I don't know enough to not ask this question. How do they know that freezing and then rethawing didn't have some effect on the outcome? Would you not have to redo the experiment the same way again to make sure that you got the same results again?
The experiment took 20 years to perform and showed separate evolutionary paths taken by the 12 initial populations. The 12 colonies were from two populations that only differed by a neutral marker - one that would not have an effect on the final results. This experiment was looking for a population to evolve the ability use citrate as a carbon source. Only the final colony and its descendants of its 20,000 generation ancestor, were capable of metabolizing citrate. That would indicate that this type of experiment may not be repeatable because it depends on the microorganism's ability to undergo mutations that culminate in the ability to use citrate. It is a common practice to freeze bacteria in glycerol. Stocks of bacteria can be stored at -80C for many years without causing mutations from the freezing process. In this case, I would bet that this population was sequenced many times along the way and also subjected to microarray or macroarray to see how the bacteria was differentially regulating its genes. I won't be able to download and read this paper until tomorrow so I can't be 100% sure about the methods used in this experiment but the initial populations are described in a paper published by the same lab in 2002.
