© 1983 by The Society for Integrative and Comparative Biology
Phenotypic Plasticity in Life-History Traits: Demographic Effects and Evolutionary Consequences1
Biology Department, Woods Hole Oceanographic Institution Woods Hole, Massachusetts 02543
Although much life-history theory assumes otherwise, most life-history traits exhibit phenotypic plasticity in response to environmental factors during development. Plasticity has long been recognized as a potentially important factor in evolution, is known to be under genetic control, and may or may not be adaptive. The notion of adaptive plasticity contrasts with the idea that developmental homeostasis is a major evolutionary goal. The conflict was resolved in principle by Ashby's cybernetic analysis of homeostasis, which showed how plasticity in "response variables" might act to screen "essential variables" from the impact of environmental disturbance. To apply this analysis to life-history plasticity, it must be incorporated into a demographic model. An approach is presented here using life cycle graphs and matrix projection models. Plasticity in response to temporal variation leads to time-varying matrix models: plasticity in response to spatial variation leads to models structured by criteria other than age. The adaptive value of such plasticity can be assessed by calculating its effects on a suitable measure of fitness: long-term growth rate for time-invariant models, expected growth rate discounted by variance for time-varying models. Three examples are analyzed here: plasticity in the rate of development from one instar to the next in a stage-classified model, plasticity in multiplicative yield components, and plasticity in dormancy as a response to environmental cues. Development rate plasticity is adaptive if reproductive value increases from the instar in question to the next, maladaptive otherwise. Plasticity in yield components reduces fitness variance, and hence is adaptive, if the responses of successive developmental steps (e.g., flowers/stem, seeds/flower) are negatively correlated. Plasticity in dormancy is adaptive if it responds to the same factor(s) influencing mortality, but with opposite sign. A number of important problems, including trade-offs between genetic and phenotypic adaptation and the distinction between continuous and discontinous plasticity remain to be solved.