Some people are still having trouble with DGS as a technique. They think of it as something that is virtually guaranteed to spook a reconstruction. Instead of increasing confidence that parts have been correctly identified, they have no confidence in work that has the taint of DGS.
Here’s a step-by-step run through DGS on a familiar specimen, Eudimorphodon ranzii. Using DGS enabled the recognition of some oddly long posterior ribs (that were always visible, just ignored) and a wider than deep torso in a pterosaur for which these traits were not otherwise recorded.
Eudimorphdon ranzii (Zambelli 1973, Wild 1978) s a Late Triassic pterosaur known from an articulated crushed skeleton missing feet, tail and most of each wing (Figs. 1-3). Some parts are easy to see and trace, like the skull and sternal complex. Some parts are more difficult like the two pubes (Wild 1978 only found one by combining the two into an oddly broad prepubis), the pelvis, and the odd arrangement of the posterior ribs.
Figure 1. Eudimorphdon ranzii with post cranial bones colorized.
Step one: Colorize the bones (Fig. 1)
Darren Naish seems to think this is okay if you know which bone is which ahead of time when looking at the specimen and you’re just making a visual presentation. I like to take it one step further and use DGS to segregate bones that are more difficult to identify. Here the pelvis is found. The dorsal ribs will precisely transferred to the reconstruction, not generically applied. As we’ve learned earlier, sometimes pterosaurs have the cross section of a horned lizard.
Figure 2. The colorized bones on a fresh canvas. Most tetrapods have shorter posterior dorsal ribs, but not here in Eudimorphodon. Lighter tones on the pelvis represent overlying bones, in this case vertebrae. It is important to put a numeral on each vert and rib because it is otherwise easy to become confused.
Step two: Transfer the colorized bones onto a fresh white background (Fig. 2)
Here we’re just trying to put the bones on a fresh canvas. You’ll note some bones are estimates based on vague clues as they appear beneath the sternal complex.
Figure 3. Moving colorized bones into a rough reconstruction or Eudimorphodon. Here both pelves are shown as they appeared in situ. In figure 1 I jumped the gun and put the parts together.
Step three: Move the colorized bones into a rough assembly (Fig. 3)
Here we’re just trying estimate a body shape to make tracing the colored bones easier.
Figure 4. Lateral, dorsal and cross-sectional views of Eudimorphodon ranzii. Note the overlap of the posterior ribs over the hind limbs and the very wide torso. The cross section shows the 2nd dorsal ribs and the 23rd. Note the small ischium which could only produce small eggs. A little taller and wider than we thought before. The forelimbs are pretty short relative to the torso.
Step four: Tracing the colorized bones for the final reconstruction. (Fig. 4)
If I just attempted a lateral view I would have missed out on the very broad posterior torso based on the length of the posterior ribs. So I create both a dorsal view and a cross section view. Note that the sternal ribs, rarely found on most pterosaurs, extend laterally to meet the dorsal rib tips in Eudimorphodon. This give it a slightly wider body anteriorly, increasingly wider posteriorly. This is an odd autapomorphy, but it is based on many ribs, so it can’t be ignored. As you can see from the in situ image (Fig. 1) those long posterior ribs were there all the time. They were simply ignored by myself and others.
Eudimorphodon: a little odder than we thought
That torso is odd. Rather than tapering toward the pelvis, as in many other pterosaurs and tetrapods in general, the posterior torso is flat and wide, roofing the femora. My guess it provides a greater volume for eggs or respiration. With such small eggs, more eggs could have been carried by the mother. Note that the predecessor of E. ranzii, MPUM 6009, has a much deeper pelvic opening, likely to produce one large egg at a time. Note the reduction of the pelvis is also reflected in the reduction of the number of sacrals to four or five depending on the connection to the posterior pelvis.
If there is anything wrong with the results here, please let me know. If not feel free to use the technique yourself. I think it works pretty well.
I also don’t make these identifications without entering the taxa into a phylogenetic analysis that typically finds the same traits in sister taxa. Unfortunately posterior ribs are virtually unknown among Triassic and Early Jurassic sisters.
Pterosaur workers haven’t produced too many Eudimorphodon reconstructions, and certainly none that have recovered the oddly long posterior ribs. My earlier reconstructions were given generic ribs. So I did a bad thing. I went along with the paradigm of a tubular pterosaur body without testing that paradigm. While it takes a lot of work for small discoveries such as this, and the results are minor changes, well, I had nothing better to do on a quiet Sunday.
Wild R 1978. Die Flugsaurier (Reptilia, Pterosauria) aus der Oberen Trias von Cene bei Bergamo, Italien. Bolletino della Societa Paleontologica Italiana 17(2): 176–256.
Zambelli R 1973. Eudimorphodon ranzii gen.nov., sp.nov. Uno Pterosauro Triassico. Rendiconti Instituto Lombardo Accademia, (rend. sc.) 107: 27-32.