Basal tetrapod cladogram: Marjanovic and Laurin 2016, PeerJ

Recently I added
several basal tetrapod taxa to the large reptile tree (LRT, now 950 taxa) in order to better understand the origin of the clade Reptilia (= Amnlota). Along the way, the software recovered some contra-traditional nestings which revived typically cordial correspondences with Drs. David Marjanovic and Jason Pardo, both of whom have studied basal tetrapods extensively. I don’t have all of the latest literature and I appreciate that these researchers open doors I may not have seen.

Less recently
Marjanovic and Laurin (2016) reexamined a earlier report on lissamphibian origins by Ruta and Coates (2007). Marjanovic and Laurin (ML) report “thousands of suboptimal scores due to typographic and similar errors and to questionable coding decisions: logically linked (redundant) characters, others with only one described state, even characters for which most taxa were scored after presumed relatives. Even continuous characters were unordered, the effects of ontogeny were not sufficiently taken into account, and data published after 2001 were mostly excluded.”

Figure 1. Click to enlarge. Wait 10 seconds for animation to begin. Basal tetrapod tree form Marjanovic and Laurin 2016.

Figure 1. Basal tetrapod tree form Marjanovic and Laurin 2016. After 10 seconds those moving lines that appear on the right will make sense when you CLICK TO ENLARGE and see how they connect taxa on competing trees.

ML document and justify all changes
to the earlier matrix, then add 48 taxa to the original 102. They report,  “From the late19th century to now, the modern amphibians have been considered temnospondyls by some (refs omitted), lepospondyls by others and polyphyletic yet others, with Salientia being nested among the temnospondyls, Gymnophionomorpha among the lepospondyls, and Caudata either in the lepospondyls (all early works) or in the temnospondyls (works published in the 21st century).”

“The present work cannot pretend to solve the question of lissamphibian origins or any other of the controversies in the phylogeny of early limbed vertebrates (of which there are many, as we will discuss). It merely tries to test, and explain within the limitations of the dataset, to what degree the trees found by RC07 still follow from their matrix – the largest published one that has been applied to those questions – after a thorough effort to improve the accuracy of the scoring and reduce character redundancy has been carried out to the best of our current knowledge. However, we think this effort forms a necessary step towards solving any of those problems. Further progress may come from larger matrices…”

“Our matrix has only 276 characters, a strong decrease from the 339 of RC07. For the most part, this is due to our mergers of redundant characters and does not entail a loss of information.”

After all that work and all those changes and additions,
ML report their repaired tree “topology is identical to Ruta and Coates 2007.”

Unfortunately that tree is vastly different
from the one recovered in the LRT, which has far fewer taxa, but an equal or greater gamut. Let’s figure out why the topologies differ and are similar. I’ll start slow with the similarities and the metaphorical ‘low-hanging fruit.’ The difficult topics we will handle later. I took the last few weeks (far too little time) to better understand basal tetrapods having zero knowledge of most taxa before starting. I have not been able to cover all the taxa employed by the ML tree.


  1. Both trees include fish and fish-like tetrapods at the base
  2. Both trees include microsaurs, reptiles and extant amphibians as derived taxa
  3. Both trees agree on the inclusion set for microsaurs and holospondyls
Figure 6. Subset of the large reptile tree focusing on basal tetrapods, updated with Gerrothorax.Figure 6. Subset of the large reptile tree focusing on basal tetrapods, updated with Gerrothorax.

Figure 2. Subset of the large reptile tree focusing on basal tetrapods, updated with Gerrothorax.


Due to taxon exclusion,
the ML tree nests several reptiles as non-reptiles. These include:

  1. Sivanerpeton – basal reptile/amniote
  2. Gephyrostegus (bohemicus) –  basal reptile/amniote
  3. Bruktererpeton – Lepidosauromorpha
  4. Solenodonsaurus (+ Chroniosaurus, Chroniosuchus) – Archosauromorpha
  5. Tseajaia – Lepidosauromorpha
  6. Limnoscelis – Lepidosauromorpha
  7. Orobates – Lepidosauromorpha
  8. Diadectes – Lepidosauromorpha
  9. Westlothiana – Archosauromorpha

Where each taxon nests in the LRT follows each dash.

Due to taxon exclusion,
the ML tree nests several taxa as ‘Sauropsida’ a clade that no longer has utility based on the new basal reptile dichotomy Archosauromorpha and Lepidosauromorpha. These include:

  1. Captorhinus – Lepidosauromorpha
  2. Paleothyris/Protorothyris – two distinct Archosauromorpha
  3. Petrolacosaurus – Archosauromorpha

ML report, “Chroniosaurus has a fully resolved position one node more crown ward than Gephyrostegidae, Bruktererpeton or Temnospondyli and one node more rootward than Solenodonsaurus.”  This is a similar nesting to the LRT except that all listed taxa other than Temnospondyli nest within the Reptilia. ML are missing several taxa that would have changed their tree topology (see the LRT for that list).


I caught a little heat for not using the latest drawings of Microbrachis earlier. The new tracings (Fig. 3) come from Vallin and Laurin 2004. Note the tracings of the in situ specimen (color) do precisely match the freehand reconstruction they offered. All scoring changes further cemented prior LRT relationships.

Figure 3. Microbrachis images from Vallin and Laurin 2008. Color added here.

Figure 3. Microbrachis images from Vallin and Laurin 2004. Color added here.

More later.

Marjanovic D and Laurin M 2016. Reevaluation of the largest published morphological data matrix for phylogenetic analysis of Paleozoic limbed vertebrates. PeerJ. Not peer-reviewed. 356 pp.
Ruta M and Coates MI 2007
. Dates, nodes and character conflict: addressing the lissamphibian origin problem. Journal of Systematic Palaeontology 5-69-122.
Vallin G and Laurin M 2004. Cranial morphology and affinities of Microbrachis, and a reappraisal of the phylogeny and lifestyle of the first amphibians. Journal of Vertebrate Paleontology: Vol. 24 (1): 56-72 online pdf



6 thoughts on “Basal tetrapod cladogram: Marjanovic and Laurin 2016, PeerJ

  1. Figure 1. Basal tetrapod tree form Marjanovic and Laurin 2016

    For fuck’s sake, man. That’s not our tree. It’s the tree of Ruta & Coates (2007)! We’re presenting their tree for comparison to ours, which are presented in later figures.

    You didn’t read any farther than the first figure, did you.

    After all that work and all those changes and additions,
    ML report their repaired tree “topology is identical to Ruta and Coates 2007.”

    That’s an alternative fact.

    The topology we find is identical to the ones they found if and only if we analyze their matrix without any changes! That’s our Analysis O1.

    Check out the analyses with an R instead of an O. Those result from our modified matrix. You’ll find the topologies are quite different.

    I refuse to even read your post any further. Read our preprint first. There’s a reason it has 350 pages and 37 figures, not just the 5 pages and 1 figure you’ve looked at.

    Honest question: do you have a presidential attention span?

  2. I notice that you repeatedly argued that taxon exclusion is either the largest, or culprit responsible for the discordant results seen between the LRT and all other trees. With that in mind, have you tried excluding those specific taxa from the LRT to see if it then produces the same topology seen in these studies (e.g., the Marjanović and Laurin topology you compare above)? It seems like the easiest way to test the robustness and validity of the LRT when compared to other studies. If the LRT does show the same results as prior studies, after taxon exclusion, then you would have a strong argument for greater taxa inclusion in future studies. If it doesn’t, then it might be time to re-evaluate why your trees are consistently coming out differently.

    • the Marjanović and Laurin topology you compare above

      Again, it’s the Ruta and Coates topology he’s comparing to his above.

      We repeated their analysis to check if they had done it right. That may sound strange, but 1) it has happened several times before that people didn’t let their analyses run for long enough to find all, or sometimes even any, of the shortest trees for their matrix; 2) Ruta & Coates used a trick to cut down on calculation time, and we wanted to see if the trick actually worked. (It did. Our trick-free reanalysis O1 found the same number of trees of the same length and topology.)

      Concerning 1), in the same preprint we also reanalyzed the matrix of Maddin, Jenkins & Anderson, 2012. Lo & behold: the shortest trees are one step shorter than what they had reported, and the number of trees with the length they reported doesn’t have two digits, but six. The topology of the shortest trees is very similar to what they reported, though, so it happens not to be a big deal in this case.

    • The taxon sample of the LRT isn’t simply larger. It’s wider, but it’s not denser. Several taxa in our sample aren’t in the LRT, and some (like Pholidogaster) were only added to the LRT very recently.

  3. BTW, you miscited the preprint. It has not been published in PeerJ (where it’s currently undergoing the second round of peer review). It’s been published in PeerJ PrePrints, which isn’t the same thing. Specifically, it’s in volume 4, and its article number is e1596v2 (v2 because it’s the second version, the first having been published in volume 3 in 2015).

    You could have simply copied & pasted the “Cite this as” box. Why didn’t you?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.