MANDIBULOHYOID CONNECTION IN CIRRHITOID FISHES 



99 



rence of a mandibulohyoid linkage in bony fishes (see Tcher- 

 navin, 1953, and references cited by Verraes, 1977, Springer 

 et al., 1877, Lauder & Liem, 1980, and above) certainly 

 seems to support the views of functional anatomists with 

 regard to its involvement in the mechanics of jaw opening. It 

 also refutes the apparently widespread view (see reviews in 

 Lauder & Liem, 1980; Lauder, 1982) that the linkage may be 

 a primitive character of neopterygian fishes, one lost in 

 higher teleosts (but see also Lauder & Liem, 1989, for later 

 views). However, although the mandibulohyoid connection 

 may be functionally homologous in both 'higher' and 'lower' 

 bony fishes, there are indications that it may not be homolo- 

 gous in an ontogenetical and hence phylogenetic context (see 

 below). Nevertheless, the diversity of mandibulohyoid con- 

 nections already known in but a few teleost fishes strongly 

 suggests that the structural, functional and ontogenetic 

 aspects of this system need to be reevaluated. 



Any attempt to establish or refute the homology of man- 

 dibulohyoid connections in cirrhitoid fishes with those in 

 other bony fish groups (see below) is hampered by a lack of 

 information on the ontogeny of the linkage in the various taxa 

 involved. Indeed, this problem also arises with the different 

 mandibulohyoid linkages found within the cirrhitoids them- 

 selves, namely those in the Cirrhitidae (p. 93) and that in the 

 latrid Acantholatris monodactylus (p. 97). 



The cirrhitid linkage type in the Paracirrhites species 

 examined (p. 93) strongly suggests that the connection 

 between the mandible and the ceratohyal in these fishes is 

 derived from an extension of the central aponeurosis of the 

 adductor mandibulae muscle's Aw portion onto the hyoid 

 arch (with, in addition, a partial insertion on the quadrate; 

 Fig. 1C and p. 93). In another cirrhitid group (viz. Cyp- 

 rinocirrhites polyactis, Cirrhitichthys oxycephalus and Cirrhi- 

 tops fasciatus) the connection also has a linkage with the 

 aponeurosis of adductor mandibulae Aw. Here it is effected, 

 somewhat indirectly, by a branch from the major mandibulo- 

 hyoid connection joining the maxillary tendon of adductor 

 mandibulae A, muscle, which tendon itself is derived from 

 the aponeurosis of the Aw portion of that muscle. This 

 association with the Aw aponeurosis in both cirrhitid groups 

 raises the possibility that ontogenetically, the mandibulohy- 

 oid linkage is through a tendon and not a ligament as it 

 appears to be in the salmonid Oncorhynchus mykiss (see 

 Verraes, 1977). It also raises the question whether or not the 

 so-called mandibulohyoid ligament (see below) in other 

 teleosts (and in the semionotiform Lepisosteidae; see below) 

 is truly a ligament. A similar problem arises with the third 

 type of mandibulohyoid connection found in cirrhitoids, 

 namely that in the latrid Acantholatris monodactylus. Here 

 the linkage is not associated with the Aw muscle, and has 

 both its origin and its insertion entirely on bone, thus 

 appearing to be a true ligament. 



There is some indirect support for the idea that in members 

 of the Cirrhitidae the mandibulohyoid connection could be 

 derived ontogenetically from the adductor mandibulae 

 muscle bloc (sensu Edgeworth, 1935) of the early embryo. 

 This stems from the considerable posterior extension of the 

 adductor mandibulae Aw aponeurosis onto the bones of both 

 the palatoquadrate arch and the interoperculum in certain 

 other perciform fishes (see also discussions in Winterbottom, 

 1974; Elshoud-Oldenhave & Osse, 1976; Anker, 1978;Green- 

 wood, 1985) and, indeed in other cirrhitoids such as the 

 cheilodactylids. 



An origin of the mandibulohyoid connection from the 



adductor mandibulae Aw tendon system seems less likely in 

 the latrid Acantholatris monodactylus. Here the linkage 

 extends from the posterior tip of the dentary's lower arm 

 (not, as in the cirrhitids, from its coronoid process or the 

 anguloarticular) to the upper part of the ceratohyal's lateral 

 face (Fig. 3). At no point has this apparent ligament in 

 Acantholatris any association with the adductor Aw muscle or 

 any part of its tendon system. With regard to its attachment 

 to the lower aspect of the dentary, the connection is compa- 

 rable both with the loosely compacted and fibrous linkage 

 between the dentary and ceratohyal identified by Aerts et al. 

 (1987) in the cichlid Astatotilapia elegans, and with Osse's 

 (1969) ligament XXIV in the percid Perca fluviatilis . In both 

 these species, however, the tissue has insertions on certain 

 branchiostegal rays as well as on the ceratohyal, and in 

 neither species does it have the ligament-like appearance of 

 the connection in Acantholatris monodactylus. 



Aerts et al. (1987:97) describe in some detail the histology 

 of the hyoid-dentary connection in Astatotilapia elegans, 

 which seemingly is derived from the anterior, tendinous part 

 of the geniohyoideus muscle, with whose dorsolateral aspect 

 it is closely associated over much of its length. These authors 

 conclude (op. cit.: 99) that 'In fact, the rostral part of the 

 interconnection can be interpreted as a parallel elastic com- 

 ponent of the protractor hyoidei' (=geniohyoideus). The 

 posterior attachment of the connection is on the epi- and 

 ceratohyals dorsally, with, as noted above, a number of small 

 strands merging into the dermal layers of the skin-fold 

 between the hyoid and interoperculum. A mandibulohyoid 

 connection, superficially like that in A. elegans also occurs 

 (pers.obs.) in another haplochromine cichlid, Thoracochro- 

 mis buy si (Penrith); although its histology was not studied, 

 the linkage appears to originate from within the geniohyoid 

 muscle, and to attach to the hyoid arch at the epi-ceratohyal 

 suture. 



At least with regard to its superficial features, Aerts et a/.'s 

 description of the dentary-hyoid connection in Astatotilapia 

 elegans does not resemble the condition seen in Acantholatris 

 monodactylus. Here, the interconnecting tissue is clearly 

 separated from the geniohyoideus muscle over virtually its 

 entire length, and is much more compact and ligament-like. 

 However, posteriorly it does appear to fuse with the tendi- 

 nous insertion of the geniohyoideus at the point where both 

 elements attach to an elevation on the anterior margin of the 

 certohyal. The insertion of the geniohyoideus muscle then 

 extends down along the lateral face of the ceratohyal, but that 

 of the mandibulohyoid connection does not. Thus in adult 

 Acantholatris monodactylus the only suggestion of the con- 

 nection being derived from the geniohyoideus muscle is a 

 partially shared insertion with that muscle on the ceratohyal. 

 That suggestion is, unquestionably, far less convincing than 

 the evidence provided by the situation in Astatotilapia 

 elegans, but is one that could be clarified if studied ontoge- 

 netically in Acantholatris monodactylus. 



A distinct mandibulohyoid ligament, superficially like that 

 in Acantholatris monodactylus, has been described by Wiley 

 (1976) in the semionotiform gars Lepisosteus and Atrac- 

 tosteus. The connection is labeled as a tendon in figure 9 of 

 Wiley's paper, but is referred to, I believe correctly, as a 

 ligament in the accompanying text. The ligament in gars 

 differs from the ligament-like mandibulohyoid connection in 

 Acantholatris monodactylus in its points of attachment (epi- 

 hyal and retroarticular in the gars, ceratohyal and dentary in 

 A. monodactylus). Again, without ontogenetic information 



