Figure 2. Graphical hypotheses for the relationship between environmental properties and benefits of plasticity for adaptive evolution. Panel A [i-iii] describes the benefit of plasticity for adaptive evolution across an increasing rate of environmental change, Panel B [i-iii] describes this dynamic across increasing environmental variance, and Panel C [i-ii] across increasing temporal autocorrelation.
Hypothesis A[i]: The benefit of plasticity increases with increase in the rate of environmental change, eventually plateauing. Selection is weak when the mean environmental change is slow, and phenotypic lag is small. Population growth is consequently high, and heritability of fitness-related traits is high. In this scenario, plasticity adds little to adaptive tracking, and thus the costs of plasticity outweigh the benefits in decreasing the phenotypic lag. Conversely, when the mean environment changes too fast for adaptive evolution to track, and phenotypic lag is high, plasticity helps to ‘catch up’ with the moving optimum by allowing for the population to increase in size, and thus maintain the genetic diversity.
Hypothesis A[ii]: The benefit of plasticity decreases with increases in the rate of environmental change. Contrary to Hypothesis A[i], when selection is weak, lag load can increase because the population evolves more slowly. In this scenario, plasticity can bring the population phenotypic mean close to the selection peak at a low rate of environmental change. Conversely, as rate of mean environmental change increases, the limits of plasticity set by its costs (physiological toll and erosion of genetic diversity) may result in a limited role of plasticity for adaptive tracking. The population size may be small at high rates of environmental change, and plasticity may increase the chance of extinction due to drift by shifting the phenotypic average and thus shading the genetic variation from selection. Moreover, a high rate of environmental change can limit the efficacy of plasticity given the low predictability of the future environment.
Hypothesis A[iii]: The benefit of plasticity is maximised at an intermediate rate of environmental change, above (following A[i]) and below (following A[ii]) which its benefit decreases. Following moving optimum theory, there is an intermediate rate of environmental change at which the balance between selection strength and population persistence is optimal.
Hypothesis B[i]: The benefit of plasticity to adaptive evolution increases with increasing environmental variation. As the environment becomes more variable, plastic responses in physiology, life history, phenology, and/or behaviour can dampen the detrimental effects of unpredictable fluctuations, thereby preventing extinction. This buffering would afford the population more time to reach its adaptive peak via adaptive evolution. This benefit would eventually cross a point of diminishing returns, as when the environment becomes too variable, the costs involved in plastic responses may outweigh their benefits in part due to the lack of predictability in the temporal environment. Moreover, at a highly variable environment with a stationary mean, evolution may serve to be nonadaptive, and plasticity can allow the genotypic mean to remain near the environmental mean amidst the environmental variability.
Hypothesis B[ii]: The benefit of plasticity to adaptive evolution decreases with increasing environmental variation. In an environment with a small amount of variation, plasticity works together with evolution to fix traits helpful in the new environment. As the environment becomes more variable, phenotypic plastic responses may drive a disconnect between phenotypic selection and genotypic selection, ultimately making the genetic variation in the population maladapted to future environmental conditions. In other words, plasticity might help a population that is stuck in a valley or a local peak to find a global peak on a fitness landscape when the environment is moderately variable; if the environment is too variable, peak searching might be disrupted too much—and peaks themselves would be shifting on the landscape.
Hypothesis B[iii]: The benefit of plasticity to evolution is highest at low and high amounts of environmental variability. The ability of the trait mean in the population to reach the peak of fitness landscapes via adaptive evolution may be optimal at an intermediate level of environmental variance. In this case, the facilitative role of plasticity would be low at an intermediate level of environmental variance if it masks genetic variance of the population from selection, or shifts the phenotypic average.
Hypothesis C[i]: The benefit of plasticity to adaptive evolution increases with increasing temporal autocorrelation. Higher autocorrelation in the environment corresponds to higher reliability of temporal cues and thus higher predictability of future environmental states. Therefore, plastic responses can more accurately track moving selection targets, and aid adaptive tracking. In addition, adaptive evolution may be less likely to occur in isolation in highly autocorrelated environments.
Hypothesis C[ii]: The benefit of plasticity to adaptive evolution decreases with increasing temporal autocorrelation. Autocorrelation at temporal lags that are not in resonance (similar lengths of time) with paces of life history can increase extinction risk. This could lead to a population existing in an unfavourable environment for long periods (reducing temporal refugia), and plasticity could decrease the genetic variation upon which selection can act. Thus, the ability for plasticity to help adaptively track moving optima decreases.