[Taxacom] Molecular symplesiomorphies (frogs, invertebrates, etc.)

Tue Apr 1 21:41:12 CDT 2014

Hi John,
      Sorry I couldn't get back to this earlier.  Reading your post on frog phylogeny absolutely baffles (even shocks) me, coming from someone who lectures primatologists for putting too much emphasis on molecular (vs. morphological) data regarding humans and great apes (chimpanzees in particular).                     
       Now we have a case where two frog families apparently diverged a hell of a lot longer ago, and you pick this as a case where molecular data convinces you, even though there are NO morphological synapomorphies AT ALL shared by these frogs.  The consensus actually seems to be that their shared morphological characters are primitive (symplesiomorphies), so these "DEEP" shared molecular characters are very likely to be symplesiomorphies as well.   If there is a case where the molecular data should be questioned, it is these frog families (not chimps and humans).  
        Therefore, I would side with Green and Cannatella (1993) and Cannatella and Ford (1993) who actually found several morphological synapomorphies uniting Leiopelmatidae with more recent frogs (to the EXCLUSION of Ascaphidae, which lacks those morphological characters).  Thus they named a new clade Leiopelmatanura as sister group to the most primitive living family of frogs (Ascaphidae).    
       Not surprisingly, it is noted that one of the co-authors in Ford et al., 2006 (namely Green), disagreed with their decision to lump Ascaphidae and Leiopelmatidae together.  Keeping them as separate families would have made perfect sense, given the conflict between morphological and molecular data.  However, Green was overruled by the consensus among the authors of that particular paper.    
       So John, I would suggest that if you are so concerned about molecular data being questionable, this is clearly a case where you should be focusing your attention.  If Ascaphidae split off first, and then Leiopelmatidae split off next (before the rest of the extant frogs families), one would expect those two primitive families to not only share morphological  (primitive) symplesiomorphies, but that their shared molecular sequences would be symplesiomorphies as well.  Especially as these cladistic splits happened so long ago. 
       Frankly, it is this type of ancient divergence which has made it so difficult to use molecular data to establish the relationships of invertebrate phyla, and the relationships within some those phyla (such as Arthropoda).  If you don't pick just the right molecules to sequence, you are likely to end up confusing molecular symplesiomorphies as molecular synapomorphies (and this has no doubt happened often). 
        Looks to me like the cladistic relationships between the earliest frogs is just another case where we haven't found the right molecular sequences to discriminate between symplesiomorphy and synapomorphy to explain the similarities.  CONCLUSION: Although morphology can obviously be confusing when it comes to convergence and parallelism,  the further back in time one goes, we are just as likely (sometimes even more likely) to be confused by convergence and parallelism in molecular sequences if we don't pick the right molecules.                                                                    ---------------------------------Ken Kinman
--------------------------------------------------------------------------------------------------
> Date: Wed, 2 Apr 2014 06:43:13 +1300
> From: calabar.john at gmail.com
> To: bckcdb at istar.ca
> CC: taxacom at mailman.nhm.ku.edu
> Subject: Re: [Taxacom] Frog biogeography
> 
> Fred,
> 
> 'consensus' means everyone, but not everyone takes the view expressed in
> the Wikapedia page. Interestingly, the three references given show (4)   Frost,
> Grant, Faivovich, Bain, Haas, Haddad, de S?, Channing, Wilkinson,
> Donnellan, Raxworthy, Campbell, Blotto, Moler, Drewes, Nussbaum, Lynch, and
> Green, 2006, Bull. Am. Mus. Nat. Hist.,
> 297<http://research.amnh.org/vz/herpetology/amphibia/index.php//Bibliography/F/Frost-Grant-Faivovich-Bain-Haas-Haddad-de-Sa-Channing-Wilkinson-Donnellan-Raxworthy-Campbell-Blotto-Moler-Drewes-Nussbaum-Lynch-and-Green-2006-Bull.-Am.-Mus.-Nat.-Hist.-297>,
> discussed why *Leiopelma* and *Ascaphus* should be associated and found
> this inclusive taxon to sit phylogenetically as the sister taxon of all
> other frogs, as did Roelants, Gower, Wilkinson, Loader, Biju, Guillaume,
> Moriau, and Bossuyt, 2007, Proc. Natl. Acad. Sci. USA,
> 104<http://research.amnh.org/vz/herpetology/amphibia/index.php//Bibliography/R/Roelants-Gower-Wilkinson-Loader-Biju-Guillaume-Moriau-and-Bossuyt-2007-Proc.-Natl.-Acad.-Sci.-USA-104>:
> 887-892states:
> 
>  ref 5 is a 2009 paper that does not present a phylogenetic analysis. Ref 6
> links to TOL that states:
> 
> *Ascaphus* and *Leiopelma* have usually been placed in the same family,
> sometimes called Ascaphidae or Leiopelmatidae (the latter name is the
> nomenclaturally correct one). This arrangement has been recognized as based
> on shared primitive characters (Green and Cannatella, 1993), because no
> morphological synapomorphies have been proposed to unite the two taxa.
> 
> However, analyses of molecular data have indicated that *Ascaphus* and
> *Leiopelma* are sister-groups, but the divergence between them is very
> deep. Hay et al. (1995) analyzed relationships among families of frogs
> using 12S and 16S rRNA and found *Ascaphus* and *Leiopelma* to be
> sister-groups.  The same arrangement was found by Roelants et al. (2005)
> and San Mauro et al. (2005).
> 
> So there would seem to be no 'consensus' supporting dissociation and Ken's
> assertion that the leiopelmatid clade is outdated does not appear to be
> substantiated. In place of monopphyly some studies I saw years a go
> appeared to provide for one of the two genera being basal to all other
> frogs, and then the other genus was basal in all other frogs and this would
> still place the two basal-most groups straddling the Pacific.
> 
> John Grehan
> 
> 
> 
> 
> On Wed, Apr 2, 2014 at 5:29 AM, Fred Schueler <bckcdb at istar.ca> wrote:
> 
> > On 4/1/2014 11:25 AM, John Grehan wrote:
> > > I did not see any evidence from Ken refuting a leiopelmatid clade
> > > encompassing NZ, W Nam and Argentina (fossil). Some on this list may find
> > > the biogeography of this group in relation to all other frogs of some
> > > interest.
> >
> > * I haven't chased this further, but... "While some taxonomists have
> > suggested combining the North American frogs of the genus Ascaphus in
> > the Ascaphidae family with the New Zealand frogs of the genus Leiopelma
> > in the Leiopelmatidae family, the current consensus is that these two
> > groups constitute two separate families.[4][5][6] ...The family
> > Ascaphidae, of the same suborder, shares these primitive
> > characteristics, and hence the two have often been described as related,
> > or even part of the same family."
> > http://en.wikipedia.org/wiki/Leiopelmatidae
> >
> > fred.
> > ==========================
> >
> > >
> > > Heads notes that "The family is sister to all other modern frogs...With
> > its
> > > western North American records, Leiopelmatidae cannot be Gondwanan. San
> > > Mauro et al (2005) described the leiopelmatid genera as 'merely relicts',
> > > but this does not explain the overall allopatry between the family,
> > > straddling the central Pacific basin, and all other frogs, which are
> > absent
> > > from the Pacific east of Fiji...The split between the Pacific
> > > Leiopelmatidae and all other frogs dates back to the Paleozoic or early
> > > Mesozoic (San Mauro et al., 2005), before the Jurassic formation of the
> > > Pacific plate and the Pacific basin."
> > >
> > >
> > > John Grehan
> > > _______________________________________________