Evolutionary Biology 31, 105-154 (2000).
The Origin of the Mineral Skeleton in Chordates

Jerzy Dzik
Instytut Paleobiologii PAN, Twarda 51/55, 00-818 Warszawa, Poland. e-mail:

Conclusions. There are several anatomical traits shared by the most primitive Early Cambrian chordates Yunnanozoon, the Middle Cambrian probable westergaardodinid Odontogriphus, and the Vendian problematicum Dickinsonia, and their body plans resemble that of the nemerteans. The rhynchocoel of the nemerteans and (at least) the most anterior dorsal chamber of Dickinsonia may appear thus homologous, which is consistent with the idea of its homology with the notochord and myocoel of the chordates. The phosphatic elements of the westergaardodinid ‘paraconodont’ apparatuses seem to represent a stage in the development of mineral skeleton transitional between nemertean stylets and conodont elements. Odontogriphus had an oral grasping apparatus composed of such mineralized, probably phosphatic elements. Some kind of oral apparatus of unknown internal structure also occurred in Yunnanozoon, which probably also bore complex sensory head organs. Dickinsonia, Odontogriphus, Yunnanozoon, and Pikaia may thus represent a developmental series, from muscular dorsal chambers to the organization of myomeres typical for all later chordates (Fig. 13). In this respect even cephalochordates and tunicates are anatomically highly derived. Amphioxus and the tunicates share the presence of the filtratory basket surrounded by the atrium, which indicates that they are closely related and evolutionarily late. The earliest, more or less reliable, fossil evidence of this structure is in the Carboniferous Typhloesus. It shares with Recent salps a very characteristic concentration of the alimentary tract in a globular body (although bilobed in Typhloesus). Notably, the latest of the anaspid agnathans with weakly mineralized body covers, the Late Devonian Legendrolepis, shows a very elaborate gill apparatus (Arsenault and Janvier, 1991), suggestive of filtratory adaptations. The time order in appearance of these anatomies is consistent with the idea that salps and Amphioxus derived from chordates related to the anaspids. This implies a more complex anatomy of the common ancestor of the tunicate and cephalochordate evolutionary branches than usually assumed, but is consistent with neontological data by Lacalli (1996) and Williams and Holland (1996). The mode of secretion of the conodont crown tissue, with its early mineralization, and distribution of shapes of the secretory cells suggest that localized intensity of secretion of calcium phosphate, instead of cell migration, was the main factor controlling the morphology of elements. Calcium phosphate secretion was extremely high at denticle tips which requires (as phosphate and calcium ion transport took place above the surface of the element) a relatively thick cover of soft tissue to fulfill supply needs. In typical conodonts, this tissue was probably protected on the outside with horny caps. The primitive panderodontid conodonts had a grasping apparatus composed of at least seven element pairs morphologically similar to that of the Cambrian and Recent chaetognaths. Any homology of their crown tissue or the organic cap, with the grasping spines of the chaetognaths (‘protoconodonts’) is unlikely. The typical organization of the conodont apparatus originated in the early Ordovician within the clade of coniform Protopanderodontida. Their apparatuses were composed of a pair of incisor-like elements in front, a set of four pairs of relatively gracile elements connected into a single unit by one posteriorly located symmetrical element, and two pairs of robust elements hidden within the throat. This apparatus architecture did not undergo any basic modifications in the ozarkodinid and prioniodontid clades. In the most advanced prioniodontids (e.g., Promissum) the apparatus was probably evertible, with the throat being sinuously folded at the resting position. Elements of some locations were doubled in these conodonts. A homology between the organic caps of elements of advanced conodonts and horny jaws of Recent hagfish is likely, which would make the Myxinoidea close relatives of the conodonts or even members of their class. The homology between the crown and basal filling tissues of the conodonts and the enamel and dentine of vertebrates, respectively, implies that the developmental mechanisms which arose at the origin of the conodonts were subsequently used by the agnathans to build up their protective dermal scales.