Skull and dentition
Bears do not have a skull or dentition that is extreme in any particular regard. They do not have the short snout, long canines, and sharp cheek teeth (i.e., carnassials) that typify the highly specialized cats, nor do they have the high-crowned teeth of herbivores that are adapted to processing browse and forage. Nonetheless, the skulls and teeth of bears have distinctive features.
A number of researchers have investigated the ways in which skulls and teeth vary among carnivores and speculated on the implications and reasons why. Leonard Radinsky was perhaps the first to take on this task in a systematic, quantitative way. The most recent--in some ways definitive--analyses have been done by Borja Figueirido. The adapted results of one of his studies are shown in the two figures at left, expressed in terms of what Figueirido and others call "morphospace." The two graphics show a series of stylized skulls and jaws that progressively vary in shape along two key dimensions, each. The features that define each dimension are described by synoptic phrases at each corner, meaning that skulls or jaws progressively trend towards more extreme expressions of these features the closer they are to one end of the gradient or the other. The skull or jaw profiles match these trends.
So, what differentiates bears? The morphospace within which you can find bear skulls and jaws is bounded by a brown line containing a brown-shaded filling; a space denoted by the bears' family name: Ursidae. Bear skulls occupy the corner typified by a combination of being wide, domed towards the back, with shorter canines, and with dentition otherwise better adapted to grinding. Bear jaws occupy a corner typified by a forward-facing jaw hinge, a shallow underbite, small canines, and generalized dentition.
These features signify a group of species--the bears--with skulls and jaws that are adapted to a generalized diet, meaning that these features facilitate not only modest efficiency at grinding vegetation, but also modest efficiency at capturing and processing mammalian prey. This is in contrast, for example, to the more highly specialized skulls and jaws of other carnivores, including the elongate long-snouted skulls of dogs (Canids) or the powerful skulls and jaws, large canines, and slicing dentition of cats (Felids). Although not exclusively the case, both Felids and Canids tend to be more carnivorous than Ursids.
The figure immediately above provides a bit more information on the distribution and magnitude of the bite force generated by the teeth of various carnivores; work that was done primarily by Per Christiansen. A distinction is made between the bite force delivered by the canines versus the bite force delivered by the cheek teeth (i.e., molars/carnassials). Bite force is represented both in absolute terms (Newtons) as well as in terms standardized to the size of the associated animal. Each dot corresponds to a species, with bears denoted by a brown dot, and each bear species identified by a 3-letter acronym. The line represents a condition where the force delivered by the cheek teeth is equal to the force delivered by the canines.
Viewed in absolute terms (Panel B), bite force becomes ever more loaded on the molars or carnassials the greater the force that is delivered. Hence there is a trend downward in the ratio of canine force to molar force as force increases. Also of note, bears along with some of the largest cats deliver the greatest force of all the carnivores. Much of this is a function of size, which means that size matter when it comes to absolute bite force.
When standardized to size (Panel A), bears look less extreme. They actually tend to cluster lower down the curve, which means that, given their size, they are less well built to deliver bite force compared to many other carnivore species, especially of the Felids and Canids. The noteworthy exceptions are the polar bear (Uma) and the giant panda (Ame), which is not surprising. The polar bear is the most carnivorous of all bears (meaning, it needs to grasp and hold prey), whereas the panda is the most herbivorous (meaning it needs a powerful bite to grind the bamboo that comprises most of its diet).
Focusing in on molars, the above graph shows the surface area of the molar teeth relative to body size for various herbivore species, plus the grizzly bear. The point here is that grizzlies have less molar surface area per unit body mass than do most of the herbivorous species treated in this graph (the exception being cattle relative to horses). The implication is: although grizzlies do graze, and sometimes ingest large volumes of foliage, their primary grinding teeth (the molars) are not particularly well-adapted to the task. Again: generalized teeth for a generalized diet.
Another take on the configuration of carnivore skulls is offered by Dorren Covey and Walter Greaves. They measured dimensions related to the capacity of the skull to resist torsion, which is a measure of robustness. The results of their work is shown immediately above. Again, bear species are denoted by brown dots.
The noteworthy result with regard to bears is: they are not particularly noteworthy (consistent with the results of other analyses, including Figueirido's). On the other hand, Canids (dogs) are consistently deviant in that their long-snouted skulls are perhaps the least resistant to torsion. Hence, unlike the cats, dogs tend to rely less on shear strength of bite to grasp and subdue prey.