Spatial Analysis of Landscape Diversity

This study presents a spatially explicit method for assessing landscape diversity by integrating Landscape Character Analysis (LCA) with multivariate statistical techniques. A two-level landscape characterization approach was employed: Level 1 Landscape Character Types (L1-LCTs) were clustered using Principal Component Analysis and k-means clustering, resulting in 13 distinct landscape character clusters. Level 2 Landscape Character Types (L2-LCTs) were then used to analyze landscape diversity within each cluster through the Shannon Diversity Index (LDiv), Landscape Evenness (LEven), and Mean Nearest Neighbor Distance (MNND). Results show that topographic and climatic heterogeneity are the main drivers of diversity. A key contribution of the study lies in its two-tiered framework, where regional-scale structural patterns (L1-LCTs) define broader landscape clusters, and detailed sub-regional data (L2-LCTs) quantify internal heterogeneity. This hierarchical approach enables the integration of upper-scale landscape structure with lower-scale variation, enhancing the understanding of spatial complexity. The study also introduces a novel boundary redefinition strategy that maintains ecological continuity by extending analysis beyond administrative borders. By linking clusters of landscape structures and functions, the method provides a more nuanced interpretation of landscape diversity and offers a replicable model for supporting landscape-based planning and characterization efforts.