| Considering and optimazing the interrelationship between forest/vegetation and water resources is one of the important subjects for ecological reconstruction and vegetation restoration in the semi-arid zone of Northwest China. The Liupanshan Mountains as an important water-retention area is representative for mountain ecosystem in the semi-arid zone. Studying the hydrological effect in the Liupanshan Mountains is very crucial for understanding the interrelation between forest/vegetation and water resources, and for providing a theoretic basis for ecological reconstruction and vegetation restoration in semi-arid zone. This dissertation focused on the hillslope hydrological effect of forest/vegetation in the Diediegou Catchment, which lied in the north side of Liupanshan Mountains, and converting farmland to forest was started here in 2000. The main research results are shown as following: 1. Background information of the Diediegou Catchment Diediegou Catchment is located in the City of Guyuan of Ningxia of China (E106°09′-106°30′, N35°15′～35°41′), with an area of 25.4 km~2, elevation variation of 1795-2615m asl. The valley lies in a direction from south to north. This catchment can be divided into 3 sections according to its landform. The area of the upper, middle and lower section is 7.2, 11.3, 6.9 km~2 respectively. In the upper section, the slope is smaller and within 20o, the soil types are loess, old red earth and culticated soil, the vegetation is mainly crops; In the middle and lower section, the slope is steeper and distributed within the range of 10-35o and 20-40o respectively, the soil type is mainly gray-cinnamonic, the vegetation includes grassland, shrubs and Larix principis-rupprechtii plantation. The slope degree has an effect on the land use and soil thickness, with shallower gray-cinnamonic soil directly above bedrock on steep slopes and mountain ridge, while loess (or no loess) and slope deposit underling gray-cinnamonic soil one the site with thicker soil. The propotion of grassland in the catchment is the biggest one with a value of 58.7% and distributed on any landforms, other vegetation types are inlayed in grassland. Since the mankind's disturbance, the zonal typical grassland was degraded and instead of the grassland mainly composed of Artemisa vestita and Artemisia giraldii. After mountain closure, the grassland has been gradually recovering, and the typical zonal grass species such as Bothriochloa ischaemum and Stipa bungeana Trin. have appeared. The area ratio of schrub is 18.4%, among which 2.7% is planted Hippophae rhamnoides, 0.8% of natural Hippophae rhamnoides, 1.4% of natural Ostryopsis davidiana, and 13.5% of natural mixed shrubs. The farmland mainly locaded in the upper section covers 15.4% of the catchment. The area ratio of forest now amounts to only 6.1%, of which 3.9% of Larix principis-rupprechtii, 2.0% of planted poplar, 0.2% of willow. The residential area ratio is 1.5%, the smallest one of landuse form. 2.The main characters of plantation and soil physics in the Diediegou Catchment The soil physical properties in depth of 40-60 cm were measured and compared among the main soil types in the upper section of the catchment. The bulk density is 1.11, 1.15, 1.29, 1.57 g/cm3 for the cultivated loess (converted to forest), loess (grassland), cultivated loess (still as cropland), old red soil (converted to forest) respectively. The corresponding total porosity is 60.17%, 56.36%, 50.00% and 39.59%. The corresponding capillary porosity is 52.74%, 51.33%, 47.17% and 36.34%. The corresponding non-capillary porosity is 7.43%, 5.03%, 2.83%, and 3.25%. The final infiltration rate is 1.4, 0.9, 0.8, 0.2 mm/min. The soil physical characters decrease from cultivated loess, loess to old red soil, but converting to forest can improve the soil physical characters. In the lower section of the catchment, 12 plots of grassland, 8 plots of differend schrubs, 5 plots of Larix principis-rupprechtii were set up and the vegetation characters were measured in Aug.-Sep. 2004. 1) Grassland: the averaged aboveground biomass is 822 g/m2. The biomass is affected by slope aspect and position, in generally, higher on shade slopes than sunny slopes, and decreasing from upper, middel to lower part of slope. There is no clear diffeence of ground coverage among the grassland plots with different slope positions, but a difference exists between slope positions, 95% for slopes of shade, half-shade and ridge, while 85-90% for sunny and half-sunny slopes. The height of grass is affected by slope position and species. It varies between 10-30 cm on upper slope and ridge, 10-40 cm on middle slope, and 20-45 cm on lower slope. 2) Shrubs: the plot of Hippophae rhamnoides has a density of 1300 stems/ha, a height of 2.3m, a canopy density of 0.5, and an aboveground biomass of 0.96 kg/m2. The 3 plots of Ostryopsis davidiana have an averaged density of 50.7 stems/m2, a height of 1.01 m, a canopy density of 0.86, and an aboveground biomass of 2.05 kg/m2. The plot of Spiraea sp. has a density of 4.8 stems/m2, a height of 1.6m, a canopy density of 0.9, and an aboveground biomass of 2.24 kg/m2. The plot of Corylus heterophylla has a density of 8.6 setms/m2, a height of 2m, a canopy density of 0.95, and an aboveground biomass of 2.47 kg/m2. 3) Plantation of Larix principis-rupprechtii: The averaged tree hight of 5 plots on shade or half-shade slopes varies between 4.8-7.7 m, the averaged DBH between 5.6-8.3cm, the tree density between 950-1975 poles/hm2, and canopy density of 0.5-0.7. Soil physical parameters, such as bulk density and porosity, were investigated in the middle and lower sections of the catchment. There is no clear difference betweenslope aspect and position, but they are different with soil depth and vegetation types. 1) Grassland: The averaged bulk density of 12 plots for the soil layers of 0-20cm, 20-40cm, 40-60cm, >60cm is 1.00,1.20,1.18,1.27 g/cm3 respectively. The corresponding averaged total porosity (%) is 60.01, 53.84, 54.46, 51.79. The corresponding averaged capillary porosity (%) is 53.19, 45.97, 46.32, 39.69. The corresponding averaged non-capillary porosity (%) is 6.83, 7.87, 8.14, 12.1 (8.03). 2) Shrubs: The averaged bulk density of 6 plots for soil layers of 0-20cm, 20-40cm, 30-50cm, >60cm is 0.92, 0.92, 1.17, 1.33 g/cm3 respectively. The corresponding total porosity (%) is 64.45, 63.80, 55.52, 46.37. The corresponding capillary porosity (%) is 55.54, 58.67, 48.94, 40.76. The corresponding non-capillary porosity (%) is 9.21, 5.13, 6.58, 5.61. 3) Plantation of Larix principis-rupprechtii: The averaged bulk density for soil layers of 0-20cm, 20-40cm, 40-60cm, >60cm is 0.94, 1.02, 1.04, 1.12 g/cm3 respectively. The corresponding total porosity (%) is 60.99, 57.31, 56.21, 54.35. The corresponding capillary porosity (%) is 50.69, 51.01, 47.74, 44.90. The corresponding non-capillary porosity (%) is 10.30, 6.30, 8.47, 9.44 respectively. In summary, the soil depth being improved by vegetation decreases from forestland, shrubland to grassland, while the soil physical properties decline from shrubsland, forest land, to grassland. 3. Canopy interception and evapotranspiration of Larix plantation A plantation stand of Larix principis-rupprechtii with an age of 20 years on shade slope was selected to measure the canopy interception and water cycling. Canopy interception was observed in the time periods of Aug. to Oct. 2003 and May to Oct. 2004. The interception ratio was 9.8% for the period of June to Sep. 2004, when the bulk rainfall outsite of forest amonted to 408.1mm. A canopy interception model was developed. This model represents the interception process by running with small time steps. It also considers the effects on canopy interception capacity and evaporation during rainfall by the surface area of trunk and leaves, and the dryness of canopy and trunk. The outputs of this model are free throughfall, canopy drainage, stemflow, rainwater held on canopy and trunk, rainwater evaporated during rainfall event. So the process of canopy interception can be reproduced. The model parameters were determined, with the freely throughfall coefficient of 0.28, canopy interception capacity of 1.4mm, trunk interception capacity of 0.05mm, stemflow ratio of 0.09, leaf aera index of 4.19, and trunk area index of 0.2. The interception process of 34 individual rainfall events from Aug. to Oct. 2003 and from May to Sep. 2004 was simulated using this model with a time step of 10 seconds. The simulation results were shown well. According to the simulated calculation, the stemflow is very small with a stemflow ratio of only 0.3% of rainfall. Evaporation during rainfall event is one relative important fraction (26.6%) of canopy interception. However, the structure of interception model is required to be further improved for more detailed consideration of interception process. By using a sap-flow meter, the transpiration on individual larix trees was measured during June –Sep. 2004. It shows that, 1) The sap-flow density curve in a day is bimodal, and higher in south side than north side of trunk. Sap-flow in night was often observed, but its size varies with weather. However, the night sap-flow is notneglectable in calculating daily transpiration. 2) The transpiration of the stand from June to Sep. was calculated by upscalling from single trees to stand, based on the relation between transpiration rate and DBH of individual tree. The transpiration amounts to 239.8 mm, it means an averaged daily transpiration of 1.97mm in growing season. The monthly transpiration was 61.3, 56.7, 63.3, 58.5 mm for June, July, August and September respectively, i.e. corresponding daily transpiration of 2.04, 1.83, 2.04, 1.95 mm. The averaged daily evapotranpiration of forest floor from June to September was 1.64mm measured by Self-made lysimeters. For the period from June to September, the total evapotranspiration of this stand amounted to 474.1mm, with a daily average of 3.89 mm. Of which, the biggest fraction is transpiration of trees (50.58%), the second is evapotranspiration from floor (41.19%), and the smallest is canopy interception (8.23%). 4.Evaporation of Hippophae and evapotranspiration of ground vegetation A plot of natural Hippophae rhamnoides on shade slope was set up for measuing the transpiration by a special type of sap-flow meter of Dynagage, which was designed for shrubs with thin stem, during June to September 2004. Self-made lysimeters were used for measuring the evapotranspitation from floor and from grassland. It shows that: 1) Night sap-flow also exists in Hippophae plants, and varies with weather. The night sap-flow is not neglectable and can amount to 12.87 -13.84% of the total daily volume. 2) The daily sap-flow curve for Hippophae is unimodal. 3) The tranpiration of this Hippophae community during June to September was estimated as 164.8mm based on upscalling calculation with LAI, i.e. a daily transpiation of 1.35 mm, which is lower than that of Larix trees. The monthly transpiration was 52.2, 50.3, 32.1, 30.2 mm for June, July, August, September respectivey, with daily transpiration of 1.74, 1.62, 1.04, 1.01 mm. 4) During the period from later July to beginning October, the averaged daily evapotranspiration from floor of the Hippophae community was 1.43 mm, lower than that of Larix plantation floor. 5) During the period from later June to beginning October, the averaged daily evapotranspiration of 4 grassland plots was 2.30 mm, lower than that from shrubs (Hippophae rhamnoides) and trees (Larix principis-rupprechtii). 5.Comparision of soil moisture on hillslopes The soil water measuring instrument of TRIME was used to monitor the soil moisture dynamics of some permanent plots. It shows that: the soil water content in upper soil layer varies more strongly, the affected depth of soil water by Larix and Hippophae is deeper than grass vegetation. Soil auger was used to measuring soil water content of many sites with different landform and vegetation in the catchment. A clear difference exists amoung slope aspects, decreasing from sunny, half-sunny, half-shade, and shade slope. The soil is wetter in valley than on ridge. The soil moisture increases from upper to lower position on a sunny slope. The profile averaged soil water content varies with vegetation type, in the order of Hippophae > Corylus mandshurica > Ostryopsis davidiana (shade slope) > grassland (shade slope) > Larix plantation > Spiraea sp. > farmland >grassland (sunny slope) > Ostryopsis davidiana(sunny slope). For the soil layer of 0-20 cm or 0-40cm, the order of soil water content is Hippophae rhamnoides > Corylus mandshurica > Ostryopsis davidiana (shade slope) > Larix > grassland (shade slope) > Spiraea sp. > farmland > grassland (sunny slope) > Ostryopsis davidiana (sunny slope) 6.Water balance of permanent hillslope plots During the growing season (21 June to 3. Octomber) in 2004, no surface runoff in an important quantity was observed on all 5 hillslope plots, even in the extreme rainfall event with a daily rainfall of more than 100 mm. After integrating the research results of interception, evapotranpiration, soil water variation etc., the water balance of soil layer of 0-90 cm was analysed for the 5 permanent hillslope plots. With the background of a rich rainfall year in 2003, the 5 plots were classied according to their potential contribution to water yield of catchment by their water balance (export or import; or in other words: leaching or absorbing) in the growing season in 2004 (21 June to 3. Octomber). Grassland (both in shade and sunny slope), shrub land (Hippophae in half-shade slope) can be classified into a type of water-yielding, with a water leaching from the soil layer of 161.5, 120.1 and24.1mm respectively. The forest land (Larix plantation) can be a type of water-balancing (Larix in the steep shade slope) or water-consuming (Larix on the gentle slope foot, where a water inflow from upper slope is possible), the corresponding water balance is 3.9 and -57.9mm. 7.Some character of stream runoff in the Diediegou Catchment A certain effect of water evaporation from stream on the daily stream runoff volume was observed in the days with sunshine. The discharge was decreased during day-time and recovered in night. The runoff loss due to stream evaporation in sunny days was estimated as 2.87-4.78% of stream runoff volume during June to September 2004. The streamflow in rainy season was gradually increasing with accumulated rainfall amount. There are two main contributors for streamflow. The first one is the direct runoff generated on the saturated area or water area along stream, they are the absolute most part of peakflow. The second one is the base flow composed of deep interflow or the flow from bedrock cranny. The stream runoff observed is mainly composed by base flow, therefore the month with highest runoff volume was delaied about 2 monthes than the month with highest rianfall depth. The steamflow is affected by rainfall event both in the same year and in the last year. 8 . Simulation of hillslope hydrological process and scenarios analysis of hydrological effect of different vegetation The lumped hydrological model BROOK90 was used to simulate the water balance of Larix principis-rupprechtii plantation (steep shade slope), Hippophae rhamnoides community (half-shade slope), grassland (sunny slope), and grassland (half-shade slope). The model parameters were determined mainly based fitting the observed and simulated soil water content. The calibration of moel shows an acceptable high accuracy of simulation. Using the determined parameters, thehydrological effect of different vegetation on each plot is simulated. It was concluded that, the evapotranspiration of Larix principis-rupprechtii plantation and Hippophae rhamnoides community is higher than grassland, this means more water resource is required for maintaining their stabilization.
|
PDF Full Text Request