Synergies and trade-offs in forest carbon pools: separating universal drivers from forest type-specific controls.

Introduction: Understanding the synergies and trade-offs among tree, understory, and soil carbon pools is critical for optimizing forest carbon sinks. However, the mechanisms regulating these relationships, particularly how they differ between natural and planted forests, remain unclear. This study aims to deconstruct these complex interactions in subtropical forests to provide a scientific basis for enhancing ecosystem carbon storage. Methods: Based on data from 440 plots covering six major forest types in subtropical China, we employed linear mixed-effects models (LMMs) to quantify universal and context-dependent driver effects. We then used structural equation models (SEMs) to test and compare the mechanistic pathways of carbon allocation in natural versus planted forests. Results: Our LMMs revealed a universal trade-off, with tree layer carbon density (TCD) strongly suppressing understory carbon density (UCD) (β = -0.22, P < 0.001) while synergistically promoting soil organic carbon density (SOCD) (β = 0.36, P < 0.001). SEM analysis (natural forests: CFI = 0.986, RMSEA = 0.064; planted forests: CFI = 0.960, RMSEA = 0.076) revealed divergent regulatory mechanisms. In natural forests, tree diversity directly buffered the suppressive effect of TCD on UCD via a significant positive path (β = 0.22, P < 0.01). This buffering pathway was absent in planted forests, leading to an amplified TCD-UCD trade-off (β = -0.54, P < 0.001). Discussion: Our findings demonstrate that forest carbon allocation is governed by a vertical trade-off and an above-belowground synergy, with tree diversity acting as a key modulator. Compared with complex natural forests, the carbon allocation mechanism in planted forests is simplified with more acute trade-offs. We conclude that enhancing structural and species diversity in plantations is a critical pathway for synergistically optimizing the entire ecosystem's carbon sink capacity.