Basal Euteleosts: Relationships 

 W. L. Fink 



AS mentioned in the introduction to this section of the sym- 

 posium, the order Salmoniformes has had a history of 

 attrition, such that today I would recognize it as coextensive 

 with the Salmonidae. Previously included taxa are now scat- 

 tered, primarily as unresolved lineages at or near the base of 

 the Euteleostei. What follows is a preliminary analysis, a sketch 

 of alternative hypotheses of interrelationships of the basal eu- 

 teleosts. Fully resolving these problems will take more time and 

 more material than I have had available to me, and I hope that 

 work stimulated by this symposium will provide insights which 

 have not been forthcoming using traditional material and char- 

 acters. 



Unfortunately, very little information of a comparative nature 

 is available on the larvae of basal euteleosts, and when these 

 larvae have been discussed, only rarely have characters or char- 

 acter transformations useful at large clade levels been men- 

 tioned. Since adult specimens are more easily available in most 

 collections, that is what I have relied on, with examination of 

 larvae when possible. 



Results 



The Euteleostei is a large group of modem teleosts which is 

 poorly diagnosed in terms of unique traits, and most more phy- 

 logenetically advanced members lack some of the diagnostic 

 characters. Patterson and Rosen (1977) considered the following 

 as euteleostean traits: 1) an adipose fin, 2) nuptial tubercles, and 

 3) an anterior membraneous component to the first uroneural. 



Near the "base" of the Euteleostei, Fink and Weitzman(1982) 

 recognized several lineages, including the Esocoidei, Ostario- 

 physi, Argentinoidei, Osmeroidei, Salmonidae, and Neoteleos- 

 tei. All were considered monophyletic, but the interrelations of 

 these large clades were left unresolved (Fig. 106). Below is a 

 review of each of the groups, with new information included 

 when possible. 



Esocoidei or Esocae.—The^ fishes have been a continuing 

 problem for ichthyologists. They are considered as euteleosts 

 on the basis of an anterior membraneous component to the first 

 uroneural, although it is not extensive. No esocoids can have 

 an adipose fin as the dorsal fin is posteriorly situated. Neither 

 do they have breeding tubercles. Rosen (1974) provided diag- 

 nostic characters documenting monophyly of the group. Fink 

 and Weitzman (1982) suggested that esocoids could be the sister 

 group of all other euteleosts based on the lack in the latter of a 

 toothplate on the 4th basibranchial, a bone which is present in 

 esocoids and other primitive teleosts (see those authors for a 

 discussion of the distribution of this character). Wilson and 

 Veilleux ( 1 982) have recently reviewed interrelationships in the 

 Umbridae, and they place Umbra and Dallia as sister taxa, with 

 Novumbra as their sister group; all these together are placed as 

 the sister group of Esox. This corroborates the hypothesis of 

 Nelson (1972). 



Rosen (1974) considered Lepidogalaxias to be a member of 

 this assemblage, which he termed the Esocae. Fink and Weitz- 



man (1982) questioned that hypothesis, leaving the genus un- 

 placed. I have further comments and a new hypothesis of its 

 relationships below. I have nothing to add to what Fink and 

 Weitzman (1982) did with esocoids sensu stricto. and until more 

 is forthcoming, consider them the likely sister group to other 

 euteleosts. 



Ostariophysi— In terms of numbers of species and morpholog- 

 ical diversity, this is the dominant basal euteleostean group. 

 Fink and Weitzman (1982) did not consider the relations of 

 these fishes to other euteleosts, primarily because their survey 

 was intended to establish the placement of stomiiforms, and 

 there was no evidence suggesting relationship between the two 

 groups. No phylogenetic examination of ostariophysan rela- 

 tionships to other teleosts has been done since Rosen and Green- 

 wood (1970) expanded traditional concepts of the group by 

 adding the previously protacanthopterygian gonorynchifonns. 

 Fink and Fink ( 1 98 1 ) examined relationships within the group, 

 placing siluroids and gymnotoids as sister taxa (order Siluri- 

 formes), these the sister taxon of characiforms, and these to- 

 gether the sister group of cypriniforms (the Otophysi, inclusive); 

 sister group relationship of the gonorynchiforms to the Otophysi 

 was corroborated. This entire assemblage was considered mono- 

 phyletic on the basis of numerous characters, including lack of 

 a dermopalatine, unique gasbladder morphology, specializa- 

 tions of the vertebrae, and adductor mandibulae anatomy. 



Argentinoidei.— Cvetn'wood and Rosen (1971) combined the 

 alepocephaloid and argentinoid fishes into an expanded Argen- 

 tinoidei, in the Salmoniformes. Fink and Weitzman (1982) agreed 

 with the combination of the two groups and used the formal 

 subordinal name to include both subgroups. However, Fink and 

 Weitzman (1982) were unable to provide evidence bearing on 

 relationships of these fishes, even though their cladogram (Fig. 

 23, Fig. 106 herein) showed them as the sister group of the 

 osmeroids. I have similarly been unable to place them, in part 

 because of lack of adequate material. 



Osmeroidei— Thii group, which includes the northern and 

 southern smelts, galaxiids (here including Lovettia and Aplo- 

 chiton), Plecoglossus, and salangids, can be diagnosed as mono- 

 phyletic based on several characters, including presence of one 

 or more rows of teeth near the medial border of the mesopter- 

 ygoid, loss or appearance late in ontogeny of the articular bone, 

 and presence of a foramen in the posterior plate of the pelvic 

 bone. Some subgroups of osmeroids have lost various of these 

 diagnostic characters, but the patterns of loss allow other fea- 

 tures to provide evidence of relationship in the group. 



Nevertheless, relationships within the suborder remain prob- 

 lematical. The following review is based upon examination of 

 specimens, the literature, and the contributions to this sym- 

 posium. Incidentally. I have not attempted to diagnose the var- 

 ious genera, but McDowall's comments (this volume and 1 969) 

 indicate that such needs to be done. The phylogenetic hypoth- 



202 



