96 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



Fig. 46. Heads of leptocephali of Dysommatidae (above) and Syn- 

 aphobranchidae (below), showing telescopic eye. 



pharyngidae, Eurypharyngidae, Monognathidae), or as cryptic 

 forms with modified lateral-line and gill-arch characters (My- 

 rocongridae, Xenocongridae, Muraenidae). Two clear associa- 

 tions are evident within this group. One contains the Myrocon- 

 gridae, Xenocongridae, and Muraenidae. These three families 

 are relatively generalized externally but share a marked reduc- 

 tion in gill-arch elements and in the lateral line. The second 

 association contains the three families Saccopharyngidae, Eu- 

 rypharyngidae, and Monognathidae, the so-called gulper eels. 

 These are highly modified midwater eels with a greatly enlarged 

 mouth and an elongated, posteriorly directed suspensorium. 

 The gulpers show extreme reduction in all the skeletal elements, 

 and their relationship to other eels is difficult to determine. 

 Among the remaining families, the Anguillidae is quite primi- 

 tive morphologically, but it seems to have no advanced char- 

 acters clearly linking it to any of the other families. The Mo- 



ringuidae and Heterenchelyidae are fossorial forms that never- 

 theless show substantial internal differences from each other 

 (Smith and Castle, 1972). Their resemblances may simply be 

 convergent adaptations to a similar way of life. The Nemichthy- 

 idae and Cyematidae both have prolonged, nonocclusible jaws 

 studded with liny recurved teeth, but they differ markedly in 

 almost every other character; their traditional association must 

 be questioned. 



Larval characters have so far proved more useful in eluci- 

 dating relationships within families than between them. Some 

 examples will illustrate the contribution that larvae have made 

 to systematics. 



The Moringuidae consists of two genera, Moringna and Neo- 

 congcr. Although both are basically fossorial forms, they differ 

 enough in external appearance that for more than a century they 

 were placed in different families. It was only the striking simi- 

 larity of the larvae (Fig. 45) that prompted a critical comparison 

 of the adults (Smith and Castle, 1972). In this case, the larvae 

 show the relationship much more clearly than do the adults. 



The close relationship between the Synaphobranchidae and 

 Dysommatidae is supported by a unique feature of the larvae— 

 the telescopic eye (Fig. 46). 



The genus Hoplimnis has long been placed in the family Mu- 

 raenesocidae because of its possession of a pectoral fin and its 

 enlarged median vomerine teeth. Saiirenchelys was always con- 

 sidered a nettastomatid because it lacked a pectoral fin. Smith 

 and Castle (1982) showed that the larvae of these genera are 

 indistinguishable (Fig. 47). On that basis and because of many 

 similarities in the adults, Hophinnis and Saurenchelys were shown 

 to be closely related and to belong in the Nettastomatidae. The 

 two characteristic swellings in the gut of larval Hoplunnis and 

 Saurenchelys are also found in the larvae of Nettastoma and 

 Nettenchelys. 



The major problem in eel systematics today is the relationship 

 between the families, and here larvae provide little help. Sim- 

 ilarities occur between larvae of families which otherwise show 

 no evidence of close relationship. For example, the larvae of 

 the Anguillidae and Derichthyidae are quite similar (the larva 

 of Derichthys was even named Lcplocephalus angiulloides). but 

 the two families do not seem especially close and fall on opposite 

 sides of the fused-frontals vs. divided-frontals dichotomy. The 

 larvae of the Heterenchelyidae resemble those of certain con- 

 grids, but heterenchelyids have divided frontals and congrids 

 have fused frontals. Larvae of the congrid genus Acromycter 



Fig. 47. Leptocephali of Hoplunnis tenuis (above) and Saurenchelys sp. (below) (Nettastomatidae). 



