In order to capture the phenotypic variation of the internal skull structures, such as
the sinuses or the brain, it is necessary to perform CT scans in a large number of
specimens, which is difficult and expensive. Therefore, while the external morphology of
the mammalian cranium has been the subject of many morphometric studies, the internal
structures of the cranium have been comparatively less studied. Here, we explore how the
variation of external shape reflects the morphology of internal structures. We use the
family Ursidae (Carnivora, Mammalia) as a case study because bears have a wide
variability of cranial morphologies in part associated with different trophic ecologies. To
do this, we digitized a set of landmarks in 3D with a Microscribe G2X from the external
surface of the cranium in a wide sample of bears. Additionally, the crania of seven bear
species were CT-scanned and prepared digitally to visualize the 3D models of the
external cranium morphology and of internal structures. Subsequently, we divided the
landmarks into two modules, splanchnocranium and neurocranium, and we perform a
two-block partial least squares analysis (2B PLS) to explore the intraspecific (static)
morphological changes associated with the covariation between them. These
morphological changes were visualized using the morphing technique with the 3D
models, looking at both the external shape and the internal structures. In addition, we
inferred the volume of the sinuses and of the brain in each hypothetical model. Our
results show that the first two PLS axes are associated externally with changes in the
basicranial angle, face length and cranium height and width. Concerning the internal
structures, there are parallel changes in dorso-ventral and medio-lateral expansion of
sinuses and brain, accompanied by their corresponding changes in volume. In contrast,
the third PLS axis is related to opposite changes in the volume of sinuses and brain.
These preliminary results suggest that the opposite relationship between sinuses and brain
volumes in the bear cranium is not as evident as expected, at least at intraspecific level.