The seasonal dominance mechanism of temperature and runoff on the carbon residence time and the dynamic imbalance of carbon cycle in the Pearl River Basin

The seasonal dynamics of carbon residence time (CRT) and its regulatory mechanisms under climate change remain insufficiently quantified, particularly regarding the interactions between temperature and runoff on CRT. This study systematically quantified the seasonal dynamics of CRT in vegetation-litter-soil continuum ecosystems across spring, summer, and winter in the Pearl River Basin, China. An integrated approach combining structural equation modeling, threshold detection, and scenario prediction analysis was used. The associations among monthly average temperature (AMT), surface runoff (SFR), ecosystem CRT, net ecosystem productivity (NEP), and ecosystem carbon use efficiency (CUEe) were evaluated to examine a potential pathway linking hydrothermal variability with carbon-cycle disequilibrium. The results reveal how seasonal CRT variation is associated with changes in NEP and CUEe, providing evidence that carbon-cycle disequilibrium in the Pearl River Basin is organized by season-specific hydrothermal conditions rather than representing a simple input-output mismatch. Scenario simulations indicated season-dependent CRT responses under future hydrothermal changes, with the strongest increase projected in winter. Winter carbon loss was markedly intensified under RCP8.5, as indicated by a NEP decline of more than 5.58%. A 23%–29.7% increase in CRT was associated with a 4.13%–5.51% higher probability of NEP transitioning from carbon source to sink, whereas high-CRT scenarios were associated with a 36.1%–58.7% reduction in CUEe.