324 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



Fig. 166. Cladogram showing proposed phylogenetic relationships of the major subgroups of the Lophiiformes. Black bars and numbers refer 

 to synapomorphic features discussed in the text. 



those of lophioids; smallest larvae are certainly less than 50%, 

 and probably less than 30% the size of those of lophioids; size 

 at transformation to the prejuvenile stage is less than 60% that 

 of lophioids); (15) Head of larvae proportionately large rel- 

 ative to body (always greater than 45% of standard length, com- 

 pared to less than 30% in lophioids); (16) Reduction in the 

 number of dorsal fin spines from a primitive number of six in 

 lophioids to three or less (Pietsch, 1981:409, figs. 36-38); and 

 (17) Loss of pharyngobranchial IV (present and well-toothed in 

 lophioids; Pietsch, 1981:401, figs. 1 1, 28-32). 



Monophyly for a group containing the suborders Chauna- 

 cioidei, Ogcocephalioidei and Ceratioidei is supported by two 

 synapomorphies: (18) Second dorsal spine reduced and embed- 

 ded beneath skin of head (Pietsch, 1981:410, figs. 36-38); and 

 (19) Gill filaments of gill arch I absent (but present on proximal 

 end of ceratobranchial I of some ceratioids; Bradbury, 1967: 

 408; Pietsch, 1981:415). 



That the Ogcocephalioidei is the primitive sister-group of all 

 ceratioid families is supported by three synapomorphies: (20) 

 Second dorsal spine reduced to a small remnant (well developed 

 in the ceratioid family Diceratiidae. and in all other lophiiforms; 

 Bertelsen, 1951:17; Pietsch, 1 98 1 :4 1 0, fig. 38); (2 1 ) Third dorsal 

 spine and pterygiophore absent (present in all other lophiiforms; 

 Bertelsen, 1951:17; Bradbury, 1967:401; Pietsch, 1981:410, fig. 

 38); and (22) Epibranchial I simple, without ligamentous con- 

 nection to epibranchial II (in batrachoidiforms and all other 

 lophiiforms epibranchial I bears a medial process that is liga- 



mentously attached to the proximal tip of epibranchial II; Pietsch, 

 1981:401, figs. 28-32). 



Of the possible cladograms that could be constructed on the 

 basis of the data provided in this study, the one shown in Fig. 

 166 is by far the most parsimonious. But at the same time, 

 acceptance of this revised hypothesis of relationships of lo- 

 phiiform fishes requires evolutionary convergence or reversal 

 in three derived character states previously used by me (Pietsch, 

 1981:415, fig. 41) to support a hypothesis of sister-group rela- 

 tionship between the Chaunacidae and Ogcocephalidae: ( 1 ) Pos- 

 teriormost branchiostegal ray exceptionally large (all four pos- 

 teriormost branchiostegal rays approximately equal in size in 

 batrachoidiforms and all other lophiiforms; Pietsch, 1981, fig. 

 27); (2) Gill teeth tiny, arranged in a tight cluster at apex of 

 pedicel-like tooth plates (in all other lophiiforms gill teeth, if 

 present, are relatively large, and either single, or associated with 

 a flat, rounded tooth plate; but tiny, and at apex of elongate 

 pedicel-like tooth plates in at least some batrachoidiforms, e.g., 

 Poriclithys; Pietsch, 1981, figs. 31,32) and (3) Illicial bone, when 

 retracted, lying within an illicial cavity (an illicial cavity is absent 

 in all other lophiiforms; however, the illicium and esca lie within 

 a shallow groove on the dorsal midline, sometimes enveloped 

 by folds of skin, in the antennariid genus Histiophryne, Pietsch, 

 1981, fig. 39; Pietsch, 1984:40). 



The cladistic relationships of the Lophiiformes are summa- 

 rized in the following revised classification. While the ranking 

 of taxa is not dichotomous (see methods in Pietsch, 1981:388), 



