Evolution of the eye

The evolution of the eye stood for years as a paradigmatic example of independent evolutionary paths fulfilling the same need. Vertebrates and mollusks have single lens eyes (though the photoreceptive cells under the lens have opposite orientation) while insects have compound eyes. These differences had been taken to imply that the eye evolved (independently) numerous times. Now it looks (pun!) as if the large morphological differences share a common developmental pathway for eye morphogenesis. The evidence for commonality in these developmental pathways comes from looking at similar proteins in mammals and flies (the Pax6 proteins). This particular protein, called eyeless for its mutant phenotype in fruit flies, was shown to give rise to eyes on wings and legs when expressed in those locations. So it must be near the top of the developmental hierarchy for eye development. A mutation in a similar protein in mammals (so similar that it is regarded as being derived from the same protein in the common ancestor of flies and mice) results in abnormal formations of the eye. The mouse protein, when expressed in unusual locations in the fly, also results in production of fly eyes! This suggests that these two proteins have very similar functions. It also suggests that either the common ancestor of flies and mice also had working eyes whose development used this same protein (or their common ancestor) or that whatever this protein was doing in the common ancestor it facilitated the evolution of eyes in other lineages (it is found in squid and octopus too). Does this mean that there is no selection - of course not! But it does sensitize us to the importance of the lineage. It suggests that if there is even the slightest backdrop that can be utilized in a new context it will be. If the Pax6 cascade had not been extant in the lineage that led to insects, chordates and mollusks would there have been eyes in these lineages?

As I stood waiting for the bus after a winter snow I imagined what a selectionist theory of snow melt might look like. We would have the hyperselectionists sagely reminding us that the snow that is left is what has, heroically, not melted yet. True enough. Reformulating this in terms of melt resistance (read fitness) might be a solution, but only if we have an external way to measure it – temperature, salt concentration in the snow, density, etc. Even then it may still be the case that to get all of this information we would have to destroy the snow bank – which would make it difficult to use the theory in a predictive way.

Evolutionists when working in the adaptationist mode talk as if teleological principles are at work. The presence of a particular trait is due to what it does. But this is just short hand for the full selectionist account. In the same way, when evolutionists talk about the archetype or bauplane that facilitates and inhibits particular changes this is just short hand for an account that tries to understand the dynamics of development. Developmental geneticists are starting to understand morphospace (i.e. which phenotypes are close to which other onese) given the developmental unfolding of a lineage. For example, is loss of appendages easier than gain? Will the losses be paired when and if they do come? Ironically this research agenda may be more tractable and productive than sifting through the uncountable measured and posited selection regimes. Understanding the contours of constraint may tell us more about extant forms than attempting to invent histories for each trait.