Mecistocephalus guildingii Newport , 1843

Bonato, Lucio, Cupul-Magaña, Fabio G. & Minelli, Alessandro, 2009, Mecistocephalus guildingii Newport, 1843, a tropical centipede with amphi-Atlantic distribution (Chilopoda: Geophilomorpha), Zootaxa 2271 (1), pp. 27-42 : 28-35

Mecistocephalus guildingii Newport, 1843

= Mecistocephalus punctilabratus Newport, 1845 (new synonymy)

= Lamnonyx leonensis Cook, 1896 (new synonymy)

= Mecistocephalus janeirensis Verhoeff, 1937

= Mecistocephalus maxillaris guadeloupensis Demange and Pereira, 1985 (new synonymy)

Full lists of citations

Mecistocephalus Guildingii Newport 1843a: 179 ; 1843b: 500; 1845: 429, figs 18–19 of pl. XXXIII; 1856: 82. Meinert 1870: 96. Saussure and Humbert 1872: 206. Pocock 1893a: 123; 1893b: 470. Latzel 1895: 101.

Mecistocephalus punctilabratus Newport 1845: 302 .

Geophilus Guildingii. Gervais 1847: 311 .

Mecistocephalus punctifrons (partim). Meinert 1886: 213. Haase 1887: 104. Bollman 1888: 337. Pocock 1891: 423.? Brölemann 1897: 136; 1898: 254. Chamberlin 1914a: 210. Bücherl 1942a: 356. Foddai et al. 2000: 65.

Mecistocephalus guildingii . Bollman 1887: 82; 1889: 128. Pocock 1899: 63. Attems 1929: 156. Crabill 1959: 189, figs 1–6. Demange and Pereira 1985: 182. Lewis 1986: 24. Lewis and Wranik 1990: 69. Pereira et al. 1997: 80. Foddai et al. 2000: 63. Bonato et al. 2003: 543. Bonato and Minelli 2007: 166. Barber 2008: 39; 2009: 46, fig. 75.

Mecistocephalus Guildingi [sic!]. Brölemann 1896: 528.

Lamnonyx leonensis Cook 1896a: 39 ; 1896b: 61; 1896c: 74. Attems 1903: 210.

Mecistocephalus guildingi [sic!]. Kraepelin 1901: 201. Attems 1903: 209. Chamberlin 1913: 122; 1914b: 85; 1918: 164. Verhoeff 1902 –25: 697. Bücherl 1942b: 209. Negrea et al. 1973: 237. Matic et al. 1977: 286. Negrea 1977: 304.

Lamnonyx maxillaris (partim). Silvestri 1919: 61.

Mecistocephalus leonensis . Attems 1929: 141.

Mecistocephalus maxillaris (partim). Brolemann 1930: 80, figs 61–66. Chamberlin 1920a: 185; 1920b: 278; 1924: 42; 1930: 67. Wang 1951: 46; 1962: 89.? Chamberlin 1958: 14. Jeekel 1964: 116. Schubart 1967: 45. Demange and Pereira 1985: 182. Foddai et al. 2000: 64.

Mecistocephalus janeirensis Verhoeff 1937: 230 ; 1938: 382; 1939: 84. Bücherl 1942a: 356; 1942b: 164. Attems 1947: 101.

Mecistocephalus maxillaris guadeloupensis Demange and Pereira 1985: 195 , figs 59–70. Pereira et al. 1997: 80. Foddai et al. 2000: 65.

Mecistocephalus insularis (partim).? Ashmole and Ashmole 2000: 327.

Description. A Mecistocephalus species (for a diagnosis of the genus, see: Bonato and Minelli 2004; Uliana et al. 2007) with body length up to ca. 4 cm long. Colour (fig. 1) yellow, without dark pigmentation on the trunk; head and most anterior trunk segments reddish brown. Cephalic capsule ca. 1.9–2.0 times as long as wide. Intermediate part of the frontal line slightly concave forward or almost straight. Antennae ca. 5 times as long as the head width. Antennal apical sensilla ca. 10 µm long, with a flat crown-like projection at ca. midlength. Club-like sensilla on the external sides of the ca. 8 most distal antennal articles, and on the internal sides of the ca. 4–6 most distal articles. Areolate part of the clypeus (fig. 2) ca. 2 times as long as the plagulae, without non-areolate insulae; 3 pairs of setae in the areolate clypeus, arranged in an almost transverse line, and a pair of setae just anterior to the plagulae and close to the mid-line; plagulae without dense micropores, and with anterior margins rounded. Buccae with setae on the posterior half only. Posterior margin of labral side-pieces entire, without any conspicuous medial projection; internal margin of the anterior ala shorter than the internal margin of the posterior ala. Mandible with ca. 6–7 pectinate lamellae, the first lamella bearing ca. 5–7 teeth, an average intermediate lamella bearing ca. 10–12 teeth, which are similar to each other. Coxosternum of maxillae I with poorly projecting antero-lateral corners, and with a distinct incision on each lateral margin. Coxosternum of maxillae II without non-areolate insulae; telopodites slender; claw uniformly tapering into a pointed tip. Forcipular coxosternum (fig. 3) 1.1–1.2 times as wide as long, without condylar processes; cerrus absent. Scapular points short. Forcipular article I ca. 1.8–2.1 times as long as wide, with a small proximal tubercle and a small distal tubercle, similar to each other; intermediate articles each with a small tubercle; two smaller tubercles at the basis of the tarsungulum, one dorsal to the other. Poison calyx reaching the distal part of article I. Sterna of leg-bearing segments of the anterior half of the trunk with furcate mid-longitudinal sulcus, the angle between the two branches approximately rectangular in a few most anterior sterna and gradually more obtuse to almost flat in the following sterna. Invariantly 49 leg-bearing segments. Tergum of the last leg-bearing segment subrectangular, ca. 1.2–1.4 times as long as wide. Sternum of the last leg-bearing segment (fig. 4) subtrapezoidal, slightly wider than long, with a short pillowlike posterior process, without notches on the lateral margins. Legs of the last pair with a single apical spine.

Differential diagnostic features between M. guildingii and other similar species, all with 49 pairs of legs, are given in Tab. 1 .

Taxonomical remarks. Mecistocephalus guildingii was originally described by Newport (1843a). Indeed, two identical versions of the paper containing the original description were issued ( Newport 1843a, 1843b), but Newport (1843a) was published earlier and should be taken as providing the first valid description of the species ( Crabill 1957). The year of the original description has been often erroneously reported as 1842 or 1844 or even 1845, but the true publication dates of the relevant papers by G. Newport have been clarified definitively by Crabill (1957).

M. guildingii was described upon five specimens from the island of Saint Vincent in the Lesser Antilles. The syntypes were collected by L. Guilding and were found by G. Newport in the collection of F.W. Hope ( Newport 1843a, 1843b, 1845). They are presently preserved in the Hope collections (“Newport myriapod collection”) at the Oxford University Museum of Natural History .

Other specimens were identified as M. guildingii by Meinert (1870), Bollman (1887, 1889), Pocock (1893b), Latzel (1895), and Chamberlin (1918), but most of these authors subsequently synonymised M. guildingii under either M. punctifrons Newport, 1843 ( Meinert 1886; Haase 1887; Bollman 1888; Brölemann 1898; Bücherl 1942a) or M. maxillaris ( Gervais, 1837) ( Silvestri 1919; Chamberlin 1920a, 1920b; Brolemann 1930; Jeekel 1964). However, these synonomies were based on the uncertain or misunderstood identity of M. punctifrons and M. maxillaris at that time, as later clarified ( Crabill 1970; Bonato and Minelli 2004; Bonato et al. 2004). Actually, M. punctifrons is distinct from M. guildingii in many morphological characters (tab. 1) and, despite many putative records from most tropical regions throughout the world, it has been reported reliably only from the Indian peninsula ( Bonato and Minelli 2004). Instead, the true identity of M. maxillaris still remains unknown because it was described upon specimens introduced in Paris , the original description is very vague, and the original material is most probably lost. Anyway, relying on the number of legs given by Gervais (1837) (46 pairs, most probably excluding the last pair, which is differently shaped), the possibility that M. maxillaris is identical to M. guildingii can be confidently ruled out, as M. guildingii has invariantly 49 pairs of legs. Even the species traditionally named M. maxillaris after Silvestri (1919), and inhabiting at least the Hawaiian islands (Bonato et al. 2004), is distinct from M. guildingii in some morphological characters (tab. 1); moreover, its putative occurrence in the Americas, Africa and the Indian Peninsula has not been confirmed by recent investigations ( Bonato and Minelli 2004). Worth noting is that Crabill (1959) did not find any specimen referable to either M. punctifrons or M. maxillaris among the specimens that he examined from the Americas.

Out of the other known species of Mecistocephalus , M. guildingii seems very close to Mecistocephalus insularis ( Lucas, 1863) , as already highlighted by Crabill (1959). M. insularis was originally described from Réunion Island and was putatively reported from many tropical regions throughout the world, but recent investigations suggest that it may be actually limited to the western part of the Indian Ocean ( Lewis 1986, 1996; Lewis and Wranick 1990; Bonato and Minelli 2004; Bonato et al. 2004). Main differential characters between M. guildingii and M. insularis are in body length (no more than 4 cm in M. guildingii vs. up to 9 cm in M. insularis ), colour (always without dark patches in M. guildingii vs. usually patched in M. insularis ), clypeal insulae (invariantly absent in M. guildingii vs. sometimes present although inconspicuous in M. insularis ), and the relative extension of the clypeal plagulae (distinctly shorter than the areolate clypeus in M. guildingii vs. about as long as the areolate clypeus in M. insularis ).

The name Mecistocephalus punctilabratus appeared only in the legend to the figures actually illustrating Mecistocephalus guildingii in Newport’s (1845) monograph on chilopods, obviously in place of the name M. guildingii , which is explicitly used as the valid name of the species in the text of that publication (cf. page 429, figures 18–19 of plate XXXIII, and page 302). At the best of our knowledge, the name punctilabratus was never used or cited again in the literature. Despite the fact that Mecistocephalus punctilabratus was most probably introduced inadvertently, it should be construed as an available species-group name, as it satisfies all requirements of Art. 11 of the International Code of Zoological Nomenclature (ICZN 1999), and was published before 1931 accompanied by an illustration that serves as an indication of the taxon that it denotes (Art. 12.1; Art. 12.2.7). Mecistocephalus punctilabratus Newport, 1845 is a junior objective synonym of Mecistocephalus guildingii Newport, 1843 , as the specimen illustrated was one of the syntypes of Mecistocephalus guildingii .

Lamnonyx leonensis was described upon a series of specimens, 20–35 mm long, from Sierra Leone and Liberia ( Cook 1896a, 1896b, 1896c). The name was made available by Cook (1896a) by providing a brief morphological description of the species, even though not citing it explicitly as a new taxon. Indeed, O.F. Cook cited the species name in two other papers published in the same year ( Cook 1896b, 1896c) but without providing any diagnostic information. Probably referring to those papers only, Attems (1903) erroneously stated that the nominal species L. leonensis lacked any diagnosis. As the original description was poorly detailed and not accompanied by illustrations, and no supplementary accounts were provided subsequently, the identity of the species remained unknown and no other specimens were referred to it. After that, Silvestri (1919) redescribed M. maxillaris upon a composite series of specimens, most probably not conspecific, from different localities throughout the world tropics (see Bonato et al. 2004), L. leonensis was often considered a junior synonym of M. maxillaris ( Silvestri 1919; Chamberlin 1920a; Wang 1951, 1962; Foddai et al. 2000), or otherwise provisionally retained as a valid species of uncertain identity ( Attems 1929). By examining three syntypes of Lamnonyx leonensis Cook, 1896 preserved in the National Museum of Natural History , Smithsonian Institution, Washington (an adult female 38 mm long, an adult male 33 mm long, and a subadult male 24 mm long; mounted on microscopic slides, with the head detached from the trunk; all collected in “ Sierra Leone ” in “ XII.1893 ” according to the labels pasted on the slides; figs 5–8), we found that these syntypes are fully consistent with M. guildingii in all characters recognised here as diagnostic of the species (including elongation of the head, pattern of clypeal areolation and setae, arrangement of setae on the buccae, shape and size of tubercles of the forcipular segment, aspect of the sternal sulcus, shape of the sternum of the last leg-bearing segment). Therefore, L. leonensis is considered here confidently identical to M. guildingii .

Mecistocephalus janeirensis was described upon a single specimen, 26 mm long, of unknown sex, from Brazil, probably from near Rio de Janeiro ( Verhoeff 1937, 1938). The name was first introduced validly by Verhoeff (1937) in a key, even though it was explicitly described as a new taxon only in a later paper ( Verhoeff 1938). The species was rarely cited in the literature, and no other specimens were referred to it. After examining specimens from Rio de Janeiro confidently representative of M. janeirensis, Crabill (1959) recognised it as identical to M. guildingii . The synonymy was not questioned by subsequent authors.

Mecistocephalus maxillaris guadeloupensis was described upon five specimens, of both sexes, up to 39 mm long, from the islands of Guadeloupe in the Lesser Antilles ( Demange and Pereira 1985). No other specimens were subsequently referred to it, but its taxonomic status was never discussed, and it was registered as such in subsequent catalogues ( Pereira et al. 1997; Foddai et al. 2000). Based on the detailed description and illustration of the female holotype and a representative male provided by Demange and Pereira (1985), we found that M. maxillaris guadeloupensis is fully consistent with M. guildingii in all the characters described here as diagnostic of the species. Therefore, M. maxillaris guadeloupensis is confidently considered here identical to M. guildingii . Worth noting is that, in describing M. maxillaris guadeloupensis, Demange and Pereira (1985) explicitly contrasted the new taxon with M. maxillaris and M. insularis , but quite surprisingly they did not discuss its status with respect to M. guildingii , which they considered a valid, distinct species.

Detailed descriptions and valuable illustrations of representative specimens of M. guildingii are provided by Brolemann (1930: 80, figs 61–66; as M. maxillaris ), Crabill (1959: 189, figs 1–6; as M. guildingii ), and Demange and Pereira (1985: 195, figs 59–70; as M. maxillaris guadeloupensis ). Conversely, other published accounts are erroneous or at least misleading in some detail: the head of the holotype of M. janeirensis was described as 2.5 times as long as wide by Verhoeff (1937), but such a figure appears most probably overestimated, as it is higher than ever observed to date in any other centipede, including Mecistocephalus species ; a single basal tooth was detected on the forcipular tarsungula by previous authors ( Meinert 1870; Crabill 1959; Demange and Pereira 1985) but this was most probably due to subjective interpretation, as the presence of a couple of shallow prominences in many Mecistocephalus species has been often overlooked, as demonstrated by Crabill (1970). Some slight sexual dimorphism was repeatedly claimed in the literature: the sternum of the last leg-bearing segment and the coxopleura could be more elongate in females than in males ( Crabill 1959; Demange and Pereira 1985), and the ultimate legs could be slightly swollen and covered with shorter setae in males ( Meinert 1870). However, our observations failed to provide evidence in favour of this.

Geographical distribution. After examining new material and re-evaluating all published findings of Mecistocephalus specimens from the Americas and Africa, we ascertained reliable records of M. guildingii from a total 26 localities, based on specimens identified directly by us or by other reliable authors, as well as on specimens that are unambiguously recognisable as M. guildingii according to the published accounts even though originally misidentified in the literature. Other published records of Mecistocephalus species from another 14 localities can be assigned to M. guildingii with high probability, based on some evidence. Other records from 6 additional localities could be tentatively assigned to this species. All these localities are listed in the Appendix and mapped in fig. 9.

Records of M. guildingii have been so infrequent that it is not possible to distinguish confidently between established populations and occasionally translocated specimens. If we take repeated findings in years, or findings of multiple specimens together, as suggestive of the stable presence of the species in a locality or area, we can guess that established populations of M. guildingii occur at least along the African coasts from Gambia to Liberia ( Cook 1896a; new records), in the Cape Verde archipelago ( Silvestri 1919; new records), in the Bermuda islands ( Bollman 1889; Chamberlin 1920b; new records), in coastal sites in Florida ( Chamberlin 1958; Crabill 1959), in some islands in the Lesser Antilles ( Newport 1843a; Crabill 1959; Demange and Pereira 1985), in at least a locality on the Pacific coast of Mexico (new records), and possibly also in Jamaica ( Pocock 1893b; Chamberlin 1918) and near Rio de Janeiro ( Verhoeff 1937; Crabill 1959).

Other reliable records of M. guildingii , however hitherto only occasional, are from Madeira ( Brölemann 1897; Silvestri 1919), Greater Antilles including Cuba and Hispaniola ( Bollman 1887; Chamberlin 1918; Silvestri 1919), other islands in the Lesser Antilles ( Meinert 1870; Chamberlin 1924; Crabill 1959), Atlantic coastal sites and nearby islands of the Americas from Panama to Brazil ( Brölemann 1898; Chamberlin 1914a, 1930; Crabill 1959), and some inland localities in Brazil ( Chamberlin 1914a; Silvestri 1919; Bücherl 1942a). Moreover, specimens have been occasionally found in hothouses or gardens in European cities ( Latzel 1895; Brolemann 1930; Barber 2008).

Almost all records of M. guildingii are from Atlantic islands or coastal regions of both the American and African mainlands. Outside of this amphi-Atlantic core area, the species has been found only in a single site on the Pacific coast of Mexico, occasionally in European cities, and possibly in a few inland sites in Brazil.

Environmental conditions are unknown for most of the collection sites, thus preventing any general analysis of the species’ habitat. However, records from urban or human-disturbed sites as well as from natural or semi-natural grounds come from throughout the geographical range documented here in the Americas and Africa. Conversely, the few records from Europe are strictly from urban sites. Worth notice is that specimens have been found in Puerto Vallarta, Mexico, both in the urban area and in the tropical forests surrounding the city.

Up to now, M. guildingii had been recorded from the Americas only, in particular from islands and coastal sites on the Atlantic side ( Crabill 1959), but for the occasional findings from European cities. Actually, the geographical distribution of M. guildingii is significantly wider than previously thought, including the most western part of the northern sub-tropical Africa, and therefore spreading between the two sides of the Atlantic Ocean in its tropical portion.

West of the area inhabited by M. guildingii , many islands in the Pacific Ocean are colonised by different species of Mecistocephalus . Among these islands, the closest to the American continent is Clipperton Island, which lies ca. 1200 km west of the coast of Mexico. Clipperton is inhabited by a species of Mecistocephalus which is easily distinguished from M. guildingii by its less elongate head (ca. 1.6 times as long as wide vs. ca. 2.0 times); it was described by Chamberlin (1914b) as a distinct species Mecistocephalus parvus Chamberlin, 1914 . Correspondingly, east of the area inhabited by M. guildingii , most of tropical Africa is colonised by a species of Mecistocephalus that is easily distinguished by M. guildingii , above all for the less elongated head and the presence of clypeal insulae; it has been identified by different authors under different names, including M. punctifrons and M. insularis .

While other species of Mecistocephalus are therefore surely present in areas contiguous to those inhabited by M. guildingii , it is not clear whether other species occur within the amphi-Atlantic core area of M. guildingii . Indeed, besides M. guildingii and the taxa here synonymised under it, three other nominal species have been reported to date from the Americas and some Atlantic islands, namely M. maxillaris ( Gervais, 1837) , M. punctifrons Newport, 1843 , and Mecistocephalus insularis ( Lucas, 1863) . However, published records of M. maxillaris and M. punctifrons from the Americas, Madeira and Cape Verde islands have been demonstrated or are strongly suspected to be due to misidentification of M. guildingii (see under “Taxonomic remarks”). This also could be the case for the vague citation of the species from the Canary Islands by Silvestri (1919), even though apparently not based on his own examination. Instead, M. insularis has been reported only once from Ascension Island ( Duffey 1964; Ashmole and Ashmole 2000), but the true identity of the specimens needs confirmation because M. guildingii can be easily misidentified with M. insularis (see under “Taxonomic remarks”). Both Silvestri (1919) and Crabill (1959) referred all the specimens of Mecistocephalus they examined from the Americas to a single species, and explicitly suspected that a single species inhabited the Americas, even though they applied different names to it, L. maxillaris and M. guildingii respectively. Our revisionary work is consistent with this view, even though we cannot rule out the possibility that other species of Mecistocephalus may have been established after introduction within the amphi-Atlantic tropical range of M. guildingii .