Jeffery E. Barrick and his colleagues have published the results of one of the most interesting evolution experiments I have ever read.1 Actually, the genius behind this experiment is Richard E. Lenski, who is on the author list as well. Lenski started an experiment with E. coli almost 20 years ago, and it is still producing excellent results. Essentially, the experiment followed twelve populations of E. coli over all those years. The focus of the paper was one of those twelve populations.
In the experiment, the bacteria were grown on a minimal medium with glucose as a limiting nutrient. Each day, a small sample of the culture was removed and placed in a fresh medium. Periodically, samples were frozen so that they could be analyzed in detail at any time.
Thanks to the wonderful technology we have today, the entire genome of E. coli can be sequenced in a “reasonably short” amount of time. So this paper reports on the results of comparing the genome of the original bacterium to that of the bacteria after 2,000, 5,000, 10,000, 15,000, 20,000, and 40,000 generations. The results were “rather surprising,” according to the authors.
The main thrust of the experimental results is given in Figure 2 of the paper. In that figure, we see that the number of mutations accumulated over about 25,000 generations was pretty linear. Then the mutation rate shot up considerably. To give you an idea of just how much the mutation rate changed, there were just under 45 mutations in the population at generation 25,000, but by generation 40,000, there were more than 600!
The fact that the mutation rate jumped was easily understood, and it is the result of a mutation in a gene called mutT. This mutation causes very specific kinds of subsequent mutations, which were 90% of the new mutations found in generation 40,000. Thus, this result was not surprising. It is well understood.
The “rather surprising” result is what happened to the fitness of the population while mutations piled up at a rather constant rate. The authors found that with the first few mutations, fitness increased rapidly. After that, however, mutations accumulated at a more or less constant rate (at least for the first 25,000 generations), but the fitness did not. Instead fitness increased more slowly as time went on, even though the mutation rate stayed constant.
The authors tell us the expected result: mutations and fitness should increase or decrease hand-in-hand. If lots of mutations are being preserved in the population, that should mean they are beneficial to the organism in that environment. Thus, the organism’s fitness should increase rapidly. If few mutations are being preserved in the population, that should mean there are few beneficial mutations for the organism in that environment, so the fitness should increase slowly. Thus, the rate of preserved mutations should go along with rate of change in fitness. This is not what the data showed. Instead, the preserved mutations increased linearly, but the fitness increased rapidly at first and then leveled off as time went on.
After proposing a couple of hypotheses and showing how they could not be the right explanation for the “rather surprising” result, the authors finally said that the leveling off of fitness increase was due to the fact that once the initial large increases in fitness were gained by certain mutations, the only possibilities left were mutations that led to small increases in fitness. As a result, there would be a lot of competition between subpopulations that had SOME of those mutations and other subpopulations that had OTHER versions of those mutations. Thus, competition between subpopulations that each had SOME small-but-beneficial mutations would keep mutations rates up but would limit overall gains in fitness.
In addition, some mutations produce both positive and negative effects for the bacteria. When a mutation produces a large positive effect and a few small negative effects, subsequent mutations can work on getting rid of those small negative effects while preserving the large positive effect. Once again, however, that results in only minimal fitness gains.
In other words, the authors conclude that the changes you can expect in a genome are limited. If you put bacteria in a stressful environment, they will begin to mutate, because their genome seems designed to do that. However, since there is a limit to how much the genome can change, the fitness gains that come as a result of mutation cannot continue to increase at the same rate as preserved mutations.
In other words, mutation and natural selection can “tinker” with the genome, making minor improvements. However, they can’t do much more than that. As a result, mutations will continue to accumulate if the bacteria are stressed, but because there are limits to the flexibility of the genome, those mutations will result in smaller and smaller fitness gains.
This, of course, is what creationists have been saying for many years. God created specific kinds of creatures, and He built in them machinery that would allow them to change over time so as to adapt to changes in their surroundings. However, each genome is limited to the specific kind of organism created, and at some point, you reach a limit at which the genome cannot be changed anymore, at least not in a way that will promote further adaptation.
This, of course, is diametrically opposed to what is needed for the evolution of one kind of organism into another kind of organism. If all it takes is mutation, natural selection, and time to turn a fish into an amphibian, the genome must be almost infinitely flexible, and there should be virtually no limit to the changes that can be produced by mutations. That’s not what these data (or other data2) indicate.
It is nice to see that 40,000 generations of E. coli data support the creationist view of the genome.
1. Jeffery E. Barrick, et al., “Genome evolution and adaptation in a long-term experiment with Escherichia coli,” Nature, 461:1243-1247, 2009
Return to Text
2. A. A. Hoffmann, et al., “Low Potential for Climatic Stress Adaptation in a Rainforest Drosophila Species,” Science, 301:5629-34, 2003
Return to Text