Description and DNA barcoding of Tipula ( Pterelachisus) recondita sp. n. from the Palaearctic region (Diptera, Tipulidae)

Abstract Tipula (Pterelachisus) recondita Pilipenko & Salmela, sp. n. is described. The new species is collected from two localities: Finland, Kittilä (North boreal ecoregion) and Russia, Primorski kray (Zone of temperate broadleaf and mixed forests). Although variation in the structure of male hypopygium between the Finnish and Russian populations is observed, DNA barcode sequences differ only by three nucleotides (0.2 % K2P distance), supporting presence of one widespread species. K2P minimum distances between the new species and 17 other species of the subgenus range from 5.3 to 15.8 % (mean 8.8 %). The new species is forest-dwelling, known from an old-growth herb-rich forest (Finland) and Quercus mongolica forest (Russia). The new species is perhaps closest to Tipula (Pterelachisus) imitator Alexander and in lesser extent to Tipula (Pterelachisus) pauli Mannheims; the inner gonostylus of both species are illustrated.

introduction Tipulidae (Diptera, Nematocera), or long-palped crane flies, are medium to large sized true flies. Globally, 4269 tipulid species and subspecies are known, of these 1322 occupying the Palaearctic region (Oosterbroek 2011). In general, northwest European tipulid fauna is rather well known (e.g. Salmela 2010Salmela , 2011. The majority of the species have large European or Palaearctic ranges, only a few species are known from Fennoscandia or Russian Karelia alone. On the other hand, certain species have disjunct occurrences in northern Fennoscandia and the East Palaearctic region (viz. Tipula kaisilai Mannheims, T. subexcisa Lundström, T. tchukchi Alexander).
Tipula (Pterelachisus) Rondani is a northern hemisphere subgenus, totaling over 200 species and subspecies (Oosterbroek 2011). Despite taxonomic monographs covering Russia (former USSR, Savchenko 1964) and Europe (Theowald 1980), the Palaearctic fauna of the subgenus includes several elusive species, known from the type locality or female specimens only. Finnish Tipula (Pterelachisus) species were reviewed by Salmela (2009) and those of the Central European territory of Russia were listed by Pilipenko (2009). The subgenus Pterelachisus is closely allied to Lunatipula Edwards and Savtshenkia Alexander, but is diagnosable due to the bare squama, grayish coloration, patterned wings and structure of male hypopygium (Theowald 1980). Savchenko (1964), dealing with the fauna of former USSR, recognized 11 species groups and two mixed groups within Tipula (Geotipula) and T. (Oreomyza). These subgenera were synonymized to Pterelachisus by Alexander (1965) and later Theowald (1980) named 18 species groups from the West Palaearctic region. These species groups are mainly based on differences in the structure of male hypopygium (Theowald 1980), but no cladistic analysis or phylogeny of the species groups was provided by Theowald or authors after him.
DNA barcoding is a molecular-based method used in the identification and delimitation of species, having usually considerable congruence with morphologybased identifications (Ward et al. 2006, Hausmann et al. 2011, Park et al. 2011. Furthermore, barcoding has revealed cryptic diversity within a morphospecies (Smith et al. 2006, Huemer and or indicated a presence of one species despite morphological variation within studied specimens (Memon et al. 2006). DNA barcoding has its disadvantages (Meier 2008, Skevington et al. 2007, Taylor and Harris 2012, but it may be used as an additional, and apparently very powerful, method in taxonomy (Schlick-Steiner et al. 2010). Despite the wide use of DNA barcodes in the current taxonomy and biodiversity studies, the method has been only rarely used in taxonomic studies of crane flies (Ujvárosi et al. 2009, Ujvárosi andBálint 2012).
In the present article we provide a description of Tipula (P.) recondita Pilipenko & Salmela sp. n. collected from Europe (Finland) and Asia (Russian Far East). Both sexes of the new species are richly illustrated. In addition, mtDNA sequences (COI) were used to assess (i) the conspecific status of disjunct Finnish and Russian populations and (ii) genetic divergence between the new species and 17 consubgeneric species.

Material and methods
Total DNA of Tipula (P.) recondita Pilipenko & Salmela sp. n. specimens was extracted using a modified non-destructive salt extraction method (Aljanabi & Martinez 1997, Gilbert et al. 2007. Whole holotype (JES-20110034) and one paratype (JES-20110035) adult specimens and one leg from a paratype (JES-20110036) were placed on 250 μl 96-plate wells. Ethanol-stored samples were briefly dried at 60 °C. First 118 μl of sterile salt homogenizing buffer (0.4 M NaCl, 10 mM Tris-HCl pH 8.0, 2 Mm EDTA pH 8.0 and 2% SDS) containing 8 μl of 20 mg/ml proteinase K (400 μg/ml final concentration) was added into each well. The samples were incubated overnight in the buffer at 55-65 °C. After the incubation, the intact samples were removed from the buffer and placed into 99.5% ethanol to stop further digestion. Type specimens JES-20110034 and JES-20110035 were finally preserved in 70 % ethanol. Then 80 μl of 6 M NaCl (NaCl saturated H 2 O, pH 8) was added to each well. Samples were vortexed for 1 min at maximum speed, and centrifuged for 20 minutes at 4000 rpm. Thereafter 100 μl of supernatant was transferred to wells on a new plate. An equal volume (100 μl) of isopropanol was added to each sample and the plate was briefly vortexed. Then the plate was placed into freezer (-20 °C) for 1 hour. After freezing, the samples were centrifuged for 20 minutes at 4000 rpm. The supernatant was discarded and the pellet was washed by adding 150 μl of ice-cold 70% ethanol and centrifuging for 20 min at 4000 rpm. The ethanol was then carefully pipetted out and the pellet was dried for overnight at room temperature. The next day, DNA pellet was dissolved in 50 μl of previously warmed ultrapure water.
For other species (totaling 17 species and 26 specimens, Table 1) DNA barcodes were obtained at the Canadian Centre for DNA Barcoding. Legs or 2-3 abdominal segments of the specimens were placed in 96% ethanol in a 96-well lysis microplate and dispatched to the Biodiversity Institute of Ontario where DNA was extracted and sequenced using standard protocols and primers (deWaard et al. 2008). Resultant sequence data were placed into a project (HOLPT) on BOLD (http://www.boldsystems.org, Ratnasingham and Hebert 2007). These sequence records are now publically available on both BOLD and on GenBank.
In order to assess the COI divergence between the new species and 17 Holarctic Tipula (Pterelachisus) species, we calculated Kimura two-parameter (K2P) (Kimura 1980) distances between all sequenced specimens. Based on K2P distances we also produced Neighbor-Joining (NJ) tree to visualize similarity of the Pterelachisus species.
We also inferred relatedness of the species with character based Maximum Likelihood (ML) method (GTR + gamma as evolutionary model, 1000 Bootstrapping replicates). However, because the NJ and ML trees were practically identical, only ML tree is presented ( Fig. 1). K2P distances, NJ and ML were produced by using MEGA5 program (Tamura et al. 2011). Because one gene is far too little for reasonable phylogenetic analysis (Gatesy et al. 2007), the ML tree presented here do not reliably illustrate evolutionary relationships among the sequenced taxa. The morphological terminology used here mainly follows Alexander and Byers (1981). Terminology of some special parts of male genitalia was taken from Frommer (1963). If not otherwise stated, measurements are given in μm. The following acro-nyms for museums and collections are used in the text: ZMKU -Zoological Museum of National Museum of Natural History, National Academy of Science of Ukraine, Kiev, Ukraine; NCBN -Netherlands Centre for Biodiversity Naturalis, Leiden, the Specimens were studied with a Zoom Stereo Microscope. Photographs were taken with a Canon PowerShot A640 camera and processed using Combine ZP software. All drawings were prepared from photographs.
[ Diagnosis. Rather small yellowish brown Tipula species (body length: 11 mm male, 12.3 mm female; wing length 11-12.6 mm male, 12.5-13.5 mm female). Scape, pedicel and base of 1 st flagellomere yellowish, other flagellomeres brown. Caudal margin of male 9 th tergite with a median notch, bearing no tooth or other elevated structures. Outer gonostylus narrow, about as long as inner gonostylus, slightly bent subbasally. Lower beak of inner gonostylus apically rounded, black. Outer basal lobe of inner gonostylus with 3-4 stout black spines.
Etymology. The species epithet is from reconditus (Latin, adjective) meaning hidden, concealed. This word refers to the rarity and apparent low detectability of the new species, so far known only from two sites in the Palaearctic region.
Distribution and ecology. Tipula (P.) recondita Pilipenko & Salmela, sp. n. is known from North Europe (Finland) and Asia, Russian Far East. The Finnish collecting site in Kittilä, Iso Mustavaara, is a state-owned Nature Reserve (Lehtojensuojelualue), included in the Natura2000 network of conservation areas. It is part of the biogeographical province of Lkoc (Lapponia kemensis pars occidentalis) and lies in the North boreal vegetation zone. The collecting site is an old-growth mixed forest, dominated by birch (Betula pubescens), goat willow (Salix caprea) and Norway spruce (Picea abies), with scattered aspen (Populus tremula) trees. Lower vegetation is characterized by herbs and shrubs such as Calypso bulbosa, Daphne mezereum, Actaea erythrocarpa, Ribes spicatum, Filipendula ulmaria and Geranium sylvaticum. Decaying trees, especially goat willow and birch, are abundant in the site. The Russian collecting site is located in the Kedrovaya Pad' Nature Reserve, within the temperate broadleaf and mixed forest zone, in an oak forest (Quercus mongolica) growing on limestone outcrops on the southern slope of a mountain range. Lower vegetation is characterized by Lespedeza bicolor, Spodiopogon sibirieus, Astra ageratoides, Carex siderosticta, Artemisia keiskeana, Lathyrus davidii and Calamagrostis brachytricha.
Morphologically the new species is perhaps the closest to two Palaearctic species, namely T. (P.) imitator and T. (P.) pauli. The former species has a median notch in 9 th tergite, but also a distinct tooth at the midpoint (Fig. 4g); the outer basal lobe of inner gonostylus bears one conspicuous black spine, not 3-4 smaller ones (Fig. 4e). For other differences, see Figures 4f, h, i. Tipula pauli also has a median notch in 9 th tergite and a small but discernible tooth in the midpoint; the lower beak of inner gonostylus is roundish and black, but the outer basal lobe bears no stout, black spines (Fig. 4d). Tipula (P.) imitator is known from Japan and Kuril Islands and T. (P.) pauli from Europe, Altay and Russian Far East (Oosterbroek 2012, V. Pilipenko pers. obs.).
Based on COI divergence, the new species is apparently rather isolated from the members of the subgenus Pterelachisus (Fig. 1). Among the other species vs. the new species, interspecific distances varied from 5.3 % (Tipula winthemi Lackschewitz) to 16.1 % (Tipula laetibasis Alexander). Mean of the minimum interspecific distances was 8.8 %. According to K2P divergence, the new species is closest to T. winthemi (5.3 %), T. jutlandica Nielsen (5.5 %), T. stenostyla Savchenko (6.6 %) and T. pauli (6.8 %); distances between the other species range from 7.4 to 16.1 %. In other words, no very close relatives were present in the pair-wise comparisons of COI sequences. For example, much shorter interspecific K2P distances were found between T. varipennis/T. pseudovariipennis (1.5 %), T. mutila/T. wahlgreni (2.2 %), T. stenostyla/T. winthemi (3.7 %). However, it must be emphasized that T. imitator was not included in COI analysis, due to the lack of fresh material. Given to the morphological similarity of the new species and T. imitator, it is likely that their barcoding distances would be similar to those three comparisons given above.
There are some morphological differences (9 th tergite, inner gonostylus) between Finnish and Russian specimens, perhaps due to the long distance and lack of gene flow between the populations. These differences, however, are here considered to be intraspecific variation. Very small K2P divergence of COI gene (0.2 %) between Finnish and Russian specimens also substantiates the presence of one widespread, but disjunct, species. In rare cases (see Burns et al. 2007) differences of only one to three nucleotides may be observed between otherwise (for example morphologically and ecologically) distinct species. However, in this case we were able to produce 524 bp of high quality sequence, instead of <400 as in the problematic cases of Burns et al. (2007). Moreover, the known biologies of the Finnish and Russian individuals seem alike. To say more of the COI variation, it would be essential to collect more individuals which is rather difficult, given the rarity of the species.
The new species is most probably a very rare tipulid. Despite the rather long tradition of crane fly taxonomy and faunistics in North Europe, this species has hitherto remained unnoticed. One of the authors (JS) has within 12 years identified some 70 000 crane flies from a Finnish Malaise trapping material consisting of 476 sampling sites and ca. 1670 Malaise trapping months. Thus, despite this relatively large sampling effort, only three specimens from a single locality have been caught. The true range of the species is Palaearctic, whether disjunct or not remains to be seen. In Northwestern Europe the species is likely to occur in the north boreal zone (for further information on boreal ecoregions or vegetation zones, see e.g. Ahti et al. 1968). Tipula (P.) recondita Pilipenko & Salmela, sp. n. may be confined to old-growth forests, and its rarity is perhaps due to the narrow habitat niche. On the other hand, the new species may be hard to collect using traditional methods. Larval associations of this species are unknown, but some T. (Pterelachisus) species are saproxylic, i.e. dependent on decaying trees. Such species are e.g. T. (P.) pseudoirrorata Goetghebuer and T. (P.) stenostyla Savchenko (Salmela 2009), both of them also encountered in the type locality.
To our regret we were not able to examine the holotype male of T. imitator (D. Furth, pers. comm.). Description of that species was based on a single male specimen (Alexander 1953). We have however examined other material (two male specimens, see above) that very likely represents T. imitator. Despite morphological similarity of T. (P.) recondita Pilipenko & Salmela, sp. n. and T. imitator, we are confident that these are separate taxa, due to the differences in the structure of male hypopygium.