Evolution of a native blueberry (Vaccinium elliottii) in a spatially and temporally heterogeneous landscape

 

 

Species that inhabit different positions along environmental gradients can be subject to divergent selection. Divergent selection, in turn, can lead to the evolution of three possible strategies: 1) specialization, where fixed ecotypes are adapted to a narrow range of abiotic and/or biotic conditions; 2) generalization, where an intermediate fixed phenotype persists in all environments; and 3) adaptive phenotypic plasticity, where phenotypes shift in response to changing environmental conditions. This study addresses how the scale of environmental variation affects patterns of population differentiation in a dynamic landscape with interhabitat spatial variation (upland vs. floodplain forests) and intrahabitat temporal variation (flood pulses in floodplain forests). The objectives are to determine whether upland and floodplain populations of Vaccinium elliottii (Ericaceae) are locally adapted specialists, and whether plasticity is more likely to evolve in temporally dynamic floodplain forests than in the more static upland forests. Contrary to predictions, plants from floodplain populations do not exhibit higher levels of phenotypic plasticity than those from upland populations. Rather, data from greenhouse and reciprocal transplant experiments suggest that individuals from different populations share similar plastic responses in traits likely to confer flooding and drought tolerance. To date, I can detect little evidence for local adaptation to the stressful floodplain habitat. In the field, plants had significantly greater mortality in floodplain than upland sites over two growing seasons, and upland genotypes outperformed floodplain genotypes in both habitats. Observational studies indicate that upland populations are more robust, and have significantly greater seed set. I hypothesize that gene flow may be primarily unidirectional from upland to stressful floodplain populations, and that source-sink dynamics may restrict adaptive evolution in the floodplains.