| The water cycle is a key ecosystem function that links other processes in forest ecosystems. Meanwhile, hydrological processes are difficult to study because they are influenced by a myriad of biophysical factors. Forest hydrology, an important science addressing the relationship between forest ecosystems and hydrology, has received significant attention globally.Evergreen broad-leaved forest (EBLF), covering a lot of area in China, is the zonal vegetation type in subtropical region. However, under long-term human disturbances, this forest is shifting to include much more degraded area dominated by secondary forests, shrubs and human plantation. Although succession could be expected to influence forest hydrology, there are a few reported studies. An important need for such studies is the influence of global warming, in that the vegetation climax may change, thus affecting the hydrology of large regions. The utilization of a chronosequence of forest is a common practice for study ecosystem dynamics. A major advantage of this technique in forestry is that it provides data on long term changes in the structure and function of the ecosystem within a reasonable time.To understand the impacts of the succession of EBLF on the hydrology, we chose mature EBLF as the reference climax forest, secondary forest, and shrubs to represent three different successional stages of EBLFs. The experiments studied the eco-hydrological features of canopy, litter and soil in the three types of forests by the method of the positioning observation in the field. Also, the research studied the chemistry characteristics of different components of rainfall partitioning. The results of the present study could provide theoretical proofs for natural forests protection at Tiantong as well as technological supports for local forest management.Main research results were following:(1) Coefficient variations (CVs) of throughfall decreased along the increasing amount of rainfall under different successional stages. The number of gauges required to estimate throughfall are calculated using two methods.1) Based on the Kimmins equation, a sample size of 16 at shrub,5 at secondary forest, and 4 at climax forest would be enough for throughfall estimates at an accepted error of 10% of 0.1 significance level, respectively.2) Monte Carlo resampling approach was used to find mean values and confidence intervals of 90 and 95% of a variable number of collectors ranging from 2 to 24. For the whole data set, with 9 collectors at shrub,5 at secondary forest, and 5 at climax forest, the error in the mean individually throughfall is around 10%, respectively. This error is reduced to 5% when using 16,10 and 10 collectors at shrub, secondary and climax forests, respectively. Therefore, we concluded that the 25 of collectors used in present study were sufficient to estimate the throughfall value at an accepted error of 10% at 90 and 95% confidence levels, even for those small rainfalls in eastern China.(2) Throughfall inside forest accounted for 55-77% of rainfall among three successional stages of EBLFs. There was an obvious correlation between throughfall and rainfall with equation of TF= a+b·PG (P<0.01).Stemflow was 9,14 and 6% of gross precipitation in shrub, secondary and climax forests, respectively. Results showed that stemflow had significant linear relationships with gross precipitation at each successional stage of EBLFs (P<0.01).The orders of canopy interception rates were:Secondary forest (31%)> shrub (25) > climax forest (17%). The results proved that canopy interception (mm) was found to be best represented as a power function of incident precipitation (mm) in each forest type.(3) The storage of litter layer was 6.42,6.45, and 5.71 t·hm-2 at shrub, secondary forest, and climax forest,respectively. The water holding capacity of shrub, secondary forest and climax forest were 13.38,11.93, and 9.39 t·hm-2, and the corresponding effective water holding capacity of litter layer in these three types were 8.09,6.64 and 5.03 t·hm-2, respectively. During the process of water holding of litter layer, the water holding capacity and absoption speed of the first 2 h were superior to the rest of time. The equations of water holding capacity and water absorption speed with immersed time are O= a·ln t+b and V= k·tn.(4) The average soil bulk density was 1.17 g·cm-3 at shrub, which was greater than that of secondary forst (0.99 g·cm-3) and climax forest (0.99 g·cm-3). The soil bulk density increased along the deeper of the soil layer, but soil porosity and soil non-capillary porosity were both decreased. The water holding capacity of soil (0-30 cm) was at the range of 131.66 to 145.02 mm, which was highest at climax forest and lowest at secondary forest, respectively. The sum of the water storage capacity of the soil were:shrub (16.43 mm)> climax forest (12.61 mm)> secondary forest (12.59 mm).The seasonal variation of soil moisture can be divided into moisture accumulation (November 2008-April 2009) and moisture consumption (May October 2009) period. Atmospheric precipitation is the main control of soil moisture content.The soil infiltration characteristics of the three successional chronosequence of EBLFs appeared consistent regularity, which showed that the infiltration rate decreased along with the increase depth of soil. The experienced formula of Kostiakov and Horton were used to simulate the water infiltration processes of soil well.(5) Water chemistry of throughfall and stemflow were changed after the rainfall passage through the canopy. pHs of throughfall and stemflow, which had lower pHs than gross precipitation, varied significantly along with seral chronosequence on a rain-event basis. EC and ion concentrations of throughfall and stemflow both increased along successional stages. The solute inputs in stemflow were lower than that of throughfall, because the volume of stemflow only accounted for 8%-25% of throughfall. The fluxes of K and Mg in net precipitation (throughfall+stemflow) were consistently higher than that in rainfall. The other three nutrients fluxes (Ca, Na, S) in gross precipitation were absorbed by canopy surface and stem after passage through the forest. The nutrient fluxes of stemflow were lower than that of throughfall, due to the amount of stemflow only accounted for 8-25% of throughfall. |
PDF Full Text Request