684 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



characters and are confident that the Pleuronectidae as currently 

 defined are not monophyletic. In fact, four of the pleuronectid 

 subfamilies are not bothoids as we define the group. However, 

 what the true relationships of these groups are is unknown. We 

 discuss these subfamilies individually: 



Poecilopsettinae.— We have examined radiographs of speci- 

 mens ofPoecilopsetta and Nematops. These genera have hypural 

 pattern 1, at least one free epural, 20 caudal rays, and what 

 appears to be a haemal-arch remnant on the parhypural. The 

 caudal structure here is primitive compared to the bothoids and 

 these fishes do not belong to that group. Poecilopsettines are 

 poorly known and character states defining the group or relating 

 it to others have not been investigated. 



?ar?i\ic)M\iod.mdie:. —Paralichthodes algoensis has hypural pat- 

 tern 1 (Ahlstrom, pers. observ.) and does not belong to the 

 bothoid group. Its relationships to other groups are unknown. 



Samarinae.— Since Hubbs' (1945) removal oi Brachypleura and 

 Lepidoblepharon from this group, it has been composed of Sa- 

 maris and Samanscus. We have not done a detailed study of 

 these genera, but some characters we have examined are worthy 

 of note: ( 1 ) These genera show a unique hypural pattern (5; Fig. 

 364 upper). We interpret this pattern as derived relative to 

 pattern 1 and as indicative that the group is monophyletic. Using 

 this pattern to relate the group is more difficult: however, one 

 of us (Ahlstrom) noted that in late-stage larvae of Samanscus. 

 hypural pattern 1 is present, and fusions resulting in pattern 5 

 must occur very late in development. This is evidence that 

 pattern 5 may have evolved directly from pattern 1 and does 

 not represent a modification of the bothoid pattern 6. (2) Sa- 

 marines are the only pleuronectiforms known other than the 

 Bothidae to have intermuscular bones, although they do not 

 have the two series of myorhabdoi as found in bothids. We have 

 not done a detailed study of these bones in samarines, but they 

 appear very similar to the epimerals, epicentrals. and hypom- 

 erals of bothids. (3) Samarines, cynoglossids, and soleines have 

 an anterior pair of well-developed transverse apophyses on many 

 vertebrae. Two pairs of these structures are found in the Both- 

 idae and Scophthalmidae. (4) The Samannae, Soleoidei, and 

 Mancopselta lack postcleithra, at least in adults. How to inter- 

 pret these last three character states is open to question. Are 

 three of the series of intermuscular bones homologous in sa- 

 marines and bothids? Are the anterior vertebral transverse 

 apophyses homologous between all of the groups? Do some of 

 these character states indicate a close relationship between sa- 

 marines and some soleoids (i.e., cynoglossids and soleines)? Our 

 tentative hypothesis is that the samarines are a line that is at 

 least independent from the bothoids. Here we are obviously 

 stressing caudal characters. The corollary of this is that we are 

 interpreting similarities between samarines and bothoids in in- 

 termuscular bones and vertebral transverse apophyses as hom- 

 oplasies. 



Rhombosoleinae. — The main character states used by Norman 

 (1926, 1934) to define this subfamily were the high degree of 

 asymmetry in the ventral fins and the absence of pectoral radials. 

 The ocular ventral fin is on the midventral line and its base is 

 considerably extended. The blind ventral fin is short based or 

 missing. Another interesting characteristic of this group is that 



several genera show high numbers of fin rays in the ocular ven- 

 tral fin. There is a great deal of morphological diversity in rhom- 

 bosoleines. Some genera appear fairly generalized in many char- 

 acters (Oncopterus. Psammodiscus. Rhombosolea. Azygopus. and 

 Pelotretis); others are more specialized (Colistiuin. Peltorham- 

 phits, and Ammotretis). Many of the specializations in the latter 

 genera are similar to those in some soleoids. This has been 

 interpreted as parallel evolution (Norman, 1934; Hubbs, 1945). 

 Norman apparently had some doubts about aligning this group 

 with the Pleuronectinae. He realized that Parker's (1903) ex- 

 amination of one specimen of Oncopterus darwinii in his survey 

 of optic chiasmata did not prove the group to be monomorphic 

 in this character. This group has still not been studied in detail. 

 It may be monophyletic, but its relationship to other flatfishes 

 is unknown. 



We have examined the caudal skeleton of all rhombosoleine 

 genera except Psammodiscus. They show hypural patterns 1 and 

 4 (Fig. 363 upper and lower). Assuming the group is monophy- 

 letic, there are two implications here: (1) The primitive pleu- 

 ronectiform hypural pattern 1 is also plesiomorphic for the 

 Rhombosoleinae, and the derived pattern 4 arose within the 

 group independently from the same pattern in the Soleinae, 

 Cynoglossidae, and Eucttharus. (2) The Rhombosoleinae are 

 not bothoids and should not be aligned with the Pleuronectinae. 



The possibility has recently become apparent that Mancop- 

 setla may be most closely related to the Rhombosoleinae. All 

 known specimens of Mancopselta are sinistral and it has been 

 considered a bothid. However, it shares certain character states 

 with at least some rhombosoleines. This genus has ventral-fin 

 ray counts of 7 on the ocular side and 5-7 on the blind side. 

 Although not strictly limited to the rhombosoleines, these high 

 counts, at least in the fin of the ocular side, are characteristic of 

 at least four rhombosoleine genera. The eyes are densely scaled 

 in Mancopsetta and in Azygopus and Pelotretis. However, scaled 

 eyes are found in some genera of other groups also (e.g., some 

 pleuronectines). Andnashev (1960) and Penrith (1965) have 

 both remarked on a fleshy lip-like structure which overhangs 

 the anterior end of the upper jaw in .Mancopsetta. One of the 

 soleoid-type characteristics exhibited by the more specialized 

 rhombosoleines is the dorsal fin originating in a rostral hook 

 that overhangs the mouth. In the more generalized genera, there 

 is no rostral hook and the dorsal fin originates at some posterior 

 position. In at least one of these generalized genera (Azygopus. 

 the only one examined for this character) there is a fleshy struc- 

 ture (possibly a precursor to the rostral hook?) overhanging the 

 anterior end of the upper jaw which is very similar to that in 

 Mancopsetta. Obviously more comparative work needs to be 

 done here. However, it is possible that Mancopsetta and the 

 Rhombosoleinae may form a monophyletic group with an in- 

 discriminately dextral or sinistral common ancestor. 



Pleuronectinae. — Norman (1934) stressed two character states 

 in defining this subfamily: ( 1 ) lateral line well developed on both 

 sides of the body; and (2) olfactory laminae parallel (except in 

 Atheresthes), without rachis. A well-developed lateral line on 

 both sides of the body is plesiomorphic for the order and both- 

 oids. We have not examined olfactory laminae or attempted to 

 analyze distributions of states for the character. 



We have shown that the Pleuronectidae is probably not mono- 

 phyletic, due to the inclusion of the four non-bothoid subfam- 

 ilies. The subfamily Pleuronectinae is the only bothoid group 



