Variation of soil organic matter with particle size in the wind erosion region of northern China

•SOM was measured in eight particle size for 84 samples in northern China.•The distribution of SOM in each size class basically decreased from east to west of the study region.•Though SOM mostly fell with size class, 25 % of the samples showed an increase or no monotonic trend.•Predictive models were built for the relationships between SOM and particle size. Soil wind erosion is one of the most important surface processes in the arid and semi-arid regions of northern China, which serves as an important driver for the migration and turnover of soil organic matter (SOM) and should be given full attention in regional as well as in global carbon cycles. Wind tends to transport fine soil particles typically enriched with SOM, which, however, has been frequently ignored and the mean SOM of the bulk soil is instead used in the carbon budgets at various scales. The objective was then to characterize the variations of SOM with soil particle size and to establish the predictive models for SOM in the wind erosion region of northern China. A total of 84 soil samples were collected in the wind erosion region of northern China, covering 3 climatic regions, 17 erodible soil types, and 5 land uses. Each sample was sieved into 8 size classes of 0–63, 63–100, 100–150, 150–200, 200–250, 250–300, 300–400 and 400–880 μm using the classical dry sieving method, and the potassium dichromate volumetric method was applied to determine SOM for each class. No matter for which size class, SOM basically decreased from east to west, which was consistent with the spatial distribution trends of precipitation and the normalized difference vegetation index (NDVI). The magnitudes and variations of SOM with particle size, moreover, differed among soil types and land uses. For most of the soil samples, i.e., 63 out of the total 84 (75 %), SOM generally decreased with particle size. Whereas for the remaining samples, SOM tended to increase or remained rather consistent as the particle size increased. The relationships between SOM and particle size highly depended on the climatic factors of precipitation and temperature, as well as the vegetation type and health as indicated in the magnitudes of NDVI. The corresponding models developed exhibited high accuracies for SOM prediction, which were capable of explaining 70–90 % of the total SOM variations. These findings hold important implications for understanding the mechanisms of carbon loss in the wind erosion region of northern China.