Mutation, and the ability to repair it, are incredibly important drivers of evolution on just about every level.
Last month's PNAS has two cool articles on bdelloid (pronounced DELL-oyd) rotifers and their ability to repair mutations, which demonstrate this extremely well. Bdelloids are the group of rotifers that reproduce only asexually, unlike the Monogonont rotifers I study, which alternate between sex and asex. Asex doesn't allow natural selection to remove deleterious mutations nearly as effectively as does sex (i.e. Muller's ratchet), so most species that go asexual quickly build up an enormous mutational load and die out after some hundreds or thousands of generations. But bdelloids have apparently been happily asexual for billions of generations. So how have they avoided Muller's ratchet? By not allowing it to turn in the first place, apparently. One doesn't need natural selection to remove mutations if one can repair them one's self.
Bdelloids apparently are degenerate tetraploids, meaning instead of two copies of each chromosome, at some point in their evolutionary past they had four, but those four then diverged somewhat into two pairs. Still, this means they have four copies, on separate chromosomes, of most of their genes. And it appears they can use these four copies as templates to repair each other. If one copy might have a mutation, check it against the other three, find the differences and correct them.
The utility of of this system in the short term (on the time scale that natural selection functions) is demonstrated by two other super-powers of bdelloids. First, they can dry out completely, at any life stage, and when rehydrated will repair all the damage to their chromosomes and resume life where they left it. Second, they can continue reproducing at radiation levels five times higher than what most anything else can stand, because every time the radiation damages their DNA, they just fix it. Bdelloids don't need to worry about cancer, apparently.
So with all these advantages, why haven't bdelloids taken over the world? Why doesn't everything do the bdelloid? Presumably because there are disadvantages in other contexts. Bdelloidism removes mutations so effectively, it seems unlikely very much macro-evolution could take place. After all, the repair mechanisms remove pretty much all mutations, and have no way of knowing if that particular mutation would have been advantageous. Once a bdelloid, always a bdelloid. Bdelloids might also be slow on the micro-evolution side of things. If selective pressures shift, having genetic variation is essential to any sort of adaptive response. Mutations are the ultimate source of genetic variation, so if they are all repaired out of existence, it may be hard to adapt. Finally, I would guess (not knowing the details of the repair mechanism) it is physiologically expensive to do all that checking and repairing all the time.
Now I find myself wondering about the demography of bdelloids. Hmmmm.