JOHNSON: GIGANTURIDAE 



201 



(Bertelsen, 1 95 1 ) or the larvae of certain scopelarchids (Johnson, 

 1974b, 1982) to the elongate, shallow, slender shape of the gi- 

 ganturids. The head while still massive is proportionately much 

 less so ('/« vs 'A SL in Rosarua) and the dorsal head profile is 

 essentially horizontal rather than steeply oblique (Fig. 105). (B) 

 Eyes. Eyes in larvae are round, small and directed laterad; eyes 

 in adults are fully tubular and directed rostrad. (C) Fins. Dis- 

 tinct, partly-stalked, 5-rayed pelvic fins are present in larvae, 

 resorbed or shed during transformation, and lacking in adults. 

 The line of insertion of the pectoral-fin rays is obliquely vertical 

 in larvae, essentially horizontal in adults. In larvae the pectoral 

 insertion is behind the gill slit, in adults (especially prominent 

 in Gigantura) the pectoral insertion is substantially above the 

 gill slit. A distinct dorsal adipose fin is present in larvae, absent 

 in adults. Procurrent caudal fin rays number (3)4(5) in larvae 

 and are prominent, in adults procurrent caudal rays are fre- 

 quently embedded in the skin, difficult to see, and number 

 (0)1(2,3). (D) Teeth. Among the most striking changes occurring 

 dunng transformation is the total loss of all larval teeth (in- 

 cluding basihyal teeth). Transforming specimens are character- 

 ized by a scalloped, irregularly-emarginate jaw edge (upper and 

 lower) which is edentulous. None of the 40 known transforming 

 specimens shows development of adult teeth and the smallest 

 known post-transformation specimen (36.4 mm SL, G. chuni; 

 47.9 mm SL, Rosaura indica) possess a full complement of 

 adult teeth. (E) Color. Larvae are essentially translucent with 

 very little development of pigment, adults are entirely blackish/ 

 brown (often with the development of an iridescent finish in 

 Gigantura). Onset of transformation is indicated by the "sud- 

 den" widespread development of pigmentation. (F) Loss of skel- 

 etal elements. Larvae possess at least the following skeletal ele- 

 ments not seen in adults: symplectic, coracoid, cleithrum, 

 posttemporal, supracleithrum, branchiostegal rays. 



Relationships 



The first association of "Rosaura" with the giganturids was 

 by Ahlstrom and Berry about 1960 (letters and mss material 

 made available by H. G. Moser) with the first published sug- 

 gestion made in Berry and Perkins (1966). Key characters sug- 

 gesting relationship included the very high pectoral-fin ray count 

 and the highly unusual 10-1-6(7) distribution of principle caudal 

 rays, apparently unique to "Rosaura" and the giganturids. The 

 disparities between "Rosaura" larvae and adult giganturids— 

 briefly outlined above— left doubt in many minds, but the cap- 

 ture of essentially complete transformation series (to be de- 

 scribed and illustrated in detail elsewhere) make it unquestion- 

 able that "Rosaura" is the larval form of the giganturids. With 

 a caudal peduncle depth of ca 9.9% of SL (Tucker, 1954:168) 

 there is likewise no doubt that the type of Rosaura rotunda 



represents a larva of "Balhyleptus," requiring recognition of the 

 more elongate, shallow-bodied species as Rosaura indica (Brauer, 

 1901). The deeper-bodied species is Gigantura chuni Brauer, 

 1901 (other species have been described but the characters used 

 to distinguish them do not work, nor has other evidence been 

 found to support the hypothesis of more than two species). Of 

 the two, Walters (1961, 1 964) argued for the more apomorphous 

 condition of Gigantura but his characters need to be re-exam- 

 ined in light of outgroup comparisons and in conjunction with 

 other characters. 



Vanous authors have allied giganturids with such disparate 

 groups as Stylephoridae, Saccopharyngiformes and "... a line 

 [leading] from a subiniomous group such as the esocoids toward 

 the synodontoid inioms, and this line later may have given rise 

 to the Cetunculi . . ." (Walters, 1961). Rosen (1973:438-441) 

 has offered evidence that the original placement by Regan (1925: 

 57) of giganturids with synodontoids was correct. Rosen calls 

 particular attention to similarities in upper jaw and infraorbital 

 configuration with synodontoids and the presence of a retractor 

 dorsalis (=RAB in Rosen, 1973; see Winterbottom, 1 974b) mus- 

 cle configration state characteristic of the synodontoid/alepi- 

 sauroid line (Johnson, 1982:85, 95). An important character 

 (Johnson, 1982:71; Okiyama, this volume) uniting synodon- 

 toids with alepisauroids is the presence in larvae of multiple (3 

 or more) peritoneal pigment sections. Uniting synodontids and 

 harpadontids (sensu Sulak, 1977) is the fact that in larvae of 

 these fishes the sections are paired . . . and not connected over 

 the gut. The condition in "Rosaura" is that seen in aulopids, 

 chlorophthalmids, primitive scopelarchids, and ipnopids, viz. a 

 single section situated over the gut. This is the state thought 

 primitive for inioms. Also distinguishing the giganturids is a 

 unique conformation of the gut. In larvae the gut arises from 

 the pylorus, descends round the left margin of the abdominal 

 cavity, crosses transversely midventrally, reascends the right 

 side, turns abruptly mediad, then turns again, descending abruptly 

 and obliquely to the vent. In adults the intestine arises mid- 

 ventrally, makes a few small twists, ascends the right side, and 

 passes posteriad above the dorsal contour of the expanded stom- 

 ach, only descending to the vent posterior to the terminus of 

 the stomach. In all the inioms I have examined the intestine 

 arises midventrally and passes essentially straight back to the 

 vent along the midventral wall of the abdominal cavity. For the 

 time being, the available evidence suggests that the giganturids 

 are neoteleosts (retractor dorsalis muscle), allied with the inioms 

 (discrete peritoneal pigment section), diverging early from the 

 rest and acquiring characters making them among the most 

 specialized and distinctive of teleosts. 



Field Museum of Natural History, Roosevelt Road at 

 Lake Shore Drive, Chicago, Illinois 60605. 



