Abstract
The developmental gene expression, morphogenesis, and population variation in mammalian molar teeth has become increasingly well understood, providing a model system for synthesizing evolution and developmental genetics. In this study, we estimated additive genetic covariances in molar shape (G) using parent-offspring regression in Cryptotis parva, the Least Shrew. We found that crown shape had an overall h2 value of 0.34 (±0.08), with higher heritabilities in molar cusps than notches. We compared the genetic covariances to phenotypic (P) and environmental (E) covariances, and to the covariances in crown features expected from the enamel knot developmental cascade (D). We found that G and D were not strongly correlated and that major axes of G (evolutionary lines of least resistance) are better predictors of evolutionary divergences in soricines than is D. We conclude that the enamel knot cascade does impose constraints on the evolution of molar shape, but that it is so permissive that the divergences among soricines (whose last common ancestor lived about 14 million years ago) do not fully explore its confines. Over tens of millions of years, G will be a better predictor of the major axes of evolution in molar shape than D.
Acknowledgements
Special thanks to Gregg Gunnell, Jerry Hooker, and Gareth Dyke for their efforts in organizing this volume, as well as to Percy Butler for insights he provided over the years about tooth development, function, and evolution. I thank S. Gilmour, J. Head, J. Jernvall, B. Khoruzhenko, N. MacLeod, R.A. Nichols, I. Salazar-Ciudad, and two anonymous reviewers who provided useful ideas and criticism that improved this study, and J. Eger (Royal Ontario Museum), T. Gorog (University of Michigan), J. Hausser (University of Lausanne), H. Price-Thomas (Queen Mary, University of London), and N. Slade (University of Kansas), who provided access specimens in their care.