Biochar’s potential impact on soil erosion has gained attention, but remained inconsistent, particularly lacks the results from continuous field monitoring under diverse rainfall conditions. This study aimed to explore the long-term variation in soil erosion induced by biochar addition under distinct natural rainfalls. Six runoff plots, treated with different amounts of biochar derived from apple tree branches (0, 1%, 2.5%, 4%, 5.5% and 7%), were established. Over a four-year period from 2017 to 2020, runoff and soil loss from these plots were continuously monitored and analyzed. Among 215 recorded rainfall events, 27 events resulted in runoff and sediment production were classified into three rainfall patterns (R-I, R-II and R-III) using K-means clustering. R-III characterized by high mean rainfall depth (138.80 mm), duration (3443 min), and erosivity (879 MJ.mm (ha. h)(-1)), but infrequent occurrence, followed by R-II and R-I with moderate and low mean depth, duration and erosivity, but higher frequency. Biochar addition, relative to control, resulted in an average annual runoff reduction, ranging from 9% to 36%, with a mean reduction of 28%, and reduced runoff for each rainfall, ranging from 4% to 28%, with a mean reduction of 12%. Similarly, biochar addition also reduced average annual soil loss, ranging from 43% to 79%, with a mean reduction of 61%, and reduced soil loss for each rainfall, ranging from 29% to 74%, with a mean reduction of 52%. The reduction of runoff and soil loss was positively correlated with biochar addition rates. Notably, the soil loss reduction was exceeded that of runoff. Soil erosion was negatively linked to surface area of soil particles (p < 0.05), total carbon, and total organic carbon (p < 0.01), but positively correlated with sand content (p < 0.01) and bulk density (BD) (p < 0.05) under biochar addition. Comparison of the mean reduction of runoff and soil loss among three rainfall patterns, the order was as follows: R-II>R-I>R-III. The optimal controlling of soil erosion was observed in 7% biochar addition under R-II. Runoff-soil loss relationships well-fitted by power equation, showed soil incorporated with biochar treatments exhibited heightened susceptibility to erosion under R-III with high intensity compared to R-II and R-I with moderate or low rainfall intensity. Thus, the findings indicate biochar is a potential soil amendment to improve a degraded or degrading sloping farmland, however, the rainfall characteristics need to be considered when it is applied to the sloping farmland prone to erosion.
