A reason for glucose-cycling in Lenski's E. coli experiment
Each daily population of the 12 lines of bacteria was raised in 10 ml of medium, and one-percent -- i.e., 0.1 ml -- of each old population is transferred to start the next population. That ratio of the medium transfer allows a 100:1 growth in population to maintain the same final population density in the populations. Assuming a doubling of population with each generation, a 100:1 population growth would be reached in 6-7 (6.64, to be exact) generations. 6-7 generations occur in just a few hours (the populations are started in the morning and the glucose is exhausted by the afternoon). If more generations were allowed, the population density would become too high, which would limit capacity for population growth when the next population is started, so glucose supply is limited to limit the number of generations of glucose-eating-only bacteria. I presume that the 10 ml and 0.1 ml quantities are limited by the physical limitations of the lab equipment, e.g., the incubator size and the means for transferring the medium. If desired, the glucose starvation period could be eliminated by restarting the populations more frequently, but that would make the experiment even more labor intensive and I think it was desirable to have a glucose starvation period, anyway. The medium recipe says that the glucose supply could be varied, and the supply might have been varied to control population growth -- I don't know. The experimental procedure is described here.
Also, as I noted before, Blount did not give straight answers to my question about whether Cit+ evolution was a goal of the experiment. If, say, Cit+ evolution was a secondary goal or a longshot goal, he would be misleading people by going around flatly telling them that Cit+ evolution was not a goal of the experiment. In a sane world, Blount's failure to answer my simple, basic questions would be widely condemned -- but this is not a sane world.
Another somewhat minor concern I have is the potential loss of mutations during transfer, but that loss is inevitable. Mutations occurring in one of the last generations in a daily population might exist in only one, two, or just a few bacteria and so would have only a small chance of being in the one-percent of the medium that is transferred to start the next generation.
The glucose supply was unbelievably small -- only 25 mg per liter or just 0.25 mg for the 10 ml medium for each population!
Anyway, what is the likelihood that anything similar to the carefully controlled conditions in this experiment could occur in nature? Even with these carefully controlled conditions that favored Cit+ evolution, the Cit+ evolution was a rare occurrence, occurring in only one of 12 lines of bacteria in 20 years.
Anyway, unlike Andy Schlafly over at Conservapedia, I see no reason to suspect error or fraud in the claims of Cit+ evolution. Lenski et al. claim that they have the Cit+ E. coli bacteria -- which are very unusual -- and also claim that they have E. coli bacteria that have a strong tendency to evolve the Cit+ trait (bacteria of 20,000 generations or later). It would be easy to verify or disprove the claims about Cit+ evolution.
Labels: Citrate-eating E. coli