Molecular phylogeny of the Trechus brucki group, with description of two new species from the Pyreneo-Cantabrian area (France, Spain) (Coleoptera, Carabidae, Trechinae)

Abstract A molecular phylogeny of the species from the Trechus brucki clade (previously Trechus uhagoni group)based on fragments of four mitochondrial genes and one nuclear gene is given. We describe Trechus (Trechus) bouilloni sp. n. from the western pre–Pyrenees: Sierras de Urbasa–Andía, Navarra, Spain. The species was collected in mesovoid shallow substratum (mss), a subterranean environment. Molecular as well as morphological evidences demonstrate that the new species belongs to the Trechus brucki clade. A narrow endemic species of high altitude in western French Pyrenees merged with Trechus brucki Fairmaire, 1862a, Trechus bruckoides sp. n., is described. A lectotype is designated for Trechus brucki and Trechus planiusculus Fairmaire, 1862b (junior synonym of Trechus brucki). The species group is redefined based on molecular and morphological characters, and renamed as the brucki group, as Trechus brucki was the first described species of the clade. A unique synapomorphy of the male genitalia, a characteristic secondary sclerotization of the sperm duct, which is shared by all the species of the brucki group sensu novo, is described and illustrated. The Trechus brucki group sensu novo is composed of Trechus beusti (Schaufuss, 1863), Trechus bouilloni sp. n., Trechus brucki, Trechus bruckoides sp. n., Trechus grenieri Pandellé, 1867, T. uhagoni uhagoni Crotch, 1869, T. uhagoni ruteri Colas, 1935 and Trechus pieltaini Jeannel, 1920. We discuss the taxonomy of the group and provide illustrations of structures showing the differences between the species, along with distribution data and biogeographical comments.

In this paper we describe a species collected by traps in a MSS (mesovoid shallow substratum, "Milieu Souterrain Superficiel" sensu Juberthie et al. 1980, Giachino andVailati 2010) in the Sierras de Urbasa-Andía (Western pre-Pyrenees, Navarra, Spain) and a second orophilic species from the French Central Pyrenees. We study the phylogenetic relationships of the new species and provide a molecular phylogeny of the group, including all known species but four.

Historical background
Trechus uhagonii was described by Crotch in 1869 and dedicated to S. de Uhagon with whom he visited caves in the Alsasua area in June 1869. Trechus bruckii was first described under the name T. politus by Fairmaire (1862b). He renamed it one year later to bruckii because politus was already in use for an American species. The two names were corrected to uhagoni and brucki by subsequent authors, and recently renamed uhagonii and bruckii in catalogues (Lorenz 1998, 2005, Moravec et al. 2003, Queinnec and Ollivier 2011. As the names uhagonii and bruckii have not been used since their description, we choose to keep the prevailing usage of uhagoni and brucki in accordance with the article 33.3.1 of the International Code of Zoological Nomenclature on incorrect subsequent spellings (ICZN 1999).
In the Monographie des Trechinae,  erected the uhagoni group for the seven species of Trechus from the Pyreneo-Cantabrian area quoted above. Español (1970) described T. ortizi from a cave of Burgos province (Spain), and included it in the uhagoni group, close to T. bordei.
The uhagoni group sensu , although poorly defined morphologically, was enriched with 5 species by Casale and Laneyrie (1982) in their catalogue of species of world Trechinae: T. pecoudi Colas & Gaudin, 1935 (described first as a subspecies of T. brucki), T. ortizi, T. escalerae Abeille de Perrin, 1903(considered by Jeannel (1927 to belong to the T. angusticollis group), T. enigmaticus  and T. aubryi Coiffait, 1953. Trechus uhagoni was here considered as subspecies of T. grenieri. By describing Trechus baztanensis from a cave of Navarra, Dupré (1991) suggested that the peculiar genital morphology of T. bordei, T. navaricus, T. bonvouloiri and the new species should lead to their removal from the uhagoni group, and he created the bonvouloiri group for these species, opinion followed by Queinnec and Ollivier (2011). Toribio and Rodríguez (1997) added one species from Cantabria to the uhagoni group, T. carrilloi, a species collected in the MSS. Sciaky (1998) described T. jeannei from Cantabria, and included it in the uhagoni group, close to T. bordei and T. ortizi. Hernando (2002) described T. comasi from a cave of Navarra and suggested that it should be considered as sister species of T. brucki. Molecular and morphological evidence suggest that this species should be removed from the uhagoni group (Faille et al. 2010a, 2011a, Ortuño and Arribas 2010. Ortuño and Toribio (2005) described a new species of Trechus belonging to another group of species, indicating that 11 species belong to the uhagoni group in the Iberian Peninsula. Reboleira et al. (2010) considered that 10 species belong to this group in the Peninsula, without providing the list of taxa included.
T. brucki is an alpine species located at high altitude in the central and western Pyrenees, and it is until now not recorded from the Spanish slope of the chain (Serrano 2003). Colas and Gaudin described T. pecoudi in 1935 from the western Pyrenees (Pic d´Orhy) as a subspecies of T. brucki.   described 3 subspecies of T. brucki; T. brucki vandeli, T. b. truilheti and T. b. microthorax. The subspecies vandeli and truilheti were later related to T. pecoudi (Casale & Laneyrie, 1982) so that T. pecoudi counts three subspecies in the Catalogue of Palearctic Coleoptera (Moravec et al. 2003). Queinnec and Ollivier (2011) considered T. pecoudi as a subspecies of brucki restricted in the Anie and Orhy massifs, whereas the subspecies vandeli, described from Anie, was considered a synonym of T. brucki brucki together with the subspecies truilheti and microthorax.

Taxon sampling, Morphological study, DNA extraction and sequencing
Specimens were collected by hand or by means of pitfall traps containing water saturated in salt or propylene glycol, known to preserve DNA (Rubink et al. 2003, López and (Table 1). The protocol is detailed in Faille et al. (2010b): Extractions of single specimens were non-destructive, using the DNeasy Tissue Kit (Qiagen GmbH, Hilden, Germany). After extraction, specimens were mounted on cards and genitalia stored in water-soluble dimethyl hydantoin formaldehyde resin (DMHF) on transparent cards, pinned beneath the specimen. Vouchers and DNA samples are kept in the collections of ZSM, IBE and MNHN.
We included examples of most species of the T. uhagoni group, with the exception of T. bruckoides sp. n., T. carrilloi and T. sharpi and some examples of Trechus of the angusticollis group sensu  and Casale and Laneyrie (1982) (Table 1). The tree was rooted with Aphaenops leschenaulti Bonvouloir, 1862, which is known to belong to a different group of Pyrenean Trechini , Faille et al. 2010a.
We amplified fragments of four mitochondrial genes: 3' end of cytochrome c oxidase subunit (cox1); a single fragment including the 3' end of the large ribosomal unit (rrnL), the whole tRNA-Leu gene (trnL) and the 5' end of the NADH dehydrogenase 1 (nad1); and one nuclear gene (internal fragment of the large ribosomal unit 28S rRNA, LSU) (see Table 2 for primers used). Sequences were assembled and edited using Sequencher TM 4.8 (Gene Codes, Inc., Ann Arbor, MI). Parts of the sequences for 14 of the species were taken from Faille et al. (2010a) and Faille et al. 2011a (Table 1).
New sequences have been deposited in the EMBL database with Accession Numbers HE817887-HE817940 (Table 1).

Phylogenetic analyses
We aligned the sequences using the MAFFT online v.6 and the Q-INS-i algorithm (Katoh and Toh 2008), a progressive pair-wise method with secondary refinement. We used Maximum Likelihood as implemented in the on-line version of RAxML (which includes an estimation of bootstrap node support, Stamatakis et al. 2008), using GTR+G as the evolutionary model and three partitions corresponding to the cox1, rrnL+trnL+nad1 and LSU fragments.
Diagnosis. Large size (ca 5 mm) and round shape (Fig. 1). Median lobe of aedeagus slender, in lateral view (Fig. 15) the basal third curved, the central part straight and the apex with a curved hook assymetrical in dorsal view (Fig. 8). Inner sac of aedeagus (=endophallus) with an elongate and well-sclerotized piece, forming a gut and armed with internal scales. Characteristic secondary sclerotization of the sperm duct ( Fig. 15: CP2) forming a kind of second copulatory piece outside base of the median lobe.
Description of the holotype. Habitus as in Fig. 1. Elongated, round-sided. Body surface with a very thin, hardly visible, dense microreticulation, with more distinguishable meshes on the head.
Colour. Dorsal surface dark brown, moderately shiny. Antennae, palpi and legs light brown.
Chetotaxy. Surface of elytra glabrous with the exception of a periscutellar seta, two discal setae on the third stria, four humeral setae, four setae along lateral margin and two preapical setae. Marginal setae of pronotum present, the anterior ones located before the first third of the length. Ventral pubescence limited to one seta on each half sternite.
Head. Eyes reduced, flat; ommatidia well defined; maximum diameter of about eight ommatidia, temples approximately twice the length of eyes, strongly wrinkled to the neck. Frontal furrows deeply impressed. Antennae moderately long, five antennomeres extend beyond the pronotal base. Antennomere III distinctly longer than antennomeres II and IV, which are similar in length.
Pronotum. Proportions (M-F): WP/LP = 1.3-1.28, WP/WPB = 1.3-1.3, WP/ WH = 1.38-1.3, WE/WP = 1.57-1.53. Transverse, with lateral margins finely bordered; wider in anterior part, narrower than elytra; posterior part much narrower than base of elytra. One seta in the marginal gutter at about a third of pronotum length, another one close to hind angle. Sides evenly rounded and straight just between hind angles and insertions of posterior setae. Hind angles well developed, salient.
Elytra. Proportions (M-F): WE/LE = 0.65-0.69. Oval, broadest almost at midlength; surface moderately convex, flattened on disc. Shoulders distinct but rounded. Striae very finely punctuated, sixth inner striae deeply impressed on disc, but reduced at apex and base; seventh striae shallower, but distinct, the eighth reduced to the posterior quarter of elytra. Apical striola strongly impressed continuing the fifth stria.
Hind wings. Very reduced, not functional. Male genitalia. Median lobe of aedeagus slender, in lateral view (Fig. 15) the basal third curved, the central part straight and the apex showing a curved hook; assymetrical in dorsal view (Fig. 8). Parameres slender, each with 4 to 6 setae at apex. Internal sac of aedeagus with an elongate well-sclerotized piece, forming a symmetrical gut and armed with internal scales (Fig. 16). Characteristic secondary sclerotization of the sperm duct forming a kind of second copulatory piece out of the base of the median lobe ( Fig. 15: CP2).
Size. Mean length (5 exemplars): 5.25 mm (male), 4.56 mm (female). Etymology. The new species is dedicated to Michel Bouillon, Pyrenean speleologist, who was the first to discover the existence of cave beetles in MSS.
Affinities. Trechus bouilloni sp. n. is a representative of the T. brucki group sensu novo as defined in the present paper. It shares with T. grenieri, T. uhagoni, T. beusti, and T. pieltaini the same kind of aedeagus morphology, especially the apex with a curved hook in lateral view, and an internal sac showing two sclerotized parts, the internal copulatory piece and another triangular piece forming a kind of second copulatory piece (CP2, Figs 17-24), also existing in T. brucki and T. bruckoides sp. n. . Similar secondary sclerotized structures of endophallus are known in some groups of insects including Coleoptera, and described as a "sperm pump" (Beutel and Leschen 2005, Hünefeld and Beutel 2005, Jäch and Delgado 2010. In the T. brucki group, however, the structure is too rudimentary to play the same role in sperm transfer, and its function -if any-remains obscure. Although never observed before, this structure is also present in the others species of the group and is the main synapomorphy of the clade. The lack of this internal sclerotized structure in T. carrilloi and T. sharpi (Figs 30-32) casts doubt on their affinities.
Distribution and ecology. Trechus bouilloni sp. n. is only known from the type locality, the MSS of Lizarraga pass (Navarra, Spain) (Fig. 36). The type locality is a MSS located on a northern slope at the eastern extremity of the Sierra de Andía-Urbasa, close to the Lizarraga pass.
Trechus were collected by means of traps in a zone of scree (altitude: 900 m) extending from east to west at the feet of cliffs of Albian limestone lining the northern slope of the plateau of the Sierra de Andía-Urbasa. This scree slope consists of a mass of fallen rocks resulting from the erosion of calcareous cliffs and constitutes a steeply sloped (45°) MSS, filling one of the numerous gullies of a beech forest covering the entire northern side of the plateau lining the southward depression of the Río Arakil (Sakana valley).
On this unstable ground, beeches are replaced by grassy and mossy vegetation dotted with shrubs. The layer of humus is irregular and very thin and only partly covers the blocks of white, angular, medium-sized limestone, rarely exceeding the size of 1 dm³.
The traps were placed 50 centimeters deep in a "C-type" horizon (sensu Juberthie et al. 1981), constituted mainly by stones of 5 cm³, not sealed by the ground and not welded, leaving numerous spaces between them and forming a layer several meters thick above the compact rock.
Some specimens of Trechus bouilloni sp. n. were parasitized by an undetermined Ascomycete.
Trechus bouilloni sp. n. was not found in caves of the area north of Larraona (cueva de los Cristinos, cuevas de Erbeltz, Txintxoleze, Noriturri, Akuandi, del Queso, Iniriturri, Arleze, Laminatitur), suggesting that it is strictly located in MSS (CB personal observation).   (printed)]. This specimen could be the reference specimen of T. oblongus Schaum, 1862. Reference of the name comes from Schaum (1862: addenda, p. 119): "P. 14 col. 2 Trechus oblongus Schaum;" only the name is mentioned, without any description, number of exemplars studied or locality. It should then be considered as nomen nudum.  indicates that T. oblongus is a synonym of T. brucki with type locality: "Pyrén. occ." We were unable to find the specimen or reference where Jeannel found the type locality.

Taxonomic comments
The study of specimens of T. brucki pecoudi from Orhy and of numerous exemplars of T. brucki, including types of the previously described subspecies of T. brucki, demonstrated that none of the characters quoted either by Colas and Gaudin (1935) or by Queinnec and Ollivier (2011)  Trechus politus and T. planiusculus were described by Fairmaire in the volume of the Annales de la Société Entomologique de France of 1861 published in 1862 (Fairmaire, 1862b). As the name Trechus politus was already used for an American species (today Trechisibus politus Brullé, 1842), Fairmaire (1862a) changed the name of this species to Trechus bruckii. Trechus planiusculus was considered synonymous with T. brucki by , and, moreover, the name planiusculus was preoccupied as it was used by Costa (1858) in a work on Italian fauna. In his works on Trechini,  illustrated the genitalia of a male from the Ossau Valley. Recently, in a revision of the french fauna of Carabidae, Queinnec and Ollivier (2011) suggested that the drawing of  was incorrect, and that the male genitalia of T. brucki has a homogeneous shape throughout the distribution area. By examining the types of Fairmaire, we noticed that the drawing of  does not actually match with T. brucki. However, by studying specimens from Ossau Valley we found that the drawing of Jeannel actually corresponds to another undescribed species, very narrowly located in the area of Pic de Montagnon (Bielle-Pyrénées Atlantiques).
Here we describe this new species as T. bruckoides sp. n.  Elytra. Proportions (M): WE/LE = 0.64. Subrectangular, broadest after the midlength; surface moderately convex, flattened on disc. Shoulders distinct but rounded. Striae almost impunctuate, sixth inner discal striae distinct, but reduced at apex and base, especially in callus area; seventh striae shallower, nearly indistinct, the eighth only distinct close to apex of elytra. Apical striola well impressed continuing the fifth stria.

Trechus bruckoides
Hind wings. Very reduced, not functional. Male genitalia. Median lobe of aedeagus slender, in lateral view (Fig 25) basal third curved, central part straight, parallel and elongated towards apex. Nearly symmetrical in dorsal view (Fig 13). Parameres slender, each with 4 setae at tip. Inner sac of aedeagus armed with scales with an elongate well sclerotized piece, forming a twisted gut (Fig 26). Characteristic secondary sclerotization of the sperm duct forming a kind of second copulatory piece out of the base of the median lobe (Fig 25: CP2).
Female genitalia. Not examined.

Size. Mean length (4 exemplars): 4.78 mm (male).
Etymology. The specific epithet refers to Trechus brucki, species with which the new species was merged.
Affinities. Trechus brucki and T. bruckoides sp. n. are externally very similar but strong differences isolate the two taxa especially in shape of male genitalia (Figs 25,  27). The aedeagus shape of T. bruckoides sp. n. is exactly as indicated in  for T. brucki.
Distribution and ecology. Trechus bruckoides sp. n. is only known from the calcareous plateau of Esturou located at 1860 m, north of Montagnon peak (1973 m) and Mailh Massibé (1973 m), at the northern extremity of the massifs separating Aspe and Ossau valleys (Fig. 37). South of this area (Sesques and Gaziès peaks (2600 m)), it is replaced by Trechus brucki which occurs together with Trechus distinctus. During Pleistocene glacial cycles, this plateau was covered by a névé which shaped an area of sinks of nivo-karstic origin (Auly 2008). After winter, snow remains in these sinkholes (July-August) and allows the preservation of a nivicolous fauna, which is unusual at these medium altitudes. Trechus bruckoides sp. n. lives exclusively in the masses of fallen rocks of sinkholes and follows the withdrawal of the snow. When the snow thaws it likely seeks refuge underground.
This mid altitude nivicolous environment could have led to isolation of populations of the species from southern glaciated areas and glacial tongues of the northern slope of Ossau glacier and led to the differentiation of this population of cryophilic and highly hygrophilic Trechus. Such a hypothesis could also explain the presence of the hypogean Trechini Aphaenops bessoni Cabidoche, 1962, endemic to this karstic plateau (pits of Col d'Aran), and closely related to Aphaenops loubensi Jeannel, 1953, an endemic species of the Pierre Saint Martin massif, western to the Aspe Valley. Some other endemic nivicolous Carabidae with morphologically distinct populations occur in the area, like Carabus (Iniopachus) pyrenaeus Audinet-Serville, 1821 (the population of Sède de Pan was first described as a distinct subspecies, C. pyrenaeus cephalicus Csiki, 1927), Nebria lafresnayei Audinet-Serville, 1821, Pterostichus (Cryobius) amoenus mascarauxianus Pupier, 2008, Pterostichus (Lianoe) nadari mascarauxi Jeannel, 1928 and Pterostichus (Lianoe) dufourii (Dejean 1828). The peculiarities of this fauna suggest that this restricted area is an important center of diversification.
Two species, T. pieltaini and T. beusti, were included by  in the angusticollis group. Trechus carrilloi and T. sharpi are provisionally not included in the group because of the absence of sclerotization of the sperm duct (CP2). Moreover, concerning T. carrilloi, the apical hook is not a synapomorphy, as several other Trechus groups have this kind of hook (i.e. T. aubryi or the Tibetan species T. bastropi Schmidt, 2009 andT. damchungensis Deuve, 1997). As expected by Dupré (1991) for some of the species that he included in a "bonvouloiri group", the following regional species previously considered part of the brucki group are clearly excluded here: T. escalerai, T. navaricus, T. bordei, T. jeannei, T. bonvouloiri. Although not included in our analyses, we also exclude of the group T. ortizi, T. baztanensis and T. enigmaticus, as those species were put into the uhagoni group because of close morphological affinities with T. bordei and T. navaricus (Español 1970, Dupré 1991. Trechus jeannei, which was said to be close to T. bordei (Sciaky 1998), does not belong to the T. brucki clade and is not clearly related with the T. bordei clade. A study including more Iberian species should clarify its phylogenetic affinities. Trechus aubryi is excluded from the T. brucki clade and shares strong affinities with T. distinctus.
Our molecular results as well as genital morphology, in dorsal and lateral view and shape of copulatory piece (Figs 10, 19, 20), suggest that T. uhagoni could be considered a distinct species from T. grenieri. Morphological differences between the two species are the following: -T. grenieri: aedeagus in dorsal view ( Fig. 9) with subparallel sides, round apex with a short triangular tip; in lateral view (Fig. 17) basal third strongly rounded, medi-an lobe slightly angular in the middle; apical hook with a thin tip. The copulatory piece is an asymmetrical gut slightly tapering and filled with a densely scaly area (Fig. 18).
-T. uhagoni: aedeagus in dorsal view (Fig. 10) with the left side narrowed or sinuate from the middle to apex, the left side of apical quarter deeply narrowed, forming a long triangular tip; in lateral view (Fig. 19) only the basal quarter rounded, median lobe without dorsal angle in the middle, short with apical hook with massive tip. The copulatory piece is similar to the one of grenieri but the gut is parallel and shortened in its apical part (Fig. 20).
With 7 subspecies recognized in the last catalogues (Moravec et al 2003, Queinnec andOllivier 2011), T. grenieri currently lives in humid forests (1000 m) from Espinal to Iraty (ssp. ruteri), then from Gave de Pau river (north side of Pic de Montaigu) to the Neste d'Aure valley (ssp. grenieri). From Aure valley to the Salat, it is replaced by the subspecies despaxi, which crosses the Garonne river near Saint Béat (Haute Garonne). An isolated subspecies (prepyrenaeus) was described by Coiffait in the high Arize valley (Andronne and Bosc forests around 1000 m) (Coiffait 1974). Along the axial ridge, T. grenieri lives above 1500 m from high Garonne valley (ssp. bepmalei) to Mont Valier (ssp. aulaensis Aubry, 1981). Study of numerous specimens suggests that Trechus uhagoni and all the specimens of the subspecies ruteri are morphologically close, especially in the shape of the male genitalia, and should be considered as a distinct species. Moreover, we studied specimens from various localities of the Pyrenean range (see Distribution) and established that they share some morphological characters (color pale, brown, pronotum transverse with lateral margin regularly rounded, elytral striae superficial, weakly impressed) that justify keeping the status of ruteri as a subspecies of T. uhagoni. The subspecies ruteri should then be considered as belonging to uhagoni, so that grenieri is restricted to the area between Gave de Pau and Ariège valley, northern slope of Pyrenees. T. uhagoni ruteri n. comb. could be distinguished from T. uhagoni uhagoni by its color, usually paler brown, the pronotum transverse with lateral margin regularly rounded and the elytral striae superficial, weakly impressed. It is restricted to the western Pyrenees.
The study of specimens from the whole range of Trechus grenieri including all the subspecies, most of the types and material from intermediate localities (see distribution) leads us to conclude that the characters used to discriminate the subspecies (size, eyes size, shape of elytra and pronotum) are inconstant and overlapping between populations. The shape of the male genitalia is similar for all the populations between Gave de Pau and Ariège valley, including the one (ssp aulaensis) which was said to be different (Queinnec and Ollivier 2011). We consider then Trechus grenieri as a single species without any valid subspecies: Trechus grenieri grenieri Pandellé, 1867 = Trechus grenieri bepmalei = Trechus grenieri despaxi Jeannel, 1922= Trechus grenieri aulaensis Aubry, 1981= Trechus grenieri prepyrenaeus Coiffait, 1974 T. beusti was described by Schaufuss (1863). The type series is located in the Schaufuss collection in the Museum für Naturkunde, Berlin (M. Jaeger pers. com.). T. pieltaini was described by  from a cave of the Basque country, Cueva de Mairruelegorreta. Bolívar y Pieltain and  suggested that the peculiar morphology of the aedeagus of these two species indicates clear affinities with T. uhagoni and T. grenieri. Surprisingly,  in his Monographie des Trechinae considered that these two species belong to another group of species, the T. angusticollis group. This opinion was followed by subsequent authors (Español 1965, Casale and Laneyrie 1982, Ortuño and Marcos 2003. However, and in accordance with the morphology of the median lobe of the aedeagus, molecular results support Bolívar y Pieltain and  point of view and confirm the close affinities between T. beusti, T. pieltaini and the species of the T. brucki clade. Differences between the two species are weak: the apical part of the aedeagus is longer in T. pieltaini (Fig. 12, 23) than in T. beusti (Fig. 11,  21). The copulatory pieces are almost identical (Fig. 22, 24). Trechus beusti is larger, with elytra more oval and elytral striae less impressed (Fig. 4); T. pieltaini is smaller, narrower, and elongate with subparallel elytra and striae more impressed (Fig. 5).
Trechus brucki and T. bruckoides sp. n. do not have the peculiar hooked apex of the median lobe observed in the other species of the clade, but the apex is nevertheless strongly curved (Figs 25, 27).
The case of two further species remains doubtful: T. carrilloi was included by its descriptor in the uhagoni group especially because of the structure of the aedeagus, with an apex with an apical hook (Fig. 31). However, the secondary sclerotization of the ejaculatory duct is lacking in this species and it is characterized by a homogenous elytral pubescence which is present in other species of the area (Ortuño and Jimenez-Valverde 2011), but lacking in all the species of the T. brucki clade sensu novo. The presence of a hook at the apex of the aedeagus is also known in other Pyreneo-Cantabrian species like T. arribasi Jeanne, 1988, currently included in the T. fulvus group (Toribio 2001, Reboleira et al. 2010 or T. aubryi from Ariège. This character led its descriptor to include T. aubryi in the T. uhagoni group. Queinnec and Ollivier (2011) included the species in the T. angusticollis group. The species appears to be the sister species of T. distinctus (Fig. 34).
Finally, T. sharpi was included in the T. uhagoni group by , but the shape of the median lobe of the aedeagus and the copulatory piece, that shares some similarities with the T. bordei group, together with the lack of the sclerotization of spermiduct present in all the species of the T. brucki group sensu novo, cast doubts on its real phylogenetic affinities.

Biogeography of the T. brucki clade
If we use the standard mitochondrial mutation rate for insects of 2.3% divergence per Myr (0.0115 substitutions ⁄ site ⁄ Myr) (Brower 1994, Papadopoulou et al. 2010, Pons et al. 2010, the isolation between the T. navaricus and T. brucki group seems to have occurred at the end of the Pliocene (Faille et al. 2011a). Pliocene climate was much warmer than the Present (Uriarte 2003): the interval between 3.3 Myr to 3 Myr was called Mid Pliocene warm Period, with an average temperature of about 3°C higher than at present and an annual average precipitation between 400 to 1000 mm higher than present. The transition to the Pleistocene (ca 2.7 Myr) is marked by the onset of marked climatic variability; the radiation of the T. brucki clade occured during the Pleistocene, following the rhythm of alternations of cold, warm/humid and dry periods that led to changes in biome composition (Salzmann et al. 2008).
Strong erosion leading to a deep excavation of Pyrenean valleys associated with climate variations led to the dispersal and diversification of the brucki clade. The main events are (Barrère 1963, Campos 1979, Serrat and Ventura 1993, Calvet 2004): 1. Persistence of the Ebro depression between the Basque-Pyrenean area and the Iberian central plateau. The persistence of the Ebro salty basin from the late Oligocene (25 Ma) until the late Miocene (6 Ma) isolated groups with an Iberian distribution from those with a Pyrenean or Basque-Pyrenean distribution. This flat and shallow lagoon area received the tributaries of the Ebro river, from Reinosa to the Mediterranean Sea.
2. Impact of Quaternary erosion on karst fragmentation. On the northern slope, the folds which have an east-west orientation are narrow and divided by north-south valleys. On the southern slope, orogenesis caused the formation of two folds with an east-west orientation (internal and external "sierras") parallel to the axial chain. Similarly, Quaternary erosion separated these sierras by narrow north-south valleys. Near the Atlantic, these "sierras" meet with Basque folds which have a complex northwest/south-east orientation, divided by narrow north-south valleys, from Bilbao to Alsasua. Between Vitoria and Pamplona, these Basque "sierras" are separated by the Pre-pyrenean middle depression, a broad valley excavated by the Zadorra (westward) and Arakil (eastward) rivers (Fig 35). These rivers flow into the Ebro Basin, separating the northern massifs of Aralar, Urquilla and Gorbea from the southern Sierra of Urbasa-Andía. The hydrographic system was set mainly by significant erosion due to numerous glaciation cycles during the Pleistocene (2.5 Ma).
Our molecular study suggests that the brucki lineage could have originated in the area delimited by the northern sierras of Gorbea and Urquilla and the edge of the sierra de Andía. The sierras de Andía, Urbasa and Entzia form the exact border between the hypogean fauna of the Pyrenees and Iberia. North of this limit occur Trechus bouilloni sp. n., Troglorites breuili Jeannel, 1919 (Carabidae, Pterostichini) -Urbasa-Andía-Entzia, Aralar, Ernio and Pagoeta massifs, between the Deba and Urola rivers )-, Euryspeonomus eloseguii (Español, 1948, Bathysciola rugosa (Sharp, 1873), -Leiodidae, Cholevinae, Leptodirini which also belong to a clade of Basque-Pyrenean distribution )-whereas south of this area (Sierras de la Demanda and Lóquiz, surrounding the Ebro basin) is characterized by a lack of Leptodirini and Troglorites. The only cave Coleoptera is Trechus schaufussi comasi (Basaula cave in Baríndano, south of Urbasa). Trechus schaufussi Putzeys, 1870 is a model of Iberian extensive distribution: it is widespread in the Iberian Peninsula, from Algarve in Portugal to Cantabria, Iberian Central System and the pre-pyrenean massif of Guara in Spain (Jeanne and Zaballos 1986, Zaballos and Jeanne 1994, Serrano 2003. This species is known to have separated early from Trechus sensu stricto (Faille et al. 2010a(Faille et al. , 2011a. Trechus bouilloni sp. n. has a subterranean lifestyle among the scree-covered northern slope (900 m) of Sierra de Andía, whereas the type locality of T. uhagoni is the Orobe doline (700 m), located at the eastern limit of Sierra de Urquilla. Early Pleistocene climate variations could have led to drastic changes in biome composition, limiting dispersal possibilities and leading to the isolation of the population of T. bouilloni sp. n. (potentially forestal), south of the Arakil River. One hypothesis could be that the hygrophilous species were colonizing high altitude or hypogean habitats during interglaciar warming as observed in other species of Coleoptera (Hernando et al. 1999, Faille et al. 2011b). These climate fluctuations might also have led to the western subterranean colonization of the two hypogean species, T. beusti (Sierra de Urquilla) and T. pieltaini (Sierra de Gorbea) while the group colonized the Pyrenean chain and diversified in numerous forms living in humid forests and alpine zones, from the Iraty Valley to the Ariège Basin. Migration toward East could have been possible along the small sierras of Tajonar and Labia, which link the Basque Mountains to the Pyrenees.
Trechus brucki lives in the alpine zone (above 1700 m) of the axial ridge from Pic d´Orhy to Col du Pourtalet, in the high Ossau Valley. On the north ridge, T. brucki can be encountered in the same biotopes, near snow tongues melting on scree-covered slopes, from Aspe to Gave de Pau Valleys. As for T. grenieri and T. uhagoni, both are mainly forestal and occur at lower altitude except for T. grenieri in the eastern part of the range (Mount Valier area). The Ariège Valley is the eastern limit of the group.
The distribution area of the T. brucki group coincides with the one of the Basque-Pyrenean Leptodirini clade (Fig 35). In the Pyrenees, both groups are made up of forestal, endogean, humicolous, lapidicolous or orophilous, but not hypogean, species. It is only in Basque relief, the western part of their distribution, that both groups include subterranean species. Regarding Leptodirini, the basal group of the Basque-Pyrenean clade is the Bathysciola schiodtei group (endogean/humicolous elements); its distribution area is extended from Ariège, B.  Jeannel, 1919 (between the Urola and Orio Rivers). In the Pyrenees, the species of the Trechus brucki clade are epigean, forestal (T. grenieri, T. uhagoni ruteri) or orophilous (T. brucki, T. bruckoides sp. n.). Pyrenean speciation events in the group are more recent and are probably closely related to late Pleistocene climatic changes, as already observed in alpine Trechus (Lohse et al. 2011). Troglobiomorphic features (depigmentation, microphthalmy) only occur in the hypogean T. beusti and T. pieltaini, both of them located in the Basque area. The two other species of this geographical area are located in wet and cold dolines (T. uhagoni) or subterranean environments (T. bouilloni sp. n.). Their general appearance (pigmented, well-developed eyes), similar to other epigean species, could be an indication of the recent colonization of this reduced habitat.