Ecosystem engineers that modify landforms can be valuable tools for restoring habitat, but their use has frequently resulted in unanticipated outcomes. Departures from expectations might arise because applications discount the possibility that geomorphic processes are influenced by heritable phenotypic variation. We conducted a field-scale common garden experiment to assess whether shoreline erosion reflects intraspecific variation in the landform engineer Spartina alterniflora. Replicated plots on a shoreline denuded by the Deepwater Horizon oil spill were revegetated using plants from four genetically distinct sources: the local population, a nonlocal population, and two nursery stocks. We assessed variation in biomass, tissue nutrients, and functional traits alongside soil shear strength, surface elevation, and shoreline erosion rates over 2 years. We found that productivity, traits, nutrient content, and erosion rates varied according to plant provenance. Erosion reflected traits like root architecture more so than coarser metrics of growth. Erosion was significantly higher in plots with nonlocal plants that exhibited lower productivity, likely due to nitrogen limitation. Our results indicate that restoration practices should account for intraspecific variation in landform engineers and that in situ trials should be performed at sites slated for restoration to evaluate donor source suitability, particularly if introductions might modify local populations.
Keywords: Deepwater Horizon oil spill; cultivars; ecosystem genetics; erosion; extended phenotype; functional traits; marsh loss.
© 2020 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd.