HENSLEY AND AHLSTROM: PLEURONECTIFORMES 



683 



Preorbital on blind side.— This bone is absent in the Bothidae. 

 It appears to be present in all other bothoids (Pleuronectinae 

 not examined for this character). Based on this comparison, we 

 interpret the loss of this bone a derived state within the bothoids 

 defining the family Bothidae. 



Infraorbital bones of the ocular side. — All bothids have an ocu- 

 lar preorbital bone but lack the remainder of the series. The 

 presence or absence of the ocular preorbital has not been sur- 

 veyed in most bothoid groups. However, an ocular infraorbital 

 lateral line is present in most bothoids. In addition to the Both- 

 idae, it is missing in Brachypleura and the Cyclopsetta group. 



Intermuscular bones.— We interpret the presence of at least two 

 of the series of these bones (myorhabdoi) as a derived state 

 unique to and defining the Bothidae. 



First neural spine.— Although the first neural arch is present, 

 the neural spine is missing in the Bothidae. It is present in all 

 other bothoids except the Cyclopsetta group. 



Vertebral transverse apophyses.— All bothids have two pairs of 

 transverse apophyses on most vertebrae. As previously dis- 

 cussed, how to interpret these on the pieuronectiform level and 

 within the bothoid group is questionable. Within the bothoids 

 well-developed and very similar structures occur only in the 

 Bothidae and Scophthalmidae. Very small transverse apophyses 

 also occur in the Cyclopsetta group. 



Based on these characters, the Bothidae appear to be mono- 

 phyletic and definable by synapomorphies in at least three char- 

 acters or complexes: (1) loss of the preorbital on the blind side; 

 (2) presence of myorhabdoi; and (3) asymmetrical states of ven- 

 tral-fin morphology. 



Since Amaoka's ( 1 969) work, we have examined the remain- 

 ing genera not examined by him that have been considered 

 bothids (i.e., Graminatobothus. Lophonectes, Pelecanichlhys, 

 Mancopsetta). All of these except Mancopsetta are bothids. 

 Mancopsetta exhibits the following character states: ( 1 ) hypural 

 pattern 1, i.e., the most primitive type; (2) presence of pleural 

 and epipleural ribs, but no myorhabdoi or other intermuscular 

 bones in the caudal region; (3) at least one free epural (none in 

 adult bothids); (4) anus on midventral line (clearly on blind side 

 in bothids); (5) no vertebral transverse apophyses; and (6) seven 

 rays in the ocular ventral fin, 5-7 in that of the blind side (six 

 in both fins in bothids). These are all characters in which Man- 

 copsetta differs from the Bothidae. Due to the primitive hypural 

 pattern, it is not a bothoid (see Rhombosoleinae). 



Amaoka (1969) analyzed intergeneric relationships of Jap- 

 anese bothids. However, his analysis was eclectic and did not 

 include all genera (i.e., Engyophrys. Trichopsetta, Monolene, 

 Perissias. Graminatobothus. Lophonectes. and Pelecanichlhys 

 were not examined). He recognized two subfamilies, the Tae- 

 niopsettinae and Bothinae. He erected the first subfamily for 

 Taeniopsetta. Hensley (1977), Futch (1977), Evseenko (1977, 

 1981), and Amaoka (1979) implied that Engyophrys and Tri- 

 chopsetta should be included in the Taeniopsettinae. This was 

 done on the basis of larval characters and ventral-fin morphol- 

 ogy. Most of the slates used to define the Taeniopsettinae were 

 considered by Amaoka (1969) to be plesiomorphic at the family 

 level. Three characters were emphasized: (1) degree of anterior 



extension of the base of the ocular ventral fin; (2) shape of the 

 ventral (sciatic) area of the urohyal; and (3) number of suborbital 

 bones on the blind side. In the taeniopsettines, the origin of the 

 blind ventral fin is at the same transverse level as the second 

 ray of the ocular ventral fin, i.e., the base of the ocular fin is 

 only slightly elongated. In the Bothinae, extension of the base 

 of the ocular fin is greater and the origin of the blind fin is on 

 the same transverse level as the third or fourth ray of the ocular 

 fin. Obviously, the taeniopsettine state here is the more plesio- 

 morphic. Engyophrys, Trichopsetta, Monolene, and Perissias 

 show this state. Taeniopsetta has a broad, truncate margin on 

 the sciatic part of the urohyal. In bothines, this area of the 

 urohyal is pointed. Amaoka (1969) clearly showed that the ple- 

 siomorphic state for bothoids is closer to the condition shown 

 in taeniopsettines. Engyophrys, Trichopsetta, and Perissias show 

 the taeniopsettine condition, Monolene the bothine state. 

 Amaoka (1969) noted an apparent trend among bothoids in 

 reduction of the number of suborbital bones of the blind side. 

 This reduction may have occurred in several bothoid groups 

 and interpretation of this character is not clear. Thus, infraor- 

 bital counts for bothoids are as follows (preorbital -t- suborbit- 

 als): Scophthalmidae 1+5; Brachypleura 1 + 0; Paralichthys 

 group 1 + 4-5; Pseudorhombus group 1 + 5-7; Cyclopsetta 

 group 1 + 5-6; and Bothidae + 3-5. Pleuronectines were not 

 examined for this character. The most common count in both- 

 oids other than bothids is 1 + 5-7. Thus, there is some evidence 

 that the basal or plesiomorphic count for bothids may be five 

 suborbitals on the blind side. Among bothids this count appar- 

 ently occurs only in Taeniopsetta and Pelecanichthys. Engyo- 

 phrys, Trichopsetta, Perissias, and Monolene have three sub- 

 orbitals on the blind side. In summary, there is good evidence, 

 at least for the first two characters discussed above, that the 

 Taeniopsettinae show states that are plesiomorphic for the fam- 

 ily and may not be monophyletic. 



Pleuronectidae.— Norman (1934) considered this family to be 

 one of the "higher" flatfish groups, i.e., those with a mono- 

 morphic optic chiasma. Hubbs (1945) basically followed this 

 interpretation, but showed that two of Norman's pleuronectid 

 genera, Brachypleura and Lepidoblepharon. possessed some 

 primitive states not shown in other pleuronectids. These two 

 genera were removed by Hubbs and placed in his family Cith- 

 aridae. 



Norman (1934) defined the Pleuronectidae as being dextral 

 and having eggs without oil globules. Basic to his concept of 

 this family were the assumptions that all members were mono- 

 morphic in regard to the optic chiasma and that nearly all species 

 were discriminately dextral. He divided the family into five 

 subfamilies. All members of the Poecilopsettinae, Paralichthod- 

 inae, Samarinae. and Rhombosoleinae, as presently interpreted, 

 are discriminately dextral, i.e., sinistral individuals occur so 

 rarely in any one species that they can be considered anomalies. 

 Most species of Pleuronectinae are also discriminately dextral. 

 The few exceptions have probably returned to indiscriminate 

 ocular asymmetry secondarily (Hubbs and Hubbs, 1945). We 

 have no reason to doubt Norman's or Hubbs' assumption that 

 the Pleuronectinae have a monomorphic optic chiasma. How- 

 ever, as previously discussed, there are no data showing this for 

 the other pleuronectid subfamilies. Uniting these groups in the 

 family Pleuronectidae appears to have rested only on ocular 

 asymmetry. We have surveyed these subfamilies for various 



