Elops

Hoffmann, Matthias & Britz, Ralf, 2006, Ontogeny and homology of the neural complex of otophysan Ostariophysi, Zoological Journal of the Linnean Society 147 (3), pp. 301-330 : 320-321

Elops Oncorhynchus Chanos Puntius Rhabdalestes Apteronotus Silurus

1936) among nonteleostean actinopterygians. Similar cartilages are also known from Dipnoi (see Arratia & Schultze, 2001), but their homology has not been resolved.

It is surprising that no previous author has realized that supradorsals present a major part to the formation of the neural complex of Otophysi during development. Fink & Fink (1981: 325–326) doubted the presence of supradorsals altogether, stating that

Rosen & Greenwood (1970) and others have used the term ‘supradorsal’ for the dorsal part of the neural arch. However, supradorsal properly refers to a separate median cartilage dorsal to the neural arch (basidorsal) element in elasmobranchiomorphs and has been only tentatively applied to the paired cartilages in many actinopterygians ( Goodrich, 1958: 34). Since these paired cartilages appear not to be separate elements, but simply cartilage of the arches along their midline synchondral joint, the term supradorsal seems inappropriate.

A number of authors have followed Fink & Fink’s (1981) conclusion, such as, for example Coburn & Futey (1996) and Coburn & Chai (2003). de Pinna & Grande (2003: 843) added to the confusion with the following incorrect statement: ‘ Rosen & Greenwood (1970) homologized the claustrum with a dissociated dorsomedial portion of the first neural arch (which they called supradorsal, following François, 1966), an idea endorsed by Fink & Fink (1981) and apparently first proposed by Hora (1922).’ This statement contains two errors. Rosen & Greenwood (1970) considered supradorsals to be elements separate from the neural arches, and Fink & Fink (1981) specifically argued against this hypothesis, as noted above.

Fink & Fink’s (1981) statement refuting the presence of supradorsals in Ostariophysi consists of two parts: (1) the first deals with the homology of supradorsal cartilages of teleosts and those of elasmobranchiomorphs; (2) the second with the absence of supradorsals as autogenous anatomical entities. We will address both in sequence.

The term ‘supradorsal’ was created by Gadow & Abbott (1895: 171). These authors specifically mentioned Stannius (1849), Müller (1853), and Grassi (1883), who previously described the structures they named supradorsals in Esox , Salmo , and Coregonus . Thus, the term supradorsals was not specifically restricted to cartilaginous structures in elasmobranchiomorphs, but was equally applied to the paired cartilages at the dorsal tip of neural arches in a variety of osteichthyans. Fink & Fink (1981) are thus mistaken that supradorsals occur only in elasmobranchiomorphs.

Fink & Fink (1981) were apparently unaware of a number of previous papers clearly demonstrating that supradorsal cartilages are autogenous structures ( Stannius, 1849; Müller, 1853; Goette, 1879; Grassi, 1883; Scheel, 1893; Gadow & Abbott, 1895; Bloch, 1900; François, 1966). Our observations in our ostariophysan and outgroup developmental material confirm that supradorsals arise as separate cartilages after the neural arches have ossified, demonstrating unambiguously that they are autogenous cartilages and not part of the cartilage of the neural arch itself. A useful landmark to delimit supradorsals is their relationship to the dorsal longitudinal ligament that extends between the braincase and the caudal area. They are situated invariably on the medial side of the arches and below this ligament, as pointed out previously by a number of authors ( Stannius, 1849; Goette, 1879; Grassi, 1883; Scheel, 1893; Gadow & Abbott, 1895).

Having established the significance of supradorsals 3 and 4 for the formation of the neural complex in otophysans, we now turn to the different hypotheses concerning the homology of this structure and discuss them in light of our findings. We consider it essential to review all previous hypotheses, even if those are considered out of date now. This is especially important, because a large number of key papers were published in languages other than English and therefore often do not receive the attention they deserve in many of the recent papers on the Weberian apparatus.