Paper short abstract:

Crop rotation affects the spatial variability of soil properties, and may offer considerable potentials for farmland SOCD mapping in plains. This paper obtained crop rotation maps using time-series remote sensing images and incorporated rotation information in improving SOCD mapping.

Paper long abstract:

Accurate mapping of farmland soil organic carbon density (SOCD) is crucial for precision agriculture and carbon sequestration potential evaluation. Topographical factors are the most commonly used and effective variables to assist in mapping soil properties. However, they perform poorly when applied to plains. The impact of human activities, especially farmland management, on soil properties have drawn increasing attention. Crop rotation that involves crop species and multiple cropping index affects the spatial variability of soil properties, and may offer considerable potentials for farmland SOCD mapping in plains. The objective of this paper is to obtain crop rotation maps and incorporate rotation information in mapping SOCD. Specifically, using 230 topsoil samples collected in Jianghan Plain, China, we (i) obtain crop species and multiple cropping index maps using time-series HJ-1A/1B satellite images; (ii) investigate the difference of SOCD under different crop rotations, and (iii) explore the effectiveness of incorporating rotation information in mapping SOCD. The results of two-way analysis of variance showed significant difference of SOCD under different crop species. The SOCD under monoculture was higher than that of multiculture. Post hoc test results suggested that the SOCD under single cropping rice and soybean was high, while SOCD under maize-artemisia selengensis and sesame-wheat was low. The effectiveness of regression Kriging model that integrated rotation information was favored by the model validation results. The findings reveal the importance of single cropping rice and soybean in reducing the carbon loss and the effectiveness of crop rotation in improving farmland SOCD mapping in plains.