The millipede genera Amblyiulus Silvestri, 1896 and Syrioiulus Verhoeff, 1914 in the Caucasus, with notes on their distributions (Diplopoda, Julida, Julidae)

Abstract In the Caucasus, the genera Amblyiulus Silvestri, 1896 and Syrioiulus Verhoeff, 1914 are shown to include two and four species, respectively: Amblyiulus georgicus Lohmander, 1932, from Georgia and Armenia, A. hirtussp. nov., from Azerbaijan and Dagestan, Russia, Syrioiulus adsharicus (Lohmander, 1936), from Georgia, S. continentalis (Attems, 1903), from Azerbaijan and Iran, S. taliscius (Attems, 1927), from Azerbaijan, and S. armeniacussp. nov., from Armenia. All these six species are described, illustrated, and keyed, and their distributions are mapped and discussed, based on the literature data and abundant new samples.


Introduction
The very large family Julidae, of the basically Holarctic order Julida dominates the millipede faunas of Europe and the Mediterranean, marginally extending into the Oriental realm as well (Enghoff et al. 2015;Kime and Enghoff 2017). The subfamily Pachyiulinae, often referred to as the tribe Pachyiulini, encompasses between 15 and 20 genera, or 16-22 genera or subgenera, according to Antić et al. (2018) and diagnoses and scopes. Both these genera appear to be very similar, but differ clearly in the structure of the posterior gonopods: each of these being strongly divided into two branches ("bipartite"), i.e., a frontal (= mesomeral) and a caudal (= opisthomere) branch in Syrioiulus spp., vs. "tripartite", with a third branch, a flagelliform rod, in Amblyiulus spp. (Golovatch 2018). A further refined account of the main differences between both these genera compared is given below.

Materials and methods
All material has been shared between the collections of the Zoological Museum of the Moscow State University, Russia (ZMUM), the Senckenberg Museum of Natural History in Görlitz, Germany (SMNG), and the Institute of Zoology, University of Belgrade, Serbia (IZB). The specimens are stored in 70% ethanol. Some parts of males (antennae, gonopods, legs, etc.) and females (vulvae and leg pairs 2) were dissected and mounted in glycerol on temporary microscopic slides. Photographs were taken using a Zeiss StereoDiscovery V.20 microscope and processed with Zeiss ZEN software. Line drawings were executed using a camera lucida attached to a Radical light-transmission microscope. Scanning electron micrographs were taken with a Zeiss CrossBeam 340 (Rostov-on-Don State Technical University, Rostov-on-Don, Russia) or a JEOL JSM-6510LV (SMNG) scanning electron microscope (SEM). For some SEM micrographs, the gonopods were glued to a small sticky plastic triangle, placed on an SEM-stub, air dried for two days in a glass filled with Silica gel and finally coated with gold. After examination, material was removed from stubs and returned to alcohol. Live animals were photographed in situ using a Canon PowerShot SX120 IS digital camera.
The distribution map was created using Google Earth Pro 7.3.3 and processed in Adobe Photoshop CS6.
A "body ring formula" indicates the number of podous (including the gonopodbearing segment/ring) and apodous segments/rings in an individual. This formula is p+a+T where p is the number of podous body rings, a the number of apodous body rings, and T represents the telson (Enghoff et al. 1993). Only adults have been analysed in the present study. For morphological descriptions, we largely used the terminology from Minelli (2015), for descriptions of the gonopods, that of Enghoff (1992) with changes in Vagalinski (2020).
The biogeographic regionalisation of the Caucasus follows the botanical one by Menitsky (1991).
No type material of the previously described species has been revised (mostly stored in the Zoological Institute, Russian Academy of Sciences, St. Petersburg) because of the 2020-2021 COVID pandemic, and the descriptive accounts and illustrations available in the literature are sufficiently complete and clear to allow a safe species identification. Colouration is largely described from preserved material. In the catalogue sections, D stands for a description or descriptive notes, R for new or repeated records, while M is a mere mention.

Refined characteristics of Amblyiulus vs. Syrioiulus
As shown below in the descriptions of individual species, the use of SEM allows for the distinctions between both genera concerned to be further refined. It is the opisthomere, not the entire posterior gonopod, that is bifid in Syrioiulus: a solenomere (with a distinct fovea or a deep saddle-like structure on top) and an anterior process (an anterior, lamellar branch adjacent to the solenomere) (Figs 7D, 10D, 12D-F, 14D). Using standard light microscopy, this anterior process often remains unnoticed, since it is very tightly appressed to the solenomere. In contrast, a third, anteromesal or lateral process/rod of the opisthomere is characteristic of Amblyiulus (Figs 3D, F, 5A, D-F, I, J).
As a result, it is only the structure of the opisthomere of the posterior gonopods that allows for the genera Amblyiulus and Syrioiulus to be more or less confidently diagnosed and separated. At the same time, species of Syrioiulus are mostly very similar in gonopodal conformation (Figs 7A-C, E, F, 10A-C, E, F, 12A-C, I, J, 14A-C, E-G), but they differ well in somatic characteristics, such as the presence or absence of eyes, frontal setae, and hairs on the rings (see also Key below). Similarly, within the genus Pachyiulus, several species were synonymised based solely on shared gonopodal characters (Mauriès et al. 1997), but later some have been revalidated on the basis of somatic features such as colouration (Frederiksen et al. 2012).
Diagnostic remarks. Here we follow Tabacaru's (1978) opinion that the genera of the subfamily/tribe Pachyiulinae/-ini are best to be diagnosed using a complex of characters, both gonopodal and somatic. In the latest review of this tribe (Mauriès 1982), all genera are divided into three groups depending on the structure of the apical part of the opisthomere, viz., the presence/absence of a fovea and the presence/absence of a pseudoflagellum. He mistakenly assigned the genus Amblyiulus to group 3 (along with many other genera like Dolichoiulus), which have neither a fovea nor a pseudoflagellum.
However, in accordance with our and earlier descriptions (e.g., Golovatch 2018), Amblyiulus has a fovea, however small, on the top of the solenomere. By the presence of a fovea and the absence of a pseudoflagellum, the genera Parapachyiulus Golovatch, 1979 andDangaraiulus Golovatch, 1979 also join this group (Golovatch 1979(Golovatch , 2018. According to a number of other characters, such as the presence of frontal setae, apicoventral lobes on the male mandibles, and the mesomeral process being as high as the opisthomere, Amblyiulus belongs to Mauriès' subgroup 3aa, together with the genera Syrioiulus and Promeritoconus. However, it seems noteworthy that sometimes frontal setae can be absent, while male mandibular stipites can remain unmodified.
The promere in Amblyiulus is narrowed in the basal third, in contrast to that in Promeritoconus, which is narrowed apically; in the apical part it may have one or two denticles, but sometimes none. The head can be with or without frontal setae. The eyes are mostly absent. The opisthomere of the posterior gonopod is tripartite: a solenomere, an anterior process, and an anteromesal or lateral rod, vs. bipartite in Syrioiulus.
Female. First two leg pairs unmodified. Vulva rounded, operculum and bursa equal in height (Fig. 15A). Operculum at apical margin oblique, undivided. Bursa asymmetric, lateral valve higher than mesal one. Each valve with two rows of long setae. Median field of bursa very short, narrow; emargination of median field suboval.
Remarks. This species was described from Borjom (= Borjomi), Georgia (Lohmander 1932a). The above samples represent the first formal records of this species from Armenia. It seems to populate high-montane deciduous forests in the western part of the Caucasus Minor (= Lesser Caucasus) (Fig. 16 Diagnosis. Assigned to the genus Amblyiulus primarily because of the presence of a rod on the posterior gonopod opisthomere. Differs from A. georgicus, perhaps the most similar congener known to date, by the following combination of somatic and gonopodal characters. Head with frontal setae; collum and metazonae of body rings each with a posterior whorl of setae. Promere narrow, with two side ridges. Solenomere apically with small filament-like processes. Rod of opisthomere relatively long.  adults, 45-67+1-3+T (♂♂); or length 27-28 mm, width 1.1-1.3 mm, number of body rings, 46-55+2-3+T (♀♀). Body subcylindrical (typical of Julidae), metazonae and prozonae yellowish grey (Fig. 1B). Head, a few postcollum rings and telson slightly lighter than other body rings. Collum slightly more vividly reddish. Antennae, mouthparts, and legs yellow ( Fig. 4A-C). Eyes absent. Metazonae with weak, dense, and regular striations, 21-23 striae per quarter of metazonal surface, i.e., that between dorsal axial line and ozopore ( Fig. 4A-G). Ozopores relatively large, with a stria in front, lying behind suture without touching it (Fig. 4H).
Remarks. This species seems to be endemic to the eastern part of the Caucasus Major within both northeastern Azerbaijan and the Republic of Dagestan, Russia (Fig. 16).
It is the presence of a laterally positioned rod that brings both A. georgicus and A. hirtus sp. nov. particularly close together. However, the rod in these two species is located laterally, whereas that in A. barroisi anteromesally (Enghoff 1992: fig. 11; Golovatch 2018: fig. 10C). These differences seem to be quite important, but because those three species share not only the presence of a rod, but also a small, but discernible fovea on top of the solenomere, for the time being it seems best to regard the trio as members of Amblyiulus. Genus Syrioiulus Verhoeff, 1914 Type species. Dolichoiulus polyzonus Attems, 1910, by subsequent designation of Jeekel (1971).
Diagnosis. All characters as in Amblyiulus, except as follows. Promere usually with two denticles in apical part. Head with or without frontal setae. Eyes present or absent. Opisthomere of posterior gonopod bipartite: a solenomere (with a distinct fovea on top) and an anterior process, vs. tripartite in Amblyiulus.
Female. First two leg pairs unmodified. Vulva elongated, covered with long setae (Fig. 15C). Operculum relatively low, deeply divided. Bursa asymmetric, lateral valve higher than mesal one. Median field of bursa narrow; emargination of median field narrow and elongated.
Remark. This species was originally described from Batumi, "Bortschacha" (Lohmander 1936). Our new record from near Adigeni is evidence of the species likely to be endemic to the southern part of the Colchidan biogeographic province, all within Georgia (Fig. 16). Figs 1D, 8A , B, 9, 10, 15D, 16 Pachyiulus (Dolichoiulus) continentalis Attems, 1903: 147, 148, figs 82-84 (D). Diagnosis. Differs from all congeners by the following combination of somatic and gonopodal characters. Head with frontal setae. Collum and each metazona of following body rings with a whorl of long setae at caudal margin. Eyes present. Solenomere with a group of small spines on top. Anterior process of opisthomere subtriangular apically. This species is clearly distinguished in the field from all other millipedes by its characteristic greyish yellow colouration with a yellow stripe dorsally, and its particularly strong odour clearly resembling that of Pachyiulus krivolutskyi from the western Caucasus (= Colchis).
Female. First two leg pairs unmodified. Operculum of vulva without setae on caudal surface, apical margin relatively flat (Fig. 15D). Bursa subsymmetrical, lateral valve slightly larger than mesal one. Each valve with two rows of long setae. Median field of bursa narrow; emargination of median field suboval.
Remarks. Probably one of the most common and apparently the largest species of the genus. The unusually strong odour and the chemical composition of the repugnatorial secretion are similar to those of Pachyiulus krivolutskyi (Makarov et al., pers. obs.). This species inhabits various deciduous forests in Azerbaijan, also occurring in northern Iran (Lohmander 1932b), endemic to the Hyrcanian biogeographic province (Fig. 16). (Attems, 1927) Figs 1E, 8C, D, 11, 12, 15E, 16 Amblyiulus taliscius Attems, 1927: 243, 244, figs 336-338 (D). Diagnosis. Differs from all congeners by the following combination of somatic and gonopodal characters. Head without frontal setae. Collum and metazonae of following body rings without setae. Eyes absent. Solenomere in apical part with a group of small spines. Anterior process of opisthomere subtriangular apically.
Female. First two leg pairs unmodified. Operculum of vulva without setae on caudal surface, apical margin poorly divided (Fig. 15E). Bursa mostly symmetric, lateral valve slightly larger than mesal one. Each valve with two rows of long setae. Median field of bursa narrow; emargination of median field suboval.
Remarks. This species was described from the Talysh Mts, Lenkoran, Azerbaijan (Attems 1927). In the Caucasus, this is probably one of the most common and widespread congeners. Like S. continentalis, it inhabits various deciduous forests, but it can only be considered as subendemic to the Hyrcanian biogeographic province (Fig. 16).  Diagnosis. This new species belongs to the genus Syrioiulus because of the presence of only two apices on the opisthomere. Differs from all regional congeners by the following combination of somatic and gonopodal characters. Head with frontal setae. Collum and metazonae of following body rings without setae. Eyes absent. Solenomere with a pointed process apically. Anterior process rounded on top.
Female. First two leg pairs unmodified. Vulva rounded, operculum higher than bursa (Fig. 15F). Operculum slightly divided at apical margin. Bursa asymmetric, lateral valve higher than mesal one. Each valve with two rows of long setae. Median field of bursa narrow; emargination of median field elongated and suboval.
Remark. This species seems to be endemic to the Caucasus Minor within Armenia, but most likely it also occurs in the adjacent parts of eastern Azerbaijan and northwestern Iran (Fig. 16).
Brief description. Body grey, head and collum dark yellow, antennae and legs yellow. The following somatic characteristics seem to be the most important: absence of eyes, presence of frontal setae, presence of caudal whorls of setae on metazonae, absence of an epiproct, and setose anal valves.
Remark. Unfortunately, the only female does not allow a closer generic allocation. Brief description. Body greyish yellow. Head, collum, a few postcollum rings and telson slightly lighter than other body rings. Ommatidia and frontal setae absent. Collum and each metazona of following rings with a whorl of long setae at posterior margin. Epiproct undeveloped. Anal valves densely setose.
Remarks. These specimens differ from Syrioiulus adsharicus in the absence of frontal setae and ommatidia. The absence of males makes it impossible to definitively identify the above samples. At least the taxonomic significance of frontal setae must not be overestimated, as they may be present or absent even within the same species, e.g., S. aharonii (see Golovatch 2018).
Since the two unidentified species may well prove to represent Amblyiulus or Syrioiulus, we map their records in Figure 16, but omit them from the key below.
Key to Amblyiulus and Syrioiulus species known to occur in the Caucasus (based on males)

Discussion
Two allopatric species of Amblyiulus, both likely endemic, are found to populate the Caucasus. Amblyiulus georgicus inhabits western and central Transcaucasia, while A. hirtus sp. nov. seems to be confined to northeastern Transcaucasia and the eastern Caucasus, i.e., occurring on both macro slopes of the Caucasus Major (Fig. 16).
The genus Syrioiulus is more diverse and widespread, but presently it seems to be restricted to Transcaucasia. Thus, S. adsharicus has a rather narrow distribution in the southwestern parts of the Colchidan biogeographic province. Two most widespread species, S. continentalis and S. taliscius, are often sympatric to even syntopic within the Hyrcanian biogeographic province, but the latter species also occurs in the Caucasus Minor and the eastern part of the Caucasus Major. Syrioiulus armeniacus sp. nov. inhabits the Caucasus Minor within southern and central Transcaucasia (Fig. 16).
Only two species of Syrioiulus, S. continentalis and S. taliscius, are endemic or subendemic, respectively, to the Hyrcanian biogeographic province within the Republic of Azerbaijan and Iran, while the remaining Amblyiulus and Syrioiulus spp., however provisionally, are formally strictly endemic to the Caucasus sensu lato, including Hyrcania (Fig. 16). Generally, the problem concerning the origins of the Amblyiulus and Syrioiulus species could be approached through analysing the distribution areas of both these genera. However, because their species compositions are far from settled, including new congeners and records certainly ahead, their exact ranges cannot be outlined at the moment. Based on the highest species diversity estimates, one of the main centres of Amblyiulus and Syrioiulus speciation could have lain in the Middle East, whence members of both genera might have reached the Caucasus sensu lato. The roles that both Colchis and, especially, Hyrcania, two major, relictual, meso-to hygrophilous biogeographic provinces of, and refugia in, the region concerned, could have played in the evolution and secondary speciation seem paramount. It is hardly random that most pachyiuline species in the Caucasus are encountered in the Caucasus Minor, the immediate northern peninsular continuation of Asia Minor.
Both unidentified species seem to be endemic to the Caucasus, with Pachyiulini gen. sp. 1 being confined to Ciscaucasia, and Pachyiulini gen. sp. 2 to deciduous forests in southern Colchis, occurring sympatrically with Syrioiulus adsharicus (Fig. 16).
The above picture is definitely very far from final, but it agrees well with the biogeography of the Caucasus (e.g., Abdurakhmanov 2017). Given that the Pachyiulini in the faunas of Turkey (Enghoff 2006) and Iran (Vagalinski 2020) are likewise quite poorly known, each amounting to only a handful of species, there can hardly be any doubt that future progress in the study of pachyiulines in the Caucasus region and adjacent countries will reveal numerous novelties. In addition, given the presence in Crimea of a troglobitic species, Syrioiulus kovali (Golovatch, 2008), finding cave species of Pachyiulini in the Caucasus could also be expected (Golovatch 2008;Golovatch et al. 2017;Turbanov et al. 2018). The present taxonomy and distributions as outlined above must be clarified and refined through future research, both morphology-and molecular-based.