676 



ONTOGENY AND SYSTEMATICS OF FISHES- AHLSTROM SYMPOSIUM 



NS(EP?) UN 



PHYP 



Fig. 359. Caudal skeleton oi Psetlodes bennctti. Hypural pattern 1. 

 EP = epural, HY 1-5 = hypurals 1-5, NS = neural spine, PHY = par- 

 hypural. PHYP = parhypurapophysis, PL) 2, 3 = prcural centrum 2, 3, 

 THC = terminal half centrum. UN = uroneural. Redrawn from Monod 

 (1968). 



flatfishes. It can be characteinzed as follows (Fig. 359): a par- 

 hypural with a haemal arch and parhypurapophysis; five au- 

 togenous hypurals; two pairs of uroneurals, i.e., pairs of stegurals 

 and splinter bones; two epurals, the first between the neural- 

 arch remnants of the second preural centrum; terminal half 

 centrum, i.e., fusion of two ural centra and the first preural 

 centrum; haemal spine of the second preural centrum autoge- 

 nous; haemal spine of the third preural centrum fused; and 24- 

 25 caudal rays, 17 principal, 15 branched. The caudal skeleton 

 of Psetlodes has been labelled as basically percoid (e.g., Wu, 

 1932; Monod, 1968; Amaoka, 1969). It should be noted here 

 that the neural spine of the second preural centrum is interpreted 

 as probably a captured epural, and that apparently only one free 

 epural remains. This is one of the more important differences 

 between Psetlodes and all other pleuronectiforms, which have 

 a neural spine on the second preural centrum and apparently a 

 basal number of two epurals. There are at least two hypotheses 

 which may explain this difference: (1) The earliest pleuronec- 

 tiforms may have had three free epurals, the anteriormost be- 

 coming wedged in the neural-arch remnant on the second preur- 

 al centrum (i.e., captured) and, thus forming a secondary neural 

 spine. In Psetlodes the remainmg epurals were fused (Amaoka, 

 1 969) or one was lost, while both were retained in the remaining 

 flatfishes, at least primitively. (2) The earliest pleuronectiforms 

 had two epurals, the anteriormost being captured in Psetlodes, 

 leaving one free epural. In the remaining flatfishes a neural spine 

 on the second preural centrum was acquired by fusion of this 

 vertebra with an anterior one bearing a spine. Rosen (1973) has 

 discussed the second hypothesis to account for secondaiy ac- 

 quisition of a neural spine on the second preural centrum and 

 offered as evidence the frequent occurrence of double spines on 

 the second preural centrum. Such anomalies are frequent in 



pleuronectiforms (see Cole and Johnstone. 1902; Barrington, 

 1937;Chabanaud, 1937;Amaoka, 1969; Okiyama, 1974;Futch, 

 1977; Fig. 360H). However, although a detailed survey for these 

 doubled spines has never been done, it appears that doubled 

 neural spines on this vertebra are just as frequent as doubled 

 haemal spines. 



In spite of the work that has been done on pleuronectiform 

 caudal osteology, there is still little agreement on interpretation 

 of some structures. We cannot solve these problems here or 

 discuss them in great detail. Most of these differences in inter- 

 pretation concern certain epaxial elements. More detailed com- 

 parative work needs to be done on these elements before ho- 

 mologies can be determined. For example, there is one 

 interpretation that uroneurals occur only in Psetlodes (Ahl- 

 strom). However, what appear to be remnants of a stegural may 

 remain in Cilharoides, Lepidoblepharon, Scophlhalmus. and 

 some achirines (Fig. 361; Amaoka, 1969; Hensley, pers. ob- 

 serv.). Although sufficient comparative work has not been done 

 to treat these dorsal structures across all lines of ffatfishes, within 

 certain groups we can be fairly sure of homologies, due to certain 

 consistent patterns of placement and shape and to some larval 

 work where fusions have been observed. 



In regard to neural and haemal spines of the second preural 

 centrum, the parhypural, and hypurals, our knowledge rests on 

 firmer ground. Characteristics of these structures have been 

 widely surveyed and there is much more agreement on inter- 

 pretation of homologous states. We interpret autogenous neural 

 and haemal spines on the second preural centrum, retention of 

 a parhypurapophysis and haemal arch on the parhypural. and 

 articulation of the parhypural with the terminal half centrum 

 as plesiomorphic for the order. 



Several patterns of fusions occur in regard to hypurals 1-4. 

 Hypural 5 moves to an epaxial position during ontogeny in 

 flatfishes (Figs. 360, 362), and its fate is more properly discussed 

 in reference to fusion (or lack of it) with epurals. The most 

 primitive condition is where hypurals 1-4 are not fused to the 

 terminal half centrum or among themselves (pattern 1; Figs. 

 359, 363 upper). 



There are three patterns which are slightly different from each 

 other. The interpretation of these is not so obvious, and we are 

 hesitant here to make statements concerning homologies be- 

 tween groups. One of these (pattern 2) is where hypurals 3 and 

 4 are fused to the terminal half centrum (Fig. 36 1 ). This pattern 

 is shown by Citharoides and apparently some Achirinae. In 

 some achirines, a somewhat different pattern (3) occurs where 

 hypurals 2, 3, and 4 are fused to the terminal half centrum (Fig. 

 363 middle). A fusion of hypurals 1-4 to the terminal half 

 centrum (pattern 4) is found in the Soleinae, Cynoglossidae, one 

 cithand (Eucilharus), and two genera of Rhombosoleinae (Pel- 

 torhamphus. Rhoinbosolea\ Figs. 362, 363 lower). Caudal-fin 

 development in a soleine is illustrated in Fig. 362. 



Another pattern of hypurals (5) is unique to the Samarinae 

 (Fig. 364). There are two ways to interpret this pattern. Here 

 the central hypurals (2 and 3 or 2-4) are fused to the terminal 

 half centrum. However, unlike the patterns previously de- 



Fig. 360. Caudal-fin structure of Engyophrys senla larvae (A-F). juveniles and adults (G-H). Standard lengths of specimens: (A) 4.6 mm; (B) 

 5.5 mm; (C) 7.0 mm; (D) 7.6 mm; (E) 7.7 mm; (F) 15.3 mm; (G) 45.7 mm; (H) 82.4 mm. NC = notochord. other abbreviations as in Fig. 359. 

 Redrawn from Hensley (1977). 



