“A cave was recently found where they discovered two new species,” Meredith said. “One of them is pure white and blind! Once a fish has lost its eyesight through evolution, do you think it would be able to re-evolve it if the species were exposed to light?”
My first thought was that that a blind fish in lighted environment would be like a fish out of water. With every other species having such a supreme advantage as the gift of site, our poor cave fish would soon be extinct. However, let’s say that light were somehow introduced to this sealed off cave. All the cave species – which are ostensibly blind – would be on equal footing, but now light is part of the equation. Would their dead eyes hold any chance of regaining vitality?
“This is a job for genetic algorithms,” I shout, fist pointed to the air. We were then asked to leave the theater, which was no skin off of our noses as it was already like four hours into the Da Vinci Code and it showed no signs of stopping.
I had to go to work, so rather than constructing a GA, I did some thought experiments.
How easy would it be for a lost, de–evolded trait to re-evolve? It is obviously heavily dependent on the fitness function, but my thought is that once something is gone it is nigh impossible to get it back.
Think of mutation as always happening and being largely destructive. Imagine a gene as a painting and mutation as putting a random color dot at a random spot on the canvas. Now, on rare, rare instances that color dot will actually improve the painting, but by-in-large that dot is going to look totally out of place and actually be detrimental to the painting.
It is because of this generally destructive property of mutation that evolution has the “use it or lose it” (UIOLI) property. For example, put some fish in a cave with no light. If some fish offspring have eye mutations which are a detriment to their eyesight, it won’t affect these baby fish’s ability to survive in the least. It’s a completely dark cave; what do they need with good eyesight? These bad eye genes will propagate and eventually mutation will again strike and make the eye genes even worse. Again, what do those fish care? They can’t see anyway because, as I may have mentioned, the cave is ttly dark. This process continues from fish generation to fish generation until – heavens to Betsy! – the fish are blind!
Now, imagine trying to reverse this process. Let’s go back to the painting metaphor. Let’s say that the painting was beautiful except a mutation caused on very unattractive yellow dot to appear right in the middle. The only way you have to fix the painting is with a button that says “mutate.” If you push that button, it is your hope that the mutation will happen right on top of that yellow dot and it will change it to just the right color to make the painting as it was. Do you think that will happen? What seems more likely to happen is that the new mutation dot will land on some other part of the painting making the panting even more ugly. If you keep hitting the “mutate” button, determined to change that original yellow dot, by the time luck has it that you correct that mutation, he rest of the painting is now covered with randomly colored dots.
I’m thinking it is a similar story with our cave fish’s eyes. Let’s say that there is just one mutation (which is a lot more likely to correct than two, three or more mutations) causing their blindness. Each of the fish have this mutation as all of them are blind. By the time random mutation has corrected the original mutation, it has decimated the rest of the eyesight gene.