Earlier in six prior posts we looked at some new basal amniotes revealed by phylogenetic bracketing and phylogenetic analysis. Data was gleaned by DGS, Digital Graphic Segregation, a technique that is currently used by a few paleontologists and should be used more often by more of them as you’ll see in the present demonstration.
Figure 1. Gephyrostegus watsoni as traced by Carroll 1970 using traditional methods. Here just the most prominent bones are identified leaving many unknown. Where are the gastralia? Where are the vertebrae?
DGS – Digital Graphic Segregation
has been getting a bad rap for a long time. Here, once again, I was able to find more bones than did prior workers not using DGS. Instead they examined these basal amniotes first hand and created tracings or sketches in their own manner, often without great precision and too often leaving out bones that were indeed present (Fig. 1).
Here’s a good chance to judge the results for yourself.
If this is voodoo, if this is useless, ignore it. If you think it has value, embrace it. Click here to see a rollover image of Gephyrostegus watsoni, both in situ and with bones colorized. The original image was 600 dpi. The presentation on the web is at 72 dpi. Even so you’ll have trouble seeing everything. Sometimes it takes awhile. I can only share my results and encourage you to experiment on your own.
Figure 2. Click to enlarge and see rollover image. Here DGS, digital graphic segregation, enabled the identification of many more bones than firsthand observation, including the displaced carpals and tarsals, along with a few insects and egg-shapes. Originally the some bones were on one layer, others on added layers. Remember, reconstruction is also part of this process. Reconstruction reminds you which bones are missing and need to be found.
Gephyrostegus watsoni
is a crushed Westphalian (310 mya) amniote currently considered to be an anamniote juvenile of Gephyrostegus bohemicus. It was traced by Brough and Brough (1967) and Carroll (1970, Fig. 1). Brough and Brough determined that it was sufficiently distinct from the holotype of G. bohemicus to erect a new species. Carroll did not recognized those differences and so considered it a juvenile lacking carpals and tarsals, having a large skull with short rostrum and other traditional juvenile traits. Klembara et al. (2014) agreed.
DGS found more bones than firsthand observation and enabled a precise reconstruction (Fig. 3). Tracing the bones in color enables one to lift those bones, as they are, to create a more accurate reconstruction while minimizing handwork that could introduce error.
Figure 3. Reconstruction of G. watsoni as a distinctly different genus, nesting with Eldeceeon rather than G. bohemicus. DGS was key to recovering this data.
Phylogenetic analysis nests G. watsoni with Eldeceeon (Fig. 4), not with G. bohemicus. So this specimen is not a juvenile and it needs a new generic name. DGS was key to recovering the data found here. If you take a look at the specimen with colorized bones, you’ll soon realize that the several layers would leave a pencil and a prism in the dust. On the computer monitor tracing becomes simpler pulling bones out of the chaos on the matrix layer by layer.
Figure 4. Click to enlarge and see the rollover. Eldeceeon with a strangely expanded belly (defined by gastralia/scales) that could have contained a load of eggs, traced in green here.
And here’s a second example
Eldeceeon is a Viséan amniote known from another crushed skeleton (Fig. 4). Here I was able to find more bones than in prior tracings (Fig. 5) and create a more accurate reconstruction (Fig. 6) than created by prior workers (Fig. 7).
Figure 5. Eldeceeon as traced by Smithson 1994. Colorized manus and pes added by me.
Note that drawings of bones often unlabeled, don’t tell the whole story. By colorizing each bone and using the same color for the left and right counterparts the chaos is reduced and reconstructions can be created with ease.
Figure 6. Two specimens attributed to Eldeceeon that nest together. The holotype is the one in figure 4. Compare this reconstruction to one produced earlier, shown in figure 6.
These two Eldeceeon specimens (Fig. 6) nest together, but would clearly be distinct genera if they lived in the modern world. This also means that if you use the skull of one on the body of the other, you will create a chimaera, which only leads to phylogenetic trouble. See the family tree of basal amniotes here. See basal amniotes to scale here.
Figure 7. Eldeceeon as reconstructed by Smithson 1994 (gray area added). Anterior skull is based on the referred Eldeceeon specimen. Even the rib count is off. Note the large size of the pelvis and too short torso, traits that would be errors if entered into phylogenetic analysis.
Data from the literature
While we all have to rely on specimen drawings and reconstructions, that’s not always a good idea, as this little exercise demonstrates. After DGS I have more confidence that the reconstruction is more accurate.
The upshot is
with DGS I was able to more accurately nest these taxa on this side of the anamniote/amnote transition and shed new light on this important stage in the evolution of amniotes/reptiles, including you and me. Making discoveries like this is richly rewarding. The extra effort used to create DGS is definitely worth the extra effort.
I hope
this demonstration puts an end to the bad rap that DGS has been getting.
And a big hello
to all the paleontologists in Berlin attending the SVP convention there.
References
Brough MC and Brough J 1967. The Genus Gephyrostegus. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 252 (776): 147–165.
Carroll RL 1970. The Ancestry of Reptiles. Philosophical Transactions of the Royal Society London B 257:267–308. online pdf
Klembara J, Clack J, Milner AR and Ruta M 2014. Cranial anatomy, ontogeny, and relationships of the Late Carboniferous tetrapod Gephyrostegus bohemicus Jaekel, 1902. Journal of Vertebrate Paleontology 34:774–792.
Smithson TR 1994. Eldeceeon rolfei, a new reptiliomorph from the Viséan of East Kirkton, West Lothian, Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences 84 (3-4): 377–382.
The Pterosaur Heresies 
But how could it, when you apply it to a Carroll drawing? In my experience and that of at least one other person, Carroll’s drawings – even his specimen drawings, not just his reconstructions – are often idealized, sometimes to the point of being outright unreliable.
Um, David, I applied DGS to a photo, a high rez photo of the specimen — not the drawing by Carroll. Carroll examined the specimen firsthand and did not find many of the bones found here an ocean and half a continent away from the specimen. That’s the value DGS brings to 2-D specimens.