Species that exhibit adaptive phenotypic plasticity alter their phenotypes in response to environmental conditions, thereby maximizing fitness across spatially and temporally heterogeneous landscapes. However, when individuals of a species show significantly greater fitness in one habitat than another, selection is thought to favor traits that enhance fitness in the high quality or source habitat at the expense of fitness in the ecologically marginal habitat.   Phenotypic plasticity is not expected to evolve under these conditions.   Elliott’s blueberry (Vaccinium elliottii) occurs in dry upland and flood-prone bottomland forests through the Southeastern United States.  These contrasting habitats differ in a variety of environmental characteristics which impose divergent natural selection and favor alternate phenotypic optima.  Nevertheless, I found limited evidence for local adaptation to bottomland and upland forests in a previous study.  Instead, V. elliottii exhibited patterns consistent with source-sink dynamics: higher fitness in upland relative to bottomland forests and asymmetrical gene flow from upland into bottomland populations.  The objective of this study was to assess whether families of V. elliottii seedlings and cuttings displayed a phenotypically plastic response to a complex environmental gradient in a reciprocal transplant experiment and long-term drought vs. flooding in a greenhouse experiment. 

Results/Conclusions

In contrast to predictions from source-sink models, we found a high degree of phenotypic plasticity in foliar traits (specific leaf area, leaf size), rooting architecture (depth of roots) and root:shoot ratios.  The phenotypic plasticity was concordant with environmental conditions, e.g. thicker leaves (low specific leaf area) were expressed in upland relative to bottomland transplant sites.  Nevertheless, V. elliottii did not express several traits known to improve flood tolerance like lenticels, adventitious roots or high root porosity.  Furthermore, phenotypic plasticity in foliar traits was greater in the field experiment than the greenhouse experiment, which suggests that differences in water stress were not the primary drivers of the foliar plasticity that was observed in the field.  I hypothesize that foliar plasticity is driven by light level differences and potentially edaphic conditions.  This evidence for phenotypic plasticity might not contradict predictions of source-sink models if the plasticity itself is favored within upland systems (the source).  Indeed, selection analyses revealed that plasticity in foliar and root traits could be advantageous under drought conditions.  Thus, in systems driven by source-sink dynamics, landscape-level plasticity can be maintained by selection in the source habitat.