I was working on a homepage graphic (look for it toward the bottom) listing dozens of basal reptile taxa on a timeline. I discovered that Romeriscus (Figs. 1, 2) was improperly nested chronologically and therefore, at least in this case, phylogenetically.
Romeriscus is represented by a crushed fossil
that has proven notoriously difficult to interpret (Fig. 1) and nest by all prior workers (see below). I reinterpreted the elements yesterday and now it seems to make more sense, both time-wise and morphologically. It now appears to be a basal archosauromorph reptile, a sister to Gephyrostegus watsoni (Fig. 2; which is not a sister to the holotype of the genus, G. bohemicus, and so G. watsoni needs a new name.)
I nested Romeriscus between Emeroleter and Lanthanosuchus, reconstructing it with a rather flat, wide skull. Today, I reconstruct the skull a bit less wide, but still >2x wider than tall. Back in the Westphalian, there was not so much diversity in reptiles, so what Romeriscus would look like, using DGS or not, was rather limited and therefor greatly simplified. Skull width can often be determined from occipital and/or palatal elements, both of which are difficult to find on this fossil. Here (Fig. 2) skull width was determined by measuring the distance between the bases of the two still parallel quadrates, assuming that they did not move much during taphonomy.
is reconstructed like and nests with another basal archosauromorph reptile, G. watsoni. Distinct from some, but not all basal gephyrostegids the limbs of Romeriscus are subequal in length. More below.
Romeriscus perialllus (Baird and Carroll 1967, Early Pennsylvanian (Westphalian A, 306 mya, YPM-PU16 482) was originally considered a limnoscelid reptile and the earliest and most primitive reptile known in 1967. Laurin and Reisz (1992) considered it a tetrapod of unknown affinity and they considered limnoscelids as anamniotes. They were unable to distinguish the various skull elements, but reinerpreted several of the interpretations of Baird and Carroll.
Using DGS I was able to identify most of the bones on the plate (Fig. 1). In the large reptile tree the resulting reconstruction (Fig. 2) nests with Gephyrostegus watsoni (Fig. 3).
Distinct from G. watsoni,
Romeriscus was twice as large overall and with a more erect occiput. The orbit was 2x longer than tall. The naris was larger. The fore and hind limbs were similar in size. The frontals were wider and the parietals were narrower. The hind limb was narrower than the forelimb.
To those who think ReptileEvolution.com should still be ignored, the challenge is still out there… which taxa on the tree are improperly nested. I just found one, but I guess I’m finding ‘the low hanging fruit.’ the easy mistakes to pick out. Not sure if there are any more errors hidden in the tree. Only time and a critical eye will tell. Not happy finding any errors here, but I am happy correcting them.
Science is not okay with errors.
Science is better with corrections, but first the errors have to be uncovered. Look for things that just doesn’t fit established patterns. It could be that the established pattern is incorrect.
Baird and Carroll were correct in 1967.
Romeriscus was one of the most primitive reptiles known in that era. However, I don’t think they would have considered it a reptile if they knew Romeriscus retained an intertemporal bone, like many of its sisters. The loss of the intertemporal was considered an amniote trait, but the advent of the amniote egg preceded that loss according to the large reptile tree.
You can’t tell
what is a reptile and what is not by a list of traditional skeletal traits that somebody decided without a phylogenetic analysis. That sort of problem has hung up paleontologists for decades. You can only tell which taxa are the most basal reptiles by recovering the last common ancestor of dinosaurs and lizards in phylogenetic analysis. Presently that taxon is a sister to Gephyrostegus bohemicus, a late survivor (310 mya) of a much earlier (345 mya) reptile evolution and radiation from seymouriamorph sisters like Silvanerpeton (335 mya).
Speaking of errors,
if you see any here, with regard to Romeriscus or any other taxa, let me know ASAP, so I can clear those up.
Baird D and Carroll RL 1967. Romeriscus, the oldest known reptile. Science. 1967 Jul 7;157(3784):56-9.
Laurin M and Reisz RR 1992. A reassessment of the Pennsylvanian tetrapod Romeriscus. Journal of Vertebrate Paleontology 12(4): 524-527.