Revision of Xyalophora Kieffer and description of Xyalophoroides gen . n . ( Hymenoptera : Figitidae : Figitinae )

The genus Xyalophora (Giraud, 1860) is revised herein. The revision includes the type species of Xyalophora (Figites clavatus Giraud, 1860), the type material and the original descriptions of all the species of Xyalophora included in the Weld catalogue, and long series of undetermined material. Xyalophora clavata (Giraud, 1860) and X. singularis (Ashmead, 1896) are the only currently recognized species that should be included in Xyalophora. Four new species are described: X. belizini sp. n., X. giraudi sp. n., X. provancheri sp. n. and X. zarazagai sp. n. The type species of Ceraspidia, Ceraspidia japonica Belizin, 1952, corresponds to males of a species within Xyalophora. Thus, Ceraspidia is a new synonymy of Xyalophora, which results in Xyalophora japonica comb. n. Xyalophora impatiens (Say, 1836) and Xyalophora picea (Spinola, 1853) being considered like incertae sedis, as the latter does not belong to the Figitinae but to the Eucoilinae and is probably a species within Acantheucoela Ashmead, 1900. Xyalophora aciculata Benoit, 1956 is transferred to the genus Figites Latreille, 1802: Figites aciculata comb. n. Xyalophora leviventris Kieffer, 1908 is a synonym of Xyalophora quinquelineata (Say, 1836), which is transferred to Xyalophoroides gen. n., a new genus here described. The differences between all the genera of Figitinae with a scutellar spine are discussed and illustrated.


INTRODUCTION
The Figitinae are a cosmopolitan subfamily belonging to the Figitidae (Hymenoptera: Cynipoidea), a family that is biologically characterized as being parasitoids of the larvae of other insects, principally Diptera Cyclorrhapha (Ronquist, 1999).Two large morphological groups can be differentiated inside the Figitinae, according to the presence or absence of a scutellar spine (Weld, 1952).The systematics of the Figitinae with a scutellar spine has always been problematic, with four genera (Neralsia, Xyalophora, Xyalosema and Solenaspis) in a state of taxonomic chaos.
Neralsia was created by Cameron (1883) to include a species from Guatemala, N. rufipes Cameron, 1883, described as having a closed radial cell.Ashmead (1887) described the genus Solenaspis to include a North American species of Figitinae with a open radial cell, S. hyalinipennis.This generic name was preoccupied by a genus of Syrphidae (Diptera) described by Osten-Sacken (1881).Due to this, Dalla Torre & Kieffer (1910: 94-95) created the genus Xyalosema to include the two species of Cynipoidea previously described as Solenaspis (Xyalosema hyalinipennis and X. singularis).Weld (1930) studied the type species of Neralsia, N. rufipes, and noticed that this species has a open radial cell, not closed as cited in the original description (Cameron 1883); for this reason, he synonymized Xyalosema to Neralsia and transferred all the species of Xyalosema to Neralsia, except Xyalosema singularis (Ashmead, 1896), which was transferred to Xyalophora Kieffer, 1901.Xyalophora was described by Kieffer (1901) to include a European species of Figitinae, Figites clavatus Giraud, 1860 (now Xyalophora clavata); the original description indicates that this species has a scutellum with a longitudinal sulcus and a distal spine, and pubescent eyes, in contrast to Neralsia (= Xyalosema).This erroneous description (see redescription of X. clavata below), together with the fact that Neralsia is considered by Dalla Torre & Kieffer (1910) to be a member of the Aspicerinae, confused several authors studying American species of Xyalophora.
In addition, Belizin (1952) erected the genus Ceraspidia Belizin, 1952 to include a Japanese species with scutellar spine: C. japonica Belizin, 1952.This genus is included by Ronquist (1999) in the Aspicerinae, but a recent study of Ros-Farré (2007) concludes that Ceraspidia belongs to the subfamily Figitinae.Thus, currently the Figitinae with a scutellar spine are represented by 3 genera: Neralsia, Xyalophora and Ceraspidia.
In this paper, after examining long series of undetermined Xyalophora and studying the type material of Ceraspidia and Xyalophora, the status of these genera and the species mentioned above is resolved.The consequences of this revision are: (1) the description of Xyalophoroides gen.n., which includes a species transferred from Xyalophora, (2) the synonymy of Ceraspidia with Xyalophora, and (3) the description of four new species of Xyalophora.The differences between the genera of Figitinae with a scutellar spine (Neralsia, Xyalophora and Xyalophoroides gen.n.) are discussed and illustrated, and figures and descriptions of all the species included in Xyalophora and Xyalophoroides gen.n. are presented.

MATERIAL AND METHODS
The unidentified material was loaned, mostly, from the Canadian National Collection of Insects (CNCI, Ottawa, Canada), the United States National Museum of Natural History (USNM, Smithsonian Institution, Washington DC, USA), the California Academy of Sciences (CASC, San Francisco, California, USA), the Museo de la Plata (MLPA, La Plata, Argentina) and the Museo Entomológico de León (UNAN, León, Nicaragua).Part of this material has been deposited in the Pujade-Villar collection and is indicated as UB (Universitat de Barcelona, Barcelona, Spain) in the material studied.The types of Figites clavatus Giraud, 1860, deposited in MNHN (Muséum National d'Histoire Naturelle, Paris, France), Xyalophora leviventris Kieffer, 1908, and Xyalophora armata var nigricornis Kieffer, 1907, deposited in CASC, Solenaspis singularis Ashmead, 1896, deposited in USNM, Xyalophora aciculata Benoit, 1956, deposited in MRAC (Musée Royal de l'Afrique Centrale, Tervuren, Belgium) and Ceraspidia japonica Belizin, 1952, deposited in ZIN (Zoological Institute of the Russian Academy of Sciences, Saint Petersburg, Russia) were also studied.
The terminology of the morphological structures comes from Gibson (1985) and Ronquist & Nordlander (1989); sculptural terminology from Harris (1979).The measurements and abbreviations used include: F1-F11, first and following flagellomeres; POC (postocellar distance) is the distance between the internal margins of the posterior ocelli; OOC (ocello-ocular distance) is the distance between the external margin of the lateral ocellus and the internal margin of the compound eye; COC (ocellar distance) is the distance between the lateral and frontal ocelli; the diameter of the lateral ocellus is its greater diameter; the transfacial line is the distance between the internal margin of the compound eyes measured at the level of the antennal sockets (toruli).The relation between the scutellum (spine included) and the spine is measured in its dorsal projection.
The features mentioned in the descriptions are present in both sexes; the differences present in the males are indicated in the subheading "Males" of the description; the morphological characters of the male antenna are not relevant to the diagnosis so they are not detailed in each species, only in the generic description.
Scanning electron microscopy (SEM) images were taken on a Leica 360 SEM, at low voltage (700 V) without gold coating in order not to damage the specimens.The images are available from the databank www.morphbank.com.
Thirteen taxa of Figitinae were studied in the phylogenetic analysis, including all the Xyalophora species, four Neralsia species in order to cover its intrageneric variability, and the only species of Xyalophoroides gen.n.; Amphitectus areolatus (Hartig, 1840), a figitine without a scutellar spine, was included as an outgroup.The analysis was based on a data matrix of 14 characters (Appendix 1); multistate characters were marked as additive only if the states appeared to form a linear transformation series (characters 1, 3, 8).The analysis was run using uniformly weighted parsimony with TNT v. 1.1.The heuristic search was conducted with 1000 replications and the support of the branches was calculated using jacknife analysis.

RESULTS AND DISCUSSION
The examination of long series of undetermined specimens of Xyalophora and the study of the type material of all the species included within Xyalophora and Ceraspidia, together with the previous studies on Neralsia (Jiménez et al., 2005b(Jiménez et al., , c, 2006;;Pujade-Villar et al., 2006), reveals several things: (1) Xyalophora aciculata Benoit, 1956 corresponds to Figites: Figites aciculata (Benoit, 1956) comb.n.
(3) Xyalophora leviventris Kieffer and X. quinquelineata (Say) are synonyms, resulting in only one valid species: X. quinquelineata (Say) (see redescription and comments on the synonymy below).This species has some characters present in Xyalophora but not in Neralsia, and vice versa (Table 1).The combination of characters makes the assignment of this species to any of the two genera impossible.Furthermore, X. quinquelineata has three unique characters, not present in either Xyalophora or Neralsia: (1) radial cell half open, (2) forewing glabrous, without setae, and (3) F1 in male antenna longer than F2.These characters, together with the mixture of characters of Xyalophora and Neralsia, justify the erection of a new genus, Xyalophoroides gen.n., to include this species: Xyalophoroides quinquelineata comb.n.Phylogenetic analysis confirms the erection of the new genus (Fig. 8, see comments following the descriptions).The unique apomorphy of this new genus is the glabrous forewing (see details in the diagnosis of Xyalophoroides and the redescription of X. quinquelineata below).
(4) Some characters mentioned by Weld (1952) to differentiate Neralsia and Xyalophora are inconsistent (Jiménez et al., 2005a), and some new characters must be used to differentiate these genera and Xyalophoroides gen.n.The characters used to define these three genera are detailed in Table 1 and in the descriptions.These are the same characters that constitute the data matrix (Appendix 1) used for the phylogenetic analysis.A key to differentiate the three genera follows the descriptions.
(6) Four new species of Xyalophora are described and a key to all Xyalophora species is provided.
Xyalophora aciculata Benoit 1956: 377.Examination of the type material reveals that this species does not have a scutellar spine, though it was said to have one in the original description (Benoit, 1956: 378).Actually, the scutellum is margined laterally and the posterior margin is curved upwards, a state that occurs in some species of Figites.The absence of a scutellar spine and of transverse sculpture inside the notauli, and the presence of strigae on the metasomal tergum II, makes it clear that X. aciculata must not be included in Xyalophora.It cannot belong to Xyalophoroides gen.n., either, not only because it lacks the scutellar spine but also because the radial cell is completely closed in X. aciculata.Thus, we conclude that X. aciculata must be transferred to Figites, which results in Figites aciculata (Benoit, 1956)

Diagnosis
Xyalophoroides gen.n. can be distinguished from the other two genera of Figitinae with a scutellar spine by the radial cell half open (completely open in Neralsia, com-pletely closed in Xyalophora), the forewings glabrous (all the species of Neralsia and Xyalophora have abundant setae on the wing margin and at least some setae on the blade) and the male antenna with F1 longer than F2 (F1 shorter both in Neralsia and Xyalophora).In addition, it is differentiated from Neralsia by its scutellum always rugose, ending in a very short spine, at most 1/10 of the total length of the scutellum, interfoveal carinae always at the same level as the foveae, mesopleura completely striated, strigae strong, and male antenna with F1 larger than the rest of segments.Xyalophoroides gen.n. can be distinguished from Xyalophora by its mesoscutum without sculpture, an occiput irregularly carinated, and a metasomal tergum II striated at the base; in addition, the females of Xyalophoroides gen.n. have moniliform antennae and the males have two smooth areas on the face; this is not the case in Xyalophora.

Description
Coloration black.Face of the females (Fig. 1h) with irregular carinae that emanate from clypeus or centre of face towards toruli; surface around eyes coriaceous; interocellar space smooth; genal sulcus present and strongly costulate; occiput with strong irregular carinae (Fig. 2h).Face of male same as females except for two smooth areas (Fig. 7e).Female antennae shorter than body, moniliform, with 13 segments; male antennae much longer than body, filiform and with 14 segments.Pronotal plate weakly incised dorsally; lateral areas of pronotum (Fig. 4h) with sharp carinae in antero-ventral and dorsal part, rest of pronotum smooth.Mesopleura completely sculptured, with broad carinae, transverse, continuous and parallel (Fig. 4h).Scutum (Fig. 5h  notauli percurrent, internal surface smooth.Scutellar disk slightly humped, posterior margin with very short spine (Fig. 4h); scutellar disk (Fig. 5h) with irregular carinae; scutellar foveae (Fig. 5h) with carinae inside and separated by an interfoveal carina at same level as foveae (in lateral view).Tarsal claw simple, long and very arcuate.Forewings glabrous on margin and blade; radial cell half open, areolet absent.Metasoma oval in females, truncated in males; metasomal tergum I finely and densely carinate, metasomal tergum II densely and uniformly carinate at base (Fig. 7c); metasomal tergum III mostly smooth and punctate distally; metasomal tergum VIII with posterior margin concave in lateral view.
Etymology.Related to the genus Xyalophora.Gender feminine.

Description
Coloration.Black.Antennae dark brown.Legs and ventral area of metasoma reddish.
Head.Face of the females (Fig. 1h) with irregular carinae that emanate from clypeus or centre of face towards toruli; surface around eyes coriaceous; intero-cellar space smooth; genal sulcus present and strongly costulate; occiput with strong irregular carinae (Fig. 2h).Face of males same as females except for two big smooth areas below toruli (Fig. 7e).
Antennae.Female with F1 1.3 times longer than F2, last segments 1.4 times longer than wide.Males with F1 slightly larger than following segments.
Forewings (Fig. 6j).Glabrous, no setae on wing margin and blade; even if a number of dots can be seen on the wing blade that resemble sockets of broken setae, all specimens of X. quinquelineata lack completely setae on the forewings.Radial cell half open, 1.7 times longer than wide, length of vein R1 along the forewing margin reaching half way the width of the cell or beyond; areolet absent.
Variability.The length of vein R1 along the forewing margin is variable.Depending on the specimens, R1 can cover only 1/3, ½ or even ¾ of forewing margin of radial cell.The scutellar spine is also variable.In all the specimens it is short, but its shape, size and length is variable.Nevertheless, scutellar spine length is always 1/8-1/10 of total length of scutellum.The variation observed in these characters is considered here as intraspecific variation.We have not been able to separate groups of specimens with completely different morphological models.For this reason, previous to the erection of the new genus, Xyalophora leviventris was synonymyzed to X. quinquelineata after examining hundreds of undetermined specimens (see studied material below).
Distribution.Holarctic.Present in USA and Canada.

Biology
The only published data corresponds to that in Weld (1944), who cites this species from puparia of Sarcophaga lherminieri (Diptera: Sarcophagidae).In the studied material this species is also cited from Ravinia querula (Diptera: Sarcophagidae).From all the label data it can be concluded that X. quinquelineata attacks cyclorrhaphan Diptera (Sarcophagidae, Muscidae) specifically in cow dung, even though it is cited from Laspeyresia nigricana (Lepidoptera, Tortricidae) and Metator pardalinus (Acrididae, Orthoptera); these two hosts are highly improbable and should be confirmed.

Diagnosis
Xyalophora can be distinguished from the other two genera of Figitinae with scutellar spine (Neralsia and Xyalophoroides) by several characters: radial cell closed, mesopleura strongly striated and irregular in anterior area, interfoveal carina below foveae level, scutum sculptured (at least in areas adjacent to notauli), internal surface of notauli with sculpture, occiput with strong and continuous carinae, F1 of male antenna clearly shorter than F2, sculpture on face of the females homogeneous (without radiating carinae), that on face of the males present all over the face (without two big smooth areas as in Neralsia and Xyalophoroides).

Taxonomic comments
The only previously described species that have the generic characters mentioned above are: X. clavata (Giraud) and X. singularis (Ashmead).Ionescu (1969) described for the first time the males of Xyalophora clavata, which show substantial sexual dimorphism in the antennae.The study of Ionescu (1969) and the examination of a long series from CNCI and USNM confirm that Belizin's (1952) description of Ceraspidia corresponds to the males of Xyalophora.Thus, we have no doubt in considering Ceraspidia a syn.n. of Xyalophora, which results in Xyalophora japonica (Belizin) comb.n.Xyalophora armata var nigricornis Kieffer was considered by Weld (1951: 598) to be a synonymy of Xyalophora quinquelineata (Say); after the examination of this material, we confirm this synonymy as valid; X. quinquelineata (= Xyalophora leviventris Kieffer, syn.n.) is now included in the new genus Xyalophoroides gen.n.In addition, Xyalophora aciculata Benoit belongs to Figites and the correct name is F. aciculata (Benoit) comb.n.Moreover, Xyalophora impatiens (Say, 1836) and X. picea (Spinola, 1853) are considered incertae sedis.The reasons for all these taxonomical changes are detailed below.Finally, Xyalophora armata (Say, 1836), which was included in Neralsia by Weld (1930), is a nomen dubium according to a recent study (Jiménez et al., in press).

Biology
Unknown.There are no records of hosts for Xyalophora.The only host ever cited for Xyalophora was Sarcophaga Meigen 1826 by Ionescu (1969) for the species X. quinquelineata, but this species was transferred to Xyalophoroides gen.n. in this work.However, Xyalophora may be a parasitoid of Diptera: Cyclorrhapha in habitats such as dung and carcasses.

Diagnosis
The females of Xyalophora belizini sp.n., as in Xyalophora singularis and X. giraudi sp.n. have transverse and continuous carinae more or less parallel over the entire occiput (Figs 2e-g), as well as moniliform antennae, with the last segment subglobular (Figs 3d-f).In Xyalophora belizini sp.n., the occipital carinae are parallel and abundant (Fig. 2e), and the mesopleura is transversely carinate without any other type of sculpture (Fig. 4e), while in X. singularis and X. giraudi sp.n., the occipital carinae are subparallel and not so abundant (Figs 2f, g), and mesopleural carinae are present anteriorly with an alveolate sculpture (Figs 4f, g).Xyalophora belizini sp.n. and X. giraudi sp.n. are the only species in which the radial cell of the males is longer than in the females (Figs 6e-f, h-i).Finally, in X. belizini sp.n., the interfoveal carina is clearly visible (Fig. 5e), while in X. giraudi sp.n. it is much below the fovea level (Fig. 5g) and nearly invisible, resulting in apparently only one scutellar fovea.
Forewing.Radial cell closed, 1.7 times longer than wide (Fig. 6e).Setae present on surface and margin.

Diagnosis
The females of Xyalophora clavata have clavate antennae (Fig. 3g) like X. japonica, X. zarazagai sp.n. and X. provancheri sp.n. (Figs 3b, c, h); the males and females of these species have irregular and discontinuous occiput carinae just behind the ocelli (Fig. 2a-d).Xyalophora clavata can be differentiated from the other species of this genus by its very short scutellar spine, shaped like an equilateral triangle (in dorsal view) in both sexes (Fig. 5a).In addition, X. clavata is the species with coriaceous sculpturing more extended on the scutum (Figs 5a, 7a).
Variability.The specimens collected in the Nearctic have the centre of the scutum smooth, differing from the Palaearctic specimens, which have more or less extended coriaceous sculpture.Two males deposited in the USNM (collected in Arizona and Mexico) have the scutellar spine slightly different than the rest of the specimens; the spine is very short but not pointshaped.

Diagnosis
The females of Xyalophora giraudi sp.n., X. singularis and X. belizini sp.n. have transverse and continuous carinae over the entire occiput (Figs 2e-g), as well as moniliform antennae in females, with the last segment subglobular (Figs 3 d-f).In Xyalophora giraudi sp.n. and X. belizini sp.n., the radial cell is longer in males than in females (Figs 6h-i, e-f); in the other species of this genus this dimorphism is not present.Xyalophora giraudi sp.n. is the only species of Xyalophora with an interfoveal carina much below the foveae level, nearly invisible, resulting in apparently only one scutellar fovea (Fig. 5g).
Mesosoma (Figs 4g,5g).Pronotal plate weakly incised dorsally; lateral areas of pronotum with sharp carinae on antero-ventral and dorsal part, rest of pronotum slightly coriaceous.Mesopleura completely carinate, anteriorly with an alveolate sculpture.Scutum with weak sculpture next to notauli.Median sulcus of scutum clearly defined, with carinae inside.Notauli percurrent, internal surface with transverse parallel carinae.Interfoveal carina much below level of foveae, resulting in apparently only one large fovea, weakly carinate inside.Scutellar disk humped, with few, irregular, strong carinae, surface between them smooth.Scutellar spine long, slightly longer than 1/3 of total length of scutellum.
Forewing.Radial cell closed, 1.6 times longer than wide (Fig. 6h).Setae present on surface and margin.
Distribution.Neotropical.Known from Argentina, Nicaragua and Ecuador.
None of the above confirm that Figites impatiens is a Xyalophora, nor that this species even belongs in the Figitinae.As a result of this study, Xyalophota impatiens (Say, 1836) is considered incertae sedis.

Diagnosis
Xyalophora japonica comb.n. is morphologically similar to X. clavata, X. zarazagai sp.n. and X. provancheri sp.n. (Figs 3b, c, g), since the females of all them have clavate antennae (Fig. 3h), unlike other Xyalophora species.Xyalophora japonica is easily differentiated from X. clavata by the shape of the scutellar spine (Fig. 5b), which is long in X. japonica but very short in X. clavata (Fig. 5a).Xyalophora japonica (Fig. 2b) has stronger and more irregular carinae on the occiput than X. zarazagai sp.n. and X. provancheri sp.n. (Fig. 2c-d).In addition, the frons and interocellar area have a strong and alveolated sculpture in X. japonica (Fig. 1b) but this is weak and less defined in X. zarazagai sp.n. and X. provancheri sp.n. (Fig. 1c-d)
Forewing.Radial cell closed, 2 times longer than wide (Fig. 6b).Setae present on surface and margin.
Males.As in females except: shorter in length (2.2-2.5 mm), antennae filiform, with 12 flagellomeres, F1 shorter than others (3 times longer than wide), remaining flagellomeres subequal (3.9 times longer than wide).Taxonomic comments.Ceraspidia japonica was described by Belizin (1952) from one male collected in Japan.After studying the material from CNCI and determining for the first time females of this species, we can affirm that this species should be included in Xyalophora.Thus, we establish X. japonica comb.n.
Type material.Holotype % from JAPAN, deposited in the ZIN with the following labels: "Misaki, Kiu Shiu, Japan,
The type material of this species was not located in the Spinola collection, deposited in the Museo Regionale di Scienze Naturali di Piemonte, thus we consider it is lost.However, the original description of Figites picea Spinola, 1853 suggests that this species cannot be a Xyalophora as the absence of setae on the compound eyes, the presence of whitish spots on the lateral areas of the pronotum, the darkened strips in the forewings and the ring of whitish setae at the base of the metasoma make this placement highly unlikely.These characters, together with the shape of the scutellum (tridentated), the type of carinae on the propodeum and the body coloration (brown polished and bright), described by Spinola (1841: 42), lead us to believe that this species belongs in the Eucoilinae, possibly in Acantheucoela Ashmead, 1900.In addition, Figites picea, according to the original description, has the scutellar foveae fused.In the revision of Acantheucoela done by Díaz (1987), one species was included with foveae fused, A. brevidens (Kieffer, 1909); in addition, this species has a neotropical distribution (Peru and Bolivia), as Figites picea (Brasil).Despite all the characters mentioned, we cannot affirm the correct placement of this species without examining the type material.Thus, in this study Xyalophora picea (Spinola, 1853) is considered an incertae sedis.

Diagnosis
The females of Xyalophora provancheri sp.n. have clavate antennae (Fig. 3b), as X. clavata,X. japonica comb. n. and X. zarazagai sp. n. (Figs 3c,g,h); the males and females of these species have the carinae on the occiput irregular and discontinuous, just behind the ocelli (Figs 2a-d).The females of Xyalophora provancheri sp.n. are distinguished easily from X. clavata by the shape of the scutellar spine, which is very short in the latter (Figs 4a, d).The females of X. provancheri sp.n. can be confused with those of X. japonica, but the parallel carinae on the occiput, truncated and interrupted by a smooth area in X. provancheri sp.n. (Fig. 2d) and the shape of carinae on the frons (Fig. 1d) differentiate them (Figs 1b,2b).Xyalophora provancheri sp.n. has a longer scutellar spine than X. zarazagai sp.n. and an interocellar area smooth, while X. zarazagai sp.n. has this area sculptured.
Head.In anterior view, triangular; in dorsal view, 2 times longer than wide.Face with barely visible carinae and with wrinkles, emanating from centre of face to frons, reaching ocelli (Fig. 1d); interocellar space smooth; transfacial line 1.5 times longer than eye's height.Relation POC : OOC : COC is 8 : 5 : 4, diameter of lateral ocellus 2.5.Genal sulcus costulate, projecting behind compound eyes.Occiput with some arcuate and parallel carinae in basal area, and with dorsolateral carinae towards lateral ocelli; centre and apical part of occiput smooth (Fig. 2d).

Diagnosis
Xyalophora singularis, Xyalophora belizini sp.n. and X. giraudi sp.n. have transverse and continuous carinae on the entire occiput (Figs 2e-g) and the antennae of the female moniliform, with the last segment subglobular (Figs 3 d-f).In Xyalophora singularis and X. giraudi sp.n. the mesopleura is transversally carinate and anteriorly with alveolate sculpture (Figs 4f, g), which differs from X. belizini sp.n., which has a mesopleura completely carinate, without alveolate sculpture (Fig. 4e).The position of the interfoveal carina separates X. singularis from X. giraudi sp.n., since in the former the interfoveal carina, even if placed below the foveae level, clearly delimits two foveae (Fig. 5f), while X. giraudi sp.n. has apparently only one scutellar fovea (Fig. 5g) since the interfoveal carina is placed at an extremely low level.
Distribution.Nearctic and northern Neotropical.Known from USA (Texas, Colorado, Alaska, Illinois), Mexico and Nicaragua.

Diagnosis
The females of Xyalophora zarazagai sp.n. have clavate antennae (Fig. 3c), as in X. clavata (Giraud), X. japonica comb.n. and X. provancheri sp.n. (Figs 3b,g,h); the males and females of these species have irregular and discontinuous carinae on the occiput just behind the ocelli (Figs 2a-d).The females of Xyalophora zarazagai sp.n. can be distinguished easily from X. clavata by the shape of the scutellar spine, which is very short in X. clavata (Fig. 4a, c).The females of X. zarazagai sp.n. can be confused with those of X. japonica comb.n., but the strong carinae and the alveolate sculpture on the face, the frons and the occiput in the latter (Figs 1b, 2b) distinguishes them (Figs 1c,2c).Xyalophora zarazagai sp.n. has a shorter scutellar spine than X. provancheri sp.n. and an interocellar area sculptured, while this area is smooth in X. provancheri sp.n.

Phylogenetic analysis
The results of the phylogenetic analysis presented in figure 8, consists of a strict consensus of three trees with a length of 26 steps, CI: 0.80 RI: 0.91, with the jacknife supports given.Two clear clades are separated: one with all the species of Xyalophora (well supported, jacknife value 92), and another with Xyalophoroides + Neralsia.The only species of Xyalophoroides, X. quinquelineata, was not included in Neralsia because this genus is well supported (jacknife 92) and well defined by three synapomorphies: a radial cell open (1 : 0), a scutellum carinate (7 : 1) and a F1 in male antenna only slightly shorter or subequal to F2 (10 : 2).In addition, the support value for the Xyalophoroides + Neralsia branch is weak (jacknife 48).The unique apomorphy of Xyalophoroides is the glabrous forewing (9 : 0).Thus, Xyalophoroides is phylogenetically well distinguished from both Neralsia and Xyalophora, and the erection of a new genus for X. quinquelineata is in our opinion justified.

TABLE 1 .
Characters used to define and differentiate Xyalophora, Xyalophoroides and Neralsia.