Testing the adaptive plasticity hypothesis: Density-dependent selection on manipulated stem length in Impatiens capensis

In plants, stem elongation at high density in response to vegetation shade is hypothesized to be an example of adaptive phenotypic plasticity. Elongated stems may increase the light capture for plants in dense stands, while nonelongated stems may be favored for plants in low density. We tested the adaptive value of plastic stem elongation in Impatiens capensis by manipulating the controlling light cue, red to far red ratio, to produce elongated and nonelongated plants. These plants were then transplanted into high and low densities in a natural population. The results supported the adaptive plasticity hypothesis; elongated plants were more fit at high density, and suppressed plants were more fit at low density. Phenotypic selection analysis revealed selection for increased height in high density and for decreased height relative to leaf length in low density. Elongated plants showed less growth of the second internode at 2 wk after transplantation in both densities, which suggests a cost of elongation. Direct selection on height explained the fitness differences between suppressed and elongated plants at high density, but it did not completely explain the lower fitness of elongated plants at low density, which implies an intrinsic cost of elongation independent of selection on morphology.