Stability domains in barrier island dune systems

Early ecological descriptions of barrier island dune landscapes recognized the importance of biogeomorphic feedbacks and thresholds. However, these dynamics have not been formally linked to complexity theory. In this article, I develop models of dune landscape phase states, or stability domains, based on a synthesis of these prior studies and statistical analyses. Data for these analyses were obtained from compositional and topographic sampling along Sapelo Island, Georgia and South Core Banks, North Carolina. These undeveloped barrier islands are at opposite ends of a regional meteorological gradient in stormforced overwash disturbance along the southeastern US Atlantic coast. The topography and plant functional group abundances on each island were supportive of earlier studies noting the propensity for distinct self-organizing biogeomorphic feedbacks to emerge along barrier coasts. On South Core Banks, where storm forcings of overwash are more frequent, the recovery processes initiated by burial-tolerant species may reduce topographic resistance to overwash exposure and contagion. On Sapelo Island, a greater abundance of dune-building and burial-intolerant swale species may confer a higher topographic roughness and more resistance to overwash exposure and contagion. Conceptually, the biogeomorphic modifications of overwash forcings on each island perpetuate the habitat conditions and dominant dune plant species in a weak positive feedback. To provide more empirical evidence for these models, I tested a suite of hypotheses built from the premise that the biogeomorphic interactions on each island engender unique landscape patterns. These patterns originate in each island’s distinctive coupling of extrinsic overwash disturbance processes with intrinsic plant-mediated topographic modification. Via this premise, the properties proposed for each domain model were subsequently supported, through: (1) the structure of compositional variance along principal coordinate ordination axes; (2) differences in biotopographic thresholds measured in cross sections; and (3) the distinctive elevational boundaries and spatial scales over which elevation exerted its influence on species diversity. These two domain models can be used as conceptual templates for the development of dune computational simulations. # 2005 Elsevier B.V. All rights reserved.