Blind Flight? A New Troglobiotic Orthoclad (Diptera, Chironomidae) from the Lukina Jama – Trojama Cave in Croatia

The genus Troglocladius Andersen, Baranov et Hagenlund, gen. n. is erected based on T. hajdi Andersen, Baranov et Hagenlund, sp. n. collected at 980 m depth in the Lukina jama—Trojama cave system in Croatia. Morphological features such as pale color, strongly reduced eyes and very long legs make it a typical cave animal. Surprisingly, it has also retained large wings and appears to be capable of flight which would make T. hajdi the first flying troglobiont worldwide, disproving previous beliefs that bats are the only animals capable of flying in complete darkness. Morphologically the new species does not readily fit within any described genus, but shares characteristics with genera both in the tribes “Metriocnemini” and “Orthocladiini”. Bayesian molecular phylogenetic analysis using the markers COI, 18S rDNAs, 28S rDNA, CADI, and CADIV groups it with the genera Tvetenia, Cardiocladius and Eukiefferiella in the tribe “Metriocnemini”. Troglocladius hajdi may be parthenogenetic, as only females were collected. The discovery confirms the position of the Dinaric arch as a highly important hotspot of subterranean biodiversity.


Introduction
Terrestrial species adapted to cave life are generally characterized as either troglophiles or troglobionts. While troglobionts are obligatory cave-dwellers, spending their entire life in the caves, troglophiles are essentially epigean species that are subdivided to eutroglophiles (able to maintain permanent subterranean populations) and subtroglophiles (dependent on epigean habitats for some biological functions). A third category of cave-dwelling organisms, the trogloxenes, are only accidentally found in caves [1]. Out of more than 21,000 terrestrial cave taxa worldwide [2] all troglobionts are ground dwellers. Bats fly in the complete darkness of caves using echolocation for navigating, they are however, troglophilic, and there are no known flying troglobionts.
The Dinarides or Dinaric Alps stretch along the western part of the Balkan Peninsula in southern Europe and are a part of the Alpine system along the east coast of the Adriatic Sea. According to the most recent census, the Dinaric karst harbors 600 terrestrial and 330 aquatic troglobionts [3]. Comparison of this number to the 1138 obligate cave taxa in United States [4] reveals the Dinarides as the world's richest area of obligate subterranean fauna when compared to other regions of approximately the same size [4]. The extreme richness of the area can also be portrayed via other fauna, including the world's sole troglobiotic representatives of Porifera, Cnidaria, Bivalvia and Polychaeta [5].
Although hundreds of different Diptera species have been recorded from caves, it has been stated that Diptera contains few true cave species, most are occasional guests or at most troglophiles (see e.g. [6,7]). However, some Diptera have evolved morphological and physiological adaptations for a troglobiotic life [7]. These adaptations are possibly most pronounced in the African Mormotomyia hirsuta Austen, 1936 (Mormotomyiidae), which is associated with bats in caves and regarded as troglophilic as it sometimes is found outside of caves. It lacks eyes, wings and pigments and has developed elongate, spindly legs densely covered with sensory setae [8]. In the family Sphaeroceridae, several species and subspecies are found in caves and some of these, such as Crumomyia absoloni (Bezzi, 1914), which is found in caves in northeastern Herzegovina, are pale and have reduced eyes, long legs and short wings and halteres [9][10]. Some of these adaptations are also found in other taxa including Allopnyxia patrizii Freeman, 1952 (Sciaridae) from caves in Italy, the obligate bat parasitic families Nycteribiidae and Strebliidae, and several species in the families Culicidae, Psychodidae, Mycetophilidae, Keroplatidae, Phoridae, and Heleomyzidae [7,11].
Chironomidae (Diptera) have been recorded from caves in many parts of the world (e.g. [12][13][14][15][16][17]). Most of these records are from the first few hundred meters from the cave entrance and, if identified beyond family level, the species recorded appear mostly to be trogloxenes. Reeves [18] categorizes Zavrelimyia nr thryptica (Sublette, 1964) from caves in the Great Smoky Mountains, U.S.A., as a troglophile, but until now there seems to be no records of any true troglobionts among the chironomids.
During expeditions to the Lukina jama-Trojama cave system in the Velebit Mountain Range in Croatia in 2013, several females of a pale chironomid belonging to the subfamily Orthocladiinae were collected in a chamber at 980 m below the surface. The female is described below and placed in a new genus based on morphological and molecular data. Furthermore, its habitat and natural history is outlined based on collection data and in situ observations.

The Lukina Jama-Trojama cave system
The Lukina jama-Trojama cave system is situated in the Hajdučki and Rožanski kukovi Strict Reserve, Velebit National Park, a 109 km 2 park in the northern parts of the Velebit Mountain in Croatia. There are several deep caves in the park with some of the world's largest subterranean vertical drops. The Lukina jama-Trojama cave system is 1.431 m deep and thus the deepest cave in southeast Europe and the 14 th deepest cave in the world.
The cave system has two vertical entrances at 1.438 and 1.475 m above sea-level ( Fig 1A) and in winter snow falls into the pits making an icy layer that covers the walls and extends approximately to a depth of 350 m in Lukina jama and 200 m in Trojama; some years, the Lukina jama entrance is completely sealed off by an ice cap. Air temperature gradually sinks and becomes more stable during the first 200 m of the cave system, as influence from the outside decreases (at -180 m it is around -0.7°C). From there, the temperature is very stable year round and slowly rises again to 2.5°C in the middle reaches (around Second Camp, Fig 1), in the chamber at 980 m depth it reaches 3.5°C, and at the bottom of the pit the temperature is around 4.9°C [19]. The first 200 m of the cave system has substantial presence of ice and snow year round. Water and a strong draught of cold air constantly flow down the shaft. These conditions, along with the cold temperatures make this part of the cave system unsuitable for subterranean fauna.
The new species has only been observed at depths between 800 and 980 m. Though it cannot completely be ruled out that the species is present in the upper parts of the Lukina jama-Trojama cave system, it is highly unlikely given the inhospitable conditions. The chamber at 980 m depth (Fig 1B-third camp) is about 85 x 70 meters in size and contains several habitats including dry walls, hygropetric environments with water seeping on the walls, small streams with sediment and a small pond with standing water-potential habitats for the immature stages.
A total of 54 animal taxa have been recorded from the Lukina jama-Trojama cave system, of which 32 are true cave dwellers [20]. Among these, the cave system houses the endemic snail Zospeum tholussum Weigand, 2013 and one of the largest known colonies of the subterranean leech Croatobranchus mestrovi Kerovec, Kučinić et Jalžić, 1999 [21][22].

Collection of specimens
Permission for collecting cave fauna was issued by The Ministry of Environmental and Nature Protection, Croatia. Altogether 16 females have been collected in the Lukina jama-Trojama cave system. During an expedition between 29 th July and 3 rd August in 2010 one female was found on the wet wall of the cave at about 800 m depth. During a second expedition between 1 th and 3 rd August 2011, two females were collected on the wall in the chamber at 980 m depth ( Fig 1C). During a third expedition between 8 th and 12 th August in 2013 altogether 13 females were collected using sticky traps placed directly on the walls of the cave. The specimens were preserved in alcohol and later mounted in Canada Balsam following the procedure outlined by Saether [23].

Morphological description
The general morphology follows Saether [24]. The measurements are given as ranges, followed by the mean when four or more specimens were measured, followed by the number of specimens measured in parentheses.

DNA extraction, PCR amplification and sequencing
The thorax and part of the abdomen of five specimens were kept in alcohol and used for DNA extraction. The DNA was extracted non-invasively using the QIAGEN DNeasy 1 Blood and Tissue kit and amplified with PCR following protocols as specified in S1 Text and S1 Table. The markers 18S rDNA, 28S rDNA and COI were amplified using primers as specified in Cranston et al. [25]; GenBank accession numbers and voucher collection numbers for the two successfully sequenced specimens are listed in the S2 Table. Two further markers, CADI and CADIV, failed to amplify in our study, but were nevertheless included in the final analysis for topological stability with respect to Cranston et al. [25].
The gene regions were aligned in Mega v6.0.6. [26] using a combination of MUSCLE [27] and manual alignment. Ambiguous regions were excluded in GBlocks v0.91b [28][29], with all the options set for a less stringent selection. This method was sufficient for 28S rDNA; however, for 18S rDNA the amount of hypervariable regions made it necessary to align manually according to secondary structure. The alignment and reference sequence are available in S2 and S3 Texts.

Molecular phylogenetic analysis
The sequences were compared with all Orthocladiinae sequences from Cranston et al. [25] using Diamesinae and Prodiamesinae as outgroups. The final data set consists of 3.709 characters, 1.560 of which are parsimony informative, in a matrix of 67 taxa. Analyses were carried out in PAUP Ã , MEGA and mrBayes v3.2.3, using Bayesian methods [26,[30][31]. Models were selected and tested using JModeltest [32] in combination with PAUP Ã . Bayesian analysis was performed using 10 million generations, 4 MCMC chain reactions and the model GTR+G+I as selected by JModeltest for all genes.

Type material
The holotype and 3 paratypes are deposited in the Croatian Biospeleological Society (CBSS) collection, which is a part of Croatian Natural History Museum collection in Zagreb, Croatia. Three paratypes, along with DNA extracts are deposited in the University Museum of Bergen, Bergen, Norway (ZMBN) and an additional paratype is deposited in the Zoologische Staatssammlung München, Munich, Germany (ZSM).

Nomenclatural acts
The electronic edition of this article conforms to the requirements of the amended International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix "http://zoobank.org/". The LSID for this publication is: urn:lsid:zoobank.org:pub:5CA9C540-440A-4439-8961-4D7BB2FBB781. The electronic edition of this work was published in a journal with an ISSN, and has been archived and is available from the following digital repositories: PubMed Central, LOCKSS, and Cristin (www.cristin.no). Etymology. From the Greek trogle, hole, and -cladius, a common ending of related Orthocladiinae genera. The gender of the name is masculine.

Troglocladius
Diagnosis. The pale color in combination with hairy eyes without, or with only a few ommatidia, reduced palp with palpomeres 2+3 and 4+5 most often fused, very long legs, and large wings and halteres will separate the genus from all other Orthocladiinae genera.
Description. Female. Medium sized species, wing length 1.8-2. Head. Eye reniform, strongly hairy, without or with few reduced ommatidia. Temporal setae in single row consisting of inner and outer verticals, postorbitals present or absent. Clypeus with few setae. Frontal tubercle absent. Tentorium and stipes normal. Cibarial pump with anterior margin straight. Palp short, palpomeres 2+3 and 4+5 fused in most specimens; third palpomere with few sensilla clavata apically.
Thorax. Antepronotum with lobes meeting medially at anterior margin of scutum, with few lateral antepronotals. Acrostichals short, starting some distance from antepronotum; dorsocentrals absent or few; prealars few; supraalar absent. Scutellum with few setae in single row. Haltere large.
Legs. All legs very long. All tibia with only one spur, hind tibial comb normal. Tarsal pseudospurs and sensilla chaetica absent. Pulvilli vestigial.
Systematics. The species is difficult to place systematically based on morphological characters alone, partly because only the female is known, and partly because it shows several morphological adaptations to cave life. Using the preliminary key to the females of Orthocladiinae in Saether [33] the species will key to Cricotopus van der Wulp, 1874 if the dorsocentrals are regarded as short and decumbent or to Paratrichocladius Santos Abreu, 1918 if the dorsocentrals are regarded as long and erect. It clearly has a divided gonapophysis VIII with a large ventrolateral lobe covering most of the dorsomesal lobe, as can be found in the Cricotopus group of genera. Further, tergite IX is divided in two setigerous protrusions, which also can be found in the Cricotopus group.
The phylogenetic analysis based on DNA sequences yielded a tree which overall is very similar to that given for Orthocladiinae by Cranston et al. [25]. The two specimens from the Lukina jama-Trojama cave grouped together and were well separated from all other genera in the analysis. They did not however, group with Cricotopus in tribe "Orthocladiini", but fell within tribe "Metriocnemini" together with the genera Tvetenia Kieffer, Cardiocladius Kieffer and Eukiefferiella Thienemann (Fig 2). Posterior probabilities for the other higher nodes within this clade were all significant (>98%), thus excluding the specimens from these sub-clades with high confidence. Posterior probability for the node within tribe "Metriocnemini" containing the new species, together with Tvetenia, Cardiocladius and Eukiefferiella was however, only 61% possibly due to missing data on CADI and CADIV. The mutual arrangement of the genera within this node is thus uncertain. One should also note that many Orthocladiinae genera remain to be sequenced and the internal relationships within tribe "Metriocnemini" may change.
Morphologically the new species does not group with the genera Tvetenia, Cardiocladius and Eukiefferiella, as all three have an undivided gonapophysis VIII [33] while the new species clearly has a divided gonapophysis VIII. A divided gonapophysis VIII with a large ventrolateral lobe covering most of the dorsomesal lobe and a divided tergite IX are however, traits found in many Orthocladiinae, including both the Cricotopus group and several genera within "Metriocnemini". Based on our current knowledge, the new species apparently does not fit into any described genus, thus supporting our placement of the species in a new genus within tribe "Metriocnemini".  Etymology. Named after the Hajdi, a group of winged, dwarf-like creatures from Slavic mythology, where they acted as messengers of fate and were said to dwell in caves. The name is to be regarded as a noun in apposition.

Discussion
True troglobionts are generally characterized by several morphological features like lack of integumentary pigment; most species are whitish or even translucent. Further, since there is no light in the caves, troglobionts often show some degree of eye reduction. They often have elongated legs and antennae; in cave environments long legs and appendages are clearly an advantage and are often coupled with a well-developed sensory apparatus. Most cave insects also lack wings or the wings are reduced [34].
The new species described here shares several of these adaptations. The body is pale yellowish ( Fig 5) and weakly sclerotized. The eyes are strongly reduced with only 0-4 ommatidia present. The legs are comparatively very long (Fig 6). The antenna appears short, but have few, long setae on segments 4-6 and strong sensilla on segments 1-5. Remarkably however, the wings are well developed, long and broad with an excavated anal lobe, and the halteres are also unusually large.
The combination of strongly reduced eyes and large, broad wings appears to be unique among troglobiotic organisms and might indicate that the species is able to fly slowly or hover in the total darkness of the cave. The long forelegs might serve as "feelers" if they are stretched forward during flight and the large halteres might help the insect maintain balance. During the 2013 expedition T. hajdi was never observed flying, but some of the specimens collected in the sticky traps were sitting in the middle of the strips, suggesting that they fly at least occasionally. During future expeditions, efforts should thus be made to observe the species' behavior in the cave. Of all free-living troglophilic or troglobiotic Diptera recorded so far, T. hajdi has by far the most troglomorphic characteristics. Furthermore, while other taxa are found close to the entrance or in comparatively small caves, T. hajdi inhabits a deep and highly isolated part of a large cave system with very little possibility to communicate with the outside environment. The discovery of the most troglomorphic winged Diptera described so far provides further support for the Dinaric karst as an area of extreme subterranean biodiversity.
The new species might well be parthenogenetic as only females were collected. In insects, parthenogenesis is most often found in harsh environments, where mating can be difficult. Parthenogenesis is not uncommon among chironomids and is most often found in extreme or isolated habitats [35]. Examples include Antarctic midges like Eretmoptera murphyi Schaeffer, 1914 or isolated island populations like populations of Limnophyes minimus (Meigen, 1818) on Gough and Nightingale Islands [36][37]. Parthenogenesis can also occur in habitat specialists restricted to fragmented habitats such as Polypedilum parthenogeneticum Donato et Paggi, 2008 living in water held in the leaf axils of Eryngium pandanifolium Chamb et Schlecht (Apiaceae) [38]. Although too few specimens of T. hajdi have been collected to confidently rule out the existence of males, we regard parthenogenesis as the most likely reproductive system considering the species' isolated habitat.
The larvae and pupae of T. hajdi are not known. During the expeditions in 2013 several larvae of the mycetophilid Speolepta leptogaster Madwar, 1937 were collected on the walls in the chamber at 980 m depth and wrongly thought to represent the larvae of T. hajdi. However, in the 980 m chamber there are several shallow streams with fine sediment, which were not sampled during the 2013 expedition and the larvae of T. hajdi might inhabit these streams. During coming expeditions to the Lukina jama-Trojama cave system or to other deep caves in that area, the immatures of T. hajdi should therefore be searched for. However, if endemic to the Lukina jama-Trojama cave system, great care should be taken not to collect too many specimens, as the populations might be very small and thus vulnerable.   Table 1. Lengths (in μm) and proportions of legs of Troglocladius hajdi Andersen, Baranov et Hagenlund, gen. nov., sp. nov., female (n = 4-5). Morphological abbreviations follow Saether [24].