Review of the millipede genus Levizonus Attems, 1898, with description of a new species from the Far East of Russia (Diplopoda, Polydesmida, Xystodesmidae)

The genus Levizonus Attems, 1898 is rediagnosed and shown to contain eight species from Russia (Far East), North Korea, Japan and North-East China. One species is described here as new to science: Levizonus nakhodka sp. nov. A new formal synonym is proposed: Levizonus circularis Takakuwa, 1942 = Levizonus variabilis Lokschina & Golovatch, 1977 syn. nov., the valid name being the former. Levizonus circularis Takakuwa, 1942 is recorded for the fauna of China for the fi rst time. All currently known species of Levizonus are included in a key, mapped and discussed.

LEGS. Relatively long but stout, increasingly strong in anterior part of body. Coxa and prefemur (either only prefemur or only coxa) of all postgonopodal legs (or only hind legs) distoventrally with a small short outgrowth becoming longer posteriorly toward telson. Coxa 2 with a large setose outgrowth terminating the vas deferens.

Female
BODY. Usually stouter than in male. 20 rings including telson.
LEGS. Usually slenderer than in male. Coxa 2 with a small, setigerous process. Sternites with two tiny outgrowths between leg pair 3 only. VULVAE. Setose. Bursa composed of two valves, operculum, one setose receptacle. Valves almost identical in size and shape.

Diagnosis
The main distinguishing characters of the species are as follows: apex of gonopod telopodite (Fig. 2C) not like two plates placed perpendicular to each other as in L. malewitschi; gonopod SL, more or less strongly curved mesad (vs not strongly curved mesad in other congeners); the central edge of the gonopod telopodite apex with outgrowths and spinules ( Fig. 3A-D) (vs without outgrowths and spinules in other congeners excluding L. thaumasius, L. nakhodka sp. nov. and L. malewitschi); metaterga with low bosses either everywhere ( Fig. 1)    COLOUR. Shining whitish. Tergites often patterned with lateral marbled brown spots and transverse brown stripes.
HEAD. Epicranial suture not quite reaching the level of antennal sockets. Lateral excavation of head relatively deep for accommodation of antennae. Antennae rather slender, varying from short (reaching to front edge of ring 2) to long (reaching to front edge of ring 4).
BODY. Metaterga with bosses ( Fig. 1) increasingly evident from ring 5 posteriorly, especially so in males in contrast to females and juveniles. The number and location of the bosses varies from a few (or one) only in lateral parts above peritremata to relatively numerous as transverse rows throughout each metatergite (Fig. 1B). Paraterga rounded laterally; starting from ring 12-13, they have rounded, though never clear-cut, caudal corners.
LEGS. Each coxa and prefemur of postgonopodal legs distoventrally with a short conical knob becoming spineshaped in coxa toward telson. Coxa of leg 2 with a large setigerous process curved forward and terminating with gonopore.
GONOPODS. Distal part of coxite with three strong setae laterally (two) ( Fig. 2A) and mesally (one) (Fig. 2E). Telopodite laterally with longitudinal striae along almost its entire length. Distal part of telopodite with an external dentiform process (= SL) of different length. SL more (Figs 2F, 3A-I) or less (Figs 2B, 3J) strongly curved mesad and located lower or higher. Central edge of telopodite apex with outgrowths and spinules of varying shape and size (Figs 2C, 3 A-J).
BODY. Metatergal bosses either relatively well developed or traceable, same as in juvenile.

Remarks
This species was originally described by Takakuwa (1942) from Zyônanri (modern names: village: Sangnam-ri; county: Yeonweon-gun; province: Pyeongannam-do; North Korea). However the original description is far too fragmentary. That L. circularis might prove to be a senior synonym of L. variabilis has long been suspected (Golovatch 1979;Mikhaljova et al. 2000). So, the above male from North Korea, kept at FSCB, can be considered as near topotypic, because both localities are only a few airkm away from one another. Unfortunately, information about the holotype of L. circularis could not be found; the holotype is probably lost. Maybe it was deposited in the private collection of Y. Miyosi as Takakuwa's other xystodesmid samples and was destroyed (see Tanabe 1994). Originally described from Lazovskiy (old name: Sudzukhinskiy) and Kavalerovskiy districts, Primorskiy krai, Russia (Lokschina & Golovatch 1977), L. variabilis appears to be widespread in and confi ned to the south-eastern and eastern parts of the Primorskiy krai (Mikhaljova 2017). The gonopods of L. variabilis are variable (Mikhaljova 1981a;Mikhaljova et al. 2000). The telopodite apex carries outgrowths and spinules of varying shape and size ( Fig. 3A-D). The SL is located somewhat lower and less strongly curved mesad in specimens from North Korea (Fig. 3J) and North-East China ( Fig. 2A-B) as compared to that of Russian samples.
Supported by the geographical evidence and the morphological variability of the gonopods of L. variabilis, the following new formal synonym is proposed: Levizonus circularis Takakuwa, 1942 = Levizonus variabilis Lokschina & Golovatch 1977 syn. nov., the valid name being the fi rst.
Levizonus circularis and accordingly its junior synonym L. variabilis have never been recorded from China. The species is new for the Chinese fauna. Mikhaljova, 1990 Figs 4, 13

Diagnosis
The species differs from its congeners mainly by the always smooth central edge of the gonopod telopodite apex (vs with one to numerous outgrowths and spinules in L. distinctus, L. thaumasius, L. circularis, L. nakhodka sp. nov., L. malewitschi) and the strong dentiform SL (Fig. 5A-E).

Diagnosis
The species differs from its congeners mainly by the confi guration of the gonopod telopodite apex, like two plates placed perpendicular to each other, the central one with a seminal groove, the second one being serrate at its outer margin (Fig. 6A-B) (vs without two plates in all congeners).

Distribution
Japan: central part of Hokkaido.

Remarks
This species was originally described by Takakuwa (1941) from Hokkaido Island with several collection points (Sapporo, Rumoi, Mt. Daisetu). Unfortunately he did not denote the precise type locality.
Nevertheless Levizonus montanus appears to be spread in central part of Hokkaido (see Tanabe 1994).

Diagnosis
The species differs from congeners mainly by the confi guration of the gonopod telopodite with subapical large K laterally ( The species seems to be particularly similar to L. thaumasius but differs in a telopodite apex strongly curved caudad and as a result a large well-developed lateral subapical K (Fig. 8C) (vs low, not kneeshaped outgrowth (K1) in L. thaumasius, Fig. 10D), as well as the narrower apex of telopodite, permanent presence of a central P on telopodite apex and by a stick-like SL (Fig. 8B-D).

Etymology
The specifi c epithet refers to the type locality, a noun in apposition.   COLOUR. Shining whitish in alcohol, tergites often beige.

Material examined
HEAD. Epicranial suture not quite reaching the level of antennal sockets. Lateral excavation of head relatively deep for accommodation of antennae. Antennae rather slender, in situ reaching to the front edge of ring 4.
TELSON. Caudal dorsal projection with relatively long sparse setae, blunt apically, often apex with tiny excavation.
LEGS. Each coxa of hind legs distoventrally with a small conical outgrowth. Coxa of leg 2 with a large setigerous process curved forward and terminating a vas deferens.
GONOPODS. Distal part of coxite with three strong setae mesally (one) and laterally (two) (Fig. 9A). Telopodite laterally with longitudinal striae along almost its entire length (Fig. 8A). Telopodite apex relatively sharply curved mesad, as result distal part of the telopodite laterally carrying a large subapical K (Figs 8A, 8C, 9B). Apical part of telopodite with an external stick-like process (= SL) curved mesad as well as always with central P carrying numerous, small outgrowths (Figs 8B-C, 9B). Inner edge of telopodite apex with several small outgrowths.

Diagnosis (after Tanabe 1994)
The species differs from all congeners mainly by the shorter gonopod telopodite carrying a longitudinal blade and by a trifurcate apex of telopodite.

Distribution
Japan: western part of Hokkaido.

Diagnosis
The species differs from congeners mainly by the confi guration of the central edge of the gonopod telopodite apex, which is smooth (rarely) or with one to numerous small outgrowths (Figs 10B, 11Bseveral outgrowths) (vs without outgrowths in L. laqueatus, L. montanus, L. takakuwai), the gonopod telopodite with a low subapical not knee-shaped outgrowth (K1) (Fig. 10D) (vs without the outgrowth in other congeners excluding L. nakhodka sp. nov. and L. circularis; with K in L. nakhodka sp. nov.; not knee-shaped in L. circularis) and with a curved SL (Fig. 10B)

Remarks
Originally described from the environs of Vladivostok, Primorskiy krai, Russia (Attems, 1898), this species is common and abundant in the western and central parts of the Primorskiy krai (Mikhaljova 2017). The dorsum is often distinctively patterned (Fig. 12A-B). Its abundance ranges from 24 to 86.5 ind./m 2 in montane forests (the maximum abundance recorded is 172.5 ind./m 2 ), but this species is not numerous (4 ind./m 2 ) in fl oodplain forests (Mikhaljova 1988). Vulva as in Fig. 10E.

Discussion
At the moment, the genus Levizonus is represented by eight species (including a new species described here) in the Far East of Russia (Primorskiy krai), Japan (Hokkaido Island), North Korea and North-East China (Fig. 13). Five of the species seem to be endemic to Russia, and two are endemic to Japan. Only L. circularis is known from Russia, North Korea and North-East China. Only three species appear to be relatively widespread in the Far East of Russia. Levizonus thaumasius seems to be confi ned to the central and central-western parts of the Primorskiy krai. Levizonus circularis is distributed mainly in the eastern territory, whereas L. malewitschi occurs chiefl y in the north-eastern part of this region. The remaining Russian congeners tend to be more local in distribution: L. distinctus is recorded only in the Sikhote-Alin State Nature Biosphere Reserve, L. laqueatus in the Lazovskiy State Nature Reserve and L. nakhodka sp. nov. in the environs of Nakhodka. Japanese species of Levizonus are found in the central (L. montanus) and western (L. takakuwai) parts of Hokkaido Island (see Tanabe 1994). Members of the genus live mainly in various forests at low and middle altitudes. Russian species of Levizonus are characterized by an extended period of reproduction, with the dominance of soil-dwelling juveniles over adults that prefer to live in litter (Mikhaljova 1983). As a rule, they are dominant in millipede communities in forests with mature soil.