We describe 2 new and closely allied tettigoniid species, recently discovered on mountain slopes in southwestern continental Greece. Though they differ in habitus, coloration, morphology (modified paraprocts) and habitat from most members of Rhacocleis, the new species have been arranged tentatively in this genus. Unknown morphological characters and faunistic data for some other poorly known Greek Rhacocleis species (R. ferdinandi Willemse & Tilmans 1987 and edentata Willemse 1982) are included.
During a visit to Greece in 2003, a new bushcricket was discovered on a mountain slope in Ipiros, western continental Greece. This discovery was surprising, as the species was found underneath rocks and amidst scree, an unusual habitat for this group of tettigoniids. The peculiar habitat and its nocturnal lifestyle, the day spent hidden away underneath rocks and stones, may explains why this species was not discovered earlier. In 2004 several more populations were located in Greece in similar habitats throughout the southern Pindhos range, including places visited on earlier occasions. The populations found in 2004 were of another undescribed species, closely allied, but distinct from that found in 2003.
For song recordings we used a Tascam DA P1 tape recorder and Sennheiser K6 module with ME62 microphone: this system has an audio-limited frequency response of 50 to 20 kHz. Oscillographic analysis used BIAS Peak and SoundEdit software. Given our audio-limited equipment, and because the carrier frequencies of other Rhacocleis spp. lie mainly above 20kHz (Heller1988), we have not attempted to carry out spectral analysis.
The specimens were kept indoors at room temperatures (19 to 25°C). The cages used were either plexiglass, 50 × 30 × 25 cm, the top covered with mosquito netting or wooden-framed, 20 × 15 × 15 cm, the vertical sides covered with metallic screen. Optimal singing activity was commonly around midnight and the hours following. Recordings were made at irregular intervals of days, weeks or months. While recording, artificial light was kept to a minimum. The distance between subject and microphone was usually about 10 to 15 cm, but sometimes 5 cm or less when recording the incidental sounds of tremulation (shaking of the body). (Though no transducer of substrate vibrations was available, we recorded cagerelated sounds that accompanied tremulation.) When appropriate, other specimens were removed from the studio. The specimens were kept either isolated from each other, as a couple, or several together in one cage. Recorded specimens are labelled as such and deposited as indicated below. The song terminology followed is that commonly used by European bioacousticians. For an extensive discussion see Ragge & Reynolds (1998).
The tribe Platycleidini (s.s.) was defined by Zeuner (1941) (nec Harz 1969, as referred to in Otte, Eades & Naskrecki 2004) for the genera Platycleis and Metrioptera. Neither new species belongs to these genera, but to the Platycleidini (s.l.), as defined in a much wider sense in Rentz & Colles (1990).
Based on traditional characters (e.g., Brunner von Wattenwyl 1882, 1893, Caudell 1908, Harz 1969), the new species keys out to Rhacocleis Fieber 1853: prosternum bispinose, fore tibia with upper outer margin presenting an apical spur, hind tibia with single pair of large ventral apical outer spurs, pronotum with no median keel, meso- and metasternum unarmed and hind tarsus with free, relatively long, plantula.
However, compared with other members of Rhacocleis: both new species are characterised by 1) their unique tridentate and quite robust male cercus, 2) male paraprocts more sclerotised and, in one of them (R. lithoscirtetes sp.n.), remarkably elongate, 3) the general coloration being pearl grey to pale buff instead of brownish (particularly in R. crypta sp.n.), 4) the comparatively large and wide tegmina, which are mainly solid black, and 5) by a general habitus less slender than typical of most Rhacocleis species.
Another distinctive feature is the habitat: Rhacocleis species occur from the lowlands up to moderate altitudes and prefer habitats with sufficient vegetation cover, consisting of a mix of shrubs, herbs, grasses etc. However, the new species, as far as known, occurs only in mountain habitats, where they are found hiding in cavities and crevices between stones and rocks.
Both new species bear some resemblance to Antaxius Brunner 1882, particularly to the general color and habitus of A. pedestris (Fabricius 1787). Moreover as in the new species, in some species of Antaxius the male paraprocts are modified, but only slightly, the ventral side being spoon-shaped, not cylindrical or as long as in R. lithoscirtetes, where they extend beyond the epiproct. It is noted here that the modification of the paraprocts in R. lithoscirtetes is not as dramatic as the ‘pseudocerci’ with clasping function found in the tettigoniid tribe Nedubini. However, Antaxius is made quite distinct from both new species by its flattened male cercus, the female subgenital plate and the unarmed upper inner margin of the fore tibia.
The difficulties in generic assignment lead to the question of the status of Rhacocleis vs Pterolepis. Both genera are very similar and even assumed to be synonymous (Heller et al. 1998). A study of the armature of the fore tibia showed that the new species fits Rhacocleis rather than Pterolepis. The relevant data of this study, together with an updated checklist of presently recognized taxa of Rhacocleis and Pterolepis, are presented in Willemse & Willemse (this issue). As long as a critical review of the generic definitions in the Platycleidini (s.l.) is unavailable, and because the armature of the legs of both new species fits Rhacocleis rather than Pterolepis, we propose for the time being to assign both new species to Rhacocleis. In case a separate supraspecific taxon for the new species becomes necessary, the proposed species name lithoscirtetes (a noun, and as far as known not preoccupied, (see http://www.ubio.org/NomenclatorZoologicus) seems an appropriate choice.
Material studied.— ♂ holotype, 14 ♂♂ and 14 ♀♀, paratypes, labeled: “HELLAS from Kastania (Evrit., Proussos) to Aghiro Pighadi (Aet-Akarn.), bel. summit Mt.Triandafillia (1817m); 1380–1460m; 9 aug. 2004; D.&F.&J.&L. Willemse & T. Blauw”, “WGS84: N38°42′36.2-34.1″ E021°41′35.3-19.5″; stenige hellingen, verspreide coniferen, spaarzame vegetatie”.
Additional material.— 4♂♂ and 7 ♀♀: “HELLAS (Aet.-Akarn.) 1200-1560 m Mt. Tsakalakis, S of Ano Chora, & Kendriki; 6 & 7 aug. 2004, D.&F.&J.&L. Willemse & T. Blauw”, “WGS84: N38°33′36.7″ E021°53′26.9″ steen helling, struiken, bomen, kruiden, grassen”; “Ano Chora – na 2 km afslag Nafpaktos, na 7 km afslag Kendriki, na 5 km afslag A. Kiriaki, 3.5 km na deze afslag”; 2 ♂♂ and 1 ♀: “HELLAS (Evritania) 1320m Mt. Panaetolikon, pas tussen Labini & Prousos, 1.5 km west; 8 aug. 2004 D.&F.&J.&L. Willemse & T. Blauw”, “WGS84: N38°43′26.8″ E021°37′00.4″, steenhelling met coniferen, spaarzame vegetatie”; 3 ♂♂ and 2 ♀: “HELLAS (Evritania) 1250 m Mt.Timfristos boven Karpenisi 10 aug. 2004; D.&F.&J.&L. Willemse & T. Blauw”, “WGS84: N38°55′30.7″ E021°48′07.2″; geul tussen 2 hellingen, veel puin”. (All deposited in collection Willemse [in the future a public collection]; 1 ♂ 1 ♀ paratype in collection Heller).
Diagnosis.— Both sexes of this species are easily recognised among congenerics by general habitus and coloration. General color in Rhacocleis varies between shades of brown to greyish brown, including the color of the forewings. The general color of the new species, however, is pearl grey and most of the forewings commonly solid black. The general habitus in Rhacocleis is slender, with relatively long to very long hind legs, while those of the new species are less slender. Both characters of the new species are reminiscent of e.g., Antaxius pedestris, rather than any member of Rhacocleis.
Other remarkable differences in the male sex are the short, wide and basally strongly toothed cerci, instead of the commonly slender elongate ones seen in other Rhacocleis species. However, the most peculiar difference refers to a pair of terminal abdominal processes, easily visible to the naked eye. Closer view of the latter reveals these are modified paraprocts, unknown for this genus as well as for most Platycleidini.
Another obvious difference lies in the stony and rocky mountain habitat: other members of the genus, as far as known, prefer shrubby areas with much vegetation, occur commonly in lowlands and are not restricted to any altitude.
Description.— Male (Figs 18, 21, 24, 26). Medium size. Integument silky, glossy, smooth. Fastigium of vertex parallel sided, not grooved or depressed, as wide as greatest width of scape or slightly wider. Pronotum subcylindrical, edges rimmed, in profile straight; metazona relatively short, with dorsum very slightly flattened, without distinct median keel, widely rounded and merging with lateral lobe; hind margin obtuse, almost straight or with very weak median impression; lateral lobes almost vertical, along fore and lower edges slightly impressed. Prosternum on either side with a tiny spine; mesosternum with low triangular projections; similar projections of metasternum wider, but less produced. Thoracic tympanum widely open, strongly elongate. Forewings: (Figs 39, 42) reaching beyond hind margin of 1st tergite, at most up to hind margin of 2nd tergite; left forewing slightly longer than wide; fore edge from straight to weakly convex, apex broadly obtuse; hind edge with well-developed cubital sinus; Sc, R and M veins well raised; Sc and R reaching edge and more or less parallel; M not reaching edge but well curved; distal part of Cu2 strongly inflated. Stridulatory file, as seen from below, relatively short, raised on the swollen Cu2; shortest distance between proximal and distal ends ∼1.8 mm, proximal part strongly curved, remainder of file weakly arched, in both horizontal and vertical directions; total number of teeth about 80, of which 50 are strongly sclerotised; widest spacing between teeth approximately in mid-file, 16 teeth per 0.5 mm, i.e., in this part of the file, distance between the middle of successive teeth is 0.03 mm; greatest width of these middle teeth 0.05 mm. Right forewing with apex wide and almost transverse, mirror well delimited by veins, as wide as long, almost square, with apical margin slightly bowed. Abdominal tergites scarcely folded along midline, no true median keel. Abdominal terminalia: (Figs 37, 38, 40, 41, 43–45) last tergite slightly hairy as are distal tergites; hind margin with shallow parabolic median excision, the lobes not pointed, but smoothly rounded and sometimes provided with a minute sharp, downwards pointing tubercle at lower edge, at either side of median excision. Cercus (Figs 37, 38, 40) hairy, robust, conical, basal part thickset and tapering into an apical part that is inward- and downward-curved, terminating into a sharp tip; inner side of cercus with minute basal tooth, immediately followed by a large, strong subbasal tooth that is dorso-ventrally flattened and points inwardly; distance, as seen from above, between base and tip of cercus slightly longer than distance between lateral surface of cercus and tip of sub-basal tooth (1.8–2.0 mm and 1.5–1.9 mm respectively). Epiproct as usual, a small and weakly sclerotised triangular flap located below hind margin of last abdominal tergite (Fig. 40). Paraprocts (Figs 41, 43–45) modified: remarkably long (1.5 mm), well-sclerotised, cylindrical, tip obtuse; paraprocts far exceeding tip of epiproct and hind margin of last tergite, extending between the bases of the cerci, as a pair of parallel, but commonly slightly diverging, finger-shaped processes, commonly curved outwards and pointing backwards and/or upwards. Subgenital plate elongate, slightly tapering with low median keel, its hind margin from slightly to deeply v-shaped emarginate; styli slender, cylindrical, of normal length.
Titillators (Figs 7, 8) slender, thin, well-sclerotised; basal fusion of right and left titillator membranous; basal and apical parts of about equal length and pointing in same vertical direction; basal part slightly inflated, slender, smooth; apical part towards tip flattened and slightly curved, tip hook-shaped, surface smooth. There is a second pair of short, simple, and less strongly sclerotised structures, with a membranous connection to each other.
Legs not very slender, but comparatively of normal length and width; hind femur length 5.1 × greatest width of hind femur (5.4 × in female); fore femur as long as pronotum, hind femur a little more than 3 times as long as pronotal length. Tibial tympanum slit-like. Free plantulae of hind tarsus 2/3 as long as first tarsal segment. Armature of legs: fore femur sometimes with a few tiny spines along lower inner margin, mid femur unarmed, hind femur commonly with 4 tiny spines along the mid part of the lower inner margin; genicular lobes of all femora with 1 or 2 tiny spines; upper inner (= anterior) margin of fore and mid tibia with 2 spurs (in the fore tibia a proximal one at the distal end of the tympanal opening and a distal one halfway on the tibia); upper outer (= posterior) margin of fore and mid tibia with 3 or 4 spurs (in the fore tibia, a proximal one at the distal end of the tympanal opening, 1 or 2 halfway on the tibia and always an apical one); both lower margins of fore and mid tibia with 5-7 spurs; tip of hind tibia with a pair of ventral apical spurs (= outer spurs) and sometimes without, but commonly with, a pair of pits in between these spurs, often bearing a hair.
Coloration: (when alive). General color pearl grey to pale buff (turning to dark grey or brown after dry preservation), distinctly mottled with dark brown. Antennae with flagellum uniformly pale brown, scape and pedicel sometimes with black markings. Head and pronotum of general color, commonly marbled with brown, sometimes forming a longitudinal pattern on the dorsum of the metazona or occiput; face often with a pair of short black median vertical lines and symmetrical black lateral points; antennal sockets and mouth parts sometimes with black dots; the middle of the pronotal dorsum and lateral lobes commonly with some symmetrical black or dark-brown patches, lateral lobes commonly uniformly colored, exceptionally with a black patch and pale stroke along hind margin; pro, meso- and metapleurae with black spots. Forewing black, small apical part between R and M from pale yellowish to white, transparent parts not colored. Abdominal tergites of general color, first tergites dorso-laterally black, hind margins with a tiny median black dot. Pro-, meso-, metanotum, sternites and subgenital plate, all uniformly pale grey white; pleura sometimes with black dots. Male cercus pearl grey or buff, with a brownish black median stroke extending between apical part and sub-basal and basal teeth, this stroke in its middle sometimes reddish brown; lower side of cercus dark or even blackish brown, particularly at the base; teeth dark brown to black. Legs of general color, variegated with dark brown and black dots or stripes, insertion of spurs commonly black, spurs brown. All femora with black pregenicular incomplete ring; upper side of hind femur of male with large black postbasal spot, often extending onto dorsal face as series of transverse black stripes; outer side often with a few black stripes; fore and mid-knee yellowish white. All tibiae with postgenicular blackish mark; very apex of fore and mid tibia yellowish white. Plantulae of first tarsal segment black or dark brown; second segment cream or ivory white; third tarsal segment and claw dark brown to solid black.
Female (Figs 22, 28, 29, 31, 33). Head, thorax and legs as in male. Micropterous, forewings lateral, almost completely covered by pronotum, oval, reaching just beyond fore margin of 1st tergite. Abdominal sternites, including the 7th, not modified.
Cerci basally inflated, apical part conical. Ovipositor (Fig. 33) ensiform, relatively short and robust, 3/4 as long as hind femur; towards apex very weakly curved upward, margins smooth. Subgenital plate (Fig. 31) divided by transverse fold into apical and basal sclerite; apical sclerite elongate triangular, not or scarcely, convex, sometimes with low median keel, but often with center near its anterior margin weakly sclerotised and somewhat depressed, lateral margins strongly bent upward and straight, tapering to tip with very short u- or v-shaped excision; basal sclerite narrowly transverse, weakly sclerotised, with transverse folds on either side. Gonangulum elongate, simple.
Coloration as male but paler. Tegmina from solid black to unicolorous pale brown. Head, thorax, abdomen and terminalia more uniformly colored. Legs as in male; hind femur with fewer black markings.
Measurements.— See Table 1.
Variation.— The upper inner margin of the left and right fore tibia commonly presents 2 spurs, but in 5 of 47 specimens this margin of either the left or right fore tibia (in one specimen even bilaterally), has 3 spurs. In one specimen out of 47, this margin of one of the fore tibia presents but 1 spur, the other tibia the usual 2 spurs. The tip of the hind tibia shows invariably no true inner spurs. In 2 of 47 specimens there is, just unilaterally, a single very short spur between the outer spurs. The minute basal tooth of the male cercus varies slightly in size, ranging from more reduced to more developed.
Bioacoustics.— (Fig. 13A,B) Observations were made on 5 males and a number of females. The songs of these males, one from Mt.Timfristos, one from Mt.Tskakalakis and 3 from Mt.Triandafilia, were recorded in sessions of 15 min to 2 h during the nights of 26 August, 10, 15, 19 and 24 September, and 11, 13 and 14 November, 2004. Recordings were of isolated males, of a male together with a female in a cage, or of a male with other males and/or females kept together in one cage.
A calling song consists of series of 3 to 30 echemes, repeated at irregular intervals (Fig. 13A). Echemes (Fig. 13 B) last 600 to 1100 ms and consist of 10 to 20 syllables (Fig. 13B). Intervals between echemes within the series last 200–600 ms. Syllables, excepting the first 1 or 2, are equally loud to human listeners, and are produced with a repetition rate of about 19/s (19°C). An alternation of 2 hemisyllables is clearly visible throughout the oscillogram (Fig. 13B). Presumably the shorter hemisyllable (about 12 ms) corresponds to wing opening, the longer (35 to 40 ms) to the closing. The first emission of an echeme is usually different from the following opening and closing hemisyllables. The closing hemisyllable consists of about 35 individual pulses.
In acoustic interaction with 1 to 3 other males, the series of echemes tends to be shorter, but individual echemes tend to last longer (up to 2.5 s, with up to 36 syllables), especially the first one in a series. By contrast, in the presence of a female, both echemes and the series of echemes, last longer.
In both this species and R. crypta (see below), a series of echemes could be followed by a series of faint tremulation sounds, accompanying the shaking of the body (Fig 13D,F). This body shaking also happened independently of the presence of other specimens. There was no preference among available substrate: the wall of the cages versus the netting or some vegetation present in the cage.
Although we cannot characterize the body-movement pattern of these tremulations, nor its vibratory form within the substrate, on the basis of the accompanying faint sounds (Fig. 13 D) it incorporates 5 to 10 elements. The origin of these tremulation sounds was unclear: perhaps some part of the substrate (cage) on which the animal stood. During tremulation it appeared that only the legs made contact with the substrate and never the ventral side of the abdomen; but the rapidity of the movements left us unsure whether the forewings were at rest.
Discussion.— The song of R. lithoscirtetes is different from that of the majority of species of Rhacocleis, Anonconotus, Pterolepis and Antaxius. Yet the overall structure of the echeme shows affinity with that of Rhacocleis germanica and Anonconotus alpinus (Heller 1988). In comparison with the syntope R. germanica, the syllable repetition rate in R. lithoscirtetes is slow (about 30 to 35/s in R. germanica at 20°C) and the intervals between echemes are usually shorter than the echemes themselves (in contrast to R. germanica).
The calling song shows a considerable amount of variation in the number of echemes in a series. It is not yet clear whether this variation expresses differences in populations or has to be judged as an expression of individual singing activity.
Tremulation is a common phenomenon in captive Rhacocleis. In this particular case its function seems obscure, being perhaps inefficient upon a rocky substrate. However, there are no observations of the species' nocturnal life and it is unknown whether the males leave their daytime stay, moving nocturnally onto more suitable substrate, such as vegetation.
Distribution.—This species is only known from 4 localities in continental Greece, all in the southern part of the Pindhos range: Mt. Tsakalakis, near Ano Chora in the province of Aetolia-Akarnania and the Mts Panaetolikon and nearby Triandafillia and Timfristos, all in the province of Evritania (Fig. 14) The type locality is the Evritanian slope, below the summit named Triandafillia (= rosebush). The spot was accessible via a very rough mountain road that connects the village of Kastania in Evritania (east of Proussos) via a pass (∼1550 m) over Mt. Triandafillia, with that of Aghiro Pighadi in Aetolia-Akarnania. The type locality is at 1380 m elevation, a few kilometers inside the border of Evritania.
Etymology.— lithoscirtetes = stone-jumper, a masculine noun composed from litho[-s] = stone in Greek and scirtetes = jumper from the ancient Greek verb scirtao = to jump.
Ecology.— The habitat (Figs 15–18) consists of dry mountain slopes, predominantly covered by rocks and stones between 1200 to 1500 m elevation. Vegetation is comparatively poor, with few, often coniferous, shrubs or trees, few herbs and tufts of grasses or locally a group of ferns. Most remarkable are the large number of more or less loose stones and firmly embedded rocks with abundant cavities and crevices. Some of these spots are almost devoid of vegetation or with only a few scattered tufts of, e.g., Thymus sp.
In daytime we found the specimens commonly hiding under stones. When disturbed they are very apt at finding cavities and small crevices that make relocating them difficult and commonly to no avail (Figs 17,18). The number of specimens observed was always low.
In early August, most individuals were not yet adult. Kept in indoor cages, most became adult after a couple of weeks and started to sing and mate from the mid or end of August. The high season is the month of September. The species is nocturnal, most activity taking place around midnight: moving, feeding, singing, mating, oviposition. In the daytime, when resting in dark corners of their cage, often several individuals, mostly females, but also males and females, clustered together, a behavior also observed in the field, where specimens seemed to occur in small groups. From observed behavior we found no indication of the function of the elongated paraprocts of the male.
Among diverse substrates for digging and laying eggs, most females preferred a firm but light foam, as used by florist's shops, over fine gravel, sandy ground, clay or peat-like soil. As far as observed, eggs were laid separately, in small numbers and close together. When gutting the female abdomen, up to 40 eggs could be counted. Specimens held in captivity died in the second half of November, a few females lasting even to the beginning of December.
While cultivation in 2005 of the offspring of Rhacocleis werneri and insularis, also collected in 2004, was easy, that of the new species was not successful: eggs did not hatch, at least within this period of 1 y.
Everywhere R. lithoscirtetes occurred together with Rhacocleis germanica, the latter always in very low densities. Adult specimens from both species could easily be distinguished, as well as the juveniles, by the overall brownish coloration of R. germanica, by (often) a pale band along the hind margin of the pronotal lateral lobes, by their more slender habitus and strongly different cerci, and by the short paraprocts in the male.
In these habitats, but particularly incorporating a wider area with more vegetation, the new species was also found together with other tettigoniids and acridids.
We abstain from giving a full list of the local faunas. Some species, however, are worth mentioning. On Mt. Tsakalakis occur Poecilimon zimmeri Ramme, Decticus verrucivorus (Linnaeus), Pholidoptera femorata (Fieber), Eupholidoptera megastyla (Ramme), Platycleis (s.str.) intermedia (Serville), P. (s.str.) albopunctata grisea (Fabricius), Gryllomorpha sp., Oedipoda germanica (Latreille), O. coerulescens (Linnaeus), Celes variabilis (Pallas), Dociostaurus maroccanus (Thunberg), Paracaloptenus c. caloptenoides Brunner, Stenobothrus fischeri (Eversmann), S. rubicundulus Kruseman & Jeekel, Chorthippus bornhalmi Harz, C. biguttulus euhedickei v. Helversen, C. pulloides (Ramme), C. willemsei Harz, C. dorsatus/dichrous group, C. paralleleus tenuis (Brullé), Stauroderus scalaris (Fischer de Waldheim).
In Mt. Triandafillia we observed most of the same species, except Gryllomorpha sp., Paracaloptenus c. caloptenoides Brunner, Chorthippus pulloides (Ramme), C. dorsatus/dichrous group, and Celes variabilis (Pallas); but in addition we found Oropodisma lagrecai Willemse, Chorthippus m. mollis (Charpentier) and Myrmeleotettix m. maculatus (Thunberg).
Material studied.— ♂ holotype, 4 ♂♂ and 8 ♀♀, paratypes, labelled: “Hellas Arta (southeast) 1500–1750 m, Mt. Gavrogo below summit Gerabi above Megalochari village, 3–4.viii.2003, D., L. & F. Willemse & T. Blauw; WGS84: N39°11-12′46.4-1″ E21°16–17′16.4-51.8″ (deposited in collection Willemse); 3 ♂♂, 4 ♀♀: Greece, Arta, Mt. Gavrogo, 1600–1800m, 4–5.8.2003, 39°13′N 21°15′E, K.-G. Heller, M. Heller & M. Volleth (CH6175 ♂, CH6176 ♂, CH6177 ♀, CH6178 ♀, CH6179 ♀, CH6265 ♂, CH6266 ♂, collection Heller).
Diagnosis.—This species is easily recognisable among its congenerics by similar characters as noted under R. lithoscirtetes. The species differs from the latter in the male cercus and the degree of modification of the paraprocts. The cercus is still more robust, but the paraprocts are less strongly modified, extending scarcely beyond the tip of the male abdomen and thus not as conspicuously visible as in lithoscirtetes.
Description.— Male. Size, habitus (Figs 19, 23, 25), shape of tegmina (Fig. 36) and stridulatory file, hind margin of last abdominal tergite, titillators (Figs 9, 10), armature of legs, all similar to Rhacocleis lithoscirtetes. Differs from the latter only in cercus (Figs 5, 6) and paraprocts (Fig. 35)). Paraprocts well sclerotised, inflated, cylindrical, robust, but short, only slightly longer than tip of epiproct and not pointing up and backwards from in between the cerci as large finger-shaped processes. Cercus similarly shaped, robust but conspicuously shorter: basal part strongly inflated, conically tapering into a short in- and slightly downcurved apical part, with a narrowly obtuse tip; inner side also with a small basal tooth, immediately followed by a strong and flattened sub-basal tooth, pointing medially; distance between base of the cercus and its tip slightly less than distance between lateral surface of the cercus to the tip of the subbasal tooth (1.3 mm and 1.5 mm respectively).
Coloration.—As in R. lithoscirtetes, but the general coloration when alive, is much paler, a pale pearl grey (Fig. 19), turning darker to grey or brown when dry preserved.
Female (Figs 20, 27, 30, 32) as in R. lithoscirtetes, excepting the basal sclerite of the subgenital plate (Fig. 30) presenting on either side a more solid transverse ridge, converging and fusing into a low median bar reaching the middle of hind margin of the unmodified 7th sternite. Coloration as for the male, but paler and the dark pattern less conspicuous.
Measurements.— See Table 1.
Variation.— The upper inner margin of both fore tibia has commonly 2 spurs. In 4 out of 20 specimens, there is unilaterally but 1 spur, and in 1 specimen a single spur on both sides. The apex of the hind tibia lacks a pair of inner spurs, except in one specimen (1 of 20) where a pair of very short spurs is present, bilaterally, in between the outer spurs, and which may be considered as rudimentary inner spurs. The small basal inner tooth of the male cercus is somewhat variable in size, occasionally arising at the base of the large sub-basal tooth.
Bioacoustics.— (Fig. 13 C-F) Three males and 2 females were observed during 2 sessions, each for about 1 to 2 h during the nights of 12 and 31 August 2003. In the first session, recordings were made from one male kept in a cage, separated from the other specimens, which were moved to another room. During the second session the sounds of a male in the presence of a female in one cage were recorded.
The calling song consists of series of 10 to 20 echemes (Fig. 13C), repeated at irregular intervals. Echemes last 500 to 800 ms and consist of 12 to 20 syllables. Intervals between echemes last 300 to 4000 ms. Syllables, except the first one or two, are equally loud to the human ear and produced with a repetition rate of about 24 to 26/s (24°C). An alternation of 2 hemisyllables is clearly visible throughout the oscillogram (Fig. 13E). The shorter hemisyllable probably coincides with the opening movement of the wings and lasts about 10 ms, whereas the presumed closing hemisyllable lasts about 30 ms. The very first sound (syllable) of an echeme is usually different from subsequent syllables (Fig. 13 E). The closing hemisyllable consists of about 40 rapidly decaying pulses.
In acoustic interaction with other males, the series of echemes tends to be shortened and echemes to last longer, up to 3.5 s, with up to 60 syllables; this is especially true of the first in the series.
In the presence of a female, echeme series tend to last up to several minutes, while individual echeme durations are 1 to 3 s. With a female in close proximity the interval between echemes was 400 ms on average and the syllable repetition rate 15 to 16/s (20°C). A series of echemes is often followed by a series of tremulation signals (Fig. 13 D, F; see also above under R. lithoscirtetes).
Discussion.— The song of this species is very similar to that of R. lithoscirtetes. But considering our small sample of specimens and recordings, we cannot conclude if there are any consistent differences.
Distribution.—Known only from the type locality, Mt. Gavrogo above the village of Megalochari, in the southeastern part of the district of Arta in western continental Greece (Fig. 14). This mountain belongs to the Valtou range, a western offshoot of the Pindhos range, separated from the latter by the deep valley of the Acheloos river.
Etymology.— crypta = hidden, from the ancient Greek adjective kryptos, latinized to cryptus (Rhacocleis = gender feminine).
Ecology.— The habitat at 1500 to 1700 m altitude is similar to that of R. lithoscirtetes. The species was observed mainly along the stony banks of a rough mountain road, commonly hiding under stones and, exceptionally, found in the open. The coloration of the specimens (Figs 19, 20) fits perfectly with the white grey color of the local limestones. But few specimens were observed and most of them subadult on 2–3 August. In captivity they became adult after about 2 wk and started about 10 d later to sing and mate.
At the only known locality of R. crypta we found no Rhacocleis germanica. In a wider area, outside the typical habitat, we observed a number of other species: Decticus verrucivorus (L.), Platycleis (Parnassiana) sp., Pholidoptera sp. macedonica group, Eupholidoptera megastyla Ramme, Poecilimon zimmeri Ramme, Oropodisma sp., Gryllomorpha sp., Troglophilus sp., Pezotettix giornae (Rossi), Aryptera labiata (Brulle), Oedipoda germanica (Latreille), Myrmeleotettix m. maculatus (Thunberg) Stauroderus scalaris (Fischer de Waldheim), Chorthippus m. mollis (Charpentier), Chorthippus bornhalmi Harz, Omocestus rufipes (Zetterstedt), Stenobothrus rubicundulus Kruseman & Jeekel and Stenobothrus fischeri, (Eversmann).
Rhacocleis crypta differs clearly from R. lithoscirtetes in the size of the male cercus and length of the paraprocts, both being shorter in the former. Among the diverse populations of R. lithoscirtetes, these structures are of similar length and appear to be a consistent character. Particularly, the obvious difference in length of the paraprocts in R. crypta and R. lithoscirtetes seems to us quite an important character.
The general differentiation of both new species from other members of Rhacocleis is already summarised above under the paragraph of generic distinction and the diagnosis. Some additional details are specified below.
The general habitus of both new species resembles that of R. distinguenda Werner, 1934 and R. uvarovi Ramme, 1939, which are also less slender than most congenerics. However, both male and female abdominal terminalia are quite distinct. Of note, the coloration of the pronotal lateral lobe of R. distinguenda is much the same as in both new species: without a clear pale band along the lower and hind margins.
The male cercus in most Rhacocleis species is slender, much longer than wide with one basal or sub-basal tooth, and thus quite distinct from the stout and tridentate cercus in R. lithoscirtetes and crypta. The male cercus of R. tyrrhenica [!! = thyrrhenica auct.] La Greca 1952 (l.c. figs II 1–2), reminiscent of Pterolepis spoliata, is also robust and strongly bent inward, but bidentate. R. tyrrhenica differs from the new species by the male paraprocts being only slightly sclerotised, not cylindrical and certainly not as elongated as in R. lithoscirtetes, and by differences in the titillator (La Greca 1959, l.c. fig. 32) and female terminalia (La Greca 1952, l.c. fig. II 3; Nadig 1962, l.c. fig. 5b), the habitus (slender in R. tyrrhenica) and the coloration of the pronotal lateral lobes, which resemble most Rhacocleis species in being brown with a pale band along the lower and hind margins.
The male cercus of R. ferdinandi L. Willemse & Tilmans 1987 (l.c. figs 3, 4) is also strongly bent inward, but is less robust, biden-tate and of different shape. Originally this species was known and described only after the male sex, but now the female is available from recently collected material of the locality of the male type. The 6th and 7th abdominal sternites of the female of R. ferdinandi are unmodified, as in both new species, but the subgenital plate of R. ferdinandi is different: tip similarly excised but apical sclerite heart shaped, as long as basal width, and basal sclerite plate-like, well sclerotised, wide, transverse, almost flat, without or with an indistinct median ridge (Fig. 12). Further features distinguishing between this species and both new ones, are the general appearance and coloration, unmodified male paraprocts and habitat.
The male cerci in R. derrai Harz 1983 (l.c. fig.1) (Fig. 4), ayali Karabag 1974 (l.c. fig. 2), bidens Uvarov 1924 (l.c. figs 14, 15), dernensis Salfi 1926 (l.c. figs 50, 51; Karabag 1974, l.c. fig. 7), lagrecai Fontana & Massa 2004 (l.c. figs 7–10) and of andikithiraensis Tilmans 2005 (l.c. figs) are relatively short, and particularly in the last species, robust, but always bidentate and of different shape. The females of R. derrai, bidens and andikithiriensis are unknown, and those of ayali (Karabag 1974, l.c. fig. 6), dernensis (Salfi 1929, l.c. fig. II-b) and lagrecai (Fontana & Massa 2004, l.c. figs 19, 20) differ in the subgenital plate.
The distinction between both new species and R. edentata Willemse 1982 (l.c. figs 9, 10, 24, 25, 43, 44) is also obvious. Its general appearance, coloration and habitat are typically Rhacocleis. Variation and distribution of R. edentata were insufficiently known. Apart from the island of Levkas, the species is also known from the area of continental Aitolia-Akarnania (a land bridge between Levkas and the continent, 2.vii.1986, lat 38°50′N, long 20°43′E, K.-G. Heller; Aetolia-Akarnania, 15 km N of Astakos, 50m 23.vi.1986, F. Willemse; Mt. Akarnanika above Voustrio, 1100m, lat 38°48′N, long 20°59′E, 3–4.viii.2003 & above Thirion, 1100–1400m, lat 38°50′N, long 20°58 and 59′E, 3.viii.2003, K.-G. Heller, M. Heller & M. Volleth) and the Ionian islands of Kefallinia (Kouvalada, 28.ix.1980, H. Malicky; Mt. Ainos 1000–1300m, 2–3.vii.1986 & surr. Dhavgata, 7 km N of Argostolion, 4.vii.1986, L. Willemse) and Zakinthos (Kato Gerakari 50m, 29.vi.1986, L. Willemse).
Apparently the male cercus of R. edentata does not, as its name would suggest, always lack an inner tooth, but may show an indication of such a tooth as a small tubercle, ranging even to a well-developed, though short and obtuse tooth (Figs 1–3). This variation of the male cercus is not just local but also individual. The male paraprocts are unmodified. The female of this species, previously unknown, is now at hand. The last abdominal sternites (Fig. 11) are unmodified and the basal sclerite of the subgenital plate is weakly sclerotised as in both new species; but the transverse ridges of the latter, converging towards the middle of the 7th sternite, are much stronger. In addition, the apical sclerite of the subgenital plate is not elongate triangular, but wide, heart shaped, well sclerotised, with a low but distinct median keel and a V-shaped, moderately deep apical incision.
We are much indebted to Baudewijn Odé (Amsterdam, The Netherlands) who described the song, provided oscillograms and bioacoustic comments. Dr. David Ragge, Dr. Klaus-Gerhard Heller, Dr. Glenn Morris and an anonymous referee are gratefully acknowledged for their comments on this paper. We are grateful to Klaus-Gerhard Heller and his family (Magdeburg, Germany) for their data and photos of R. crypta and loan of some material. We want to express our warm appreciation to our family members, Dominique Willemse and Tjeerd Blauw (Middelburg, The Netherlands) and Manuel Willemse and his son Joost Willemse (New Orleans, USA) for their unfailing enthusiasm and help in collecting during our joint field trips.