Climate-sensitive spatial variability of soil organic carbon in multiple forests, Central China

The soil organic carbon (SOC) content is determined by multiple factors in terrestrial ecosystems. This study explored the SOC content and its influencing factors in multiple forests of subtropical, temperate, and transitional climate zones in Central China. The quantile regression forest algorithm (QRF) was employed to model the SOC contents along the soil depths up to 100 cm. The models had the highest explained variance score of 0.621 for the topsoil layer (0–20 cm), which decreased with soil depths. In addition to parent material, bulk density, mean tree height, NDVI and topography, climate factors, especially summer precipitation, and surface soil temperature, were identified as the most influential factors for SOC accumulation and spatial variability in these forests. SOC was positively affected by summer precipitation from zero to 100 cm. A higher surface soil temperature was beneficial only for forest SOC in the top layer. The highest SOC > 50 g kg−1 was found under the cover of mixed coniferous forests and evergreen broad-leaved forests at high elevations approximately 1000 m asl, which were dominated by high summer precipitation. At the vertical depth, SOC largely accumulated at a depth of 40 cm but decreased thereafter. Owing to the positive impact of precipitation on forest SOC and the urgent demand for increasing carbon sinks, forest management could focus on mixed evergreen broad-leaved forests and mixed coniferous forests, afforestation, and reforestation of mixed forests based on local species in vast areas. [Display omitted] •Quantile regression forest algorithm was used for predicting forest SOC.•Precipitation was the dominant climatic indicator for spatial variability of SOC.•Mixed coniferous and evergreen broad-leaved forests contributed the greatest in forest soil carbon sinks.•Adaptative management on mixed forests is recommended for forest productivity, carbon sequestration and conservation.