Quantifying spatial structures of managed and unmanaged coastal dune systems using grid-based landscape metrics

Abstract The spatial structures of coastal dunes, represented by the composition and configuration of roughness elements such as vegetation, wrack, and sand fences, reflect geomorphic dynamics influenced by natural and anthropogenic factors. Despite widespread recognition of the structural differences between managed and unmanaged systems, quantitative measurements and comparisons of their spatial characteristics remain limited. This study employs an optimized approach for grid-based landscape metric analysis using very high-resolution (VHR) aerial imagery to quantify dune structures across various managed and unmanaged systems. Sensitivity analysis of varying hexagonal grid sizes revealed that a 15 m sample unit was optimal for minimizing the percentage of saturated grid cells, stabilizing landscape metric values, and reducing the influence of grid position. After performing Principal Component Analysis and intercorrelation analysis to assess the redundancy of landscape metrics, patterns of Percent Cover (PC), Number of Patches per Class (NPC), Largest Patch Index (LPI), Edge Density (ED), and Perimeter-Area Ratio (PAR) were examined to capture the size, degree of subdivision, and shape complexity of dune structures. Using a 15 m sample unit size and the selected metric set, grid cells were classified based on PC, enabling quantitative comparisons between managed and unmanaged systems. Our findings demonstrated that unmanaged systems generally exhibited more fragmented and complex dune structures compared to managed systems. However, some beach-dune management practices, such as sand fence installation and nourishment, could promote structural characteristics resembling those of unmanaged dunes. This approach provides valuable insights that can inform adaptive coastal protection strategies in response to ongoing environmental changes and anthropogenic disturbances.