New Palaearctic species of the tribe Thalassaphorurini Pomorski, 1998 (Collembola, Onychiuridae)

Abstract The paper is devoted to a taxonomic revision of the genus Sensillonychiurus Pomorski & Sveenkova, 2006. Five new species of this genus, i.e. Sensillonychiurus mirus sp. n., Sensillonychiurus taimyrensis sp. n., Sensillonychiurus vegae sp. n., Sensillonychiurus vitimicus sp. n., and Sensillonychiurus amuricus sp. n., as well as three new species of the related genus Allonychiurus Yoshii, 1995, i.e. Allonychiurus subvolinensis sp. n., Allonychiurus elikonius sp. n., and Allonychiurus unisetosus sp. n. are being described from various regions of Eurasia. The diagnoses of both genera are amended to include described species. Two genera, Tantulonychiurus Pomorski, 1996 and Thibaudichiurus Weiner, 1996, are treated as junior synonyms of the genus Allonychiurus. Agraphorura eisi (Rusek, 1976) is transferred to Sensillonychiurus; Tantulonychiurus volinensis (Szeptycki, 1964) and Tantulonychiurus asiaticus Babenko, 2007 to Allonychiurus. A review of morphological peculiarities of Sensillonychiurus is performed, comparisons with the other genera of Thalassaphorurini given, and a key to the known species provided.


Introduction
This paper has been prompted through the discovery of a new species on the Barents coast of Kola Peninsula. This species from the tribe Thalassaphorurini is characterized by the combination of morphological features that fails to completely fit into any of the known genera of the tribe. Unfortunately, the tribe's generic classification, as well as that of the whole subfamily Onychiurinae, is still far from perfect. Starting from the pioneering papers by Bagnall (1948Bagnall ( , 1949 and until recently, about 15 genera have been proposed for various representatives of the tribe. Most of these genera are entirely valid, yet some are considered junior synonyms. For instance, at least four synonyms are known for the type genus of the tribe, Thalassaphorura Bagnall, 1949 alone (see Sun et al. 2010). At present, according to the database of Collembola of the World (Bellinger et al. 1996(Bellinger et al. -2011 the tribe includes nine widely accepted genera: Thalassaphorura, Micronychiurus Bagnall, 1949, Uralaphorura Martynova, 1978, Allonychiurus Yoshii, 1995, Spinonychiurus Weiner, 1996, Tantulonychiurus Pomorski, 1996, Agraphorura Pomorski, 1998, Detriturus Pomorski, 1998, and Sensillonychiurus Pomorski et Sveenkova, 2006 In addition, there are several generic names that are only used occasionally. Thus, the provisional synonymy of Thibaudichiurus Weiner, 1996 with Allonychiurus was recently rejected by Sun et al. (2011).
The only character uniting all of the members of the tribe Thalassaphorurini is the structure of the furcal remnant which forms a finely granulated area in mid-section of Abd.4 with 4 small setae arranged in two posterior rows. The second character shared, i.e. distinct antennal and tergal sensilla, is probably present in all genera, but not all species of the tribe. Taking this into account, evidently the genus Uralaphorura is to be excluded from the tribe in having nothing in common with the other Thalassaphorurini, being characterized by a quite different structure of the furcal remnant with four posterior setae arranged in a line. Thus, Uralaphorura is probably closer to Onychiurini than to Thalassaphorurini (see also Babenko 2009).
According to R.J. Pomorski (personal communication), two of the remaining eight genera, i.e. Micronychiurus and Agraphorura are to be considered as synonyms. Nevertheless we do not follow here this suggestion as it was never officially published and a discussion on the status of these genera is beyond the scope of our paper. In his draft Synopsis on Palaearctic Onychiuridae, Pomorski also intended to synonymize the genus Tantulonychiurus (and also Thibaudichiurus) with Allonychiurus. Most probably, this latter suggestion was dictated by its practical usefulness, as well as by the impossibility to unity the known species of all these "genera" into more or less natural groups, based only on our present knowledge. For instance, according to Sun et al. (2009) only five of about two dozen known species of Allonychiurus completely correspond to the diagnoses as given by Yoshii (1995) and Weiner (1996). Later the generic diagnosis was amended by Sun et al. (2011) to include some species showing a partly reduced tibiotarsal chaetotaxy and smooth sensillar clubs in AO. In this scope, Allonychiurus differs from both Thibaudichiurus and Tantulonychiurus only in having several rows of manubrial setae posterior to a furcal remnant, and more numerous labral setae. Two latter genera were separated by the position of MVO and a different number of distal setae on the tibiotarsi, namely seven setae in two so far known species of Tantulonychiurus whereas the type species of the genus Thibaudichiurus has not eleven (Sun et al. 2011) but nine setae in distal whorl (personal communication of L. Deharveng). Such a generic classification of the complex partly simplified the situation, but did not completely solve the main problem existing in the group, i.e., the existence of many inadequately described species which can only provisionally be placed in this or that modern genus. Thus, six of 14 species treated as representatives of the genus Allonychiurus in the last paper of Sun et al. (2011) were marked by an asterisk indicating that "species assignment requires confirmation". Furthermore, it is rather difficult to apply this division to some known species as well, since some are purely parthenogenetic or just lack modified setae in reproductive males, e.g. Tantulonychiurus asiaticus Babenko, 2007, which is in other respects virtually identical to the members of the Thibaudichiurus/ Tantulonychiurus complex. The degree of reduction of the tibiotarsal setae appears to be a rather variable character in some genera of Thalassaphorurini, still being unknown for many described forms. As a revision of all these related genera is beyond the scope of the present paper, we tentatively accept here a broadened conception of Allonychiurus (see diagnosis in the end of the paper) and use the following generic classification of Thalassaphorurini as consisting of seven genera: Spinonychiurus (d 0 on head absent, PAO lobes compound, sternum of Abd.3 clearly subdivided, tibiotarsi with 11 setae in distal whorl), Detriturus (d 0 absent, PAO lobes compound, sternum of Abd.3 not subdivided, Abd.5-6 fused, tibiotarsi with 11 setae in distal whorl), Sensillonychiurus (d 0 absent, PAO lobes compound, sternum of Abd.3 not subdivided, Abd.5-6 clearly separated, tibiotarsi with 7 or 9 distal setae), Thalassaphorura (d 0 present, PAO lobes simple), Micronychiurus (d 0 present, PAO lobes compound, Abd.6 with 1+1 prespinal microsetae, multiplication and unusual position of anterior pso on head and on abdominal tip, AS present or absent), Agraphorura (d 0 present, PAO lobes compound, Abd.6 with 1+1 prespinal microsetae, low number of dorsal pso in usual position, AS absent) and Allonychiurus (d 0 present, PAO lobes compound, Abd.6 with 2+2 prespinal setae, AS present). The latter genus includes two rather distinct species-groups, the flavescens-group and the volinensis-group, which clearly differ habitually, as well as in the number of labral setae and setal rows in the manubrial zone of Abd.4, but both latter characters may be size-dependent. At the same time, we are sure that this generic system requires additional attention, while the scopes of some genera might be cardinally changed in future.
The new species mentioned in the beginning of Introduction appears to be especially similar to the known representatives of the small eastern Asiatic genus Sensillonychiurus. A study of the available material from M. Potapov's and authors' collections reveals a whole number of closely related forms and shows that the original diagnosis of the genus must be somewhat amended. Thus, the present paper includes a brief review of the morphological peculiarities of Sensillonychiurus as compared to the other genera of the tribe, a slightly changed diagnosis and a key to all of the known species of this genus, as well as descriptions of five new species. In addition, three further new species habitually similar but, according to the accepted system of Thalassaphorurini, assignable to the genus Allonychiurus, have also been described and used for comparative purposes. Types of all the new species are deposited in the collection of the Department of Zoology & Ecology, Moscow State Pedagogical University (MSPU).

A review of the main morphological characters of Sensillonychiurus Pomorski et Sveenkova, 2006
The present review is based on the morphological peculiarities of five new species described in this paper, as well as on published data on all four so far known species of the genus. Three of them were described by the authors of the genus (Pomorski and Sveenkova 2006), the fourth one, S. eisi (Rusek, 1976), comb. n., has hitherto been treated as a species of Agraphorura (see Pomorski 2004;Arbea 2005). Nevertheless, there is good evidence of its similarity to Sensillonychiurus, for instance, in the absence of d 0 on the head and in the presence of only three guard setae in AO (see figs 11A and 12D in Rusek 1976).
Body shape and size. All of the so far known species of Sensillonychiurus are among the smallest Onychiurinae, with body size ranging between 0.4 and 0.7 mm. The body is slender and elongated (Figs 8-9), with rather short antennae and clearly club-shaped Ant.4 (Fig. 10). Area antennalis is not distinctly demarcated.
Sensillar armature of the antennae. Pomorski and Sveenkova (2006) considered the presence of only three guard setae in AO as the main diagnostic feature of the genus Sensillonychiurus. These authors, based on an examination of all three species then known, found this character as being unique not only to Thalassaphorurini, but to all other Onychiurinae as well. They wrote that it "may indicate that the new taxon is monophyletic". However, our study reveals that not all of those species, albeit indeed strikingly similar, are characterized by such a deep reduction of the number of guard setae in AO. Thus, a far more usual number (4) of guard setae was found in the European S. mirus sp. n., as well as in two eastern Asian species, S. vitimicus sp. n. and S. amuricus sp. n., thus correlating with a full set (5) of papillae. Only one congener, S. geminus Pomorski et Sveenkova, 2006, has AO with five papillae but three guard setae. That is why not only the diagnosis of the genus has to be slightly amended, but its distinctions from the other genera of the tribe must be reconfirmed, although most of Thalassaphorurini are characterized by a complete set (5) of guard setae in AO. Apart from Sensillonychiurus, species with less numerous (4) guard setae are known only among Agraphorura. Discarding this character, the sensillar chaetotaxy of the antennae in Sensillonychiurus is not genus-specific, being more or less similar to that in the other genera of Thalassaphorurini: Ant.4 always bears two distinct thickened sensilla (a dorsal one subapically and an inner one in the mid-section of the segment), a small subapical organite (or) and a subbasal microsensillum (ms) which is clearly larger than that on Ant.3 (see, for instance, Figs 1, 10). The latter character is also typical of Micronychiurus and Agraphorura. The position of ms on Ant.4 in relation to ordinary setae slightly varies between different species  and can be used in their identification. Sensorial elements in AO of different species of the genus are similar: clubs are smooth, more or less roundish, with or without clear ribs. A different type of sensorial clubs in AO (distinctly granulated, morula-like) is known among Thalassaphorurini only in some Thalassaphorura and in the flavescens-group of Allonychiurus.
Structure of the PAO. All species of Sensillonychiurus show a relatively wide PAO consisting of few (6-8) vesicles with numerous secondary lobes. As a whole, it usually looks like a single mass with only traces of vesicle divisions (Fig. 3).
Labrum. All congeners are characterized by a constant number (7) of labral setae, four distal ones being longer and clearly thicker, and two or four prelabral setae. The variant with two prelabral setae seems to be more common (see Table 1), but this character is still unknown in S. eisi, S. virginis Pomorski et Sveenkova, 2006 and S. geminus. Such a slightly reduced number of labral setae is also typical of all Thalassaphorura known for this character, as well as of the volinensis-group of Allonychiurus, but not of the flavescens-group, at least some of which showing nine labral setae (Yoshii 1995;Sun et al. 2009Sun et al. , 2011. This feature is completely unknown in Detriturus, Spinonychiurus, and Micronychiurus, whereas among Agraphorura the existing information concerns only A. calvoi Arbea, 2005, which has nine labral setae (a presumed basal set for Onychiurinae), and A. sangelensis Kaprus' et Stebaeva, 2006, with two prelabral and seven labral setae (our data).
Labium. The type of labium most frequently seen in the genus is AC, with the ABC-type is found only in two species, S. mirus sp. n. and S. vitimicus sp. n. The number of setae on the proximal, basal and laterobasal fields of the labium is more or less stable, although individual variations and some asymmetry are visible in some specimens. The number of distal guard setae of the labial palp corresponds to the most common (and also complete) set found in Onychiurinae (Fjellberg 1999): seven long guard setae (b 3-4 , d 3-4 , and e 1-3 ) and four shorter (a 1 , b 1-2 and d 2 ) ones set on papillae. The only notable peculiarity of the labium in the study group is the unusual length of a 1 seta which is clearly longer and thicker than b 1-2 or d 2 (Fig. 4). Unfortunately, the fine structure of the labium is known only for a few representatives of the tribe, this not allowing for serious comparisons to be made. We can only state that all three types of labium (А, АС and АВС) are known in Thalassaphorura (Sun et al. 2010), with АС being the most common. In the genus Allonychiurus, two types (AC and ABC) are found among species of the volinensis-group (Fjellberg 1999, our data), while only the AC-type is known in two species of the flavescens-group (see Sun et al. 2009Sun et al. , 2011. The A-type is observed in Spinonychiurus epaphius Kaprus' et Tsalan, 2009 and, according to Pomorski andSveenkova (2006), in the genus Detriturus. The AB-type seems to be most characteristic of the genera Agraphorura (Pomorski 2004;Arbea 2005;Kaprus' and Stebaeva 2006) and Micronychiurus (Pomorski, pers. communication). The presence of a complete number of distal guard setae on the labial palp in such small-sized species as Sensillonychiurus is rather unexpected, as, for instance, all of the so far studied Thalassaphorura and members of the volinensis-group of Allonychiurus, being usually larger, have only ten guards (e 2 absent) (Fig. 41). The same is probably characteristic of the flavescens-group of Allonychiurus (Sun et al. 2009(Sun et al. , 2011 although the authors believe that not e 2 but one of the b-setae is absent. A relatively long a 1 -seta could be suggested as a possible apomorphy of the genus, but there is not enough information concerning the other groups of Thalassaphorurini for such an assertion. Dorsal and ventral pso. Contrary to the majority of Onychiurinae, the number of dorsal and ventral pso does not significantly vary within the genus, being almost always as following: 32/133/33343 (dorsal) and 1/000/0000 (ventral). There are only two exceptions: S. virginis, with a lesser number of pso on thoracic terga (32/022/33343 as a whole), and S. geminus, with some pso on two abdominal sterna. The ventral pseudocellar formula of the latter species was given differently by Pomorski and Sveenkova (2006) in the original description (1/000/0101) and in their comparative table of diagnostic characters (1/000/10010). The former version is probably the correct one. Apart from this, S. mirus sp. n. often lacks the anteriormost pso of the postantennal group on a head. Such a dorsal formula (32/133/33343) is rather common in two other gen-era of Thalassaphorurini, namely, Agraphorura and Allonychiurus, known also in some Thalassaphorura, as well as in different genera of Onychiurini and Oligaphorurini. The absence of pso on abdominal sterna as the most usual character of Sensillonychiurus can also be found among Spinonychiurus, Allonychiurus and Detriturus.
Parapseudocelli. The complete absence of parapseudocelli (psx) on the subcoxae, femora and abdominal sterna is characteristic of most of the studied species of the genus, except for S. vegae sp. n. which sometimes possesses a pair of psx on Abd.4. Such a weak development of psx is rather frequent among Thalassaphorurini, also known in Micronychiurus, Agraphorura, Allonychiurus (in both flavescens-and volinensis-groups), and some Thalassaphorura. Probably it at least partly correlates with the small size of specimens. Some intraspecific variations of psx numbers are likely (see, for instance, description of S. vegae sp.n.) and need further attention.
Dorsal chaetotaxy. The chaetotaxy in the genus was originally described as follows: "Seta d 0 on the head absent. Abdominal terga of IV, V and VI with 2, 1 and 1 medial setae, respectively". It can be added that these unpaired setae (m 0 and p 0 on Abd.4, p 0 on Abd.5 and a 0 on Abd.6) are meso-or macrosetae probably belonging to the primary chaetotic set, but not microsetae which can appear during ontogeny. Terga of Th.2-3 in adults with 3+3, of Abd.1-4 with 2+2 and of Abd.5 with 1+1, axial microsetae, additionally each tergum with 2+2 mesosetae in the axial group set out of line with microsetae (see, for instance, Fig. 8). The same pattern is found in all studied species which appear to have an almost symmetrical (especially in the mid-section of terga) and virtually identical dorsal chaetotaxy. This pattern seems to be unique to Thalassaphorurini. Thus, Sensillonychiurus shares the absence d 0 with only two genera of the tribe, Spinonychiurus and Detriturus. Known representatives of both these genera show different distributions of unpaired setae on the abdominal tip (Arbea and Jordana 1985;Palacios-Vargas and Diaz 1995;Pomorski 1998;Kaprus' and Tsalan 2009), the most similar but yet not identical is that in D. jubilarius (Gisin, 1957) (see fig. 97G in Fjellberg 1998). In the group with d 0 on the head, species of Micronychiurus and Agraphorura with known chaetotic patterns possess a medial seta only on Abd.6 (Palacios-Vargas and Deharveng 1982;Beruete et al. 1994;Pomorski 2004;Arbea 2005;Kaprus' and Stebaeva 2006), Allonychiurus has quite a different chaetotaxy of Abd.6 with one or two medial setae and 2+2 prespinal microsetae (Figs 40,49), unpaired setae on Abd.4 and 5 are microsetae if present (Lee 1973;Weiner 1989;Sun et al. 2009Sun et al. , 2011. A similar pattern is typical of most Thalassaphorura. Tergal and sternal sensilla. The lateral microsensillum in all studied species is always present on Th.2, but usually absent from Th.3, except for two species, S. minusculus Pomorski et Sveenkova, 2006 and S. geminus. Several thickened macrosensilla in certain parts on terga and sterna are also very typical of Thalassaphorurini and of Sensillonychiurus as well. The most usual number of such sensilla in the studied species is as follows, 1/011/222111 from head to Abd.6 ( Fig. 8), additionally two ventral sensilla are usually distinguishable on the anterolateral part of the head and one sensillum on each ventrolateral side of Abd.4 (Fig. 33). Variations are not frequent and somewhat obscure; the only clear exception being the European S. mirus sp. n. which shows more dorsal sensilla (2/022/222221 as a whole). The described variability of the character in various genera of Thalassaphorurini permits to suggest that it can hardly be used in separating the genera. Moreover, the degree of sensillum differentiation varies widely both between and within species, being clearly age-dependent; sometimes the sensilla look like slightly thickened macrosetae distinguished only due to their positions. Some level of population variability of the character is not improbable either.
Ventral chaetotaxy. Most of the species of the genus lack setae on thoracic sterna. The only exception is S. vitimicus sp. n., with 0-1-1 setae on each side of the linea ventralis on the thorax (Fig. 33). Among Thalassaphorurini, the complete absence of ventral setae on the thorax is only observed in some species of the genus Micronychiurus (Pomorski, pers. communication) and Agraphorura (Pomorski 2004;Arbea 2005). All studied species also show no setae at the base of VT and a rather stable number of setae on its distal lobes (usually 6+6). These latter characters are not unusual in Thalassaphorurini, known, e.g., in some Micronychiurus, Agraphorura and Allonychiurus.
Tibiotarsal chaetotaxy The pattern characteristic of all studied species of the genus can be described as follows: seven or nine setae in the distal whorl (all or two T-setae absent), 7-7-6 setae in B-whorl, Y-seta present, but M-seta absent (Figs 20,(29)(30). The same pattern with 9 distal setae was previously found in S. eisi by Fjellberg (1991). It is noteworthy that the latter character (absence of M-seta) only rarely occurs in Poduromorpha. Nevertheless, the same is probably characteristic of some Agraphorura (Palacios-Vargas and Deharveng 1982; Arbea 2005) but the number of tibiotarsal setae in the latter genus is rather variable, with both distal and proximal whorls being partly reduced. For instance, in A. sangelensis Ti.1-3 bare only 13-13-13 setae, respectively (seven in distal whorl, five B-setae and a slightly longer Y-seta set virtually in B-whorl, M-seta absent). Species of the genus Micronychiurus are known as having 7 or 9 distal setae and 8-8-7 setae in proximal whorls (Beruete et al. 1994), and so probably possess M-seta and lack one of the B-seta on Ti.3. All of the studied species of Thalassaphorura, as well as all Allonychiurus from the volinensis-group (also showing 7-9 distal setae on tibiotarsi), are characterized by a complete B-whorl on all legs (7-7-7) and the presence of both setae M and Y (Figs 47-48). The same pattern but with few additional proximal setae in C-whorls was known for Allonychiurus antennalis Sun, Chen et Deharveng, 2011 from the flavescens-group but the data for A. megasomus Sun, Yan et Chen, 2009 is different, with 11 distal setae, 8-7-7 setae in B-whorls, and 2-2-1 additional setae involved. All other genera of the tribe feature a complete set of distal setae; in addition, at least some of them, for instance Detriturus jubilarius, has M-seta (see fig.  389 in Pomorski 1998). These differences are evidently a good reason to complete the descriptions of tibiotarsal chaetotaxy in such oligochaetotic forms of Onychiurinae.
Subdivision of sterna. Among Thalassaphorurini there is a genus, Spinonychiurus, characterized by such a unique feature as a secondary division of Abd.3 sternum. Some traces of such division can also be seen in all well preserved specimens of Sensillonychiurus ( Fig. 6), as well as in some other small-sized species of various group of Onychiurinae. Nevertheless, the anterior subsegment in Sensillonychiurus is narrow and, contrary to Spinonychiurus, lacks setae.

Furcal remnant position.
In complete agreement with the main diagnostic character of Thalassaphorurini, the furcal remnant in all studied Sensillonychiurus is in the form of a finely granulated area in the mid-section of Abd.4, with four small setae arranged in two posterior rows. Individual variations in number and position of these setae are not frequent, but have been noted. The number of setal rows on manubrial area is also more or less stable: usually two rows (mm and mp according to Weiner (1996) with 4 setae in each can be distinguished (Fig. 33) although some variations especially in their position are also seen. Additionally 1+1 setae (ma?) usually present at a level with finely granulated area. The most significant is the anterior position of the latter area at contact with the border between Abd.3-4 sterna (Figs 6, 33). According to the available, mainly illustrative data (Fjellberg 1998 Sun et al. 2010), but possess more manubrial setae arranged in several rows; sometimes a few additional setae are present (Weiner 1996;Sun et al. 2011). This difference was used as a main diagnostic feature in separation of Allonychiurus from Tantulonychiurus and Thibaudichiurus by Sun et al. (2011). However, it can also be considered as a result of polychaetosis clearly seen on fig. 1B in Sun et al. (2011). More investigation including a study of juveniles is probably needed to evaluate the significance of these differences.
Anal spines. A full spectrum from complete absence to strong spines set on low papillae is found among the studied species, but an intermediary situation is most frequent. The same is characteristic of Spinonychiurus and Micronychiurus, but not of Detriturus and Agraphorura (complete absence of spines), Thalassaphorura (AS absent as an exception) and Allonychiurus (spines always present).
Based on this review of the morphological features, the following can be concluded: Regardless of one's opinion on the status of the genus Sensillonychiurus, all studied species represent a rather homogeneous group of closely related forms, characterized by many common morphological features and seemingly congruent distributions mainly covering the northern parts of eastern Asia with insulated records from North America and Eastern Europe.
The genus Sensillonychiurus shares many characters with representatives of other genera of Thalassaphorurini, but a combination of characters seems to be unique for the tribe. The only features, which set the genus apart from all other Thalassaphorurini, appear to be not the number of guard setae in AO but dorsal chaetotaxy and anterior position of furcal remnant at a contact with border between Abd.3 and 4 although the data concerning other genera is still rather limited for a final decision.

Description of species
Affinities. S. mirus sp.n. clearly differs from the all previously described species of the genus first of all in having not three but four guard setae in AO. Nevertheless it is not a unique character for the group. The same structure of AO (5 papillae and 4 guards) as in S. mirus sp. n. is known in two other species of the genus, S. vitimicus sp. n. and S. amuricus sp. n. (see descriptions below). All these species which are characterized by only a weak reduction of AO with a highest possible number of papillae and 4 guard setae have many other characteristics in common (see Table 1.). Nonetheless, S. mirus sp. n. can easily be distinguished from S. vitimicus sp. n. by the complete absence of setae on thoracic sterna, from S. amuricus sp. n. in the different type of labium (ABC in S. mirus sp. n. versus AC in S. amuricus sp. n.), and in four prelabral setae (S. amuricus sp. n. possesses only two prelabral setae which are more common in the genus).
Etymology. Initially, the name mirus (odd, strange, unusual in Latin) reflects both an isolated position of the new species within the genus and the gap between its typelocality and the distributions of the other known species of the genus which are pure Asiatic or American. The level of morphological uncommonness has lowered after the performed survey of all available material, but the geographical isolation still exists.
Affinities. Apart from S. taimyrensis sp. n., only two known species of the genus, i.e. S. minusculus and S. vegae sp. n., completely lack all T-setae on tibiotarsi (distal whorl with 7 setae). S. minusculus clearly differs in having lateral ms on Th.3 and Abd.6 without AS. Two other species, S. vegae sp. n. and S. taimyrensis sp. n. are very similar, sharing many common characteristics (see Table 1). Nonetheless S. taimyrensis sp. n. can be easily distinguished due to stronger AS set on low papillae (cf. Figs 18 and 19), more distal position of ms on Ant.4 (cf. Figs 10-14 and 15-17) and clear differences in the mutual position of microsetae p 1 and mesosetae p 2 on Abd.3 (cf. Figs 21 and 22).
Variability. The types of S. vegae sp.n. completely lack psx as well as all so far studied species of the genus. Nonetheless, at least some of the specimens collected on Vitim Plateau possess 1+1 ventral parapseudocelli on Abd.4 (Fig. 23) being otherwise identical to the types. This population may represent a separate species, but its reliable distinction is hardly possible. Anyway, more material from different points of the distributional range is needed to evaluate the constancy and significance of this character.
Affinities. Virtually all of the main morphological characteristics of S. vegae sp. n. (structure of AO and PAO, labrum and labium, dorsal and ventral chaetotaxy, number and distribution of pso, presence of ms only on Th.2, number of setae on subcoxae, tibiotarsi and VT) are identical to those of sympatric S. taimyrensis sp. n. Concerning the differences of S. vegae sp. n. from S. taimyrensis sp. n. see description of the latter.
Etymology. The new species was initially collected during the joint Swedish-Russian expedition arranged in 1994 in order to commemorate A.E. Nordenskiöld's first trip on "Vega" board along the Northern Sea Route (1878-1879). That is why it is named after Nordenskiöld's famous steamship "Vega".
Affinities. The same structure of AO (five papillae and four guard setae) as in S. amuricus sp. n. is only known in two species of the genus, S. mirus sp. n. and S. vitimicus sp. n. All these species which are characterized by only a weak reduction of AO with a full number of papillae and 4 guard setae also show the highest number of setae (9) in the distal tibiotarsal whorl. Both can easily be distinguished from S. amuricus sp. n. in having a different type of the labium (ABC versus AC in S. amuricus sp. n.) and four prelabral setae (S. amuricus sp. n. possesses only two prelabral setae, which occurs more commonly in the genus). Apart from this, S. amuricus sp. n. is the largest congener.
Two other species of the genus, S. virginis and S. geminus, are characterized by the most complete set of tibiotarsal setae (17-17-16) but against the background of a pronounced reduction of AO.

Sensillonychiurus vitimicus
Affinities. Due to the presence of four guard setae in AO, S. vitimicus sp. n. is the most similar to S. mirus sp. n. and S. amuricus sp. n. All these three species have many other characteristics in common (see Table 1), but S. vitimicus sp. n. can easily be distinguished by the presence of setae on thoracic sterna (a presumed apomorphic condition within Onychiuridae according to Fjellberg (1998).
Etymology. The new species was named after its terra typica. Distribution. Known from several biotopes in vicinity of the type locality.
One more species of the genus Sensillonychiurus was found on Kamchatka (vicinity of Petropavlovsk, sandy sea beach with weed debris, leg. L. Lobkova). It differs from S. virginis in having setiform anal spines, from S. geminus by the absence of lateral ms on Th.3. The lack of material (only a single female is available) did not allow us to describe it, but it is listed in the key and in Table 1  AO with 4 papillae and 3 guard setae (as in Fig. 10 AS strong, set on low papillae (Fig. 18), ms on Ant.4 clearly above proximal setae (Figs 10-14), microsetae p 1 set anteriorly to mesosetae p 2 on all terga from Abd.1 to Abd.3 (Fig. 21) Pomorski, 1996, syn.n. Syn. Thibaudichiurus Weiner, 1996 Type-species. Onychiurus flavescens Kinoshita, 1916: 458, by original designation. Diagnosis. Small-or medium-sized Thalassaphorurini with compound vesicles in PAO; labrum with 7 or 9 setae, labium of AC or ABC-type; AO with 4-5 papillae and 5 guard setae, smooth or granulated sensory clubs; antennal and tergal sensilla usually distinct, d 0 on head present, Abd.4 and 5 usually with some axial microsetae, Abd.6 dorsally with 2+2 prespinal microsetae and 1-2 medial macrosetae; distal whorl on Ti.1-3 with 7, 9 or 11 setae, B-whorl usually complete on all tibiotarsi, M seta present; no tendency to dorsal pso multiplication, head and abdominal sterna with ventral pso, dorsal pso on Th.1 usually present; psx not numerous or absent; sternum of Abd.3 not subdivided, furcal remnant situated at some distance from border between Abd.3-4 sterna, with one or several rows of manubrial setae posterior to dental setae; MVO present or absent; AS present.
Remarks. As it was already stressed in Introduction the genus is accepted here in a wider scope than it was proposed by Sun et al. (2011) to include species described below. In this scope the genera Tantulonychiurus Pomorski, 1996 andThibaudichiurus Weiner, 1996 recognized as valid by Sun et al (2011) are placed here in synonymy of Allonychiurus. In our opinion, the generic value of main differentiated character of these genera, i.e. the number of setal rows on manubrial area, appears to be size and age dependent and needs further attention to be proved. The genus in the accepted scope is rather heterogeneous but completely analogous to Thalassaphorura which mainly differs in having simple vesicles in PAO. Here we only deal with the representatives of so called volinensis-group of the genus characterized by small size (less than 1.0 mm), smooth sensorial clubs and usually four papillae in AO. According to the generic classification proposed by Sun et al. (2011), the species described below should probably be assigned to the genus Tantulonychiurus since all of them are characterized by only one row of manubrial setae posterior to dental microsetae and by the position of MVO on Abd.4 sternum if present. In this case, the degrees of reduction of the tibiotarsal setae found in the new species completely fill up the gap between Tantulonychiurus and Thibaudichiurus (7-9 versus 9 setae, respectively) and make their distinction rather problematic, taking also into account that not all of these species possess a MVO. That is why we are inclined to leave a decision concerning the status of all these genera pending a complete revision of the complex. Paratypes: 22 females on slides and more than 300 specimens in alcohol, same data as holotype; 11 females, same region, 1,480 m alt., plant community with predominance of Dryas sp. on slope, 07.vii.2002; 7 females, same region, 1,430 m alt., herbaceous meadow on south-facing slope, 07.vii.2002; 14 females on slides and more than 800 specimens in alcohol, same region, greenhouse of "Vostochnaya" Meteorological Station, 1,287 m alt., 24.vii.2002, leg. O. Makarova (MSPU).
Affinities. The main morphological features of A. elikonius sp. n. are similar to those of A. volinensis, A. subvolinensis sp. n. and A. asiaticus (Babenko, 2007), comb. n. (see Table 2). Thus, all four species are characterized by virtually identical dorsal chaetotaxy and similar numbers of pso on all terga, sterna and subcoxae. The presence of a complete set of B-setae and M-seta on all tibiotarsi is also shared by them. A. elikonius sp. n. has a different type of the labium (AC in A. elikonius sp. n. versus ABC in three other species) and differs from A. volinensis and A. subvolinensis in the mutual position of antennal pso (cf . Figs 40 and 45). There are also some variations of the number of distal setae on the tibiotarsi in these four species (7 setae in A. volinensis and A. asiaticus, 9 in A. elikonius and A. subvolinensis). A. asiaticus is the only species in the group showing five papillae in AO (found in elikonius only in exceptional cases), and only A. subvolinensis is characterized by the presence of setae on all thoracic sterna (absent from Th.1 in all other species).
It is more difficult to distinguish A. elikonius sp. n. from three Korean and one Chinese species of the group, namely A. mariangeae (Thibaud et Lee, 1994), A. donjiensis (Lee et Kim, 1994), A. jindoensis (Lee et Kim, 1994), and A. foliatus (Rusek, 1967), because their descriptions are incomplete and probably not fully correct in certain details. The most complete description is that of A. mariangeae. It is rather similar to A. elikonius sp. n. in having an almost identical chaetotaxy, the same number of dorsal pso and tibiotarsal setae (see Table 2). The only difference of the sternal pso formula is the presence of true pseudocellus on Abd.1 in A. mariangeae instead of an elongated parapseudocellus without clear cuticular ring in A. elikonius sp. n. However, these organs are homologous and sometimes difficult to distinguish. The most characteristic feature of A. mariangeae is the presence of MVO in mature males. Unfortunately, A. elikonius sp. n. in the region under study is only represented by parthenogenetic populations: among more than 100 specimens checked, all were females. Formally, these species differ in size (0.75-0.83 mm in A. elikonius sp. n. versus 0.5-0.65 mm in A. mariangeae), in the absence of ventral setae on Th.1 in elikonius, in the different number of setae on Ant.1 (9 in A. elikonius versus 8 in A. mariangeae), by unguiculus length (equal to or slightly longer than unguis in A. elikonius versus 0.75 of U 3 in A. mariangeae), and by the absence of a 0 on Abd.5 in A. mariangeae, but all these characters are probably variable. ‡ Slightly different formula of ventral pso, i.e. 11/000/0112, is given by Fjellberg (1998) Three remaining species of the volinensis-group were described as having a lesser number of dorsal and ventral pso (see Table 2). Yet this probably needs verification. In any case, clear differences in the ecological preferences of compared species confirm the specificity of A. elikonius sp. n. The monsoon subtropical climate of southern Korea (the habitats of A. mariangeae, A. donjiensis, and A. jindoensis are sand beaches) and central China (vicinity of Shanghai, the only known locality of A. foliatus) has nothing to do with the extremely continental conditions of mountainous Yakutia (about 160 km from Oymyakon, one of the coldest places on Earth), where A. elikonius sp. n. was found. Nevertheless, the probability that some of these nominate species can probe to be conspecific with A. elikonius sp. n. cannot be completely ruled out until their adequate redescriptions.
Etymology. The new species was named after its type-locality, Elikon River. Distribution. Still known only from the region of the type-locality, where it inhabits a number of different communities up to 1,500 m alt. Description. Colour white. Size 0.55-0.62 mm. Body slender and elongated, slightly wider in region of Abd.4. Antennae about as long as head, antennal area not clearly demarcated. Ant.4 rather short and wide, 2 usual sensilla not especially thickened but distinct, a subapical organite and a basal microsensillum present. Ant.3 organ consisting of 4 low papillae, 2 sensory rods, 2 smooth sensory clubs without clear ribs, 5 guard setae, and a lateral microsensillum. Ant.1 and 2 with 8 and (12)13 setae, respectively. PAO wide (length/width ratio ca 1.5), with about 7-10 composed vesicles set close together. Labrum as a rule with 7 setae and 4 prelabral ones, but holotype with an abnormal number of setae set asymmetrically. Apical part of labium with thick  7 setae in row B, setae M and Y present. Unguis simple, without teeth, unguiculus narrow, gradually tapering, with fine filament reaching tip of unguis. Anal spines curved and rather thin, set without papillae. MVO in reproductive males probably identical to that in A. volinensis but in both available mature males only thickened setae present in mid-ventral section of Abd.4 (Fig. 53).

Allonychiurus subvolinensis
Affinities. A. unisetosus sp. n., A. volinensis and A. subvolinensis sp. n. constitutes a rather homogeneous subgroup among the known species of the volinensis-group of Allonychiurus. All of them are characterized by identical position of antennal pso with b-pseudocellus set close to midline and out of antennal area (cf. Figs 50-51 and 40). Such a position is unique for the group. A. unisetosus sp. n. shares equal number of tibiotarsal setae (9) with A. subvolinensis sp. n. and identical ventral chaetotaxy of thorax (0-1-1) with A. volinensis (see Table 2) but differs from both species in having only two prelabral setae, one ventral pso on Abd.4 as a rule, only one axial macroseta on dorsal side of Abd.6 (cf. Figs 54 and 49), and clearly thinner AS.
Etymology. The name reflects the presence of only one axial macroseta on Abd.6 in the new species separating it from similar congeners.

Allonychiurus asiaticus
The above new material collected from an area lying far south (more than 1,000 km) of the terra typica of the species differs from the original description in having more clearly differentiated tergal sensilla, but otherwise being very similar. These specimens may even represent a separate species, but material from intermediate areas is needed to evaluate the significance of these differences.