The origin(s) of extant amphibians (Marjanovic and Laurin 2013)

This post was updated February 8, 2017 with new scores for several basal tetrapod taxa and new taxa. 

A new paper by Marjanovic and Laurin 2013 reviews the various hypotheses regarding the enigmatic origin of modern amphibians: frogs, salamanders and caecilians, together considered (by some, but not all paleontologists) members of a monophyletic clade, the Lissamphibia.

From their abstract (condensed)
“The origins of the extant amphibians (frogs, salamanders, caecilians) remain controversial after over a century of debate. Three groups of hypotheses persist in the current literature: the “temnospondyl hypothesis”… the “lepospondyl hypothesis”, and the “polyphyly hypothesis” according to which the frogs and the salamanders are temnospondyls while the caecilians are lepospondyls. We present a review of recent publications and theses in this field, several of which show more support for the Lepospondyl hypothesis.”

They conclude, “Still, a consensus will not be reached soon, despite the increasing range of data and types of analysis that are used.”

Amphibians in the large reptile tree. Green taxa are most closely related to living amphibians. Pink taxa are diadectomorphs deep within the Reptilia

Figure 1. Amphibians in the large reptile tree. Green taxa are most closely related to living amphibians. Pink taxa are diadectomorphs deep within the Reptilia

Results of the Large Reptile Tree
The large reptile tree includes 32 non-reptile outgroup taxa (Fig. 1), from the Devonian early tetrapod, Ichthyostega, to the derived microsaur, Rhynchonkos.

Also included at ReptileEvolution.com are Rana (the extant bull frog) and its ancestors, Triadobatrachus (a Triassic pre-frog) and Gerobatrachus (an early Permian pre-pre-frog and pre-salamander.)  A sister to Gerobatrachus is Doleserpeton, also from the Early Permian.

An ancestral caeceilian, Eocaecelia, is also featured at ReptileEvolution.com.

All three find an ancestor in sisters to Utegenia (Fig. 2) then Seymouria and Eldeceeon. 32 is not a complete listing, nor a full gamut listing, by any means. Still, these nestings all appear to be good sister taxa without any “strange bedfellows” that just don’t seem to fit. In other words, the large reptile tree found a plausible solution with Utegenia near to the origin of modern amphibians plus two clades of microsaurs, each with 6 or 7 taxa.

Utegenia nests as the common ancestor of frogs, salamanders, caecelians and microsaurs.

Figure 2. Utegenia nests as the common ancestor of frogs, salamanders, caecelians and microsaurs.

Utegenia
Utegenia (Fig. 2) does not play a big part in the Marjanovic and Laurin (2013) study. Neither does Gephyrostegus. Marjanovic and Laurin (2013) separate ‘Amniota’ from Diadectomorpha, Solenodonsaurus and Westlothiana, but these are all amniotes in the large reptile tree.

Microsaurs
Microsaurs, like Utaherpeton, nest as the sister taxon to modern amphibians. Microsaurs took on a very reptilian appearance by convergence.

Figure x. Celtedens, an albanerpetontid close to salamanders.

Figure 3. Celtedens, an albanerpetontid close to salamanders.

Albanerpetontids
Celtedens (Fig. 3) is one of the few complete skeletons attributed to a clade of near salamanders known as the Albanerpetontidae. First time I’ve heard of them was in the Marjanovic and Laurin (2013) paper.

No jugal?
Frogs and salamanders lack a jugal. The maxilla connects directly to the quadratojugal, if it connects to anything at all.

The lysorophian lissamphibian, Brachydectes, likewise has no jugal, but this snake-like amphibian nested closer to caecelians (in the large reptile tree) and apparently has lost its jugal by convergence.

Utegenia retains a jugal that separated the maxilla and quadratojugal, but it is also ancestral to microsaurs, which have a similar cheek pattern. Amphibamus retains a similar jugal. Cacops, however, reduces the jugal and strongly connects the maxilla and quadratojugal. Doleserpeton is more similar to modern amphibians in having a much smaller jugal. Gerobatrachus likewise has a gracile jugal and a lateral temporal fenestra. Celtedens (Fig. 3), an albanerpetontid close to salamanders, also has a jugal.

Palatal tusks
Marjanovic and Laurin (2013) report, “Be that as it may, palatal tusks are absent throughout the amniote-diadectomorph-lepospondyl clade (the tetrapod crown‑group according to the LH), with only three reversals (all of them among “microsaurs”)”. The large reptile tree found tusks absent in reptiles and lissamphibia + microsauria, but present in other tested non-amniotes, including Cacops.

Lepospondyli
According to Wikpedia , “Lepospondyli are a group of small but diverse Carboniferous to early Permian tetrapods. Six different groups are known, the AcherontiscidaeAdelospondyliAïstopodaLysorophiaMicrosauria and Nectridea, and between them they include newt-like, eel- or snake-like, and lizard-like forms, along with species that do not fit any current category.” Here (fig.1) Adelospondyli and Lysorophia are related to Caecelians. Nectridea nest within the Microsauria and these are sisters to the ancestors of frogs and salamanders, Gerobatrachus and Doleserpeton. The other clades have not been tested.

As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.

Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.

References
Marjanovic D and Laurin M 2013. The origin(s) of extant amphibians: a review with emphasis on the “lepospondyl hypothesis.” GEODIVERSITAS • 2013 • 35 (1) online here

Caecelians by Darren Naish

5 thoughts on “The origin(s) of extant amphibians (Marjanovic and Laurin 2013)

  1. Whoa! Thanks for the publicity! :-)

    frogs, salamanders and caecilians, together considered (by some, but not all paleontologists) members of a monophyletic clade, the Lissamphibia.

    AFAIK, the last paleontologists who sticks to the polyphyly hypothesis is Carroll. Anderson, as we mention in the paper, gave up in 2012, when two analyses coauthored by him found lissamphibian monophyly.

    Rana (the extant bull frog) and its ancestors, Triadobatrachus (a Triassic pre-frog) and Gerobatrachus (an early Permian pre-pre-frog and pre-salamander.)

    1) Rana is one of those ever-shrinking genera. Linnaeus lumped all anurans into Rana (just like he lumped all urodeles into Lacerta salamandra and all crocodylians into Lacerta crocodilus). The type species of R. is R. temporaria, a European species of “brown frog” (a terrestrial clade). Many people these days restrict R. to the Eurasian “brown frogs” and their sister-group, a small North American clade that contains the red-legged frogs like R. aurora. In this scheme, the North American bullfrog is no longer R. catesbeiana, but Lithobates catesbeianus.
    2) Whether Gerobatrachus has anything to do with Lissamphibia is very much an open question. We discuss this in our paper at some length, and we’ll have a bit more to say in the next paper (of which I’ll submit a revised draft soon).
    3) Sure, Triadobatrachus is an unquestioned salientian (total-group frog). But why do you think it’s an ancestor of anything known? Have you found it to lack autapomorphies?

    A sister to Gerobatrachus is Doleserpeton

    1) The only situation where a clade can have more than one sister-group is when they form a hard polytomy.
    2) In Fig. 1, Doleserpeton is the sister-group of Gerobatrachus.
    3) In some of my analyses, D. and G. likewise come out as sister-groups, though in others, G. clusters with Micropholis instead. In the analyses coauthored by Anderson, however, the sister-group of G. is Batrachia ( = frogs + salamanders), which is why its description calls G. a “stem batrachian”.

    Eocaecelia

    Eocaecilia. You get it right in the URL and in Fig. 1.

    (In Fig. 1, however, there are other mistakes: “Oradectes” should be Orobates, and “Romeria primus” should be Romeria prima.)

    Still, these nestings all appear to be good sister taxa without any “strange bedfellows” that just don’t seem to fit.

    I disagree strenuously on this one. I hardly even know where to start, so weird is your tree! Utegenia is an obvious seymouriamorph, you have the amphibamid dissorophoid temnospondyls paraphyletic with respect to the lepospondyls, the remaining temnospondyl – Eryops – just threw the towel and said “screw you, guys, I’m going home”, Adelospondylus looks like it came out where it did just because it’s long and possibly limbless (hint: add colosteids and a bunch of extra characters to your matrix), the obvious near-diapsid amniote Anthracodromeus fell out as the sister-group to Utaherpeton – one of the least amniote-like lepospondyls, probably close to lysorophians, lissamphibians, and holospondyls –, Silvanerpeton and Gephyrostegus are way too close for comfort to Amniota, Cephalerpeton is in a somewhat odd place, the relationships of the diadectomorphs are beyond description (see below), the synapsids are in two outright bizarre places, and the really basal Solenodonsaurus is deeply nested in the amniotes. Strange bedfellows? It’s a whole orgy of strange bedfellows.

    Diadectomorpha does not just contain the taxa underlain in pink, that’s just (part of) Diadectidae. Diadectomorpha also contains Tseajaia and Limnoscelis.

    Figure 2. Utegenia nests as the common ancestor of frogs, salamanders, caecelians and microsaurs.

    *literal, not metaphorical, bellyache*

    The fuck it does! You actually get it almost right in the text: “All three find an ancestor in sisters to Utegenia“. You got U. as the sister-group to your temnospondyl/lepospondyl/lissamphibian clade; that clade and U. share a common ancestor which was not Utegenia.

    I understand that you wanted to shorten the legend of Fig. 2. “Things should be made as simple as possible, but not any simpler”, said Einstein. You made them simpler than possible; you changed the meaning, and you didn’t even notice.

    Utegenia (Fig. 2) does not play a big part in the Marjanovic and Laurin (2013) study.

    Of course not; it’s an obvious seymouriamorph, and therefore no more relevant than any other seymouriamorph.

    Neither does Gephyrostegus.

    We couldn’t find a reason to think it was close to Lissamphibia… Also, the existing literature on it is quite unsatisfactory; at the SVP meeting of 2011, Jozef Klembara presented a poster from his evidently ongoing redescription of G., but that paper still hasn’t appeared. (I checked in Google Scholar a few minutes ago.) So, all I can really say is that Klembara’s reconstruction of the skull differs quite a bit from all previous ones – don’t rely on them.

    Marjanovic and Laurin (2013) separate ‘Amniota’ from Diadectomorpha, Solenodonsaurus and Westlothiana, but these are all amniotes in the large reptile tree.

    That’s a problem with your tree, not with our review paper.

    I’m sittng here in the Museum für Naturkunde in Berlin. Not only have I read the redescription of S.*, I’ve seen the material several times. There is no way that that beast is an amniote.

    If you compare the specimen drawings with the reconstruction in the monograph on Westlothiana, you’re bound to notice a few things. I conclude that W. is even less amniote-like than people used to suspect. My analyses consistently find it as the sister-group to all other amphibians ( = lepospondyls plus lissamphibians minus adelospondyls).

    * Marylène Danto, Florian Witzmann & Johannes Müller (2012): Redescription and phylogenetic relationships of Solenodonsaurus janenschi Broili, 1924, from the Late Carboniferous of Nýřany, Czech Republic. Fossil Record 15(2): 45–59.
    Warning: in a few places their coding contradicts their text or their figures, and they used the matrix by Ruta et al. (2003) instead of the updated version by Ruta & Coates (2007).

    Microsaurs, like Utaherpeton

    Calling U. a microsaur may not even make sense. But I’m still working on this.

    a clade of near salamanders known as the Albanerpetontidae. First time I’ve heard of them was in the Marjanovic and Laurin (2013) paper.

    *bellyache again*
    *drop in blood pressure*

    You talk about the origin of frogs, salamanders and caecilians, but didn’t even know about the fourth group of modern amphibians?!?!?

    *headdesk*

    I’m out of words.

    So I’ll comment on the other topic: as we explain at length in the paper, it’s by no means certain that the albanerpetontids are “near-salamanders”. Probably they’re not, and indeed most of the literature says they’re not. They could be near-batrachians, or near-caecilians, or they could lie (just) outside Lissamphibia altogether! A lot more research is needed.

    …Just let me mention that the name Albanerpetontidae (which should probably be Albanerpetidae, actually) dates from 1982, as do I, and the name Albanerpeton dates from 1976. It’s not like they were all discovered yesterday while you were looking in the other direction.

    Figure 3. Celtedens, an albanerpetontid

    Yes.

    close to salamanders.

    Or not.

    No jugal?

    A whole subsection of our paper deals with this question. That’s all I need to say.

    Gerobatrachus likewise has a gracile jugal and a lateral temporal fenestra.

    That’s the first time I hear of G., or any temnospondyl, having a temporal fenestra. Are you aware the skull is preserved in ventral view? …Your page on G., where you misspell the species name G. hottoni as “G. hotteni”, shows temporal fenestrae that exactly match the subtemporal fenestrae in the palate. Somehow you’ve arrived at the conclusion that they had no roof at all, and the jaw muscles had nowhere to attach. I think you’re wrong… epically so.

    The large reptile tree found tusks absent in reptiles and lissamphibia + microsauria, but present in other tested non-amniotes, including Cacops.

    That’s precisely our point: Cacops is a dissorophoid temnospondyl. Even you (Fig. 1) find it next to the dissorophoid temnospondyl Amphibamus.

  2. Sorry, I forgot: Eoserpeton, which shows up in Fig. 1, is a junior synonym of something; I don’t even remember of what – it was sunk so long ago that it hasn’t appeared in the literature for several decades.

  3. “the remaining temnospondyl – Eryops – just threw the towel and said “screw you, guys, I’m going home””

    In Peters’ defense, Eryops IS known for being fat…

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s