Effects of land use/cover changes on soil organic carbon stocks in Qinghai-Tibet plateau: A comparative analysis of different ecological functional areas based on machine learning methods and soil carbon pool data

Abstract

Understanding the process of land use/cover changes (LUCC) can provide experience on the enhancement of soil organic carbon (SOC) stocks and carbon sequestration potential for different areas. This study is uniquely to divide different ecological functional areas, and originally combine the machine learning method and soil carbon pool dataset for regional comparative analysis, to compare and quantitatively analyze the drivers of LUCC and the changes in SOC stocks effected by LUCC over 30 years. The results show that topography and climate changes are the main drivers affecting LUCC in four natural areas, while soil factors and population changes do not cause significant effects. The total SOC stocks in Qinghai was increased by 71.18 Tg C and 107.19 Tg C in 0–30 cm and 0–300 cm layers, respectively, and the highest SOC stocks within 0–300 cm were in Pastoral area. Desert and Gobi area had the lowest SOC stocks in both 0–30 cm and 0–300 cm layers. SOC stocks increased in both 0–30 cm and 0–300 cm layers only in Sanjiangyuan Natural Reserve, while the Desert and Gobi area showed a decrease in both over 30 years. This study emphasizes the significant impact of grassland changes on SOC stocks, indicating the importance of considering these changes in land management and ecological protection policies. The initial and original SOC stocks of pre-LUCC may influence the SOC stocks in post-LUCC. The response of SOC stocks changes to LUCC was varies in different areas. The heterogeneity of different ecological functional areas is affected by multiple factors and SOC stocks will become more complex among these areas in the future. These findings contribute to the development of ecological protection policies and the enhancement of regional land management strategies.