| The Qinling Mountains region is a major water conservation area in China, flourishingmany streams and serving as headwaters for several important rivers. Conservation practiceshave led to great improvement in forest conditions. Some stands have reached near-matureand mature stage with substantial change in species composition, stand structure and functionin contrast to young stands. To understand the ecohydrological functions and processes offorest ecosystem at different growth stages, the naturally generated Pinus tabulaeformisforest stands located in the middle elevation of the Qinling Mountains were selected for thisstudy. The research approaches included field observations, controlled lab experiment andmodel simulation. Specifically, the microclimate was measured by micrometeorologicalstations, the ecohydrological functions of canopy layer, litter layer and soil were studied toprovide evidence for management of this important water conservation forest. The mainresults are as follows:(1) The mean annual air temperature and relative air humidity inside forest were higherthan those outside forest for the observation period. The prevailing winds were fromNortheast or south. For a given micrometeorological factor, the variation was usually smallerunder canopy than that in the open area. There were interactive influences betweenmeteorological factors of outside and inside forest. The annual downwelling radiation fluxwere14849.89MJ/m2、14272.99MJ/m2and14454.91MJ/m2, and the annual upwellingradiation flux were11902.17MJ/m2、11751.33MJ/m2and11732.89MJ/m2, respectively,from2006to2008.(2) The annual PAR5(Photosynthetically active radiation, PAR), PAR4, PAR3, PAR2, andPAR1of2007and2008were7604.68and7939.77mol·m-2,1436.70and1389.98mol·m-2,1503.42and1392.12mol·m-2,1000.68and917.11mol·m-2, respectively. The RPFD3(Relative photosynthesis photon flux density, RPFD), RPFD2and RPFD1of2007and2008were18.89%and17.51%、19.77and17.53%、13.16%and11.55%, respectively. Theintensity of daily PAR under canopy reduced gradually as crown thickness increased. Theaverage PAR values of growing season were higher than non-growing season. (3) There was a discrete separation between wet and dry seasons, with the wet seasonincluding months from May though October and dry season from November though April ofcoming year. Low-intensity rainfall events were frequent, but accounting for a small amountof the annual total rainfall. Mean annual rainfall interception ratios were32.9%,33.0%and33.6%for2006,2007and2008, respectively. Canopy interception was a dynamic process,interception ratios varied with rainfall size class and rainfall intensity. Interception ratiodecreased when the rainfall size class or rainfall intensity increased. The higher the meaninterception rate was, the shorter it took the canopy to saturate. The ability of canopy tointercept rainfall decreased while canopy approached its storage capacity.(4) Based on direct measurement of100rainfall events, the gross precipitation,throughfall, stemflow, and interception loss were1576.4,982.9,69.5, and524.0mm,respectively. Rainfall was partitioned as follows:62.4%throughfall,4.4%stemflow and33.2%canopy interception. The redistribution of precipitation by forest canopy was closelyrelated to rainfall size classes. For a discrete rainfall event, the stemflow and throughfallratios were positively related to its size class. A lag existed between rainfall onset time andthe production of throughfall and stemflow. This time lag decreased with increased rainfallclasses.(5) A positive correlation existed between the dependent variable throughfall and eachof the following independent variables-rainfall, the interaction terms of rainfall and rainfallduration, and relative humidity. A negative correlation existed between throughfall and therainfall duration, also between throughfall and temperature. The stemflow was positivelycorrelated with rainfall, interaction terms of rainfall and temperature, negatively correlatedwith air temperature. There were positive correlations between interception loss and rainfall,rainfall duration, and temperature. There were negative correlations between canopyinterception loss and the interaction terms of rainfall with rainfall duration, air temperature.Non-significant correlations were found between throughfall, stemflow, and interception losswith antecedent dry time, wind direction, wind speed, photosynthetic available radiation, andnet radiation, respectively.(6) The direct measured canopy interception ratio was33.2%, the estimated canopyinterception ratio was35.9%from the Gash model and53.6%from the Fan model, and themodeled values were higher than the measured values. The differences between measuredand modeled values from the Gash model ranged from-0.3to+7.1%for different rainfallamounts and from+1.9to+3.2%for different years. The Fan model satisfactorily simulatedinterception for large rainfall events (>50mm) with differences from-3.4to+1.3%, butsubstantially overestimated interception loss for smaller rainfall events (+21.2to+37.2%). The differences between measured and modeled values from the Fan model ranged from+15.4to+24.1%for different years. The Gash analytical model adequately simulated thecanopy interception of Pinus tabulaeformis forest. The parameterized Fan model comparedfavorably to the Gash model in simplicity but not in precision, especially for the smallerrainfall classes.(7) The total thickness and amount of litter layer were9.3±2.8cm and27.94±9.81t/hm2for P. tabulaeformis and8.3±3.6cm and16.04±3.60t/hm2for Q. aliena var. acuteserrata,respectively. The amount of semi-decomposed layer was higher than non-decomposed layerfor both forests. Litter water-holding function was mainly reflected in the early stage of therainfall. A negative correlation existed between water-absorbing rate and soaking time.Water-absorbing rate declined and tended to smooth gradually for semi-decomposed andnon-decomposed layer when the time of immersions increased. The water-absorbing rate ofsemi-decomposed layer was higher than non-decomposed layer. Maximum water-holdingcapacity, interception volume, and effective interception volume were high insemi-decomposed layer than non-decomposed layer for both forest stands. For a given metric,it was higher for P. tabulaeformis stand than Q. aliena var. acuteserrata stand. Thenon-capillary water-holding capacity, capillary water-holding capacity, and maximumwater-holding capacity of soil layer (0-60cm) for P. tabulaeformis were919.66t/hm2、1667.33t/hm2and2586.98t/hm2, respectively and614.61t/hm2、1416.16t/hm2and2030.76t/hm2for Q. aliena var. acuteserrata stands, respectively. The water-holding capacity of soillayer (0-60cm) was higher for P. tabulaeformis stand than Q. aliena var. acuteserrata stand. |
PDF Full Text Request