Figure 2.

Distribution of the size α₁ of the first adaptive substitution (left) and for the entire adaptive walk (right) in the direction of the moving optimum, as a function of phenotypic complexity n for different rates of environmental change v₁. Symbols in the left-hand panel show the predicted mean of the first adaptive substitution when assuming a uniform distribution of mutational effects (environmentally limited regime, eq. 19). This approximation produces a good match as long as the predicted does not exceed the (mean) standard deviation of the effects of new mutations (). Beyond this mark, the realized step size is reduced due to limited availability of large-effect mutations. Compared to the first step, the increase of with n is less pronounced when considering the entire adaptive walk. The reason is that subsequent substitutions will often compensate for pleiotropic side effects of previous steps rather than follow the moving optimum. Boxplots are based on 10, 000 replicated adaptive-walk simulations. The box contains 50% of the data. Horizontal white bars indicate the mean step size . Whiskers extend to maximally 1.5 times the size of the box. Outliers are not shown. Parameters: σ² = 10, ρΣ = 0, Θ = 1, m² = 1, ρM = 0; the scaled rate of environmental parameter γ = 10 × v₁.