Fig. 1. The adaptive surface model of the contribution of phenotypic plasticity to successful colonization of a new environment. In this example the bold line is the adaptive surface (i.e. mean fitness curve). The solid thin line is the phenotypic distribution. For illustration, an environmental change is assumed to affect phenotypic plasticity without affecting the mean fitness curve, and to cause a shift in the mean of the phenotypic distribution without changing the variance. Broken lines indicate three levels of plasticity, from small (1) to moderate (2) to large (3). In (1), plasticity is insufficient to bring the population into the realm of attraction of the higher peak. This then leads to genetic evolution back towards the first peak, a process termed genetic compensation (Grether, 2005). In (2), plasticity brings the population into the realm of attraction of the higher peak, and there is then directional selection resulting in genetic change. In (3), plasticity is large, and the population is immediately centered under the higher peak; in this case there is no genetic change. A moderate amount of plasticity (2) is thus optimal for genetic assimilation.