Impacts of agricultural intensity on soil organic carbon pools in a main vegetable cultivation region of China

• We map the SOC density distribution of a main vegetable cultivation area in two periods.
• SOC density increased as the increase of agricultural intensity.
• Agricultural intensity index is better than single indictors in interpreting SOC changes.
• An S-curve model was the best fit for SOC densities with the agricultural intensity index.

Agricultural intensity, a function of agricultural input and output, impacts agricultural soil organic carbon (SOC) pools. Investigating the agricultural intensity and SOC density relationship supports understanding of anthropogenic activities on SOC pools in agricultural regions. Cangshan County, one of the most important vegetable cultivation counties in China, was selected as the study area for this investigation. By using soil survey data compiled in 1980 and 2008, 21 towns in Cangshan County were used as study units to investigate the relationship between agricultural intensity indicators and index, with SOC density. Results demonstrate that single agricultural intensity indicators could not reflect SOC densities changes well. Conversely the agricultural intensity index, a composite measure of agricultural input, output and input–output indicators (AI1, AI2 and AI3, respectively), was a more reliable measure reflecting SOC density variability. An S-curve model, SOCd (t ha−1) = exp (α-β/AIi), was the best fit for the plot of SOC densities with the agricultural intensity index. Relationships between AI3 and SOC density were statistically significant for 1980, 2008 and their variation during 2008–1980 (p < 0.05), indicating that AI3 had a more robust relationship with SOC density compared with AI1 and AI2. Though the relationships have low precision (R2 = 0.24–0.40) for not including the natural factors and more indictors relative to variation of SOC density should be selected to improve the interpretability, it is still valuable to introduce the agricultural intensity index when detecting dynamics of SOC pools at a large regional scale.