More data on the holotype of Pappochelys

Earlier we looked at Pappochelys (Figs. 1-4), touted as a stem turtle, but nesting in the large reptile tree at the base of the Placodontia and Sauropterygia.

I found online
a pretty good photo of the holotype of Pappochelys (SMNS 91360, Fig. 1). This taxon is also known from about 19 other less complete referred specimens, many including bones that fill in gaps left by the holotype. These specimens also document a variety of ontogenetic ages and sizes, as noted earlier.

Originally Pappochelys was considered a stem turtle
with no more than nine dorsal vertebrae and ribs (Fig. 2). Here, in the large reptile tree, it was recovered as a basal placodont close to Palatodonta, not related to turtles like Proganochelys and Odontochelys.33-37 steps are added when Pappochelys is forced to nest with turtles.

This is the holotype
of Pappochelys, SMNS 91360 (Fig. 1). Click to enlarge. It seems like there are more ribs here and the number of gastralia appears to suggest a longer torso than originally imagined (Fig. 2). The holotype may not represent the entire dorsal series, whether in vertebrae or ribs.

Figure 1. Schock and Sues counted a maximum of nine dorsal vertebrae, but is that the total number of dorsal vertebrate in Pappochelys. It is if you want it to be a turtle. It may not be if you reconstruct the holotype like this. Click to enlarge. Note the differences in gastralia tip shapes. Note the difference in sizes of the dorsal vertebrae. The longest dorsal ribs appear to be skewed toward the posterior torso. Certainly these images are not 100% correct. Instead they represent a best guess based on the data.

Figure 1. Schock and Sues counted a maximum of nine dorsal vertebrae, but is that the total number of dorsal vertebrate in Pappochelys. Nine is a good number if you want Pappochelys to be a turtle. It may not be turtle if you reconstruct the holotype like this or add Pappochelys to the large reptile tree. Click to enlarge.  Note the difference in sizes of the dorsal vertebrae. The longest dorsal ribs appear to be skewed toward the posterior torso, so I did the same with the scattered gastralia. The color tracings of the gastralia were originally traced from the photo and transferred to the original drawing, then transferred again into a regular order. Certainly these images are not 100% correct. Instead they represent a best guess based on the data. The gastralia are so scattered that one cannot determine that some tips point anteriorly, especially since Pappochelys is not related to turtles.

It is always better to use just one specimen in analysis. 
That is why I have revised the reconstruction based on the holotype. In this case we trust Schoch and Sues with regard to their list of referred specimens, as they trust their own judgement. There are no other vaguely similar taxa recognized as present in these strata and referred specimens preserve key elements not found in the holotype, like the skull.

What Schoch and Sues identified
as a femur on the holotype (Fig.1) is the shape and size of a posterior dorsal rib (#6, Fig. 3). A smaller, better femur is found on another specimen and it has an offset proximal head (Fig. 3) not illustrated by Shoch and Sues in situ, only in their reconstruction (Fig. 2).

In the Schoch and Sues reconstruction (Fig. 2)
the dorsal view reconstruction shows a much larger gastralia basket than the lateral view shows (Fig 2). That’s not scientific. The Schoch freehand drawings also indicate the pectoral girdle migrating beneath the anterior dorsal ribs in lateral view based on this incongruity. That’s what turtles do, but Pappochelys is not related to turtles. So that’s imaginary and hopeful.

The Shoch and Sues tracing (Fig. 1)
shows a straight rib (#3) with a T-shaped cross section, but the Schoch reconstruction does not show any large straight ribs. The dorsal ribs all bend posteriorly in the Shoch reconstruction except the small, short anterior rib.

None but
the first and second of the in situ caudal vertebrae appear to have any transverse processes, yet Schoch illustrated transverse processes on all the caudals and scored them as appearing beyond the fifth one caudally.

The coracoid
identified by Schoch and Sues in the holotype (Fig 3) is the same size and shape as the ischium, but it could still be a coracoid.

The pubis
identified by Shoch and Sues in the holotype (Fig 3) has no articular surfaces that fit the ilium. That ‘bone’ is here identified as the pubis AND ischium with appropriate articular surfaces that fit the ilium.

The ilium
of the 91895 specimen does not match the ilium of the holotype, which is narrower in all respects.

Figure 2. Pappochelys according to R. Schoch. Note the mismatch in the gastralia (ghosted area in lateral view, red bones otherwise).

Figure 2. Pappochelys according to R. Schoch. Note the mismatch in the gastralia (ghosted area in lateral view, red bones otherwise). Schoch and Sues presented Pappochelys as a turtle ancestor, but created distortions like those above to do so. Compare to bone tracings in figure 1.

Schoch and Sues provided a freehand drawing
of Pappochelys (Fig. 2), uniting parts from several specimens and filling in gaps where necessary. Freehand drawings are always biased (see above). There’s no way to get around it. It’s better Science to trace the original elements precisely (Fig.1), no matter if the bones are crushed or broken. Using DGS permits one to segregate some bones from others and lift, rotate and shift them, as is, to create a reconstruction.

Figure 3. Comparing elements from different Pappochelys specimens to scale. The ilia do not match. Nor do the femora. The 91360 specimen is identified here as a dorsal rib. What Schoch and Sues identified as a pubis would not fit on the ilium of the same specimen. I think they show two bones, the pubis and the ischium as illustrated here.

Figure 3. Comparing elements from different Pappochelys specimens to scale. The ilia do not match, even accounting for breaks. Nor do the femora. The 91360 specimen is identified here as a dorsal rib. What Schoch and Sues identified as a pubis would not fit on the ilium of the same specimen. I think they show two bones, the pubis and the ischium as illustrated here with appropriate articular surfaces.

More precision
was needed in the Schoch and Sues study. More taxon inclusion was also needed in their small focused analysis.

Figure 4. Pappochelys overall revised. Note that various specimen numbers may indicate different size individuals. All pectoral elements are individual specimens.

Figure 4. Pappochelys overall revised. Note that various specimen numbers may indicate different size individuals. All pectoral elements are individual specimens. Pappochelys may have had fewer dorsal vertebrae than its placodont sisters. Or not. A more complete dorsal series will tell us.

While the above rendition is more precise than before,
even more precision is necessary to complete the task of positively identifying each bone. We don’t know how many dorsal vertebrae were present. A sister taxon, Majiashanosaurus appears to have 18. See the skull in more detail here.

Based on the 92068 rib,
(Fig. 4) some Pappochelys specimens (or a species similar to it) grew to be twice as large as the holotype.

References
Schoch RR and Sues H-D 2015. A Middle Triassic stem-turtle and the evolution of the turtle body plan. Nature (advance online publication) doi:10.1038/nature14472 online

Advertisements

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