686 



ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM 



Relative time of caudal-fin formation.— In most known larvae 

 of flatfishes and other teleosts, formation of the caudal fin pre- 

 cedes or occurs with that of the dorsal and anal fins. The only 

 exceptions known in pleuronectiforms are the cynoglossids. In 

 this family, the caudal fin does not develop until the dorsal and 

 anal fins are nearly completely developed. This pattern of de- 

 velopment is considered apomorphic in pleuronectiforms. 



Eye migration and dorsal-fin position at tnetamorphosis. — Eye 

 migration has been observed in some flatfish groups. In the 

 Psettodidae, Pleuronectinae, Paralichthyidae (excluding the Cy- 

 clopsetta group), Scophthalmidae, and apparently some Sole- 

 idae, the first ray of the dorsal fin is above or posterior to the 

 eyes. At metamorphosis, the migrating eye crosses anterior to 

 the dorsal-fin origin. These types of eye migration and dorsal- 

 fin position appear to be plesiomorphic for the order. Several 

 derived states for these characters occur. In at least one species 

 of cynoglossid, a fleshy rostral beak is formed anterior to the 

 dorsal-fin origin. Eye migration takes place between the rostral 

 beak and the interorbital region. In some soleids, the dorsal-fin 

 origin projects above the snout and the eye migrates between 

 this projection and the neurocranium. In the Bothidae, the dor- 

 sal fin is anterior to the eye and attached to the ethmoid region. 

 During migration, the eye goes between the base of the dorsal 

 fin and the ethmoid region. A path for the migrating eye is 

 created by detachment of the anterior section of the dorsal fin 

 from the ethmoid region so that a narrow slit is formed, or some 

 tissue in the path of the migrating eye is absorbed. A very similar 

 type of eye migration occurs in some species of the Cyclopsetta 

 group. However, in other members of this group, the eye mi- 

 grates around the dorsal-fin origin (Gutherz, 1970; Tucker, 1982). 



Phylogenetic information provided by 

 larval characters 



Although larvae of some critical groups are unknown or poor- 

 ly known, some comments about phylogenetic relationships can 

 be made in regard to groups where our knowledge is on a higher 

 level. 



Bothoids. —Spines in the otic or frontal regions of the neuro- 

 cranium which are isolated or in small clusters appear to be 

 limited to various groups of bothoids. If these spines prove to 

 be homologous between these groups, they may be apomorphic 

 within the order. In this interpretation, they would be primitive 

 for bothoids and lost in various lines. 



Paralichthyidae.— As discussed in the section on adult charac- 

 ters, this family as currently interpreted is polyphyletic due to 

 the inclusion of Tephrinectes and Thysanopsetta. We do not 

 consider these genera bothoids as defined by the caudal-fin com- 

 plex. Their larvae are unknown. 



We have interpreted the Cyclopsetta group as monophyletic 

 based on some adult character states which are probably apo- 

 morphic. Although larvae of this group show certain states which 

 appear to be apomorphic within bothoids (e.g., elongated left 

 ventral-fin rays), not all species in this group show these. 



The Pseudorhombus group is possibly definable by adult syn- 

 apomorphies. In larvae of this group, we see no character states 

 that are presently interpretable with certainty as synapomor- 

 phies. 



In examining adult characters of the Paralichthys group, it 

 appeared likely that this group had no synapomorphies. Larvae 



tend to support this. They show the following character states 

 which appear to be plesiomorphic for the order: (1) presence of 

 preopercular spines; (2) origin of the dorsal fin behind the eyes; 

 (3) metamorphosis in a size range of 7.5-14.2 mm; and (4) eye 

 migration anterior to the dorsal fin. In addition, at least some 

 species show the following states which may prove to be ple- 

 siomorphic at least within the bothoids: (1) four or five elon- 

 gated, early-forming dorsal-fin rays; and (2) presence of otic 

 spines. 



Bothidae. — 'With the exclusion of Mancopsetta and inclusion of 

 Perissias, this family is definable by adult synapomorphies. Lar- 

 vae of the Bothidae are probably better known than for any 

 other family of flatfishes. However, larvae of many genera are 

 still unknown (i.e., Parabothus, Asterorhombus, Tosarhombus. 

 Neolaeops. Japonolaeops, and Perissias). Amaoka (1979) re- 

 viewed larval characters of most genera for which larvae are 

 known. Known bothid larvae show the following character states 

 which are interpreted as synapomorphies: (1) metamorphosis 

 at a relatively large size (ca. 15-120 mm); (2) eye migration 

 below the dorsal fin; (3) dorsal-fin origin anterior to eyes just 

 prior to metamorphosis; (4) elongated, early-forming second 

 dorsal-fin ray; and (5) lack of preopercular spines. 



Larvae of some bothid genera have various combinations of 

 otic-region, urohyal, cleithral, and basipterygial spines. It is 

 tempting to use the presence of these spines to define bothid 

 groups, and therefore, assume that they are apomorphic within 

 the family. Spines in the otic region within the Bothidae are 

 limited to the Taeniopsettinae as presently defined. However, 

 spines in this region occur in other bothoid groups. Although 

 sufficient comparative osteological work has not been done to 

 show that these spines are homologous between taeniopsettines 

 and other bothoids, use of these spines to infer close relation- 

 ships between Engyophrys. Taeniopsetta. and Trichopsetta is 

 questionable. Urohyal, cleithral, and basipterygial spines are 

 known only from larvae of nine bothid genera. They occur in 

 various combinations inter- and intragenerically. Amaoka (1969) 

 presented a model of intergeneric relationships for Japanese 

 bothids based on adult characters. Occurrence of these larval 

 spines is scattered among the bothid lines hypothesized by 

 Amaoka. This could indicate two possibilities: ( 1 ) the spines are 

 apomorphic within the family, and Amaoka's model is incor- 

 rect; or (2) Amaoka's model is correct and the spines are ple- 

 siomorphic within the family and have been lost in several lines. 

 Two major problems exist with Amaoka's phytogeny based on 

 adult characters; it was constructed using eclectic methods and 

 it did not include all genera. Interpretation of urohyal, basip- 

 terygial, and cleithral spines should await a cladistic analysis of 

 bothid interrelationships based on adult characters. 



Pleuronectidae. — Based on adult characters, we interpret this 

 family as polyphyletic. Larvae of the four non-bothoid subfam- 

 ilies are poorly known, and hence, of little aid in determining 

 relationships of these groups. However, there are certain simi- 

 larities in general body morphology between the few known 

 samarine and poecilopsettine larvae. In regard to the Pleuro- 

 nectinae, many adult states that are shared are plesiomorphic 

 for pleuronectiforms or bothoids. This also appears to be true 

 for most larval characters. The position of the dorsal-fin origin 

 (posterior to the eyes) and the type of eye migration (anterior 

 to the dorsal-fin origin) are plesiomorphic for the order. Some 

 pleuronectine larvae have preopercular spines, which again, are 



