So this week, I want to highlight a few of the amphibamiforms that are not always included in phylogenetic analyses, in part because some of them have clear destabilizing effects on the phylogenetic analysis. There are at least 16 recognized non-branchiosaurid amphibamiforms (so the classic 'amphibamids'), but I would say that only half of those are always included in dissorophoid / amphibamiform analyses, such as Amphibamus. But if you think about it, not including some taxa doesn't mean they don't exist, and it's important to be mindful that excluding taxa can skew the picture, whether intended to or not. As always, good to be constructively critical of methodologies and to think about how they can bias results!
Authors: Rainer Schoch, Andrew Milner
Lives at: University of California Museum of Paleontology, Berkeley, CA
General comments: I won't belabour this one too much since I'm the most recent one to publish it and did one of my usual blog posts covering the paper when it came out earlier this year. Nanobamus actually ties in well with the longstanding conundrum in the identification of small-bodied dissorophoids that I discussed last week – the one and only known specimen was first described in 1985 as a larval trematopid by E.C. Olson. If his identification had been correct, it would have been the only record of a larval trematopid, which would have made it quite significant, but Olson's interpretation was challenged by David Dilkes in 1991, whose interpretation as "at least not a larval trematopid" (more or less) was accepted by later workers. Rainer Schoch later mentioned it as an amphibamid (in the historical sense) in a 2002 paper as an aside, but it was never formalized as a distinct taxon until 2014 in the Paleoherp Handbook. I suspect that one reason that this taxon has remained relatively neglected is because the holotype apparently lived in my former lab since the 1990s...
Tersomius package deal! I've mentioned before that Tersomius is often regarded as a bit of a wastebasket taxon among amphibamiforms because it's somewhat questionable whether the material referred to it all belongs to Tersomius, and at least some material previously described has since been proven to belong to other amphibamiforms (Pasawioops), microlerpetids (Eimerisaurus), or even to a dissorophid (Reiszerpeton) (Maddin et al., 2013). Material of Plemmyradytes was tentatively considered to be Tersomius in cursory reports prior to its formal publication. Much of the confusion is well-summarized by Huttenlocker et al. (2007: p. 326). Suffice to say it's a sloppy mess.
Can we get a rehash?
Below are a few phylogenies that included at least some of the various taxa that I've discussed here. I colour-coded boxes around them just to highlight them and am using differently coloured circles to indicate major nodes within Amphibamiformes (and Amphibamiformes itself).
What I want to point out here is that including these various "somewhat forgotten" taxa has drastically different outcomes on the phylogeny. For example, by definition of Schoch (2018), Micropholidae is the most exclusive group including Micropholis stowi and Tersomius texensis. In the phylogeny of Anderson & Bolt (2013), that actually ends up including every terrestrial amphibamiform. Similarly, Plemmyradytes is sometimes a micropholid (Huttenlocker et al., 2007; Bourget & Anderson, 2011) and sometimes an otherwise unplaced (not in a specific family) amphibamiform (Anderson & Bolt, 2013; Maddin et al., 2013). Some of the variability probably results from the inclusion/exclusion of Branchiosauridae (only included in Maddin et al. 2013 of the studies that I listed here); other studies have often showed a close relationship to Georgenthalia when branchiosaurids are included. Between the various studies here, Georgenthalia shifts around a lot though, something that's persisted since its description in 2008 (Anderson et al., 2008).
This harkens back to my trematopid phylogeny paper from earlier this year in which I demonstrated a total loss of resolution in the tree when including every single trematopid taxon (something that hadn't been done before; see the far left figure), followed by a restoration of resolution when I restricted the taxon sample to seven species like previous studies had done (B and D). People don't like to see a lack of resolution because it means you can't say very much, but if you can only say something by excluding data, what does that mean for those conclusions? This is an area of research I'm continuing to work on, so stay tuned!
About the blog
A blog on all things temnospondyl written by someone who spends too much time thinking about them. Covers all aspects of temnospondyl paleobiology and ongoing research (not just mine).