Stages of limb development for A. sagrei

As lineages rapidly diversify, such as in the history of anoles, does their developmental-genetic architecture constrain the rate or direction of evolutionary change? In other words, could the processes controlling the production of variation, the variation that natural selection acts on, affect patterns of phenotypic evolution by generating some phenotypes more readily than others? While theoretical discussions like these have been prevalent for over a century, developmentally-based constraints were not formalized in the context of modern biology until the 1980’s, fueled by an influential paper by Maynard-Smith and colleagues and the re-synthesis of evolutionary and developmental biology. Since then evo-devologists have been testing the plausibility of developmental constraints by examining the developmental bases of traits that have independently evolved multiple times; phenotypes that have repeatedly evolved using the same mechanisms may be indicative of constraint (because the precise interpretation of these patterns and appropriate level of analysis are contentious I will leave further theoretical discussion of constraint to future conversations).

In a recent paper, for which I am the lead author, we set out to examine whether developmental constraints could have affected diversification of anole limb morphology. As most readers of this blog are likely aware, relative limb length is correlated with the ecology and habitat preferences of many Caribbean anoles and similar limb morphologies have evolved independently on each island in the Greater Antilles. We compared growth series of eight anole species from the earliest stages of limb development to adulthood. Based on our analyses it appears that the same developmental mechanism may have been repeatedly recruited to generate the diversity of limb morphologies we observe today. On each island we can trace variation in adult morphology all the way back to the point in development when the cartilaginous long bone anlagen are just forming.

Is it possible that anoles inherited common patterns of variation from their most recent ancestor leading to the co-option of the same mechanisms time and time again? From our findings, it is certainly tempting to jump on the developmental constraint bandwagon, but we first need to consider if other non-developmental factors could generate the same pattern. For example, selection on juvenile morphology could generate strong correlations between different life stages. Unfortunately we still don’t know enough about juvenile ecology to know the strength of this hypothesis. In addition, we do not yet know whether the same molecular mechanisms underlie these phenotypic patterns. Finding whether the same genes (or pathways) underlie ecomorphological patterns of variation will more conclusively test the constraint hypothesis. Further examination of early limb bud development in conjunction with comparative genomics among anole species will potentially shed light on the precise mechanisms of adaptive morphological evolution.

And what about the elephant in the room: does anole limb length evolution represent an example of convergence or parallelism? We cautiously state that we cannot rule out the possibility that this is an example of parallelism; Anolis limb length variation has potentially evolved through the interaction of selection and shared developmental constraints (but see paragraph above). Unfortunately, there is a quagmire of definitions regarding these terms floating around – no fewer than three definitions of parallel evolution are commonly used in the literature – and distinction between these concepts has become blurred. In my opinion, much could be gained from moving our discussions past fitting examples into this dichotomous classification towards analyses of the multiple evolutionary mechanisms that can generate these patterns. Shared developmental, genomic, and selective constraints could all create similar patterns in Recent species and it is rapidly becoming time to explicitly test between these alternatives and others that may be applicable to particular systems. [stepping off of my soap box]

Hope you enjoy our paper!

Thomas Sanger
Latest posts by Thomas Sanger (see all)