of the median carina of the propodeum (for L. pallidus and L. suecicus). Lysitermus talitzkii (Tobias, 1971) was synonymised with L. pallidus by Belokobylskij and Tobias (1986).
The second author reared a series of L. pallidus from Diplodoma laichartingella (Goeze, 1783) (Lepidoptera, Psychidae). After studying the reared material, plus small reared series in the National Museums of Scotland and Naturalis Biodiversity Center, it turned out that there are only two species present in the material from western Europe, with most likely a third one in south-eastern Europe, and that the characters used in the previous key (van Achterberg 1991) were too variable to be useful. In this paper, a new key is presented and L. pallidus is described and illustrated for the first time, together with a redescription of L. tritoma and L. talitzkii.

Material and methods
The specimens studied were (rarely) collected in Malaise traps and directly conserved in 70% alcohol or reared from their hosts and preserved dry. For identification of the subfamily Lysiterminae, see van Achterberg (1993), and for identification of the genus Lysitermus, see van Achterberg (1991).
Morphological terminology follows van Achterberg (1988van Achterberg ( , 1993, including abbreviations for wing venation. Measurements are taken as indicated by van Achterberg (1988): for the length and the width of a body part the maximum length and width is taken, unless otherwise indicated. The length of the mesosoma is measured from the anterior border of the mesoscutum to the posterior border of the propodeum and of the first tergite from the posterior border of the adductor to the medio-posterior margin of the tergite. An asterisk indicates a new record for the country.

Diagnosis. See van Achterberg
Notes. Lysitermus Foerster and the widespread Old World genus Acanthormius Ashmead, 1906, are very similar and should be synonymised in future if molecular data show that Acanthormius and Lysitermus are paraphyletic. Up to now, with only few species sampled, the Old World Lysitermus species sampled clusters with Afrotritermus Belokobylskij, 1995 andAtritermus Belokobylskij, Zaldivar-Riverón &Quicke, 2007 and not with the Acanthormius clade (Jasso-Martínez et al. 2021). Therefore, we refrain from synonymising both genera, despite that both have vein CU1a of the fore wing at the level of vein 2-CU1 or above (Fig. 2) and lack a parastigma (but the parastigma is also rarely absent in Aulosaphoides van Achterberg, 1995). However, in Acanthormius vein 2-SR of the fore wing is complete and the third tergite, excluding lamella, usually protruding latero-apically. In Lysitermus vein 2-SR is often largely or entirely absent, if present then nearly always its posterior third unpigmented and only the lamella of the third tergite is protruding. The development of vein 2-SR is very variable; the specimen in Figure 2 has the vein in one wing only pigmented, as figured, but in the other wing entirely sclerotised. In Aulosaphoides vein CU1a of the fore wing is situated distinctly below the level of vein 2-CU1 (at the same level or above in Lysitermus) and vein r is emitted distinctly before the middle of the pterostigma (submedially in Lysitermus). Trissarthrum Ashmead, 1900, is traditionally included in Lysitermus (Wharton 1993;van Achterberg 1995;Yu et al. 2016), but its Neotropical type species, T. maculipennis Ashmead, 1900, from St. Vincent has a complete vein 2-SR of the fore wing and the propodeal areola comparatively narrow. Therefore, it may be a different Neotropical genus near Acanthormius lacking the apico-lateral protruding part of the third tergite and having vein M+CU1 of the fore wing non-tubular and vein 1-M of hind wing much longer than vein M+CU (van Achterberg 1995). Both characters are also present in the only other described species from the New World, L. woolleyi Wharton, 1993  The position of the tribe Lysitermini is uncertain, but there is increasing evidence for a subordinate position in the Hormiinae. Recently, Lysitermini are either included as a tribe in the Rogadinae sensu lato (Chen and van Achterberg 2019) or the Hormiinae (Jasso-Martínez et al. 2021), or treated as a separate subfamily .
Variations. Antenna with 16-17 segments; length of body 2.1-2.3 mm, and of fore wing 1.6-1.7 mm; length of ovipositor sheath 0.21-0.23× as long as fore wing; vein 2-SR of fore wing varies in reared series from nearly complete to entirely absent (Fig. 2), most frequently only basal half present as pigmented but unsclerotised vein; notauli complete to posterior third obsolescent; second metasomal tergite 1.3-1.5× longer than second tergite; lamella of third tergite more or less serrate (Figs 1, 5).
Distribution. Finland, Germany, Moldova, *Sweden. Biology. Five specimens of Lysitermus pallidus hatched from ten final instar larval cases of Diplodoma laichartingella (Goeze, 1783) (Lepidoptera, Psychidae) collected in Sweden by the second author. Dissection of the final instar larval cases showed that only three had been parasitized and five specimens had hatched from them. It clearly indicates that L. pallidus is a gregarious larval ectoparasitoid of this host, but probably a facultative one perhaps depending on the size (first-or second-year stage?) of the host. In northern Europe D. laichartingella has a two-year life cycle which raises the question of whether L. pallidus is a parasitoid only of fully grown larva and, therefore, has a two-year lifecycle like its host (Kunz 1989) or whether it might be able to complete its lifecycle in a juvenile larval case as well. The D. laichartingella cases are usually found in woodland areas where the caterpillar feeds on algae/mosses as well as dead insects on tree trunks, especially, of Fagus sylvatica L. and Quercus robur L. The specimen from Finland has also been reared from D. laichartingella (det. M. Mutanen) but was referred by Gupta and Quicke (2018) as reared from an unidentified psychid host.
Variations. Antenna with 14-17 segments; length of body 1.5-1.9 mm, and of fore wing 1.3-1.5 mm; length of ovipositor sheath 0.21-0.26× as long as fore wing; vein 2-SR of fore wing varies from completely absent (Figs 13, 19), complete (Fig. 29) to small unsclerotised trace (Fig. 45); notauli complete to posterior third obsolescent; scutellar sulcus with 1-3 carinae; precoxal sulcus smooth or with few crenulae; metapleuron largely smooth or with some rugulae medially; median carina of propodeum approx. as long as anterior side of areola or much shorter; second metasomal tergite 1.3-1.5× longer than second tergite; lamella of third tergite hardly serrate (Fig. 12), with 8-16 carinae; colour of body is very variable: southern specimens are brownish yellow with only third tergite dark brown and northern specimens are largely dark brown (except first and second tergites, but in both Finnish specimens also darkened; Figs 27-29), sometimes hind coxa, femur and tibia largely brown or dark brown (Fig. 29).
Notes. The medio-longitudinal carina of the propodeum is very variable in length, from about as long as oblique anterior side of propodeal areola to much shorter and the second metasomal suture varies from distinctly sinuate (typical L. tritoma;Figs 24,25) to straight (typical L. suecicus;Figs 15,20,22). After more reared specimens became available, intermediates of both characters have been found, and there are no grounds to separate any longer the two species, as has been proposed by van Achterberg (1991).
The holotype of L. suecicus (NRS) is incorrectly figured in the original description. For instance, the second tergite is not twice as long as the third tergite but 1.5× (Fig. 22), the ovipositor sheath is not widened but subparallel, and the pterostigma is not robust but rather slender (Fig. 22).

Material. Holotype of
Redescription. Paratype ♀ of L. longiventris, length of fore wing 1.4 mm, and of body 1.6 mm.
Head. Antenna 1.1× as long as fore wing, with 15 segments, rather widened apically (Fig. 30), scapus oblique apically, length of third segment 1.1× fourth segment; third, fourth, and penultimate segments 6.6, 5.8, and 2.9× their width, respectively; face largely smooth, with long erect setae and 2.2× wider than high; clypeus smooth and distinctly convex; head in dorsal view 1.6× broader than long medially, eye 1.4× longer than temple dorsally, strongly and roundly narrowed posteriorly (Fig. 32); vertex smooth and shiny, including area near stemmaticum (Fig. 32); occipital carina strong medio-dorsally and in front of it micro-crenulate; POL:diameter of posterior ocellus: OOL = 20:12:35; eye in lateral view 1.5× as high as wide, temple distinctly broadening ventrally, smooth, and width of eye 1.3× minimum width of temple laterally; malar space 0.5× as long as height of eye and 1.7× as long as basal width of mandible; frons smooth and shiny.
Mesosoma. Mesosoma 1.5× as long as high in lateral view; mesoscutum granulate and shiny, medio-posteriorly with indistinct groove (Fig. 32); notauli only anteriorly distinct, largely smooth, but anteriorly crenulated; scutellar sulcus with one median carina and 6 weak crenulae; scutellum largely smooth and shiny, convex (Fig. 32); propodeum shiny, largely smooth between carinae and with large areola, its median carina shorter than anterior side of areola; side of pronotum and mesopleuron largely smooth; precoxal sulcus rather shallow and only its anterior half present, largely smooth; metapleuron largely smooth and with long setae. Wings. Fore wing: pterostigma elongate triangular and 5.0× longer than wide (Fig. 30), r issued from its middle and 0.7× shorter than width of pterostigma; 2-SR absent except for some pigmentation (Fig. 30); 3-SR 3.1× longer than r; SR1 straight and reaching tip of wing; m-cu rather short and largely unpigmented; subdiscal cell distally closed, CU1b far from m-cu and CU1a above level of 2-CU1 (Fig. 30). Legs. Hind coxa largely smooth; length of femur, tibia and basitarsus of hind leg 4.5, 11.1, and 8.4× their width, respectively; hind tarsus compressed.
Metasoma. Length of first tergite 0.6× its apical width, its surface longitudinally striate and with additional granulate sculpture between striae, dorsal surface evenly convex, its dorsal carinae lamelliform and medially interconnected anteriorly; medial length of second tergite 0.9× its basal width, and 1.4× as long as third tergite; second and third tergites longitudinally striate and with distinct additional granulate sculpture between striae; second transverse suture coarsely crenulate and nearly straight (Fig. 31); third tergite antero-laterally with minute tooth-shaped protuberance and gradually narrowed posteriorly (Fig. 31), with distinct and sculptured posterior rim; apical lamella of third metasomal tergite moderately protruding laterally (Fig. 31), straight and narrow medially in dorsal view, with approx. 14 carinae; setose part of ovipositor sheath 0.4× as long as hind tibia.
Biology. Unknown. Distribution. Moldova, Poland (Huflejt 1997), Russia. Notes. The male holotype of L. talitzkii (Tobias, 1971) was considered to be a synonym of L. pallidus Foerster by Belokobylskij and Tobias (1986) and van Achterberg (1991), but after examination of reared series consisting of both sexes it is obvious that the holotype male with its gradually widened hind tibia (Fig. 45) fits better with L. longiventris, as defined above, and the latter is synonymised with it. The holotype has the notauli shallowly impressed posteriorly (van Achterberg 1995) although normally the posterior half of the notauli are absent, but specimens with vaguely to distinctly indicated notauli have been examined among specimens of the closely related L. tritoma from Sweden (holotype of L. suecicus), Italy, and Portugal, and, therefore, we do not consider the more developed notauli as a valid reason to maintain L. longiventris as a species different from L. talitzkii.
Lysitermus longiventris (Tobias) was described from N. Caucasus and is very similar to L. tritoma (Bouček) but differs by having the third metasomal tergite of- ten with a minute tooth-like protuberance antero-laterally and its posterior lamella straight medially or nearly so, the metasoma slightly slenderer, the third tergite slightly less narrowed posteriorly, and vein 3-SR of the fore wing usually more than twice as long as vein r. The reduction of the longitudinal rugae on the third tergite is considered less distinctive because reduction of sculpture is common in Lysitermus males and to a lesser degree in females. Other differences given by Belokobylskij and Tobias (1986), such as the first tergite as long as wide at apex and the second tergite longer than its apical width, disagree with the only figure in the very short original description and the figures included here (Figs 31, 36, 44). Characters such as the third tergite less sculptured than second tergite and without distinct lamella apically, the face twice as wide as high, and the width of the hypoclypeal depression equal to distance from eye to depression are more or less also found in males of L. tritoma.