Spatiotemporal trends in ecosystem carbon stock evolution and quantitative attribution in a karst watershed in southwest China

Karst ecosystems serve as a vital part of the Earth's ecosystem and have a substantial influence on the worldwide carbon cycle. Revealing the features and driving factors of the spatiotemporal evolution of Carbon (C) stock in karst watersheds is critical for the in-depth exploration of the regional carbon cycle and carbon sources/sinks, as well as in regional ecological restoration. In this study, the Nanming River Basin, a representative karst basin in southwest China, was used as the subject region. Based upon data on land use change from 2000 to 2020, an Integrated Valuation of Ecosystem Services & Tradeoffs (InVEST) model was applied to calculate the C stock from 2000 to 2020 and identify the driving factors of the regional change of C stock using an optimal parametersbased geographical detector (OPGD) model. The findings indicate that: (1) the cumulative reduction of C stock in the basin from 2000 to 2020 was 3.16 x 105 t; the center of gravity of C stock increase has shifted to the southwest by 2434.16 m, and the center of gravity of the reduction has shifted to the southwest by 7260.53 m. (2) The transition of forest land into construction land had the highest contribution to the decrease of C stock (60.91%); the transition of cultivated land and grassland into forest land was conducive to a rise in the C stock, and these conversions contributed 45.93% and 35.00%, respectively, to the increase of C stock. (3) The normalized difference vegetation index (NDVI), population density, intensity of human activity, slope, and lithology all have annual average q-values greater than 10%, meaning they are the primary factors driving the spatial differentiation of C stock. The contribution of the NDVI & AND; slope direction to the regional heterogeneity of C stock was the largest among the interactive driving factors, with an explanatory power close to 30%. The interactive combinations of drivers all showed nonlinear enhancement or two-factor enhancement effects. To some extent, this study deepens the study of changes in C stock and related driving mechanisms in karst areas, intending to provide a scientific foundation for the recovery of fragile ecosystems and to support a regional lowcarbon economy.