Traceability analysis of forest carbon dynamics with a matrix-represented vegetation demographic model

Understanding carbon cycle dynamics during forest succession is essential for predicting ecosystem responses to environmental change. Vegetation Demographic Models (VDMs), which include detailed demographic processes, offer valuable insights into forest successional dynamics. However, the high complexity of model structure can obscure our understanding of simulated ecosystem carbon dynamics. To address this, we developed a traceability framework to decompose VDM simulations of carbon storage into distinct, traceable components associated with different plant functional types (PFTs). Specifically, the transient carbon storage can be partitioned into three hierarchical layers: (i) carbon storage capacity (Xc) and potential (Xp); (ii) net primary production (NPP), carbon residence time (τ_N), net carbon pool change (X'), and carbon chasing time (τ_ch); (iii) carbon allocation, transfer, and turnover rates. We applied this framework to a cohort-based VDM, Biome Ecological strategy simulator (BiomeE), and evaluated its utility using field observations of 72 species across three plots spanning 150 years of succession in a subtropical forest. The results showed that early succession exhibited high PFTs diversity, including evergreen broadleaf trees, evergreen broadleaf shrubs, evergreen needleleaf trees, deciduous broadleaf trees, and deciduous broadleaf shrubs, driving rapid increases in Xc and Xp. As succession progressed, deciduous PFTs declined, and evergreen broadleaf trees dominated carbon dynamics, with ecosystem carbon storage reaching approximately 40 kg C m^-2 during the mid-succession stage. In the late successional stage, ecosystem carbon storage stabilized at 75 kg C m^-2, closely approaching Xc, which is supported by high NPP (1.37 kg C m^-2year^-1) and long τ_N (70 years), while Xp and carbon sink strength declined. During succession, evergreen broadleaf trees contributed the most to carbon sequestration, with evergreen broadleaf trees (83.73%) > evergreen needleleaf trees (8.11%) > evergreen broadleaf shrubs (5.24%) > deciduous broadleaf trees (2.39%) > deciduous broadleaf shrubs (0.52%). These findings highlight the critical role of successional shifts in forest structure in shaping carbon dynamics in subtropical regions.