Comps prep: Unionoida evolutionary relationships


Note: This post is a work in progress that can and will change over the next few days.  I'm making it available for critique, so comment away!  MBK 2012-08-07 1629

With an oral exam coming up in a week, it's nose to the books time!  One of my committee members has posed a certain question relating to soon-to-be-published research on the phylogeny of Order Unionoida (Class Bivalvia1).  This post is an effort to figure out just what has been going on regarding the relative positions of the families in this order.  Were it a simple factual question, I might have more luck, but it happened to be one of those darn "describe the implications of…" inquiries.

According to Bogan and Roe (2008), below is the general arrangement of superfamilies and families within Unionoida.  There are other arrangements that have been made, however, notably placing Family Hyriidae in Superfamily Unionoidea (e.g., Roe and Hoeh, 2003; Walker et al., 2006) according to Simpson's (1900) original classification (Figure 1).

(Figure 1): Historical trees presented by Roe and Hoeh (2003).  Note that the Family Etheriidae is missing from these evolutionary hypotheses: where does it fit in?

  • Order Unionoida
    • Superfamily Unionoidea
      • Family Margaritifera
      • Family Unionidae
    • Superfamily Etherioidea
      • Family Etheriidae
      • Family Iridinidae
      • Family Mycetopodidae
      • Family Hyriidae 

Although the order and the families are relatively solid (according to the work so far), the contents of each superfamily (indeed, the number of superfamilies) is highly debatable.  The families themselves are not all entirely monophyletic, which leads to additional questions: what relationship do the ragged edges (those genera that belong to a different branch of the tree than the rest of the family) actually have to the families they are assigned?  Is it a poor selection of characters, or bad phylogenetic analysis, or wrong identification, or what?  

I'm digressing here, but I think it's important to point out again that the way we group taxa in order to keep things organized may not reflect the actual evolutionary relationships among those taxa.  For example, using the name "Unionidae" for a certain branch of the tree minus genus Psuedomulleria as in Figure 3 of Hoeh et al. (2009) might make sense from a classification standpoint but not from an evolutionary standpoint (Figure 2)2.  Ideally (and I can't think of a counterargument to this), classification systems would follow evolutionary relationships.

(Figure 2): One of three trees from Hoeh et al. (2009).  So what's up with Pseudomulleria?  Turns out Coelatura works better in another position (see Whelan et al., 2011).

Previous Discussion

Returning to the point of this exercise, what are the evolutionary relationships among the six families in Order Unionoida?  (Ordered as above for consistency.)  Interestingly enough as a paleontologist, the characters used for phylogenetic analyses at the family level are all soft-part anatomy and life history (some of which are listed below) rather than hard-part shell morphology.  Although shell morphology in Order Unionida is suspect due to convergence at both the evolutionary and ontogenetic levels, I'm not sure how much thought has been given to the rate at which this morphology evolves in comparison to soft-part anatomy (I spoke too soon).

Superfamily Unionoidea
Lack of "mantle fusion that results in incurrent and excurrent apertures" (Bogan and Roe, 2008).

Family Margaritiferidae
Considered to be monophyletic.  Soft-part morphology 'primitive' or basal to the rest of Order Unionoida; molecular analysis places as sister to Family Unionidae to form Superfamily Unionoidea (Graf and Cummings, 2006).  Skawina and Dzik (2011) support the basal ("least derived") hypothesis based on morphological similarity to Neotrigonia.  
Glochidial larvae.  
Brood in all four demibranchs.
Northern hemisphere (widespread) (Bogan and Roe, 2008).

Family Unionidae
Considered by different studies to be both monophyletic and paraphyletic.  Most recently, Whelan et al., 2011 claim a monophyletic Unionidae very similar to the Linnean classification produced by Bieler et al. (2010) but do not examine relationships with other families.  
Glochidial larvae.  
Brood in all four demibranchs or just outer pair.
Northern hemisphere (widespread) (Bogan and Roe, 2008).

Superfamily Etherioidea
"Some degree of fusion that results in, at least, a completely fused excurrent siphon and, often, a completely fused incurrent siphon" (Bogan and Roe, 2008).

Family Hyriidae
Monophyletic according to all studies so far.  Soft-part morphology places this family as sister to "anything with a lasidium" (Superfamily Etherioidea as listed at the top of this post), but mtDNA analysis breaks the Hyriidae out entirely and places it as sister of and basal to all other families in Order Unionoida (Graf and Cummings, 2006; Bogan and Roe, 2008).  This disagreement comes down to methods, which means it matters with whom you're speaking at the moment.  
Glochidial larvae.  
Brood in inner demibranchs.
Southern hemisphere (South America and Australasia) (Bogan and Roe, 2008).

Family Mycetopodidae
Lasidial larvae.  
Brood in inner demibranchs.
Southern hemisphere (South America) (Bogan and Roe, 2008).

Family Iridinidae
Lasidial larvae.  
Brood in inner demibranchs.
Southern hemisphere (Africa) (Bogan and Roe, 2008).

Family Etheriidae
Lasidial larvae.  
Brood in inner demibranchs.
Southern hemisphere (Africa) (Bogan and Roe, 2008).

Walker et al., 2006 summarizes the possible familial relationships as follows (although there may be others):

  • basal Margaritiferidae, paraphyletic Unionidae, derived Hyriidae+Mycetopodidae+Iridinidae.  "Analyses of morphological characters presented by Graf (2000) and Hoeh et al. (2001) both return the Margaritiferidae as the basal unionoid lineage, a paraphyletic Unionidae, and the Hyriidae + Mycetopodidae + Iridinidae as a derived lineage." 
  • basal Hyriidae, paraphyletic Unionidae. "In contrast, both the molecular (i.e., cox1 DNA sequences) and the combined analysis of morphological and molecular data presented by Hoeh et al. (1998a, 2001) (Fig. 1C) place the Hyriidae as the basal unionoid lineage, with the Unionidae returned as paraphyletic." 
  • basal Margaritiferidae, monophyletic Unionidae, sister Hyriidae+Iridinidae+Mycetopodidae. "Yet another combined analysis of morphological and molecular (i.e., cox1 DNA sequences) characters (Roe & Hoeh, 2003; Fig. 1D), this time using binary coding of the morphological data and a posteriori character weighting, returns the Margaritiferidae as basal and a monophyletic Unionidae as sister to a Hyriidae + Iridinidae + Mycetopodidae clade."  It is pointed out by Graf and Cummings, 2006 that the binary coding of characters resulted in the unforseen weighting of certain character pairs, i.e., coding "1" for one character necessitated coding "0" for another.

If we consider the methods used in each case to be equally plausible, each of these three scenarios supports the close relationship between Iridinidae and Mycetopodidae (Etheriidae, being represented by a single(?) genus, seems to be left out, but I believe is generally considered to be closely related to these two families as well).  This leaves Unionidae (paraphyletic or not), Hyriidae, and Margaritiferidae (the latter two of which are considered by different character sets as both basal and derived).  While you think on this, check out Figure 3.

(Figure 3):  Graf's interpretation of some relationships within Order Unionoida.

As a final thought on past work, click here to see what Graf and others have come up with for solutions, and a summary figure for the last several papers on this subject.  Confusing, isn't it?


New Discussion

I think I will wait a month or so before I finish this section up.  Some new information should be coming down the pipeline.

Works Cited

Many of the publications cited have been linked to in the text, often to a page on an academic social network.  By following these links, you are letting the authors know you are interested in their work.  Publications that have no official online presence are listed here.

Roe, K. J. and Hoeh, W. R. (2003). Systematics of freshwater mussels (bivalvia: Unionoida). In Lydeard, C. and Lindberg, D. R., editors, Molecular systematics and phylogeography of mollusks, chapter 4, pages p. 91–122. Smithsonian Books.






1 Why Bivalvia and not Pelecypoda?  Bivalvia has nomenclatural priority, unfortunately, so as cool as studying "hatchet-footed" animals is, we don't get to call them that in polite company.
2 This has supposedly been resolved by Whelan et al. (2011) by improving the position of Coelatura.