SMITH: ELOPIFORMES, NOTACANTHIFORMES AND ANGUILLIFORMES 



99 





Fig. 50. Caudal structure of an anguilliform leptocephalus (above) and a notacanthiform leptocephalus (below). 



The only real character seems to be the swim-bladder mor- 

 phology of the two groups (Marshall, 1962). but a critical com- 

 parison with the swim bladders of Elops and Megalops has not 

 been made. Until that is done, it cannot be determined whether 

 the swim bladders of eels and notacanthiforms represent a syn- 

 apomorphy or simply a general condition of elopomorphs. 



Larvae probably cannot resolve the trichotomy. A classifi- 

 cation based on larvae would also yield three groups, but they 

 would not be the same three groups. The three main groups of 

 larvae are the fork-tailed group, the notacanthiform group, and 

 the anguilliform group. These simply represent the condition in 

 the adults. The forked tail is a primitive condition retained in 

 the Elopidae. Megalopidae. and Albulidae. 



Larvae do not reveal much about relationships within the 

 fork-tailed group either. The larvae of Elops and Megalops re- 

 semble each other more than they do that of Albula. They are 

 smaller, the gut is shorter, and the dorsal fin is above or nearly 

 above the anal fin. Albula shows a trend toward elongation, 

 although the myomeres are no more numerous than those of 

 Elops. The gut is very long, ending under the hypural, and the 

 dorsal fin is much farther forward than the anal fin. Pterothrissus 

 is even more elongated and grows larger before metamorphosis 

 than Albula. In albulids the myomeres are more V-shaped than 

 W-shaped. If the primitive condition is small size and relatively 



short larval life, then Megalops has the most primitive larva. It 

 is the smallest known leptocephalus, metamorphosing before it 

 reaches 30 mm standard length, at an age of two to three months 

 (Smith, 1980). Larvae of Elops are closer in size and form to 

 those oi Megalops than to Albula, but this does not necessarily 

 demonstrate that the two former genera are more closely related 

 to each other cladistically than either is to Albula. It could simply 

 mean that Elops and Megalops retain a more primitive larval 

 form and that, once again, they merely lack a specialization 

 found in albulids. 



The larvae of the Notacanthiformes and Anguilliformes do 

 not indicate a particularly close relationship between the two 

 groups. The elongated form simply reflects the condition in the 

 adults, and in several respects the two groups are quite different. 

 The short-based dorsal fin and the presence of pelvic fins in 

 notacanthiform larvae immediately separate them from an- 

 guilliform larvae. Eels lack pelvic fins and their dorsal fin is long 

 and confluent with the caudal and anal fins. In both these char- 

 acters the notacanthiforms show the more primitive state. In 

 the structure of the tail, however, the notacanthiforms are more 

 highly modified. Eels, despite their elongate form, retain a caudal 

 fin complete with hypural plates and caudal fin rays. To be sure, 

 the caudal fin is greatly reduced and shows much fusion of 

 elements, but it clearly exists, in larvae as well as adults (Fig. 



