Development and morphology of rostral cartilages in batoid fishes (Ghondrichthyes: Batoidea), with comments on homology within vertebrates

The rostral cartilages of batoid fishes were examined to elucidate their development, morphology and homology. Comparison of a variety of rostral cartilages among elasmobranchs with other groups of vertebrates shows that rostral cartilages originate embryologically from the trabecula and/or lamina orbitonasalis. Because different morphogenetic patterns of the derivatives of the two embryonic cartilages give rise to a wide variety of forms of rostral cartilages even within elasmobranchs, and because morphogenesis involves complex interactions among participating structures in the ethmo-orbital area, we put forward conceptual and empirical discussions to elucidate the homology of the rostral cartilages in batoid fishes. With six assumptions given in this study and based on recent discussions of biological and historical homology, our discussions centre on: (1) recognition of complex interactions of participating biological entities in development and evolution; (2) elucidation of a set of interacting biological and evolutionary factors to define a given morphological structure; (3) assessment of causal explanations for similarities or differences between homologous structures by determining genetic, epigenetic and evolutionary factors. Examples of conceptual approaches are given to make the approaches testable. Although a paucity of knowledge of rostral cartilage formation is the major obstacle to thorough analysis of the conceptual framework, several tentative conclusions are made on the homology of rostral cartilages that will hopefully attract more research on development and evolution in vertebrate morphology. These are: (1) the rostral cartilage in each group of vertebrates examined can be defined by both developmentally associated and adult structural attributes, yet such data do not allow us to assess homology of a variety of forms of rostral cartilages at higher taxonomic categories; (2) the entire rostral cartilage in elasmobranchs is formed by the contribution of the embryonic trabecula and lamina orbitonasalis. The status of the development and homology of the rostral cartilage in holocephalans remains uncertain; (3) there is no simple picture of evolution of rostral cartilages among three putative monophyletic assemblages of elasmobranchs, galeomorphs, squaloids (possibly plus Squatina, Chlamydoselachus and hexanchoids as the orbitostylic group) and batoid fishes. It is highly likely that rostral cartilages in each subgroup or subgroups of these assemblages may be of phylogenetic significance but that it may not serve as a basis to unite these assemblages into much higher assemblages; (4) the tripodal rostral cartilage is unique in form in the group including some carcharhinoid and lamnoid sharks. The status of the analogous tripodal cartilage in some squaloids remains uncertain. The unfused tripodal cartilage of the electric ray Narke is interpreted as developmentally equivalent to, but not homologous with, the unfused or fused ones in the sharks; (5) the rostral cartilage in the electric ray Torpedo is uniquely formed because of its embryonic origin solely from the ventro-medial part of the lamina orbitonasalis, but it is regarded as homologous with the rostral cartilages which are formed by the trabecula and other components of the lamina orbitonasalis in other batoid fishes; (6) the cornu trabecula contributes to the formation of the ventral stem of the rostral cartilage at least in elasmobranchs, especially to a particular set of rostral cartilages, i.e. the tripodal rostral cartilage in the shark Scyliorhinus and dorso-ventrally flattened rostral shaft in the narcinidid electric rays; (7) there is a unique form of a rostral shaft with rostral appendix in skates and probably guitarfishes; (8) there is no rostral cartilage in adult benthic stingrays, pelagic stingrays Dasyatis violacea and Myliobatidae, although it is present in embryonic stages; (9) there is a unique form of the rostral cartilage as a rostral projection from the dorso-lateral part of the lamina orbitonasalis in pelagic stingrays Rhinopteridae and Mobulidae, which together with part of the pectoral fins, forms a pair of cephalic fins; (10) different developmental mechanisms may be responsible for the absence or loss of rostral cartilages in different groups, i.e. absence of the cartilage derived from the medial area of the trabecula in Torpedo vs absence of the rostral cartilage in benthic stingrays; (11) the rostral cartilages in some placental mammals (cetaceans and sirenians) arise only from the medial area of the trabecula because monotreme and placental mammals do not form the trabecula cranii; (12) some actinopterygians and sacropterygians possess a rostral cartilage which originates only from the medial area of the trabecula. One scombroid group, including Sardini and Thunnini, Scomberomorus, Acanthocybium, Istiophoridae and Xiphias, possesses a unique larval beak composed of the rostral cartilage, ethmoid cartilage and premaxillar bone. The development and homology of other rostral cartilages remain to be further elucidated; (13) urodeles possess a medial rostral process whose anlage is probably developmentally equivalent to that in batoid fishes but the occurrence in urodeles is either atavistic or unique (autapomorphic); (14) the upper jaw of tadpoles is unique in possessing the suprarostral cartilage; the anlage of the cartilage is probably developmentally equivalent to the outgrowth of the cornu trabecula in batoid fishes.