Sampling strategy and evolutionary transitions

Earlier this year, we published an article on eye size evolution of early tetrapods in PNAS. Dan-Eric Nilsson wrote a very nice commentary on the paper in Current Biology, and we followed up with some clarification concerning our sampling scheme. To further illustrate our point, we prepared an additional figure, which can’t be posted as part of our comment, and hence we are making it available here.

Our point is centered on the idea that the further away one samples from a given transition that represents a release from constraint, the higher the possibility that other factors dilute the signal of that release from constraint. For that reason, we were careful to sample around the water to land transition as narrowly as possible. With this narrow phylogenetic bracket, we try to make the ecological situation of these animals comparable. To understand this point, consider the following figure below.sampling_scheme.001

Figure: Potential influence of the sampling strategy on eye size comparisons.

We wanted to determine how the transition from vision through water to vision through air (a release of constraint) affects eye size evolution and visual ecology of early tetrapods. That particular transition is indicated by the red branch in the phylogeny in the above figure. To understand the influence of this release of constraint on eye size evolution, the phylogeny dictates the sampling of very closely related species from just prior and just after that transition. We highlight candidate taxa with orange and red circles. Ideally, one would sample all species that document this transition, but, especially in the fossil record, that is often impossible.

Note that in our hypothetical example, taxa with vision through air from just after the transition (red circles) have slightly larger eyes than their close relatives that still see through the limiting water medium (orange circles). Now let’s assume one would sample with a much wider phylogenetic bracket, across the phylogeny of all vertebrates. In this case, many other factors known to affect eye size evolution would diminish the signal of the transition from water to land. For example, more distantly related taxa may occupy deep sea habitats (purple), fly in open habitats (green), or don’t even rely on vision anymore (brown). The possible outcome of sampling too broadly is that the true signal is no longer discernible. Only by narrowly examining the transition one can determine the correct pattern of eye size evolution.

This entry was posted in vision and tagged , , , . Bookmark the permalink.