Abstract:Soil aggregation and transformation play an important role in soil erosion and vice versa. The quantifiably breakdown and transformation pathways of soil aggregate have not been clarified. The limitation hampers the in-depth study of sediment separation and transport mechanism of soil erosion processes. Recently, 137Cs, 210Pb, 7Be and magnetite tracer methods have been widely applied to soil erosion research at slope and small watershed scales. However, those methods are not suitable for tracking multiple erosion processes simultaneously. Six typical agricultural soils with different texture from Loess Plateau and black soil region in China were selected in this study. Rare earth elements (REE) tracer method was applied to quantify the transformation pathways of five soil aggregate fractions (5~2, 2~1, 1~0.5, 0.5~0.25, < 0.25 mm). The adsorption and desorption capacity of REE on different aggregates were analyzed with the simulated runoff disturbance cycles. Results showed that the actual adsorption concentration of REE and soil aggregates were lower than that of the application concentration. REE adsorption concentration of soil aggregates at 2~1, 1~0.5, 0.5~0.25 and < 0.25 mm were significantly positively correlated with the clay content (P<0.05). The effect of runoff disturbance on the desorption of REE adsorbed on soil aggregates was very weak, and the desorption concentration only accounted for 0.001%~0.139% of the actual adsorption concentration of REE. The transformation pathways among the 5~2, 2~1, 1~0.5, 0.5~0.25 and < 0.25 mm soil aggregates fractions were basically the same, converting to < 0.25 mm microaggregates was the main pathway. The transformation rate of soil aggregates with high sand content to 1~0.5 and 0.5~0.25 mm fractions was generally lower, compared with soil aggregates with high content of silt and clay. The REE tracer method underestimated the labeled > 0.25 mm aggregates, which ranged from -27.96% to -11.08%. However, < 0.25 mm aggregates was overestimated by 3.65%~22.73%. Based on the REE quantification values of soil aggregates of various particle sizes, a correction relationship was established, which could reduce the calculated relative error to 0.04%~16.24%.