New genus and species of lice in the Oxylipeurus -complex (Phthiraptera, Ischnocera, Philopteridae), with an overview of the distribution of ischnoceran chewing lice on galliform hosts

Here, we describe a new genus of lice (Phthiraptera, Ischnocera) in the Oxylipeurus -complex, parasitising galliform hosts in the genera Tragopan Cuvier, 1829. This genus, Pelecolipeurus gen. nov. , is separated from other members of the complex by the unique shape of the male subgenital plate and stylus, the male genitalia and other characters. The only previously-known species in the genus is Lipeurus longus Piaget, 1880, which is here tentatively re-described as Pelecolipeurus longus (Piaget, 1880), based on specimens from a non-type host, Tragopan temminckii (Gray, 1831). In addition, we describe a new species, Pelecolipeurus fujianensis sp. nov. , based on specimens from Tragopan caboti (Gould, 1857). An overview of the distribution patterns of ischnoceran lice on galliforms is presented, which suggests that host phylogeny, host biogeography and host biotope, as well as elevation of host range, may all be important factors that have structured louse communities on landfowl. We transfer the genus Afrilipeurus from the Oxylipeurus -complex to the Lipeurus -complex and include an emended key to the Oxylipeurus -complex.


Introduction
Chewing lice (Phthiraptera) in the Oxylipeurus-complex mainly parasitise gamefowl (Galliformes; Price et al. (2003)) and most species are known from Asian galliforms.Traditionally, most of the species have been placed in the one genus, Oxylipeurus (e.g.Clay (1938a); Hopkins and Clay (1952); Price et al. (2003)).However, this classification was challenged by, for example, von Kéler (1958) and Carriker (1967), who considered several groups of Oxylipeurus to be sufficiently distinct to form separate genera.Mey (2009) considered several of these genera valid and, since then, a large number of new genus-level taxa within this complex have been established (Gustafsson andZou 2020a, b, 2023;Gustafsson et al. 2020a, b).
Assessing taxon limits in this complex is difficult, as the overall chaetotaxy and morphology, including that of the male genitalia, are conserved in many genera and species are often delimited by more nebulous characters, such as head shape, mesosome shape and degree of reticulation of the cuticle (e.g.Gustafsson et al. (2020a)).Moreover, many species are poorly known and have not been fully described or illustrated; the last detailed revisions of the complex were published by Clay (1938a) and von Kéler (1958).Gustafsson et al. (2020a) tentatively considered Lipeurus longus Piaget, 1880, to belong to the genus Reticulipeurus Kéler, 1958, based on its placement by Clay (1938a), von Kéler (1958) and Złotorzycka (1966); however, they stated that they had not examined any specimens and that some aspects of the morphology of this species were aberrant for Reticulipeurus.Here, we describe this group as a separate genus, Pelecolipeurus gen.nov., based on specimens examined from two hosts in China and examination of photos and illustrations of Lipeurus longus Piaget, 1880.We tentatively re-describe the only previously-known species (L.longus) and add a second species, Pelecolipeurus fujianensis sp.nov.
Given that this new genus is the third Oxylipeurus-complex genus to be described in recent years from the same host group, we also take this opportunity to summarise what is known about host-associations amongst ischnoceran lice parasitising galliform hosts.Finally, we update the key to the genera of the Oxylipeurus-complex previously published by Gustafsson et al. (2020b).
Description.Both sexes.Male longer than female (Table 1).Head longer than wide, frons rounded to slightly flattened (Figs 3,17).Dorsal pre-antennal suture present, but often not well-defined and visible as pale band across head; suture not reaching lateral margins of head.Interior thickenings of pre-antennal head present as double, undulating carinae anterior to suture.Head chaetotaxy as in Figs 3, 17; mds may be absent in female; s2 (?) located median to s1; s5 absent; s6-8 present; mts3 only temporal mesoseta, but os may be longer than pos and mts1-2 in males.Antennae sexually dimorphic, with male scape and pedicel elongated and swollen compared to female (cf.Figs 3,4,17,18); male flagellomere I with distal, finger-like extension and intensely scaly inner surface (Figs 3,17).Temples rounded, somewhat bulging.Thoracic and abdominal segments as in Figs 1, 2, 15, 16.Legs II and III much longer than legs I; coxae I-II close together.Meso-and metasterna fused.Metepisternum long, reaching almost to mesometasternum.Pronotum with lateral and posterior setae; pteronotum with microsetae in antero-lateral corners and short seta submedianly in distal half; mms in single sublateral bunch; mths and mtrs roughly dorsal.Tergopleurites II-VII in both sexes medianly interrupted; male tergopleurites IX-XI fused into a medianly continuous plate; female tergopleurites IX-XI fused laterally, but not medianly (Figs 2,16).Male subgenital plate with lateral extensions in distal section (Figs 7,15); stylus slender, elongated, tapering, attached subterminally and extending beyond distal margin of abdomen.Female subgenital plates reduced to near vulval margin; exact extent of these often not clearly visible.Leg chaetotaxy as in Figs 9-14.
Female.Vulval margin deeply and narrowly concave (Figs 8,22).Three sets of genital setae: long, slender vms, the more median setae shorter than the more lateral setae; short, slender or lightly stout vss in median part of vulval margin; single seta on each side situated further submarginally and apart from vss.Subvulval sclerites present, slender and elongated, reaching to vulval margin.
Geographical range.All known species are from China or the Himalayas, corresponding roughly to the combined range of the known hosts.
Etymology.The name Pelecolipeurus is derived from "pélekus", Greek for "two-headed axe" and the traditional name for long slender lice, Lipeurus Nitzsch, 1818.This refers to the shape of the male subgenital plate.
Remarks.Gustafsson et al. (2020a) tentatively included Lipeurus longus Piaget, 1880, in Reticulipeurus Kéler, 1958, following von Kéler (1958) and Złotorzycka (1966).They noted that they had not examined any specimens and that this placement was doubtful, based on the illustrations published by Clay (1938a) andvon Kéler (1958).The examined collection at NNHM includes two different species belonging to the same morphological group as L. longus and these are sufficiently different morphologically from all other members of the Oxylipeurus-complex that the erection of a separate genus is warranted.
Unfortunately, no specimens from the type host of Lipeurus longus were found at NNHM and no specimens of this species have been examined from other collections.A lectotype and five paratypes are available at the Natural History Museum, London (NHML), but we had no opportunity to examine or borrow these.A photo of the lectotype female at the NHML homepage (https:// data.nhm.ac.uk/dataset) confirms that this species belongs to Pelecolipeurus, but is insufficiently detailed to compare adequately with the specimens we have examined at the NNHM.Only two modern illustrations of L. longus have been published (Clay 1938a;von Kéler 1958), both of which depict the ventral view of the distal end of the male abdomen.Allowing for individual variation and differences in illustration techniques, we cannot separate the specimens illustrated in these publications from specimens we have seen from Tragopan temminckii (Gray, 1831) (see below) and these specimens are here tentatively considered conspecific with L. longus; however, this will need to be confirmed by comparison with type specimens of L. longus and a re-description of this species.
As the type specimens of L. longus could not be examined, we select the species that could be examined as the type species of Pelecolipeurus.
Included species.Pelecolipeurus fujianensis sp.nov.Type host: Tragopan caboti (Gould, 1857).Pelecolipeurus longus (Piaget, 1880: 370)  Diagnosis.Due to the limited illustrations published for Pelecolipeurus longus from the type host (see above), we here compare P. fujianensis sp.nov.with the specimens tentatively identified as P. longus from T. temminckii, which we consider conspecific with the species illustrated by Clay (1938a) andvon Kéler (1958).A re-description of P. longus from the type host is necessary to determine additional characters separating this species from P. fujianensis.
Pelecolipeurus fujianensis can be separated from P. longus as illustrated by Clay (1938a) andvon Kéler (1958) by the following characters: male fused abdominal segment IX-XI with more or less straight lateral margins in P. longus, but with concave lateral margins in P. fujianensis (Fig. 7); proximal mesosome of P. longus with flattened anterior margin, but with medianly pointed anterior end in P. fujianensis (Fig. 6); parameres more curved in P. longus than in P. fujianensis (Fig. 6).
In addition, P. fujianensis can be separated from the population from T. temminckii described above by the following characters: frons more flattened in P. longus s. lat.(Fig. 17) than in P. fujianensis (Fig. 3); male sternal plate VI with 2 sts of more or less equal length in P. longus s. lat.(Fig. 15), but with lateral seta on each side much shorter than median seta on each side in P. fujianensis (Fig. 1); female sternal plate VI with 1 sts on each side and sternal plate VII with 3 medium-length setae and up to 2 microsetae on each side in P. fujianensis (Fig. 2), but sternal plate VI with 2 sts on each side and sternal plate VII without microsetae in P. longus s. lat.(Fig. 16); male subgenital plates of different shape (cf.Figs 7, 21) and stylus evenly tapering distally in P. longus s. lat.(Fig. 21), but with convex lateral margins in distal half in P. fujianensis (Fig. 7); female subgenital plate medianly continuous in P. longus s. lat.(Fig. 22), but medianly interrupted in P. fujianensis (Fig. 8); proximal mesosome with flattened to slightly concave anterior margin in P. longus s. lat.(Fig. 20), but with pointed anterior margin in P. fujianensis (Fig. 6); ventral sclerite of mesosome and shape of gonopore and distal mesosome also differ between species (cf. Figs 6,20).Male antennal characters may be more similar in these two species than illustrated here (Figs 3,17), as their shape is affected by mounting.However, scape appears to be broader and the distal process of flagellomere I appears to be longer in P. fujianensis (Fig. 3) than in P. longus s. lat.(Fig. 17).
Description.Both sexes.Head shape and structure as in Fig. 3; frons gently rounded.No prominent reticulation on head.Marginal carina of moderate width, not widening posteriorly.Dorsal pre-antennal suture prominent, not reaching marginal carina laterally.Head chaetotaxy as in Fig. 3; most dorsal sensilla visible as microsetae in most examined specimens.Antennae sexually dimorphic.Thoracic and abdominal segments and chaetotaxy as in Figs 1, 2.
Male.Antennae as in Fig. 3; scape, pedicel and flagellomere I swollen and modified in shape compared to female; scape with slight process in proximal third; flagellomere I with prominent distal projection and restricted rugose area, which does not extend to proximal bulbous process of segment.Abdominal chaetotaxy as in Fig. 9; inner ventral ps present on segments V-VIII; median sts on sternite VI much longer than lateral sts.Subgenital plate, stylus and terminalia as in Fig. 7; stylus broadening in distal half, not tapering evenly.Genitalia as in Figs 5, 6. Proximal mesosome with narrow median point, widening distally.Ventral sclerite small, roughly rounded-rectangular, with minute postero-lateral extensions; 1 sensillum on each side associated with sclerite; 3 sensilla on each side lateral to ventral sclerite, forming distally divergent rows.Distal mesosome oval, dominated by large oval gonopore.Parameres curved slightly medianly, with median and lateral fingers of parameral head roughly equal in size.Measurements as in Table 1.(Gould, 1857).Male habitus, dorsal and ventral views.(Gould, 1857).Female habitus, dorsal and ventral views.
Female.Antennae as in Fig. 4. Abdominal chaetotaxy as in Fig. 2; sternal plate VI with 1 sts on each side.Subgenital plate, vulval margin and terminalia as in Fig. 8; subgenital plate divided medianly.Vulval margin with 17-23 medium-length, slender vms and 6-10 short, slender vss on each side; median vms shorter than lateral vms.Measurements as in Table 1.
Pelecolipeurus longus (Piaget, 1880), comb.nov.Diagnosis.Both specimens from T. temminckii and those illustrated from the type host by Clay (1938a) and von Kéler (1958) can be separated from P. fujianensis sp.nov.by the following characters: male fused abdominal segment IX-XI with more or less straight lateral margins in P. longus, but with concave lateral margins in P. fujianensis (Fig. 7); proximal mesosome of P. longus with flattened anterior margin, but with medianly pointed anterior end in P. fujianensis (Fig. 6); parameres more curved in P. longus than in P. fujianensis (Fig. 6).Specimens from T. temminckii can be further separated from P. fujianensis by the characters listed under this species above, but examination of specimens from the type host of P. longus is necessary to establish whether the population on this host can also be separated from P. fujianensis by the same characters and whether the populations on T. satyra and T. temminckii are conspecific.
Description (of specimens from Tragopan temminckii).Both sexes.Head shape, structure and reticulation pattern as in Fig. 17; frons somewhat flattened.Marginal carina of moderate width, widening posteriorly.Dorsal pre-antennal suture prominent, reaching to or nearly to marginal carina laterally.Head chaetotaxy as in Fig. 17  Male.Antennae as in Fig. 17; scape, pedicel and flagellomere I swollen and modified in shape compared to female; scape with seemingly hyaline, broad process in proximal third; flagellomere I with intensely rugose surface and intensely rugose bulbous process near proximal base.Abdominal chaetotaxy as in Fig. 15; inner ventral ps absent on all tergopleurites; sts on sternite VI of about equal length.Subgenital plate, stylus and terminalia as in Fig. 21; stylus tapering more or less evenly towards distal end.Genitalia as in Figs 19, 20.Proximal mesosome flattened to slightly concave, with short, stout antero-lateral extensions bent slightly anteriorly.Ventral sclerite inverse V-shaped, with up to 3 sensilla on each side associated with its distal margin; 3 sensilla on each side lateral to ventral sclerite, forming distally divergent rows.Distal mesosome rounded rectangular, dominated by large, roughly rounded-trapezoidal gonopore.Parameres roughly parallel; pst1-2 as in Fig. 20.Measurements as in Table 1.Female.Antennae as in Fig. 18.Abdominal chaetotaxy as in Fig. 16; sternal plate VI with 2 sts on each side.Subgenital plate, vulval margin and terminalia as in Fig. 22; subgenital plate continuous medianly.Vulval margin with 17-22 medium-length, slender vms and 8-12 short, slender vss on each side; median vms shorter than lateral vms.Measurements as in Table 1.
Remarks.We have not seen any specimens of L. longus from the type host.The original illustrations (Piaget    (Gray, 1831).17.Male head, dorsal and ventral views; 18. Female antenna, ventral view; 19.Male genitalia, ventral view; 20.Distal male genitalia, ventral view.
1880; figs 8, 8a, 8b) are insufficiently detailed to establish its identity beyond placing it in Pelecolipeurus.Clay (1938a: 171, figs 33b, 35a) and Kéler (1958: fig. 34) illustrated the male terminalia and male genitalia of this species, confirming our placement of L. longus in the new genus Pelecolipeurus.Kéler (1958: 327-333) provided some additional morphological details, but did not consider it possible to separate this species from Reticulipeurus, even subgenerically.Złotorzycka (1966) placed P. longus in Reticulipeurus; this was followed tentatively by Gustafsson et al. (2020a), but they did not examine any specimens and noted that the species may need to be moved to a separate genus.
Piaget (1880) treated specimens from both hosts as conspecific.Specimens of Pelecolipeurus from Tragopan temminckii are similar to P. longus (Piaget, 1880) as illustrated by Clay (1938a) and Kéler (1958), but published illustrations and descriptions (Piaget 1880;Clay 1938a;von Kéler 1958) are insufficient to establish the status of these populations adequately.As in illustrations of L. longus, the male specimens from T. temminckii have largely flat lateral margins of abdominal segment IX-XI, suggesting they are conspecific.However, Clay's illustration of the male mesosome of L. longus (Clay 1938a: fig. 35a) indicates that there may be differences in the shape of the mesosome between L. longus and specimens we have examined.In our previous experience with Clay's illustrations in this publication (e.g.Gustafsson et al. (2020a)), details such as these are not always reliable when compared to specimens examined by Clay, presumably due to limitations of the microscopy and illustration techniques of the time.Clay (1938a: 172) stated that she had also examined specimens from T. temminckii and Tragopan melanocephalus and that it was "impossible to say whether [P.longus] normally occurs on these two hosts or whether Piaget's specimens are stragglers".
A photo of the lectotype female of P. longus is available online at the NHML's homepage (https://data.nhm.ac.uk/ dataset).In this photo, the distal claspers of the abdomen are more attenuated than illustrated here and the sclerotisations following the vulval margin may be narrower, but this is not clear in the photo.Moreover, these characters may be affected by mounting or be subject to individual variation within the Oxylipeurus-complex and cannot be used alone as reliable indicators of species identity.Other characters, such as vulval chaetotaxy, cannot be seen in the photo.A re-examination of the lectotype and the six paralectotypes of P. longus will be necessary to establish the identity of specimens from T. temminckii (and other hosts) listed here, but, unfortunately, we were not able to either examine the specimen at the NHML, nor borrow this specimen.
We presently consider populations from both T. satyra and T. temminckii to be conspecific, but note that P. longus from the type host is in need of re-description and that such a re-description may warrant the recognition of the specimens described here as a separate species.

Discussion
Galliforms have some of the most diverse chewing louse faunas of any bird orders.Price et al. (2003) recognised 21 genera of lice from galliform hosts, whereas Mey (2009) recognised a total of 64 genera from the same hosts.Including the new genus described here, an additional eight genera of lice have been described from galliform hosts since 2009, all except two in the Oxylipeurus-complex (Mey 2010(Mey , 2013;;Gustafsson andZou 2020a, b, 2023;Gustafsson et al. 2020aGustafsson et al. , b, 2023)).In the checklist of Price et al. (2003), lice on galliform hosts represent almost 10% (21 of 212) of all avian louse genera accepted as valid; if genera accepted by Mey (2009) and those described from other host groups since 2003 are added (in total 49 genera; for example, Mey (2004); Gustafsson and Bush (2017); Gustafsson et al. (2020a, b)), this would imply that over 23% (72 of 302) of the known louse genera occur on galliform hosts, despite Galliformes itself comprising ~ 2.66% of bird diversity (290 of 10906 spp.; Clements et al. ( 2022)).
Clearly, the diversity of lice on galliform hosts is disproportionate to the diversity of host species in this group.The reasons for this over-diversity are unclear.Galliformes constitutes an ancient lineage of birds, with fossil records going back to perhaps the late Cretaceous (Clarke 2004;Agnolin et al. 2006).However, age itself does not necessarily indicate that a host group should have a diverse louse fauna.For instance, the closely-related anseriforms comprise 180 species (~ 1.65% of bird diversity; Clements et al. (2022)), but are only parasitised by 11 (Price et al. 2003) or 14 (Mey 2009) louse genera, constituting 3.61% or 4.61% of louse genera, respectively; note that taxonomic richness of lice is lower in diving than in non-diving birds (Felsõ and Rózsa 2006), which may affect this comparison.
Probably, as more becomes known of the lice of galliform hosts, clear patterns may emerge in the distribution of these louse genera that could explain the unexpectedly high diversity of lice on landfowl.However, some patterns are already dimly visible in the known distribution of lice on these hosts.In Table 2, we list the distribution of ischnoceran lice parasitising galliforms, roughly following the classification of Mey (2009), but with some modifications, based on our own examinations of specimens (DRG, unpublished data).It should be noted that no characters are known that can consistently separate the Goniodes-and Goniocotes-complexes as currently circumscribed (Gustafsson et al. 2021) and Johnson et al. (2011) found both complexes to be paraphyletic.The structure of the male genitalia may ultimately be useful for defining the Goniocotes-complex, but this complex is likely nested inside the Goniodes-complex as defined by Price et al. (2003; as Goniodes), Mey (2009) and here.
Each of the groups of lice included in Table 2 show different patterns of distribution and the Ischnocera of galliform hosts include both generalist genera occurring over several of the major radiations and genera that are more restricted.Of particular interest are the lice of tragopans and allies (genera Lophophorus Temminck, 1813, Tetraophasis Elliot, 1871 and Tragopan).These three genera together form a monophyletic clade, with no close relatives (Meng et al. 2008;Bao et al. 2010;Liu et al. 2014;Wang et al. 2014Wang et al. , 2017;;Kimball et al. 2021).We here refer to this group as the "tragopan group" for simplicity.

Oxylipeurus-complex
The Oxylipeurus-complex is widely distributed across galliforms, being absent only from numidid hosts (Table 2); the genus Afrilipeurus Mey, 2010, was originally described from species known from numidid hosts, but this genus appears to be a member of the Lipeurus-complex (see below) and only superficially similar to lice in the Oxylipeurus-complex.Amongst the genera within the Oxylipeurus-complex, most are restricted to certain host groups.
Reticulipeurus also has wider host associations than most other genera in this complex, being known from both Cracidae and Phasianidae II-III.However, the species known from Phasianidae III belong to a different subgenus (Gustafsson and Zou 2023).The species from cracid hosts have not been revised in recent years and may represent a separate radiation.Reticulipeurus, as currently understood, may represent a plesiomorphic morphotype, from which other, morphologically more distinct, groups of Oxylipeurus-complex lice, have evolved.If not, the distribution of Reticulipeurus on two distinct host groups -mainly Asian phasianids and almost entirely Neotropical cracids -requires further study to understand.
Similarly, Gustafsson et al. (2020a) noted that the widely distributed genus Megalipeurus slightly differs morphologically amongst different host groups and may also represent several distinct lineages.Most other genera are more restricted: Eiconolipeurus and Epicolinus on odontophorid hosts, Labicotes on cracid hosts, Trichodomedea on cracid and odontophorid hosts and Calidolipeurus, Gallancyra and Valimia being known from one  Kimball et al. (2021); host taxonomy follows Clements et al. (2022).Associations are based on Mey (2006Mey ( , 2009Mey ( , 2010Mey ( , 2013)), Gustafsson andZou (2020a, b, 2023), Gustafsson et al. (2020aGustafsson et al. ( ,b, 2023) ) and here; note that the Goniodes-, Goniocotes-and Lipeurus-complexes have not been comprehensively revised since Clay (1938aClay ( , 1940) ) and von Kéler (1940) and some of these taxa may not form meaningful groups.In Goniocotes Burmeister, 1838 (sensu Price et al. (2003)), four morphologically distinct groups are denoted by Roman numerals; due to the lack of detail in original descriptions of many species in this genus, it is not possible to assess whether these groups represent distinct genera or just well-marked species groups.In Goniodes Nitzsch, 1818 (sensu Price et al. (2003)), nine morphologically distinct groups are denoted by Arabic numerals; if Goniodes is divided as suggested by Mey (2009), these groups would represent separate genera for which no genus name has ever been published.Note that Price et al. (2003) used a more conservative classification, in which all genera were placed as synonyms of Goniocotes, Goniodes, Lipeurus Nitzsch, 1818and Oxylipeurus Mjöberg, 1910, except for Pachyskelotes Kéler, 1940and Passonomedea Carriker, 1944.Some genera treated as synonyms of Goniodes by Price et al. (2003) are probably closer to Goniocotes (see Mey (1997)), based on the morphology of the male genitalia.Conversely, it seems likely that Pavoniocotes Gustafsson et al., 2023 and the groups denoted Goniocotes III-IV here are more closely related to the Goniodes-complex than to the Goniocotes-complex.For ease of reference, the position of these species follows Price et al. (2003); these genera are marked with an asterisk (*) in the list.A few species of Goniodes cannot be identified from their available illustrations and descriptions and are here entered as "unknown".

Oxylipeurus-complex
Goniocotes host genus each.The perplexing distribution of the genus Chelopistes was discussed in detail by Mey (2006).
Notably, the genera in the tragopan group are hosts to three genera of Oxylipeurus-complex that are, so far, not known from hosts outside that clade (Cataphractomimus, Pelecolipeurus, Sinolipeurus).The distribution of lice in these genera on the hosts of this radiation is summarised in Table 3.In at least one case, lice from all three genera are known from the same host species, echoing the radiation into three congeneric species of the genus Valimia on the same host species (Gustafsson and Zou 2020b).To date, there is no example of all three genera occurring on the same host individual.However, data from any galliform host are rather limited, not least because many birds in this radiation are protected.Examinations of birds in, for example, rescue centres may be necessary to establish whether the three Oxylipeurus-complex genera on hosts in the tragopan group ever co-occur on the same host individual and, if so, if they then partition the plumage amongst them.

Goniocotes-complex
Lice in the Goniocotes-complex are conspicuously absent from both the mainly New World host radiations, Odontophoridae and Cracidae, as well as from all New World genera in the other host radiations.The sole exception is the turkey, which is sometimes parasitised by Goniocotes gallinae (Linnaeus, 1758), normally found on domestic chicken.Goniocotes gallinae never seems to be reported from wild turkey in their native range (e.g.Hightower et al. (1953); Kellogg et al. (1969) 2014)) and this host association is likely based only on domestic birds which have been in contact with domestic chicken.Based on current knowledge, the Goniocotes-complex would, thus, seem to be an exclusively Old World radiation.
Based on the structure of the male genitalia, lice of the Goniodes-complex, listed from megapodiid hosts in Table 2, are likely more closely related to Goniocotes than to Goniodes.Mey (1997) circumscribed the genera on megapodiid hosts as a distinct group, but excluded the one known Goniocotes species from this group.If this group is considered part of the Goniodes-complex (as by, for example, Price et al. (2003)), it must be considered an aberrant group within this genus.The only known Goniocotes species from a megapodiid host may, as Mey (1997) pointed out, be evidence either of a secondary infestation or of a relict association.
Goniocotes sensu lato is widely distributed across Old World landfowl (Table 2) and do not show any obvious patterns of distribution.At least six morphologically different groups can be found within Goniocotes, but the relationship between these groups is unclear.Only one of these groups, Goniocotes I, is widely distributed across Numididae and Phasianidae I-III.The poorly-known Goniocotes II group is only known from numidid hosts, where it may overlap in distribution with species in Goniocotes I. Goniocotes III-IV are only known from the Congo peafowl and both groups are poorly known and may not be closely related to the rest of Goniocotes (see Clay (1938b)).A fifth group, only known from peafowl, was recently described as the genus Pavoniocotes Gustafsson et al., 2023.
The distribution patterns of the sixth group, previously called Dictyocotes Kéler, 1940, mirrors that of the three Oxylipeurus-complex genera summarised in Table 3, being found mainly on hosts in the tragopan group.However, some species of Dictyocotes are also known from hosts in the genus Crossoptilon Hodgson, 1838, another high-altitude group of birds, mainly distributed in and around China.The presence of a mesosome in the male genitalia in this group, as well as other morphological characters, suggests that Dictyocotes should be separated from Goniocotes; this will be discussed in more detail elsewhere (DRG, in prep.).

Goniodes-complex
The Goniodes-complex is by far the most diverse of the ischnoceran louse groups known from galliform hosts and almost half (28 of 60; 46.7%) of the groups identified in Table 2 belong to this complex.Of these, at least eight currently have no genus-level name and, with the exceptions of Pachyskelotes Kéler, 1940 andPassonomedea Carriker, 1944, all were treated as members of a highly polytypic Goniodes by Price et al. (2003).To discuss the Table 3. Known distribution of Oxylipeurus-complex lice on Tragopan spp., Tetraophasis spp.and Lophophorus spp.These three host genera form a monophyletic clade with no close relatives (Meng et al. 2008;Bao et al. 2010;Liu et al. 2014;Wang et al. 2014Wang et al. , 2017;;Kimball et al. 2021).Dashes ("---") indicate that no species of lice in this genus has, so far, been described from this host.Records suspected to be stragglers or contaminations (see Pelecolipeurus longus) are not listed here.distribution of morphologically distinct groups within this complex, we here follow Mey (2009) in resurrecting numerous older names within this complex and use the numbers 1-8 to denote some groups that have no available genus names.We deviate from Mey (2009) only in considering Zlotorzyckella Eichler [in Eichler and Vasjukova 1981], 1981, as a synonym of Oulocrepis Kéler, 1940.Note that, as some species in this complex have never been adequately described or illustrated, the exact limits of these proposed genera and groups is in some cases tentative.A small number of species are so poorly described that they are noted as "Unknown" genera in Table 2 and not discussed further here.

Host species
Goniodes-complex lice are unknown from cracid hosts and if the Goniodes-complex genera parasitising megapodiid hosts discussed above are moved to the Goniocotes-complex, no Goniodes-complex lice would be known from members of this host family either.Otherwise, lice in the Goniodes-complex occur across all major radiations of galliforms.However, only three groups within this complex could reasonably be said to be widely distributed: Oulocrepis Kéler, 1940, Solenodes Kéler, 1940and Stenocrotaphus Kéler, 1940.The remaining genera and groups in this complex are known only from single host families or even single host genera (Table 2).Stenocrotaphus is mainly known from numidid hosts and African and South Asian francolins and spurfowl, but has, secondarily, also become established on chicken.
Oulocrepis is more widely distributed, occurring on many different host genera in Phasianidae I-II.Morphological variation, above all, in the male genitalia in this group is large (see, for example, Clay (1940)) and, above all, the type species (Goniodes dissimilis Denny, 1842) is somewhat different from all other species in the group with regards to head shape and male genitalia; however, other characters, such as female genitalia, indicate a close relationship.The genus as circumscribed here seems to be established on hosts in different geographical regions, from the Arctic to Sub-Saharan Africa and, in many cases, seems to occur on the same host species as other Goniodes-complex lice.
Solenodes is a widely distributed group, which as circumscribed here, occurs on hosts from Odontophoridae and Phasianidae I-II.Notably, most of the hosts of species in Solenodes are associated with drier grasslands.The male genitalia of this group are more reminiscent of those of the Goniocotes-complex than those of any other group of Goniodes-complex lice; however, as these genitalia are much reduced in complexity, it is possible that the group is artificial and, in reality, comprises several different lineages.Several species here placed in this genus are poorly described and illustrated and a revision of the group is needed to establish its limits.
The tragopan group of birds is collectively parasitised by four Goniodes-complex genera, of which only one presently has a proposed name: Margaritenes Kéler, 1940; the others are here referred to as Genera 2-4.Amongst these, only Genus 2 and Genus 3 appear to be closely related, sharing similarities in the structure of the male antennae and a unique fusing of the pteronotum and tergopleurite II.Potentially, as these species are studied in more detail, further similarities may be found, but, at present, there seems to be nothing to indicate that all four genera are part of the same radiation within the Goniodes-complex.

Lipeurus-complex
Lice in the Lipeurus-complex are the most morphologically homogeneous amongst the groups of ischnoceran lice occurring on galliforms.Lice in this complex are unknown from all New World hosts, except the turkey, which is parasitised by Lipeurus caponis (Linnaeus, 1758) naturally found on domestic chicken.Two genera in this complex are known from numidid hosts only (but secondarily established on domestic chicken) and three genera are unique to the Megapodiidae (Table 2).Based on the structure of the tergopleurites, female genitalia, male subgenital plate, abdominal chaetotaxy and other characters, it seems likely that Afrilipeurus belongs in this complex (see below); thus, three different genera occur on numidid hosts, although only two genera are known to occur on the same host genus.As with the Goniodes-and Oxylipeurus-complexes, the Lipeurus-complex genera known from megapodiid hosts are unique to that radiation, highlighting the distinction of the louse fauna on megapodes.
No Lipeurus-complex species have been described from any species of Tetraophasis or Tragopan and the only species of the genus Lipeurus known from Lophophorus spp.needs verification and may represent a contamination.As both Lipeurus-and Oxylipeurus-complex lice are of the wing louse ecomorph, it is conceivable that the multitude of Oxylipeurus-complex lice on hosts in the tragopan group have prevented Lipeurus-complex lice from establishing themselves there.However, the louse fauna of many members of the tragopan group remain poorly known and the absence of Lipeurus-complex species on these hosts needs verification.Moreover, the mechanisms of interspecific competition in lice are poorly known and cases are known where the same host species is parasitised by multiple louse species of the same ecomorph (e.g. head lice on common blackbird; Oslejskova et al. ( 2020)).

Other ischnoceran genera
Several smaller groups of ischnoceran louse genera are also known from galliform hosts.Of these, Megapodiella Emerson & Price, 1972, is only known from megapodiid hosts, Colinicola Carriker, 1946, only from odontophorid hosts, Lerwoecus Mey, 2006, only from Lerwa lerwa (Hodgson, 1833) and Galliphilopterus Emerson & Elbel, 1957, only from Arborophila brunneopectus Blyth, 1855.It should be noted that Colinicola may be polytypic, based on the structure of the male genitalia and other characters, but this has no major implications for the distribution of this genus.The remaining two genera, Cuclotogaster Carriker, 1936 andLagopoecus Waterston, 1922, are more widely distributed.
Cuclotogaster is known from hosts in Odontophoridae and Phasianidae I-III; however, the species from New World odontophorid hosts needs verification and may be an introduction following the European colonisation of the Americas.Otherwise, Cuclotogaster is absent from all New World hosts, despite being widely distributed in the Old World.Species of Cuclotogaster from Arborophila spp.are morphologically different from other species, with much narrowed male genitalia and possibly some differences in the tergopleurites and the female genitalia; these characters are poorly studied.The genus has not been thoroughly revised since Clay (1938a) and the overall variation in Cuclotogaster is poorly known.Notably, most known hosts are in Phasianidae I and are associated with drier, open country (e.g.savannah, grassland).Species occurring on hosts outside this radiation also often share the same kind of habitat, suggesting that host-switching between sympatric host species may have occurred.
In contrast, the genus Lagopoecus is mainly known from hosts in the Phasianidae II radiation, with a few species known from hosts in Phasianidae I; at least the association with domestic fowl may be due to straggling in domestic settings.Species of Lagopoecus occur in both the Old and New World and are often associated with more boreal or mountain-or forest-dwelling hosts, but exceptions are known (Table 2).In general, Lagopoecus occurs on lowland hosts in the boreal area, but seems more restricted to mountain-dwelling hosts further south and is largely absent south of the Equator.
Galliforms in the tragopan group are parasitised by lice in the genus Lagopoecus, but no species of Cuclotogaster are known from these hosts.The Lagopoecus species parasitising Lophophorus spp.are morphologically distinct, lacking the dorsal pre-antennal suture, but species known from Tragopan spp.are not similar to these and do not appear to be closely related.The genus Lagopoecus has not been comprehensively reviewed since Clay (1938a) and the patterns of variation are poorly known.Nevertheless, based on our current knowledge, there is nothing to suggest that the Lagopoecus species on tragopan group hosts form a unique radiation within this genus.

Contrasting and overlapping patterns
It is clear from this brief overview that no single factor can be used to explain distribution patterns amongst the Ischnocera that parasitise galliform hosts.Overall, both host phylogeny, host biogeography and host ecology appear to influence the known host associations in the groups included in Table 2.Moreover, in some cases, it is not clear which factors are most important, as several factors overlap.
Undoubtedly, host phylogeny is an important factor structuring host associations in louse communities on galliforms.For instance, there appears to be little overlap between the lice of megapodiid hosts and other landfowl (Table 2), likely reflecting that megapodiids are the sister group of all other galliforms (Kimball et al. 2021).Similarly, many of the groups of lice occurring on numidid hosts do not occur on other host groups naturally (but some have spread to, for example, domestic chicken in domestic settings).Numerous smaller groups are also limited to one or a few closely-related genera, especially in the Goniodes-complex.
There is also a distinct difference between most of the New World and Old World galliforms, with Trichodomedea being shared by two New World host families, but absent on all Old World hosts and most Cuclotogaster and Goniocotes being absent from New World hosts despite being widely distributed across the Old World.Notably, African odontophorids are not parasitised by the same lice as New World members of this family, but by Cuclotogaster, which is widely distributed on other African hosts.
Contrasting with the large-scale biogeographical pattern, some patterns may have more to do with host biotope than with faunal regions.For instance, even if Cuclotogaster is largely limited to hosts in Phasianidae I, the genus also occurs on some members of Phasianidae II that occur in less forested areas, such as Phasianus Linnaeus, 1758 and Perdix Brisson, 1760 (Table 2).Similarly, Oulocrepis is found across both these radiations, often on birds that inhabit more open, grassy areas; Solenodes and Stenocrotaphus also appear to be distributed mainly on hosts in the same type of biotope and include at least some species in other host radiations.
Notably, some patterns cannot easily be explained and may be due to gaps in our knowledge or on incorrect classification of known species.It is, for instance, curious that the widely-distributed genus, Reticulipeurus, should occur on both Old World phasianids and New World cracids, despite all other ischnoceran lice on cracids being specific to the New World.Based on published data, there are no obvious morphological differences between the species on these host groups, although the species of cracids have not been revised in recent decades and few detailed illustrations have been published.The distribution of the genus Megalipeurus is also difficult to understand, but the genus is morphologically heterogeneous and a revision of the group may reveal that the current circumscription is artificial (Gustafsson et al. 2020a).
It is worth noting that elevation may influence distribution patterns.Several distinct genera and groups are known only or mainly to infest high-elevation hosts, such as Lerwoecus and many of the unnamed groups within the Goniodes-complex.In contrast, low-elevation hosts are often parasitised by more widely-distributed louse genera (e.g.Lipeurus, Oulocrepis, Goniocotes I).

Lice of the tragopan group
Most relevant to the taxa described here are those found in the tragopan group, all of which are high-elevation birds within Phasianidae II (Kimball et al. 2021).With the exception of an unconfirmed Lipeurus species occurring on one of the species of Lophophorus, all of the typically low-elevation groups of Ischnocera are absent from hosts in the tragopan group.However, the diversity of lice in this group is considerable.Despite comprising only ten species in three genera, the species in this group are collectively hosts to at least three Oxylipeurus-complex genera (Table 3), as well as a morphologically distinct group of Lagopoecus, four genera within the Goniodes-complex and almost all the known species of Dictyocotes.With the exception of Dictyocotes, all these genera and groups are unique to hosts in the tragopan group.This pattern may also be mirrored in the Amblycera.Price and Beer (1964) considered Colpocephalum tetraophasis Price & Beer, 1964, "rather unique", but did not detail in what way; Amyrsidea impejani Scharf [in Scharf and Price], 1983, was also described as having some distinct morphological characters, rare for the genus.
Possibly, the relevant factors structuring these host associations are a mixture of host phylogeny -all the hosts, except Crossoptilon are closely related -and habitat factors -all the hosts including Crossoptilon are high-elevation birds.The overlap in range varies between species pairs, but, in at least some cases, the hosts in these genera co-occur and may even forage together (Madge and McGowan 2002).More data are needed from all hosts in the genera Lophophorus, Tetraophasis and Tragopan, as well as from other high-altitude galliforms, to evaluate the louse community parasitising these hosts and their relationship to related lice on other galliforms.
Future collections may also help determine the extent to which the three genera of Oxylipeurus-complex lice on T. temminckii co-occur on the same host population.All three species are known from Sichuan Province, China, but from different collection events and areas within this Province.The co-occurrence of three congeneric species of lice in the Oxylipeurus-complex was recently reported from turkey (Gustafsson and Zou 2020b) and cases such as these could give insights into the process of microhabitat partitioning in chewing lice and potentially into the process of higher-level radiation of lice on hosts.
Emended key to the Oxylipeurus-complex Here, we update the genus-level key to the Oxylipeurus-complex previously published by Gustafsson et al. (2020b) and emended by Gustafsson and Zou (2023) after the description of the subgenus Reticulipeurus (Forcipurellus).We here remove the genus Afrilipeurus from this complex, based on the justification below and include the genus Pelecolipeurus.Coni elongated (Fig. 27); male mesosome with prominent V-.... or Y-shaped thickening in distal half (Fig. 35); proximal margin of mesosome with rounded lateral lobes (Fig. 35 (Fig. 48); area between margin of head and secondary carina, densely reticulated, in some species, including ventral surface of eye (Fig. 48); male parameres with pst1-2 situated close together apically (Fig. 49

Removal of Afrilipeurus
We believe it is justified to remove the genus Afrilipeurus from the Oxylipeurus-complex and,instead, place it in the Lipeurus-complex, where it is probably closely related to Numidilipeurus Tendeiro, 1955, which occurs on the same host group.This is based on the following morphological comparisons: 1. Amongst slender-bodied Oxylipeurus-complex genera, abdominal chaetotaxy consists of only sutural setae and principal post-spiracular setae dorsally and only one sternal seta on each side on segments II-V ventrally.In Numidilipeurus, intertergal plates are present on at least male segments III-V.Afrilipeurus has medianly continuous tergopleurites and intertergal plates on male segments III-IV.3. Female genitalia lack distally convergent rows of vulval oblique setae in the Oxylipeurus-complex and this set of setae is often reduced to one or a few short setae on the posterior margin of abdominal segment VII.In Numidilipeurus, these setae are numerous (> 5 per side) and form roughly convergent rows on the ventral side of segments VII-IX+X, which is the same as in Afrilipeurus.Outline of male paramere of Sinolipeurus tetraophasis (Clay, 1938), redrawn and simplified from Gustafsson et al. (2020a); 44.Outline of male terminalia and stylus of Sinolipeurus tetraophasis (Clay, 1938), redrawn and simplified from Gustafsson et al. (2020a); 45. Outline of male paramere of Reticulipeurus (Reticulipeurus) ithaginis (Clay, 1938), redrawn and simplified from Gustafsson et al. (2020a).4. Marginal temporal seta 1 is at least a mesoseta in Afrilipeurus and Numidilipeurus, but always a microseta in the Oxylipeurus-complex. 5. Despite considerable variation amongst genera, the male terminalia in the Oxylipeurus-complex are rather uniform in their basic structure, with a generally rounded ano-genital opening, anterior to which may be a transverse sclerotisation that may be continuous with the subgenital plate; several setae of varying length are situated anterior to this opening and a maximum of one seta on each side (typically none) is situated on the ventral side of the poorly-sclerotised areas postero-lateral to the ano-genital opening.Even in genera such as Pelecolipeurus, where the ano-genital opening is not clearly visible, its position can be judged by the distribution of setae and the anterior sclerotisation and this structure appears to be found even in the genus Labic-otes Kéler, 1940, in which the stylus is absent.The terminalia of the Lipeurus-complex males are more variable, but do not include a transverse sclerotisation and the non-sclerotised areas distal to the subgenital plate may form a longitudinal groove, with multiple setae on each side; this is, for instance, the case in some Numidilipeurus.In Afrilipeurus, there is no transverse sclerotisation and there are multiple small setae on each side lateral to a longitudinal groove.6.The female subgenital plate is never extended much distal to the row of sternal setae of segment VII in the Oxylipeurus-complex, but is extended distal to this row in Lipeurus-complex and in Afrilipeurus.
For these reasons, we here exclude Afrilipeurus from the Oxylipeurus-complex and transfer it to the Lipeurus-complex, where it is probably close to Numidilipeurus.
; many dorsal sensilla very small and difficult to see.Antennae sexually dimorphic.Thoracic and abdominal segments and chaetotaxy as inFigs 15, 16.

Table 2 .
Distribution of ischnoceran lice across different galliform hosts.The host groupings are based on