Saturday, August 21, 2010

Clusters

See link and download supporting Word document.
http://onlinelibrary.wiley.com/doi/10.1111/j.1420-9101.2010.02039.x/suppinfo



Notice the "birdlike cluster" (dark blue, light blue and violet) and the separate "tyrannosaur cluster" (red, orange and
pink).
Note: The "birdlike cluster also contains taxa 14, 17 and 30 (see below).


-->
Table S10 Three-dimensional coordinates of the CMDS visualization of the 2009 matrix. Number and color on left refer to the label and color of each taxon’s point in the scatter plot (Figure 2). Note that closely related ingroups are given the same color, and non-coelurosaurian ougroups are black.
1 black                   Dilophosaurus      
2  black                  Allosaurus               
3 black                   Sinraptor                
4 pink                     Guanlong               
5 pink                     Gorgosaurus           
6 pink                     Tyrannosaurus        
7 red                       Sinosauropteryx     
8 red                       Huaxiagnathus       
9 red                       Compsognathus      
10 orange               Ornitholestes          
11 gold                   Gallimimus             
12 gold                   Ornithomimus         
13 gold                   Struthiomimus        
14 yellow Falcarius               
15 yellow Therizinosaurus
+ Segnosaurus
+ Erlikosaurus
+ Erliansaurus
+ Neimongosaurus                 
16 olive                   Shuvuuia                                
17 light green          Protarchaeopteryx


-->
+ Incisivosaurus  
18 light green          Caudipteryx           
19 dark green         Citipati                  
20 dark green         IGM 100/42           
21 dark green         Ingenia                   
22 dark green         Khaan                     
23 light blue            Anchiornis            
24 light blue            Mei                         
25 dark blue           Bambiraptor         
26 dark blue           Sinornithosaurus   
27 dark blue            Microraptor           
28 dark blue           Velociraptor          
29 dark blue           Deinonychus         
30 magenta             Epidendrosaurus
+ Epidexipteryx   
31 violet                  Archaeopteryx       
32 violet                  Sapeornis               
33 violet                  Confuciusornis      


http://onlinelibrary.wiley.com/enhanced/doi/10.1111/j.1420-9101.2010.02039.x/

The 2009 matrix (Fig. 2) reflects the new discoveries of several taxa that bridge gaps between formerly separated clusters. With the addition of the basal tyrannosauroid Guanlong and the basal compsognathids Huaxiagnathus, a coherent cluster (hereafter called the Tyrannosaur Cluster) is formed by Tyrannosauroidea, Compsognathidae and Ornitholestes. A second cluster, hereafter called the Birdlike Cluster, includes basal birds and the birdlike taxa Dromaeosauridae, Troodontidae, Epidendrosaurus + Epidexipteryx, [scansoriopterids] Protarchaeopteryx [oviraptor]Incisivosaurus [oviraptorand Falcarius [Therizinosaur]

Between the least-separated members of the two clusters (Guanlong in the Tyrannosaur Cluster and Falcarius in the Birdlike Cluster) is a gap of only 0.135, which is smaller than the span of either cluster (0.143 for the Tyrannosaur Cluster, and 0.263 for the Birdlike Cluster). Distances of ≥ 0.200, indicating gaps, continue to isolate Oviraptoridae, CaudipteryxShuvuuia, Ornithomimidae and Therizinosauridae from all other taxa.


Let's begin with a precise listing of which taxa are in the "Birdlike cluster" according to Senter:
Dromaeosauridae (dark blue): 25, 26, 27, 28, 29
Troodontidae (light blue): 23, 24
Aves (violet): 31, 32, 33
Protarchaeopteryx + Incisivosaurus (light green): 17
Epidendrosaurus + Epidexipteryx (magenta): 30 
And Senter says these are a morphologically continuous group. 

Turning to the "Tyrannosaur cluster", Senter includes:
Tyrannosauroidea (pink): 4, 5, 6
Compsognathidae (red): 7, 8, 9
Ornitholestes (orange): 10 
And Senter says these are a morphologically continuous group. 

6 comments:

  1. Just to clarify something, actually the "birdlike cluster" is not just "dark blue and violet" but also green (light and dark green for oviraptors and alvarezsaurs respectively), yellow for therizonosaurs and magenta for scansoriopterids. The "tyrannosaur cluster" likewise is not just "red, orange and pink" but also includes gold for ornithomosaurs. When you take these groups into account you can see that overall the clusters come very close in places (in fact closer than between some of the groups within the clusters and much closer than the overall span of each cluster). This is consistent with the two groups being related which is what Senter actually says is the case - he does not argue that they are unrelated, that I'm afraid is a misunderstanding on your part. I can't believe that anyone who has carefully read the paper would conclude that this is what he is saying. His conclusion is that there is "morphological continuity between nonavian coelurosaurs and basal birds"

    ReplyDelete
  2. A few things. I will deal with them one at a time.
    You have said:
    His conclusion is that there is "morphological continuity between nonavian coelurosaurs and basal birds".

    I have pointed out that the words "nonavian coelurosaurs" is misleading since it includes both "Birdlike cluster" creatures and "Tyrannosaur cluster" creatures. May I suggest you read the latest posts that deal with that.

    ReplyDelete
  3. You have said that: "the "birdlike cluster" is not just "dark blue and violet" but also green (light and dark green for oviraptors and alvarezsaurs respectively), yellow for therizonosaurs and magenta for scansoriopterids". The "tyrannosaur cluster" likewise is not just "red, orange and pink" but also includes gold for ornithomosaurs."
    Can you give a link for that and copy and paste the material that supports what you are saying please?
    Then we can analyze this together.

    ReplyDelete
  4. Actually I wrote in haste last night (it was late) - it's a bit more complicated than that and I should correct my error here. I forgot that in his paper Senter argued that currently the fossil record is still incomplete for the later oviraptors (although not the basal oviraptors Protarchaeopteryx + Incisivosaurus), alvarezsaurs respectively and therizonosaurs, so he didn't include them in the birdlike cluster for the sake of this analysis - although they should be in it and in his conclusions he points to a recent alvarezsaur find that would probably close several gaps. However scansoriopterids and some oviraptors are certainly in it - can't you tell from your scatter plot? Also the fossil record for ornithomosaurs is less complete so he excluded them too even though anatomically they share much in common with members of the tyrannosaur cluster and are obviously related (for now - remember this is an evolving data set as new fossils are added - as the earlier scatterplots show gaps can be very misleading so don't read too much into them). However the main point remains that he shows, you can see this clearly in the scatterplots, that the distance between the 2 clusters is less than the span of either of them - which falsifies the hypothesis that it shsould be greater if the two groups really were seperate creations or seperately evolved.

    Remember he does also include a fuller cladogram showing how the groups are linked - this is based on more data than the scatterplots too (as he says in the paper - do you not have access to it?)

    re your point about changing the "nonavian coelurosaurs" - sorry but that's just silly. Whether it is "misleading" to you or not (actually it's perfectly clear) it's certainly clear about what Dr Senter is arguing so you can't say "he says" the 2 clusters are unrelated when what he says is clearly that they are. You are entitled to your own spin on it if you choose but not to put words in the authors mouth.

    ReplyDelete
  5. Let's begin with a precise listing of which taxa are in the "Birdlike cluster" according to Senter:
    Dromaeosauridae (dark blue): 25, 26, 27, 28, 29
    Troodontidae (light blue): 23, 24
    Aves (violet): 31, 32, 33
    Protarchaeopteryx + Incisivosaurus (light green): 17
    Epidendrosaurus + Epidexipteryx (magenta): 30
    And Senter says these are a morphologically continuous group.

    Turning to the "Tyrannosaur cluster", Senter includes:
    Tyrannosauroidea (pink): 4, 5, 6
    Compsognathidae (red): 7, 8, 9
    Ornitholestes (orange): 10
    And Senter says these are a morphologically continuous group.
    Senter does not say that these two groups are related. If you have information otherwise please give the link and copy and paste the relevant material.
    Also concerning your point that
    "the distance between the 2 clusters is less than the span of either of them - which falsifies the hypothesis that it shsould be greater if the two groups really were seperate creations or seperately evolved". If Senter is saying this, please give the link and copy and paste the relevant material.
    Concerning the words "nonavian coelurosaurs": if you are still not understanding the issue, keep thinking about it. I am not going to use up time arguing this obvious point with you.
    I will say as a general comment what I said earlier to a commenter:
    I know full well from lots of experience that people will hold onto any tiny, tiny point and argue it till the cows come home. I will not be falling for that here.

    ReplyDelete
  6. For those not acquainted with the phrase "till the cows come home", please see:
    http://www.phrases.org.uk/meanings/382900.html

    ReplyDelete