Our evolutionary developmental biology (EvoDevo) aim is to construct developmental-based models that are used to predict patterns of phenotypic variation. Our ultimate goal is to discover the 'logic' that governs the production of the phenotypic variation available for natural selection. In other words, we want to know whether we should worry about development when explaining evolution of phenotypes. Most of our work uses the mammalian dentition as a model system in the context of both micro- and macroevolution, and methods ranging from developmental biology experiments to computer models simulating development.

Kavanagh et al. (2007)

Above: Using a genetically-modified mouse we can observe the initiation and development of mouse molars in a petri dish. In the normal experiment, the second molar often does not develop. However, if the tail of the developing first molar is cut away from the developing tooth germ, it grows faster and larger than in the normal experiment.
Kavanagh et al. (2007) Nature