Dinoflagellate phylogeny revisited: reconciling morphological and molecular based phylogenies

Abstract

Ultrastructural and molecular phylogenetic data suggest that dinoflagellates diverged as a lineage possibly as early as the Precambrian. However, the fossil record is problematic before the Mesozoic. From the mid Triassic, though, the fossil record of dinoflagellates is a rich source of information on Mesozoic-Cenozoic dinoflagellates, especially the gonyaulacoids and peridinioids. From the sequence of appearance of species and tabulation types and the impression of early morphological experimentation and later stabilization, the early Mesozoic radiation of dinoflagellates appears to be a real evolutionary event: indeed, dinoflagellate morphology as we know it today may originate in that event. This would explain why it is so difficult to interpret earlier fossils as dinoflagellates. However, that the dinoflagellate lineage existed in some form in the pre-Mesozoic is supported by biogeochemical data, early results of which indicate that certain early Paleozoic acanthomorph acritarchs may belong to the lineage.

A surprising degree of consistency is observed between ultrastructural (including tabulational), coarse biochemical and molecular sequence data. For example, sequence data provided by small subunit (SSU) rRNA support the hypothesis of progressive loss of histones within the dinoflagellates. Gymnodinioids have long been considered to be polyphyletic but are thought of generally as forerunners to the strongly thecate groups such as gonyaulacoids and peridinioids. In molecular trees they appear in both early-derived and late-derived positions, but mostly the latter. SSU data clearly support the gonyaulacoid/peridinioid ordinal separation, as does the fossil record. Prorocentroids are now thought to be the among the most derived dinoflagellates (and presumably the morphologically similar dinophysoids), but SSU sequences have so far failed to resolve the relationships of most gymnodinioids, peridinioids and prorocentroids (the so-called GPP complex) to one another. However, they do suggest the origin of prorocentroids from peridinioids rather than gonyaulacoids and that gymnodinioids probably had several origins.