| The water and soil conservation measures should be water-saving and high-efficient for controlling soil erosion but without threatening the water supply safety in the dry Loess Plateau. Therefore, the effects of changes of meteorological factors and land uses on annual runoff and sediment yield should be separated and evaluated, especially in the scale of small watersheds, for guiding the integrated management of watersheds. Therefore, the small watershed of Zhifanggou, located at the Pingliang City of Gansu Province was selected for this study. The long-term monitoring data of annual precipitation, runoff, sediment module, land use, and management measures within 1955-2003 were used for this study. Based on the changing management process, the several decades of soil conservation of the small watershed was divided into 4 periods:1955-1964,1965-1976,1977-1985 and 1986-2003.The main management measures within the first 3 periods were engineering measures such as terrace and check dams; while the last period was characterized by the beginning and increasing vegetation measures including afforestation and grassland restoration. The effects of mankind activities and precipitation change on the annual runoff change were separated. The changes of the relations between annual precipitation and annual runoff and between the annual runoff and sediment yield module with time were analyzed according to the management periods of this small watershed. The multiple linear relations among the annual runoff ratio, sediment yield module and the area ratio of each land use category were fitted through regression analysis. These relations were used to quantitatively evaluate the effects of land use change on the annual runoff and sediment yield of the small watershed, and to compare the efficiency of eco-water used for reduding the sediment yield among the main land use types. The research results showed:With in 1955-2003, the annual precipitation did not have a significant change, but the anuual runoff was reduced significantly, from 59.1 mm in the 1950’s to 12.2 mm in the early beginning of the 21thcentury, a relative reduction of as high as 72.4%. The period of 1955-1964 was chosen as the base period for comparison. It showed that besides the obvious contribution of annual precipitation decrease to the annual runoff reduction in dryer years, the annual runoff variation was mainly caused by the mankind activities, and the absolute contribution from mankind activities showed an increasing tendency.Accompanied with the increasing management activities which caused a remarkable reduction of annual runoff, the sediment yield module of the small watershed was also decreased obviously within 1955-2003. Compared with the base period of 1955-1964, the dependence of annual runoff of the small watershed on annual precipitation had been decreasing because of the increasing intensity of water and soil conservation. This is related with the changes of capacities of both runoff generation and runoff interception of the small watershed. The runoff generation capacity was increased in the 2nd and 3rd periods due to the vegetation destruction, but greatly decreased in the 4th period due to vegetation restoration. The runoff interception capacity increased by engineering measures experienced a process of "increase-decrease-disappearance" with the construction and function degradation of terraces and check dams.To exclude the disturbance from annual runoff variation, the relations between annual sediment yield module and annual runoff were analyzed in each period. Within the first 3 periods, the correlation degree between runoff and sediment yield did not changed obviously, indicating the engineering measures caused a significant reduction of runoff but not the sediment yield ability. However, the sediment yield ability was reduced a lot in the last period due to the vegetation increase and runoff decrease.The multiple and linear relation between the annual runoff, annual precipitation and the area ratio of each land use category, and the relation among the sediment yield module and the annual runoff and the area ratio of each land use category was fitted. The coefficients in these relations indicated that the 2 land use categories of forestland, terrace and dam fields will decrease the annual runoff, and the decreasing intensity of forestland was 1.48 times of that of terrace and dam fields. The runoff was mainly generated from the 2 land use categories of grassland, slope farmland and others. The runoff ratio from grassland was 1.8 times of that from slope farmland and others. The forestland and grassland showed a function of reducing the sediment yield, in which the reducing intensity of forestland was 2.5 time of that of grassland. The sediment was mainly from the 2 land use categories of terrace and dam fields, slope farmland and others, in which the contribution to sediment yield from terrace and dam field was doubled high as that from slope farmland and others. If the slope farmland and others is converted to forestland, grassland, terrace and dam field in an intensity of 10% of the total area of watershed, the annual runoff will be changed by-9.6,+4.2,-8.1 mm; and the sediment yield module will be changed by-6367.0,-2532.4,+300.5 t/km2/a, respectively.The effects of vegetation measures (afforestation, grassland restoration) for water and soil conservation is expected to be longer than that of engineering measures (terrace and check dams). The forestland can well reduce the sediment yield module, but consumes much more water, although its water use efficiency for reducing sediment yield is the highest among the 4 land use categories. The grassland has lower sediment yield reduction and water use efficiency compared with forestland, but it can increase or maintain the water yielding ability of watersheds. The terrace and dam field has also an effect of reducing runoff, which is just slightly lower than the forestland. The effect of terrace and dam filed on sediment is mainly expressed by the sediment interception, rather than erosion prevention. However, the sediment interception ability will gradually lower or even disappear with the silt up of check dams and the degradation of terraces. Moreover, the water use efficiency of terrace and dam fields for reducing sediment is the lowest one among the 4 land use categories. In summary, to meet the requirements of water-saving and high-efficiency in soil erosion control in the dryland regions, the most suitable measures should be the grassland restoration. In this way, we can get a long-lasting service of sediment reduction and erosion control, an unreduced or even enhanced water yield, and higher eco-water use efficiency in erosion control. |
