K.U.Leuven, Celestijnenlaan 200E, BE-3001 Leuven. Belgium;. Institute .... 1:1 soil:water suspension was determined with a pH and EC meter (Mettler Toledo;.
ASSESSMENT OF SOIL ERODIBILITY IN TALEGHAN DRAINAGE BASIN, IRAN, USING MULTIVARIATE STATISTICS
Kazem Nosrati Department of Physical Geography Faculty of Earth Sciences Shahid Beheshti University Tehran, 1983963113 Iran Sadat Feiznia Department of Reclamation of Arid and Mountainous Regions Faculty of Natural Resources University of Tehran Karaj, 3158777878 Iran Miet Van Den Eeckhaut Physical and Regional Geography Research Group K.U.Leuven, Celestijnenlaan 200E, BE-3001 Leuven Belgium; Institute for Environment and Sustainability Joint Research Centre (JRC)—European Commission, 21027 Ispra (VA) Italy Sjoerd W. Duiker Department of Crop and Soil Sciences Pennsylvania State University 116 ASI Building, University Park, Pennsylvania
Abstract: Soil erosion has been recognized as one of the major forms of human-induced soil degradation. Due to land use changes in Iran, erosion has increased 800% between 1951 and 2002, calling for urgent action. But erosion research and policy development are hampered by a lack of information on the underlying factors controlling erosion. Soil types vary in their inherent susceptibility to erosion; but, like most countries, Iran lacks a network of field plots where erodibility is measured. A proxy for erodibility based on existing data and supplemented by an easily measured minimum data set is therefore needed. In this study, we use geological mapping and cluster, principal component, and factor analysis to group soils in the Taleghan Drainage Basin in Iran and subsequently determine their erodibility. First, a geological map of the area was prepared by photogeological methods and on-the-ground verification. Then, three soil profiles were investigated within similar landform units of each geological formation, and soil samples were taken. Physical and chemical properties that might impact soil erodibility (soil texture, pH, electrical conductivity, CaCO3, and soil organic matter) were used to create a matrix of soil properties and parent material. Application of cluster analysis and factor analysis to the data allowed identification of three geological (parent material) clusters. To investigate the mutual effect of land
78 Physical Geography, 2011, 32, 1, pp. 78–96. DOI: 10.2747/0272-3646.32.1.78 Copyright © 2011 by Bellwether Publishing, Ltd. All rights reserved.
SOIL EROSION IN TALEGHAN DRAINAGE
79
use and parent material on soil erodibility, a soil erodibility factor was obtained for three land use types in each cluster: rangeland, cropland (irrigated), and dry-land farming (nonirrigated). Geological cluster 1, consisting of marl, gypsum, and gypsiferous mudstone, was the most erodible; geological cluster 2, consisting of recent alluvium, alluvial fan, and landslip deposits, was of intermediate erodibility; and geological cluster 3, consisting of igneous rocks, dolomite, and conglomerate, was the least erodible. Within each geological cluster, dry-land farming was the most erodible, cropland was medium erodible, and rangeland was least erodible. The study suggests that geological and land use maps provide a useful framework for assessing soil erodibility. This work can guide future soil erosion studies and direct soil conservation policy to areas most susceptible to erosion. [Key words: soil e rodibility, geological cluster, land use, factor analysis, cluster analysis, Taleghan.]
INTRODUCTION Accelerated soil erosion is a serious problem in Iran, leading to degradation of soil and water resources, reduction of soil fertility, destruction of range and agricultural lands, desertification, recurring floods, sedimentation of reservoirs, and pollution of fishery habitats. Studies of floods over the period 1951 to 2002 in Iran showed that, in addition to life and financial losses, a significant volume of fertile soil was transported to reservoirs, lakes, and the ocean. Average soil erosion in Iran in 1951, 1961, 1981, 1993, 1999 and 2002 was estimated to be approximately 3.0, 4.6, 9.1, 15.2, 21.2 and 24.3 tons ha–1 year–1, respectively, representing an increase of 800% in the rate of soil erosion between 1951 and 2002 (Ahmadi, 1999; Sharifi and Heydarian, 1999; APERDRI, 2002). Sediment accumulation behind dams and in reservoirs at about 120 × 106 m3 year–1 causes a loss of dam reservoir capacity of 1–2% per year in Iran (IWRM, 2009). Hence, the increase in the rate of soil erosion in Iran necessitates a better understanding of the soil erosion process and application of conservation measures to combat this problem. Soil erodibility is the ease with which soil is detached by rainfall splash and surface flow, as affected by soil properties (Romkens et al., 1997). Ozdemir et al. (1998) found that soils formed on limestone were less erodible than soils formed on basalts in Turkey. Similarly, Sfar Felfoul et al. (2003) concluded that soil erosion was most affected by the lithological characteristics of the drainage basins. Although land use effects on soil erosion would be captured under the cover management (C) factor, soil erodibility is not independent of land use because of the influence of land use on soil organic matter content, soil structure, and permeability. Ozdemir and Ashkin (2003) found that the erodibility of soils derived from gypsum, alluvium, andesite, and basalt decreased as they were under maize, clover, and grass cultivation, respectively. Erodibility depends on soil physical and chemical properties such as organic matter content, soil permeability, aggregate stability, and cation types and concentrations (Morgan, 2005). A variety of different parameters and statistical tests has been proposed to describe and determine susceptibility of soil to erosion. The soil erodibility factor (K factor) in the Revised Universal Soil Loss Equation (RUSLE) is a well-known approach, based on soil texture, soil organic matter content, soil structure, and permeability (Wischmeier and Smith, 1978; Lal and Elliot, 1994; Renard et al., 1994, 1997; Tejada and Gonzalez, 2006; Perez-Rodriguez et
80
NOSRATI ET AL.
al., 2007; Zhang et al., 2008). Others have suggested the ratio of percent dispersed soil aggregates to percent stable soil aggregates (Bryan, 1968; Lal and Elliot, 1994). Shirazi and Boersma (1984) suggested a simple formula only involving the geometric mean particle diameter and standard deviation that has been used to estimate the soil erodibility by other studies (e.g., Salvador Sanchis et al., 2008; Zhang et al., 2008). The relationships of soil erodibility indices and soil properties have also been investigated by others, including Douglas and Goss (1982), Fullen et al. (1993), Sepaskhah and Mahdi-Hosseinabadi (2008), and Vaezi et al. (2008). In Iran, little information exists on the underlying factors of soil erosion, and there is no network of field plots where soil erodibility is measured. But a limited number of soil erosion studies in Iran showed the significant relationship of soil erodibility factor (K) in the RUSLE method with various factors such as land use and soil management (Bahrami et al., 2005), soil texture and organic matter content (Ghasemi and Mohammadi, 2003), and organic matter and lime contents (Vaezi et al., 2008). Therefore, a proxy for soil erodibility based on existing data and supplemented by an easily measured minimum data set is needed. Parent material and land use may both affect soil erodibility. In this study, we used geological and land use maps, field surveys, and multivariate statistical techniques to evaluate spatial variations of geological formations and soil erodibility. Multivariate statistical techniques, such as cluster analysis (CA), principal component analysis (PCA), and factor analysis (FA), help in interpreting complex data matrices of soil properties (Webster, 1979; Huggett, 1985). The objective of the study was to group soils according to common properties, using mapping and multivariate statistics to assist in sampling site selection, and to determine erodibility of Iranian soils using available datasets. MATERIAL AND METHODS Study Area The study was performed in the Taleghan Drainage Basin, located between 50º 20' and 51º 10' East Longitude and 36º 5' and 36º 23' North Latitude, in the Southern Alborz Mountains, 90 km northwest of Tehran, Iran (Fig. 1). The Taleghan River, the main river of the area, originates in the Asalak Mountains and, after combining with some tributaries, flows into the Taleghan Dam Reservoir. Geological formations from the oldest Pre-Cambrian to the youngest Quaternary formations are exposed at the surface in the drainage basin (Fig. 2). These geological formations have different lithological characteristics. Figure 3 is a soil map (according to the FAO classification) of the Taleghan Drainage Basin (IWMP, 1999). Sampling and Soil Analyses The geological map of the Taleghan Drainage Basin was extracted from 1:100,000 scale geological map of Shakran (Anenells et al., 1977) and was improved using aerial photo imagery and field investigations (Fig. 2). In the study area, 16 major geological formations were selected. Forty-eight soil samples (three samples in each geological formation) were collected from the top 20 cm of the soil in similar landform (slope,
SOIL EROSION IN TALEGHAN DRAINAGE
81
Fig. 1. Location of Taleghan Drainage Basin, Iran (Landsat ETM image, bands 4, 3, 2 in RGB). The dots show the locations of the 48 sampling sites used for the cluster analysis in the study area.
aspect and elevation) or uniform topographic units (Fig. 1). For each sample of airdried soil (
