The phyletic position of Dinogeophilus
The anatomical and molecular data presented here concur in providing compelling evidence that Dinogeophilus originated within the Schendylidae rather than within the Geophilidae as previously thought. The phyletic position of Dinogeophilus has been hitherto remarkably misunderstood, because the two families are only very distantly related according to the consensus phylogeny of the Geophilomorpha derived from modern analyses of molecular and morphological data (Edgecombe and Giribet, 2007; Murienne et al., 2010; Bonato et al., 2014): after the basal emergence of the Mecistocephalidae, the separation between Geophiloidea (including Geophilidae) and Himantarioidea (including Schendylidae) was most probably the deepest split within the remaining geophilomorphs (Adesmata) (Bonato et al., 2014), and this separation has been estimated to date back between 350 and 250 million years ago (Murienne et al., 2010).
The long-lasting misconception appears even more remarkable if we consider that Silvestri (1909a) hypothesized that Dinogeophilus could be close to Apogeophilus Silvestri, 1905, which is undisputedly a geophilid, although still very poorly understood. Silvestri’s hypothesis was eventually dismissed, because it was based on an erroneous interpretation of the articular structure of the legs of the ultimate pair in Dinogeophilus (Pereira, 1984). While most authors maintained Dinogeophilus in the Geophilidae, a very different arrangement was elaborated by Verhoeff (1925), who included Dinogeophilus in a very heterogeneous subfamily Chaetechelyninae inside a family Scolioplanidae, together with other genera currently recognised in different lineages of Geophilidae. Verhoeff’s proposal was based on erroneous interpretations of the structure of the labrum that are now largely superseded (Koch and Edgecombe, 2012).
Many of the morphological characters supporting the evolutionary emergence of Dinogeophilus within the Schendylidae had been already described and illustrated correctly, at least for D. oligopodus (Pereira, 1984): denticle-like projections on the labral margin, branching pretarsi on the second maxillae, longitudinally elongate pore-fields, additional minute accessory spines on the leg pretarsi, peculiarly modified legs of the ultimate pair, uni-articulate and separate gonopods in the female. However, some of these characters have been hitherto underestimated for their value in inferring phylogeny and driving classification, or even fully ignored (especially the structure of the female gonopods; Bonato et al., 2014). Conversely, a major role in perpetuating the misplacement of Dinogeophilus within the Geophilidae has been played most probably by the primary diagnostic value so far attributed to the shape of the mandibles, especially to the number and diversity of the lamellae (e.g., Attems, 1929). Our observations confirm previous reports (Silvestri, 1909a; Pereira, 1984) that the mandibles of Dinogeophilus resemble more the unilamellate mandibles of the Geophilidae than the bilamellate mandibles of the Schendylidae. Therefore, we should hypothesize that the unilamellate mandible found in Dinogeophilus derives from the bilamellate mandible of Schendylidae, independently from the similarly unilamellate mandible of all Geophiloidea, which instead derives from a multilamellate mandible (Bonato et al., 2014).