Reconstructing the Anomalodesmata (Mollusca: Bivalvia): morphology and molecules.

The extant anomalodesmatan bivalves have always proved rather enigmatic and difficult to interpret, both in terms of their relationships to other bivalve taxa and the interrelationships of the constituent families. These difficulties stem from their diverse and often highly specialized life habits which have resulted in a wide array of disparate morphologies, and also from the fact that many are extremely rare. Classifications based on morphological characters have been dogged by fears that convergent and parallel evolution has masked phylogenetic signals. Molecular surveys of members of 12 of the 15 constituent families, using the 18S rRNA gene, have revealed that anomalodesmatans are robustly monophyletic and lie within the basal heterodonts. The Anomalodesmata should no longer be regarded as a subclass, but as a part of the Heterodonta. Here we present an enhanced analysis of 32 anomalodesmatan species (representatives of 12 families). Our results, subjected to Maximum Parsimony, Maximum Likelihood and Bayesian analyses, challenge our understanding of the internal relationships within the Anomalodesmata. In particular they indicate the need for a re-distribution of the families traditionally placed in the Thracioidea and Pandoroidea into a ¿thraciid¿ lineage (Thraciidae + Cleidothaeridae + Myochamidae) and a ¿lyonsiid¿ lineage (polyphyletic Lyonsiidae + Clavagellidae + Laternulidae + Pandoridae). The endolithic Clavagella and endobenthic Brechites and Penicillus form a robust clade. The hypothesis that the carnivorous septibranchs are monophyletic can, thus far, be neither supported nor rejected. Mapping critical morphological characters onto our molecular results provides evidence of multiple loss of some characters (e.g. prismato-nacreous shell microstructure and shell spicules) and also multiple gain of others (e.g. chondrophores).