A new genus of the millipede tribe Brachyiulini (Diplopoda: Julida: Julidae) from the Aegean region

A new genus of the julid tribe Brachyiulini, Enghophyllum gen. nov., is described, comprising two species from Greece. The type-species, E. naxium (Verhoeff, 1901) comb. nov. (ex Megaphyllum Verhoeff, 1894), appears to be rather widespread in the Aegean: it is known from Antiparos Island and Naxos Island (the type locality), both in the Cyclades, as well as East Mavri Islet, Dodecanese Archipelago (new record). The vulva of E. naxium is described for the fi rst time. In addition, E. sifnium gen. et sp. nov. is described based on a single adult male from Sifnos Island, Cyclades. The new genus is distinct from other genera of the Brachyiulini mainly by its peculiar gonopod structure, apparently disjunct and at least mostly apomorphous: (1) promeres broad, shield-like, in situ protruding mostly posteriad, completely covering the opisthomeres and gonopodal sinus; (2) transverse muscles and coxal apodemes of promere fully reduced; (3) opisthomere with three differentiated processes, i.e., lateral, basal posterior and apical posterior; (4) solenomere rather simple, tubular. The evolution and biogeography of the new genus are briefl y discussed, both suggesting its profoundly long isolation in the Aegean region from the contribal genera in the adjacent Balkans and Anatolia.


Introduction
The present paper continues our efforts in revising the large julid tribe Brachyiulini (Lazányi et al. 2012;Vagalinski et al. 2013), this time being devoted to the description of a new oligotypic genus from several Aegean islands of Greece. Its type species, E. naxium (Verhoeff, 1901), previously referred to as Brachyiulus naxius, Chromatoiulus (Diaxylus) naxius or Megaphyllum naxium, has been found to occur both in the Cyclades and Dodecanese archipelagos, whereas the second congener, E. sifnium gen. et sp. nov., is a new species described below. The gonopod and, to some degree, the vulval structures of these two species appear to be of great interest by shedding additional light on the evolution and biogeography of the entire fauna of eastern Mediterranean Brachyiulini.

Material and methods
The present work is based on the examination of material from the Non-insect Invertebrate Collection, National Museum of Natural History (NMNHS), Sofi a, and the Natural History Museum of Denmark, Zoological Museum, University of Copenhagen (ZMUC). Types of the new species are deposited in ZMUC. Specimens are stored in 70% ethanol and small dissected parts have been placed in genitalia vials which are kept together with the dissected individuals. Female vulvae were mounted on slides in Faure-Berlése medium and incubated at room temperature (around 25°C) for several hours. After having been studied, the slides were put into distilled water, vulvae were removed from the Faure-Berlése medium and put into 70% ethanol.
Line drawings were made with a Leica M125 (HNHM) stereo microscope and an Olympus BX40 light microscope.
The apical end of the opisthomere is distinguished as the solenomere; all processes emerging from the opisthomere's main body are named simply according to their position as lateral, basal and apical posterior processes.

Etymology
This genus is named in honor of Prof. Henrik Enghoff from the ZMUC, not only for his vast contribution to our knowledge of diplopods, but also for his continuous encouragement and help during our work with Brachyiulini.

Males
Mandibular stipes slightly protruding ( Fig. 1F-G). First leg-pair like simple, rounded, somewhat converging hooks; the two hooks converging in an obtuse angle (Fig. 1H). Second leg-pair ( Fig. 1I) with two ventral pads: on postfemur (pf) and tibia (t). Penis (Fig. 1J) signifi cantly small, deeply hidden in penis sac, with two very short lobes and two rounded small lamellae running parallel, i.e., not diverging (rl and ll for right and left lamella). Pleurotergum of the 7 th body ring protruding like a simple shovel (broken in the unique holotype of E. sifnium gen. et sp. nov., thus investigated only in E. naxium, see  Gonopods noticeably big compared to body size, directed mostly posteriad. Promere bevelled in laterobasal corner to fi t into the inner curve of the 7 th body ring; latero-basal corner of promere only with thin triangular lamella (tl) (Fig. 2D-E). Promeres attached to each other in their meso-basal corner with a chitinous bridge. A short, oblique ridge (r) on posterior surface of promere (Figs 2A, D-E, 4A, C). Flagellum emerging from a well-developed sinus (fsi) (Fig. 2D-E). At the level of this sinus a deepening on anterior surface of the promere (d) (Fig. 4B). Opisthomere entirely concealed under the promere, protruding postero-ventrad, the two promeres tightly closing the ventral opening on 7 th ring like a pair of shutters. Opisthomere not embedded in the groove formed by the oblique ridge on the promere's posterior surface (unlike in Megaphyllum), but "riding" on the ridge (

Females
The only available female of E. sifnium gen. et sp. nov. is in stadium IX (with 8 rows of ocelli) and proved to be subadult, but females of E. naxium in stadia VIII-IX (with 7-8 rows of ocelli) had fully developed vulvae, which we used here for the description of female sexual characters. First two legpairs slightly swollen. Vulva ( Fig. 3): Subcylindric in shape, the mesal half shorter, the opening large, oval, apical. Operculum shorter than bursa, with around 20-25 setae. Mesal and lateral sclerites with 2-3 setae each. Apodematic tube not opening apically into a longitudinal median cleft as in, e.g., Megaphyllum species, but into a wide sinus (si). Around this sinus a thin wall, formed by the apical part of the bursa. These walls with small dot-like pores; most setae (around 20/side) emerging from this apical region. Apodematic tube (at) ending in two receptaculi seminis or ampullae. Central ampulla (ca) drop-like. Distal, globular ampulla (da) joining the apodematic tube through a twisted connecting tube (ct). Connecting tube very long, thus the distal ampulla hanging out of the bursa; easily broken off during preparation.

Remarks
The everted penes of the adult E. naxium male from Antiparos was quite soft, amorphous and huge, contrary to the stout, minute penes, deeply hidden under the 2 nd leg-pair coxae, observed in the E. naxium male from Mavri Islet and the E. sifnium gen. et sp. nov. male from Sifnos. It is possible that penes vary in size between copulating and non-copulating periods.

Diagnosis
Differs from the only congener, E. naxium (Verhoeff, 1903), by the apical margin of the promere being slightly convex and the opishtomere being signifi cantly compressed meso-laterad.
COLOUR. Basic colour yellowish-beige (obviously faded) with a narrow brownish middorsal band, divided by a white median line. Two wider, diffused brownish bands at ozopore level, more intensely coloured around the ozopores.
TELSON (Fig. 1D). Anal valves covered with scarce, long setae and a row of much shorter setae along each valve's caudal margin. Preanal process without any setae in the adult male, but with a few setae in the two subadults; straight, stout, protruding near to the level of the longest anal setae (for the adult male). Subanal scale triangular, with a rounded tip, slightly protruding behind rear contour of anal valves; with two pairs of long setae laterally. GONOPODS ( Fig. 2A-G). Promere (Fig. 2D-E and P on Fig. 2A-C) wide, slightly narrowing apically towards mesal margin. Apical margin convex, mesal and lateral margins almost parallel, the lateral one slightly sinuous. Opisthomere with three processes: a fl at lateral process (lp on Fig. 2A-C, F), a slightly curved, rod-like basal posterior process (bpp on Fig. 2A-C, F-G) and an elongate, massive, rod-like apical posterior process (app on Fig. 2A-C, F-G); all processes well-developed, tapering and not furcated. A micro-spinose pillow (msp) around the base of solenomere (s) ( Fig. 2A-C, F-G). Entire opisthomere strongly compressed meso-laterad (compare Fig. 2C with Fig. 4A).

Diagnosis
Differs from the only congener E. sifnium gen. et sp. nov. by the apical margin of the promere being concave, the opisthomere not compressed meso-laterad, e.g., the lateral process (lp on Fig. 4A-D) parallel to the promere, the lateral process (lp) much broader and the apical posterior process (app) broader and blunter.  This species was described on the basis of two males and four females and has only been found once since its description, viz., by Mauriès & Karamaouna (1984) from Antiparos. The new specimens from east Mavri belong to the collection of Karamaouna, but they are slightly different: the apical posterior process (app) of the solenomere is not bifurcated, but undivided and slightly serrated (serration visible in posterior view) (Fig. 4A, C-D).

Discussion
Enghophyllum naxium was hitherto treated as a Megaphyllum species (or under its former synonyms Chromatoiulus and Brachyiulus s. auct.) (Lazányi et al. 2012). It was placed in the subgenus Diaxylus Attems, 1940(type species: Chromatoiulus anatolicus Attems, 1927 on the basis of having a simple promere, a simple solenomere, a posterior process at right angles to the opisthomere and lacking a bristle-row along the fl agellum channel (Attems 1940). After thorough investigations of this species and of E. sifnium gen. et sp. nov., supplemented with the examination of Megaphyllum argolicum (Verhoeff, 1900), M. asiaeminoris (Verhoeff, 1898), M. euphorbiarum (Verhoeff, 1900) and M. anatolicum (Attems, 1927), all referred to Diaxylus, it became apparent that E. naxium and E. sifnium gen. et sp. nov. form a group neither with the aforementioned species, nor with any other Megaphyllum subgenus or Brachyiulini genus. The combination of external morphological characters together with male and female genitalia characters supports the establishment of a new brachyiuline genus: Enghophyllum gen. nov. Considering gonopodal structure, some similarities with Megaphyllum byzantinum (placed in the monotypic subgenus Byzantorhopalum Verhoeff, 1930) can be discerned, namely the well-developed lateral process of the opisthomere and the cylindrical vulvae with a mostly apically positioned opening. However, some external and internal features of the vulvae, the shape of the penes, and notably the unique arrangement of the promeres leave no doubt that these species belong to a different genus. The known distribution of Enghophyllum gen. nov. is so far restricted to three islands in the Cyclades (Naxos, Antiparos and Sifnos) and one small islet in the westernmost part of the Dodecanese Islands, Greece (east Mavri Islet, Fig. 5). There are currently 10 other species of Brachyiulini reported from the Aegean Islands: Brachyiulus stuxbergi (Fanzago, 1875) (Crete), Megaphyllum bicolor (Loksa, 1970) (Rhodos, Naxos), M. brachyurum (Attems, 1899) (Thasos), M. chiosense Lazányi & Korsós, 2012 (Chios), M. creticum (Strasser, 1976) (Crete), M. mueggenburgi (Verhoeff, 1901) (Karpathos, Kasos), M. rossicum (Timotheew, 1897) (Samothraki), M. sapphicum (Strasser, 1976) (Lesvos), M. syrense (Verhoeff, 1903) (Syros) and M. taygeti (Strasser, 1976) (Crete) (Felesaki et al. 2010;Lazányi et al. 2012 Heller (1976). He found most Aegean species of the terrestrial gastropod family Enidae to occur only on one or several adjacent islands, which is considered to refl ect the lack of active immigration in recent geological times. These biogeographic similarities could be explained with the very limited dispersal abilities characteristic of millipedes and land snails in general. Another reason may be the rather weak ecological specialization common to both groups. Thus, their current distribution would be less affected by the presence/absence of particular plant communities or habitat degradation, unlike the case with, e.g., the central Aegean Oniscidea (Sfenthourakis 1996), and would more clearly refl ect the geological events that took place in the Aegean region. In this respect the record for E. naxium from east Mavri may present interesting information in terms of further revealing the biogeographic affi liations of this group of several small islands (Kinaros, west and east Mavri and Levitha), positioned between the Cyclades and the Dodecanese although offi cially ascribed to the latter. Sfenthourakis (1996) suggested that Mavri is biogeographically related to several scattered islands in the south Cyclades, while Levitha has more in common with the other Dodecanese islands. However, the overal scarce faunistic data hamper any fi rm conclusions being reached about the islets' paleogeographic history.
In spite of the rather few distribution records for the new genus, as well as the insuffi ciently investigated millipede fauna of the Greek islands in general, we may suppose that the genus described here represents a distinct lineage of Brachyiulini that has developed in relatively long-lasting isolation, as there are currently no apparent closely related forms in the Balkans or in Anatolia. The relict character of the Aegean fauna has already been demonstrated for other invertebrate groups like tenebrionid beetles (Fattorini & Fawles 2005) and land snails (Heller 1976;Welter-Schultes & Williams 1999).
In conclusion, it can be assumed that Enghophyllum gen. nov. confi rms the importance of the Aegean Islands together with south Greece and west Turkey for the evolution of the millipede tribe Brachyiulini, being inhabited by both highly distinct local forms, as well as by members of (sub)genera and species groups found in other regions -evidence of successive processes of migration and vicariance resulting in intense speciation. On the other hand, the dominant position of the tribe in terms of species richness among Diplopoda from the region, combined with a high rate of endemism, provide good reasons for using the group as a model for further elucidation of the complicated paleogeographic history of the Aegean Archipelago.