Key points are not available for this paper at this time.
• Modeling scenarios revealed the impacts of climate and LUCC on the sustainability of ESs. • The variation in land use conversion ratios can reflect the sustainability of different climate development scenarios. • Temperature, precipitation, and topographic factors are the primary individual drivers influencing ESs. • In ESs management, the synergistic effects of human activities and scale effects should be considered. Human activity has transformed natural land use patterns and exerted a significant impact on the mechanisms underlying ecosystem service (ES) provisioning. This has disrupted the equilibrium between human development and ecological protection and hasled to further degradation of desert ecosystems endemic to arid regions. In this study, we employed the three SSP-RCP scenarios provided by CMIP6 to represent varying rates of development within the Tarim River Basin (TRB) and modeled land use in 2050 based on the coupled system dynamics (SD) + patch-generating land use simulation (PLUS) + InVEST model. A dynamic simulation of the trade-off/synergy between key land use types and ESs in the ecologically vulnerable areas of the TRB was performed to identify the impacts of the individual effects of single factors and the scale effects of multiple factors on ESs. The results showed that (1) in the SSP126, SSP245, andSSP585 scenarios, the ratio of outflows to inflows of converted unused land gradually increased in the following periods:79.28 %, 79.42 %, 79.54 %in 2020–2030; 96.19 %, 91.62 %, 66.07 %in 2030–2040; and 93.76 %, 91.65 %, 60.96 %in 2040–2050, respectively.(2) A high water yield (WY) was observed in the highland mountains; high habitat quality (HQ) was observed in the highland mountains, along the Tarim River, and around cities; high carbon storage (CS) was observed in the highland mountains, along the Tarim River, and in urban areas. (3) Under the three scenarios, WY, HQ, and CS were mainly affected by environmental factors, with the highest single-factor tests represented by temperature (0.594; 0.595; 0.596), precipitation(0.582; 0.589; 0.590); NDVI(0.312; 0.295; 0.299), soil erosion(0.227; 0.221; 0.219); and NDVI(0.298; 0.282; 0.310), soil erosion(0.122; 0.125; 0.132), respectively. Additionally, the interaction factors with the highest explanatory power for WY, HQ, and CS were precipitation ∪ NDVI (0.73); elevation ∪ NDVI (0.37); and soil ∪ NDVI (0.31), respectively. Moreover, the human aggregation level exerted positive (0.117), negative (−0.173), and negative (−0.286) impacts on WY, HQ, and CS, respectively. The results show that the combination of GeoDetector and the partial least squares structural equation modeling (PLS-SEM) provides a complementary and creative framework for investigating the driving mechanisms of ecosystems.
Song et al. (Tue,) studied this question.