• Ostariophysi;
  • Gonorynchiformes;
  • Cypriniformes;
  • Characiformes;
  • Siluroidei;
  • Gymnotoidei;
  • Siluriformes;
  • phylogenetic systematics

The history of ostariophysan classification is summarized and it is noted that traditional concepts of relationships have never been supported by characters found to be unique to the taxa. We present a new hypothesis of relationships among four of the five major ostariophysan lineages: Cypriniformes, Characiformes, Siluroidei, and Gymnotoidei (Otophysi). Cypriniforms are the sister-group of the remaining three (Characiphysi), and characiforms are the sister-group of siluroids plus gymnotoids (Siluriformes). Placement of the Gonorynchiformes as the sister-group of the Otophysi is supported by additional evidence. Each of the five lineages is monophyletic. Analysis was concentrated upon species thought to be the least specialized within each lineage; choices of these species are discussed.

Chanos is determined to be a relatively primitive gonorynchiform morphologically and the sister-group of all other Recent members of the order. Opsariichthys and Zacco are found to be morphologically primitive cypriniforms. We propose that a monophyletic group comprising the Citharinidae and Distichodontidae forms the sister-group of all other characiforms. Within the two families, Xenocharax is the least specialized. We suggest that Hepsetus, the erythrinids, and the ctenoluciids are more derived than the distichodontids and citharinids, and may form a monophyletic group within die characiforms. The traditional hypothesis that Diplomystes is the primitive sister-group of all Recent siluroids is substantiated. Our evidence suggests that Sternopygus is the most primitive gymnotoid morphologically; but rather than being the sister-group of all other gymnotoids, it is the primitive sister-group within a lineage called the Sternopygidae by Mago-Leccia.

Previous explanations of otophysan distribution have been based on notions of relationships which are unsupported by the evidence presented herein. Our own analysis of relationships serves primarily to make clear the extent of sympatry, and therefore the probability of dispersal, among the major ostariophysan lineages. The extent of sympatry, together with the widespread distribution of ostariophysans, suggests that the group is older than previously supposed, and our hypotheses of relationships among the characiforms implies that many of the extent characiform lineages evolved before the separation of Africa and South America. Further understanding of ostariophysan distribution must await phylogenetic analysis within each of the five major lineages so that distributions linked with vicariance patterns and dispersal events can be sorted out.