Interactions between ecosystem services and their causal relationships with driving factors: A case study of the Tarim River Basin, China

• Provided a new approach to studying causal relationships between ESs and driving factors. • Synergistic effects among ESs dominate in the TRB. • The CS bundle and WY bundle face degradation risks. • The dominant direction of bidirectional asymmetric causality between different ESs and driving factors was revealed. Clarifying different ecosystem service (ES) interactions and their primary driving factors is essential for effective ecosystem management. Grassland degradation, interrupted river flow, and intensified human activities pose serious threats to the ESs of the Tarim River Basin (TRB). However, there is insufficient research on the between ES interactions and their causal relationships with drivers in the TRB. Therefore, this study measured four key ESs in the TRB: water yield (WY), carbon sequestration (CS), soil conservation (SC), and habitat quality (HQ). Correlation analysis and bivariate local spatial autocorrelation were employed to uncover trade-offs and synergies between different ESs from both holistic and spatial perspectives and ES bundles were identified using self-organizing maps. Geographic convergent cross-mapping was utilized to investigate the cause-and-effect relationships between ESs and their influences, pinpointing the main drivers. The findings revealed that: (1) from 2000 to 2020, WY and SC decreased, whereas CS increased markedly. HQ initially declined but then improved, with an overall insignificant change. Spatially, low-value ES regions were in the central and eastern desert areas, high WY and SC values occurred in mountainous regions, and high CS and HQ values were found in oases and mountainous areas; (2) ESs exhibited significant synergy throughout the watershed. Spatially, trade-offs and synergies coexisted, with high-high synergy predominating in mountainous regions and low-low synergy occurring primarily in the central and eastern desert areas. Trade-off effects were limited, mainly occurring in oases and parts of the Kunlun Mountains. ES bundles exhibited signs of change or deterioration, and the CS regulation bundle and WY supply bundle in particular face degradation risks; (3) the dominant direction of bidirectional asymmetric causality differed across ESs and drivers. Overall, the dominant direction of WY and drivers was that WY influenced drivers (WY → drivers), whereas SC was typically influenced by drivers (drivers → SC). The dominant orientation of CS and HQ concerning drivers is that natural factors influenced these ESs (natural factors → ESs), while human factors were influenced by ESs (ESs → human factors). The main drivers for WY and SC were precipitation, temperature, potential evapotranspiration, and elevation. The main drivers for CS and HQ were land use intensity, followed by precipitation, potential evapotranspiration, and temperature. The results of this study provide a reference for the conservation and management of ESs in the TRB.