Dryland agricultural expansion leads to lower content and higher variability of soil inorganic carbon in topsoil

Soil inorganic carbon (SIC) is of key importance to the global carbon cycle and mitigating climate change. However, decades of agricultural expansion in the drylands have resulted in vast and fragmented agricultural landscapes and have potentially affected SIC. Previous studies investigating the impact of agricultural expansion on SIC have primarily focused on cultivation impacts, while ignoring edge effects from agricultural land boundaries. Here, we combined cultivation impacts with edge effects and fully evaluated the impacts of agricultural expansion on SIC. We conducted spatial statistical and buffer analyses using a SIC distribution map, which was developed using 388 sample points in 0 - 20 cm topsoil over the Sangong River watershed in drylands of China. We found that decades-long agricultural expansion resulted in lower content and higher variability of SIC in topsoil. Specifically, cultivation decreased SIC content by 28.57% (1.64 g kg-1) and increased SIC variability by 95.71% (0.67 g kg- 1) in agricultural lands. Edge effects altered SIC content by - 1.95-11.51% (-0.11 to 0.54 g kg- 1) and increased SIC variability by 39.14% (0.34 g kg- 1) at edges. In addition, we found that fewer sampling points in areas with high SIC variability (e.g., agricultural lands and edges) decreased the estimation accuracy. Our study highlights the need for SIC conservation during dryland agricultural expansion. Also, it suggests that as many sampling points as possible should be set in high variability areas, which is especially useful for accurately estimating SIC distribution in drylands.

Automated summary

Dryland agricultural expansion leads to a decrease in soil inorganic carbon (SIC) content and an increase in variability, with edge effects also influencing SIC content and variability. It is crucial to conserve SIC during dryland agricultural expansion and to increase the number of sampling points in areas with high variability to accurately estimate SIC distribution.