The Hillslope Hydrological Effect Of Vegetation And Simulation In The Small Watershed Of Diediegou, Liupan Mountains

The effects of vegetation types, terrain and soil on hillslope hydrological process were studied in the small watershed of Diediegou at the northwest part of Liupan Moutains. The contents of the dissertation are comprised of analysis upon spatial investigation of soil hydraulic characteristics, vegetation canopy interception, the hydrological function of litter, transpiration of arbor and shrub, infiltration of soil water, soil water variation, water balance in permanent plots, the character of runoff in watershed and the applications of BROOK90 model for scenarios simulation. The main results are as follows:1. The spatial distribution of soil hydraulic characteristics in small watershedThe special distribution of soil depth in watershed presented a middling degree of variation. The soil depth at shady and semi-shade slopes was much thicker than at the sunny and semi-sunny slopes. The soil depth (cm) in different slope positions showed an order as follows: the slope top (34.8) < the up-slope (47.7) < the middle-up slope (75.9) < the middle slope (83.4) < the down-slope (131.2) < the slope foot (286.4). In the small watershed, much of the shrubs were laid at sites with thin soil, while the arbor woodlands were basically laid at site with thick soil.The average of rock fragment content in all soil layers of all 93 soil plots investigated in this study was 5.96%, and the content in diameter range of 2-6mm occupied 62.39% of the gross rock fragment. The rock fragment content (%) in the surface layer (0~20 cm) at different slope aspects showed an order as follows: sunny slope (3.07) > shady slope (2.89)> semi-shady slope (1.99). The rock fragment content (%) in the surface layer at different slope positions showed an order as follows: down-slope (3.46) > middle-up slope (3.05) > slope top (2.46) > middle-slope (2.20) > up-slope (2.08). The rock fragment content increased with soil depth basically in an S-curve.2. The interception of vegetation canopyThe averaged interception ratio in Larix principis-rupprechtti stand was 20.03% in growing period of 2008 (April to October) with a precipitation of 389mm. The averaged interception ratio in May (23.73%) was the highest and that in September (18.12%) was the lowest. The spatial distribution of throughfall ratio under tree canopy was influenced by the variation of leaf area index (LAI). The averaged stem-flow from 5 Larix principis-rupprechtti trees for 18 rainfall events was 0.09mm, accounting for 0.46% of the gross precipitation.In the research period, the throughfall ranged from 1.1 to 42.4mm in Ostryopsis davidiana community plot, and from 0.7mm to 45.0mm in Hippophae rhamnoides community plot. The canopy interception capacity of these 2 shrubs was obtained by using the statistical analysis between throghfall and open-field precipitation over 5mm. The canopy interception capacity was 0.54 mm for Ostryopsis davidiana, and 1.58 mm for Hippophae rhamnoides. The averaged interception ratio of herbage per unit LAI was 0.02 g/cm2 which was measured after one 51.6 mm rainfall event, accounting for 1.64% of the total rainfall depth.3. The hydrological function of litterBy soaking the litters from 4 different vegetation type plots, the saturated water-holding ratio of litter was in the order as follows: Larix principis-rupprechtti (332.2%) > Hippophae rhamnoides (288.6%) > Ostryopsis davidiana (220.6%) > grassland at semi-sunny slope (219.5%), and saturated water-holding capacity of litter in depth was in the order as follows: Larix principis-rupprechtti (2.2mm) > Ostryopsis davidiana (1.3mm) > Hippophae rhamnoides (0.9mm) > grassland at semi-sunny slope (0.6mm).By the field measuring of the water-content variation process of litter in Larix principis-rupprechtti plot during one series of continuous rainfall events, it was obtained that the peak water-holding ratio of litter ranged from 159.9% to 246.6%, the average was 219.4% and the corresponding water-holding capacity of litter in open field was 16.01 t.hm-2, which equals only 74.3% of it obtained by soaking litter.The relation between litter evaporation rate and litter moisture is logarithm. The litter evaporation rate increased with increasing litter moisture, approaching the potential evaporation rate from free water surface. The litter moisture in the Larix principis-rupprechtti plot varied in the range from 6.0% to 192.3% in the research period from Jun. to Oct. in 2008.4. Transpiration of Larix principis-rupprechtti and Hippophae rhamnoidesIn sunny days, the daily variation of sap flow velocity (SFV) of sample trees of Larix principis-rupprechtti was a single-peak curve. The sap flow velocity fluctuated tunelessly in cloudy days. The SFV was dropped down to the night level in rain times and appeared complex curve with much peaks. The peak SFV and the daily averaged SFV in sunny days after rainfall events were higher than the sunny days before rainfall events, and the difference after and before rainfall increased with rising rainfall depth. The peak daily SFV was going to be reduced within continual drought days. The sap flux possesses a seasonal variation, the monthly averaged sap flux of Larix principis-rupprechtti in the growing period of 2008 showed an order as follows: May > June > July > August > September > October.The daily variation curve of sap flux for Hippophae rhamnoides is a single-peaked in sunny days, but double-peaked or complex-peaked in cloudy or rainy days. A night sap flow also existed in Hippophae rhamnoides. The night sap flow can amount to 17.78% of the diurnal volume and the percent in drought period is higher than in wet period.5. Infiltration of soil waterA study on the final infiltration rate was carried out for 4 typical vegetation forms in the small watershed of Didiegou, by using a method of overflowing in slope plots. The final infiltration rate (mm/min) of different vegetation types were: natural grassland (5.33) > shrub of Ostryyopsis davidiana (5.13) > shrub of Hippophae rhamnoides (4.29) > plantation of Larix principis-rupprechtii (3.81). The final infiltration rate was influenced by landform and there was clear difference among slope aspects, slope positions. Taking grassland as an example, the final infiltration rate in different slope aspects showed an order as follows: shady slope (5.40) > semi-shady slope (5.23) > semi-sunny slope (5.12), and the final infiltration rate in different slope positions showed an order as follows: midst slope position (5.92) > lower slope position (5.36) > upper slope position (4.96).6. The variation of soil water condition in permanent plotsThe spatio-temporal difference of soil water condition in the Larix principis-rupprechtii plot was very remarkable. According to the soil water dynamics during growing period in 2008, it can be divided into relatively wetness period (4 May to 21 May), durative consuming period (22 May to 7 Sep) and fast accumulating period (8 Sep to 1 Nov). According to soil water utilizing character, the vertical soil water dynamics with soil depth can be divided into 3 layers: faintness-utilizing layer (0-20cm), mostly-utilizing layer (20-60cm) and adjusting layer (below 60cm).The soil water availability in different plots in growing period of 2008 showed the order as follows: Hippophae rhamnoides shrub > grassland in sunny slope > grassland in semi-sunny slope > Ostryyopsis davidiana shrub > Larix principis-rupprechtii on steep slope. The soil water condition in plots of Larix principis-rupprechtii on steep slope and Ostryyopsis davidiana shrub appeared even an invalidation state from June to September, but other plots were unwieldy state.7. The water balance in permanent plots and the watershed streamflow charactersFor all the plots with different vegetation types in this study, the surface runoff was only about 0.5% of precipitation, and the inter-flow was below 0.5% of precipitation. So they the surface runoff and interflow can be neglected in plot water balance analysis. The results analysis indicated that the evapotranspiration was the biggest item in plot water balance. Grassland (both in sunny and semi-sunny slopes) can be classified into a type of water-yielding, while the Larix principis-rupprechtii plantation (both in steep slope and slow slope) and the Hippophae rhamnoides shrubland in semi-sunny slope can be classified into a type of water-consuming.The seasonal variation of streamflow in the small watershed was very uneven. The streamflow was mainly composed of base flow in initial period of growing season and after November, but composed of fast runoff with peak discharge caused by rainstorms in the rainy season from August to October.8. The applications of BROOK90 model to simulate the hydrological effect of vegetationThe simulation with the calibrated model of BROOK90 showed that the evapotranspiration can amounts to more than 70% of gross precipitation for all the vegetation plots in growing season. The evapotranspiration of 3 vegetation types showed an order as follows: Larix principis-rupprechtii > Hippophae rhamnoides > grassland, but the plot water yield as: grassland > Hippophae rhamnoides > Larix principis-rupprechtii. For all the vegetation plots investigated in this study, the evapotranspiration in sunny slope was higher than in shady slope, but the water yield in sunny slope was less than in shady slope. The effect of slope aspect on plot water yield was different with vegetation types and the magnitude of effect decreased with the order of: grassland > Hippophae rhamnoides > Larix principis-rupprechtii. The steeper slope can yield more water resources than that from gentler slope. The evapotranspiration appeared to be decreased with increasing slope gradient for all plots.Soil depth is very important factor for the evapotranspiration, but this effect varied with vegetation types. The main response to changing soil depth was the soil evaporation in grassland plot, in the plots of shrub and forestland. When the soil depth increased from 20cm to 90cm, the transpiration increased by 37.1% in Larix principis-rupprechtii plot and26.3% in Hippophae rhamnoides plot. The evapotranspiration and water yield from plots decreased with decreasing initial soil water potential. Taking an integrated analysis of the hillslope hydrological effect from vegetation types, slope aspect and soil depth, the suitable vegetation restoration models was discussed for the research areas.