Meeting Abstract
Modularity enables complex morphologies to evolve, by coordinating variation of functionally coupled traits within modules, which, being quasi-autonomous, can vary independently of other modules under stabilizing selection. Consequently, modular complexes can be optimized without interfering with adaptations of other modules. The rodent mandible has long served as a classic example of modularity, but the hypothesis of modularity for the mandible is of two “functional modules”: the tooth-bearing front and the muscle-bearing back. That is difficult to reconcile with the idea that the mammalian mandible can be modeled as a beam because a beam is not divisible into two functional complexes along its length. It is even more difficult to reconcile with the rodent mandible because the incisor runs through it, well past the boundary between the modules, and the masseteric muscle overlaps the border in the other direction. The question is: why would the front/back model fit so well, if it does? We find, using the Covariance Ratio, that it does fit well in eight species of squirrels from both tree- and ground-squirrel lineages. However, if we extend the beam of the horizontal ramus posteriorly through the back (to the bases of the posterior processes), and analyze correlations between the shapes of those regions, we find a high correlation between that posterior segment and the anterior (diastema) and/or middle (molar) in most of the species we analyzed. The exceptions are species that have high correlations between those anterior regions and at least one posterior process. On average, regions of the front may be more highly integrated than they are with the back but in no case are front and back even quasi-autonomous.
