182 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



STERNOPTYCHIDAE 



GONOSTOMATIDAE 



PHOTICHTHYIDAE 



ASTRONESTHIDAE 



IDIACANTHIDAE 



MALACOSTEIDAE 



MELANOSTOMIIDAE 



STOMIIDAE 



CHAULIODONTIDAE 



Fig. 95. Weitzman's (1974) hypothesis of relationships of the sto- 

 miiform fishes. The Gonostomatidae and Stemoptychidae comprise the 

 Gonostomata and the remaining families comprise the Photichthya. 



sister group to the Eurypterygii. Fink and Weitzman (1982) 

 agreed with this placement, provided more characters to sub- 

 stantiate it, and demonstrated monophyly of the stomiiforms. 

 Steyskal (1980) has presented arguments that the root of the 

 family-group names demands that these be altered from Sto- 

 miatidae and Stomiatiformes to Stomiidae and Stomiiformes, 

 respectively, and I use these forms throughout this paper. 



As recognized by Weitzman (1974), there are two major sto- 

 miiform lineages, Gonostomata and Photichthya, both classified 

 at infraordinal rank, with families Gonostomatidae and Ster- 

 noptychidae in the former and families Photichthyidae, Sto- 

 miidae, Chauliodontidae, Astronesthidae, Melanostomiidae, 

 Malacosteidae, and Idiacanthidae in the latter (Fig. 95). I have 

 no disagreement with Weitzman's hypotheses of monophyly of 

 the Stemoptychidae, but our recent work on Diplophos (Fink 

 and Weitzman, 1982) caused us to question the monophyly of 

 the Gonostomatidae and Photichthyidae, and my work on the 

 barbelled stomiiforms, comprising the remaining families, has 

 cast doubt on the entire traditional arrangement of the included 

 26 genera as well as on the monophyly of the Photichthya. I 

 have found features which support new hypotheses of relation- 

 ship within the stomiiforms and will present some of these ideas 

 below. Some are more tentative than others. Weitzman is cur- 

 rently working on the genera he placed in the Gonostomatidae 

 and Photichthyidae. 



First, I have found no evidence that Diplophos is the sister 

 group of any other genus of stomiiform and it may be, as Fink 

 and Weitzman (1982) suggested, the sister group of the rest of 

 the order. Specializations in the adductor muscles indicate that 



Diplophos 



GONOSTOMA 



Cyclothone 



Margrethia 



BONAPARTIA 



Triplophos 



STERNOPTYCHIDAE 



PHOTICHTHYA 



Fig. 96. Hypotheses of stomiiforms as discussed herein. See text for 

 explanation. 



Gonostoma. Cyclothone, Margrethia. and Bonapartia form a 

 monophyletic group, but what relationships within that group 

 are I cannot say, and presumably this will be treated by Weitz- 

 man. These hypotheses would cause a redefinition of the Gon- 

 ostomatidae, restricting it to the four genera mentioned just 

 above. Relationships of Triplophos are also unclear, and there 

 is evidence in the hyoid apparatus that it may be related to some 

 of the "photichthyans," rather than the gonostomatids, as 

 Weitzman ( 1 974) supposed. Weitzman ( 1974) established mon- 

 ophyly of the Stemoptychidae, and 1 have nothing to add to his 

 conclusions. Nevertheless, since he did not deal with monophyly 

 of the Gonostomatidae or with the sister group relationship of 

 the Stemoptychidae, there is no current evidence that the latter 

 is more closely related to some subset of the former, and I leave 

 that part of the phylogeny unresolved. These hypotheses are 

 summarized in Fig. 96. See also the paper by Ahlstrom, Rich- 

 ards, and Weitzman (this volume) on the Gonostomatidae, Ster- 

 noptychidae and other stomiiforms. 



Within the "Photichthya," we have the same problem as with 

 the Gonostomatidae; that is, there is a diagnosable monophy- 

 letic unit (the barbelled forms) and an undiagnosed grade group, 

 the Photichthyidae. 



My own efforts have been on the barbelled forms, currently 

 distributed in six families, as listed above. There have been no 

 strictly phylogenetic studies of relationships within the group, 

 but they were examined in a traditional sense by Parr (1927), 

 Regan and Trewavas ( 1 929, 1 930), and Beebe and Crane (1939). 



