Files
Abstract
Phenotypic trait variation is ubiquitous among plant populations. Multiple evolutionary and ecological processes may produce trait variation, including divergent selection due to heterogeneous environmental conditions driving local adaptation, phenotypic plasticity, historical genetic structure, genetic drift, and barriers to gene flow. Consequently, disentangling the principal cause driving trait variation is challenging, since multiple processes may result in similar outcomes. Understanding the underlying source of phenotypic trait variation is particularly critical for informing restoration and conservation decisions. Uniola paniculata is a perennial grass that occurs on southeastern U.S. coastal dunes and is federally protected due to its role in stabilizing dune habitats. Across a single dune system, U. paniculatas range spans a localized environmental gradient from the dynamic dunes closest to the shoreline to the more stabilized dunes farther inland. We characterize the shoreline-to-landward environmental gradient and determine that variation in morphological and physiological traits in U. paniculata mirrors variation in the underlying abiotic conditions in a mensurative field study. We find no evidence that populations are locally adapted to microhabitats along the environmental gradient using reciprocal transplants of individuals from each habitat. We employ a comparative approach using a greenhouse common garden to examine variation in quantitative traits and genetic analyses (allozymes) to examine neutral genetic variation. This study includes populations from four Georgia barrier islands. We find evidence of divergent selection on both aboveground and total biomass which appears to be driven primarily by inter-island differences rather than intra-island differences from habitats along the shoreline-to-landward environmental gradient. Lastly, we explore the genetic structure of a widely dispersed beach annual, Cakile edentula, which is comprised of three subspecies associated with particular geographic distributions. This study has three key findings to contribute to our understanding of C. edentula: genetic diversity (Hep) is fairly low; taxonomic subspecies designations are supported by UPGMA-dendrogram clusters and AMOVA analysis indicating that variation among regions explains 42% of total variation; and pairwise FST estimates suggest that there are significant rates of migration between populations.