| Pinus massoniana is the main timber forest tree shelter forest tree species in southern China, which has the characteristics of strong adaptability, easy reproduction and growth rapidly, has been widely used. In the early1980s, beginning with soil and water conservation in the Yangtze basin, Pinus massoniana became important pioneer tree species for afforestation of mountain in the Yangtze basin. Because it is positive tree species, suitable for high and dry clay and clay soils, can grow in poor gravel and drying the waste mountain. But in a lot of pinus massoniana Forest, the hierarchy is single, the structure is irrational and there were less coverage or no coverage under the surface, with the result that the dammed-up precipitation ability of forest decreased, resistance was reduced, and the ecological and economic benefits droped."Far look green, near the water and soil loss" and "soil and water loss under Forest" are very serious problems.The research of "soil and water loss under Forest" occurrence, mechanism and influence factors research is uncommon.Proceeding from the main control factors of soil erosion--rainfall factor in this article, the author constructed natural slope runoff test plots of Pinus massoniana secondary forest in hilly red soil region of Taihe County and ganxian in jiangxi provice, and studied on the the relationship of runoff, Sediment and rainfall characteristics(Rainfall (P), the maximum rainfall intensity of30minutes (I30). the maximum rainfall intensity of60minutes (I60), rainfall duration (T), prophase total rainfall, rainfall composite factor (PI10, PI30. PI60, PIavc), rainfall kinetic energy composite factor (E, El10, EI30and EI60)) under different configuration mode of the test plots in2009-2011. According to analyses, this study reached conclusions as follows:(1) The relationship of runoff, Sediment and rainfall(P):The correlations between runoff、sediment and rainfall (P) of each plot were significant, whose optimal regression models were mainly the quadratic regression functions, the regression models of individual plots were the power functions and logarithmic functions. the coefficient of determination of bare test plots under the forest was the largest in all the regression models. moreover.The correlations between runoff and sediment of each plot were significant, the regression models were the quadratic regression functions.The correlation coefficient of runoff and sedimnet of plots with surface coverage was larger than them with no surface coverage except the first group of plots(GX101～103).(2) The relationship between runoff, Sediment and prophase total rainfall:In test area of gan county, the correlations between runoff、sediment and prophase total rainfall were poor, through the comparison of similar rainfall cases,we found that the prophase total rainfall influenced the runoff and sediment of the two moderate rain in certain extent. For the heavy rain and the strom, the total rainfall before n (n<4)days impacted greatly on the runoff and sediment of the current rainfall; In test area of taihe county, the correlations of the runoff and the total rainfall before2days were significant, the total rainfall before2-7days had the significant correlations with the runoff in bare under the Pinus massoniana forest as well. In addition the total rainfall before7days had the significant correlations with the runoff in the test plots (Pinus massoniana+Lespedeza); in the sediment of each plot, the total rainfall before1day had only the significant correlations with the sediment in the test plots (Pinus massoniana+Lespedeza) and the test plots(Pinus massoniana+Bahiasgrass), in the other plots, the correlations between the sediment and prophase total rainfall were poor.(3) The relationship between runoff, sediment and rainfall:the correlations between runoff and rainfall duration were significant except GX103,GX201and GX301plots in test area in gan county, the regression models were the quadratic regression functions except GX401and GX301plots; The sediment of each plots had the significant correlations with rainfall duration, the regression models were all the simple cubic regression functions. Accompanied by rainfall duration increased, especially in the typical rain conditions, Increased runoff carried sediment, which increased the amount of sediment, But judging from the regression models, when rainfall reached a certain level, the sediment of plots will had to fall.(4) Typical rainfall runoff characteristics:In test area of gan county, under the same conditions of canopy density, Except to the plots with canopy density of0%, the runoff sorting in the plots were as follows:Pinus massoniana surface exposed> Pinus massoniana+Bahiasgrass>Pinus massoniana+Lespedeza. Also found, along with rainfall increasing, runoff coefficient of the plots as a whole is an increasing trend, especially in strom and cloudburst, but there were also the situation that the runoff coefficient of small rainfall was larger than the big raninfall in the same rainfall type and the runoff coefficient of light rain type was larger than heavy rain. Except to the plots with canopy density of0%and individual plots with canopy density of24%, the sediment sorting in the plots were as follows:Pinus massoniana surface exposed> Pinus massoniana+Lespedeza>Pinus massoniana+Bahiasgrass. In test area of taihe county, the runoff and sediment sorting in the plots were as follows:the naked plot>pure forest of pinus massoniana>lespedeza plot> Bahiasgrass plot.(5) The relationship between runoff, sediment and rainfall intensity:The correlations between runoff and I30, I60were significant, The correlations between sediment and I30,I60were significant. Through prioritizing regression equation, Separately constructed the quadratic regression equation between rainfall intensity and runoff,sediment.(6) The relationship between runoff, sediment and rainfall composite factors:The runoff and rainfall composite factors were significant correlated.among them,the correlation coefficient of runoff and PI30was maximu,the correlation coefficient of runoff and PIave was minimum.The sediment and rainfall composite factors were significant correlated except PIave.the correlation coefficient of sediment and PIave was lower, the correlation coefficient of others were quite. Through simulating linear regression about the runoff and sediment of each plot between the rainfall composite factors constructed the quadratic regression equation.(7) The relationship between sediment and rainfall kinetic energy composite factors:The sediment of each plot was significant correlated with rainfall kinetic energy composite factors, Among them, the correlation coefficient of EI30, EI60and rainfall kinetic energy composite factors were the largest, followed by EI10and E.and then constructed the quadratic regression equation between sediment and rainfall kinetic energy composite factors.(8) Comparison of the effects of soil erosion caused by rainfall in different region:In test area of taihe county, the coefficients of determination of regression equation between rainfall and runoff were higher,which were more than0.87. In test area of gan county, the coefficients of determination of regression equation were between0.57and0.80, smaller compared with the coefficients of determination of taihe test area. The coefficients of determination of regression equation between sediment and rainfall in taihe test area were between0.78and0.90.and the coefficients of determinationbetween them in taihe test area were0.40and0.76,which were smaller than them in taihe. Associated with increased rainfall, the runoff and sediment had a increasing trend, among them in taihe test area, the runoff and sediment of bare plot were maximum,followed by pure pinus massoniana and the runoff and sediment of the plot(Pinus massoniana+Bahiasgrass) were minimum.Howev-er in gan county test area,the runoff increased with the increase of rainfall except GX101plot, the sediment of bare plot had a significant increasing trend, and the other plots’ regularity was not obvious.(9) Using rainfall characteristics data in gan county test area, through stepwise multiple regression method, we constructed the prediction models of the soil and water loss of each plot in gan county. |
PDF Full Text Request