EPAXIAL MUSCLES AND ACANTHOMORPH RELATIONSHIPS 



133 



Fig. 14 Epaxial musculature of blennioids: a, Tripterygiidae, 

 Enneanectes pectoralis (MPM 22463, 26.5 mm SL), insertions to 

 ninth pterygiophore; b, Chaenopsidae, Acanlhemblemaria 

 greenfieldi (MPM 24876, 30.4 mm SL), insertions to 13th 

 pterygiophore; c, Blenniidae, Entomacrodus nigricans (MPM 

 18256, 55.4 mm SL), insertions to 11th pterygiophore; d, 

 Labrisomidae, Labrisomus bucciferus (MPM 31163, 57.0 mm SL), 

 insertions to 13th pterygiophores. F, fan-shaped anterior slip of 

 epaxial to skull; other abbreviations and methods of presentation 

 as in Figs 1, 3. Scale bars = 1 mm (a,b), 5 mm (c,d). 



Ion has an almost completely separate series of muscle fibres 

 that insert on to the third to ninth pterygiophores (Fig. 5). 

 Type 1 appears to be the primitive condition for the cirrhi- 

 toids (Fig. 17), with a secondary change to an epaxial/ 

 pterygiophore association resembling more closely a Type 2 

 morphology among some cirrhitids, which could be indicative 

 of close relationship (Table 1). Among sciaenids both epaxial 

 muscle Types and 2 occur, although their distributions are 

 difficult to interpret with our current understanding of sci- 

 aenid relationships (Table 1; Sasaki, 1989). Within scor- 

 paenoids there is variation in epaxial morphology among the 

 higher taxa. More extensive surveys within these and other 

 groups with epaxial/pterygiophore insertions could help to 

 elucidate some of their intrarelationships. 



Basal taxa (Embiotocidae, Pomacentridae, and Cichlidae) 

 of the suborder Labroidei (Kaufman & Liem, 1982; Stiassny 

 & Jensen, 1987) exhibits Type morphology, whereas some 

 labrid taxa exhibit Type 2 (Table 1). It is most parsimonious 

 to interpret Type 2 epaxial muscle as independently derived 

 within labrids. This interpretation places Bodianus, Choero- 

 don, and Tautoga as basal genera among the Labridae, and 

 might be helpful for determining the polarization of other 

 characters for phylogeny reconstruction in this confusing 

 group. 



Some tetraodontiforms exhibit epaxial insertions on to the 

 distal tips of the dorsal-fin pterygiophores that resemble Type 

 2: Balistidae [Rhinecanthus, pers. obs.; probably Batistes, 

 Balistapus, Melichthys, and Odonus from figs 78, 86, 88 and 

 90 in Winterbottom, 1974b), Monacanthidae (Pervagor, pers. 

 obs.; probably Aluterus, Cantherines, Chaetoderma, Paralu- 

 teres, Paramonacanthus , and Stephanolepis from figs 100, 

 102-105 and 108 in Winterbottom, 1974b), probably Tria- 

 canthidae (Triacanthus, Tripodichthys, Trixiphichthys from 

 figs 66, 76-77 in Winterbottom, 1974b), and perhaps some 

 Triacanthodidae (Triacanthodes, Tydemania, and Mac- 

 rorhamphosodes but not Hollardia or Parahollardia from figs 

 49, 57-58, 61 and 64 in Winterbottom, 1974b). Consideration 

 of the overall anterodorsal morphology of balistids, mona- 

 canthids, and triacanthids suggests that these insertions are 

 likely to have been derived independently of (and non- 

 homologous with) those found in the Perciformes. In these 

 tetraodontiforms, the anterior spinous dorsal fin is closely 

 associated with the back of the skull and separated from the 

 soft dorsal fin. It seems that the robust pterygiophores of the 

 spinous dorsal fin act functionally as a supraoccipital crest 

 and that the epaxial musculature inserts on to these elements 

 as it would to such a crest. If triacanthodids, which possess a 

 more conventional arrangement of spinous dorsal fin and 

 posterior skull, do have epaxial/dorsal pterygiophore inser- 



Fig. 15 Type 1 epaxial musculature in Acanthistius sebastoides (USNM 246689, 96.5 mm SL). A weak tendon extends from a myoseptum to 

 the first pterygiophore and a short and not highly differentiated muscle slip inserts on to the second pterygiophore. Abbreviations and other 

 methods of presentation as in Fig. 1. Scale bar = 5 mm. 



