Abstract: Soil organic carbon (SOC) in croplands plays an important role in stabilizing the ecological environment and food security. Exploring the spatial variations and influencing factors of SOC is of great significance for promoting the sustainability of regional agriculture. We analyzed the spatial differentiation characteristics and influencing factors of cropland SOC using redundancy analysis (RDA) and structural equation model (SEM), with 150 sampling points from cropland topsoil across the Yarlung Zangbo River, the Nyangqu River, and the Lhasa River region (the YNL river region) in Tibet. The results showed that cropland SOC in the YNL river region was low in the west and high in the east, and low in the mainstream valley of the Yarlung Zangbo River and high in the source areas of the north and south tributaries. Overall, the average SOC content in the YNL river region was relatively low. Soil silt content was the dominant factor in the regional spatial variations of SOC, explaining 14.0% of total variation. Mean annual evaporation (MAE) could better explain the spatial variations of SOC than mean annual temperature (MAT) and mean annual precipitation (MAP). The SOC content in this region mainly showed vertical and meridional zonality. Altitude affected SOC levels indirectly by affecting temperature and evaporation, while longitude affected it indirectly by affecting precipitation and soil pH. The spatial variations of SOC content were indirectly affected by climate, vegetation, and soil properties. This study provides a scientific basis for the prediction of spatial variations of SOC and regional agricultural development.

Key words: soil organic carbon, spatial differentiation, Qinghai-Tibet Plateau, redundancy analysis (RDA), structural equation modeling (SEM)