GEOPHILOMORPHA

Koch, Markus & Edgecombe, Gregory D., 2012, The preoral chamber in geophilomorph centipedes: comparative morphology, phylogeny, and the evolution of centipede feeding structures, Zoological Journal of the Linnean Society 165 (1), pp. 1-62 : 52-54

publication ID	https://doi.org/ 10.1111/j.1096-3642.2011.00803.x
persistent identifier	https://treatment.plazi.org/id/8A1EA639-FFA5-5066-FED1-FBF05282B94B
treatment provided by	Marcus
scientific name	GEOPHILOMORPHA
status

Treatment

IN GEOPHILOMORPHA View in CoL View at ENA

Our character formulations pinpoint some of the uniquely derived characters of Geophilomorpha compared to other Chilopoda. Features of the preoral chamber can be added to the list of autapomorphies of Geophilomorpha , such as the presence of conical sensilla (rather than button-like sensilla) on the lips and distal tip of the hypopharynx (character 85), and a unique structure of the pharynx, which in all studied geophilomorphs has a double anchor cross-section with massive dilator muscles attaching to its sides (character 92). The standard depiction of geophilomorphs as sucking partially liquefied food has been based foremost on data for representative Adesmata. Our observations on the preoral chamber of Mecistocephalidae confirm the generality of these specializations across the order as a whole. In addition to the consistent absence of multicusped teeth on the mandibles, a strongly sclerotized pharynx with robust dilator muscles followed by an elongate, pipetteshaped oesophagus (see Koch et al., 2011), and extensive hypopharyngeal glands as the probable main production site for saliva and enzymes ( Manton, 1965) are also shared by Mecistocephalidae as well as Adesmata.

Another likely apomorphic character of Geophilomorpha is the reduction of the buccal glands to a single median pair (‘glandulae buccales mediales’ of Fahlander, 1938). This was demonstrated for species of Mecistocephalus (Mecistocephalidae) , Geophilus , and Pachymerium (Geophilidae) by Fahlander (1938), and for Orya barbarica (Oryidae) by Manton (1965). The presence of only the median pair of buccal glands is now confirmed by our histological studies for Dicellophilus carniolensis (Mecistocephalidae) and Hydroschendyla submarina (Schendylidae) ( Fig. 8 View Figure 8 ). Our SEM studies of the epipharynx furthermore revealed a single pair of openings of probably the median buccal glands in Himantarium and Haplophilus ( Fig. 9A View Figure 9 ) ( Himantariidae ), in Eucratonyx (Gonibregmatidae) , as well as in Ribautia and Zelanophilus ( Fig. 22E View Figure 22 ) ( Geophilidae ). Whether the entire absence of buccal glands in Strigamia hirsutipes (Linotaeniidae) ( Fahlander, 1938) is more widespread within Adesmata remains to be clarified.

MANDIBLES

The transformation of the mandibles in the Geophilomorpha is basically the continuation of a shortening and simplification of the mandibular gnathal edge that already started in the lineage comprising Pleurostigmophora as a whole. Previous considerations of mandible evolution in centipedes (e.g. Verhoeff, 1918 in Verhoeff, 1902 –1925; Attems, 1929; Manton, 1964, 1965) consistently refer to the state in Scolopendromorpha as the ‘basic type’ of mandible, from which all other states in non-scolopendromorph centipedes are derived. Based on recent SEM studies of the mandibular gnathal edge in centipedes (e.g. Edgecombe, Richter & Wilson, 2003; Edgecombe & Giribet, 2004) and our current insights, we refute this idea and revise the evolution of the mandibles in centipedes as follows ( Table 3). The original composition of the mandibular gnathal lobe is represented in Scutigeromorpha by (1) a pars molaris that forms the lamina condylifera bearing a hook-shaped condyle and contributing the proximalmost part of the gnathal edge in the form of an elongate molar plate ( Edgecombe et al., 2003: fig. 2C–H) surrounded by a large pulvillus (‘Haarpolster’; a ‘tuft of plumose hairs’ fide Minelli, 2011); and (2) a pars incisiva that provides, from proximal to distal, a second but short pulvillus, multicusped teeth (‘dentate lamina’ fide Minelli, 2011), and multiple pectinate lamellae (‘pectinate lamina’ fide Minelli, 2011) arranged in an imbricate series. Attems (1929: 4) already pointed out the significance of the presence of two pulvilli in scutigeromorphs in that only the shorter one in front of the proximalmost tooth is maintained in Pleurostigmophora. Edgecombe et al. (2003) and Edgecombe & Giribet (2004) favoured the view that the pulvillus in pleurostigmophorans is a remnant of the pars molaris in scutigeromorphs. We now endorse Attems’s interpretation and consider the loss of the molar plate and its surrounding pulvillus in Pleurostigmophora as the first step in the shortening of the gnathal edge in centipedes; a pars molaris is nevertheless maintained, forming a lamina condylifera that no longer contributes any armature to the gnathal edge ( Fig. 33 View Figure 33 ). Some uncertainty remains on whether the second, shorter pulvillus formerly belongs to the pars incisiva; in scutigeromorphs the second pulvillus is externally demarcated not only from the pars molaris, but also from remaining parts of the pars incisiva by a ‘membranous [i.e. non-sclerotized] band’ ( Edgecombe & Giribet, 2004: character 149, fig. 4A) and accordingly may rather represent an ‘intermediate area’ fide Edgecombe et al. (2003), as in millipedes. In pleurostigmophorans, the second pulvillus forms the proximalmost part of the gnathal edge, and varies in size (generally short in Lithobiomorpha , variably enlarged in Craterostigmus and some Scolopendromorpha ; see, e.g. Minelli, 2011: fig. 3.3) but is continuous with (fused to?) remaining parts of the gnathal edge. Changes of the multicusped teeth and pectinate lamellae within Pleurostigmophora were discussed by Edgecombe & Giribet (2004), including a potentially homoplastic increase in the number of multicusped teeth (under slow transformation) from three ( Scutigeromorpha , Craterostigmus ) to four/five in Lithobiomorpha and Scolopendromorpha ( Edgecombe & Giribet, 2004: character 128). In accordance with Attems’s interpretation of the pulvilli we now refute that the ‘expansion of the dentate lamina [...] is associated with a reduction in the size of the Haarpolster ’ ( Edgecombe & Giribet, 2004: 97). Interpreting a series of imbricated pectinate lamellae as general for Chilopoda, Pleurostigmophora, and Epimorpha, the shared presence of one or two rows of sickle-shaped aciculae in Lithobiomorpha and Craterostigmus ( Edgecombe & Giribet, 2004: character 151) may be a homoplastic transformation of the pectinate lamellae. In Geophilomorpha , the loss of proximal parts of the gnathal edge is continued in that both the second pulvillus and all multicusped teeth (‘dentate lamina’ fide Minelli, 2011) are no longer differentiated. The most obvious evidence for such increased shortening of the gnathal edge is the displacement of its distal part that in other centipede orders is primarily positioned below the hypopharynx but is shifted towards the mouth in geophilomorphs; the pectinate lamellae of geophilomorphs are accordingly no longer exposed in ventral view when the maxillae and maxillipedes are removed, but are ‘vertically situated in the pre-oral cavity’, as already pointed out by Manton (1965: 340).

The original syndetic joint between the mandibular base and gnathal lobe as commonly inferred for the myriapod ground pattern (e.g. Manton, 1964; Kluge, 1999: fig. 3B; Kraus, 2001; Edgecombe & Giribet, 2002; Koch, 2003) is most distinct in Scutigeromorpha , whereas pleurostigmophorans only maintain traces of this joint, if at all. In Lithobiomorpha and Craterostigmus the gnathal lobe is mostly fused to the mandibular base except for a fissure along the lamina condylifera that dorsally indicates the former position of the joint ( Edgecombe & Giribet, 2002: character 123). In Scolopendromorpha this fissure forms a part of cruciform sutures ( Edgecombe & Giribet, 2004: character 150) that subdivide the mandible into at least four sclerites (see Manton, 1965: fig. 64; Desbalmes, 1992: fig. 113B, Borucki, 1996: figs 1, 2, 4). The cruciform sutures were already considered as ‘secondary’ by Attems (1929), blurring the original joint between gnathal lobe and mandibular base, and are unambiguously optimized as a scolopendromorph autapomorphy in cladistic analysis (e.g. Edgecombe & Giribet, 2004). In geophilomorphs, the mandibular base is strongly shortened and mostly reduced to a rod ( Attems, 1929: fig. 121; Brolemann, 1930: fig. 194; Pereira & Coscarón, 1976: fig. 1J), thereby contrasting against the broader shovel-shaped gnathal lobe; a joint between the rod and gnathal lobe is usually indistinct or absent.

The new interpretation of the mandibular gnathal edge as only composed of pectinate lamellae throughout Geophilomorpha ( Fig. 33 View Figure 33 ) implies that geophilomorphs primarily lost the capability to bite, the mandibles only being capable of sweeping food into the mouth; the function of the pectinate lamellae alternatively may involve the formation of filter(s) that prevent large food parts entering the gut. A re-gain of sclerotized teeth is recognized as based on a transformation of the proximalmost pectinate lamella into a presumably biting or rasping organ (‘dentate lamella’ or ‘dentate lamina’ of previous authors). The recognition of a dorsal sclerotized lamella as an ancestral feature of Adesmata ( Bonato et al., 2003: character 45) is not supported by our cladistic analyses, which retrieve the gain of a sclerotized lamella as unique unambiguous support for a clade of Schendylidae , Ballophilidae , Himantariidae , Gonibregmatidae , and Oryiidae ( Fig. 32 View Figure 32 , character 17:1) within Adesmata, including a subsequent division of the formerly entire sclerotized lamella into groups of simple teeth within Schendylidae (character 18:1). This optimization, however, forces a reversal (loss of a dorsal sclerotized lamella) in Gonibregmatus and Oryidae on our cladogram topology and presently ignores a similar transformation in some Mecistocephalus species ( Fig. 5C View Figure 5 ), indicating that the (re-) gain of putative biting or rasping teeth by transformation of the proximalmost pectinate lamella underlies some degree of convergence within Geophilomorpha .

Further reversals of mandibular features within Adesmata concern the arrangement of pectinate lamellae in a single row (character 16), the transformation of the mandibular condyle into a ridge or hump (character 87), and the suppression of the delimitation of the lamina condylifera by a suture (character 88), each of which is forced to reverse by our optimal cladogram topologies, irrespective of weighting regime, in such taxa as himantariids and schendylids. The homology hypotheses incorporated in our codings consider the arrangement of pectinate lamellae in an imbricate series, the presence of a hook-shaped condyle on the gnathal lobe, and the delimitation of the sclerite bearing this condyle as plesiomorphic states within Geophilomorpha (based on states in non-geophilomorph chilopods) and account for a repeated simplification of the mandibular gnathal edge within Adesmata. These latter character polarities imply a substantial degree of homoplasy in mandibular evolution within Geophilomorpha .