A review of the European planthopper genus Trirhacus and related taxa , with a key to the genera ofEuropean Cixiidae ( Hemiptera : Fulgoromorpha )

Ten species of Cixiidae, formerly placed in Trirhacus Fieber, 1875, are redescribed and one additional species is described. They belong to seven genera: Trirhacus s. str. with T. setulosus Fieber, 1876 (type species), T. dubiosus Wagner, 1959, T biokovensis Dlabola, 1971 and T. peloponnesiacus sp. n., Apartus gen. n. with A. michalki (Wagner, 1948) comb. n. (type species) and A. wagnerianus (Nast, 1965) comb. n., Nanocixius Wagner, 1939 stat. n. with N. discrepans (Fieber, 1876) comb. n., Neocixius Wagner, 1939 stat. n. with N. limbatus (Signoret, 1862) comb. n., Sardocixius gen. n. with S. formosissimus (Costa, 1883) comb. n., Simplicixius gen. n. with S. trichophorus (Melichar, 1914) comb. n. and Sphaerocixius Wagner, 1939 stat. n. with S. globuliferus (Wagner, 1939) comb. n. The phylogenetic relationships of these genera to other Cixiidae are briefly discussed and a key to the genera of European Cixiidae is provided.

As the species from far-eastern Asia, formerly placed in Trirhacus (cf.Vilbaste, 1968), were already re-assigned to Cixius (Tsaur et al., 1991;Holzinger, 1999 b), Trir hacus s. l. contained ten species from the central and mediterranean parts of Europe at the beginning of this study.
According to Wagner (1939Wagner ( , 1948Wagner ( , 1959) ) and Dlabola (1971Dlabola ( , 1977Dlabola ( , 1980)), two groups of characters were used to define the genus.First, the presence of setiferous tuber cles not only on the veins of the fore wing, but also on the apical border of the wing between the veins, and second, the reduction of the hind wing between costa and cubitus including the presence of a distinct incision at the end of the cubital vein(s).
The number and arrangement of these setiferous tuber cles are very different in different species (see Dlabola, 1977, and Figs 19, 22, 42, 66, 67, 84, 92, 104, 123 and 126).Moreover, these tubercles are also present in Tachycixius Wagner, 1939 and in at least one species of Cixius (see Wagner, 1939;Dlabola, 1980;Remane & Holzinger, 1998).Thus, multiple independent origins of this char acter are conceivable, e. g. by the reduction of the apical part of the fore wing and subsequent fusion of the apical end of the veins and their tubercles with the apical margin of the wing.Therefore, I regard this complex of charac ters as oflimited phylogenetic value.
Both groups of characters formerly used to define Trir hacus s. l. are merely diagnostic, and obviously have a rather limited value for a phylogenetic analysis.On the other hand, the taxa included in Trirhacus s. l. are mor phologically far more heterogenous than those included in many other (monophyletic) cixiid genera.Therefore, a revision of Trirhacus s. l. based on a phylogenetic analysis is needed.

MATERIAL AND METHODS
Almost all species of Trirhacus s. l. are very rare; thus, only few specimens were available for this study.
The material was provided by: Almost all drawings except those of the structures of the inner female genitalia were made from dried specimens.Inner female genitalia were prepared and stained with "Chlorazol E endocuticular staining" as described by Bourgoin (1993), and drawings were made in distilled water.The morphological terminology is that used in recent taxonomic papers (e. g., Tsaur et al., 1988;Lariviere, 1999, in female genitalia Bourgoin, 1993).

MORPHOLOGICAL CHARACTERS OF TAXONOMIC RELEVANCE
The descriptions of several species of Trirhacus s. l. are insufficient, at least for revising the genus.Therefore, redescriptions of all species seemed necessary.
The characters used both for species recognition and classification are shown in Figs 8-145.As in other cixiids, important characters can be found in general habitus (proportions and size), head (shape and propor tions, keels, coloration), pronotum (shape of caudal mar gin), mesonotum (shape of keels), fore wings (proportion, venation, colour pattern), hind wings (venation), and in male genitalia (shape of anal segment, genital styles and aedeagus).

PRELIMINARY PLACEMENT OF TRIRHACUS S. L. WITHIN CIXIIDAE
The female abdomen of cixiids of several taxa narrows evenly towards the anal segment, with the ovipositor curved dorsad, adjacent to the genital segment (see Fig. 2).This configuration is considered to be plesiomorphic within cixiids, as it is also present in other basal Fulgoromorpha taxa.In taxa with the apomorphic state of this character, tergite IX is raised steeply inclined or perpen dicular to the body axis and the abdomen is therefore more or less truncate, with the ovipositor either curved dorsad, but not adjacent, or directed more or less straight caudad (see Figs 1,3,4).At the moment, there is no evi dence for a multiple origin of this configuration.There fore, all cixiids with a truncate female abdomen are likely to be monophyletic.All species of Trirhacus s. l. belong to this taxon.
In many cixiids with a truncate female abdomen, wax plates are present on tergite IX.It seems probable, that species with wax plates have evolved from species with a truncate abdomen.With regard to their ultrastructure, at least two different types of wax plates are present in adult cixiids (see Holzinger et al., 2002).
Within Trirhacus s. l., at least seven species do not have a wax plate, whereas two species belong to the group bearing a wax plate with "Cixius-like" wax pores on tergite IX.If the hypothesis presented above is true, it is evident, that Trirhacus s. l. is polyphyletic.

TAXONOMIC CONSEQUENCES
In contrast to all other species of Trirhacus s. l., females of Neocixius michalki Wagner, 1948 andTrir hacus wagnerianus Nast, 1965 have a wax plate at the end of the abdomen.This wax plate is evenly, covered with a regular pattern of uniform wax pores (Fig. 5, 6), and is divided medially by a flat, dorsoventral keel.The pattern of wax pores is similar to that in other Cixiini (Cixius nervosus, Cixius similis, Tachycixius pilosus, Notocixius magellanicus) and significantly different from that of Pentastirini (see Holzinger et al., 2002).The wax plate is smaller as in Cixius and Tachycixius, and the tran sition between the plate and the lateral abdomen is not more or less at a right-angle (as in Cixius and Tachycixius, see Fig. 1), but smoothly rounded (see Fig. 10).
This special configuration might represent a very primi tive type of wax plate, where the tergite IX is still rounded laterally as in taxa without a wax plate.In this case the median keel can be interpreted as a relict of the melting of the lateral borders of the tergites between the anal segment and the ovipositor.Following this scenario, in "higher Cixiini" the wax plate widened laterally, devel oped a right-angled border, and the median keel vanished completely.
Several character states of the male genitalia of these two species are plesiomorphic: The shape of the anal seg ment (almost symmetrical, with a ventral process forming two small lobes) and of the aedeagus (phallosoma tubular, without keels, apical part comparatively simple and only slightly bent).In contrast, the special shape of the genital style with a longitudinal ridge on the stem, orientated ver tically towards the expanded apical part of the style (see Figs 13,14), is a character state that should be interpreted as apomorphic.This longitudinal ridge obviously helps the male to remain fixed to the female during mating, as it is clamped between gonoplac and gonapophysis VIII during copulation (see Fig. 10).
Based on the above characters, it is evident, that Neo cixius michalki Wagner, 1948 andTrirhacus wagnerianus Nast, 1965 represent a monophyletic taxon apart from other cixiid genera; therefore a new genus, Apartus gen.nov.(type species: Neocixius michalki Wagner, 1948), is erected to accomodate these species.
Currently there is no evidence in support of at least a provisional integration of these taxa into other cixiid gen era.Therefore, all three subgenera are reinstated as sepa rate genera.Autapomorphies of Sphaerocixius Wagner, 1939 stat.n. with Sphaerocixius globuliferus (Wagner, 1939) comb.n. are the unique structures both in the male anal segment (apical ventral process almost completely reduced, new formation of a basal globular process; Fig. 94) and in the aedeagus (special shape of the apical part and in the proc esses ofthe phallosoma, see Figs 97-99).
Autapomorphies of Nanocixius stat.n. are the reduction of the median keel on the clypeus and the presence of a longitudinal ridge on the stem of the genital style (Wag ner, 1939, see Figs 25, 31).The latter is similar to, butfollowing this interpretation -convergent with the con figuration in Apartus.
Neocixius stat.n. is a taxon endemic to Corsica.It is easily determinable by the colour pattern on the head and fore wings.The diagnostic characters mentioned by Wagner (1939) -anal segment simple, three movable spines at the distal end of the phallosoma, phallosoma without dorsal keels -are known from several other cixiid taxa, too, and are therefore of little phylogenetic value.
Trirhacus trichophorus Melichar, 1914, is a large spe cies with very simple male genitalia.The phallosoma is tubular without keels, only a lateral widening is visible on its right side.One movable spine is present on each side.The apical part of the aedeagus is strongly s-shaped in dorsal view.The genital style is very simple.The anal segment is very long and slender with an almost symmet rical ventral process.In females, the abdomen is shaped like an inverse "V", similar to the situation in Sphaero cixius (Fig. 3) and distinctively different from Nanocixius and Neocixius.Again, all characters are excellent from a diagnostic point of view, but either have little phyloge netic value (male genitalia) or are currently not polarisable (female genitalia).As evidence in support of integrating this species in another cixiid genus is not available, provisionally a new genus, Simplicixius gen.n., is erected for this species.
Three previously known and one new species of Trir hacus s. l., all occuring on the Balkan peninsula (see Fig. 7), have several characters that are conspicuously similar and are placed provisionally in Trirhacus s. str.: T. setulosus Fieber, 1876 (type species), T. dubiosus Wagner, 1959, T. biokovensis Dlabola, 1971 and T. peloponnesiacus sp.n.The unifying characters are the presence of a ventral process or crest on the phallosoma, a dorsolateral widening on the left side of the phallosoma, a subapical process on the apical part of the aedeagus, a short, asym metrical anal segment and comparatively short fore wings.
It must be kept in mind, that all these characters are also found in other cixiid taxa.Thus, they could equally be synapomorphies, symplesiomorphies or convergences and are not strong evidence for the monophyly of Trir hacus s. str.
The phylogenetic relationships of all these genera and to other cixiid taxa are unknown at present and a compre-hensive phylogenetic framework of the Cixiini and Cixiidae is awaited.
A key to the genera of the family Cixiidae in Europenot including the Balearic and Azorean Is. -is given below.Type species.Neocixius michalki Wagner, 1948 Included species.Apartus michalki (Wagner, 1948) comb.n., Apartus wagnerianus (Nast, 1965) Etymology.The name is derived from the name of the grant financing this study, APART.Gender: masculine.

Diagnosis.
Oval-shaped species.Female abdomen trun cate, bearing a distinct, dorsoventrally divided wax plate.Male phallosoma tubular, dorsally with a distinct process on the right side.Two movable subapical spines are pre sent on each side.Genital style with a longitudinal ridge on the stem, orientated vertically to the expanded apical part of the style.Anal tube in males dorsoventrally flat tened, ventral process apically forming two small, slightly asymmetrical lobes directed ventrocephalad (Fig. 12).Radius and media in hind wings apically forked.(Wagner, 1948) (Wagner, 1948).8-9 -aedeagus (from Holzinger, 1999a); 10 -mating position (laterocaudal view of female, lateroventral of male); 11 -male anal segment; 12 -ductus receptaculi and bursa copulatrix.Scale bar = 0.25 mm.
Phallosoma tubular, dorsally with a smaller triangular process on the left and a larger triangular process on the right side.Two movable subapical spines are present lat erally on each side (Figs 8, 9).
Female abdomen truncate with a distinct wax plate, divided in two parts by a median keel.Ductus receptaculi with six windings (Fig. 12).
Distribution and ecology.Endemic in the Southern Lime Alps (Austria, Italy, Slovenia), living in thermophilic pine for ests.Adults can be found in spring (Holzinger 1999 a).
Face, vertex, pronotum and mesonotum more or less light-brown, keels whitish.Fore wings hyaline, with dark markings as shown in Fig. 19.
Phallosoma tubular, dorsally with a distinct subapical widening on the left side and a basal lobe on the right side.In addition, two movable subapical spines are pre sent on each side.
The female genitalia of the only female studied are very similar to A. michalki (wax plate divided by median keel, ductus receptaculi with six windings; with minor differ ences in the number and size of the sclerotisized plates in the wall of the vagina).
Face, vertex, pro-and mesonotum brownish, keels lighter.Fore wings hyaline or yellowish-brown, with or without distinct dark bands and markings.
Phallosoma tubular, in some specimens with one or two small teeth ventrally.One movable subapical spine on the right and two on the left side.Apical part of aedeagus bearing a slender tooth on the right side.
Distribution and ecology.Described from Italy, recorded there from Lazio (e. g.Castellani, 1953).Widespread in thermo philic forests of southern and southwestern France, Spain and Portugal (e. g.Ribaut, 1958, and unpublished records in several Museums).Also recorded from Tanger, Morocco (Melichar, 1907).Adults can be found in early spring.*Note.Neocixius Wagner, 1939 was treated as genus by Ribaut (1958) and Servadei (1967).This change of status was not done according to the Nomenclature code, and subsequent authors (Wagner, 1959, Nast, 1965, 1972, 1987, D'Urso, 1995) did not accept this "new" status.Therefore, this change might be considered as proposing a new status or as restoration of an already published status.Description.(See also Wagner, 1939) Total length of male 5.2-5.4mm, of female 5.7 mm.
Ventral part of phallosoma with a large basal ridge bearing two slender processes and a small median process.Two long and slender curved movable subapical spines on the left side, and one on the right side.
Female abdomen truncate, evenly rounded, without wax plate.Inner genitalia see Fig. 41.Etymology.The only known species is restricted to Sardinia.Gender: masculine.
Diagnosis.Unmistakable taxon.Apical veins in fore wings elevated, with distinct dark markings.Cubitus in the hind wing not forked.Junction between frons and vertex smooth, without distinct keel.Median keel of frons very weak, disappearing at the junction with the vertex.Lateral keels of vertex and postclypeus laterally strongly produced.Postclypeus swollen.
Description.(See also Costa, 1888) Total length of male 5.6 mm, of female 6,0 mm Upper part of face black, keels brownish.Clypeus and adjacent part of frons brownish.Only one (weak) keel on transistion between vertex and frons.Dorsal half of genae black, ventral half brownish.Vertex and pronotum medi ally light-brown, laterally darker brownish.Mesonotum dark brown.Fore wings milky white, veins brownish, api cally with very characteristic black markings.
Phallosoma ventrocaudally with a distinct ridge.One movable subapical spine on the right, two on the left side.Ventral process of anal tube almost symmetrical, roundish.
Distribution and ecology.Described from Mt. Narba (on meadows at the end of April, not very rare; Costa, 1888), endemic to Sardinia.The single record from continental Italy (Servadei, 1967) is erroneous (Holzinger, 2000).Ecology unknown.
Etymology.The male genitalia are, compared to other Cixiini species, very simple.Gender: masculine.
Diagnosis.Phallosoma truncate, with a distinct lateral widening on the right side.Two short symmetrical proc esses at the distal end of the phallosoma, one slightly curved movable subapical spine on each side.Apical part of aedeagus in dorsal view distinctly s-shaped.Genital styles spoon-shaped, simple.Anal tube in males long and slender, ventral process smoothly rounded apically.Genital segment of female truncate, shaped like an inverse "V".
Vertex and frons dark brown, keels yellowish brown.Ventral part of face (clypeus and adjacent part of vertex) yellowish brown.Upper part of genae dark brown, ven tral of antennae whitish to yellowish brown.Pronotum and mesonotum medially light brown, laterally darker.Fore wings hyaline with dark granulae and more or less distinct dark markings (Fig. 84).
Phallosoma tubular without teeth or keels, but with a distinct subapical widening on the right side.At its distal end two very short, symmetrical processes.Laterally, one evenly curved movable subapical spine on each side.Apical part of the aedeagus strongly bent when viewed from the side and above.
Phallosoma with a long and slender movable subapical spine and a doughnut-shaped process on the right and a broad movable subapical spine on the left side.Base of apical part of the aedeagus has a bulbous widening ven trally.Diagnosis.Fore wings ovoid, thus specimen oval shaped.Abdomen of females truncate, without wax plate.Phallosoma ventrally has a rigid bifurcate process.Apical part of aedeagus with a large subapical process on the left side.Anal segment in males short and stout, apical process asymmetrical, bearing two large lobes.
Note.In T. dubiosus and T. biokovensis female speci mens were unavailable.The placement in Trirhacus is preliminary; it is based on the above mentioned characters of the male genitalia.It must be remembered, that these characters are mainly of diagnostic importance; their homology and synapomorphy and, derived from that, phylogenetic value cannot be evaluated at this time.(Wagner, 1939).87-88 -head, lateral and frontal view; 89 -male genital segment, ventral view; 90-91 -female abdomen, lateral and ventral view; 92 -habitus, male; 93 -inner female genitalia.Scale bar = 0.5 mm.Nast, 1965) Total length of male 4.5 mm, of female 5.2-5.6 mm.
Phallosoma not elevated dorsally, but has distinct lat eral widenings.Two short symmetrical processes at the distal end of the phallosoma.Phallosoma has one straight, movable subapical spine on each side.Ventrally a slightly s-shaped, forked, rigid process is present.Apical part of Figs 94-99.Sphaerocixius globuliferus (Wagner, 1939), male.94 -anal segment; 95 -right genital style, inner maximum view; 96 -left genital style, lateral view; 97-99 -aedeagus.Scale bar = 0.2 mm.aedeagus slightly curved, with a distinct, long and thin subapical spine on the left side.
Ventral side of phallosoma with a very large process (Figs 116,117).One small and two large movable subapical spines present on the right and one on the left side of the phallosoma.Apical part of aedeagus strongly bent, with a large median process on the left side.bola, 1958, 1971).

Trirhacus peloponnesiacus sp. n. (Figs 135-146)
The male was discovered among unidentified speci mens in the Moravian Museum in Brno.One additional female collected by Prof. Sakis Drosopoulos in the same locality is obviously conspecific.
Etymology.The name is derived from the peninsula where the specimens were collected.
Description and differential diagnosis.Total length of male 4.3 mm, of female 4.8 mm.
Head, pro-and mesonotum uniform brownish, keels yellowish.As in T. setulosus, the shape of the fore keel of the pronotum is slightly different in males and females.Fore wings very similar to T. setulosus: ovoid, hyaline, with irregular dark markings mainly near the veins, apical parts of apical cells brownish.Phallosoma has ventrally a thorny crest strongly bent towards the right side, dorsally has a whitish widening on the left side.Two movable subapical spines are present: a shorter, slightly bent one on the right and a very long, curved one on the left side (see .Apical part of aedeagus on the left side with a broad, flat, basally pointed process.Anal segment asymmetrical, short and stout, apical process ending in two acute lobes . The helix of the ductus receptaculi in the female has more inner and outer windings than the (only one dis sected) female of T. setulosus.(Fig. 146).
This species is slightly smaller than the other species of Trirhacus.It can be distinguished from these and other Cixiidae by the unique shape of the aedeagus, as described above and shown in coll.Asche & Hoch, Museum für Naturkunde, Humboldt Uni versität Berlin, Germany coll.Drosopoulos, Agricultural University of Athens, Greece coll.Holzinger, Ökoteam -Institut für Faunistik und Tierök ologie, Graz, Austria Dipartimento di Zoologia, Universita Napoli, Italia (DZU Napoli) Hungarian Natural History Museum, Budapest, Hungary (HNHM Budapest) Moravian Museum, Brno, Czech Republic (MM Brno) Museum National d'Histoire Naturelle, Paris, France (MNHN Paris) Natural History Museum, London, Great Britain (NHM Lon don) Naturhistorisches Museum, Vienna, Austria (NHM Vienna)