Abstract: The environment of Horqin Sandy Land has gradually improved with the implementation of ecological engineering such as the ThreeNorth Shelter Forest Program and the Conversion of Cropland to Forest Program. Exploring the trade-offs and synergies among ecosystem services and their driving factors in Horqin Sandy Land is of great significance for ecological environment protection and governance. We quantitatively assessed four ecosystem services in Horqin Sandy Land in 2022 using the InVEST model, including water yield, soil retention, carbon sequestration, and habitat quality. We analyzed the spatial distribution and driving factors of the trade-offs and synergies among the four services based on Bayesian network and bivariate spatial autocorrelation methods. The results showed that (1) The spatial distribution of soil retention, carbon sequestration, and habitat quality was relatively similar, showing a distribution pattern of high in west and low in east, while water yield showed a pattern of high in east and low in west; (2) Rainfall, rainfall erosion, land use, evapotranspiration, slope, NDVI, NPP, and nighttime light were important influencing factors for the four ecosystem services; (3) At the regional scale, water conservation showed a tradeoff relationship with soil conservation, carbon sequestration services, and habitat quality. There were synergistic relationships among soil conservation, carbon sequestration, and habitat quality. At the county scale, the synergistic relationships between soil conservation and carbon sequestration, as well as between soil conservation and habitat quality, showed spatial heterogeneity and scale effects; (4) The Bayesian network scenario simulation indicated that land use and slope were the driving factors of the trade-off and synergy relationships among the four ecosystem services. Our results can provide a reference for the efficient utilization of ecosystem services in Horqin Sandy Land.

Key words: Horqin Sandy Land, ecosystem services, trade-offs and synergies, Bayesian network, driving factor