Asaphestera: the earliest amniote? …No

Summary if you’re in a rush:
Mann et al. 2020 mistakenly reassessed their ‘microsaur’, Asaphestera platyris (Fig. 1), as at ‘the earliest synapsid’. The LRT nests this taxon as a microsaur after demonstrating interpretation and reconstruction errors.

Mann et al. 2020 bring us
their view of ‘microsaurs’ from Joggins, Nova Scotia (Westphalian, Late Carboniferous) with the recognition that Asaphestera platyris as a synapsid provides the earliest unambiguous evidence of ‘mammal-like reptiles’ in the fossil record.”

Unambiguous?
No. Just because they say so, does not mean it is true.

By contrast (1):
In the large reptile tree (LRT, 1685+ taxa) the earliest amniote (determined by the last common ancestor method) is Silvanerpeton, from the Viséan (Early Carboniferous) at least 15 million years earlier. Gephyrostegus is more primitive in the LRT, but appears as a late survivor in the Westphalian (Late Carboniferous, coeval with Asaphestera platyris) of an earlier radiation. Archaeothyris, another slightly younger Westphalian taxon, was widely considered the earliest known synapsid, and remains so in the LRT. These three taxa are not mentioned in the Mann et al. text.

Figure 1. Asapehestera platyris in situ, traced by Mann et al. 2020, then traced and reconstructed using DGS methods.

Figure 1. Asapehestera platyris in situ, traced by Mann et al. 2020, then traced and reconstructed using DGS methods. There is no tall dorsal process to the maxilla, contra Mann et al. The ‘palbebral’ (PB) is below several loose dentary teeth, so it is a palate or mandible element. The ‘dorsal process’ of the maxilla is not represented by bone. The reconstruction nearly matches Kirktonecta (figure 2).

By contrast (2): 
When added to the LRT, ‘Asaphestera platyris’ (RM 2.1192, Steen 1934) nests with and is not much different from the microsaur, Kirktonecta (Fig. 2), far from any amniotes or synapsids. Kirktonecta is mentioned only once in the Mann et al text as part of a list that “do not fit clearly into this [microbrachomorph] framework.”

Figure 2. Kirktonecta is a Viséan taxon nesting with Gephyrostegus in the LRT.

Figure 2. Kirktonecta is a Viséan taxon nesting with Asaphestera platyris in the LRT.

From the abstract:
‘‘Microsaurs’ are traditionally considered to be lepospondyl non-amniotes, but recent analyses have recovered a subset of ‘microsaurs’, the fossorially adapted Recumbirostra, within Amniota.”

The LRT does not support this nesting.

Recumbirostra = pantylids, gymnarthrids, brachystelechids, ostodolepids, and rhynchonkids. In the LRT all these taxa are in the clade Microsauria, a sister clade to the Reptilomorpha. Kirktonecta is basal to the microsaur clade that ultimately produced the extant caecilian, Dermophis.

From the Mann et al. description:
“Most of the right maxilla and portions of both temporal regions are known only from impressions of the bones that have weathered away; nevertheless, valuable information is present in what remains. Parts of the dorsal margins of both temporal fenestrae are preserved on either side of the cranium, but the morphology is more completely represented on the right side.”

A reconstruction (Fig. 1) based on the same specimen does not support this description. There is no tall dorsal process to the maxilla, contra Mann et al. The ‘palbebral’ (PB) is below several loose dentary teeth, so it is a palate or mandible element. The ‘dorsal process’ of the maxilla is not represented by bone. The reconstruction nearly matches Kirktonecta (Fig. 2).

From the text:
“As a result, we tentatively attribute RM 2.1192 (Fig. 1) to the Eothyrididae. If this identification is correct, RM 2.1192 would extend the record of eothyridids substantially.”

Co-author, B Gee,
writing on his blogpost (link below) reported, “Among synapsids, this specimen most closely resembles the eothyridids, although it shares a number of features with acleistorhinid parareptiles, which were often confused for eothyridids in their earlier history of study (perhaps they still are eothyridids?).”

In the LRT, even eothyrids are not synapsids. They are basal caseasauria derived from the lepidosauromorph, Milleretta.

In summary:
Mann et al. 2020 mistakenly reassessed their microsaur, Asaphestera platyris, as a synapsid. The LRT nests it as a microsaur close to Kirktonecta, a taxon essentially overlooked by the authors. Nearly coeval Archaeothyris remains the earliest known synapsid, but several synapsids are more primitive, indicating an earlier radiation. So, they’re out there somewhere! Mann et al. did not find them…yet.

Postscript:
A reader (J) wondered how I was able to reconstruct Kirktonecta if, given the limitations provided by another reader (DM) that only the inside of the skull bones were visible. Here (Fig. 3) I show the method and the data, a crushed skull in which the bones are slightly separated along their sutures and sometimes split during taphonomic crushing. I traced the skull bones of Kirktonecta, then reassembled them using the DGS method (color tracing using Photoshop). The first step was to invert the colors (creating a negative) of the original image, something a paleontologist with firsthand access to the specimen would be unable to do without repeating this method. The original image had higher resolution, reduced here for online publication. Apparently the insides were little different from the outsides given the two-dimensional, plate-like shapes of the skull bones with few-to-no complex curves in the bones of this taxon. I leave it to the reader to decide whether or not the DGS method was successful in this case, whether inside or not.

Figure 3. Kirktonecta in situ and traced using the DGS method.

Figure 3. Kirktonecta in situ and traced using the DGS method.

 

References
Mann A et al. (7 co-authors) 2020. Reassessment of historic ‘microsaurs’ from Joggins, Nova Scotia, reveals hidden diversity in the earliest amniote ecosystem. Papers in Palaeontology 2020:1–17.
Steen MC 1934. The amphibian fauna from the South Joggins. Nova Scotia. Journal of Zoology, 104, 465–504.

wiki/Kirktonecta
wiki/Asaphestera
wiki/Asaphestera2
wiki/Carboniferous

https://bryangee.weebly.com/blog/new-publication-reassessment-of-historic-microsaurs-from-joggins-nova-scotia-reveals-hidden-diversity-in-the-earliest-amniote-ecosystem-mann-et-al-2020-papers-in-palaeontology

13 thoughts on “Asaphestera: the earliest amniote? …No

  1. The ‘dorsal process’ of the maxilla is not represented by bone.

    Indeed not. It’s represented by an impression. The bone has fallen off; or maybe it’s on the lost counterslab.

    The dentary you’re interpreting into the rock, on the other hand, is represented neither by bone nor by an impression.

    How you identify loose teeth as belonging to the dentary is beyond me, too, as is your certainty that teeth from the dentary can’t have moved above the eye during compaction of the fossil. BTW, palpebral.

    Kirktonecta is mentioned only once in the Mann et al text

    That’s because very little can be done with it in its present state. It is split through the bone; we’re seeing all the bones from the inside, their surfaces are not exposed. The specimen should either be glued back together and prepared from the outside, or µCT-scanned. There is no point in trying to include it in a phylogenetic analysis or drawing pretty much any conclusions from it.

    So, they’re out there somewhere!

    Well, as we mention in the paper, there’s Protoclepsydrops from the same site that has been interpreted as a possible synapsid and may even be junior synonym of Asaphestera. Unfortunately it’s too incomplete for comparison.

    • Glad you weighed in, David. The dorsal process of the maxilla is not represented by bone that was EVER there, impression or not. Excluding taxa remains the issue. Kirktonecta is the best match in the LRT. And it’s not a synapsid.

      • How is Kirktonecta the best match if, as David Marjanovic mentioned, you are only seeing the inside of the skull bones? Based on these criteria, Kirktonecta shouldn’t even be in your analysis as you shouldn’t be able to find more than a few characters that you can code for your matrix.

      • Glad you weighed in, David. The dorsal process of the maxilla is not represented by bone that was EVER there, impression or not.

        The impression is there, so there must have been bone at some point.

        In case you’re wondering about cartilage, the maxilla is a dermal bone. It is not preformed in cartilage and cannot have cartilaginous extensions.

        Kirktonecta is the best match in the LRT.

        How do you score Kirktonecta for anything but general body proportions and monospondyly if you can’t see its bone surfaces?

        Even if you find sutures in the skull (criss-crossed by breaks as it is), are you aware that sutures are three-dimensional surfaces? They don’t always match on the inside and the outside. For instance, postparietals are routinely longer on the ventral than on the dorsal side of the skull table, intertemporals are much smaller ventrally than dorsally, the preopercular of whatcheeriids is entirely superficial and doesn’t reach the medial side of the skull roof at all, and so on.

      • Take a look at the text again. I’ve added a postscript to the bottom which includes the data I used to reconstruct the skull of Kirktonecta. Hopefully it will answer your questions.

      • I just checked the figures in the paper that you did DGS on.

        Their resolution is appalling. Enlarged to 300%, the pixels of each photo are the size of four pixels of my screen. By inverting the colors, you’ve turned a whitish blur into a blackish blur, but that doesn’t reveal any details that aren’t in the photos to begin with.

      • You’re saying it’s impossible. I’m showing it’s possible. When we find another Kirktonecta, I suppose that will make things easier to judge.

      • Understood. I’ve been using the Photoshop method since 2003 (Jeholopterus) and I continue to both succeed and make mistakes with it. Every specimen or image is different. I am like a naive freshman with every new image, not knowing what I’m going to get. No one likes the idea that an amateur on the other side of the world can glean data from a specimen that is presently under the watchful gaze of a PhD. However, as workers are coming to realize (using µCT scans) this method of coloring bones reduces the chaos of crushed mono-color fossils by picking them apart piece-by-broken, over/underlapping piece and then uses those unaltered pieces to create a reconstruction with maximum accuracy (= minimum freehand). Kirktonecta is one of those crushed specimens that has been subjected to this technique. And now Asaphestera joins that list.

      • A wee bit of sSupporting evidence: Wikipedia reports, “It [Kirktonecta] represents the first microsaur to be discovered in the United Kingdom and the earliest occurrence of a microsaur in the fossil record. The next oldest being Utaherpeton.” Not surprisingly, both entered the LRT at basalmost nodes within the clade Microsauria. Doesn’t prove anything, but if the reconstruction of the skull was wrong, the nesting should or would have been elsewhere.

    • I just found this thread again by googling Stegotretus, oddly enough.

      Doesn’t prove anything, but if the reconstruction of the skull was wrong, the nesting should or would have been elsewhere.

      I’m afraid that logic is circular. Sure, if the reconstruction were wholly random, Kirktonecta would come out in a random place; but if it’s only off in like 10% of the scores, it could come out in some reasonably close but still wrong place.

      Your confident assertion that the skull bones were for practical purposes two-dimensional is not currently testable. Someone will really need to glue the specimen back together and prepare it from the outside, or µCT-scan it, as i said.

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 )

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.