New publication: The amphibamiform Nanobamus macrorhinus from the early Permian of Texas (Gee & Reisz, Journal of Paleontology)

Title: The amphibamiform Nanobamus macrorhinus from the early Permian of Texas
​Authors: B.M. Gee, R.R. Reisz
Journal: Journal of Paleontology 
Link to paper and DOI: 10.1017/jpa.2019.72

Holotype skull of Nanobamus macrorhinus in dorsal view (figure 1 in our paper).

The history of Nanobamus
The holotype and only known specimen has only been known for a few decades, but this taxon's already accrued a fairly lengthy taxonomic history.

• 1985: American paleontologist E.C. Olson describes the specimen for the first time. He identifies it as a larval trematopid. Trematopids are the largest dissorophoids in the early Permian and are recognized by a particularly long nostril, the main feature which Olson uses to tie this specimen to the clade.
• 1991: American paleontologist (and Reisz lab alum) David Dilkes re-examines the specimen in light of subsequent studies on trematopid ontogeny and systematics. He points out that while the naris of this specimen is indeed very long, like in trematopids, the configuration of skull bones around the nostril is not the same as in any known trematopid. He also invalidates the other two features (internarial fontanelle, absence of semilunar curvature of the squamosal) that Olson used to support trematopid affinities. Based on this, Dilkes concludes that the specimen is definitely not a trematopid, but within Dissorophoidea, further resolution is not possible.
• 2002: German paleontologist Rainer Schoch notes that the specimen is definitely a relatively mature amphibamiform (amphibamid in the traditional sense) but makes no further taxonomic arguments. 
• 2014: Schoch and British paleontologist Andrew Milner formalize the specimen as the holotype of a new taxon, Nanobamus macrorhinus, in their Paleoherpetology Handbook volume. They diagnose the taxon and present a reconstructed line drawing derived mostly from Olson's and Dilkes' illustrations.
• 2018: I "rediscover" the holotype in my own lab, with an extremely outdated collections card that still bears its UCLA VP institutional designation (this collection is no longer active and is mostly in the UCMP at Berkeley now), which suggests that the specimen has never left the lab since David Dilkes was here (i.e. 1990) and that nobody has seen it since then (that's 29 years!). That leads to the present study, which fully describes the anatomy (with photographs) in light of its amphibamiform affinities.

The peculiarities of long nostrils
The classic poster child for weirdly long nostrils are trematopids, which often have a nostril that's at least as long as their eye (see Acheloma on the bottom right). Based on the position of the septomaxilla, a bone that sits within the external naris, we know that it wasn't just that their nasal organ got longer. So in other words, they probably didn't have a particularly acute sense of smell relative to other dissorophoids. However, it's not entirely clear what filled the back end of this long nostril. There have been a few suggestions (see Dilkes, 1993 for a summary), such as a salt gland, but there isn't much of a precedent for such a feature in modern amphibians that would provide a good analogue, and the inside of the nostril isn't particularly complex; most of the internal bony surface is smooth. Since all trematopids have a long nostril, it might relate to their generally large body size and predatory habits, but the independent appearance of a very similar nostril in the much smaller amphibamiforms is then even more peculiar. The nostril of Nanobamus is longer than that of Georgenthalia, but they both are formed by an incision into the lacrimal that still excludes the prefrontal from the naris (the prefrontal frames the naris dorsally in trematopids). 

Both lateral views of the holotype skull of Nanobamus macrorhinus ('en' = external naris in part 4; figure 2 from our paper).

Skull of an exemplar trematopid, Acheloma, with the extent of the naris indicated by the red bracket (modified from Polley & Reisz, 2011).

Illustration and interpretive line drawing of Georgenthalia clavinasica from Germany (figure from Anderson et al., 2008).

​How good is amphibamiform phylogeny?
Our analysis of Nanobamus in the recently published matrix of Schoch (2019) illustrates some good points for phylogenetics, namely that you can get a fully resolved topology with a low number of MPTs and high consensus values but weak support for most nodes of interest. In Schoch's original analysis, he recovered 10 most parsimonious trees (MPTs), which is pretty good, and most nodes in the tree are resolved, as you can see in the figure to the right. However, the taxon sample utilized by Schoch is already restricted to only the best-known amphibamiforms - it doesn't include more uncertain taxa like Tersomius dolesensis or more incomplete taxa like Plemmyradytes shintoni​, for example. This can be a good way to achieve resolution in your tree, but it is naturally incomplete because it intentionally omits known taxa and their associated data, even if those data are more incomplete than those of the included taxa. 

As you can see, the best supported nodes are those that aren't directly related to terrestrial amphibamiforms (Doleserpeton down through Eoscopus). The best nodes are nodes like crown Lissamphibia (Triadobatrachus, Karaurus, and Eocaecilia) or the ones for branchiosaurids (Branchiosaurus through Schoenfelderpeton). Virtually all of the terrestrial amphibamiforms are not well-supported beyond modest support for the newly defined Amphibamidae (Amphibamus + Doleserpeton) and Micropholidae (Pasawioops, Tersomius, Micropholis​). 

What explains this? Amphibamiforms often have interesting mixtures of primitive and derived traits. Some of these might be convergent, even within amphibamiforms. For example, both Nanobamus and Georgenthalia have an elongate naris that is distinct from that of trematopids and essentially identical to each other, but they are not recovered as being closely related, which either indicates that the phylogeny is biased by some methodology (e.g., taxon sampling) or that this feature appeared twice in amphibamiforms, which would be quite interesting. The phylogeny also doesn't accord with the stratigraphic occurrence of taxa (not that it has to, but it is interesting to note). For example, Micropholis is one of the most primitive amphibamiforms, but it's the youngest one in the fossil record by a long shot (Early Triassic of South Africa). In contrast, Amphibamus is one of the most derived amphibamiforms, but it's one of the oldest (Carboniferous of North America). This can indicate many things, such as long ghost lineages with missing data from the fossil record, or problems with the existing character selection, or convergence that's being interpreted by the algorithms as the product of shared ancestry. All exciting things to keep exploring in the future!

---

Refs

• ​Anderson JS, Henrici AC, Sumida SS, Martens T, Berman DS. 2008. Georgenthalia clavinasica, a new genus and species of dissorophoid temnospondyl from the Early Permian of Germany, and the relationships of the family Amphibamidae. Journal of Vertebrate Paleontology, 28(1):61-75. doi: ​10.1671/0272-4634(2008)28[61:GCANGA]2.0.CO;2
• Dilkes DW. 1991. Reinterpretation of a larval dissorophoid amphibian from the Lower Permian of Texas. Canadian Journal of Earth Sciences, 28(9):1488-1492. doi: 10.1139/e91-130
• Dilkes DW. 1993. Biology and evolution of the nasal region in trematopid amphibians. Palaeontology, 36:839-839. [link]
  
• Olson EC. 1985. A larval specimen of a trematopsid (Amphibia: Temnospondyli). Journal of Paleontology, 59(5):1173-1180. [link]
• Polley BP, Reisz RR. 2011. A new Lower Permian trematopid (Temnospondyli: Dissorophoidea) from Richards Spur, Oklahoma. Zoological Journal of the Linnean Society, 161(4):789-815. doi: 10.1111/j.1096-3642.2010.00668.x
• Schoch RR. 2002. The evolution of metamorphosis in temnospondyls. Lethaia, 35(4):309-327. doi: 10.1111/j.1502-3931.2002.tb00091.x
• Schoch RR. 2019. The putative lissamphibian stem-group: phylogeny and evolution of the dissorophoid temnospondyls. Journal of Paleontology, 93(1):137-156. doi: ​10.1017/jpa.2018.67
• Schoch RR, Milner AR. 2014. Handbook of Paleoherpetology: Temnospondyli I (Part 3A2). München: Verlag Dr. Friedrich Pfeil; 150 pp.