Another Lacerta ID

[Mario Schweiger]

Now they are online!
PDF-5692 is only the paper, PDF-5692_full includes the appendices.

Yes, no bili/viri stuff in the paper.
But Werner wanted to do nearly the same ("trili"). The bili/viri are a "side-effect" only. Much to less data.
But I dont realy see a problem with the 2 Austrian samples.
The tree shows two different recolonication ways. One along the Danube (Hundsheim is situaded east of Vienna) and one along the Dinares. Although one might expect, there should be bilis or Adria clade in Arnoldstein (at the Italian/Austrian border from Udine to Villach).

Mario

[Jeroen Speybroeck]

The bili/viri are a "side-effect" only. Much to less data.

Would be a pity if this never becomes further substantiated.

But I dont realy see a problem with the 2 Austrian samples.

I'm a bit surprised by this, as I'm sure you already know the stuff I am about to write. The subspecies/clades are not monophyletic - any branch has to retain members of the same clade/taxon, only of that clade/taxon and nothing but members of that clade/taxon. This is not the case here.
1 - the branch with viridis also contains meridionalis (= you cannot cut out a subtree for viridis with a single cut)
2 - not all meridionalis belong to the same branch
To retain monophyletic clades/taxa, you could attribute Hundsheim, Arnoldstein and Doirani to meridionalis OR cancel meridionalis alltogether, and attribute Bolu to bilineata(!) (or something new?).

What data was used for these trees?
What are the numbers with the branching in the second tree?

[Peter Oefinger]

As far as I remember Werners presentation in Gersfeld 2011 he says that meridionalis (in Europe) is no valid subspecies but is viridis viridis. If so, there should be no problem with a paraphyletic subspecies...

[Jeroen Speybroeck]

As far as I remember Werners presentation in Gersfeld 2011 he says that meridionalis (in Europe) is no valid subspecies but is viridis viridis. If so, there should be no problem with a paraphyletic subspecies...

Aha... Then only the Bolu sample would remain problematic.

[Mario Schweiger]

Sorry for beeing late, but I was not at home.
Yes, the (sub)species codes are only given by distribution - or - what (sub)species is given for those localities in literature.
Therefore, only the Bolu specimen is problematic.
Werner used 2 more samples (after he sent me these data), he had from me from NW Anatolia (Abant and Sünnet lake), and both are very narrow related to the Bolu specimen.

I only have the dataset (ClustalX) from 2008 with the specimens shown in the tree.
The online-docu for Treeview says:

Code: Select all

Why won't TreeView display bootstrap values for ClustalX trees?

The online documentation for ClustalX states:
     "By default, the bootstrap values are correctly placed on the tree branches of the phylip format output tree. The toggle allows them to be placed on the nodes, which is incorrect, but some display packages (e.g. TreeTool, TreeView and Phylowin) only support node labelling but not branch labelling. Care should be taken to note which branches and labels go together." [my emphasis]    

In fact there is little consensus about where in a tree description one should place bootstrap values, and hence different programs support different solutions, such as replacing branch lengths with bootstrap values (PHYLIP, optional in PAUP*), using node labels (TreeView, ClustalW, optional in PAUP* and ClustalX), placing them inside brackets "[ ]" after the branch lengths (ClustalX default), and so on. In the absence of a standard, TreeView uses node labels. It will ignore labels inside "[ ]" as any such text is treated as a comment.

Mario

[Jeroen Speybroeck]

Thanks, Mario. Interesting.
However
- the inserted explanation confuses me even more...
- you still haven't told us the nature of the data upon which the trees were built.

[Mario Schweiger]

Sorry, I wanted to answer just during the server clinch

I have this CLUSTAL file since autumn 2008 and, as far as I remember, these are mtDNA data only.

Mario