Fraction-specific soil carbon responses to decadal warming in two contrasting permafrost grasslands of the Qinghai–Tibet Plateau

Permafrost regions store vast amounts of soil organic carbon (SOC), yet SOC responses to warming have remained inconsistent across studies. In particular, previous research has primarily focused on bulk SOC measurements, an approach that obscures fraction-specific responses and hinders mechanistic understanding. We therefore investigated responses of SOC and its fractions in two representative permafrost grasslands contrasting in their hydrological and biogeochemical conditions on the Qinghai–Tibetan plateau. Open-top chamber warming experiments were conducted in an alpine meadow and a swamp meadow to yield average warming of + 2.4 °C (future 2 °C scenario) and + 4.9 °C (future 4 °C scenario) over 9 and 12 years, respectively. We analyzed SOC and its fractions, particulate (POC), mineral-associated (MAOC), calcium-bound (Ca-OC), iron-chelated (Fe 3 ⁺-OC), and reactive mineral–associated carbon (OC DP ) at 0–10 and 10–20 cm soil depths. Additionally, we characterized the chemical composition of SOC fractions (derived from UV and fluorescence spectroscopy), aboveground biomass, and soil mineralogical properties to identify the drivers of SOC responses to warming. In alpine meadows, 12-yr high warming significantly increased bulk SOC, Ca-OC, Fe 3 ⁺-OC at both depths ( p < 0.05). Swamp meadows showed stable bulk SOC, but increased Ca-OC and decreased OC DP . Regression and redundancy analyses revealed alpine SOC accumulation via enhanced plant inputs and metal associations. In contrast, swamp SOC was constrained by Ca-OC formation offset by OC DP decomposition under waterlogged conditions. Ecosystem-dependent SOC responses to warming are mediated by fraction-specific dynamics. Incorporating fraction-specific responses into carbon cycling models is critical for improving predictions of permafrost soil carbon dynamics under future climate change.