Detecting Zone-Type Thresholds for Soil Organic, Inorganic, and Total Carbon Pools in China's Drylands.

Growing concerns about the accelerating global changes in drylands have intensified interest in understanding the impacts of diverse environmental factors on various soil carbon components, particularly their potential threshold effects, which may lead to abrupt nonlinear changes in both the quantity and composition of soil carbon. However, most research has predominantly focused on identifying key environmental drivers of either organic or inorganic carbon separately, often neglecting the presence and range of multiple critical thresholds. This study addresses this gap by analyzing extensive field data, including soil carbon measurements and ecosystem variables, collected across a ~4000 km transect in China's drylands. Utilizing a gradient forest model combined with threshold analyses, we assess the impacts of key drivers, including sand content, total nitrogen, aridity, and cation exchange capacity, on soil carbon pools. Our findings indicate that nitrogen content is the most influential factor for soil organic carbon, which was sensitive to low levels of nitrogen (0.07%-0.08%), with a slower response observed at higher levels. Aridity significantly affects both organic and inorganic carbon pools, with identified threshold zones for organic carbon at aridity levels of 0.48-0.52 and 0.75-0.85, and for inorganic carbon at 0.82-0.88. Threshold zones of sand content for total carbon are identified at lower levels and a wider range (51.4%-64.1% and 87.3%-88.1%), due to its negative effects on both organic and inorganic carbon, impacting 20% of the dryland area. Spatial variations in threshold effects revealed a trade-off between nitrogen and pH in regulating soil total carbon. The combined threshold effects of climate warming and aridification may pose a greater threat to soil organic carbon in high-latitude regions. This research enhances understanding of soil carbon dynamics in arid environments and offers novel approaches and insights for identifying thresholds in ecosystems that are increasingly at risk of reaching tipping points.