Mussel of the Month Pseudomulleria dalyi is a cementing freshwater mussel currently classified on this web site in the family Etheriidae. That classification has been controversial for more than 20 years. Since its first inclusion in a molecular phylogenetic analysis by Bogan & Hoeh (2000), the monophyly and classification of the traditional family Etheriidae has been open for debate. We have previously discussed this issue multiple times: when Pseudomulleria dalyi was Mussel of the Month back in December 2005, the first time Mulleria rivoli was Mussel of the Month in February 2015, and in the peer-reviewed literature (Graf & Cummings, 2010). With a new generation of freshwater mussel systematists joining the conversation, the time is ripe to revisit the phylogeny of the asymmetrically-shelled, cementing/wedging freshwater mussel species — the “freshwater oysters.”
Traditional Classification
Since the pre-cladistic era through the 20th century, the traditional family Etheriidae has been composed of four species: Etheria elliptica from Africa, Mulleria rivoli and Bartlettia stefanensis from South America, and Pseudomulleria dalyi from India. Etheria is a fully-cementing, footless though dimyarian (i.e., with two adductor mussels) mussel, while both Mulleria (also known as Acostaea) and Pseudomulleria are even more fully oysterized and monomyarian. Bartlettia is more of a “wedger” than a cementer to hard substrates. Parodiz & Bonetto (1963) argued that Bartlettia was actually a highly modified mycetopodid, and this tracked with the anatomical observations of Woodward (1898), Yonge (1962, 1978) and Heard & Vail (1976).
Comparative Anatomy & Larvae
Etheriids like Etheria and Pseudomulleria share homologies with the family Mycetopodidae that have been used to recognize family-group level taxa. Among these characters are the structure of the posterior apertures to the mantle cavity and the presence of lasidium-type larvae. There are six families of freshwater mussels, and these can mostly be distinguished based on just these two traits. For this story, we only need to consider the conditions seen in the Unionidae, Mycetopodidae, and Etheriidae.
The overwhelming majority of freshwater mussels belong to the family Unionidae, and many readers of the MUSSELp Mussel of the Month posts might regard theirs as the “typical” condition. Unionids have parasitic bivalved larvae called glochidia, whereas etheriid and mycetopodid freshwater mussels belong to a group of families with univalved lasidium-type larvae.*
The posterior apertures of unionid mussels are also distinct. There is no mantle fusion between the incurrent and excurrent apertures. The two apertures are only separated by the fusion of the ctenidia to the mantle. There is, however, a short stretch of fusion between the left and right sides of the posterior mantle dorsal to the excurrent aperture. This results in a third opening, the supra-anal aperture. In both mycetopodids and etheriids (including Pseudomulleria according to Woodward, 1898), there is mantle fusion between the incurrent and excurrent apertures and there is no supra-anal aperture. These characters are discussed in detail for all six families in Graf & Cummings (2006): unionids (like Lampsilis) have glochidium larvae and no fusion between the posterior apertures, and etheriids (like Pseudomulleria) and mycetopodids (like Anodontites) have lasidia and apertures separated by mantle fusion. It should be emphasized that the hard and soft anatomy of Pseudomulleria in India and Mulleria (+ Acostaea) in South America are nearly identical (Yonge, 1962, 1978).
Phylogenetic analysis of the available morphological data supports a monophyletic Etheriidae based not only on the presence of asymmetrical oyster-like shells, reduction of the foot, and other characters of the shell and ctenidia (see Graf & Cummings, 2006 for a more detailed discussion). Until 2000, most authorities regarded the family Etheriidae as a valid taxon composed of four monotypic genera (e.g., Mansur & da Silva, 1990), although it had also been argued that etheriids might really just be cementing mycetopodids (e.g., Heard & Vail, 1976).
Molecular Phylogenetic Results
Bogan & Hoeh (2000) published the first and really only molecular phylogenetic dataset useful for testing the monophyly of the Etheriidae. Their dataset consisted solely of COI mitochondrial DNA and their phylogeny was reconstructed using parsimony. Their COI sequences of Mulleria (+ Acostaea) rivoli (AF231739) and Pseudomulleria dalyi (AF231750) remain the only nucleotides reported for those species on Genbank. The topology that they recovered placed Etheria and Mulleria in a paraphyletic grade basal to the Mycetopodidae (not too surprising) but Pseudomulleria landed among the Unionidae (very surprising). From this result Bogan & Hoeh (2000) concluded that the Etheriidae was not monophyletic and that Pseudomulleria should be classified among the Unionidae. That result has been confirmed in subsequent re-analyses: the single COI sequence attributed to Pseudomulleria is consistently recovered with the Unionidae (Hoeh et al., 2009; Graf & Cummings, 2010).
More extensive molecular phylogenetic data sets combining multiple mitochondrial and nuclear nucleotides that could better resolve the phylogenetic position of the Etheriidae are only available for Etheria elliptica, and (as can be seen on the Etherioidea cladomics page on this web site) the multiple analysis that have included Etheria have been inconsistent in their results: Etheria has been recovered as sister to several different clades among the lasidium-bearing mussels. Moreover, with the exceptions of Combosch et al. (2017) and Pfeiffer et al. (2019), etherioideans were included in these various combined datasets of mitochondrial and nuclear DNA as part of the outgroups for analyses of unionoid ingroups.
Conclusions
We currently maintain the Etheriidae as a valid taxon because cladistic analysis of morphology supports monophyly the core genera of Etheria, Mulleria, and Pseudomulleria,** and molecular phylogenetic results to-date are based on incomplete data and have produced inconsistent results.
One question that needs to be answered is, Does Genbank Accession AF231750 actually represent Pseudomulleria dalyi? It has been argued by some that morphological similarity between Pseudomulleria in India and Mulleria in South America is simply an example of convergent evolution of cementing habits. Afterall, asymmetrical shells have evolved multiple times among freshwater bivalves, and modern authors cite Prashad’s (1932) discussion in favor of this argument. But, Prashad (1932) was arguing from a time well before the acceptance of plate tectonics — not to mention the Neo-Darwinian synthesis and the cladistic revolution. A disjunction between Pseudomulleria and Mulleria is difficult to explain without the break-up of Gondwana over the last 200 million years.
A hypothesis of convergence is flimsy on anatomical grounds alone. Pseudomulleria dalyi and Mulleria rivoli do not simply have similar shells. As emphasized by Yonge (1962, 1978), they have identical shells and develop in such similar ways as to convince him of their homology. Moreover, both species share the same soft anatomical characters to support the placement of both among the lasidium-bearing Etherioidea. Modellnaia siamensis from Thailand has an etheriid-type shell, but that is clearly convergent evolution because its soft-anatomy and larvae place it squarely in the Unionidae (Heard & Hanning, 1978). What are the unionid characters exhibited by Pseudomulleria to support its classification in that family?
The only characters that place Pseudomulleria dalyi among the Unionidae are the nucleotides of Genbank Accession AF231750. We know of no voucher specimen reported for this sample, and, in two decades, no one has replicated this result. It would be REALLY EXCITING if Pseudomulleria dalyi is a unionid, but it all hinges on this one data point and provides no explanation for all the contradictory data.
Whether Etheria, Mulleria, and Bartlettia (with or without Pseudomulleria) comprise a clade is a different question that is still waiting on a robust phylogenetic test. So, we are keeping the baby and the bathwater and maintaining the family Etheriidae as the best (although not only) hypothesis of freshwater mussel relationships.
--
* The actual larval type of many of the lasidium-bearing mussels has not been observed for many species, but their larval type is inferred based on phylogenetic topologies (Graf & Cummings, 2006).
** Bartlettia has never been included in a phylogenetic analysis. |
