Laelaps

The problem with spinosaurid skulls is that they display a mosaic of characters that show similarities to both crocodylians and other theropod dinosaurs. The presence of a secondary palate, a "rostral terminal rosette" (or a rounded, cup-like region at the tip of the snout), a less blade-like and more rounded tooth structure, and reduced antorbital fenestrae all show some convergence with crododylians. Such adaptations are typical strength-increasing changes in the skull, the closing of the antorbital fenestrae and the development of a secondary palate both making the skulls of living crocodylians more resistant from torsional stresses during feeding. In order to test whether the same might be true in spinosaurid dinosaurs, the team created computer models of the snouts of Baryonyx, an American alligator, a gharial, and a "generalized" (or "stylized," as they sometimes call it) theropod, the crocodylians being chosen as they represented extremes for crocodylian skull morphology (alligators have a broad, flat snout and gharials have a very narrow, tubular snout). The researchers also state that they did not recreate entire-skull models due to time constraints and the fact that the back of the skull of Baryonyx was incomplete, so the results are based entirely upon snout morphology.

What the researchers found when they carried out their tests was that the alligator and generalized theropod were functionally closer to each other than either was to the gharial and Baryonyx in terms of stress resistance and distribution, the gharial and Baryonyx also being closer to each other than either was to the alligator or theropod. In the case of the generalized theropod and alligator, for instance, torsional stresses became greater as their tooth rows diverged from the midline (or moving backwards from the front of the jaw as the jaws make a V shape). The gharial and Baryonyx, by contrast, had minimized torsional stress by having a narrow jaw where stress was spread out more evenly. Likewise, the secondary palate provides alligators strength against torsional stress while in the gharial and Baryonyx the secondary palate provided resistance against bending, so the function of a secondary palate as a reinforcing feature of the skull can have different functions depending on skull shape. The study also did something else that was interesting; they gave the alligator and gharial skulls antorbital fenestrae at one point, and it seems that such large holes in a gharial skull definitely weakened it (the alligator skull didn't seem to experience much of a difference). This suggests that the reduction and closing up of antorbital fenestrae was likely a consequence of a narrow snout and the stress load distribution that comes with it.

The researchers do pose an interesting question, though; if it is advantageous to develop a secondary palate and to close up antorbital fenestrae in relatively narrow skulls, why didn't large theropods do so? There could be other aspects of theropod skulls that could deal with stresses involving bite forces, especially since the study focused on stressed in the rostral region and did not take the rest of the skull/musculature into account. Also, theropod skulls are compressed laterally rather than dorso-ventrally, so such an arrangement may have its own strengths and weaknesses that will merit further research. As far as dealing with stress loads in the rostral part of the skull, though, Baryonyx seems to be functionally convergent with gharials, although further study (especially of other crocodilians and spinosaurids) is required.

References;

Rayfield, E.J.; Milner, A.C.; Xuan, V.B.; Young, P.G. (2007) "Functional Morphology of Spinosaur 'Crocodile-Mimic' Dinosaurs." Journal of Vertebrate Paleontology, Vol. 27(4), pp. 892-901