The material was collected by Pavel Stoev and Boris Sket from the cave Kaptarhana (see details below) in the course of a rapid speleobiological assessment of the caves of Koytendag State Nature Reserve of Turkmenistan undertaken in 2015. The mission was carried out under the Memorandum of Understanding between the State Committee on Environment Protection and Land Resources of Turkmenistan and the Royal Society for the Protection of Birds to protect birds and other biodiversity in Turkmenistan. For sampling, pitfall traps, with Ethylene glycol and smelly cheese as bait, were set along the main gallery of the cave, mostly in humid places, in close proximity to large boulders and guano heaps. Pitfall traps were exposed for several days in the last week of May. Subsequently, all the captured animals were transferred to a container with 95% alcohol solution and properly labelled. Despite the fact that both collectors spent 7–8 hours altogether in the cave, no specimens were found by visual observations, all specimens being caught by pitfall trapping. The cave was very spacious, with large boulders and passages at different levels, making the process difficult for the collection and discovery of cryptic animals (such as diplurans) by methods other than pitfall trapping.

Specimens were washed in distilled water and inserted between slides and glass coverslips to be examined under a phase-contrast optical microscope (Leica DMLS) using Marc André II solution. The illustrations were made with a drawing tube and the measurements were taken with an ocular micrometer. For measuring the body length, the specimens were mounted “in toto” and were measured from the base of the frontal process distal macrochaetae to the abdomen’s supra-anal valve. For measurement of the sensilla and for examination of some minute anatomical parts, six specimens were coated with palladium-gold and studied in a Hitachi S-4100 scanning electron microscope.

The morphological descriptions and abbreviations used in this paper follow Condé (1956a). Although its function is still unknown, the term ‘gouge sensilla’ is used for the concavo-convexly shaped sensilla located on the antennae (as described by Bareth and Condé 1981). For the position of macrosetae on the occiput, ma, la and lp, Wygodzinsky (1944) was followed.

NMNHS National Museum of Natural History, Sofia.

For notal macrosetae: ma, medial-anterior, la, lateral-anterior and lp, lateral-posterior; for urotergal macrosetae: post: posterior.

Turkmenocampa mirabilis is hitherto known only from the cave Kaptarhana situated near the village Gurshun Magdany, Koytendag District, Lebap Province, Turkmenistan (Figs 20, 21). Kaptarhana (also spelt Kaptar-Khana) means ‘house of pigeons’ in Turkmen language as its entrance is used by pigeons for nesting. The cave is located at the foot of Koytentag Mountain (also known as Koytendag, Köýtendag, Kugitang, Koitendag, Kugitangtau, Kugitang-Tay, Kugitangtou) in the northern part of Hojapil Sanctuary, on the left bank of the Koyten river. The cave is approximately 450 m long and is situated in Late Jurassic gypsum (Birstein and Ljovuschkin 1965, Ljovuschkin 1969). There are two main galleries starting from the entrance, one orientated west (160 m long) and the second, approximately 300 m long, in a northeastern direction. Part of the cave is occupied by lakes with saline water. According to Birstein and Ljovuschkin (1965), the salinity is 11.68‰ and the pH is 7.8. The same authors provide a detailed chemical analysis of the water comparing it with the neighbouring river Amu Darya, Aral and Caspian seas and the World Ocean. The ionic composition of the lakes shows a high similarity to that of the ocean water and little resemblance to those of the neighbouring water basins. Additional support to the relict origin of the lakes comes from their rich foraminiferan, harpacticoid, isopod and gastropod fauna comprising a number of endemic species of marine origin.

Figure 21. 

Entrance of the cave Kaptarhana. Photo credit: Aleksandr Degtyarev.

Aquatic: Foraminifera: Entzia macrescens (Brady, 1870) =? Entzia zernovi (Schmalhausen, 1950), =?Entzia polystoma (Bartenstein & Brand, 1938) subsp. caspica Mayer, 1974 (see Filipescu and Kaminski 2011), Birsteiniolla macrostoma Yankovskaya & Mikhalevich, 1972, Trochamminita sp., Miliamina sp. Nematoda: Oncholaimidae spp.; Gastropoda: Hydrobiidae: Pseudocaspia ljovuschkini Y.I. Starobogatov, 1972; Isopoda: Microcharon halophilus Birstein & Ljovuschkin, 1965; Maxillopoda: Harpacticoida: Halectinosoma abrau (Krichagin, 1877), Ectinosoma sp., Schizopera paradoxa (Daday, 1904), Nitocra sp.

Terrestrial: Isopoda: Oniscoidea gen. sp.; Pseudoscorpiones; Araneae (new record); Collembolla (new record); Diplura: Turkmenocampa mirabilis; parasitic Diptera; Coleoptera: Cryptophagidae; Chiroptera: Rhinolophus bocharicus.

Although Turkmenocampa mirabilis has so far been found only in the larger gallery of the cave, some 200–250 m inside the cave, it might well be that it also inhabits the other main passage of the cave. The species is a troglobiont, all records deriving from the aphotic zone of the cave. No specimens were however observed during the exploration of the cave, those that were trapped being found in humid locations, rich in guano.

The classifications of Campodeidae (Condé 1956a; Paclt 1957) and Diplura (Pagés 1959) are rather outdated and badly in need of revision. Only some of the higher taxa proposed in Campodeidae seem natural as they appear to receive geographical support. This is particularly so for the diplurans from the Holarctics but, outside this region and towards its edges, most phylogenetic groups appear more or less artificial. The traits currently applied in the taxonomy of the group are of little help in clarifying the natural grouping and for developing a sound phylogenetic hypothesis. The shape and distribution of macrosetae and setae, the shape and complexity of pretarsal structures, as well as the secondary sexual characters, are the major taxonomic traits used for classifying the existing campodeid taxa. Molecular methods have only recently been applied to the group (Sendra et al. 2012) and the data are still insufficient for drawing a more robust phylogenetic analysis. Despite all these taxonomic weaknesses in the current classification, the traits demonstrated by Turkmenocampa mirabilis are solid enough to justify the description of a new species and genus within Campodeidae. The new taxon possesses a combination of several features not present in the other genera, one of which – the specific morphology of the olfactory chemoreceptor sensilla of the cupuliform organ – is unique in the whole family.

Plusiocampinae seems to be a paraphyletic taxon with regard to its only diagnostic character – the additional macrosetae on the pronotum – as suggested by Paclt (1957). Nevertheless, many of its genera can be considered monophyletic. This refers to the genera Cestocampa Condé, 1956, Condeicampa Ferguson, 1996, Hystrichocampa Condé, 1948, Patrizicampa Condé, 1956, Plusiocampa Silvestri, 1912, Plutocampa Chevrizov, 1978, Simlacampa Condé, 1956 and Vandelicampa Condé, 1956, all of which have lateral crests in their telotarsus. The absence of this important taxonomical trait in Silvestricampa Condé, 1950 and Turkmenocampa clearly distinguish them from the other Plusiocampinae. With regards to the subfamily position of the new genus, the presence of more than 3+3 macrosetae on pronotum (vs. up to 3+3 in Campodeinae Condé, 1956) and the absence of scales covering part of the body (vs. present as in Hemicampinae Condé, 1956 and Lepidocampinae Condé, 1956), places T. mirabilis within Plusiocampinae. However, the absence of lateral crests in T. mirabilis and all members of genus Silvestricampa, the latter being solely known from the Afrotropical realm (Silvestri 1913; Condé 1950), raises doubts about their placement in Plusiocampinae.

The high number of macrosetae in Silvestricampa, with 7+7 macrosetae on the pronotum and presence of medial and lateral intermediate macrosetae on meso- and metanotum as well as the absence of telotarsal processes, clearly differentiate Turkmenocampa from Silvestricampa. Close examination of the genera Plusiocampa and Cestocampa, reveals that several species do not match their original diagnoses. This is the case with subgenus Dydimocampa of Plusiocampa defined by Paclt (1957) with the presence of two dorsal femoral macrosetae. The subgenus is known with two species from China (Condé 1993; Silvestri 1931): the soil-dwelling Plusiocampa (Dydimocama) sinensis Silvestri, 1931 and the cave-dwelling Plusiocampa (Dydimocampa) lipsae Condé, 1993. Their chaetotaxy show close similarities with T. mirabilis except for the additional macrosetae on II-VII urosternites present in P. lipsae. Furthermore, P. lipsae has laminar telotarsal processes although with a short sternal pubescence, while they are setiform in P. sinensis. P. sinensis has clear lateral crests while they are very small in P. lipsae.

Another allied species, the soil-dwelling Plusiocampa kashiensis from West China (Chou and Chen 1980), was recently transferred from Cestocampa (Sendra et al. 2012). Despite its poor original description, P. kashiensis shares some similarities with T. mirabilis in the distribution of macrosetae on the nota and abdomen and also in having barbed laminar telotarsal processes and claws without crests. However, the species can readily be distinguished from T. mirabilis by the short and abundant clothing setae (vs. long and thin in T. mirabilis) and the lack of extension of the claws or side-shoot (at least not mentioned in its original description). It might well be that P. kashiensis is actually a member of Turkmenocampa but, until a new or type material is studied, a formal transfer has not been suggested at this time. This also refers to the genus Anisuracampa Xie & Yang, 1990 which was described from subtropical China (Xie and Yang 1991) and shows morphological similarities with Dydimocampa.

Finally, worthy of special mention is the presence of the latero-outside side-shoot of the claw in T. mirabilis. It can be considered as a convergent character as it is known in Metriocampa (Notocampa) afra Condé, 1950 and is also present basally in Oreocampa minutella (Silvestri, 1918), as well as in Haplocampa Silvestri, 1912. It has never been found in combination with telotarsal processes (Condé 1956a) until the discovery of T. mirabilis.

It should be emphasised that all Diplura are without external eyes, although George (1963) presumably gave a light-perceptive function to its lateral sense organs, each one being below the integument in the latero-ventral position in the head. Furthermore, Diplura and Campodeidae, in particular show thin integument with no pigment or sclerotized cuticle. These traits are in all soil- and cave-inhabiting species. Nevertheless, the features which clearly define the troglobiotic campodeids are: increased body size, elongation of appendages and body, more numerous antennomeres and cercal articles, as well as specialisation of sensory organs. Considering all these features, Turkmenocampa mirabilis is undoubtedly a strict cave-dweller showing slightly elongated body and appendages (including the cercal articles and the antennomeres); and a moderate increase in the number of antennomeres reaching up to 33 and up to 10 elongated gouge sensilla in the medial and distal antennomeres. Furthermore, the species possesses three striking features which could also be related to its subterranean living environment (see also Condé 1956a; Sendra et al. 2017).

Firstly, the olfactory chemoreceptors within the cupuliform organ, which are usually present in troglobiotic campodeids. In T. mirabilis, the shape of these olfactory chemoreceptors have no analogue in any other Campodeidae (Figs 8–12). All troglobiotic campodeids show an increase in size and complexity of the olfactory chemoreceptors and their surface. Troglobionts have more and larger pores than soil-dwellers (Juberthie-Jupeau and Bareth 1980). The increase in complexity demonstrates a similar pattern showing a multiplication of tiny folds (or collarets) around a spheroid sensilla (Condé 1956a). In the most complex cases (Bareth and Condé 1984; Condé 1974; Condé and Sendra 1989), these folds have a sheet-like shape, in Plusiocampa dallai Bareth & Condé, 1984 and Plusiocampa alhamae Condé & Sendra, 1989; or digit-shape, in Campodea (Paurocampa) pretneri Condé, 1974.

Inside the cupuliform organ, T. mirabilis has three types of olfactory chemoreceptors, about forty sensilla tightly packed in a shallow cuticular invagination perforated by tiny pores (Figs 8–9). These sensilla are produced by microscopic evagination at the bottom of the cupuliform organ producing three different types of olfactory chemoreceptor sensilla. Type I and II have a wide and very short base, difficult to observe, which increases in size into two oval-shaped structures completely covered by pores. In type I, the oval structures are 7–8 µm long by 5–6 µm wide with pores of 0.1–0.25 µm diameter (Figs 9–10); in type II, the oval structures are 3.5–4 µm by 3–4 µm with pores of 0.25–0.35 µm (Figs 9, 11). Type III are tree-shaped sensilla covered by irregular pores (0.05–0.08 µm diameter) that start from the cupuliform organ base with a trunk-shaped structure 4.5–6 µm high and 1.1–1.4 µm wide, that it is divided into 2, 3 or 4 branches of 1.1–2.6 µm long extending into 1-2-3-4 spines (Figs 9, 12). The type III sensilla slightly overhang the types I and II and are mostly (Fig. 8) in the centre of the cupuliform organ surrounded generally by types I and II.

Thus in T. mirabilis, the increase in number, complexity and porous surface in olfactory chemoreceptors sensilla follow another evolutionary path, different from the pattern observed in numerous troglobiotic species in Plusiocampa and Cestocampa, where these sensilla have been examined (eg., Condé 1956a).

The other two remarkable features of T. mirabilis refer to the telotarsus and both could be an adaptation for walking on wet and soft surfaces in subterranean environment. The claws are 50–60 µm long and curved only at the distal end where they are also slightly thinner. The whole surface of the claws is marked with fine longitudinal and semi-transversal striate of 0.3 µm thickness. At approximately 15 µm from the base on the lateral external side of the claw, there is a pointed side-shoot process of 12–15 µm length (Figs 17–18). If side-shoots of the claws have a similar function to the function of lateral crests present in several troglobiotic species of Plusiocampa, Cestocampa, Paratachycampa and Juxtlacampa, which are usually regarded as adaptations to subterranean lifestyle (see, for example Condé 1956a; Bareth and Pagés 1994), then they should also be considered derived traits resulting from the long cave evolution. The third interesting feature is the shape of the barbs (about 120 thin barbs 3–18 µm long) on the ventral side of the laminar telotarsal processes, ending mostly in a hook but in a flat extension in the apical barbs (Fig. 17). These two types of laminar barbs have also been observed in other troglobiotic campodeids (Sendra et al. 2012, 2016) and were considered as an adaptation to facilitate movement on wet surfaces (Sendra et al. 2017).

The cave fauna of Central Asia is outstanding with its poverty of terrestrial troglobionts (Birstein and 1967, Ljovuschkin 1969). Kniss (2001) enumerated from the Central Asian caves altogether 80 species, of which 27 were stygo- and troglobionts. However, out of 27 strict cave-dwellers, only the springtail Pseudacherontides stachi (Ljovuschkin, 1972), found from the Amir-Temir Cave on the western spur of Zeravshan Range of Uzbekistan, is considered troglobiont (Turbanov et al. 2016b). All others, including one fish, inhabit underground waters. Likewise, in adjacent Iran, the cave fauna comprise only 89 species, of which 16 are strict cave-dwellers. Of these, only three are terrestrial troglobites – the spider Trilacuna qarzi Malek Hosseini & Grismado, 2015, the millipede Chiraziulus troglopersicus Reboleira, Malek Hosseini, Sadeghi & Enghoff, 2015 and the isopod Protracheoniscus gakalicus Kashani, Malek Hosseini & Sadeghi, 2013 (Malek-Hosseini and Zamani 2017).

Based on climatic, lithological and soil characteristics, Turkmenistan is divided into thirteen ecological regions. Koytendag Mountains form a region of its own and is characterised by desert landscapes on mountainous relief, highly dissected by ravines, foothills with ridges and cuestas and fan plains. Karst processes are well developed in the region. The average annual temperature is about 17°C and annual precipitation is approximately 150 mm (Babaev 1994). The flora contains more than 1,900 species, including 332 endemics (see Rustamov et al. 2009). Lying at the intersection of three biomes – the Eurasian high mountains (Alpine and Tibetan), the Irano-Turanian mountains and the Sino-Himalayan temperate forests – the area supports a high faunal and floral diversity with a number of endemic plant, fish and invertebrate species (UNESCO Nomination dossier 2015).

Due to its remote location, difficult accessibility and restricted border control, as well as the lack of active speleobiologists in Turkmenistan, the biological aspect of the Koytendag caves has only been marginally studied. Despite the great number of caves in the area (some estimates give 300), until now only the invertebrate fauna of the caves Kaptarhana, Gap-Gotan, Hashym Oyuk and Gulshirin, an unnamed cave near v. Svincovyi rudnik, have been explored from the biological viewpoint (Birstein and Ljovuschkin 1965, Ljovuschkin 1969, Starobogatov 1972, Kniss 2001, Turbanov et al. 2016a–c). Kniss (2001) reviewed the existing knowledge in his catalogue “Fauna of the caves of Russia and adjacent countries” and Turbanov et al. (2016a, b, c) provided a checklist of all cave species known from Russia and the former Soviet republics. Until the present time, terrestrial troglobionts have not been registered in Koytendag.

The specific composition of the brackish lake in the cave Kaptarhana and its unique stygofauna comprised of species of marine origin suggest a completely different geological history of the cave compared to the rest of the region. It is very likely that the saline waters belong to another hydrographic entity, without connection to the subterranean waters of the neighbouring Koyten and Garlyk areas where the stygofauna is represented by other species such as Troglocobitis starostini, Stenasellus asiaticus, Bogidiella ruffoi, Gammarus spp., copepods, etc. (Turbanov et al. 2016a, b, Boris Sket, unpublished).

It is noteworthy that the terrestrial fauna of the cave also shows differences compared to the other caves of Koytendag. The authors’ attempts to find T. mirabilis in any of the other caves were unsuccessful despite the fact that the same collecting method (baited pitfall traps) was applied in the cave Gap-Gotan. Furthermore, the ptinid beetle Niptus hololeucus (Faldermann, 1835), which is otherwise very abundant in the Gap-Goutan cave (mostly on porcupine scats), is missing in Kaprahana, where the group is represented by an unidentified species of family Cryptophagidae. The unique character of the fauna of Kaptarhana is supported by the finding of very likely new species of Collembola (L. Deharveng, in progress). It may well be that the existing hydrological barrier between Kaptarhana and the other caves also prevents distribution of terrestrial organisms.

Taking into consideration that the caves of Central Asia are poorly studied, the possibility is not excluded that this taxon or new species of Turkmenocampa will be found in future in other caves in Koytendag or in the neighbouring parts of Uzbekistan and Afghanistan.