The effects of slope and fertilizer rates on nitrogen losses in runoff from red soil and paddy soil during simulated rainfall

Purpose Runoff and soil erosion reflect the interactions of soil properties and rainfall. However, few researchers have investigated the forms of nitrogen lost, the first flush, and the relationships between nitrogen losses and rainfall, runoff duration, different slope gradients, and fertilizer rates on red soils and paddy soils. Materials and methods This study examined the nitrogen losses (NH4+-N, NO3--N, and TN) in runoff under simulated rainfall conditions with an intensity of 80 mm/h. The slope angles were set at 0°, 5°, and 15° for the red soil, and 0° and 5° for the paddy soil. The fertilizer was applied to the soils at 3 rates, i.e., 60, 180, and 300 kg N/hm2. Results and discussion The results showed that the cumulative rainfall required to generate runoff differed significantly for different slope gradients, and the value for the red soil was 46.46 mm at 0°, which was 3.46 and 4.62 times of the rainfalls at 5° and 15°. The value for the paddy soil was 20.09 mm (average of 0° and 5°), which was half of red soil’s value. Of the TN lost in a 90-min event, 57.25 ± 12.62% was lost in the first 20 min of runoff generation, with the losses decreasing as the runoff generation time increased. The NH4+-N/TN and NO3−-N/TN differed significantly at different fertilizer N levels (P < 0.01). As the fertilizer application rate increased, the NH4+-N/TN tended to increase while the NO3−-N/TN tended to decrease. The NH4+-N/TN decreased exponentially, while NO3−-N/TN increased logarithmically, as the runoff duration time increased. A model, which included parameters to predict the initial losses of precipitation, was established to simulate the processes driving TN losses under different conditions. It is useful to analyze the TN losses from different soil types, slope gradients, and fertilizer rates. Conclusion The threshold slope gradient, first flush effect, and the TN loss process of the red and paddy soil were determined in this study. This information derived can potentially contribute to developing measures for reducing agricultural diffuse pollution and improving resource allocation.