Scale-dependent drivers of ecosystem service supply-flow-demand couplings in the Shanxi Yellow River Basin, China

Integrating ecosystem service (ES) supply-flow-demand relationships across multiple scales is crucial for effective ecological management. Using the Shanxi Yellow River Basin as a case study, we evaluated six ESs at the grid, watershed, and county scales in 2020. By applying redundancy analysis and extreme gradient boosting, we identified key drivers and simulated ES flows. Results show: (1) overall ES supply meets demand, with most ESs clustering in the northwest and southeast, while crop production dominates central lowlands; (2) supply-demand mismatches exhibit scale-dependent effects, showing fine-grained details at grid level but amplified heterogeneity at broader scales; (3) while main drivers (temperature, precipitation, population density, GDP) remain constant across scales, their influence intensity varies markedly; (4) ES flows display distinct directions, with recreation and crops moving urban-to-rural, while other ESs oppositely. This framework offers a transferable approach for watershed-scale sustainability planning in regions facing ecological-economic trade-offs.