Abstract: Forest carbon stock (FCS) is a key component of the terrestrial carbon cycle, and accurately quantifying its spatial patterns and drivers is a critical prerequisite for carbon accounting, particularly when distinguishing between natural (NF) and planted forests (PF). Using 7,980 field plots across China, we integrated Geographical Detector analysis, growth-curve modeling, structural equation modeling and Random Forest algorithm to investigate the determinants, growth dynamics, and spatial patterns of above-ground biomass carbon (AGBC), below-ground biomass carbon (BGBC), and total vegetation carbon storage (TC) across China’s forests. Our results showed that stand age emerged as the key factor of FCS in both NF and PF and exhibited the strongest direct effect when compared to climate, soil, and topographic factors. Growth models further indicated that NF follow a nonlinear saturation trajectory, best described by the Logistic model under low density, with asymptotic TC being higher under high density (103.86 t ha^-1) than under low density (98.95 t ha^-1). In contrast, FCS in PF increased linearly, with the highest accumulation rate (1.43 t ha^-1 yr^-1) observed at high stand density. Spatial predictions from the Random Forest model showed that AGBC, BGBC, and TC exhibited similar distribution patterns, with high values in northeastern, southwestern, and Southeastern China, and low values in central China. The national predicted FCS ranged from 12.80 to 137.67 t ha^-1 for AGBC, 2.72 to 31.62 t ha^-1 for BGBC, and 16.70 to 198.05 t ha^-1 for TC. These findings underscore the critical role of forest origin in FCS estimation and provide a robust reference for improving carbon stock assessments in China.

Key words: Total vegetation carbon storage, above-ground biomass carbon, below-ground biomass carbon, forest origin, stand age