Spatiotemporal variation characteristics of sediment nutrient load from the soil erosion of the Yangtze River Basin of China from 1901 to 2010

Over the past 70 years, the Yangtze River Basin-the third longest river basin in the world-has been under the influence of augmented human activities. Regarding environmental protection and policy design, the total ni-trogen and phosphorus (TN and TP, respectively) loads from soil erosion are objectively and quantitatively critical parameters while assessing the spatio-temporal soil erosion changes over the 1901-2010 period. First, soil erosion in our study area was calculated using the Revised Universal Soil Loss Equation (RUSLE) model. Second, the TN and TP loads from soil erosion were assessed using the constructed Nutrient Loss Empirical Model (NLEM). Third, we conducted spatial autocorrelation analysis, Mann-Kendall (M-K) trend tests, and wavelet analysis of the soil erosion, and related TN and TP loads. At Datong Station over the 1901-2010 period, the average annual TN and TP loads from soil erosion were 1.77 and 0.56 million tonnes, respectively. Moran's I values of the average annual soil erosion, and related TN and TP loads from soil erosion indicates the existence of positive correlations, while clustering was the prevailing spatial distribution pattern. High-high (H-H) cluster areas were mainly evident in high-altitude region of the western and southern Yangtze River Basin, conversely, low-low (L-L) cluster areas were scattered primarily in regions with high population density and intense human activities. Soil erosion increased rapidly around 2001; hence, 2001 was a changing point. According to the M-K test, the time intersection of soil erosion, and related TN and TP loads at Datong Station was around 1990; hence, 1990 may have been a changing point, likely due to the operation of Gezhouba Dam during 1981-1986. This Dam trapped large amounts of sediments. If the sediment load data of control hydrological station of 12 sub-basins is available, SDR, sediment load, AN and AP calculation could be more accurate and interpretable.