An updated classification of the jumping plant-lice (Hemiptera: Psylloidea) integrating molecular and morphological evidence

The classification of the superfamily Psylloidea is revised to incorporate findings from recent molecular studies, and to integrate a reassessment of monophyla primarily based on molecular data with morphological evidence and previous classifications. We incorporate a reinterpretation of relevant morphology in the light of the molecular findings and discuss conflicts with respect to different data sources and sampling strategies. Seven families are recognised of which four (Calophyidae, Carsidaridae, Mastigimatidae and Triozidae) are strongly supported, and three (Aphalaridae, Liviidae and Psyllidae) weakly or moderately supported. Although the revised classification is mostly similar to those recognised by recent authors, there are some notable differences, such as Diaphorina and Katacephala which are transferred from Liviidae to Psyllidae. Five new subfamilies and one new genus are described, and one secondary homonym is replaced by a new species name. A new or revised status is proposed for one family, four subfamilies, four tribes, seven subtribes and five genera. One tribe and eight genera / subgenera are synonymised, and 32 new and six revised species combinations are proposed. All recognised genera of Psylloidea (extant and fossil) are assigned to family level taxa, except for one which is considered a nomen dubium.


Introduction
Jumping plant-lice or psyllids constitute the superfamily Psylloidea Latreille, 1807 of the hemipterous Sternorrhyncha Duméril, 1806 with world-wide about 4000 described and at least as many undescribed

Results
Taxonomy Class Insecta Linnaeus, 1758Order Hemiptera Linnaeus, 1758Suborder Sternorrhyncha Duméril, 1806 Superfamily Psylloidea Latreille, 1807Percy et al. (2018 presented two mitogenome (mtg) phylogenies that we refer to here as the AN tree ('allnucleotide' tree) and the CC tree ('conserved-codon' tree), as well as a much reduced taxon sampling using a nuclear genome analysis, and a combined mitochondrial and nuclear data analysis. Due to the greater taxon sampling for the mitogenome analyses, we refer mostly to these results here. The results of Cho et al. (2019) are similar to the AN tree. In the main, analyses in Percy et al. (2018) and Cho et al. (2019) had considerable congruence, with notable exceptions discussed below. Both mtg trees are similar and recover the same crown groups. The major difference lies in the basal groupings. The Aphalaridae Löw, 1879, as defined here, is a paraphyletic basal assemblage in the AN tree (also paraphyletic in Cho et al. 2019) and a poorly supported monophylum in the CC tree. Carsidaridae Crawford, 1911 (including Pachypsylla) and Homotomidae Heslop-Harrison, 1958 form a poorly supported sister group in the AN tree and a paraphyletic, basal assemblage in the CC tree. The former hypothesis (i.e., sister family relationship between Carsidaridae (without Pachypsyllinae  and Homotomidae) is supported by two putative morphological synapomorphies (Hollis & Broomfield 1989) and is recovered with stronger support in the nuclear genome data in Percy et al. (2018) as well as combined data in Cho et al. (2019). In both mtg trees, the Mastigimatinae Bekker-Migdisova, 1973 constitutes the sister group to a well supported (94%) clade comprising the Liviidae Löw, 1879, as defined here, and the PTCD clade (Psyllidae, Triozidae Löw, 1879, Calophyinae Vondráček, 1957 Updated classification of the jumping plant-lice Ouvrard (2012), Diaphorina Löw,1880 and Katacephala . This grouping differs from that of Burckhardt & Ouvrard (2012) who included Mastigimatinae in their artificial Calophyidae. For this reason, Mastigimatinae is removed from Calophyidae and given family rank here. This move is supported by Cho et al. (2019) although the phylogenetic placement of Mastigimatinae is not identical. The Liviidae, as defined here, is a poorly supported monophylum in the AN tree and paraphyletic in the CC tree. It is also recovered as paraphyletic in combined data analyses in both Percy et al. (2018) and Cho et al. (2019). In both mtg trees, the PTCD clade is very strongly supported (100%) (consistent with Cho et al. 2019), and Calophyidae Vondráček, 1957 (without Mastigimatinae) constitutes the sister taxon of the remainder of taxa in the PTCD clade with good (AN tree) or poor support (CC tree); notably, an alternative placement of Calophyidae as sister to Triozidae (albeit with mixed support) in combined data analyses in both Percy et al. (2018) and Cho et al. (2019) serves to emphasise that phylogenetic placement within the PTCD clade awaits robust confirmation. The support of the monophyly of Psyllidae (including Diaphorina and Katacephala) is good (AN tree) or poor (CC tree) and that of Triozidae very strong in both trees (99%). Again, due to ambiguity in the placement of Diaphorina in the combined data analysis in Percy et al. (2018), additional analyses will be required for robust confirmation. In summary, not all taxonomic groups recognized here are strongly supported as monophyla in all or any of the molecular analyses, in some cases we have erred on the side of providing a practical and stable classification, particularly where ambiguity in molecular analyses remains. A summary of family interrelationships adopted here is shown in Fig. 1.   Fig. 1. Cladogram representation of the classification of Psylloidea Latreille, 1807 adopted here, node symbols indicate families with poor to moderate support (white), or strong support (black) in molecular analyses (Percy et al. 2018).

Comments
In both mtg trees, Aphalaridae contains six strongly supported monophyla which we rank as subfamilies: Aphalarinae, Microphyllurinae subfam. nov., Phacopteroninae stat. nov., Rhinocolinae Vondráček, 1957, Spondyliaspidinae Schwarz, 1898 and a clade of seven undescribed species from New Caledonia representing an unnamed genus and subfamily. This last subfamily is not further treated here and will be described in another paper (Percy, unpublished). There is evidence (from multiple molecular analyses) that these six subfamilies are likely not collectively monophyletic, however, there is still insufficient data to clarify the phylogenetic placement of each monophyletic subfamily with respect to the others, and therefore, rather than recognize each as a separate family, we have retained them as subfamilies within Aphalaridae "sensu lato" pending further analyses. In Aphalaridae, we also place Togepsyllinae Bekker-Migdisova, 1973 andCecidopsyllinae Li, 2011 stat. nov. which were not included in the molecular analyses by Percy et al. (2018) but representatives were analysed by Cho et al. (2019). A morphological character shared by all constituent subfamilies, and putative synapomorphy for the family, is the tarsal arolium of the immatures which is either completely absent or forms a lobe lacking an unguitractor (Burckhardt & Ouvrard 2012).
Apart from strong support of the sister group relationship of Microphyllurinae subfam. nov. (as "Parapaurocephala" in Percy et al. 2018) and Phacopteroninae stat. nov., there are no consistent and well-supported relationships between the subfamilies in the molecular analyses by Percy et al. (2018). A putative morphological synapomorphy grouping the Rhinocolinae, Spondyliaspidinae and Togepsyllinae is the tubercular or knob-like meracanthus rather than horn-shaped as in the other aphalarid subfamilies and most other Psylloidea. Luo et al. (2017) listed some putative synapomorphies suggesting a close relationship of Rhinocolinae and Togepsyllinae, a relationship which was also shown in Drohojowska's (2015) trees based on an analysis of the thorax morphology, and recovered in the molecular data set of Cho et al. (2019).
Aphalaridae, in the present definition, differs from that of Burckhardt & Ouvrard (2012) in the positions of Cecidopsyllinae, Microphyllurinae subfam. nov., Pachypsyllinae and Phacopteroninae. Cecidopsylla Kieffer, 1905, was assigned to Calophyidae (Mastigimatinae) and is transferred here to Aphalaridae (Cecidopsyllinae). Microphyllurus Li, 2002, the only member of Microphyllurinae subfam. nov., was treated as a junior synonym of Peripsyllopsis Enderlein, 1926 (Liviidae: Euphyllurinae: Diaphorinini) by Burckhardt & Ouvrard (2012), whereas the "'Paurocephala' longicella group", which we consider here a synonym of Microphyllurus (see below), was referred to Aphalaridae (Rhinocolinae). Pachypsyllinae was part of Aphalaridae and is transferred here to Carsidaridae. Phacopteroninae was considered a family of basal position within Psylloidea, and a basal position for Phacopteronidae as sister to the remaining Psylloidea was strongly supported in Cho et al. (2019); this is one of the notable differences with analyses in Percy et al. (2018). It may be that the different taxon sampling strategies were critical in determining these results, but here we have elected to adopt the placement using the more comprehensive taxon sampling in Percy et al. (2018).

Comments
Aphalarini comprises the extant members of the subfamily and is probably monophyletic. It has been diagnosed by Loginova (1964), Brown & Hodkinson (1988) and Burckhardt & Queiroz (2013). The phylogenetic relationships between the 16 recognised genera were analysed by Burckhardt & Queiroz (2013). In the molecular analyses of Percy et al. (2018) six of the genera were included. The molecular analyses share with the morphological tree by Burckhardt & Queiroz (2013) the basal position of Colposcenia and the sister group relationship of Aphalara and Craspedolepta (the latter was recovered also by Cho et al. 2019).

Comments
The poorly defined Paleopsylloidini † comprises seven Eocene genera. Ouvrard et al. (2013) suggested that the tribe may be paraphyletic with respect to Aphalarini which includes only recent representatives.
In an overview of Hemiptera represented in the Insect Limestone (latest Eocene) of the Isle of Wight, UK, Szwedo et al. (2019) listed the tribes Aphalarini and Palaeoaphalarini †. They placed Paleopsylloides † Bekker-Migdisova, 1985, type genus of Palaeoaphalarini †, in the former tribe rather than in the latter, which is an obvious oversight.

Comments
The monotypic Cecidopsyllinae was diagnosed by ; for diagnoses of Cecidopsylla see also Burckhardt (1991b) and Yang et al. (2009). In the study of Cho et al. (2019), Cecidopsylla forms the sister taxon to all other psyllids, except for Pseudophacopteron (Phacopteroninae). For this reason, we transfer it to Aphalaridae and remove Cecidopsyllinae from synonymy with Mastigimatinae (Burckhardt & Ouvrard 2012). Within the Mastigimatinae as defined by Burckhardt et al. (2018b), Cecidopsylla resembles Synpsylla in the shape of the head, antennae and forewings but there are no detailed synapomorphies suggesting that the two are phylogenetically close. Awaiting new evidence, we leave the latter in the Mastigimatidae (see discussion there).

Diagnosis
Adult Head with subtrapezoidal vertex smoothly passing into genae that lack processes. Frons triangular. Antenna about as long as head width. Clypeus flattened, triangular. Propleurites with subequal epimeron and episternum. Tibiae distinctly longer than femora; metacoxa with small pointed meracanthus and small membranous lobe on metatrochanteral cavity; metatibia without genual spine, bearing an open crown of 8-9 densely spaced, sclerotised, apical spurs; metabasitarsus with 2 spurs. Forewing weakly coriaceous, covered in surface spinules; costal break and pterostigma developed; veins R and M+Cu subequal, branches of vein M, and vein Cu 1a very long; anal break close to apex of vein Cu 1b . Hindwing with costal setae not grouped; vein R+M developed. Male proctiger one-segmented; in profile, with large posterior lobe in basal half. Female terminalia cuneate.

Comments
The Phacopteroninae is strongly supported as a monophylum in both mtg trees and morphologically White & Hodkinson 1985). The subfamily corresponds to the concept of Burckhardt & Ouvrard (2012). The genera are poorly defined and their phylogenetic relationships are unknown.

Comments
The Spondyliaspidinae is strongly supported as a monophylum in both mtg trees, morphologically (Burckhardt 1991a) and by the pattern of sperm formation (Labina et al. 2014). The concept of the BURCKHARDT D. et al., Updated classification of the jumping plant-lice subfamily is the same as that by Burckhardt & Ouvrard (2012). In the mtg analyses there is a wellsupported basal split between Ctenarytaina Ferris & Klyver, 1932 and the remainder of the subfamily (Anoeconeossa Taylor, 1987, Australopsylla Tuthill & Taylor, 1955, Blastopsylla Taylor, 1985, Boreioglycaspis Moore, 1964, Cardiaspina Crawford, 1911, Creiis Scott, 1882, Cryptoneossa Taylor, 1990, Glycaspis Taylor, 1960, Lasiopsylla Froggatt, 1900and Platyobria Taylor, 1987. Tribe *Ctenarytainini White & Hodkinson, 1985, stat. rev., sensu novo Comments Burckhardt (1991a) suggested that the tribe Ctenarytainini sensu White & Hodkinson (1985) and Taylor (1990) is probably not monophyletic. The mtg analyses confirm this. Here we define the tribe in a new sense by the presence in the adult of a longitudinal comb of bristles on the mesotibia and a knob-like meracanthus, the latter character probably being a symplesiomorphy.

Comments
The mtg analyses strongly support the monophyly of a group of ten genera which lack, in the adult, a longitudinal comb of bristles on the mesotibia and a proper meracanthus. The latter character probably constitutes a synapomorphy. Within the tribe there is a strong support for a clade embracing Australopsylla, Cardiaspina, Creiis, Glycaspis and Lasiopsylla and the sister group relationship of Anoeconeossa and Cryptoneossa. The sister group relationship of Creiis and Lasiopsylla (as currently defined) is only weakly supported with Creiis paraphyletic with respect to Lasiopsylla. The two genera differ morphologically only in the shape of the forewing. There are another 11 genera which are not included in the molecular analyses.

Syncoptozus
Family *Calophyidae Vondráček, 1957 Comments Burckhardt & Ouvrard (2012) admitted the artificial nature of their Calophyidae comprising five subfamilies. Two of these, Calophyinae and Mastigimatinae were included in the molecular analyses of Percy et al. (2018) and Cho et al. (2019), which both confirmed nonmonophyly of Calophyidae. The Mastigimatinae is removed here from Calophyidae and raised to family status. The other four subfamilies lack all metabasitarsal spurs. In addition, Atmetocraniinae Becker-Migdisova, 1973, Calophyinae and Metapsyllinae Kwon, 1983 bear an internal comb of apical metatibial spurs suggesting they may be closely related. Atmetocraniinae and Calophyinae share also the one-segmented asymmetric antennal flagellum in immatures (Burckhardt & Mifsud 2003;Burckhardt & Ouvrard 2012). With this, admittedly weak, evidence we keep the four subfamilies in the Calophyidae awaiting evidence to the contrary.

BURCKHARDT D. et al., Updated classification of the jumping plant-lice
Subfamily Symphorosinae  Comments The subfamily was diagnosed by . Burckhardt & Ouvrard (2012) pointed out the similarity of the male subgenital plate of Symphorosus and Cecidopsylla (classified here in Aphalaridae: Cecidopsyllinae) but did not list detailed synapomorphies between the two genera.

Comments
In both mtg trees the monophyly of Carsidaridae and Pachypsylla is strongly or very strongly supported. Carsidaridae + Pachypsylla (Pachypsyllinae) and Homotomidae form a poorly supported sister group in the AN tree and both families are included in an unresolved basal assemblage in the CC tree. In the molecular analyses by Cho et al. (2019), Celtisaspis (Pachypsyllinae) is recovered as weakly supported sister-group of Homotomidae, rather than Carsidaridae, and the monophyly of Carsidaridae, Celtisaspis and Homotomidae is well supported. Hollis & Broomfield (1989) listed two putative morphological synapomorphies to link Carsidaridae and Homotomidae: 1) the presence of a pair of large tubercles on the metapostnotum, and 2) all three ventral sense organs of the metafemur in a basal position. These characters are present also in Pachypsyllinae, though the tubercles on the metapostnotum are relatively small in Celtisaspis. Cho et al. (2019) mention the bipartite male proctiger as a putative synapomorphy of Homotomidae + Pachypsyllinae. As in Aphalaridae, the three taxa discussed here form, depending on the type of analysis, an unresolved basal (paraphyletic) assemblage or a monophylum with contradicting relationships between the constituent groups. For reasons of consistency, we include the three groups in the single family Carsidaridae. The concept of Carsidaridae by Burckhardt & Ouvrard (2012) is broadened here to include also Homotomidae and Pachypsyllinae which is transferred from Aphalaridae. Subfamily *Carsidarinae Crawford, 1911 Prionocnemidae Scott, 1882: 466, invalid as not derived from an included genus name. Tenaphalarini : 577-578. Mesohomotomini Bekker-Migdisova, 1973: 101.

Comments
In both mtg trees the monophyly of the subfamily is very strongly supported. Hollis (1987) provided a morphological diagnosis and analysed the intra-subfamily relationships. In the mtg analyses, with five of the eight recognised genera included, Mesohomotoma is in a moderately supported basal position, whereas in the morphological tree it is the sister taxon of Paracarsidara; and apart from a reasonably well supported clade comprising Protyora + Tenaphalara + Paracarsidara, the relationships between the other genera are only poorly supported.

Comments
The subfamily and its constituent tribes and subtribes were diagnosed by Hollis & Broomfield (1989) (treated as family, subfamilies and tribes) who also analysed the generic relationships of the family using morphological evidence; the three subfamilies (here tribes) were represented in a trifurcation in their cladogram. The molecular analyses, in which only three genera were included, reflect this morphological tree but sampled only two subfamilies (here tribes). Following Ouvrard (2002), the classification of Burckhardt & Ouvrard (2012) differs from that of Hollis & Broomfield (1989) in the inclusion of Phytolyma Scott, 1882 in the Macrohomotominae White & Hodkinson, 1985 (Phytolymini) rather than in the Aphalarinae. The classification presented here reflects that of Burckhardt & Ouvrard (2012) though with reduced ranks.

Comments
Differs from other subtribes in the Macrohomotomini in the presence of a costal break in the forewing and small tubercles on the metapostnotum (Burckhardt et al. 2018a). In Cho et al. (2019), Moriphila is nested in Homotoma.

Comments
In both mtg trees, this poorly supported monophyletic or paraphyletic family contains two strongly supported monophyla and one monotypic taxon which we rank as subfamilies: Euphyllurinae, Liviinae and Neophyllurinae subfam. nov. There is no strong support for any particular sister group relationship, although in a backbone constraint analysis in Percy et al. (2018), Neophyllurinae subfam. nov. grouped more strongly with Liviinae than Euphyllurinae . The family as defined here differs from that of Burckhardt & Ouvrard (2012) in that it lacks the Diaphorinini (minus Megadicrania and Psyllopsis, which are included here in the Euphyllurinae). Adults of Liviidae often have a crown of densely spaced apical spurs and immatures have multiple lanceolate or sectasetae.

Comments
The Euphyllurinae as defined here is strongly supported as a monophylum in both mtg trees, morphologically it is, however, more difficult to diagnose. All species included here have immatures with a fan-shaped tarsal arolium bearing an unguitractor. Hosts are, as far as known, Ericaceae (Ericales), Oleaceae (Lamiales), Polygonaceae (Caryophyllales), Rutaceae and Sapindaceae (Sapindales), and Salvadoraceae (Brassicales). Its present concept, which is not further subdivided into tribes, embraces the constituents of the tribes Euphyllurini (except Neophyllura), Pachypsylloidini and Strophingiini of Burckhardt & Ouvrard (2012), as well as Megadicrania, Peripsyllopsis and Psyllopsis (from Diaphorinini).

Comments
The monophyly of Liviinae is strongly supported in both mtg trees and also morphologically (Burckhardt & Mifsud 2003). The concept of the subfamily is the same as that by Burckhardt & Ouvrard (2012). Both morphologically and in the molecular analyses Diclidophlebia Crawford, 1920 and Paurocephala are closely related. Whereas each of the genera was recovered as monophyletic in a morphological study (Burckhardt & Mifsud 2003), in the molecular analyses Diclidophlebia is paraphyletic with respect to Paurocephala.

Fifth instar immature
Caudal plate with additional pore fields.

Comments
The taxon was diagnosed by Burckhardt & Ouvrard (2012) and Burckhardt et al. (2018b). The former suggested the following sister group relationships: Bharatiana + Mastigimas and Cecidopsylla + Synpsylla, and Burckhardt et al. (2018b) indicated that Toonapsylla may be closely related to the former clade. supporting the monophyly of the three genera. Synpsylla resembles Cecidopsylla in the shape of the head, antennae and forewings but there are also important differences such as absence/presence of submedian ridges on the metapostnotum or number and arrangement of the apical metatibial spurs. Awaiting new evidence, we leave Synpsylla in the Mastigimatidae.

Comments
The monophyly of Psyllidae is well (AN tree) or only moderately (CC tree) supported in the molecular analyses. The two mtg trees share the same internal topology with most clades strongly supported (Fig. 4). The composition of Psyllidae proposed here differs from the Psyllidae of Burckhardt & Ouvrard (2012) in the addition of Diaphorina and Katacephala (from Liviidae, Euphyllurinae) as two distinct, basal clades, though inclusion of Diaphorina in Psyllidae requires further testing (Percy et al. 2018). The molecular analyses further suggest that the subfamily Psyllinae of Burckhardt & Ouvrard (2012) is polyphyletic. Here we remove Amorphicola and Platycorypha and assign them each to a new subfamily. An unnamed taxon from Madagascar, also warranting subfamily status, is not further treated here as it contains no described genus and species. Fig. 4. Cladogram representation of the classification of Psyllidae Latreille, 1807 adopted here, node symbols indicate subfamilies with poor to moderate support (white), or strong support (black) in molecular analyses (Percy et al. 2018). For Katacephalinae subfam. nov. no node symbol is given as only a single species was included in the analyses.

Comments
The monophyly of Acizziinae is strongly supported in both mtg trees.

Fifth instar immature
Body flattened, broadly oval. Antenna 7-segmented, beset with a few short club-shaped setae. Forewing pad lacking humeral lobes, bearing short marginal club-shaped setae. Margin of hindwing pad with club-shaped or capitate setae. Tarsal arolium lacking pedicel. Caudal plate with 4+4 marginal sectasetae and some moderately long club-shaped setae. Anus in ventral position; circumanal ring heart-shaped, consisting of a single row of pores, lacking additional pore fields.

Fifth instar immature
Body flattened, broadly oval. Antenna 7-segmented, beset with a few short club-shaped setae; bearing a single subapical rhinarium on each of segments 3 and 5, and 2 rhinaria on segment 7. Forewing pad lacking humeral lobes, bearing short marginal club-shaped setae. Margin of hindwing pad with clubshaped or capitate setae. Legs bearing club-shaped or capitate setae; tarsal arolium longer than claws, fan-shaped with unguitractor but lacking pedicel. Caudal plate developed, semi-circular; margin with 4+4 sectasetae and some moderately long club-shaped setae. Anus in ventral position; circumanal ring heart-shaped, consisting of a single row of pores, without additional pore fields.

Comments
The monophyly of Aphalaroidinae as understood here is strongly supported in both mtg trees and it is well circumscribed morphologically (Burckhardt 1987). The molecular analyses place Telmapsylla, which was previously included in the Aphalaroidinae (Burckhardt & Ouvrard 2012), in Ciriacreminae; we therefore transfer this genus to Ciriacreminae. Burckhardt & Ouvrard (2012) treated Primascena † as a synonym of Diclidophlebia (Liviidae: Liviinae). However, the presence of metabasitarsal spurs and a crown of spaced apical spurs on the metatibia in P. subita , the type species of Primascena †, indicate a relationship to Aphalaroidinae, as suggested by Burckhardt & Mifsud (2003). We follow the latter authors and recognise Primascena †, stat. rev., as a valid genus in the Aphalaroidinae.

Comments
The monophyly of Ciriacreminae is strongly supported in both mtg trees and its circumscription corresponds mostly to that of Burckhardt & Ouvrard (2012) with the addition of Telmapsylla Hodkinson, 1992, Caradocia Laing, 1923, Epipsylla Kuwayama, 1908and Geijerolyma Froggatt, 1903 which are included here. However, all five Old World and two additional New World genera listed by Burckhardt & Ouvrard (2012) are not included in the molecular analyses. In both mtg trees, the monophyly of Auchmerina, Euceropsylla + Heteropsylla, of Telmapsylla + 'Limbopsylla' lagunculariae and of Mitrapsylla + 'Limbopsylla' nigrivenis, as well as the sister group relationship of the first two groups is very strongly supported.
The artificial nature of Limbopsylla was acknowledged by Brown & Hodkinson (1988) when erecting the genus. The type species is a member of Platycoryphinae subfam. nov. (see there for more details). Two species of 'Limbopsylla' included in the molecular analysis are placed in the Ciriacreminae: 'Limbopsylla' lagunculariae and 'Limbopsylla' nigrivenis. We transfer here the former to Telmapsylla and assign the second to the new genus Hollisiana gen. nov. (see Table 1 and description below).
'Limbopsylla' lagunculariae (Brown & Hodkinson, 1988) comb. nov. and Telmapsylla minuta Hodkinson, 1992 constitute a very strongly supported monophyletic clade in both mtg trees. Morphologically, the two species share the head with a trapezoidal vertex, the lack of anteorbital tubercles, the hemispherical, adpressed eyes, the conical, apically pointed genal processes, the large cu 1 cell of the forewing, the BURCKHARDT D. et al., Updated classification of the jumping plant-lice presence of a M+Cu vein in the hindwing, the grouped apical metatibial spurs and the short female terminalia in the adult; and immatures with 7-segmented antenna, fan-shaped tarsal arolium with an unguitractor and a short petiole, and 4+4 marginal sectasetae on the caudal plate. Despite many morphological differences between the two species, such as antennal length, absence/presence of a genual spine and details of the male terminalia in the adult, and the number of rows of pores in the circumanal ring, we consider them congeneric and propose the following new combination: Telmapsylla lagunculariae (Brown & Hodkinson, 1988) comb. nov. from Limbopsylla.
Here, we move three genera which were not treated in the molecular analyses from Liviidae: Euphyllurinae (Diaphorinini) to the Ciriacreminae. Caradocia, Epipsylla and Geijerolyma constitute together a putative monophyletic group based on the presence of long genal processes, very long antennae, metatibia lacking a genual spine but bearing an open crown of densely spaced apical spurs, and two sclerotised spurs on the metabasitarsus. According to White & Hodkinson (1985), immatures of Epipsylla lack sectasetae on the abdominal margin. While we confirm this for an Asian species (Thailand, NHMB), this is not the case for Afrotropical taxa (material examined from Cameroon, NHMB), where the immatures are similar to those of Mitrapsylla with four sectasetae present on the abdominal margin, two grouped together and each situated on a small tubercle, and with the circumanal ring extending to the abdominal dorsum and consisting of several rows of wax pores. The combination of the last two characters can be found only in the Ciriacreminae.

Etymology
This genus is dedicated to David Hollis for his outstanding contribution to psyllid systematics.

Description
Adult See also Brown & Hodkinson 1988: figs 63-64 for illustrations. Moderately large psyllids, 2.5-5.0 mm long. Head about as wide as mesonotum, inclined at 45° from longitudinal body axis (Fig. 6A-B). Vertex trapezoidal, about 1.8 times as wide as long along midline, weakly indented around foveae; passing smoothly into genae not separated by transverse suture; genae produced into long conical processes which are covered in conspicuous long setae; median suture complete, reaching hind margin of head; lateral ocelli on small tubercles; frons forming small rhomboid sclerite, delimited by vertex and genae, almost completely covered by median ocellus; compound eyes relatively small, hemispherical, stalked on large preocular sclerite and occiput (Fig. 6C-D). Clypeus hidden by genae in lateral view, pear-shaped; rostrum short, only apex visible in lateral view. Antenna filiform, longer than forewing, 10-segmented, in some species flagellum getting thinner towards apex; flagellum beset with long conspicuous bristles; segment 3 shorter than segments 7 or 8, with a single subapical rhinarium on each of segments 4, 6, 8, and 9; terminal setae shorter than segment 10. Thorax weakly arched dorsally; lacking macroscopic setae. Pronotum transversely ribbon-shaped.
Propleurites about as broad as high, slightly oblique; proepimeron as big as or larger than episternum. Forewing oval, broadly, irregularly rounded apically, transparent, more than twice as long as wide; pterostigma lacking; costal break present, indistinct; cells m 1 and cu 1 large; anal break close to apex of vein Cu 1b . Hindwing slightly shorter than forewing; costal setae grouped; vein R and M+Cu. Metacoxa with large, horn-shaped, pointed meracanthus; metafemur slightly shorter than metatibia; metatibia bearing genual spine and 1+3+1 apical spurs. Metabasitarsus with two lateral spurs. Male proctiger unipartite, tubular or with posterior lobe. Subgenital plate elongate. Paramere slender, lamellar or digitiform. Aedeagus long and thin; distal segment shorter than paramere, inflated in apical third; sclerotised end tube of ductus ejaculatorius short, slightly sinuous. Female terminalia, in profile, cuneate, moderately short to relatively long. Circumanal ring oval, consisting of two subequal rows of pores. Valvulae triangular and lacking serrations. (Fig. 6E-H) Body elongate, about twice as long as wide (Fig. 6G-H). Antenna 9-segmented, sparsely beset with a few short setae; bearing a single subapical rhinarium on each of segments 4, 6, 8 and 9. Forewing pad small, lacking humeral lobes, bearing short marginal club-shaped setae. Margin of hindwing pad with short bristles. Legs long, with at least one moderately long capitate seta on tibiae; tarsal arolium about twice as long as claws, fan-shaped with unguitractor and pedicel. Abdomen slender; caudal plate weakly sclerotised; abdominal margin with 6+6 sectasetae, the two at the rear close together and each on a small tubercle and, in some species, distinctly larger than the remainder. Anus in terminal position; circumanal ring extending to the abdominal dorsum and consisting of several rows of wax pores.

Comments
Hollisiana gen. nov. is similar to Mitrapsylla from which it differs in the absence of a pterostigma in the adults; and the narrow abdomen with 6+6 marginal sectasetae (rather than 4+4) in the fifth instar immature.

Comments
The AN tree (but not the CC tree) shows reasonably strong support for a sister group relationship of Diaphorina to the remainder of Psyllidae, and, as noted earlier, reduced taxon nuclear genome and combined data analyses by Percy et al. (2018) place Diaphorina outside Psyllidae and sister to Triozidae. Increased taxon sampling and more analyses are required to robustly resolve this ambiguity, and therefore erection of a separate family is currently rejected in favour of inclusion in Psyllidae at this time. Based on the morphology of head, forewings and male terminalia we consider Parapsylla the sister group of Diaphorina and include it in the Diaphorininae. Subfamily *Katacephalinae subfam. nov. urn:lsid:zoobank.org:act:DE4F5C17-6685-47B9-9659-CEFD8828244D Fig. 7 Type genus Katacephala Crawford, 1914.

Diagnosis
Adult Head with genae forming conical to lobular processes; preocular sclerite developed. Antenna 10-segmented, 0.9-2.2 times as long as head width, segment 3 usually longer than segments 7 or 8. Metatibia usually without genual spine, bearing an open crown of 6-19 evenly spaced, sclerotised, apical spurs; metabasitarsus usually with 2 spurs. Forewing with costal break and large pterostigma; anal break close to apex of vein Cu 1b . Hindwing almost as long as forewing. Male proctiger one-segmented, in profile, often with posterior lobes. Paramere usually simple with stout setae on the inner face.

Fifth instar immature
Body oval to elongate, fairly robust; surface often covered in lanceolate setae or sectasetae but lacking capitate setae. Antenna 7-10 segmented with 4 rhinaria. Dorsal thoracic sclerites varying from small to large. Tarsal arolium fan-shaped, unguitractor developed, pedicel absent or present. Forewing-pads often with large humeral lobes. Anus in ventral or terminal position. Circumanal ring variable.

Comments
The monophyly of Macrocorsinae is very strongly supported in both mtg analyses. The four genera treated in the analyses are also assigned to this subfamily by Burckhardt & Ouvrard (2012) who included another eight genera. Meanwhile, Euphaleropsis and Peregrinivena were synonymised (Burckhardt et al. 2018b). Trisetipsylla was placed in the Macrocorsinae by Cho et al. (2019) rather that in Psyllinae (Burckhardt & Ouvrard 2012). At least some species of Trisetipsylla have immatures with extra pore fields on the caudal plate, supporting this placement.

Description Adult
Head, in profile, weakly to strongly inclined at 30-90° from longitudinal body axis (Fig. 8A, C, E, G). Vertex subrectangular to transversely subtrapezoidal (Fig. 8B, D, F, H); separated from genae sometimes by transverse or oblique suture, sometimes passing smoothly into genae; genae smoothly rounded (Fig. 8D, H) or forming short (Fig. 7F) or long conical processes (Fig. 8B); coronal suture fully developed or completely reduced; frons oval, almost completely covered by median ocellus; anteorbital tubercle sometimes developed (Fig. 8D: arrow). Antenna 10-segmented, filiform, ranging from slightly longer than head width to distinctly longer than forewing, segment 3 shorter or longer than segments 7 or 8. Clypeus pear-shaped, in profile hidden by genae and not visible. Rostrum usually short, only European Journal of Taxonomy 736: 137-182 (2021) tip exceeding procoxae, sometimes longer (in some Platycorypha spp.). Thorax weakly (Fig. 7G) to strongly (Fig. 8A) arched dorsally, about as wide as head; pronotum weakly or very strongly inclined from longitudinal body axis; propleurites narrow to broad, with episternum subequal to or smaller than epimeron. Legs moderately long, tibiae often shorter than femora, sometimes subequal or longer; basitarsi not much longer than broad; metacoxa with large, pointed, horn-shaped meracanthus; metatibia with large genual spine, bearing 4-5 irregularly spaced, sclerotised, apical spurs; metabasitarsus with 2 spurs. Forewing rhomboidal, broadest in apical third or in the middle, narrowly rounded or angular apically; membrane semitransparent, covered in surface spinules; costal break and pterostigma developed; vein C+Sc weakly or strongly widened; vein R longer than M+Cu; cell cu 1 large; caudal break close to apex of vein Cu 1b . Hindwing slightly shorter than forewing, membranous; costal setae ungrouped or grouped; vein R+M+Cu indistinctly trifurcating or splitting into R and M+Cu. Male proctiger one-segmented; in profile, tubular or with posterior lobe. Male subgenital plate subglobular or elongate. Paramere lamellar or complex. Female terminalia short or moderaly long; proctiger often with dorsal hump.

Fifth instar immature
Body broadly oval, lacking capitate setae. Antenna 7, 9 or 10-segmented, with 4 rhinaria. Meso and metathoracic sclerites small. Forewing pad lacking humeral lobes. Margin of hindwing pad usually with one sectaseta. Legs lacking capitate setae; tarsal arolium shorter or longer than claws, fan-shaped with unguitractor and pedicel. Caudal plate developed, semi-circular; margin with up to 3+3 sectasetae. Anus in ventral or terminal position; circumanal ring small heart-shaped to large undulate, restricted to ventral side or extended to dorsal side; consisting of a single row or multiple rows of pores, without additional pore fields. Brown & Hodkinson (1988) created Limbopsylla as a polyphyletic holding place for ten "species of the subfamilies Acizziinae and Ciriacreminae which cannot be placed in existing genera". Three species have been removed previously and three species are transferred here (Table 1). Based on adult and immature material of an undescribed species from Brazil associated with Tachigali rugosa (Fabaceae) (NHMB) which is congeneric with L. nata, the type species of Limbopsylla, we conclude that Limbopsylla is a valid genus closely related to Platycorypha, and that the other species included in Limbopsylla (Table 1) are not congeneric with the type species. Allophorina Hodkinson, 1991;Limbopsylla Brown & Hodkinson, 1988;Padaukia Hollis & Martin, 1993 (syn. Peltapaurocephala Heslop-Harrison nomen nudum, no description); *Platycorypha Tuthill, 1945 (syn. Neopsyllia).

Comments
In both trees, the monophyly of Psyllinae and the division into three monophyletic subgroups is very strongly supported. The most basal clade consists of a single Asian species, viz. Cacopsylla eriobotryae (Yang, 1984). The second clade contains four West Palaearctic genera associated with faboid Fabaceae: Arytaina Foerster, 1848, Arytainilla , Arytinnis Percy, 2003 and Livilla Curtis, 1835. The first three are monophyletic, the last is paraphyletic with respect to Arytainilla and Arytinnis, as previously shown by . Livilla ulicis Curtis, 1836, the type species of Livilla, belongs to a very strongly supported clade which is sister group to a poorly supported Arytainilla. Livilla blandula (Horváth, 1905), a species closely related to Livilla pyrenaea (Mink, 1859), the type species of Floria Löw, 1879, and Livilla radiata (Foerster, 1848), the type species of Alloeoneura Löw, 1879, belong to a clade which is sister group to the very strongly supported Arytinnis. To split Livilla s. lat. into two monophyletic genera, viz. Livilla s. str. and Floria (syn. Alloeoneura), respectively, is not practicable at the moment as only a quarter of the known species (see Ouvrard 2020 for a complete list of species) were included in the molecular analyses and no morphological characters are known reflecting these European Journal of Taxonomy 736: 137-182 (2021) groupings. The third clade comprises a very strongly supported group of North American species associated with Ceanothus L. (Rhamnaceae) Nyctiphalerus Bliven, 1955, see Table 2), previously referred to the genera Ceanothia, Euglyptoneura  and Nyctiphalerus, which is sister group to a poorly supported clade comprising one clade represented by a single species (Pexopsylla cercocarpi Jensen, 1957) and four very strongly supported clades: 1. holarctic species of Cacopsylla Ossiannilsson, 1970 associated with Elaeagnaceae, Lardizabalaceae, Rosaceae and Salicaceae (= Cacopsylla s. str.); 2. holarctic species associated with Betulaceae (= Psylla s. str., Table 3), previously referred to Baeopelma Enderlein, 1926, Cacopsylla, Chamaepsylla Ossiannilsson, 1970 andPsylla;3. palaearctic species associated with Buxus (Buxaceae) (= Spanioneura Foerster, 1848, see Table 4), previously referred to Psylla and Spanioneura; 4. North American species associated with Cercocarpus and Purshia (Rosaceae) (= Purshivora Heslop-Harrison, 1961, see Table 2), previously referred to Cacopsylla, Ceanothia, Nyctiphalerus and Purshivora. Cho et al. (2019) transferred Psylla longicauda Konovalova, 1986, an Asian species associated with Prunus, to Spanioneura and provided morphological adult characters to define Psylla s.str. and Spanioneura.
Similar to Arytaina, Arytainilla, Arytinnis and Livilla, associated with brooms (Fabaceae), which constitute a species-rich clade endemic to the Western Palaearctic, a group of Psyllinae radiated in Western North America on Ceanothus (Rhamnaceae) as well as Cercocarpus and Purshia (Rosaceae). North America authors Tuthill 1943b;Jensen 1956Jensen , 1957aJensen , 1957bBliven 1956Bliven , 1958 assigned these species to the genera Arytaina, Euphalerus and Psylla, rendering these genera very artificial, and to two monotypic genera Nyctiphalerus and Pexopsylla.  discussed the North American genera previously referred to Arytaina and, rightly, concluded that they are not congeneric with Arytaina spartii (Hartig, 1841) (= A. genistae (Latreille, 1804)), the type species of Arytaina. He erected the four genera Amorphicola Heslop-Harrison, 1961, Ceanothia, Euglyptoneura and Purshivora. His descriptions are not diagnostic and he also mixed up the figures ( fig. 2 concerns Ceanothia and fig. 3 Amorphicola, and not vice versa). Whereas Amorphicola (see Amorphicolinae subfam. nov.) is well characterised by its paramere morphology and by its host associations (Fabaceae), the other three genera are not. Hollis & Martin (1997) redefined Euphalerus Schwarz, 1904 and suggested that the Nearctic species are not congeneric with Euphalerus nidifex Schwarz, 1904, the type species, or with most of the Neotropical species. Percy et al. (2012) transferred these species to Nyctiphalerus. The molecular analyses shed much needed light on the phylogenetic relationships in this group. There are monophyletic clades associated with Rhamnaceae (Ceanothus) and with Rosaceae (Cercocarpus, Purshia). The former is characterised by immatures with a terminal anus and a large circumanal ring which extends onto the dorsum of the caudal plate, the latter has immatures with a ventral anus and a smaller circumanal ring restricted to the venter of the caudal plate. The clade of Ceanothus comprises one group with the genal processes in a lower plane to that of the vertex and lacking a genual metatibial spine (type species of Ceanothia and Euglyptoneura), and another group with genal processes and vertex flattened and in the same plane and bearing a genual metatibial spine (type species of Nyctiphalerus). The Rosaceae clade also splits into two groups: one bearing metatarsal spurs (type species of Purshivora) and one lacking metatarsal spurs (type species of Pexopsylla). Here we suggest that Ceanothia, Nyctiphalerus, Pexopsylla and Purshivora are good genera, and that Euglyptoneura syn. nov. is a junior synonym of Ceanothia (Table 2).
The Oriental genus Cornopsylla is transferred here from Liviidae, Euphyllurinae, Diaphorinini to Psyllidae, Psyllinae. The position of Cornopsylla within Psyllidae is supported by morphological (Luo et al. 2013) and molecular characters (Cho et al. 2019); in both papers, Cornopsylla was treated as a member of Psyllinae.
Psyllinae is a species-rich subfamily (ca 800 spp., Ouvrard 2020) with many species referred to Cacopsylla s. str. and Psylla s. str. that do not fit the restricted concepts of these genera provided above. Awaiting more studies on these species, we leave them in Cacopsylla s. lat. and Psylla s. lat.  Nyctiphalerus Bliven, 1955;Purshivora Heslop-Harrison, 1961;and Pexopsylla Jensen, 1957. Taxa sampled in Percy et al. (2018) are indicated with an asterisk.

Comments
With around 70 genera and over 1000 species (Ouvrard 2020), the Triozidae constitutes the second largest family of Psylloidea. Many of the genera are poorly defined and Trioza Foerster, 1848 with over 400 described species has long been recognised as polyphyletic . The molecular analyses confirm the polyphyly of Trioza and the artificial nature of genera such as Kuwayama Crawford, 1911. For a more stable and improved classification, most of the genera have to be redefined and several new genera have to be described to establish monophyletic clades, where no generic name is currently available. This task is beyond the scope of the present paper and awaits further studies.

Comments
The type of the Brazilian Labicria barbata is destroyed (D. Burckhardt, unpubl.) and we have not seen any fresh material fitting the original description. USNM, etc.) await description and poorly known faunas, such as the Afrotropical and Neotropical regions, should be explored with targeted field work.