پیش بینی سولوم خاک با استفاده از ویژگی های توپوگرافی در بخشی از اراضی تپه ماهوری کوهرنگ در زاگرس مرکزی

Introduction: Soil depth is defined as the depth from the surface to more-or-less consolidated material and can be considered as the most crucial soil indicator, affecting desertification and degradation in disturbed ecosystems. Soil depth varies as a function of many different factors, including slope, land use, curvature, parent material, weathering rate, climate, vegetation cover, upslope contributing area, and lithology. Topography, one of the major soil forming factors, controls various soil properties. Thus, quantitative information on the topographic attributes has been applied in the form of digital terrain models (DTMs). The prediction of soil depth by topographic attributes depends mainly on: i) the spatial scale of topographic variation in the area, ii) the nature of the processes that are responsible for spatial variation in soil depth, and iii) the degree to which terrain-soil relationships have been disturbed by human activities. This study was conducted to explore the relationships of soil depth with topographic attributes in a hilly region of western Iran.
Materials and Methods: The study area is located at Koohrang district between 32°20′ to 32°30′ N latitudes and 50°14′ to 50°24′ E longitudes, in Charmahal and Bakhtiari province, western Iran. The field sites with an area of 30,000 ha are located on the hillslopes at about 20% transversal slope. The soils at the site are classified as Typic Calcixerepts, Typic Xerorthents and Calcic Haploxerepts for the representative excavated profiles in summit, shoulder and backslope, respectively. The soils located at footslope and toeslope were classified as Chromic Calcixererts. Measurements were made in twenty representative hillslopes of the studied area. At the selected site, one hundred points were selected using randomly stratified methodology, considering all geomorphic surfaces including summit, shoulder, backslope, footslope and toeslope during sampling. Overall, 100 profiles were dug and described; and the solum thickness was measured for each profile. DEM data were created by using a 1:2,5000 topographic map. Topographical indices were generated from the DEM using TAS software. Terrain attributes in two categories, primary and secondary (compound) attributes; primary attributes are included elevation, slope, aspect, catchment area, dispersal area, plan curvature, profile curvature, tangential curvature, shaded relief. Secondary or compound attributes such as soil water content or the potential for sheet erosion, stream power index, wetness index, and sediment transport index. Correlation coefficients to define relationships between soil depth and terrain attributes, and analysis of variance by Duncan test were done using the SPSS software. The statistical software SPSS was used for developing multiple linear regression models. Terrain attributes were selected as the independent variables and soil depth was employed as dependent variable in the model. Thirty sampling sites were used to validate the developed soil-landscape model. In testing soil-landscape model, we calculated two indices from the observed and predicted values included mean error (ME) and root mean square error (RMSE).
Results and Discussion: The soil depth in the studied profiles varied from 30 cm to 150 cm with an average of 108.6 cm. Relatively high variability (CV = 76%) was obtained for soil depth in the study area. The linear correlation analysis of the 12 topographic attributes and one soil property (soil depth), showed that there was a significant correlation among 36 of the 77 attribute pairs. Soil depth showed high positive significant correlations with catchment area, plan curvature, and wetness index, and showed high negative correlation with sediment transport index, sediment power index and slope. Low positive significant correlations of soil depth were identified with tangential curvature, and profile curvature. Moreover, soil depth was negatively correlated with elevation. The rest of the topographic attributes including aspect, shaded relief, and dispersal area were not significantly correlated with soil depth. Many of these relationships are similar to those found in other landscapes. The results of analysis of variance showed that there are significant differences for soil depth among the selected slope positions in the studied area. The highest values of soil depth were observed in the downslope positions including footslope and toeslope. The lowest soil depth was observed in shoulder position with the highest rate of soil erosion.
Conclusions: It seems that the high variability for soil depth depends on topography of the field, and the landscape position, causing differential accumulation of water at different positions on the landscape; and moreover the soil erosion and deposition processes, resulting in high variability in the soil depth. We found relatively high correlation coefficients of soil depth with two groups of topographic attributes (erosional processes and water accumulation). Empirical model (MLR) using selected terrain attributes explains 76% of the variation of soil depth in the studied area. The terrain attributes that best predicted soil depth variability in the selected site were mainly the attributes that had significant relationships with soil depth. The dominant attributes in the MLR model included slope, wetness index, catchment area and sediment transport index.