According to Darwin's principle of "descent with modification" closely related species are expected to share similar traits. Still, Darwin's critics pointed out that unrelated species often feature strikingly similar or even identical traits, i.e. phenotypic repeatability. To explain this trend, modern theory of evolutionary biology posits that natural selection favors phenotypic repeatability in unrelated species inhabiting similar environments.

Alternative perspectives from evolutionary developmental biology (evo-devo) propose that developmental constraints imposed by organismal design could promote phenotypic repeatability by limiting variation available to natural selection. Evo-devo questions other related issues in modern evolutionary biology, such as that selection operating gradually across evolutionary time is the sole process underlying the development of novel traits, i.e. phenotypic innovation. Examination of developmental processes could provide new insights on this and other conceptual issues concerning the rapid and repeated evolution of traits.

The overarching objective of this dissertation was to employ an evo-devo framework to uncover developmental underpinnings of phenotypic innovation and repeatability. This dissertation research aimed to address this fascinating topic in three data chapters. The focal study system of these projects is the enigmatic turtle body plan, which is a classic example of phenotypic innovation and developmental constraint.

In data chapter two, I seek to reevaluate classical descriptions of embryonic development in a "model" turtle species, the painted turtle (Chrysemys picta). This observational study served as the foundation to data chapter three, which entailed the most phylogenetically comprehensive comparison of turtle development. Data chapter three seeks to identify evolutionary divergence in processes governing development of phenotypic innovation in turtles. In data chapter four, I attempt to link phylogenetic patterns of phenoptypic repeatability with underlying genetic pathways. This culminating chapter examines the evolution of a repeatedly evolving trait, shell kinesis, to address one of the most intriguing questions in modern biology: Do unrelated species employ similar genetic solutions to similar ecological problems?

Altogether, these data chapters illuminate developmental underpinnings of striking patterns of trait evolution in turtles, and corroborate that both natural selection and developmental constraints influence the origins of phenotypic innovation and repeatability in nature.