Abstract:
Worldwide, soil erosion is one of the most pressing environmental problems of present times. Particularly, soil erosion triggered by overland flow and runoff seriously affects the productivity and stability of ecosystems. The loss of fertile topsoil and soil's water storage capacity, and the discharge of sediments and associated contamination of waterbodies due to diffuse matter transport of particle-bounded agrochemicals from cropland highly elicit call a for action to combat soil erosion for a future securing of food supply and high drinking water quality. Globally, China belongs to one of those countries most affected by soil erosion. Technical problems as well as high economic off-site damages and costs resulting from reservoir siltation and thus, reduced project's lifespan due to soil erosion are typical for numerous large-scale dam projects in China. In addition to the natural disposition to soil erosion, especially, anthropogenic impacts associated to the dam construction distinctly affect the soil erosion risk potential in the adjacent ecosystems.
This can be exemplarily seen at the currently worldwide largest dam project, the Three Gorges Dam at the Yangtze River in Central China. This megaproject has been controversially discussed since its planning, and most recently since its construction and full operation in 2007. It contains the largest installed hydropower capacity worldwide, and is supposed to distinctly improve the river navigation and to secure the water supply to the northern country in the long-term. The realization of the dam project has already required massive resettlements of rural and urban population of more than one million people long before its start of operation. Additionally, large-scale land use changes, e.g., land reclamation for the road and settlement construction, for small scale subsistence farming and for cash crop production as well as shifts in land uses, on the steep sloping uphill-site above the impounded area are expected to considerably foster the soil erosion in the short- to long-term. Due to their partially direct connection to the stream network agriculturally used land with high soil erosion potential affects the water quality. Precise knowledge on the quality and quantity of soil loss, and its spatial and temporal variability can help to control the soil erosion by developing an adapted land use management and identifying conducive soil conservation measures, such as contour-aligned bench terraces. Under optimum conditions, bench terraces balance the geomorphic settings and anthropogenic use and can present a fair and sound basis for economic growth in mountainous areas.
The focus of the present thesis lies on the risk potential of soil erosion by water in the newly created reservoir of the Three Gorges Dam. Therefore, the central research questions aimed at the natural soil erosion risk potential and the effect of the dam-induced land use dynamics on the dimension and spatial and temporal distribution of soil losses. Due to the data scarcity and limited access to the terrain, a further focus of the research conducted lied on the data-based regionalization of soil erosion factors to use as input in soil erosion modeling.
The research was conducted in the subtropical Xiangxi catchment (3,200 km²) that was considered to adequately represent the Three Gorges Area in terms of physical settings and human interventions attributing to the dam project. The Xiangxi River joins the Yangtze River as a first class tributary approximately 40 km upstream the Three Gorges Dam. Due to the dam construction, the widely terraced landscape of the Xiangxi catchments is also affected by rapid, high land use dynamics with consequences on the slope stability. Particularly, the backwater area in the southern catchment area with the impounded lower reach of the Xiangxi River is characterized by steep to extremely steep sloping terrain and predominantly shallow soils with moderate to very high soil erodibility. Additionally, the very high rainfall erosivity increases the high physical vulnerability of the entire Xiangxi catchment. Between 1987 and 2007, a governmental-driven decrease of arable land and an increase of woodland and shrubland affected the northern headwater zone of the catchment. In the immediate reservoir area, the land use change from 1987 to 2007 was mainly controlled by a distinct conversion of arable land to orange orchards.
Within the framework of this thesis, methods for data survey and data processing were tested and adapted in order to evaluate the risk potential of soil erosion. In addition, comprehensive field investigations focusing on soil erosion processes and on pedological properties and further erosion-relevant factors were conducted. Relevant parameters derived from remote sensing data and land use classifications as well as the documented land use change from 1987 to 2007 were used for the parameterization of the empirical soil erosion model RUSLE. This model was applied to estimate and evaluate the spatial distribution and dynamic of the soil erosion risk potential, and to spatially localize high-risk areas. The new conceptual model TerraCE was developed and tested for the identification and spatial analysis of different terrace conditions and their causes. By means of data mining approaches, a prediction of the spatial distribution of the identified terrace conditions was computed. By integrating environmental and anthropogenic indicators on the impact of the terrain and the human influence, the causes and the strength of disturbances on the terrace conditions, and thus terrace degradation were analyzed.
During the observation period from 1987 to 2007, the Xiangxi catchment is generally characterized by a decrease of average annual soil losses and their maxima due to implemented environmental programs. However, a very high soil erosion risk potential in the entire catchment must be assumed. Frequency and intensity of soil erosion mainly concentrate in the backwater area at the lower reaches of the Xiangxi River. Here, land use changes, resettlements, and infrastructure construction have the highest impact. An inadequate construction of terraces that is not adapted to the local terrain conditions and an insufficient maintenance of the farming terraces can further strongly affect the soil erosion dynamic. Moreover, rapid ecosystem changes and an associated intensification and reclamation of terraces can lead to their degradation. The tempo of the land use dynamics hardly considers available capital and labor for the cost and time-consuming restoration and maintenance of terraces, mainly cultivated with oranges. The high increase of the reclaimed area for the orange production within very short term caused a surplus production and thus, a price decline on the local and regional markets. Due to the not very profitable sale of oranges, a lack of farmers' motivation and little or no capital are made responsible for the gradual worsening of the terrace conditions. As many of the resettled peasants, that were formerly used to farm the flat valley bottoms, are often not familiar with the new and difficult terrain settings and farming techniques, there is also a lack of knowledge on adequate terrace cultivation. Subsequently, inappropriate management of those terraces leads to an increase in the soil erosion
The findings of the present thesis suggest designating the terraces as important, sensitive ecosystem service as they present - if properly maintained - a very effective soil erosion control and enable for a sustainable land use in the mountainous Xiangxi catchment and throughout the entire Three Gores Area. Considering the data scarcity in terms of spatial and temporal resolution, the results further show that soil erosion factors can be successfully regionalized and used for a valid soil erosion modeling. Against the background of ongoing research within the 'Yangtze Project' as well as further projected large dam projects at the Yangtze River and worldwide, the research conducted offers an important starting point for further research on the soil erosion risk potential and associated environmental threats, such as water pollution.