Typhloiulus rhodopinus, Vagalinski, Boyan, Stoev, Pavel & Enghoff, Henrik, 2015

Vagalinski, Boyan, Stoev, Pavel & Enghoff, Henrik, 2015, A review of the millipede genus Typhloiulus Latzel, 1884 (Diplopoda: Julida: Julidae), with a description of three new species from Bulgaria and Greece, Zootaxa 3999 (3), pp. 334-362 : 352-358

Typhloiulus rhodopinus sp. n.

Figs 30–48

Material. Holotype ♂ (intact) ( NMNHS), Bulgaria, East Rhodopi Mts., E of Madzharovo, abandoned mine gallery on the left bank of River Arda, 15.V.1996, B. Petrov leg.

Paratypes: 1 ♂ broken into 2 pieces, gonopods dissected ( ZMUC), same locality and collecting data as of the holotype; 3 ♂ ( NMNHS) (one gonopod mounted for SEM, second gonopod missing), Bulgaria, East Rhodopi Mts., Madzharovo Distr., abandoned mine gallery, 2 km from Madzharovo, road to v. Borislavtsi, below the road, decayed wood, 20.IV.1996, P. Stoev & B. Petrov leg.; 2 ♀ (1 broken into 2 pieces, with dissected vulvae, left vulva mounted for SEM; second specimen broken into 2 pieces, not dissected) ( NMNHS), same locality and collecting data; 1 intact ♀ ( ZMUC), same locality and collecting data; 4 ♂ (1 intact specimen, 2 broken into 2 pieces, 1 broken into 3 pieces, with dissected gonopods), 4 ♀ (all broken into 2 or more pieces) ( NMNHS), Bulgaria, East Rhodopi Mts., Lyubimets Distr., abandoned mine gallery between v. Lozen and v. Cherna mogila, 4.XI.1999, B. Petrov, S. Beshkov & D. Vasilev leg.; 3 ♂ (2 dissected), 7 ♀ ( ZMUC), same locality, rotten log, 12.IV.1998, B. Petrov & B. Barov leg.; 2 ♂ (1 intact, second one broken into 3 pieces, penis dissected, gonopods mounted for SEM) ( NMNHS), Greece, Evros Distr., v. Avas, cave Avanos, 17.V.1987, P. Beron leg.

Additional material. Bulgaria: 4 juv. ( NMNHS), East Rhodopi Mts., Madzharovo Distr., abandoned mine gallery, 2 km from Madzharovo, road to Borislavtsi, below the road, decayed wood, 20.IV.1996, P. Stoev & B. Petrov leg.; 1 ♀ ( NMNHS), Kardzhali Distr., v. Oreshari, cave Razklonenata, 3.IV.1992, B. Petrov leg.; 4 ♀ ( NMNHS), Ivailovgrad Distr., cave Dupkata, 27.IV.1995, B. Petrov leg.; 2 ♂ ( ZMUC); same locality, 23.IV.1996, B. Petrov & P. Stoev leg.; 1 ♀, 1 subad. ♂ ( NMNHS), East Rhodopi Mts., Haskovo Distr., v. Dolno Cherkovishte, cave Zandana, 21.IV.1996, B. Petrov & P. Stoev leg.; 1 ♀ ( NMNHS), East Rhodopi Mts., Momchilgrad Distr., v. Kremen, cave Zlatnata yama, 27.IV.1996, P. Stoev & B. Petrov leg.; 1 ♂, 1 ♀ ( NMNHS), East Rhodopi Mts., abandoned mine gallery at the base of Kovan kaya near v. Dolno Cherkovishte, rotten log, 13.IV.1998, B. Petrov leg. Greece: 2 ♂ (1 pair of gonopods mounted for SEM), 1 ♀ ( ZMUC), Thrace, Nomos Evros, 20 km SW Dhadhia [Dadia], 80 m, stream, oak forest, 21.IV.1994, leg. W. Schawaller, H. Schmalfuss ded. 1994.

Diagnosis. T. rhodopinus sp. n. is morphologically similar to T. albanicus and T. giganteus . However, it can be readily distinguished by a combination of certain gonopodal and external body characters (see Table 1 ).

Etymology. Named after the Rhodopi Mts., the main distribution area and type locality of the species. Adjective.

Description. Holotype with 52+3+T body rings, l = 24 mm, h = 1.3 mm. Paratype males with 49–63 podous body rings, with 1–3 apodous rings, l = 17–37 mm, h = 1–1.65 mm. Paratype females with 64–72 podous body rings, with 1–2 apodous rings, l = 22–34 mm, h = 1.1–1.65 mm.

Colouration: light to dark brownish-beige, prozonae darker than metazonae, legs and antennae yellowish beige to light brown.

External structures: 4 supralabral (a central 5th seta present in one male, probably an aberration) and 14–16 labral setae. Labrum tridentate. Antennomeres 2, 3, 4 and 5 more or less equally long, somewhat longer than 6th; antennomere 5 with a dense whorl of sensilla basiconica (somewhat smaller than 4 sensilla at antennal apex) dorsally and laterally; a similarly dense whorl of very small sensilla present on antennomere 6. Male mandibular stipites not enlarged. Gnathochilarium of normal julid appearance, with 3 apical setae on each stipes, and with 5 setae in a row on each lingual plate; promentum 0.5–0.53 times as long as lingual plates, length to width ratio: 1.6– 1.8. Collum smooth, with several shallow, parallel striae at postero-lateral corner. Prozonae smooth. Metazonae with mostly straight, parallel striae, with 5–7 striae in a square with sides equal to metazonal length just below ozopore level; a dense whorl of erect setae, ca 3/5 to equal to metazonal length at metazonal hind margin. Ozopores placed behind pro-metazonal suture at ca 1/5–2/5 of metazonal length. Telson ( Fig. 31 ): preanal process moderately long, pointed, curved downwards, rather wedge- than roof-shaped, slightly surpassed by longest anal setae; several long setae present on its upper side. Subanal scale very short, rounded, not protruding behind rear contour of anal valves, with 2 rows of setae: a marginal row of 10–12 setae and an inner row of 5 setae. Anal valves densely pilose. Male pleurotergum 7 ( Fig. 32 ) ventrally with rounded protrusions directed ventro-posteriad. Male leg-pair 1 ( Fig. 33 ) normal hooks, slightly turned upwards, with or without a minute tarsal remnant. Male legs in anterior part of body ( Figs 34 & 35 ) without adhesive pads, but with a pit (p) both on postfemur and tibia. Tarsus of male mid-body leg 1.6–1.8 times longer than tibia, and 3 times longer than apical claw.

Penis ( Fig. 36 ): elongated trapezoidal, with moderately long, mostly parallel apical lobes (al) ending with long, tapering, somewhat converging terminal lamellae (tl).

Gonopods ( Figs 37–47 ): Compact, in situ entirely concealed inside gonopodal sinus, promere covering tip of mesomere, opisthomere considerably outreaching both pro- and mesomere. Promere ( Figs 37–39 ) somewhat higher than broad, with a convex lateral margin, a straight to slightly concave mesal margin and a rounded apex, strongly bent posteriad; apex and distal part of lateral margin with a tuberculate surface; parabasal internal lobe (il) short, edgy, with 2–4 apical setae; parabasal external lobe (el) elongated, distally tapering, somewhat outreaching internal one. Flagellum (f) thin, except for a very thick basal part, nearly 1.5 times longer than promere. Mesomere (M in Figs 40, 41 & 43–46 ) rather short and robust, bulging, with a tuberculate apical surface. Opisthomere ( Figs 40–47 ) with a slightly convex posterior margin, without processes; basal spine (bs) crescent-like bent; single, smaller spines present distally behind flagellum channel (fc); intermediate lamella (l) well-developed, weakly to moderately pronounced, reaching ca 2/3 of opisthomere height; velum (v) unipartite, thin, pointed, slightly to strongly bent distad; a jagged distal lamella (dl) between velum and solenomere; solenomere (s) short, with a flat, posteriorly oblique apex, with a dense row of seti-form filaments (sf).

Vulva ( Fig. 48 ): Somewhat compressed antero-caudally, mostly symmetric, mesal valve slightly broader than lateral one. Opening (o) elongated, cleft-like, nearly reaching bursal base. 4 setae in a vertical row on each valve. Operculum (op) considerably higher than bursa, broadly convex, with 4 vertical rows of 4–5 setae each. Receptaculum seminis consisting of a moderately thick central tube (ct) not forming a distinct central ampulla and of a thin, somewhat folded posterior tube (pt) ending into an egg-shaped posterior ampulla (pa).

Remarks. The examined material shows moderate variation in certain morphological characters. Morphologically the most aberrant is the population from the vicinity of Lyubimets which, unlike all remaining specimens examined, possesses 14 instead of 16 labral setae, and a considerably more strongly bent velum of the opisthomere. Also noteworthy are the observed size differences, the specimens from Greece being considerably larger than those from Bulgaria. Colouration seems to vary as well both in intensity and tincture. However, the above-mentioned differences are not sufficient to justify the erection of new subspecies, let alone more than one species.

Ćurčić et al. (2003) emphasized the similarity between T. albanicus and T. giganteus . T. rhodopinus sp. n. obviously belongs to the same species-group characterized by an opisthomere with a well-developed, but not conspicuously pronounced intermediate lamella, a slender, unipartite, spine-like velum, a jagged distal lamella between the velum and the solenomere; and a slender penis with converging sides, deeply divided apical lobes and long, parallel or converging terminal lamellae. The group is so far confined to the central and eastern parts of the Balkan Peninsula.

Typhloiulus is one of the well-defined genera of Julidae which shows a considerable number of distinctive characters, as already listed in the beginning of the systematic part. However, the grouping of species within Typhloiulus is far from settled. Many, often controversial, subgenera have been introduced in an attempt to resolve the intrageneric systematics. Two major, presumably natural, well defined groups appear from among the 33–36 presently known species. These are the subgenera Typhloiulus sensu stricto and Stygiiulus Verhoeff, 1929 . Typhloiulus s. str. was properly characterized by Strasser (1962, 1966). It comprises 15 unquestionable species: T. albanicus , T. bureschi , T. ganglbaueri , T. georgievi , T. giganteus , T. hauseri , T. incurvatus , T. kotelensis , T. longinquus , T. nevoi , T. psilonotus , T. serborum , T. strictus , T. orpheus sp. n. and T. rhodopinus sp. n. They all share a very similar gonopod conformation, and the group is further supported by certain external characters, such as: male leg-pair 1 typically hook-shaped (except for the known males of T. ganglbaueri ), the femora of male legs in the anterior part of the body without modifications on their ventral surface, the ozopores located closely behind the suture, and the preanal process straight or bent downwards. The distribution of Typhloiulus s. str. is mostly confined to the Balkan Peninsula, with several species expanding to the Carpathians and the Apennine Peninsula. In having a more elongated pro- and mesomere, the position of T. beroni , (from eastern Albania) in relation to this group needs still to be clarified. The position of T. bosniensis likewise remains unclear: this species lacks an intermediate lamella between the meso- and opisthomere. On the other hand, both latter species agree in all remaining diagnostic characters of Typhloiulus s. str.

The intermediate lamella attracts attention as a potentially important phylogenetic character within Typhloiulus and the “Typhloiulini” in a broader sense. Apart from the members of Typhloiulus s. str., the “typhloiuline” genus Lamellotyphlus Tabacaru, 1976 displays a particularly strongly developed lamella which is completely fused with the mesomere. Enghoff (1987) suggested that such a lamella may represent a transitional evolutionary stage towards the true pro-mesomeral forceps seen in the “higher julids”, and placed the genus in the tribe Leucogeorgiini which is characterized by such an intermediate gonopod conformation. If this is true, then Typhloiulus s. str. should be the most plesiotypic/basal group within the genus, and the entire genus Typhloiulus as understood here, possibly subdivided into several monophyletic genera, and possibly together with some other related genera, should be placed near the base of the “higher julids”. However, Makarov et al. (2003) argued that the lamella in Lamellotyphlus seems not to be homologous with that observed in the Leucogeorgiini, suggesting instead that it may be a secondary structure leading to a gradual fusion between the already differentiated meso- and opisthomere; this is most evident in the monotypic genus Banatoiulus Tabacaru, 1985 , from Romania.

Strasser (1962) listed four species in the subgenus Stygiiulus : T. illyricus , T. ausugi , T. maximus and T. montellensis . T. tobias seems to be another good candidate member of this group, although slightly deviating in its gonopodal features from the remaining species. The five species can be referred to as “the Alpine group”, considering their compact distribution in the southeastern foothills of the Alps.

Spelaeoblaniulus Ceuca, 1956 , is another distinctive subgenus which comprises one species, T. serbani , that possesses a very unusual opisthomere with a large, finely striated, microspinose solenomere, a vestigal velum and a conspicuous posterior process. T. motasi perhaps belongs to the same lineage given its morphological similarity to T. serbani ( Tabacaru & Gava 1992) . Their distribution in the Carpathian region provides further support of such an assumption.

The remaining subgenera are mostly weakly defined, inappropriately used or even invalidly proposed, and thus can hardly be applied to the current systematics of Typhloiulus .

The subgenus Allotyphloiulus Verhoeff, 1905 , was erected to encompass the blind Cylindroiulus vulnerarius (Berlese, 1888) alone. For reasons hard to explain, Loksa (1960) found his newly described Typhloiulus polypodus to be most similar to the above cylindroiuline species and assigned it to the same subgenus.

Haploprotopus Verhoeff, 1899 , includes only T. ganglbaueri . Although Strasser (1962) recognized the subgenus, he did not consider it justified, since the gonopods of T. ganglbaueri are typical of T. s. str., and the only definitive character of the subgenus, the unmodified male leg-pair 1 as described by Verhoeff (1899) from 3 topotypic males, is more likely evidence of “junior males” (see below for explanation) than of an outstanding member of Typhloiulus .

Two species were assigned to Inversotyphlus Strasser, 1962 , T. lobifer and T. longipes . Both share an anteriorly bent promere, the main character that defines the subgenus, but differ significantly in a number of other features. On the other hand, T. lobifer is similar to T. gellianae , as already noted by Makarov et al. (2006).

Mesoporoiulus Verhoeff, 1905 , was invalidly proposed without a type species. Later Jeekel (1971) designated Typhloiulus roettgeni (described on the basis of a female, of a dubious status, but tentatively considered as a synonym of Trogloiulus boldorii ) as the type species.

Attems (1959) listed 5 species: T. albanicus , T. incurvatus , T. bureschi , T. psilonotus and T. kotelensis , in his newly defined subgenus Smeringolophus . However, all of these are recognized as members of Typhloiulus s. str. T. psilonotus is furthermore the only species in the subgenus Xestotyphloiulus Verhoeff, 1899 .

Not only the taxonomy of Typhloiulus , but also its monophyly is somewhat questionable. As for many other julid genera, the pioneer researchers used very broad defining concepts compared to the present-day systematic standards. This has resulted in many significantly diverse species to have been treated under Typhloiulus , (almost) solely based on the lack of ocelli. Some of them, like e.g. Apfelbeckiella trnowensis (Verhoeff, 1928) and Elbaiulus chrysopygus (Berlese, 1888) , are today considered as hardly at all related to the Typhloiulini/ Leptoiulini . A number of other species from the “typhloiuline” group have been transferred to other genera, due to certain morphological differences, these mostly concerning the gonopodal apparatus.

However, even after those taxonomic changes, Typhloiulus looks too much like a heterogenous assemblage. Part of the problems lie in the rather unsatisfactory diagnoses of some of the other “typhloiuline” genera. The absence of a flagellum is in fact the only character that separates Serboiulus Strasser, 1962 , and Trogloiulus Manfredi, 1931 , from Typhloiulus . Being a reductive character that has arisen independently many times in the evolution of Julidae , its phylogenetic weight is low. Moreover, members of Stygiiulus seem to have more in common with Trogloiulus , both in gonopodal and external features, than with their nominal congeners. In the light of this observation, the southern Alpine distribution of both Stygiiulus and Trogloiulus may be the result of common origins, and the loss of the flagellum might have subsequently occurred in some of the descendants. Aberrant species like, e.g., T. longipes , T. motasi and T. serbani , further strengthen the impression that we are dealing with a para-/polyphyletic genus.

At present, the usage of the tribe Typhloiulini to encompass all blind julid genera possessing a pro-mesomerital forceps and a more or less well-differentiated velum of the opisthomere seems to lack a solid ground. Mauriès et al. (1997) regarded Typhloiulini as a “highly suspicious taxon”, emphasizing the impossibility to delimit it from the Leptoiulini in terms of gonopod characteristics. The two species of the genus Leptotyphloiulus Verhoeff, 1899 , which, apart from lacking ocelli, possess gonopod features of a proper Leptoiulus type, are a good example of that. Indeed, the existence of Leptotyphloiulus alone is evidence of polyphyletic origins of the Typhloiulini. None of the remaining mono- or oligotypic “typhloiuline” genera, viz., Serboiulus , Trogloiulus , Alpityphlus (see above under Typhloiulus seewaldi ), Buchneria Verhoeff, 1941 , Leptotyphloiulus , Banatoiulus and Lamellotyphlus , looks more similar to Typhloiulus s. str. than the latter looks similar to each of the ocellate genera Leptoiulus Verhoeff, 1894 , Xestoiulus Verhoeff, 1893 or Ophyiulus Berlese, 1884 . It thus appears logical to assume the presence of two or more lineages within a larger tribe Leptoiulini that may have independently taken parallel evolutionary courses towards troglo-, geo- or petrophilic modes of life, this leading to adaptations like reduction of ocelli, pale body colour, dense and long setation, and long legs and antennae.

There is, however, some evidence that at least some of these blind taxa may constitute a monophyletic group beyond the Leptoiulini . The molecular phylogeny of Enghoff et al. (2013) puts T. orpheus sp. n. (cited as Typhloiulus n. sp.) in an isolated clade, away from the genera Julus Linnaeus, 1758 , Pacifiiulus Mikhaljova, 1982 , Leptoiulus , Xestoiulus and Ophyiulus , that appear on the tree in a robust monophyletic clade.

The unusual ‘premature’ acquisition of morphological sexual maturity observed in males of some “typhloiuline” taxa add support to the natural ‘identity’ of the tribe in one or another generic composition. The phenomenon is known as “junior males” ( Strasser 1971b, see also Enghoff et al. 1993) in which the first leg-pair is unmodified or only partly modified, while the gonopods are (almost) fully developed, but are sometimes entirely protruding outside the gonopodal sinus. The term was coined by Strasser (1971b) as “Junior-Männchen” on the basis of specimens of T. illyricus and Serboiulus deelemanni ; we have observed junior males in Serboiulus spelaeophilus and Typhloiulus bureschi , and the same phenomenon probably concerns T. ganglbaueri . The adult males of these species have almost fully concealed gonopods and typical hook-like first leg-pair.

In any case, the systematics of the seemingly related and more or less overlapping tribes Typhloiulini, Leptoiulini and Julini present a tough riddle. Unravelling the phylogeny of Typhloiulus , together with the other, aberrant blind genera, is inevitably related to a detailed revision of Leptoiulini , especially of the speciose and notoriously problematical genus Leptoiulus .

The presently known somewhat more than 30 species of Typhloiulus are surely far below the real number of existing species. Many more are yet to be discovered, especially from undersampled areas in Serbia, Greece, Montenegro and, especially, Croatia, the latter the country in the Balkan Peninsula the richest in caves. Besides this, data on the distribution of most species is still insufficient, with nearly one-third so far known only from their type localities. Future studies would shrink these gaps in our knowledge and allow for better founded considerations on the systematic position and biogeographic affinities of the genus to be made.

Figures

FIGURES 30 – 36. Typhloiulus rhodopinus sp. n. paratype males from cave Avanos near v. Avas, Greece (30, 32 – 34 & 36), from Lyubimets (31) and from Madzharovo (35), Bulgaria (all in NMNHS): (30) habitus, not to scale; (31) telson, lateral view; (32) right pleurotergal protrusion, latero-ventral view; (33) leg 1; (34) leg 2; (35) leg 5; (36) penis. Abbreviations: al: apical lobe, p: pit, tl: terminal lamella. Scale bars: fig. 31: 0.5 mm, figs 32 – 36: 0.2 mm.

FIGURES 37 – 42. Typhloiulus rhodopinus sp. n. paratype males from v. Borislavtsi (38 & 40 – 42) and Lyubimets (39), Bulgaria (NMNHS), and non-type male from Dadia, Greece (ZMUC) (37): (37) left promere, meso-caudal, slightly top view; (38) right promere, caudal view; (39) left promere, meso-caudal view; (40) right meso- and opisthomere, lateral, slightly frontal view; (41) right meso- and opisthomere, meso-frontal view; (42) distal half of right opisthomere, fronto-lateral view. Abbreviations: bs: basal spine, dl: distal lamella, el: external lobe, f: flagellum, fc: flagellum channel, il: internal lobe, l: intermediate lamella, M: mesomere, s: solenomere, sf: seti-form filaments, v: velum.

FIGURES 43 – 48. Typhloiulus rhodopinus sp. n. paratype male (43, 46 & 47) from Lyubimets and paratype female (48) from Madzharovo, Bulgaria (NMNHS), and non-type male (44 & 45) from Dadia, Greece (ZMUC): (43) left meso- and opisthomere, mesal view; (44) left meso- and opisthomere, mesal, slightly frontal view; (45) left meso- and opisthomere, mesal view; (46) right meso- and opisthomere, lateral view; (47) distal part of opisthomere, mesal view; (48) right vulva, posterior, slightly mesal view (setae omitted). Abbreviations: bs: basal spine, ct: central tube, dl: distal lamella, f: flagellum, fc: flagellum channel, l: intermediate lamella, o: opening, op: operculum, M: mesomere, pa: posterior ampulla, pt: posterior tube, s: solenomere, sf: seti-form filaments, v: velum. Scale bars of figs 46 & 48: 0.2 mm.

Tables

TABLE 1. Diagnostic characters of T. albanicus, T. giganteus and T. rhodopinus.

setae on subanal scale	2 marginal	in 2 rows: 7 marginal and 3 submarginal	in 2 rows: 10–12 marginal and 5 submarginal
protrusions of the ventral margins of male pleurotergum 7	pointed	rounded	rounded
male walking legs in anterior part of body	adhesive pad on tibia, no modifications on postfemur	adhesive pad on tibia and a pit on postfemur	pit on both tibia and postfemur
distal lamella of opisthomere	small, considerably lower than the tip of solenomere	small, considerably lower than the tip of solenomere	pronounced, almost reaching the tip of solenomere
Discussion