That the mutations that occur in the sample are not outside of the species. There is no species jump in such mutations, any more than there is between a blonde man and a burnette man.
What do you mean by bacterial species? What specific limitations are you claiming?