Phylogenetic Relationships of Pleuronectiformes Based on Molecular Evidence

Phylogenetic analyses based on nucleotide sequence data of the 12S and 16S mitochondrial ribosomal genes were performed for representatives of the Pleuronectiformes (flatfishes). Two parsimony analyses with percomorphs and basal acanthopterygians designated as outgroups were conducted; equal weighting of all nucleotides and loop regions analyzed for transversions only. A bayesian analysis was performed under the likelihood model TrN+I+Γ. A monophyletic Pleuronectiformes was resolved in all analyses. Results are consistent with the current higher level classification, recognizing the suborders Psettodoidei and Pleuronectoidei. Psettodoidei consists of a single family, Psettodidae, that is sister to all other flatfishes. Within the Pleuronectoidei three clades are identified. Bothids plus paralichthyids (in part) form a clade in all analyses. A solelike clade containing citharids, cynoglossids, samarids, soleids, Trinectes (Achiridae), and Poecilopsetta (Poecilsopsettidae), is identified in the parsimony analyses. However, a solelike clade was not recovered in the bayesian analysis. A flounderlike clade is identified containing Pleuronectidae and Paralichthyidae (in part) in all analyses. The position of Scophthalmus (Scophthalmidae) differed among all analyses. The results further support the conclusion that eye position is not a useful source of information for the classification of flatfishes. This study provides a new framework for future studies.

Fig. 1.

(A) The Regan-Norman model of pleuronectiform relationships by Hensley and Ahlstrom (1984). (B) Hypothesis of relationships among Pleuronectiformes proposed by Chapleau (1993). (C) Hypothesis of relationships among Pleuronectiformes by Cooper and Chapleau (1998a)

Fig. 2.

Scatter-plots of the number of mutations versus uncorrected (p) distance coefficients in pairwise comparisons. (A) Transversions and transitions in 12S stem regions. (B) Transversions in 12S loop regions. (C) Transitions in 12S loop regions. (D) Transversions and transitions in 16S stem regions. (E) Transversions in 16S loop regions. (F) Transitions in 16S loop regions. (o) Transitions. (△) Transversions

Fig. 3.

The strict consensus tree resulting from 32 equally most-parsimonious trees with all nucleotides weighted equally. Numbers above nodes indicate bootstrap values (1000 pseudoreplicates) and numbers below nodes indicate Bremer decay indices. TL = 3391, CI = 0.3151, HI = 0.6849, and RI = 0.4494

Fig. 4.

The strict consensus tree resulting from eight equally most-parsimonious trees with loop regions analyzed for transitions only. Numbers above nodes indicate bootstrap values (1000 pseudoreplicates), numbers below nodes indicate Bremer decay indices, and bottom numbers indicate branch lengths. TL = 2560, CI = 0.2960, HI = 0.7040, and RI = 0.4667

Fig. 5.

Fifty percent majority rule tree produced by a bayesian analysis under TrN+I+Γ. Posterior probabilities from 19,800 trees are indicated on the branches

Received: 26 Sept 2000

Accepted: 14 Mar 2002