>With that in mind, you'd expect that fast-changing environments would select for genes that increase mutation rates I've seen the opposite in a bioreactor (ab

>With that in mind, you'd expect that fast-changing environments would select for genes that increase mutation rates I've seen the opposite in a bioreactor (about the most stable environment you can have). Variants slowly emerged and outcompeted each other until a variant emerged with a DNA repair pathway deleted. The spawns of that variant quickly out-competed the old ones and from then on there were ~10 dominant variants at any given time instead of the previous 1-3. Because the environment was so stable, the genes did not need to "remember" previous states of the environment and were free to optimize for the current state. The way to do this is to increase the mutation rate, as you potentially delete old genes that are no longer necessary and create new genes that are right for the moment. Edit: I suppose this points out a difference between "fast-changing" in a single direction and "fast-changing" that repeats states, such as seasonal changes.