Topographic and Human-Pressure Interactions in Micro-Island Ecological Quality: A Geodetector-Based Framework from Jiangsu, China

Island ecosystems are highly sensitive to climate change and intensive human activities. However, most existing assessments rely on subjective weighting and linear assumptions, which may not capture nonlinear temporal evolution and can be distorted by spatial-scale effects, including extreme density outliers on micro-islands. Focusing on five representative islands in Jiangsu Province, China (2014–2024), this study applies a data-driven evaluation framework. First, we constructed an Island Ecological Quality Index (IEQI) using principal component analysis (PCA) to integrate multidimensional ecological-state variables. Second, to reduce sensitivity to selected scale-related extremes in spatially restricted units, we employed a geographical detector model to identify explanatory factors and characterize interactive associations. The IEQI showed a nonlinear temporal trajectory, with fluctuations followed by recovery and stabilization after 2021. Under the robust adaptive Jenks classification, the factor detector showed high relative explanatory power for Artificial Shoreline Ratio (q = 0.8457), Population Density (q = 0.8339), Terrain Index (q = 0.8135), Tourist Density (q = 0.6282), and Built-up Ratio (q = 0.3189). Interaction analysis identified bi-factor enhancement between Terrain Index and Population Density (q = 0.8364) and nonlinear enhancement between Policy Count and Built-up Ratio (q = 0.6424). These q statistics describe spatial explanatory associations rather than causal effects. These findings challenge linear and generalized coastal management paradigms. We propose a “one island, one policy” framework that explicitly distinguishes micro-islands by their topographic buffering capacity and anthropogenic pressure profiles, enabling spatially targeted conservation and restoration strategies.