تحلیل میزان فرسایش و رسوب ناشی از رخساره‌های فرسایشی حوضه کاخک با مدل شبیه‌ساز باران

Analysis of Soil erosion and sediment yield from erosion facies of Kakhk watershed under rainfall simulation model

Extended Abstract

Introduction
Soil physico-chemical properties has a important impact on soil erosion. Shaly originated soils due to the high susceptibility to erosion have high erosion rates in spite of occupying relatively small areas, can make disproportionate contributions to watershed scale sediment budgets. Critical source areas are usually associated with marls, clay rocks, mudstones and shales. Additionally, few reports showed that badland landforms there are on sands or poorly consolidated sandstones.
Rainfall simulation is a good method for comparison and quantification of different runoff and erosion processes and factors that influence them. Numerous researchers have used simulated rainfall experiments on a wide range for determination of soil erodibility.
The erodible lithologies (shalls) include more than 50 percent of the area of the kakhk watershed basin. Securitizing available literatures about effective factors on soil erosion in eroded soils shows that in spite of numerous reports on different soil erosion processes, little comparative study has been considered on sediment yield originated from soils with different parent material in plot scale under different rainfall intensities. So, there is a need for more detailed investigation on soil physico-chemical and vegetation properties that effect on soil erosion. Accordingly, the present study was carried out to comprehensively compare the effects of environmental factors and rainfall intensities controlling spatial variation in soil loss in kakhk drainage watershed.

Methodology
Study Area
This research has done in kakhk watershed (3720 ha) which located in the south of Khorasan Razavi province. In this area, half of the area is made up of shale lithologies which are very susceptible to erosion. The annual precipitation is about 220 mm. The predominant lithologies are shale, sandstone and gabbro. The soil profiles are poorly developed.
Plot Locations and Characteristics
For specifying location of the plots, geology, slop, land use and erosional facies maps were prepared using 1:50,000 topography, geology and dip maps and field surveying. 4 different locations in basis of difference in geology and erosion facies were selected for these experiments. The plots located on different parent materials consist of shale, sandstone and gabbro in the. same slope (20 %) and land use (rangeland) but different lithology and erosion facies. In all of working polygons, the rainfall simulations carried out with intensity of 36 mm h-1 in autumn 2016 .

The Experiments Design
The rainfall simulator that was used in this study is a portable non-pressurized rainfall simulator which developed at the Soil Conservation and Watershed Management Research Institute (SCWMRI).
The 32 rainfall simulation experiments were performed during the autumn of 2016. All runoff and sediment data were collected and analyzed in the laboratory.
Before performing the simulations, in order to determine effective factors in sediment production and erosion, 32 soil representative samples from the first 15 cm depth of soil were taken and analyzed.

Statistical Analysis
The statistical analysis of data was conducted with the software SPSS for Windows. One-way analysis of variance techniques were used by Duncan Multiple Range Test with a level of significance of p≤0.05.
For determining the degree and type of correlation between sediment yield and soil physico-chemical properties and soil surface cover used the Pearson's correlation matrix (r) and multi-variable regression method. Stepwise multiple regression analysis was used to assess the effect of soil physico-chemical properties and soil surface cover on soil loss.

Results and discussion

The results showed that erosion and sediment yield in lithologies have meaningful differences. Jsh-RG (shale with Rill-Gully facies and Js-SR (sandstone with Sheer-Rill facies ) soil units with 68.12 and 45.12 gr/m2 have the most and the least sediment yield, respectively. It was found that the sediment yield had positive correlations with some soil properties such as silt percent, Ec, pH, and SAR and negative correlations with sand percent, OC, NPV (%), vegetation and rock fragment cover.
.
In this research, regression analysis was used to examine the relative contribution of soil physico-chemical properties on soil loss. The results present that the variables of percent of rock frogment (R.F) and Grass cover (G.C) have greater contribution in explaining the variations in soil loss.
Equation (1) with determination coefficients of 0.87 (R2) (p<0.01), selected as appropriate model for predicting soil loss.
Sediment Yield=109.112- 1.369 (R.F) -0.988(G.C) (1)
In these models, R2=0.87 indicate that 87% of the observed dissipation in dependent variables.

Conclusion
In this research, the spatial variability in soil loss for 4 representative selected soil samples derived from different parent rocks analyzed. The results revealed that rainfall simulation is well adapted to the analysis of rainfall-erosion processes within study area. Using a portable rainfall simulator revealed the effects on soil loss under rainfall intensity. Soils derived from shale with Rill-Gully facies and sandstone with Sheer-Rill facies showed the most and the least soil loss, respectively. ANOVAs showed that there are significant differences between treatments (different soils) in soil loss (P<0.01).
Multiple regression analysis revealed that rock fragment (R.F) and grass cover (G.C) are the most efficient factors determining soil loss.
Pearson’s correlation analysis showed that grass and rock fragment cover, soil vertical resistance and sand fraction are the efficient variables which have negative correlation with soil loss and the variables of silt fraction are the variables that have a positive correlation with soil loss. Meanwhile, the factors of SAR, EC and pH are the efficient chemical variables that have positive correlation with soil loss.
In this study, results of the experiments show that the magnitude of soil loss was highly controlled by some soil physical and chemical properties and soil vegetal and rock fragment cover. So, the mechanism of erosion involves the nature of the parent rocks, soil physico-chemical characteristics as well as ground cover.
Consequently, the finding of this research indicate that some physico-chemical properties of study soils and soil vegetation and rock fragment cover are suitable indicators for predicting soil loss in the study area.