Nutrient loss from slope cropland to water in the riparian zone of the Three Gorges Reservoir: Process, pathway, and flux

Abstract

Eutrophication of the surface water in the Three Gorges Reservoir (TGR) area in response to anthropogenically derived nutrients is a serious concern. Widespread cultivation in the riparian zone could lead to a large amount of nitrogen and phosphorus being transported via runoff during the cropland cultivation period or being released during the submerging period into the water of the TGR. However, little is known about the nutrient release fluxes from slope cropland into water during the submerging period. It is also not clear which process is critical. The experimental site in this study is representative of arable land on slopes in the riparian zone in the TGR area. Free-drainage lysimeters were constructed on a block of intact sandstone at a water level of 165 m in the riparian zone in the TGR area. The treatments were as follows: (1) no fertilizer (NF), (2) reduced nitrogen fertilizer treatment (RFT), and (3) conventional fertilizer treatment (CFT). The experimental plots were cropped conventionally with summer maize (Zea mays L.) from March to early September, and the experimental plots were submerged in water from October to March of the next year. Runoff water samples were collected during the 3-year experimental period (March 2016-October 2018). The results from this study indicate that nitrogen losses via interflow and phosphorus losses via overland flow are the dominant processes in the slope cropland in the riparian zone in the TGR area during the cropland cultivation period. In the CFT treatment, the average annual total nitrogen (TN) loss fluxes were 4.28 ± 0.10 kg hm−2 via overland flow and 26.17 ± 3.68 kg hm−2 via interflow. Additionally, the average annual total phosphorus (TP) loss fluxes in the conventional fertilizer treatment were 0.83 ± 0.43 kg hm−2 via overland flow and 0.15 ± 0.02 kg hm−2 via interflow. Before submerging, there were significant seasonal changes in the nutrient loss flux in every treatment. During the submerging period, the NO3--N, NH4+-N, and available phosphorus release fluxes from soil to water in the CFT treatment were 5.40 ± 0.75 kg hm−2, 4.38 ± 3.30 kg hm−2, and 12.09 ± 11.08 kg hm−2, respectively. The NO3--N release fluxes during the submerging period were higher than the NO3--N loss fluxes via overland flow, but they were lower than the NO3--N loss fluxes via interflow. Different from the pattern of NO3--N, the NH4+-N and available phosphorus release fluxes in the three treatments were much higher than the sum of the NH4+-N and available phosphorus loss fluxes via overland flow and interflow. Based on these findings, our results suggest that efforts to avoid further eutrophication of the reservoir water should take into account nutrient release from the soil of arable croplands in the riparian zone during the submerging period.