Stenotaenia

Bonato, Lucio & Minelli, Alessandro, 2008, Stenotaenia Koch, 1847: a hitherto unrecognized lineage of western Palaearctic centipedes with unusual diversity in body size and segment number (Chilopoda: Geophilidae), Zoological Journal of the Linnean Society 153 (2), pp. 253-286 : 276-280

THE GENUS STENOTAENIA

Our taxonomic revision is a contribution to the better understanding of the diversity of geophilomorph centipedes in the western Palaearctic. Indeed, the taxonomic and nomenclatural frame currently available is still largely affected by inconsistencies and redundancies, obviously complicating any effort to assess the actual biotic richness, as well as to analyse evolutionary and biogeographical patterns.

The circumscription of Stenotaenia proposed here is based on the recognition of the consistent occurrence of a well-defined combination of morphological characters, encompassing different parts of the body, including the mouth parts, the forcipules, the sternal glands, and the last leg-bearing segment ( Table 1; Fig. 1 ). Even though an adequate phylogenetic analysis would be hindered today by the inadequate level of knowledge of many other relevant geophilid taxa, we can say that the following shared characters are most probably synapomorphies of Stenotaenia : the shape of the labral margin (slightly projecting backwards into a medial, obtuse angle), the shape of the forcipules (the basal and the two intermediate articles relatively short, internal margins of the forcipules and the coxosternum without any tubercle), and the pattern of the sternal pores (arranged into a rounded to oval, medioposterior area on the anteriormost and posteriormost segments of the trunk, in two paired areas on the intermediate segments). This group of species is treated here as a distinct genus, as it can be unambiguously distinguished from all other known genera through a unique, strongly conserved combination of traits. This treatment is also consistent with the traditional, current practice in geophilid taxonomy, which assigns most of these traits a high diagnostic value at the genus level.

Out of the nominal taxa included here in Stenotaenia , some were already recognized as somehow related to each other. In particular, Verhoeff (1901, 1902–25, 1924a, 1934a, b) proposed the name Onychopodogaster , either as an informal taxonomic assemblage, or as a subgenus of Geophilus , to group abbreviatus , antecribellatus , asiaeminoris , cribelliger , fimbriatus , linearis , naxius , and palaestinus together. Attems (1929a) independently recognized the affinities of frenum , albanensis , and palaestinus , and assigned them to the poorly understood genus Simophilus . More recently, Minelli (1992) suspected strict affinities between linearis , abbreviatus , and romanus . Instead, the identity and relationships of other nominal species remained unclear or even misunderstood, as for acuneli , bosporanus , caldarium , giljarovi , and sturanyi . New genera or subgenera were introduced and maintained for some of these taxa ( Bithyniphilus , Insigniporus , Notadenophilus, Schizopleres , Simophilus , and Euronesogeophilus ), whereas the best-known species were included by most authors into already established genera, most often Geophilus and Clinopodes , but also to Nesogeophilus (which is a synonym of Tuoba ), Pachymerium , and even Himantarium . Actually, geophilid genera such as Geophilus , Clinopodes , Pachymerium , and Tuoba can be diagnosed unambiguously in respect to Stenotaenia ( Table 1; Fig. 1 ), whereas Himantarium is now regarded as belonging to another family ( Himantariidae ).

We recognized confidently nine genus-group nominal taxa and 27 species-group nominal taxa as belonging to Stenotaenia . Minor doubts remain only for Geophilus forficularius Fanzago, 1881 , because its original description is very incomplete, and we could not locate the type material (if still extant).

In addition, Geophilus ungviculatus Daday, 1889 , a nominal species described from Patras ( Greece), but still of uncertain identity ( Daday, 1889b; Attems, 1929a; Kanellis, 1959; Stoev, 1997; Zapparoli, 2002), could actually be referred to Stenotaenia , as suggested by features of the labrum, the forcipules, the sternum of the last leg-bearing segment, and the last telopodites. However, according to the original description, the arrangement of the sternal pore areas, and of the coxal pores, is unusual for a species of Stenotaenia . Therefore, we refrain from provisionally recognizing G. ungviculatus as belonging to this genus.

Two other species have been occasionally suggested to be closely related to G. linearis , but their original descriptions document unambiguously that they do not belong to Stenotaenia . Geophilus hartmeyeri Attems, 1911 , described from a few localities in Australia, was considered similar to G. linearis since its original description ( Attems, 1911), but it was later recognized as belonging to the genus Tuoba , as a synonym of either Tuoba laticeps (Pocock, 1891) or Tuoba sydneyensis (Pocock, 1891) ( Crabill, 1962, 1968; Jones, 1998). Geophilus aragonicus Daday, 1889 , described from the region of Aragon in the Iberian peninsula, was considered to be a junior synonym of G. linearis by Verhoeff (1931), but its identity remained uncertain; according to the original description ( Daday, 1889b) it clearly departs from Stenotaenia , as the sternal pores are arranged in transverse bands and only one pouch is present on each coxopleuron.

Outside the western Palaearctic, several, poorly known other species of Geophilus and allied genera have been described so far, mainly from North America. However, based on the original accounts, none of them may be recognized as belonging to Stenotaenia . As far as we know, only the northern American Mycotheres (Nemopleura) vittata Rafinesque, 1820 , currently known as Geophilus vittatus (Rafinesque, 1820) (with some synonyms; Hoffman & Crabill, 1953; Crabill, 1954), resembles Stenotaenia in some characters, such as the general shape of the forcipules and the arrangement of coxal pores. However, G. vittatus departs from Stenotaenia in other highly diagnostic traits, such as the arrangement of the sternal pores in transverse areas, and in the presence of other unusual features, e.g. dark dorsal patches on the trunk ( Crabill, 1954).

GEOGRAPHICAL DISTRIBUTION

The genus Stenotaenia is distributed within a limited part of the western Palaearctic ( Fig. 2 ). It is well documented in the central part of continental Europe, the Mediterranean part of mainland France, Corsica, Sardinia, the Italian Peninsula, Sicily, part of the African Atlas chain, the Balkan Peninsula, the Aegean islands, Crete, Anatolia, western Caucasus, and the coastal region of the eastern Mediterranean Sea. Towards the north, only scattered records are known from central and southern parts of Great Britain and the Baltic area, mainly from synanthropic sites, and therefore possibly result from recent introduction by humans. Towards the south-west, Stenotaenia is apparently absent from the whole of the Pyrenees, as well as from Iberia and the Balearic Islands, despite of the relatively high research efforts carried out in those areas ( Verhoeff, 1924b; Brolemann, 1930; Machado, 1946, 1953; Attems, 1952; Barace & Herrera, 1980; Salinas, 1990; Serra & Ascaso, 1990; Serra, Vicente & Mateos, 1996; García & Serra, 2000, 2003). It is worth noting that Attems (1926) cited Onychopodogaster generically for the Balearic Islands, but actually he almost surely was referring to Tuoba poseidonis ( Verhoeff, 1901) , a wellknown species that had been erroneously assigned to Onychopodogaster by Brölemann (1909b). Towards the east, Stenotaenia reaches the Carpathian range and the regions around the Black Sea, the most eastern records being from the eastern part of the Pontic area. No records are available from the internal regions of Ukraine and Russia, and from most of the Caucasus, but investigations in those areas have been very limited so far. Towards the south, Stenotaenia reaches Palestine, whereas there is no record for Cyprus.

SPECIES DIVERSITY

The species-level taxonomic arrangement proposed here is founded on a critical reassessment of all of the available information, as well as on comparative examination of representative specimens from throughout the range of the genus. However, it must be considered as only a preliminary arrangement, as the actual specific diversity within Stenotaenia remains understood only in part.

Some species are indeed very distinct and recognizable, as is the case of S. frenum , S. romana , and S. sturanyi , and we were able to contribute to the understanding of their morphology and geographical occurrence. Conversely, most species are treated here as valid only provisionally, pending more thorough investigations that could add to the understanding of their actual identity and their relations. This is particularly the case for S. antecribellata , S. asiaeminoris , S. bosporana , S. cribelliger , S. palpiger , and S. giljarovi .

Our revision of the actual diversity of Stenotaenia species was more effective for the western part of the range of the genus, approximately west of the Adriatic Sea. Here, extensive sampling and detailed published accounts allowed us to obtain a more satisfactory taxonomic, nomenclatural, and distributional synthesis. Major questions, however, remain: above all, regarding the taxonomic distinction and geographical limits between S. linearis and S. sorrentina , in a possible transitional area encompassing the western Alps and the nearby Mediterranean coastal regions. The arrangement proposed here is admittedly incomplete and provisional for the eastern part of the range of Stenotaenia , east of the Adriatic Sea, where sampling was less intense, where many nominal taxa were described inadequately, and where different authors followed different taxonomic opinions.

A thorough reassessment of the internal taxonomy of Stenotaenia will require testing the putative diagnostic value attributed by authors to some minor characters, such as the presence and the number of tubercles on the labrum and the size of the maxillary lappets, through the evaluation of intraspecific variation.

DIVERSITY OF BODY SIZE AND SEGMENTAL

STRUCTURE

Our morphological comparison revealed a consistent invariance in most of the anatomical traits in Stenotaenia , encompassing different parts of the body. However, despite the conservative body shape, species differ conspicuously both in the maximum size, reached by fully grown specimens, and in the number of trunk segments, which is invariant during postembryonic growth, as in all geophilomorphs. The extreme conditions are represented by S. romana , which is less than 2-cm long at full growth with as few as 43 leg-bearing segments in some specimens, and by S. sturanyi , reaching almost 8 cm in length, with up to 115 leg-bearing segments. As for the number of segments, in spite of a wide overall range of variation (72 segments), intraspecific variation is very limited, even in those species for which large numbers of specimens have been examined ( Table 3): ranges of variation observed for each sex were from six segments (in many dozen females of S. romana ) to 14 segments (in 15 males of S. frenum ; estimates for S. linearis and S. sorrentina are similar, but distinguishing between these two species is problematic).

Comparing the species of Stenotaenia , we found that the number of leg-bearing segments correlates significantly with the maximum head width ( Fig. 8 ) (Spearman correlation: for males R = 0.83, t = 10.8, P <0.000; for females R = 0.88, t = 14.0, P <0.000), revealing that the number of segments, which is fixed since hatching is associated with the maximum body size reached after growth.

The extent of variation of both body size and segment number in Stenotaenia is unusual in respect to all other recognized genus-level groups, not only within geophilomorph centipedes, but also within myriapods as a whole. As a rule, even in higher taxa where the segment number is very different between different genera, variation is very limited between closely related species included in a single genus. As far as is known, only two genera of geophilomorphs surpass Stenotaenia in the extent of variation of segment number, coupled with a comparable variation of body size, namely Henia Koch, 1847 , ranging from 43 leg-bearing segments and less than 2-cm long in Henia brevis ( Silvestri, 1896) to 153 leg-bearing segments and up to 15-cm long in Henia devia ( Koch, 1847) , and Ribautia Brölemann, 1909 , ranging from 43 leg-bearing segments and less than 2-cm long in Ribautia coarctata Ribaut, 1923 to 125 leg-bearing segments and up to 7.5-cm long in Ribautia taeniata Ribaut, 1923 ( Attems, 1929a) . However, each of these genera encompasses a more diversified assemblage of species, and the two extreme species of Henia cited above are actually separated in two distinct subgenera ( Minelli, 1982). The genus Notiphilides Latzel, 1880 has a range of variation in the segment number comparable with that of Stenotaenia (from 85 to 151 leg-bearing segments; Attems, 1929a), but this is not matched by a significant variation in body size. As for the relation between body size and segment number, preliminary analyses suggest that this may be common to other lineages and to geophilomorphs as a whole ( Foddai et al., 2003), but it has not been documented so far to the same extent as in Stenotaenia .

Therefore, interspecific comparison in Stenotaenia offers evidence of an extensive and concerted evolution of body size and segmental structure. This evolutionary pattern, together with the putative underlying developmental constraints, may contribute to the on-going research on the morphogenetic mechanisms, still inadequately understood, involved in the segmentation of the trunk of arthropods.

Figures

Figure 1. Main differential characters between Stenotaenia and related genera to which species of Stenotaenia have been assigned, to date. Drawings represent the type species of each genus (see Table 1): A, forcipular segment, ventral; B, sternum at about one third of the length of the trunk, ventral; C, last leg-bearing segment and terminal segments in the female, ventral. Abbreviations: ch, chitin line; co, condyle of forcipular coxosternum; cs, ‘carpophagus’ socket; cx, forcipular coxosternum; fb, forcipular basal article; fi, forcipular intermediate articles; ft, forcipular tarsungulum; go, gonopods; lc, claw of last leg; ls, sternum of last leg-bearing segment; op, openings of coxal glands; pc, forcipular pleurocoxal margin; po, sternal pore area.

Figure 2. Geographical distribution of Stenotaenia species. The previously published and new records have been plotted upon a grid of 30′ latitude ¥ 30′ longitude. The type localities of all nominal taxa are indicated by circles (full circles for valid species, empty circles for synonyms), with the following abbreviations: ab, abbreviatus; ac, acuneli; al, albanensis; an, antecribellatus; as, asiaeminoris; bo, bosporanus; br, brevicornis; c, caldarium; cr, cribelliger; fi, fimbriatus; ff, forficularius; fv, foveolatus; fr, frenum; gi, giljarovi; gr, graecus; li, linearis; na, naxius; or, ormanyensis; pl, palaestinus; po, polyporus; pp, palpiger; rh, rhodopensis; ro, romanus; sl, silvestrii; sm, simplex; so, sorrentinus; st, sturanyi.

Figure 8. Relationship between the adult body size, as measured by the maximum width of the head, and the number of leg-bearing segments in different species of Stenotaenia. We considered only specimens with fully developed gonopods, and only species for which at least five adults were available. Distinction between Stenotaenia sorrentina and Stenotaenia linearis is only tentative for some specimens, as it is based on a conventional threshold in the number of leg-bearing segments (see Table 3, and remarks under S. sorrentina).