Long-term carbon transfers due to land use changes in the Chinese Loess Plateau

Anthropogenic land cover change (ALCC) is a primary driver of global carbon cycle perturbation, yet its millennial-scale impacts remain poorly quantified. It profoundly alters carbon cycling through its dual effects on soil erosion and vegetation productivity. Here, we quantify for the first time, millennial-scale (5000–100 BP) carbon flux trends and their driving mechanisms in the Wei River catchment, a critical agricultural region within the Chinese Loess Plateau (CLP) and a major tributary of the Yellow River. Our coupling modeling approach, which combines landscape evolution modeling with machine-learning methods, reveals that ALCC has dominated carbon budgets for millennia by simultaneously increasing lateral soil organic carbon (SOC) export and suppressing vertical CO 2 sequestration: ALCC-induced soil erosion has caused a significant lateral SOC loss (ranging from 0.76 to 8.8 Mt C yr −1 ) since 5000 BP. It exceeds contemporaneous atmospheric carbon uptake through vegetation by a factor of 1.2 – 4.1. The latter has cumulatively declined by 35.3% in response to ALCC-induced decline in net primary production (NPP). As a result, the catchment experiences a net carbon change ranging from −5.3 Mt C yr −1 (net loss) to 0.2 Mt C yr −1 (net uptake). However, since ∼ 98% of exported carbon is stored in the lower Yellow River, and ultimately, the Bohai Sea, the overall source-to-sink river system act as a carbon sink. Critically, the effectiveness of this overall-sink has been decreasing since 5000 BP due to the vegetation-uptake decrease in response to land use change. Our work provides long-term insights for soil carbon management strategies in regions with historical agricultural legacies and highlights the necessity of integrating erosion and productivity feedbacks in assessing terrestrial carbon budgets.