A network perspective on ecosystem service trade-offs and synergies relationships and their spatial pattern optimization in arid regions

Elucidating the trade-offs and synergies (TS) among ecosystem services (ESs) is essential in arid regions. However, in arid and semi-arid regions, quantitative characterization of trade-off and synergy strengths among multiple ecosystem services and systematic analysis of their driving mechanisms remain relatively limited. In this study, we evaluated the spatiotemporal characteristics of four ESs from 2000 to 2020 in Xinjiang. A trade-off and synergy index (TSI) is introduced, and a Bayesian network (BBN) model was employed to construct a BBN-TSI model. Sensitivity analysis was used to determines the impact of key variables on the intensity of TS. Furthermore, spatial optimization zones are delineated. The results can be summarized in three critical points. First, between 2000 and 2020, the ESs exhibited spatial heterogeneity and clustering. The Ili River Valley, the Tianshan Mountains, and surrounding upland areas represented the principal distribution of high-value regions. Overall, the physical volumes of ESs increased and then decreased. Second, the relationships among ESs were particularly marked by strong TS. The trade-off zones of the soil conservation and air purification (SC_AP) node and the synergy zones of the water yield and soil conservation (WY_SC) node accounted for the largest spatial proportions, reaching 54.94% and 46.51%, respectively. Third, sensitivity analysis revealed key variables, including soil erosion, temperature, precipitation, and vegetation cover. The Ili River Valley, the Altai region, and specific mountainous regions were identified as high-value sites, prioritizing ecological conservation. These research findings provide theoretical support for elucidating the complex relationships among multiple ESs and offer scientific basis for delineating priority areas for ecological conservation. • Relationships between ecosystem services in the study area were analyzed. • Identifying key variables affecting TS intensity between ESs via BBNs construction. • Visualize key states of variables to identify optimal areas. • A new paradigm for ESs optimization has been established from a network perspective.