678 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



THC •HY3t4 



Fig. 36 1 . Caudal skeleton of Citharoides macrolepis. Hypural pat- 

 tern 2. HAR = haemal-arch remnant, other abbreviations as in Fig. 

 359. "V" on distal end of fin ray indicates dorsal- and ventralmost 

 branched ray. 



scribed, in the samarines hypural 1 does not articulate with the 

 terminal half centrum. 



The last pattern of hypurals (6) is characterized as follows 

 (Figs. 360, 364 middle and lower); hypurals 1 and 2 are fused 

 together forming one element which articulates with the pos- 

 teroventral surface of the terminal half centrum; and hypurals 

 3 and 4 are fused together and to the terminal half centrum. 

 This pattern occurs in the Pleuronectinae, Paralichthyidae (ex- 

 cept Tephrinectes and Thysanopsetta), Scophthalmidae, one 

 citharid (Brachypleura), and the Bothidae (except Mancopsella). 

 We interpret this pattern as homologous between these groups, 

 derived, and indicative of a monophyletic origin. We will refer 

 to these fishes as the bothoid group. Caudal-fin development in 

 a bothid is illustrated in Fig. 360. 



Although there is still some doubt concerning interpretations 

 of certain epaxial caudal elements in flatfishes, some patterns 

 are apparent. Most of the information indicates that at least in 

 most pleuronectiform groups, the basal epural number is two. 

 However, there is a small third element that appears in many 

 species (Fig. 36 1 ; first uroneural of Amaoka, 1 969). This element 

 does not appear to be paired and its interpretation and fate in 

 some groups is questionable. The two larger epural elements are 

 still present in some flatfishes (Figs. 361, 363 upper), the cith- 

 arids Lepidoblepharon and Citharoides and the paralichthyid 

 Tephrinectes. The fate of these from the perspective of the entire 

 order is questionable. However, it is obvious that these epurals 

 have been reduced to one or zero in several groups. Which of 

 these reductions are homologous is unknown. Within groups 

 defined by other specializations, however, we are probably jus- 

 tified in assuming these epural reductions took the same course 

 and are homologous states. 



Although space does not allow a more detailed discussion of 

 other caudal-fin characters, some obvious trends should be men- 



tioned; Symmetrization— There is a marked trend among flat- 

 fishes toward dorsoventral symmetry in the caudal fin and skel- 

 eton. This has occurred by various types of fusions, losses, and 

 secondary divisions of elements. These secondary divisions oc- 

 cur as scissures of varying depths in many caudal elements (Figs. 

 360H, 362F, 363 lower, 364 upper). Reduction of total and 

 branched caudal rays— It has long been recognized that more 

 primitive flatfishes tend to have larger numbers of total and 

 branched caudal rays. Thus, Psettodes has a total caudal ray 

 count of 24-25, 15 of which are branched. In many groups, 

 caudal rays have been reduced to less than 18 and branched 

 rays to 0-13. 



Infraorbital lateral-line canal on ocular side. — In his study of 

 sinistral flounders (i.e., Psettodidae and Pleuronectoidei) of Ja- 

 pan, Amaoka (1969) found ocular infraorbital bones present in 

 the Psettodidae, two citharid genera (Citharoides. Lepidobleph- 

 aron), and the Paralichthyidae; they were absent from Japanese 

 bothids. We have since done some survey work on this character 

 in other groups not treated by Amaoka and found ocular in- 

 fraorbital bones missing in additional groups (Table 179). 



Examination of the Regan-Norman model 

 using adult characters 



In the following discussion, the groups and classification re- 

 sulting from the current model for pleuronectiform evolution 

 will be reexamined. The limited analysis presented here sheds 

 much doubt on the monophyly of many of the currently rec- 

 ognized groups and their interrelationships. In a few cases, the 

 evidence favoring different interpretations is so strong that these 

 should be recognized in classifications. However, most of this 

 analysis has produced questions and alternative suggestions that 

 need additional study. 



Psettodoidei, Psettodidae. —Nearly all of the character states used 

 to define this group (Psettodes. two species) are symplesio- 

 morphies or have been interpreted as such. Two exceptions, gill 

 arches with groups of teeth and barbed jaw teeth, are states that 

 Hubbs (1945) proposed as synapomorphies. Although we have 

 no reason to doubt that Psettodes is a natural group, it should 

 be redefined using character states which have been shown to 

 be synapomorphies. 



Soleoidei.— The diflferences between the Soleoidei and Pleuro- 

 nectoidei were noted and expressed in important classifications 

 before the works of Regan and Norman (e.g., Jordan and Ev- 

 ermann, 1896-1900) and they are obviously evident in the cur- 

 rent model and classification. In most previous systematic re- 

 search on pleuronectiforms, the author has concerned himself 

 with one or the other group and assumed that the two were 

 related only through a common ancestor near the early pleu- 

 ronectiform line. The possibility, for example, that some so- 

 leoids may be most closely related to some pleuronectoids has 

 only rarely been addressed. In any cladistic analysis of pleuro- 

 nectiform interrelationships, character states used to unite the 

 soleoids will need to be reinterpreted. Some character states 



Fig. 362. Caudal-fin structure of Solea solea larvae (A-C), juveniles and adults (D-F). Total lengths of specimens: (A) 6.0 mm; (B) 6.8 mm; 

 (C) 8.1 mm; (D) 1 1.5 mm; (E) 18 mm; (F) 470 mm. HA = haemal arch, NA = neural arch, other abbreviations as in Figs. 359, 360. Redrawn 

 from Fabre-Domergue and Bietrix (1905). 



