Molecular phylogeny of Anomalodesmata (Bivalvia)

Research project P 14356 Molecular phylogeny of Anolmalodesmata (Bivalvia) Gerald STEINER 06.03.2000 Anomalodesmata is an important bivalve sub-class of Ordovician origins. These marine bivalves include an extremely diverse array of highly specialised forms. A recent cladistic analysis of 43 morphological characters of 14 extant families produced an insufficiently resolved tree, leaving relationships of the tube-dwelling clavagelllids and of the carnivorous taxa obscure implying that convergence and/or parallel evolution of their morphologies have masked their phylogenetic relationships within the subclass and to other bivalves. Molecular data as an independent set of characters less influenced by similar habits and ecological niches than morphological features, have the potential to overcome problems of morphological convergence. The aim of this project is to produce a molecular dataset for all major groups of Anomalodesmata that will be tested against the morphological and palaeontological data. Molecular phylogenetic markers will be the complete 18S rDNA gene, a 1200 bp long fragment of the 28S rDNA, and the exons of the Elongation Factor-1alpha (about 1400 bp coding for 460 amino acids), assessing speciations from the Ordovician to the Tertiary. Live specimens will be collected in Sydney, Australia, and Iceland. In addition, rare species are available as ethanol-preserved museum material. Standard procedures for DNA preparation and PCR are available for the 18- and 28S rDNA genes, but the EF- 1alpha gene with its introns requires reverse transcription PCR and the establishment of new primers for working with preserved museum samples. DNA and amino acid sequences will be tested for substitution rate heterogeneity and submitted to parsimony and maximum-likelihood algorithms, both separately and combined. Branch support will be quantified by bootstrap, decay, and spectral analyses. Finally, a combined morphological and molecular analysis and a calibration against the fossil record will be attempted. The specific questions we hope to find answers are: 1) What is the sister group of the Anomalodesmata? 2) What taxa actually belong to this group and how are they related to each other? 3) Which morphological characters are reliable phylogenetic markers? 4) Can we date the origin of major lineages by calibrating molecular and fossil data? The answers to these are expected to elucidate not only the phylogeny of the Anomalodesmata but also their evolutionary history.

The results of our study found the Anomalodesmata as a monophylum rooted among basal heterodont bivalves. They are not related to the polyphyletic Myoida and are molecularly characterized by homologous insertions in the 18S rRNA gene. The Anomalodesmata are a relict group of mussels descending from ancestors at least 500 million year old. Most of the Recent species live in cryptic habitats or in the deep-sea which makes them difficult to collect. Their diversity of forms and life habits have posed severe problems for their systematics and the reconstruction of their origins and evolution, resulting in several competing morphological hypotheses. In the present study, we used three ribosomal genes (18S, 28S, 16S rRNA) to unravel anomalodesmatan phylogenetic relationships and assess evolutionary pathways of the morphological characters. The molecular data from 27 anomalodesmatan species, collected between Iceland, Brazil and Australia, strongly support their common origin (monophyly). The group evolved from early heterodont mussels. Therefore, the rank of a "subclass" equal to the other major bivalve lineages (e.g. Protobranchia, Pteriomorpha, Heteroconchia) does not apply. Relationships within Anomalodesmata are not in accordance with any morphology-based classifications. The taxa Lyonsiidae and Lyonsiellidae are distinctly diphyletic, whereas there is some signal for the monophyly of Cuspidariidae, Myochamidae, Laternulidae and Poromyidae. The present data are inconclusive whether or not carnivory evolved only once or several times independently. It became obvious, however, that many morphological adaptations were acquired independently in several groups. Among these are the loss of shell spicules and the fourth mantle opening, the differentiation of septibranch gills, and valve discordance. The similarity of morphological characters caused by convergent or parallel evolution rather than common descent enhances the importance of molecular data. However, our analyses support hitherto controversial palaeontological findings and data from hinge development in some anomalodesmatan groups. Following the publication of the 18S rRNA data, we are developing an interdisciplinary approach combining all molecular markers with morphological and palaeontological data in a total evidence analysis.