206 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



Eurypterygii. — FinaWy, a few comments are due on the Myc- 

 tophoidei of Greenwood et al. (1966). This group was disman- 

 tled by Rosen (1973), and divided into two large groups, Au- 

 lopiformes and Myctophiformes. These two groups, together 

 with the Paracanlhopterygii and Acanthopterygii, were classified 

 into a new group, Eurypterygii. Aulopiformes was placed as the 

 sister group of all other eurypterygians, and myctophiforms as 

 the sister group to paracanthoptergyians and acanthopterygians. 

 All of these, together with stomiiforms, form the Neoteleostei. 

 Fink and Weitzman (1982) tentatively accepted monophyly of 

 the Eurypterygii based on the presence in its members of a 

 toothplate fused with the third epibranchial. Aulopiformes con- 

 tains a large number of families, including the Giganturidae, 

 covered in this portion of the symposium. About the latter 

 family I have little to say except that my own dissections cor- 

 roborate Rosen's placement of it. 



Summary 



A summary of the hypotheses I have discussed above is given 

 in Fig. 108. The most striking aspect of it is the degree of un- 

 certainty about relationships among the clades. This may be in 

 part due to the limitations of my study, but it does seem to me 

 to be a fair summary of the status of well corroborated hypoth- 

 eses we now have about this level of teleostean phylogeny. There 

 are certainly other arrangements that can be made, depending 

 on which characters one wishes to stress, and none of these 

 should be discarded out of hand. As examples, I will cite two 

 characters and their implications. 



First, lack of the posterior shaft of the vomer suggests that 

 salmonids and osmeroids are sister taxa. Appropriate outgroups 

 have the shaft ranging from "moderate" (e.g., Chanos) to "elon- 

 gate" (argentinoids). My own opinion, based on occipital con- 

 dyle structure of salmonids, is that the reduction in vomer length 

 has occurred independently in the two lineages (it has also been 

 reversed within both); the ultimate value of the occipital char- 

 acter remains to be seen. 



The second character, presence of breeding tubercles, is now 

 considered a euteleostean trait. Note, however, that tubercles 



are lacking in esocoids and argentinoids but are present in os- 

 tariophysans, osmeroids, and salmonids, indicating that these 

 three clades form a monophyletic group. Again, there are char- 

 acters that contradict this grouping, but it nevertheless is worthy 

 of consideration. 



It is always frustrating when one sets out to solve a particular 

 problem and then comes to the end of the allotted time without 

 a resolution. Although I have been able to shed some light on 

 several problems relevant to the goals of this part of the sym- 

 posium, I have not been able to unravel the interrelationships 

 among the major basal euteleostean clades. Clearly more work 

 is needed, especially with character suites which have been tra- 

 ditionally neglected. Almost all of our concepts of relationships 

 at this level are based on features of the adult caudal skeleton 

 and branchial basket. Some work on soft anatomy, particularly 

 the muscles of the head, has been informative at these levels 

 and one hopes that other parts of the soft anatomy will be equally 

 profitable. One area virtually untouched is larval anatomy. It 

 might be expected that not many important features will be 

 found because of the preponderance of primitive characters in 

 larvae. But larval characters have proven useful, as is shown by 

 the ontogenetic transformation in tooth types in stomiiforms 

 (from type 4 to type 3; see Fink, 1981) as well as the specialized 

 fin traits discussed by Ahlstrom et al. (this volume) for argen- 

 tinoids. It is in both these areas, ontogenetic character trans- 

 formations and presence of specializations for larval life, that 

 study of larval fishes promises rewards. The inclusion of larval 

 morphology in studies of higher level relationships should pro- 

 vide a richer data base than we currently have and perhaps will 

 reveal some crucial characters for resolving the basic questions 

 I have addressed above. This symposium has already stimulated 

 in a major way the examination of larvae for phylogenetic anal- 

 yses, and I predict that it, combined with the new ways now 

 emerging of analyzing ontogenetic information, will mark a new 

 phase in the modem study of fish classification. 



Museum of Zoology, University of Michigan, Ann Arbor, 

 Michigan 48109. 



Myctophiformes: Development 

 M. Okiyama 



MYCTOPHIFORMES is currently adopted as a distinct 

 order with intermediate affinity between the lower and 

 higher teleost groups, whereas no one feature would satisfac- 

 torily separate all of them from all Salmoniformes (Gosline et 

 al., 1966). Except Rosen (1973), recent workers agree well with 

 the familial composition of this order despite slight differences 

 in the familial or subordinal definition. 



Table 56 shows the recent classification given by Johnson 

 (1982) based on the most comprehensive knowledge now avail- 

 able. Important points of this scheme are the exclusion of Sco- 

 pelarchidae from Alepisauroidei and Pseudotrichonotidae from 

 Myctophiformes. Further details in this connection will be men- 

 tioned again in my paper on relationships (this volume). 



Exploitation of the vast hydrosphere covering the pelagic as 

 well as benthic habitat between the surface and abyssal or ul- 

 traabyssal plain by diversified members of this group is doubt- 

 lessly the important aspect in discussing the ontogenetic prob- 

 lems of the myctophiform lineage. Of the five suborders, 

 Myctophoidei and Alepisauroidei are exclusively pelagic and 

 the remaining are demersal including secondary pelagic genera 

 such as Parasudis and Harpadon. Synchronous hermaphrodit- 

 ism is common to the deep-water and offshore forms belonging 

 to Chlorophthalmoidei and Alepisauroidei with the single ex- 

 ception of Bathysauridae in Synodontoidei (Table 56). 



In general, the systematics of this order are rather well under- 

 stood except for several families or genera. As is clearly shown 



