07L Pleuronectiformes: Relationships D. A. HENSLEYAND E. H. AHLSTROM ASICS of the current working model for evolution of pleu- article, this volume) and consider it the working hypothesis to B ronectiforms were proposed by Regan (1910, 1929) and be reexamined using adult, larval, and egg characters. Norman (1934). In his monograph, Norman treated the floun- Formation of the Regan-Norman model involved an eclectic ders (Psettodidae, Bothidae. Pleuronectidae), and though he did approach, Le., a combination of phyletic and phenetic methods. not publish a revision of the remaining pleuronectiforms, his Although some of the groups currently recognized appear to be key and classification of the soleoids were published posthu- based on synapomorphies, many are clearly based on symple- mously (1966). Norman's model and classification with the siomorphies and were recognized as such by the authors. This modifications of Hubbs (1945), Amaoka (1969), Futch (1977), search for horizontal relationships among pleuronectiforms us- and Hensley (1977) represent the most recent, detailed hypoth- ing eclectic methods, with one exception, has been the only esis for pleuronectiform evolution. We will refer to this as the approach used in this group. The exception is the recent work Regan-Norman model (Fig. 358) and classification (preceding of Lauder and Liem (1983) in which a cladogram for flatfishes HENSLEY AND AHLSTROM: PLEURONECTIFORMES 67 I ma,c a, C m a, 2 mC mW -a, v) a a. 0 mC - r% m ca Lm m0 a, 0 - cz - 0 5 Psettodidae Scophthalmidae 0 m in 0 in I II 1 Bot hl dae sol Cvno ssidae Clt idae Psettodoldel Fig. 358. Current hypothesis for interrelationships of pleuronectiform fishes. Based on Norman (1934. 1966). Huhhs (1945). and Amaoka (I 969) is presented. However, these authors present this as a tentative order can be subjected to an in-depth cladistic analysis. Until hypothesis and admit that the interrelationships expressed are this work is completed, it is premature to offer a new hypothesis still problematic. Most of the character states they use are re- of interrelationships for the entire order. ductive, few characters were analyzed, and the authors were understandably unaware of recent character surveys, since much Adult characters of this information is unpublished. Several criteria were used for selecting characters for discus- We have made the assumption that the order Pleuronecti- sion: (1) amount of information available on the distribution formes is monophyletic and the sister group is the remaining of character states: (2) characters commonly used in the past to percomorph fishes (sensu Rosen and Patterson, 1969 and Rosen, define groups of pleuronectiforms; (3) those for which our 1973). Although the monophyly and origin of the group is still knowledge of distributions of states is limited, but appear to open to question and hypotheses of multiple origins have been indicate groupings different from those hypothesized in the proposed (e.g., Kyle, 192 1; Chabanaud, 1949; Amaoka. 1969), working classification and which need additional study: and (4) a monophyletic model with a percomorph sister group still ap- characters which are well known in certain groups and are po- pears to be the most parsimonious. In other words, with the tentially useful for elucidating relationships within these groups. information available, there appears to be no need to hypoth- Characters and character complexes used in this study are dis- esize multiple origins for flatfishes: to do so demands the inclu- cussed below. Characters and states are presented in Table 179. sion of a great deal of convergence. Optic chiasma. -The relationship between the optic chiasma RELATIONSHIPS and ocular asymmetry of pleuronectiforms has been investigated The following discussion of relationships within the pleuro- by several workers beginning mainly with the work of Parker nectiforms is cursory and preliminary. In fact, it asks more (1 903). Hubbs (I 945) examined this relationship further and questions than it answers and illustrates that more work (par- presented all data from previous studies. Parker found that most ticularly osteological) is needed in certain groups before the fishes have a dimorphic optic chiasma, Le., the nerve of the left 672 ONTOGENY AND SYSTEMATICS OF FISHES-AHLSTROM SYMPOSIUM or right eye is dorsal with about equal frequency (state referred truly dimorphic. Soleoids became discriminate (soleids dextral to here as truly dimorphic). Exceptions to this are species of and cynoglossids sinistral), but retained a truly dimorphic chias- paralichthyids (sinistral) and pleuronectines (dextral) where the ma. Psettodids remained indiscriminate and truly dimorphic. right or left optic nerve, respectively, is always dorsal, even in Citharids and presumably scophthalmids became discriminate reversed individuals, i.e., the optic chiasma is monomorphic. (scophthalmids and citharines sinistral and brachypleurines The Soleidae and Cynoglossidae, however, retain a truly di- dextral) but retained some ontogenetic plasticity in regard to morphic optic chiasma. Subsequent work by Regan (1910) and the optic chiasma, since reversed individuals still have the nerve Hubbs (1945) showed that in the indiscriminately dextral or of the migrating eye dorsal (basically dimorphic). The remaining sinistral Psettodes the optic chiasma is also truly dimorphic. In pleuronectoids became discriminate (Paralichthyidae and Both- addition, Hubbs presented evidence of a third state, at least in idae sinistral and Plueronectidae dextral) and evolved a mono- Citharoides (sinistral), where the nerve of the migrating eye is morphic chiasma. The only exceptions with regard to ocular dorsal even in reversed individuals. He thus interpreted the asymmetry are certain indiscriminate paralichthyids and pleu- Citharidae as having a basically dimorphic optic chiasma and ronectines. However, most of these indiscriminate pleuronec- predicted the same for scophthalmids, although apparently no toids have been shown to have a monomorphic optic chiasma one has examined a reversed scopthalmid to test this prediction. (a possible exception is Tephrinectes). It would thus appear that A truly dimorphic optic chiasma as found in Psettodes and the indiscriminate ocular asymmetry in pleuronectoids developed soleoids has been interpreted as plesiomorphic for pleuronec- secondarily from discriminate ancestors (Hubbs and Hubbs, tiforms. The type of optic chiasma found in Citharoides and 1945). predicted for scophthalmids (Le., nerve of the migrating eye Making phylogenetic interpretations from two states of ocular always dorsal) was interpreted as an intermediate state between asymmetry is difficult or impossible without corroborative evi- the truly dimorphic and the monomorphic chiasmata as found dence. Thus, a statement to the effect that two or more dextral in pleuronectoids. We agree with this interpretation of polarity. (or sinistral) pleuronectoid groups are most closely related to However, some plesiomorphic states have been used to define each other because they are dextral (or sinistral) without addi- groups: Psettodidae, truly dimorphic; Citharidae, basically di- tional evidence of synapomorphies is circular, and may lead to morphic; Scophthalmidae, predicted to be basically dimorphic; the recognition of polyphyletic groups. This reasoning was the and Soleoidei, truly dimorphic. basis for the proposed close relationship in the Regan-Norman Major problems exist with the use of the optic chiasma for model between the Pleuronectinae and the remaining pleuro- phylogenetic inference. One of these concerns the feasibility of nectid subfamilies (Poecilopsettinae, Rhombosoleinae, Samar- actually determining which state exists in a group. Demonstrat- inae, Paralichthodinae) and for treating the genera Mancopsetta ing the occurrence of truly dimorphic chiasmata is relatively and Thysanopsetfa as members of the Bothidae and Paralich- simple. All that is needed is to show that either optic nerve is thyidae, respectively. dorsal regardless ofwhich eye has migrated; reversed individuals are not necessary. To demonstrate occurrence of the basically Ribs and intermuscular bones. - In pleuronectiforms that pos- dimorphic state, reversals are needed and the nerve of the mi- sess ribs, these appear to be homologous with the pleural and grating eye must always be dorsal. Likewise, reversed individ- epipleural ribs of other teleosts, and the presence of these bones uals must be examined to show a monomorphic chiasma. Here should be considered plesiomorphic for the order. Two groups the nerve to the right eye must be dorsal in all individuals lack both series of ribs, the Achirinae and apparently the Cyn- (including reversals) of normally sinistral species and the nerve oglossidae. Chabanaud (1 940) reports epipleural ribs in some of the left eye must be dorsal in all individuals of normally cynoglossids but mentions no genera or species. We have not dextral species. When one actually examines the data for this seen them in cleared-and-stained Symphurus species or in ra- character (see Hubbs, 1945), states have been determined for diographs of several Cynoglossus species. Although it is still very few pleuronectiform groups. The occurrence ofthe basically commonly believed that all soleoids lack both series ofribs (e.g., dimorphic state in the Citharidae was demonstrated in only one Nelson, 1976; Lauder and Liem, 1983), Chabanaud (I 940, 194 1) species. Ofgreater significance, however, is the fact that a mono- found