Reinterpreting recapitulation: systematics of needlefishes and their allies (Teleostei: Beloniformes).

As needlefishes (Belonidae) grow, their jaws pass through a "halfbeak" stage that resembles the adult jaw condition of the closely related family of halfbeaks (Hemiramphidae). Based on this pattern, some authors have suggested that halfbeaks are "developmentally arrested" or paedomorphic needlefish derivatives, whereas others have supported the notion that needlefishes are descended from halfbeak-like ancestors and that needlefish ontogeny thereby recapitulates phylogeny. To test these ideas and to better understand evolutionary changes in jaw ontogeny, phylogenetic relationships among genera of needlefishes, sauries (Scomberesocidae), halfbeaks, and flyingfishes (Exocoetidae) were assessed using mitochondrial (cytochrome b and 16S), nuclear (Tmo-4C4), and morphological characters. The resultant tree provides several novel taxonomic findings: (1) flyingfishes appear to be nested within halfbeaks; (2) sauries appear to be nested within needlefishes; and (3) the Indo-West Pacific freshwater halfbeaks appear to be most closely related to the needlefish/saury clade. The structure of the tree falsifies the idea that halfbeaks are paedomorphic needlefishes. Instead, halfbeaks are basal relative to needlefishes, fitting the pattern predicted by the hypothesis of recapitulation. I discuss limitations to phylogenetic perspectives on recapitulation based on discrete character data by comparing aspects of von Baerian and Haeckelian views of the relation between ontogeny and phylogeny.