Integrating MaxEnt and PLUS for Predictive Modeling of Mangrove Suitability and Land Use Impacts in Coastal Guangdong, China.

Mangroves, as one of the most efficient carbon-sequestering ecosystems globally, provide critical ecological services including coastal protection, water purification, and biodiversity conservation. This study integrates remote sensing and ecological modeling to assess the spatial-temporal dynamics and restoration potential of mangroves in coastal Guangdong, China, under alternative socioeconomic pathways. Using Sentinel-2 imagery and hybrid classification, we mapped the 2020 mangrove distribution with high accuracy, estimating a total area of 110.28 km2. The MaxEnt model, driven by 37 climatic, hydrological, topographic, soil, and land use/land cover (LUCC) variables, identified a maximum potential habitat area of 1908 km2. However, when LULC constraints were introduced, the suitable area decreased to 1221 km2, highlighting LULC as a major limiting factor. Variable importance analysis revealed that LULC (38.2%), annual temperature range (20.8%), and distance to coastline (11.7%) jointly explained over 70% of the variation in habitat suitability, underscoring the interplay of anthropogenic and natural drivers. By coupling MaxEnt with LULC projections from the Patch-generating Land Use Simulation (PLUS) model under three SSP scenarios for 2040, we further identified dominant land-use constraints and generated multi-scenario distribution predictions. Results suggest that the SSP126 (low-emission) pathway provides the greatest restoration opportunities, whereas SSP585 (high-emission) favors limited climatic suitability gains but widespread habitat loss due to urban expansion. Our findings emphasize that integrating ecological niche modeling with advanced LULC simulation provides a novel decision-support framework that links potential habitat prediction with realistic land-use governance. The study not only identifies high-priority restoration and conservation zones but also offers actionable strategies-such as pond-to-forest conversion, hydrological connectivity restoration, and zoning-based adaptive management-to reconcile coastal development with long-term ecosystem resilience.