Revision of European species of the genus Rhabdomastix ( Diptera : Limoniidae ) . Part 1 : Introduction and subgenus

The first part of a revision of the European species of the genus Rhabdomastix Skuse, 1890 is presented. The history of taxonomic research on Rhabdomastix is reviewed, relationships of the genus are discussed, and the subgeneric classification outlined and re-assessed. A new subgenus, Lurdia subgen. n., is established for species centred around R. lurida (Loew, 1873), and Palaeogonomyia Meunier, 1899 and Sacandaga Alexander, 1911, previously considered subgenera, are synonymized with Rhabdomastix. A revision of the European species of Lurdia subgen. n. is presented. Two species are redescribed, Rhabdomastix (Lurdia) lurida (Loew, 1873) and R. (L.) inclinata Edwards, 1938, and the lectotype of the former is designated. Descriptions are provided of seven species, viz. R. (L.) mendli sp. n. (Switzerland, Germany, Italy), R. (L.) sublurida sp. n. (Czech Republic, Slovakia), R. (L.) furva sp. n. (Slovakia), R. (L.) loewi sp. n. (Switzerland, Germany, Austria, Italy), R. (L.) robusta sp. n. (Czech Republic, Slovakia), R. (L.) falcata sp. n. (Switzerland, Germany, Bulgaria) and R. (L.) tatrica sp. n. (Slovakia). Male and female terminalia are illustrated for all the species (except female@lcata), and a key to species is appended.


INTRODUCTION
The genus Rhabdomastix Skuse, 1890 is one of several highly uniform and taxonomically extremely difficult groups in the family Limoniidae.This uniformity is shown not only in external characters, but also in the structure of the male terminalia, a complex normally highly species-specific in Limoniidae.The genus is worldwide in distribution and according to the recent catalogues (Alexander, 1965;Alexander & Alexander, 1970, 1973;Hutson, 1980;Oosterbroek, 1989;Savchenko et al., 1992) includes 106 species and subspe cies.Except for those distributed in Europe, the prepon derance of the species were described by Alexander, and are recorded from all zoogeographic regions, with no spe cial species concentration.Rhabdomastix is slightly more numerous in the Palaearctic (27 species), followed by the Nearctic (26 species).It should be noted that only three species are known from the Afrotropical Region and that of the sixteen Australian/Oceanian species-group taxa, eleven occur in New Zealand.The numbers cited above, however, indicate inadequate knowledge rather than the actual richness of the faunas.
Rhabdomastix was related to various subgroups of what is currently the subfamily Chioneinae (tribe Eriopterini of most North American authors), depending on whether the wing venation or the size of the meron were believed to be of primary importance.The subgeneric classification has gradually developed into a concept of three subgenera, viz.Rhabdomastix s. str., Palaeogonomyia Meunier, 1899 andSacandaga Alexander, 1911, based exclusively on the length of the male antennae.In the pre sent revision, a new subgenus, Lurdia subgen.n., is established for species centred around Rhabdomastix lurida (Loew, 1873), and Palaeogonomyia and Sacandaga are synonymized with Rhabdomastix.Hence, European Rhabdomastix species are classified here in two subgenera, Lurdia subgen.n. and Rhabdomastix s. str.
As may be suspected in a group as uniform as Rhabdo mastix, many species have remained unnamed even in Europe.Within the two subgenera, as defined below, spe cies are mainly distinguished on the basis of structural details of the male terminalia.These features, however slight and variable, can on acquaintance be used to sepa rate species.They should, however, be considered in association with external characters.Once the nature of the difference is recognised, species limits become more evident.
In Europe, Rhabdomastix species are locally restricted to margins of brooks and larger streams, including natural lowland rivers, because they are closely associated with sandy or gravely banks.More extensive shorelines devel oped as bars on inner side of meanders, and overgrown with scattered, low vegetation, is a habitat particularly suitable for Rhabdomastix, as it is for some other chioneine Limoniidae.Some species may be observed sit ting or crawling about on plants or on the ground (cf.also Edwards, 1938: 113), but are rarely seen flying.Aggrega tions sometimes occur below branches of nearby trees, around shrubs or above vegetation.
This paper represents the first instalment of a revision of European species of the genus Rhabdomastix.The his tory of the taxonomic research on this genus is reviewed, relationships of Rhabdomastix are briefly discussed and the subgeneric classification is outlined and re-assessed.A revision is presented of the European species of the subgenus Lurdia subgen.n.In Part 2, European species of the nominotypical subgenus Rhabdomastix s. str.will be treated.
The morphological terminology adopted here essen tially follows McAlpine (1981), except for notation of the wing veins, which is in accordance with Hennig (1954) (cf.Fig. 4).Some special terms, mostly pertaining to genital structures, are referred to in figures  and/or referenced in the text.
Numbers in brackets following Czech and Slovak localities in Material examined sections refer to grid ref erences as defined by Zelený (1972).
Colour characters of species are described from drymounted specimens.Unless otherwise stated in the sec tions Material examined, the material under study is drymounted and collected by netting.Some specimens were dried after preservation in ethanol and mounted on points in the course of the study.
HISTORICAL REVIEW Skuse (1890) erected the genus Rhabdomastix to receive a new species, R. ostensackeni Skuse, 1890, from Australia (N.S.W.), based mainly on the wing venation and long male antennae, "nearly twice the length of the entire body" (Skuse, 1890: 828).Alexander (1911) described the genus Sacandaga Alexander, 1911 for a new species, S. flava Alexander, 1911, from the U.S.A. (New York), now known to be widely distributed in eastern North America (cf.Alexander, 1965), in which the male antennae are quite short, about twice as long as the head (cf.Alexander, 1911, Fig. 1).At that time, Alex ander may have not noticed the similarity in the wing venation of his species and that of R. ostensackeni.Shortly after, however, Alexander (1912) commenting on S. flava again, considered Rhabdomastix and Sacandaga closely related, as, in addition to their wing venation, they are similar in the structure of the male terminalia.He even claimed Sacandaga might well be a subgenus of Rhabdo mastix.Short diagnoses were given of the two taxa in the same paper (Alexander, 1912), with one of the distin guishing characters, listed as the first and evidently regarded as the most significant, being the length of the male antennae.Alexander (1914a) described the first Neotropical Rhabdomastix, R. illudens Alexander, 1914, from Bolivia, a species of quite bizarre appearance, in which the male antennae are extremely long, "between four and five times as long as the body", and "fore and middle trochanters long and slender" (Alexander, 1914a: 210).He classified the species in "Rhabdomastix (Rhab domastix)", which indicated that, first, he had accepted his already announced (Alexander, 1912) lowering in rank of Sacandaga and that, second, of the two possibili ties available, the subgenera Rhabdomastix s. str.and Sacandaga, he chose the former for his species.Subse quently, Alexander established several dozen Rhabdo mastix species from all zoogeographic regions, mostly classified in the above two subgenera, based again mainly on the length of the male antennae.However, in exam ining the world fauna he was faced with species highly diverse in the length of their antennae.Hence, the sub generic criterion was quite arbitrarily set by him as antennae longer than the body in Rhabdomastix s. str.and shorter than the body in Sacandaga (cf.Alexander, 1929).The matter became still more confused when in his comprehensive treatment of the Baltic amber craneflies, Alexander (1931) introduced another subgenus in Rhab domastix, Palaeogonomyia Meunier, 1899, for species in which the male antennae are "a little shorter than the body" (Alexander, 1931: 111), establishing the Lower Oligocene Gonomyia pulcherrima Meunier, 1906 as the type species of the subgenus by subsequent designation.Shortly after, Alexander (1932) assigned, with some doubts, his new species, R. omeina Alexander, 1932, from China (Szechwan = Sichuan), in which the male antennae are "about two-thirds the length of wing" (Alex ander, 1932: 395), to the subgenus Palaeogonomyia as the first (and actually the last) recent species.
The fact that a new species, R. margarita Alexander, 1940, from the U.S.A. (Tennessee), in which the male antennae are "elongate, only a little shorter than the entire body" (Alexander, 1940b: 639), was not placed in a sub genus and classified simply as Rhabdomastix, indicates uncertainty about the limits of the subgenera within the genus.Only in discussing this species was Palaeogono myia mentioned as a possible subgenus; this idea, how ever, was immediately followed by a note indicating a preference "to drop these various subgeneric groups" (Alexander, 1940b: 640).Yet subsequently, Alexander, in describing further species, still distinguished between Rhabdomastix s. str.and Sacandaga, just leaving some species unplaced as to the subgenus.
However, there are some reasons for revising the above concepts.A few apomorphies, such as a medially divided male tergite 9 and peculiarities in the structure of the male terminalia, suggest that Rhabdomastix, possibly together with Atarba, could be placed in the tribe Chioneini (= Cladurini), as delimited by Savchenko (1982) (see also the subtribe Claduraria in the sense of Alexander, 1947a, and other papers).A large meron and three sclerotized spermathecae in females may support this, although the latter feature is not an apomorphy.In any case, the struc ture of the male terminalia in Rhabdomastix is strikingly similar in general plan to that in Neolimnophila Alexan der, 1920, sharing many details, including presence and position of slender elements variously referred to as parameres or interbases in the descriptive literature (see below under Rhabdomastix).Closer study is necessary to eluci date the relationships of Rhabdomastix.

SUBGENERIC CLASSIFICATION OF RHABDOMASTIX
In his more recent papers, Alexander assigned only those species to Rhabdomastix s. str.that were distin guished by exceedingly long antennae and modified tro chanters, as represented by the Bolivian R. illudens, already mentioned and discussed further below.Such spe cies were said to be the "typical Rhabdomastix".However, the type species of Rhabdomastix, the Austra lian R. ostensackeni, indicates something else.I have examined a specimen of the latter species from the Alex ander collection (USNM) [1k, "Eukey, S. Q. / Jan. 1934./ F. E. Wilson", "Rhabdomastix osten-sackeni / Det. C. P. Alexander 193..." (last numeral of the year hand-written and illegible)], kindly loaned by H. B. Williams (USNM).The specimen fits very well the original, detailed descrip tion (Skuse, 1890: 828-829), including the structure of the male antennae, which are nearly twice as long as the body.An examination of the antennae, wing venation, male terminalia, etc., revealed unambiguously that R. ostensackeni is as closely related to the European R. hirticomis, as two species from such remote regions may be.In any case, the two species are, in my opinion, clearly consubgeneric.Similarly, R. hirticornis, except for the length of the male antennae, does not differ in any other essential character from other European Rhabdomastix species, and they definitely are consubgeneric with the type species of Sacandaga, the Nearctic R.flava.The fig ures of the type species of Palaeogonomyia, the fossil Gonomyia pulcherrima (cf.Alexander, 1931, Figs 149, 150, 154, 156), also do not show any essential differences.R3 is somewhat oblique in G. pulcherrima (cf.Alexander, 1931, Fig. 149), suggesting the condition in Lurdia subgen.n. (see below), but some other venational features, such as the length of Sc and the shape of the discal cell, clearly indicate that Palaeogonomyia cannot be identical with the latter.The position of R3 varies from vertical to oblique even in recent species of Rhabdomastix s. str.(sometimes even within a single spe cies).
Consequently, Rhabdomastix Skuse, 1890, Palaeo gonomyia Meunier, 1899 andSacandaga Alexander, 1911 are considered here to be subjective synonyms.The length of male antennae often is highly variable within genera and subgenera of the chioneine Limoniidae, and this character cannot constitute a reason for separation at this level.
On the other hand, the so-called lurida group, previ ously assigned to the subgenus Sacandaga, is described here as a new subgenus, Lurdia subgen.n. Edwards (1938) described Rhabdomastix inclinata Edwards, 1938 from Great Britain, and having examined some of Loew's type specimens stressed its close rela tionship with Gonomyia lurida Loew, 1873, described from Germany.Somewhat later, Lackschewitz (1940) classified the latter species in Rhabdomastix (Sacandaga) and listed additional limited material from Austria and northern Italy.Rhabdomastix lurida has since been recorded also from other European countries (cf.Savchenko et al., 1992).Alexander (1940a) described R. luridoides Alexander, 1940, from northern Korea, classi fying it in the subgenus Sacandaga and stating it was allied to the western Palaearctic R. lurida and R. inclinata.Here, for the first time, the "lurida group" denoted this species cluster.More recently, R. luridoides was recorded also from the Russian Far East (Savchenko et al., 1992).Alexander (1940a) specified some charac ters uniting the group, and mentioned another species from western North America.This species was described subsequently (Alexander, 1943) from the U.S.A. (Wyo ming) as R. neolurida Alexander, 1943, and a subspecies was established at the same time, R. neolurida setigera Alexander, 1943, from the U.S.A. (Colorado).The spe cies was later recorded from an additional U.S. state (Alaska) and from Canada (Alberta, British Columbia), and the subspecies elevated to full specific rank and recorded from Oregon (Alexander, 1954(Alexander, , 1965)).Another subspecies, R. neolurida flaviventris Alexander, 1964, from the U.S.A. (California), was added by Alexander (1964), possibly also a valid species.The close relation ship of this group of six species-group taxa was known.As follows from the diagnosis below, this group differs considerably from Rhabdomastix s. str. in a number of both external and genital characters, which together jus tify proposing a new subgenus.In addition, the two sub genera differ in larval characteristics (cf.Hynes, 1969;Podeniene, 2001).
Apart from Lurdia subgen.n. and Rhabdomastix s. str., which will be defined and their European species revised consecutively, there are two other morphologically sig nificant groups within the genus Rhabdomastix that should be mentioned.
(1) Species centred around the Bolivian R. Uludens (see under Historical review and in this section above) are among the most curious Limoniidae due to the enormous length of their male antennae.Although far beyond the scope of this paper, the group is mentioned here because of its exceptional morphology and to give a better insight into the classificatory changes pre sented.The group may be characterized by the following fea tures, as specified in the relevant literature and confirmed by a specimen I examined of an unidentified species near R. isabella Alexander, 1927 (1k, "Venezuela -Rio Negro / NP Tama (Tachiva) / 16.06.1987/ leg.T. Pyrez"), which was kindly pro vided by Dr. W. Krzeminski (Krakow): Male antenna extremely long, from three to five times as long as body.Scape enormously enlarged, barrel-shaped.Pedicel small, very short.First flagellomere subequal to, to longer than second; following flagellomeres progressively lengthened and more slender towards apex of antenna, distal ones very long.Subterminal flagellomere longest, subequal in length to entire body; terminal flagellomere quite minute.(Alexander did not specify the number of antennal segments; I counted 13 flagello meres in the specimen I examined, including the minute ter minal one).Wing venation much as in Rhabdomastix s. str., with R3 short, rather erect, sometimes virtually vertical in posi tion, with discal cell hexagonal and with distal sections of Mi+2 and M3 strongly arched.Legs with fore and mid trochanters slender, conspicuously elongate (about twice as long as coxae, or more, in the specimen I examined).Male terminalia usual for Rhabdomastix s. str.
The group comprises about eleven Neotropical and Oriental species, and one from the southern Nearctic, viz.R. nuttingi Alexander, 1950, from the southern U.S.A. (Arizona), all classi fied in the subgenus Rhabdomastix s. str.The status of this extralimital group cannot be solved in detail in this paper.In any case, it should be emphasized that the group is cladistically closer to Rhabdomastix s. str.than is Lurdia subgen.n., and is here considered a highly apomorphic offshoot of Rhabdomastix s. str., but not deserving of subgeneric status.
(2) Alexander (1914b) transferred "Goniomyia" (Empeda) caudata Lundbeck, 1898, described from a single female col lected in Greenland, to Rhabdomastix and classified it in the subgenus Sacandaga, listing additional material from Canada (British Columbia).Later (Alexander, 1916), he considered the British Columbian specimens to be distinct from R. caudata and described a new species, R. monticola Alexander, 1916.Subse quently, Alexander named further species, viz.R. subcaudata Alexander, 1927, from Canada (Alberta), R. subarctica Alexan der, 1933, from Canada (Quebec), and R. leptodoma Alexander, 1943, from the U.S.A. (Colorado), all classified in the subgenus Sacandaga and assigned to the group around R. caudata.Addi tional records of R. caudata, R. monticola, R. subcaudata and R. leptodoma were published by Alexander (1965), those of R. caudata in detail by Mannheims & Theowald (1971).In Europe, Lackschewitz (1933), revising Siebke's collection, examined three of the four females described as Limnobia parva Siebke, 1863 from Norway.He classified the species in "Gonomyia (Rhabdomastix)", and synonymized it with G. (R.) schistacea (Schummel, 1829), often listed in the literature at that time, but now considered a doubtful species.Edwards (1938) recorded Rhabdomastix parva from Great Britain, but this was a misinter pretation (cf.Tjeder, 1967).Tjeder (1964b) examined one of the syntypes of R. parva, designated it as the lectotype and pre sented a redescription with detailed illustrations of the female terminalia.The species, as presently known, is restricted to northern Europe, and recorded from Norway, Sweden, Finland and Iceland (Savchenko et al., 1992).The group may be recog nized mainly on wing characters, as described below, based on the relevant literature and material of R. parva (5 2, "2.7.1985So far only the above six species are known to belong to this group, which is boreal in distribution.Based on the above wing characters, the group is closer to Lurdia subgen.n. than to Rhabdomastix s. str., differing from both in that R2 is more or less preserved.This represents a plesiomorphic state.Further distinctions, may be found in the structure of the male and female terminalia.Unfortunately, only a single illustration of the male hypopygium is available (R. leptodoma : Alexander, 1943, Fig. 26).According to this, the outer gonostylus is smooth, bidentate at apex, whereas it is densely spinulose and simple at apex in other Rhabdomastix.The so-called interbases are present in R. leptodoma as pale rods, but the aedeagal com plex is not detailed in the figure.The female terminalia share distinct, sclerotized spermathecae with other Rhabdomastix, a clear plesiomorphic character, but other internal structures, such as the infra-anal (supravaginal) plate and genital fork (vaginal apodeme) are markedly different (cf.Tjeder, 1964b, Fig. 4).Although the group is related to Rhabdomastix, its rank and position remain in question.Therefore, R. parva is omitted from the present revision, mainly because only females are known.
Genus Rhabdomastix Skuse, 1890 Gonomyia pulcherrima Meunier, 1906 (fossil); subsequent designation by Alexander, 1931: 112. -Alexander, 1931: 111 (as  Diagnosis.Small to medium-sized species.Head.Male antenna 16-segmented, of variable length, from very short, about twice as long as head, to notably long, nearly twice as long as body.Flagellomeres variously shaped according to length of antenna, oval, spindle-shaped or moderately elongate, decreasing in size towards apex of short antenna, or cylindrical, rather long, subequal in length, with terminal flagellomere of long antenna very small. Thorax.Anterior pits, as termed so by Edwards (1938: 112), positioned just near anterior margin of prescutum.Pleuron bare, with meron solid and large, widely sepa rating mid and hind coxae.Wing hyaline, sometimes vari ously tinged, stigma and other wing pattern lacking in European species.Wing venation (cf.Figs 4-6): Sc1 ending from shortly beyond origin of Rs to shortly before fork of Rs.Sc2 slightly preserved near tip of Sc1, or entirely lacking.R2 (= cross-vein r) always lacking.R3 short, considerably shifted distally, rather short and straight, from oblique to vertical.Discal cell compara tively small, pentagonal or hexagonal, with m-cu attached at its mid-length.Legs with mid and hind coxae separate, corresponding to enlarged meron, and with fore and mid trochanters slightly lengthened, subequal in length to coxae.Male tarsal claws comparatively long, about half length oflast tarsomere, arolium distinct.
Abdomen.Male terminalia: Segment 9, forming socalled basal ring, interrupted in middle of its dorsal side.Gonocoxite simple, cylindrical.Outer gonostylus arched, heavily sclerotized and darkly pigmented, densely spinulose, especially at outer margin, terminating in curved apical spine, exceptionally blunt, without spine.Inner gonostylus fleshy, mostly pale, setose, generally conical, narrowed distally.Distinct interbase present, generally slender, rod-like or sword-like, variously shaped at apex, usually crossing its counterpart above aedeagal complex.Parameres short and membranous, fused to each other to produce a shelf below aedeagus.Female terminalia with cercus comparatively long and slender; internal structures distinguished by two or three distinct, sclerotized and darkly pigmented spermathecae.
Discussion.Generally, the above diagnosis is intended to cover European species, but it may well fit other Rhab domastix, except for groups (1) and (2) as defined above (Subgeneric classification of Rhabdomastix).
The male terminalia of Rhabdomastix are characterized by possessing a pair of slender elements attached to the dorsal inner base of the gonocoxites and usually crossing each other above the aedeagal complex (cf.Fig. 7, ib).These normally were referred to as the parameres (gonapophyses) in the descriptive literature.However, Alex ander (1943) was not certain whether they were homolo gous with the parameres of other Limoniidae and pre ferred to call them interbases.The latter term is used here.The term paramere is reserved for a different structure, more intimately related to the aedeagus (cf.ib,pm).
The following combination of characters appears to be significant for distinguishing the genus: Anterior pits situ ated just near anterior margin of prescutum; pleuron with meron large, widely separating mid and hind coxae; wing venation with R2 entirely lacking, R3 short, shifted distally, from oblique to vertical, and with m-cu attached at mid-length of discal cell; male terminalia with a pair of slender, usually crossed interbases; female terminalia with distinct spermathecae (2-3).Diagnosis.Predominantly medium-sized species.Head.Male antenna (cf.Figs 1-3) dark, of medium length, subequal to or slightly shorter than combined length of head and thorax.Male flagellomeres, especially proximal ones, mostly more or less spindle-shaped, narrowed at each end, decreasing in breadth towards apex of antenna.First flagellomere pear-shaped, distal ones elongate, ter minal one shorter than penultimate.Setae on antenna well differentiated into long, stiff verticils and shorter, delicate, rather dense, suberect pubescence.Female antenna with flagellomeres oval to elongate.Palpus with terminal segment distinctly longer than penultimate.
Thorax.Wing venation (cf.Figs 4-5): Sci long, ending shortly before fork of Rs.Sc2 lacking.R3 oblique, from one third to one fourth length of R4, forming with R4 angle of about 45 degrees.Discal cell generally pentago nal, with proximal section of M3+4 (forming lower side of Abdomen.Male terminalia (cf.Figs 7-9 and others): Segment 9 broadened posteriorly, excised medially at posterior margin dorsally, less so ventrally.Gonostyli generally longer than those in Rhabdomastix s. str., inner one sometimes tending to be flattened and darkened.Interbase sclerotized and blackened distally, sword-like or spine-like, curved, acute at tip, connected to its counter part through membraneous V-shaped strip at about one third length.Female terminalia (cf.Figs 10, 18 and others) with cercus and hypogynial valve considerably elongate, the former very slender, about twice length of tergite 10, slightly upturned and pointed at apex.Spermathecae three in number, spherical, darkly pigmented, subequal in size and shape.
differs in that the cross-vein m-cu is attached near mid length of the discal cell (retracted, in Gonomyia s. str., to the base of the discal cell or even proximal to it).
The male and female terminalia are of the same general type in both subgenera.The males of Lurdia subgen.n. differ from those of the other subgenus in that segment 9 is broadened posteriorly and excised medially on poste rior margin of its dorsal side (cf.Fig. 7) (generally cylin drical, without any excision in Rhabdomastix s. str.).Although tergite 9 is interrupted medially in both subgen era, the interruption is somewhat less evident in Lurdia subgen.n., the two parts of the tergite being mostly nearly contiguous and rather firmly connected to each other by a narrow membranous zone or line that may be indistinct against the generally pale colour of that region or the seg ment.In Rhabdomastix s. str., segment 9 often is largely and evenly darkened, and the interruption is wider, hence well evident.Interbases are sclerotized and blackened dis tally in Lurdia subgen.n., sword-like, curved and acute at tip, connected to each other through a membraneous strip (cf.Fig. 7) (pale, membranous rods in Rhabdomastix s. str., mostly dilated into variously shaped blades at apex, not interconnected).Female terminalia of Lurdia subgen.n. have the cerci and valves considerably elongate, the former very slender, about twice the length of tergite 10 (cf.Fig. 18), whereas the cerci are distinctly shorter in Rhabdomastix s. str., at most slightly exceeding the length of tergite 10.Generally, in fresh material or in dry-mounted speci mens, two types of colouration may be distinguished on the pleura of European Lurdia species, orange-yellow to yellowish brown (lurida, mendli, sublurida, falcata, tatrica), or various hues of brown, suffused with grey pruinosity (inclinata, furva, loewi, robusta).This distinc tion is useful as the first step in identification.However, in material preserved in ethanol the colouration rapidly unifies into various hues of yellow (according to the stage of maceration), regardless of the original tinge.No sub stantial infraspecific variation in body colouration, geo graphical, sexual or individual was observed.
Although the male terminalia in this subgenus are rela tively uniform, they are nevertheless species-specific, whereas the female terminalia are less distinctive.Internal structures of the female terminalia are here illustrated for each species (except falcata).Distinctions that may be inferred from the figures (cf.Fig. 10 and other relevant figures), such as those showing the outline of the infra anal (supravaginal) plate and the number of setae on its posterior margin, and the shape of the sternum 9 (termed fused valvae of 9th sternum by Tjeder, 1964a) and the genital fork (vaginal apodeme), including adjacent struc tures, are not sufficiently elucidated.These structures may vary to a certain degree and warrant further study.So far, the only useful genital feature in females appears to be the size of the spermathecae combined with the length of sclerotized parts of the ducts.
In general, the distribution of Lurdia species in Europe is inadequately known.Literature records of the two spe cies, R. (L.) lurida and R. (L.) inclinata (summarized by Savchenko et al., 1992) are few and largely suspect because ofthe seven new species.
All European Lurdia species are presumed to be restricted to, or at least to prefer, mountainous areas (least so perhaps inclinata).The material studied suggests a marked difference in the species composition in the Alps and some other, more southerly situated European moun tains on the one hand, and more northern mountainous territories within the Czech Republic and Slovakia, on the other.Complete substitution of species occurs in the two regions.Four species, viz.lurida, mendli, loewi and fal cata, occur in the Alps and south of there, while five spe cies, viz.inclinata, sublurida, furva, robusta and tatrica, occur north of the Alps.Surprisingly, all the four Alpine species occur syntopically at Birksau, Allgovian Alps, Germany.The restriction to mountains may account for the great deal of vicariance within European Lurdia spe cies.Diagnosis.Pleuron orange-yellow to yellowish brown.Femora darkened distally.Male antenna comparatively long, flagellomeres spindle-shaped, with verticils slightly exceeding length of respective segments.Wing moder ately broad.Male terminalia with outer gonostylus gently arched and inner gonostylus pale, rather slender.Aedeagus short, vesica very broad.Female terminalia with spermathecae small and sclerotized parts of ducts very short.
Colour.Basic colouration orange-yellow to yellowish brown.Prescutum and other dorsal parts of thorax mostly darker, yellowed laterally.Scutellum yellow throughout.Pleuron orange-yellow to yellowish brown, somewhat shiny, restrictedly patterned with pale yellow, especially on dorsopleural membrane.Coxae and trochanters yellow to yellowish brown, fore coxa distinctly darker.Femora yellowish brown, darkened distally.Rest of legs yel lowish brown.Abdomen brown.
Head.Male antenna subequal to combined length of head and thorax.Flagellomeres generally spindle-shaped, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.
Thorax.Wing moderately broad.Wing venation as for subgenus.
Abdomen.Male terminalia (Figs 7-9) of moderate size.Posterior margin of tergite 9 moderately excised.Gonocoxite slender, with sparse setae.Outer gonostylus slightly more than half length of gonocoxite, gently arched, more or less parallel-sided, somewhat tapered in distal section, with apical spine slightly recurved.Inner gonostylus pale, slender, generally conical, with sparse setae.Aedeagal complex as in Figs 8-9.Aedeagus short and slender, gently upturned.Lower (caudal) part of vesica rather broad, parallel-sided, or even slightly broad ened anteriorly, somewhat flattened, with apodeme short and broad, rounded, fan-like.Lateral (parameral) apo deme moderately extending anteriorly beyond vesica.Female terminalia with internal structures as in Fig. 10.Spermathecae small, with sclerotized parts of ducts very short, about one third diameter of spermatheca.
Other Discussion.This is the smallest of the European Lurdia species, or strictly speaking, a species in which the per centage of small specimens is the highest.From the most similar R. (L.) sublurida sp.n., R. (L.) lurida is practically indistinguishable externally, differing from it in some Figs 11-14.Rhabdomastix (Lurdia) inclinata Edwards, 1938 (k, Î : Czech Republic: Vysni Lhoty).11-13 -male terminaba, gen eral view, dorsal (11), aedeagal complex, dorsal (12) and lateral (13); 14 -female terminalia, internal structures, ventral.Scale bars = 0.25 mm.details in the structure of the male terminalia, such as the shape of the inner gonostylus (broader in sublurida) and the length of the aedeagus (considerably longer in sublu rida) (cf.Figs 7-9 and 20-22).Surprisingly, the two spe cies differ considerably in the female terminalia.Whereas R. (L.) lurida has the smallest spermathecae, i.e., of all the species studied here, with sclerotized parts of ducts very short (Fig. 10), in R. (L.) sublurida sp.n., the spermathecae are of medium size, with a diameter about one and half times that of R. (L.) lurida, and sclerotized parts of the ducts very long, exceeding the spermathecal diameter (Fig. 23).The above features validate that the two forms are separate species.
Distribution, This species was recorded from the following European countries by Savchenko et al. (1992): Austria, Swit zerland, Slovakia, Germany, France, Italy, Romania, Ukraine (modified according to the present political frontiers).Subse quently, it was recorded from Poland (Wiedenska, 1998).With respect to the species described here the above records are unre liable except for that from Germany (type locality), and those from territories north of the Alps almost definitely are for dif ferent species.The record from Slovakia pertinent to Fig. 70 (Starý & Rozkošný, 1970) is of R. (L.) tatrica sp.n.Based on the material examined, the occurrence of the species in Switzer land, Germany and Italy is confirmed and a new record for Bul garia added.R. (L.) lurida is presumed to occur in the Alps and south of there (see above and under sublurida below).The specimen of Lurdia subgen n. illustrated as "Gonomyia laeta" by de Meijere (1920, Fig. 85) is not deposisted in the ZMAN (Oosterbroek, pers. comm.).
Colour.Basic colouration brown, with greyish pruinosity.Prescutum and other dorsal parts of thorax mostly greyish brown, yellowed laterally.Scutellum yellow throughout.Pleuron mostly greyish brown, somewhat shiny, patterned with pale yellow, especially on dorso-pleural membrane.All coxae and trochanters concolorous, yellowish brown.Femora yellowish brown, darkened distally, brown at tip.Rest of legs yellowish brown.Abdomen dark greyish brown.
Head.Male antenna subequal to combined length of head and thorax.Flagellomeres generally spindle-shaped, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.
Thorax.Wing moderately broad.Wing venation as for subgenus.
Abdomen.Male terminalia  rather small compared to entire insect.Posterior margin of tergite 9 deeply and widely excised.Gonocoxite slender, sparsely covered with setae.Outer gonostylus slightly more than half length of gonocoxite, gently arched and more or less parallel-sided, with apical spine slightly recurved.Inner gonostylus pale, slender, generally conical, sparsely cov ered with setae.Aedeagal complex as in Figs 12-13.Aedeagus short, very slender, gently upturned.Vesica comparatively small, its lower (caudal) part narrow, parallel-sided and cylindrical, or very slightly narrowed anteriorly, with apodeme forming narrow, rounded flange along its anterior margin.Lateral (parameral) apodeme only slightly extending anteriorly beyond vesica.Female terminalia with internal structures as in Fig. 14.Spermathecae small, sclerotized parts of ducts long, distinctly exceeding spermathecal diameter.

Material examined.
The species was described from six males and one female, all from the same locality (Edwards, 1938: 116).I have examined three type specimens.Holotype k (original designation): Great Britain, England, [North] Yorkshire, Mulgrave Woods nr.Whitby, 23.VIII.-3.IX.1937 (F.W. Edwards leg.) (BMNH), labelled: "Type" (a red-margined circular label), "Rhabdomastix inclinata Edw./ F.W. Edwards. det. 1938" (partly printed), "N.E. Yorks., Mulgrave Woods. 23.viii.-3.ix.1937. F.W.Edwards. B. M. 1937-562" (printed).The specimen is micro-pinned on a celluloid slide, with only hind right leg missing; left wing broken off and glued to the celluloid slide.Terminalia dissected and placed in Canada balsam on another celluloid slide, pinned with the specimen.Paratypes: 1 k, 1 2 (BMNH), with the same collecting data and mounted in the same manner as the holotype.The male paratype with terminalia placed in a plastic tube with glycerine, pinned with the specimen; the female para type with a yellow-margined circular label with the inscription "Paratype", and with the preparation of terminalia between cel luloid slides (described and illustrated by Tjeder, 1964a).
Discussion.In general appearance, R. (L.) inclinata is similar to other greyish brown species (robusta, loewi, furva), differing from some of them in having compara tively long antennae (shorter in robusta and loewi), and in that, compared to the entire insect, the hypopygium is relatively small (larger in robusta and loewi).Further male genital characters include the deep excision on ter-gite 9 (shallower in robusta and loewi), the short aedeagus (longer in robusta and loewi) and the narrow vesica (broader in the other three species, especially in furva).Female terminalia have the spermathecae small (subequal in size to lurida), but the sclerotized parts of the ducts very long.
Distribution.The species was recorded from Great Britain, Romania and Ukraine (Savchenko et al., 1992).The two latter records could not be confirmed.New records are presented for the Czech Republic and Slovakia.
Rhabdomastix (Lurdia) mendli sp.n.Diagnosis.Pleuron orange-yellow to yellowish brown.Femora darkened distally.Male antenna comparatively long, flagellomeres spindle-shaped, with verticils slightly exceeding length of respective segments.Wing moder ately broad.Male terminalia with outer gonostylus gently arched and inner gonostylus pale, rather slender.Female terminalia with spermathecae large and sclerotized parts of ducts very short.
Colour.Basic colouration orange-yellow.Prescutum and other dorsal parts of thorax darker, yellower laterally.Scutellum yellow.Pleura orange-yellow to yellowish brown, somewhat shiny, restrictedly patterned with pale yellow, especially on dorsopleural membrane.Coxae and trochanters yellow to yellowish brown, fore coxa darker.Femora yellowish brown, darkened distally.Rest of legs yellowish brown.Abdomen brown.
Head.Male antenna subequal to combined length of head and thorax.Flagellomeres generally spindle-shaped, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.
Thorax.Wing moderately broad, as in most other spe cies.Wing venation as for subgenus.

Material examined.
Etymology.This new species is named in honour of Dr. Hans Mendl (Kempten/Allgäu, Germany), a specialist on the Limoniidae, who first collected it, and who generously provided the author with his extensive Lurdia material from the Allgovian Alps.A noun in genitive singular.
Discussion.In body colouration, the new species is practically indistinguishable from R. (L.) lurida and R. (L.) sublurida sp.n.It is, however, sufficiently distinctive in the structure of the male terminalia, with the aedeagus conspicuously broad, abruptly narrowed before apex, and the vesica very broad, bulbous, devoid of any apodeme (no other species with this combination of characters).Female terminalia have rather pale, large spermathecae (as in robusta and loewi), with sclerotized parts of the ducts short.
Thorax.Wing (Fig. 4) moderately broad, as in most other species, about three times as long as broad.Wing venation as for subgenus.
Abdomen.Male terminalia (Figs 20-22) of moderate size.Posterior margin of tergite 9 moderately excised.Gonocoxite slender, with sparse setae.Outer gonostylus slightly more than half length of gonocoxite, gently arched, more or less parallel-sided, somewhat tapered in distal section, with apical spine slightly recurved.Inner gonostylus weakly darkened, rather broad distally before tapered apex, sparsely covered with setae.Aedeagal com plex as in Figs 21-22.Aedeagus of moderate length and breadth, upturned.Lower (caudal) part of vesica moder ately broad, parallel-sided and cylindrical, or even broad ened anteriorly, with apodeme forming narrow, rounded, fan-like flange along its anterior margin.Lateral (parameral) apodeme moderately extending anteriorly beyond vesica.Female terminalia with internal structures as in Fig. 23.Spermathecae medium in size, with sclerotized part of ducts long, exceeding spermathecal diameter.Etymology.The name of the new species, sublurida, indi cates its close relationship to R. (L.) lurida.An adjective in nominative singular.

Material examined.
Discussion.Externally, the new species is similar to R. (L.) lurida, differing in that the male terminalia have the inner gonostylus somewhat broader and the aedeagus dis tinctly longer and more upturned.The female terminalia are well distinguished from those of the latter species by the size of the spermathecae, which are of medium size, equalling those of R. (L.) tatrica sp.n., otherwise quite a different species (distinctly smaller in lurida), and by the considerable length of sclerotized parts of the ducts (very short in lurida) (see also Discussion under lurida).
Colour.Basic colouration very dark, dark brown, with dense greyish pruinosity.Prescutum and other dorsal parts of thorax mostly dark greyish brown, sometimes yellower laterally.Scutellum largely greyish brown, yel lower only posteriorly.Pleuron mostly dark greyish brown, somewhat shiny, restrictedly patterned with paler areas, dirty yellow on dorsopleural membrane.Coxae and trochanters mostly brown, fore coxa still darker.Femora brown, darkened distally.Rest of legs brown.Abdomen dark greyish brown.
Head.Male antenna subequal to combined length of head and thorax.Flagellomeres generally spindle-shaped, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.Etymology.The name of the new species, furva (= dark), refers to its very dark body colouration.An adjective in nomina tive singular.
Discussion.The new species, although strikingly dif ferent in body colouration, is very similar structurally to R. (L.) lurida.The two species differ in various details of the aedeagal complex, such as the shape of the vesica (narrower in lurida), or in that the apodeme of the vesica is continuous with the vesica without a distinct suture (separated by a distinct suture in lurida).Female termi nalia of R. (L.) furva sp.n. have medium sized spermathecae (much as in sublurida and tatrica, which, how ever, have longer sclerotized parts of the ducts), whereas they appear smaller in R. (L.) lurida.
Colour.Basic colouration brown to dark brown, with greyish pruinosity.Prescutum and other dorsal parts of thorax mostly dark greyish brown, somewhat yellow lat erally.Scutellum obscure yellow throughout.Pleuron pre dominantly greyish brown, somewhat shiny, restrictedly patterned with pale yellow, especially on dorsopleural membrane.Coxae and trochanters yellow, fore coxa somewhat darker.Femora yellowish brown, darkened dis tally.Tibiae darkened at tip.Tarsi yellowish brown.Abdomen dark greyish brown.
Head.Male antenna somewhat shorter than combined length of head and thorax.Flagellomeres generally spindle-shaped, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.
Abdomen.Male terminalia (Figs 28-30) of moderate size.Posterior margin of tergite 9 widely but shallowly emarginate.Gonocoxite slender, with sparse setae.Outer gonostylus slightly more than half length of gonocoxite, gently arched, more or less parallel-sided, sometimes even straight and somewhat tapered distally, with apical spine recurved.Inner gonostylus weakly darkened, rather broad distally before tapered apex, with sparse setae throughout.Aedeagal complex as in Figs 29-30.Aedeagus long and slender, strongly upturned.Vesica broad, its lower (caudal) part rather long and narrowed anteriorly, with apodeme long, broad and fan-like.Lat eral (parameral) apodeme not extending anteriorly beyond vesica.Female terminalia with internal structures as in Fig. 31.Spermathecae large, with sclerotized parts of ducts slightly shorter than spermathecal diameter.
Colour.Basic colouration brown to dark brown, with greyish pruinosity.Prescutum and other dorsal parts of thorax mostly dark greyish brown, somewhat yellowed laterally.Scutellum obscure yellow, darkened anteriorly in middle.Pleuron predominantly greyish brown, some what shiny, patterned with pale yellow, especially on dor sopleural membrane.
Thorax.Wing (Fig. 5) somewhat narrower than in other species (except for loewi), more than three times as long as broad.Wing venation as for subgenus.
Abdomen.Male terminalia (Figs 32-34) very large compared to entire insect.Posterior margin of tergite 9 widely but shallowly emarginate.Gonocoxite compara tively stout, with rather dense setae.Outer gonostylus more than half length of gonocoxite, gently arched, more or less parallel-sided, sometimes straight and slightly tapered distally, with apical spine recurved.Inner gono stylus considerably darkened, very broad, flattened, bent at midlength and tapered before apex, with sparse setae throughout and with additional loose tuft of very delicate and rather long setae near base.Aedeagal complex as in Figs 33-34.Aedeagus rather long and broad, strongly upturned.Vesica broad, its lower (caudal) part consid erably narrowed anteriorly, with apodeme long, broad and fan-like.Lateral (parameral) apodeme massive, extending anteriorly far beyond vesica.Female terminalia with internal structures as in Fig. 35.Spermathecae large, with sclerotized parts of ducts subequal in length to spermathecal diameter.inner gonostylus is distinctly bent at midlength in R. (L.) robusta sp.n.The most closely related species is probably R. (L.) loewi sp.n., which is almost identical in body colouration and outline of the wings and similar in the structure of the male terminalia.In R. (L.) robusta sp.n., however, the inner gonostylus is extraordinary pow erful and darkened (not so in loewi), setosity on various parts of the hypopygium dense and conspicuous (not so in loewi) and the aedeagus is shorter and broader.Female terminalia of R. (L.) robusta sp.n. have large spermathecae, much as in R. (L.) mendli sp.n. and R. (L.) loewi sp.n., with sclerotized parts of the ducts subequal in length to spermathecal diameter (distinctly differing from mendli, which has the ducts very short, less than half the diameter of spermatheca, and similar to loewi, which has the ducts only slightly shorter than the diameter).
Colour.In the single dry-mounted specimen available (1 k, Bulgaria, Pirin Mts.), basic colouration is much as in R. (L.) tatrica sp.n., generally bright orange-yellow to yellowish brown, consisting of various hues of yellow and brown.Prescutum and other dorsal parts of thorax mostly dark brown, yellow laterally and restrictedly so elsewhere, especially in middle of scutum.Scutellum yellow throughout.Pleuron bright orange-yellow, some what shiny, conspicuously patterned with yellow and brown, pale yellow on dorsopleural membrane.Coxae and trochanters yellowish brown, fore coxa distinctly darker.Femora yellowish brown, strongly darkened dis tally, blackish at tip of fore femur.Tibiae darkened at tip.Tarsi yellowish brown.Abdomen greyish dark brown.Specimens preserved in ethanol appear entirely pale yel low.
Head.Male antenna (Fig. 3) distinctly shorter than combined length of head and thorax.Flagellomeres not markedly spindle-shaped, rather oval to elongate, longest verticils slightly exceeding length of respective segments, pubescence short, indistinct.
Thorax.Wing moderately broad.Wing venation as for subgenus.
Abdomen.Male terminalia (Figs 36-38) of moderate size.Posterior margin of tergite 9 with comparatively narrow but deep, V-shaped excision.Gonocoxite slender, with moderately dense setae.Outer gonostylus long, dis tinctly more than half length of gonocoxite, considerably arched and slightly tapered distally, with apical spine stout, recurved.Inner gonostylus weakly darkened and very broad, flattened, generally trapezoid in outline, with sparse setae throughout and with additional loose tuft of very delicate and rather long setae near base.Aedeagal complex as in Figs 37-38.Aedeagus of moderate length, rather broad and gently upturned.Lower (caudal) part of vesica more or less bulbous (but not as broad as in mendli), with apodeme reduced to narrow flange, or entirely lacking.Lateral (parameral) apodeme extending anteriorly beyond vesica.Female unknown.Etymology.The name of the new species, falcata (= scythe shaped), refers to the shape of its outer gonostylus.An adjective in nominative singular.

Material examined.
Discussion.R. (L.) falcata sp.n. is distinctive in its body colouration (much as in tatrica), in the structure of the male antennae (Fig. 3) and, above all, in the male terminalia (for details see Discussion under tatrica and Figs 36-38).
Diagnosis.Pleuron bright orange-yellow.Fore femur strongly darkened distally.Male antenna comparatively short, flagellomeres spindle-shaped, with verticils slightly exceeding length of respective segments.Wing moder ately broad.Male terminalia with outer gonostylus long and considerably arched and inner gonostylus weakly darkened, very broad, trapezoid.Aedeagus long, vesica moderately broad.Female terminalia with spermathecae medium in size and sclerotized parts of ducts rather long.
Colour.Basic colouration orange-yellow to yellowish brown, with colours bright and colour changes sharp.Prescutum and other dorsal parts of thorax mostly dark brown, yellow laterally.Scutellum yellow throughout.Pleuron orange-yellow to yellowish brown, somewhat shiny, restrictedly patterned with pale yellow, especially on dorsopleural membrane.Coxae and trochanters yellow to yellowish brown, fore coxa distinctly darker.Femora yellowish brown, strongly darkened distally, blackish at tip of fore femur.Tibiae darkened at tip.Tarsi likewise somewhat darkened.Abdomen dark greyish brown.
Head.Male antenna somewhat shorter than combined length of head and thorax.Flagellomeres generally spindle-shaped, rather broad, longest verticils slightly exceeding length of respective segments, pubescence dense, distinct, suberect.
Thorax.Wing moderately broad.Wing venation as for subgenus.
Abdomen.Male terminalia (Figs 39-41) large relative to insect.Posterior margin of tergite 9 shallowly emarginate.Gonocoxite comparatively stout, with rather dense setae, especially on distal inner margin.Outer gonostylus long, distinctly more than half length of gonocoxite, scythe-shaped, considerably arched and considerably tapered distally, evenly passing into curved, slender apical spine.Inner gonostylus weakly darkened and very broad, flattened, generally trapezoid in outline, with sparse setae throughout and with additional loose tuft of very delicate and rather long setae near base.Aedeagal complex as in Figs 40-41.Aedeagus rather long, slender, upturned.Lower (caudal) part of vesica moderately broad, parallel-sided, cylindrical, or somewhat bulbous, with apodeme mostly reduced to two lateral flanges and minute median lobe, sometimes apodeme even lacking completely.Lateral (parameral) apodeme massive, dis tinctly extending anteriorly beyond vesica.Female termi nalia with internal structures as in Fig. 42.Spermathecae medium in size, with sclerotized parts of ducts subequal in length to spermathecal diameter.Material examined.Etymology.The name of the new species, tatrica, is derived from the name of the mountains, the Tatras, which the species seems to be restricted to.The name is deemed to be and to be treated as a latinized adjective in nominative singular, in accor dance with relevant provisions of the Article 11.9. of the ICZN (1999).
Discussion.The new species is characterized by its bright body colouration with orange-yellow pleura and dark legs (colour differences more pronounced than, e.g., in sublurida), and the antennae comparatively short, thus most resembling R. (L.) falcata sp.n.It is similar to the latter species also in the structure of the male terminalia, especially in that the outer gonostylus is long and consid erably arched and the inner one flattened, trapezoid, with a tuft of delicate setae near base.These features readily distinguish the two species from the others.From R. (L.) falcata sp.n., R. (L.) tatrica sp.n. differs especially in that its hypopygium is large relative to the insect, the outer gonostylus more tapered distally and the aedeagus distinctly longer.Male antennae of R. (L.) tatrica sp.n., although slightly shorter than in, e.g., R. (L.) lurida, are longer than those of R. (L.) falcata sp.n., with flagello-meres spindle-shaped (rather oval to elongate in falcata).Female terminalia have medium sized spermathecae (much as in sublurida) and sclerotized parts of the ducts subequal in length to spermathecal diameter (longer in sublurida).
Distribution.Slovakia.The record under lurida perti nent to Fig. 70 in Starý & Rozkošný (1970) belongs here.The species has only been collected in the Tatras.