682 



ONTOGENY AND SYSTEMATICS OF HSHES-AHLSTROM SYMPOSIUM 



The family Citharidae as presently defined is a grade. Ex- 

 amination of the caudal osteology has shown two derived and 

 one plesiomorphic pattern of hypurals. Lepidoblepharon shows 

 pattern 1, which is plesiomorphic for the order. Citharoides 

 shows pattern 2, a derived pattern (Fig. 361). This pattern could 

 represent a state on a line leading toward pattern 6, which is 

 shown by Brachypleura. Eucitharus shows pattern 4, which pos- 

 sibly developed independently in some rhombosoleines. The 

 most obvious result of this is that Brachypleura belongs to the 

 bothoid group, which shares the derived hypural pattern 6. In 

 this interpretation, the character states shown by Brachypleura 

 that are primitive for the order (e.g., vomerine teeth, ventral- 

 fin spines) are also primitive for bothoids. 



Scophthalmidae.— Based on ventral-fin morphology, the Scoph- 

 thalmidae appear to be monophyletic. There are certain simi- 

 larities in ventral-fin morphology between this family and the 

 achirines, but these are probably superficial. Scophthalmids were 

 previously thought to be closely related to and derived from the 

 Citharidae (Hubbs, 1945). This hypothesis was based on certain 

 symplesiomorphies, e.g., the low degree of fusion of the gill 

 membranes and the presence of vomerine teeth. The Scoph- 

 thalmidae show hypural pattern 6 and are thus members of the 

 bothoid group. 



Paralichthyidae.— 'Norman (1934) basically defined the 

 Paralichthyinae (=Paralichthyidae with modifications) on ex- 

 ternal pelvic-fin morphology and vertebral structure (absence 

 of transverse apophyses). The group was supposed to have the 

 ventral fins nearly symmetrical in position and base lengths, or 

 the ocular fin on the midventral line and its base slightly ex- 

 tended anteriorly. Symmetries in ventral-fin position and base 

 lengths are plesiomorphic for the order and bothoids. Norman's 

 paralichthyid genera with an ocular ventral fin on the midventral 

 line and its base extended anteriorly are bothids (i.e., Trichop- 

 setta, Engyophrys, Taeniopsetta, Monolene. Perissias). 



Amaoka (1969) presented a more thorough, detailed defini- 

 tion of the family. However, many or most of the character 

 states he used appear to be plesiomorphic for bothoid fishes 

 (i.e., those defined by hypural pattern 6). A second limitation 

 of Amaoka's work on this group is that it was limited to three 

 genera {Paralichthys, Pseudorhombus. Tarphops). An important 

 change in Norman's classification was made by Amaoka when 

 he removed Taeniopsetta from the Paralichthyidae and placed 

 it in the Bothidae. Hensley (1977) and Futch (1977) did the 

 same for Monolene, Engyophrys, and Trichopsetta. 



We have now examined some characters in the remaining 

 Paralichthyidae, and additional changes are required in the com- 

 position of this group. In a survey of caudal-fin structure, it was 

 found that Thysanopsetta and Tephrinectes show the most prim- 

 itive type of hypural pattern (1; Fig. 363 upper). These two 

 genera are much more primitive than expressed in the current 

 classification and definitely do not belong to the bothoid group. 



Within the remaining Paralichthyidae another group is dis- 

 cemable. This is composed of Cyclopsetta, Syacium, Citharich- 

 thys, and Etropus, i.e., the Cyclopsetta group. States for two 

 complexes of characters, ventral-fin morphology and urinary- 

 papilla position, are unique to this group and interpreted as 

 synapomorphic. Arrangement of caudal-fin rays in the Cyclop- 

 setta group is also unique and probably apomorphic (Fig. 364 

 middle). All species have 1 7 caudal rays, none of which are 



supported by preural, neural or haemal spines. It should also 

 be noted that the fifth hypural has fused with an epural. This 

 fusion has been observed in larval development (Tucker. 1982; 

 Ahlstrom, pers. observ.). However, fusion of the fifth hypural 

 and one or more epurals has apparently occurred several times 

 in pleuronectiforms, possibly including the bothoids (e.g., see 

 Fig. 360). A detailed analysis of relationships between the Cy- 

 clopsetta group and other bothoids is not possible here. How- 

 ever, some character states may indicate a close relationship 

 with bothids (absence of first neural spine, presence of vertebral 

 transverse apophyses). 



Amaoka (1 969) and one of us (Ahlstrom) recognized another 

 group within the Paralichthyidae composed of Pseudorhombus, 

 Tarphops. and Cephalopsetla. i.e.. the Pseudorhombus group. 

 We interpret these genera as more specialized in certain char- 

 acters than most other members of the family. Species of this 

 group usually have a total caudal ray count of 17, the epural is 

 fused to the fifth hypural, and they lack a splinter ray on the 

 ventralmost caudal-fin ray. With the exceptions noted above 

 (the primitive non-bothoid genera Tephrinectes and Thysan- 

 opsetta and the Cyclopsetta group), the remaining paralichthyids 

 of the Regan-Norman classification (what we are calling the 

 Paralichthys group) have the apparently plesiomorphic states 

 of 1 8 caudal rays, at least one free epural (except in one species 

 of Hippoglossina (Sumida et al., 1979)), and a splinter ray on 

 the ventralmost caudal-fin ray (Fig. 364 lower). The splinter ray 

 is probably a remnant of a ray lost through fusion with an 

 adjacent ray (Okiyama, 1974). The Pseudorhombus group may 

 be definable by synapomorphies but a detailed analysis has not 

 been done. 



After removal of the bothids (Trichopsetta, Engyophrys, Tae- 

 niopsetta, Monolene, Perissias) and the primitive non-bothoid 

 genera (Tephrinectes, Thysanopsetta), recognition of the Cy- 

 clopsetta group as monophyletic. and recognition of the Pseu- 

 dorhombus group as possibly monophyletic, few of the original 

 paralichthyid genera remain. We have been referring to these 

 as the Paralichthys group (Ancylopsetta, Gastropsetta, Hippo- 

 glossina, Lioglossina, Paralichthys, Verecundum, and Xystreu- 

 rys). At least most of the character states known for these re- 

 maining genera are plesiomorphic for the order (e.g., symmetrical 

 ventral-fin states) or for bothoids (e.g.. usual presence of at least 

 one free epural). The Paralichthys group is probably not mono- 

 phyletic. 



Bothidae.— Norman (1934) defined the Bothinae (=Bothidae 

 with modifications) on the basis of a high degree of ventral-fin 

 asymmetry and the presence of vertebral transverse apophyses. 

 The ocular ventral fin was said to be on the midventral line with 

 its base extending anteriorly to the urohyal. Norman excluded 

 Taeniopsetta, Engyophrys, Trichopsetta, Monolene, and Peris- 

 sias from this group because the base of the ocular ventral fin, 

 although on the midventral line and somewhat longer than that 

 of the blind side, does not extend to the urohyal. 



Amaoka (1969) examined many bothid genera and redefined 

 the family using more characters. Most of the characters stressed 

 by Amaoka have now been examined in other bothoids. These 

 are discussed below: 



Ventral-fin asymmetry.— In bothids the ocular fin base is on the 

 midventral line, elongated, and has its origin anteriorly placed 

 relative to the base of the blind fin. Within the bothoids this 

 combination of states appears to be derived and unique. 



