Field-Validated Ensemble Habitat Suitability Modeling Indicates Seagrass Ecological Potential and Carbon Storage Capacity

Effective coastal management of seagrass systems requires identifying both extant meadows and degraded areas suitable for recovery. However, these decisions often rely on distribution maps that provide limited information on ecological condition or carbon potential. Here, we develop an ensemble habitat modeling (EHM) framework for seagrass meadows around Guanglu Island, China, to test whether a habitat suitability index (HSI) can serve as a proxy for ecological performance and carbon storage potential. By integrating multiple single-algorithm predictions, the ensemble model reduced algorithm-dependent variation and achieved strong predictive performance (ROC = 0.967, TSS = 0.876), with predictions capturing 92.59% to 98.98% of the 2023 field-mapped meadow area. HSI was positively associated with aboveground biomass and shoot density. A predictive model fitted directly with biomass as the response showed that aboveground biomass increased with HSI, with a leave-one-out cross-validated RMSE of 77.68 g DW m−2. High HSI samples also exhibited smaller relative deviations from the theoretical biomass–density envelope and a greater realized fraction of theoretical biomass potential. Modeled sediment carbon densities were consistent with the 2023 sediment comparison, thereby supporting conditional carbon stock screening. Our findings support the use of ensemble-derived HSI as a local spatial proxy for biomass-related ecological potential, rather than as a direct measure of habitat quality, actual biomass, or realized carbon stock. By separating predicted suitable habitat, observed meadow distribution, and model-based potential carbon stock, the framework provides a spatial screening tool for conservation prioritization, restoration opportunity assessment, and blue carbon planning under explicitly stated ecological and modeling assumptions.