A new method for mangrove health assessment based on knowledge graph and link analysis

Assessing mangrove health is critical for ecosystem protection. However, prevailing Mangrove Health Index (MHI) methods often rely on subjective variable selection and weighting based on observational data, which limits their objectivity and cross-regional applicability. To overcome these issues, this study introduces a novel assessment framework that integrates a knowledge graph (KG) with link analysis. The KG was constructed by extracting causal relationships from extensive literature, thereby identifying 24 key biotic, abiotic, and socio-economic indicators. Link analysis was then applied to quantitatively determine indicator weights, assigning greater importance to variables with higher systemic influence. We validated the framework’s practicality and sensitivity through integrated field surveys and Normalized Difference Vegetation Index (NDVI) analysis at three sites with varying disturbance levels in China’s Beibu Gulf. The results showed that the derived MHI scores accurately reflected the disturbance gradient and correlated strongly with NDVI, which decreased with increasing disturbance. Our MHI framework enhances cross-regional comparability and serves as a transferable tool for global mangrove and natural resource assessments. Synopsis: The knowledge-driven environmental assessment method establishes a robust scientific foundation for designing adaptive ecological restoration protocols and context-aware compensation mechanisms. • A novel KG and link analysis framework for mangrove health assessment. • Graph-based indicator quantification via random walks and probability computation. • Validation through a comprehensive one-year field survey in the northern Beibu Gulf.