Coastal regions in developing economies increasingly face trade-offs between ecosystem services (ESs) provision and urban expansion, while scenario-based projections of ESs remain limited. This study integrated the Patch-generating Land Use Simulation (PLUS), Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST), and machine-learning approaches to quantify five ESs—water yield (WY), soil retention (SEC), carbon storage (CS), habitat quality (HQ), and phosphorus emission (PE)—under natural development (ND), economic development (ED), and ecological protection (EP) scenarios in Fujian, China, during 2030–2060. Key findings are as follows: (1) Model validation demonstrated high predictive accuracy (Kappa = 0.86; Figure of Merit = 0.40). ND and ED scenarios resulted in approximately 7.5% forest loss, whereas the EP scenario preserved 99.17% of existing forest cover. (2) Temporal analysis revealed divergent ESs dynamics: CS and HQ decreased by 2.4% and 4.4%, respectively, while WY, SEC, and PE increased concurrently with urban expansion. Conversely, the EP scenario exhibited marginal increases in WY (0.03%), SEC (0.01%), and CS (0.71%), while PE and HQ remained stable. Spatial correlation analysis identified seven synergistic relationships among WY, SEC, HQ, and CS, and three trade-off relationships involving PE. (3) Ecosystem service bundles (ESBs) were classified into four distinct types. ND and ED scenarios substantially restructured ESBs, whereas the EP scenario maintained structural stability and high service diversity. (4) ESBs under the ND and ED scenarios exhibited alternating dominance of climatic, topographic, and socio-economic drivers, whereas the EP scenario fostered stable ESBs configurations through topography–climate synergies. Overall, this study provides robust scientific support for cross-scale governance aimed at achieving regional ecological–economic equilibrium.
Lin et al. (2026) studied this question.