Carroll does however convincingly demonstrate using models and real studies that evolution can produce new traits in a population in a relatively short time. He cites the change in the peppered moth population from light coloured to dark coloured in response to environmental pollution darkening the trees. This has probably happened on at least 4 separate occasions, as there are 4 known different mutations that cause the dark phenotype. He also gives a more recent example of pigeon populations developing a white rump. This gives them an advantage when it comes to avoiding attack by falcons; perhaps by momentarily confusing the falcon as the pigeon rolls to avoid the attack.
The rate of change in the population is proportional to the selection coefficient, and his example of mouse populations becoming all black in less than 2000 years is discussed here. He makes the point that not all individuals survive and that this is determined by the advantage/disadvantage a particular gene confers on its owner. There is also some discussion about whether natural selection acts on small or large differences between organisms. He comes down on the side of small changes. However, I would argue that it acts on both, as small changes in genes can actually cause a big difference in a characteristic. An example would be the genes controlling the sizes of tomato plants.
He introduces the idea of plasticity in species at the start of the chapter (and its role in Darwin forming his ideas of evolution) and ends with listing some of the possible types of mutation that occur – insertions, deletions, inversions, duplications, cut and paste, single nucleotide substitutions etc and will discuss some of these later in relation to evolving new functions. He also attacks the notion that all mutations are harmful. Some are actually beneficial (see above and here) and some a neutral and have no effect on function. This latter type of mutation can be used to study the effect of selective pressures on the rates of gene changes – by providing a reference point to the effect of random cumulative mutation.