Catena 156 (2017) 1–9
Contents lists available at ScienceDirect
Catena journal homepage: www.elsevier.com/locate/catena
Identification and prioritization of critical erosion areas based on onsite and offsite effects
MARK
Heidar Ghafaria,⁎, Manouchehr Gorjia, Mahmood Arabkhedrib, Ghorban Ali Roshanic, Ahmad Heidaria, Samira Akhavand a
Department of Soil Science, Faculty of Agriculture, University of Tehran, P.O. Box: 31587-11167, Karaj, Iran Soil Conservation and Watershed Management Research Institute, Agriculture Research, Education and Extension Organization, P.O. Box: 13445-1136, Tehran, Iran c Cotton Research Institute, Gorgan 49166-85915, Iran d Department of Water Engineering, Faculty of Agriculture, Bu-Ali Sina University, Hamedan, Iran b
A R T I C L E I N F O
A B S T R A C T
Keywords: Soil loss tolerance Erosion risk Erosion tolerance index Sediment-phosphorus index Sediment delivery ration
Accelerated soil erosion is considered as one of serious agro-environmental threats to sustainable development all over the word. Tolerable erosion concept is a tool for awareness and assessment of soil erosion status and its economic, social, and environmental hazards. This study was conducted to design a framework for evaluating and identifying spatial patterns of erosion hazard in Haji-Ghushan watershed, based on tolerable erosion concept, by using the SWAT model. The framework was consisted of two parts: the erosion tolerance index (ETI), and the sediment-phosphorous index (SPI) for evaluating onsite and offsite effects of soil erosion, respectively. Four hazard levels were defined for each index. The results of sediment simulation indicated that the maximum rate of erosion belonged to agricultural lands located on steep slopes in the central part of the watershed, and the minimum rate was from forest lands, despite their steep slopes. The map of spatial distribution of erosion hazard showed that, in terms of both onsite and offsite effects, a major part of the watershed (around 65%) had experienced an erosion rate lower than the erosion tolerance threshold; hence, these areas were not faced with erosion hazard. The spatial distribution of the areas exposed to the onsite erosion hazard, however, was differed from those confronted by the offsite erosion hazard. Identified high hazard areas based on the erosion offsite impacts were mainly located in sub-basins close to the watershed outlet where the sediment and phosphorous yield was high due to the high sediment-phosphorous delivery ratio. High hazard areas with high risk of soil degradation and productivity reduction are distributed throughout the watershed, depending on the magnitude of the erosion rate. These findings revealed that, in addition to erosion rate, sediment delivery ratio is also an important parameter in evaluating soil erosion hazard. For achieving the sustainable agro-environment, it is necessary to consider both the onsite and offsite effects of soil erosion to identify the high hazard areas. Also the results showed that the designed framework was capable of identifying the high hazard and hot spot areas well. The findings of this study are useful for officials and policy makers of soil conservation and environmental protection agencies in the region.
1. Introduction Globally, about 75 billion Mg of soil are eroded from agricultural lands (Pimentel et al., 1995) and around 0.3% of the agricultural production value is lost due to erosion each year (Den Biggelaar et al., 2003), which directly affects rural livelihood (Kerr, 1997; Lal, 1985) and challenges the achievement of the goal of food security (Pimentel and Burgess, 2013). Of the 75 billion Mg of soil lost worldwide, approximately two-thirds become deposited in lakes and rivers (Pimentel, 1997), which impacts aquatic resources (Bilotta and
⁎
Brazier, 2008; Eggermont and Verschuren, 2003; Clark, 1987), lake/ river sediment dynamics (Kelley and Nater, 2000; Walling, 2000), global carbon cycling (Lal, 2003), aquatic and terrestrial biodiversity (Alin et al., 2002; Harrey and Pimentel, 1996) and ecosystem services and soil quality (Lal, 2001; Pimentel and Kounang, 1998). Sediments also reduce water storage capacity, increase the maintenance cost of the dams, and shorten the life of the reservoirs (Pimentel et al., 1995). Research has shown that the intensity of soil erosion has increased considerably in Iran in the past few decades (Emadodin et al., 2012; Emadodin and Bork, 2011; Ahmadi et al., 2003; FAO, 1994) because of
Corresponding author. E-mail address:
[email protected] (H. Ghafari).
http://dx.doi.org/10.1016/j.catena.2017.03.014 Received 3 September 2016; Received in revised form 23 February 2017; Accepted 17 March 2017 0341-8162/ © 2017 Elsevier B.V. All rights reserved.
Catena 156 (2017) 1–9
H. Ghafari et al.
quantified, first, and the relationship between each of these functions and soil erosion must be determined. Therefore, scientists hope that in the near future they will be able to include all aspects of erosion in the concept of soil erosion tolerance. One of the most common limitations of existing method to determine the soil loss tolerance is that the offsite effects of soil erosion are not considered simultaneously. Therefore, the main objectives of the study were to assess the soil loss tolerance in relation to soil productivity as a main onsite impact and sediment and phosphorus delivery as a main offsite effects of erosion. The working hypothesis for this study was that estimated soil loss tolerance based on soil productivity may not be enough to control the destructive impacts of erosion from an environmental-economic perspective, and therefore another index is required.
incorrect use of soils, land-use changes, and mismanagement and that, if this current trend continues, this soil erosion will be a serious threat to the food security and the environment of this country. It has been estimated that about 500 million Mg of fertile soil are lost from the 16 million ha of agricultural lands in Iran every year (Emadodin et al., 2012). This means that the average annual intensity of soil erosion in these areas is about 32 Mg per ha. Moreover, studies on sediments yield in the watersheds of Iran indicate that storage capacity of the dams in the country declines by about 0.2 billion m3 per year (Emadodin et al., 2012). In other words, every year a relatively large dam like the Karaj Dam is filled up due to erosion. Based on the report published by Emadodin et al. (2012), losses caused by sedimentation at dams in Iran amount to about 0.6 billion U.S. dollars per year. Land management is certainly the key important factor to minimize a wide range of damaging effects of soil erosion. To implement best management practice, we need to determine soil loss tolerance, and identify and prioritize vulnerable and high-risk zones. Soil loss tolerance, which is introduced in order to give a measurement of how much soil should society allow to be eroded before experiencing excessive damage (Verheijen et al., 2009), has been recommended as an ultimate indicator in controlling onsite and offsite erosion effects (Li et al., 2009; Bhattacharyya et al., 2008; Sparovek and Jong van Lier, 1997 and it must be determined in a scientific and workable manner. Soil loss tolerance is one of the most fundamental and at the same time the most complex topics in studies on soil erosion. McCormack et al. (1982) proposed numerous criteria that should be considered in evaluating soil loss tolerance, including rate of weathering, changes in soil quality, impact on water quality, etc. Considered at a more pragmatic level, soil loss tolerance is equal to the rate of soil formation that is affected by changes in rainfall amount, average temperatures, water infiltration rate, type of soil cover and other environmental or anthropic factors (Jenny, 1941). The available data on soil formation rates are disperse and rare and the low rate of the process combined with the difficulty of its measurement at the soil-rock border makes the precise establishment not possible (Alewell et al., 2015; Sparovek and Jong van Lier, 1997). Furthermore, recent research has shown that the rate of soil production has an inverse relationship with soil depth (Dahms et al., 2012; Humphreys and Wilkinson, 2007), which means deep soils have lower thresholds of soil loss tolerance compared to shallow soils. Therefore, another criterion is proposed which limits the soil erosion to a specified range with an acceptable degree of risk associated with a soil type and social and political issues (Bazzoffi, 2009; Cole and Higgins, 1985). Soil productivity, defined as the capacity of a soil to produce a certain yield of crops or other plants with a specified system of management (Soil Science Society of America, 1997), and/or economic issues usually are a basis for determining this criterion (Lobo et al., 2005; Lal, 1998; Pierce, 1991). In a more recent study, Duan et al. (2017) developed a new method to calculate soil loss tolerance as a function of the soil productivity index (PI) for farmland on basis of the Skidmore (1982) equation. They defined T1 and T2 as lower and upper limit of the soil loss tolerance, respectively and stated that to maintain sustainable soil productivity, the soil productivity level should be higher than a threshold. In India country, a quantitative bio-physical model has been employed as a methodological approach for assessment of permissible soil loss based on soil resistance to erosion and soil depth (Mandal and Sharda, 2011; Mandal et al., 2010, 2006). Nowadays, considering increasing environmental concerns and pollution of water resources, some scientists have suggested that more than one criterion be determined for soil erosion tolerance (Bazzoffi, 2009; Li et al., 2009; Skidmore, 1982). Sparovek and De Maria (2003) expressed the view that the acceptable erosion level was a multidimensional problem that could prove successful only through a comprehensive natural process of thinking. Li et al. (2009) believed that the real soil erosion tolerance was an erosion level at which no decline or reduction happened in one or more soil functions. To calculate this soil loss tolerance all soil functions are need to be
2. Materials and methods 2.1. Site description This investigation was carried out in Haji-Ghushan, as Boustan dam drainage basin, that is a part of the large Gorgan-roud watershed, east of Golestan province between 37° 24′ and 37° 5′ N latitude and 55° 29′ and 56° 04′ E longitude with 1560 km2 area (Fig. 1). Golestan is one of the most important provinces for agricultural production in Iran. But, most of croplands located on steep lands that suffer high degree of erosion damages. The soils of the study area originated mainly from loess materials which are usually erodible because of silt particles abundance, low permeability and absence of cement materials between the particles. Accordingly, there are serious erosion problems in this area. The watershed's complex topography is characterized by mountainous, steep hillslopes and deep valleys in which two main branches of Gorgan-roud river flow down. The elevation varies from 100 m at the basin outlet to 2100 m at the highest point in the southeast of area. The climate is characterized as semiarid in north, and semi-humid in south, with average annual precipitation of 450 and 590 mm, respectively by De-Martonne classification system (Kazemi et al., 2015). The common land use of area includes cropland, rangeland, and forest land approximately 37, 32 and 29%, respectively. 2.2. Soil sampling Field surveys were employed to characterize the study watershed in terms of different soil, land use and topography based on maps, areal image and field observations. On the basis of such soil sampling zones, the watershed divided into 54 zones consisting of combinations of two slopes (2–12% and 12–25%), three land uses (cropland, rangeland and forestland) and nine soils. For each zone, representative soil profiles were described and soil samples (disturbed and undisturbed) were obtained from the 0–20, 20–40, 40–70, 70–100 cm depth intervals. Undisturbed soil samples were taken in cores (100 cm3 steel cylinders) from each layer (n = 3) for analysis of water holding capacity (WHC) and soil bulk density (BD). Initially, these soil samples were saturated from the bottom using a 0.01 M CaCl2 solution for 24 h and then placed in a pressure plate apparatus for determining WHC by the difference between soil moisture content at field capacity and wilting point, which estimated from soil samples equilibrated at a pressure of 33 and 1500 kPa on a pressure plate, respectively (Romano and Santini, 2002). In the next step, samples were weighed before and after oven drying at 105 ̊C to determine the soil's bulk density. The disturbed soil samples were air-dried and passed through a 2-mm sieve to determine selected soil physicochemical properties such as particle size distribution (Gee and Or, 2002), organic matter (OM) content (Nelson and Sommers, 1982) and soil phosphorus (P) concentration (Olsen and Sommers, 1982). In addition these soil samples, 350 recently collected topsoil samples from the soil quality monitoring project by Agriculture Organization of Golestan Province (unpublished report) were indepen2
Catena 156 (2017) 1–9
H. Ghafari et al.
Fig. 1. Location of Haji-gushan catchment and Boustan dam.
surface runoff and sediment on a daily time step by the SUFI-2 algorithm in SWAT-CUP program. Surface runoff and sediment data from 2001 to 2013 were used for calibration and the 1993 to 2000 data were used for model validation with the 1990–1993 data as a warm-up period for the model. To quantify the goodness of model performance five objective functions include coefficient of determination (R2), the Nash-Sutcliff coefficient (NSE), Percent bias (PBIAS), P-factor and Rfactor were selected based on the recommendations suggested by Moriasi et al. (2007) and Abbaspour et al. (2004).
dently used to assess the spatial distribution of the soil phosphorus in surface layers. 2.3. SWAT model description SWAT, as a large complex watershed-scale model, is a widely used model designed originally to simulate hydrology, sediment, nutrient and pesticide transport as well as crop growth over long period of time and predict the impact of management and climate change on them (Arnold et al., 1998). SWAT first simulates physiography details of watershed by describing sub-basins that are connected by a stream network and each sub-basin further delineates into hydrologic response units (HRUs). SWAT simulations are basically conducted at the HRUs level then aggregated to the sub-basin, and routed through the stream network. Surface runoff generally is estimated using curve number (CN) method, and soil erosion is estimated with the Modified Universal Soil Loss Equation (MUSLE).
2.6. Soil erosion risk assessment In order to simultaneously study both onsite and offsite effects of erosion, it is important to recognize difference between soil loss and sediment yield. Soil erosion generally refers to losing soil from a drainage area, while sediment yield is the only part of eroded soils which is reached to the basin outlet (de Vente et al., 2008). Onsite impacts are generally related to soil loss from a field, while offsite impacts commonly come from transported sediment to a particular location. Therefore, assessing onsite and offsite effects of soil erosion requires different methods.
2.4. SWAT model setup Multiple spatial datasets including available GIS maps for topography, land use, and soils were used for SWAT model parameterization. A digital elevation model (DEM) in 30 m resolution developed by National Cartographic Center of Iran was used for delineation of watershed. Soil and Land use map were obtained from the Jihad Agriculture organization of Golestan province at the scales of 1:250,000 and 1:100,000, respectively. Twenty four years (1990–2013) of daily precipitation and maximum/minimum temperature data obtained from the Meteorological organization of Golestan province contains two synoptic and four precipitation stations. Continuous daily surface runoff and limited sediment data load of the Tamer hydrology station located in outlet of the watershed were used to calibration process.
2.6.1. Erosion risk assessment based on onsite effect Soil productivity index (PI) model, as a depth-weighted soil quality index, has been widely used in diverse soils and agro-ecological regions to determine the erosion-productivity relationship. As PI requires only a limited number of variables, it is particularly useful for places where data are limited. The basic assumption in the PI model is that the soil properties within the rooting zone control crop growth and yields and other factors, such as climate, management, and genetic plant potential are constant (Duan et al., 2009; Delgado and Lopez, 1998; Pierce et al., 1983). PI value, ranging from 0 to 1, is calculated by aggregating the productivity values of each soil layer to a depth of 100 cm as following:
2.5. SWAT model calibration and validation
n
PI = The Haji-Ghushan watershed was calibrated and validated for
∑ (Ai ∙ Ci∙ Oi∙Di ∙Ki ) i =1
3
(1)
Catena 156 (2017) 1–9
H. Ghafari et al.
To interpretation of results an erosion tolerance index (ETI), defined as ratio of soil loss tolerance (T value) to the annual soil loss rate (A), was determined for each sub-basin (Sudhishri et al., 2014) and classified into four vulnerability classes include no risk (> 1), low (0.8–1), moderate (0.5–0.8), and high risk (< 0.5). The ETI value < 0.5 indicates current soil loss exceeding more than two times the threshold value.
Table 1 Soil productivity factors and scoring curve equations. Factor
Scoring curve equation
WHC
⎧ 0.05 × WHC……if …. …0 ≤ WHC ≤ 20% A=⎨ ⎩ 1……………if ………. ……WHC > 20%
CLAY
⎧ clay ……………………0 ≤ CLAY < 20% ⎪ ⎪ 20 C = ⎨ 100 − clay …………. . 40% < CLAY ≤ 100 ⎪ 60 ⎪ ⎩ 1………………………. 20% ≤ CLAY ≤ 40%
ETI =
OM
⎧ OM ⎪ ……. . …. if ……0% ≤ OM < 2% O=⎨ 2 ⎪ ⎩ 1………. . …if ……. ……OM ≥ 2%
BD
⎧ 1.87 − (0.67 × BD )……. if ……. 1.3 ≤ BD ≤ 1.5 D=⎨ ⎩ 1……. …………. …if ………………BD < 1.3 Kcum = 0.024 x0.81 Ki = Kcum , i − Kcum , i − 1
K
where i is the number of the soil layers, and n is the total number of soil layers. A, C, O, and D are water holding capacity (cm3·cm− 3), clay content (%), organic matter (%), and bulk density (gr·cm− 3), respectively. K is the relative importance of each layer in the soil profile and also the importance of soil depth. All these attributes are evaluated separately for each layer of the soil up to a depth of 100 cm, and then measured indicator values were converted into 0–1 scale using appropriate scoring curve equations (Duan et al., 2012, 2011; Mandal et al., 2010; Lobo et al., 2005; Delgado and Lopez, 1998; Pierce et al., 1983), (Table 1). Assuming the PI values changes linearly with soil depth removal, the soil's vulnerability to erosion (V), is defined as the rate of soil productivity index change with soil eroded depth and represents the proportional change in PI value after 1 cm of top soil is removed. The following equation was applied to estimate soil loss tolerance:
V=−
∆PI d
(4)
2.6.2. Erosion risk assessment based on offsite effect Reducing active storage and impairing water quality of the reservoirs due to sediment and phosphorus (P) loading are respectively two most important offsite effects of erosion (Verstraeten et al., 2003; Correll, 1998). Therefore, the design of any generalized assessment tool for potential environmental risk of erosion must be able to take into account sediment yield and sediment-bound phosphorus. The phosphorus index is a simple and widely used indicator to assess phosphorus yields and identifying critical areas (Mallarino et al., 2002; Lemunyon and Gilbert, 1993). In general, P index usually consists of three components: the transport potential of phosphorus to surface water through erosion (CE), runoff (CR) and subsurface drainage (CD). The subsurface drainage component was ignored because it is not generally recognized as an important component to transporting sediment and P in our case. In this study, phosphorus index was modified for sub-basin and basin scales and was called sediment-phosphorus index (SPI) instead of P index, as it takes into account both sediment and phosphorus parameters which can be calculated as following:
WHC: the water retained between field capacity and wilting point. OM: soil organic matter, BD: soil bulk density, and K: the relative importance of each layer in the soil profile.
R∙BD∙PI0 T value = V ∙t
T value A
SPI = CE + CR
(5)
CE = SDYi × SDRi × Enrichment factori × STPi Erosion factor
(6)
CR = Runoffi × (STPi Runoff factor)
(7)
SDR =
YO, Pi =1 − YO, Pi =0 Li, Pi =1
(8)
STPi Erosion factor = 0.7 + 0.0042 (soil test Olsen P)
(2)
(9)
STPi Runoff factor = 0.05 + (soil test Olsen P × 0.0083) (3)
(10)
and the enrichment factor (EF) are estimated using the following equations (Mallarino et al., 2002):
where T value is soil loss tolerance based on productivity (Mg ha− 1 yr− 1), R is allowable reduction rate in soil productivity (%) in the planning duration of t years, BD is topsoil bulk density (Mg m− 3), PI0 is the current soil productivity index, V is soil vulnerability (productivity loss per cm), ΔPI is the variation in PI and d is the thickness of the eroded soil in cm. The R and t values was selected 0.1% per year and 100 years, respectively, as standard (Duan et al., 2012; Mandal et al., 2010; Lobo et al., 2005; Delgado and Lopez, 1998; Rijsberman and Wolman, 1985). The reason for selecting 0.1% per year as permissible reduction rate is that soil formation factors and developing technology can restore it. Because both soil degradation and soil formation (and evolution) factors occur simultaneously. Therefore, reduction rate in soil productivity resulting in erosion should be equal or less than the increase rate of the productivity related to topsoil formation and management practices such as technical implementations, fertilization and other inputs to sustain crop productivity indefinitely (Du et al., 2013). As reported by the statistical center of Iran, the current population growth rate of the country is 1.29% that will be zero by 2050, and negative thereafter and the annual increase rate of wheat yield is estimated about 1.95% during a 30 years period of 1972–2002 and will be 1% in 2020 (Zarea et al., 2007). According to these trends, it is safe to say that the soil productivity loss rate of 0.1% per year can maintain crop productivity in the selected planning period (100 years).
⎧ 1.1. ……. . if …. Without Buffer and Tillage used EF = ⎨ ⎩ 1.3……………. . if …. With Buffer and No − Till −1
(11) −1
yr ) of sub-basin where SDY is mean annual sediment yield (Mg ha i, and STP refers to recently soil test phosphorus. SDR is sediment delivery ration for sub-basin i to the watershed outlet, YO, Pi = 1 is the sediment yield at the watershed outlet when Pi,USLE = 1, YO, Pi = 0 is the sediment yield at the watershed outlet when Pi,USLE = 0 for sub-basin i, and Li, P = 1 is the sediment yield at sub-basin i when Pi,USLE = 1 (calculated using MUSLE). This sub-basin level SDR was estimated using SWAT model proposed by Sean and Nejadhashemi (2013). Sediment-phosphorus index are eventually classified into five risk classes includes very low (0–1), low (1–2), medium (2–5), high (5–15), and very high (> 15), which soil conservation practice or management is not required for two first classes, while for next classes environmental concerns begin to become and remedial action is required. 3. Results and discussion 3.1. Model SWAT results The results of calibration SWAT model reflect goodness of modelling the surface runoff and sediment rates for the Haji-Ghushan watershed 4
Catena 156 (2017) 1–9
H. Ghafari et al.
3.2. Soil productivity index
Table 2 Summary statistics of calibration and uncertainty analysis of flow and sediment. Criteria
ENS R2 PBIAS P-factor R-factor
Surface runoff Calibration
Validation
Sediment Calibration
Validation
0.60 0.62 1.2 0.8 1.91
0.51 0.52 6.6 0.74 1.51
0.57 0.63 19.1 0.63 0.26
0.54 0.54 22.3 0.61 0.28
The values of selected soil properties representing each land use were averaged and listed in Table 3. The deep derived-loess soils of the study area are dominantly classified as Molisols and Inceptisols according to USDA Soil Taxonomy method (Khormali et al., 2009). Soil textures found in the watershed include silty loam and silty clay loam in up to 90% of the soil samples. The soil OM contents ranged between 0.85% and 2.60% in 0–20 cm, decreasing with depth gradually. Comparing the topsoil OM content across the three land use types showed it differed significantly (Table 3). Cultivated areas were found to have significantly lower organic matter of 1.18% than forests and rangeland in topsoil layers (Table 3). The lower content of OM in the topsoil of croplands than forest and rangelands indicates a reduction in OM following cultivation practices (Khormali et al., 2009). The average clay content of the 216 soil samples was sufficient to meet the optimum crop growth based on the PI model (Duan et al., 2009), ranging from 18% to 27% with average of 25%. Therefore, clay content is not recognized as a stress factor for soil productivity. The average bulk density of soil samples varied between 1.23 Mg m− 3 and 1.51 Mg m− 3. In general, BD was significantly increased with depth, probably due to decreased organic matter content and tillage traffic (Table 3). The average values of WHC varied from 0.22 to 0.11 mm and significantly decreased with increasing soil depth, in the same manner as OM. In fact, the greater BD and lower OM content of subsurface layers can result lower WHC. WHC of the topsoil (0 to 20 cm soil depths) was significantly lower in cropland than rangeland and forest about 10% and 25%, respectively (Table 3). Generally, OM and WHC are shown to be more influencing factors on PI in this study. The soil productivity index was only calculated for croplands and rangelands in two slope classes (2–12% and 12–25%), because > 95%
based on range recommendations of the objective functions (Table 2). NSE equal to 0.6 and 0.51 for surface runoff calibration and validation, respectively based on daily data, which both are higher than the minimum NSE value (0.5) for flow calibration (Moriasi et al., 2007). Obtained NSE values for sediment delivery calibration and validation are 0.57 and 0.63, respectively, which are within the recommended limits. The coefficient of determination R2 for daily flow calibration and validation was 0.62 and 0.52, respectively, and for daily sediment yield calibration and validation 0.63 and 0.54, respectively, which indicates reasonably satisfactory (R2 > 0.6) (Santhi et al., 2001). The PBIAS for both flow and sediment calibration and validation are also within the acceptable limits ( ± 25% for stream flow, ± 55% for sediment). The uncertainty in the predictions of surface runoff (Fig. 2) and sediment (Fig. 3) quantified by P-factor and R-factor also were satisfactory. The P-factor of 1 and R-factor of zero indicate a simulation that exactly corresponds to measured data. Acceptable suggested Pfactor for discharge is value of > 70%, and for sediment, a smaller Pfactor could be acceptable. The P-factor and R-factor obtained here showed acceptable performance of the model for flow and sediment simulations (Abbaspour et al., 2007, 2004).
Fig. 2. Observed, best estimated and 95PPU flow for calibration and validation period.
Fig. 3. Observed, best estimated and 95PPU sediment yield for calibration and validation period.
5
Catena 156 (2017) 1–9
H. Ghafari et al.
Table 3 Averages of measured soil properties for each land use. Land use
Depth
O.M. (%)
AWC (mm)
Cropland
0–20 20–40 40–60 60–100 0–20 20–40 40–60 60–100 0–20 20–40 40–60 60–100
1.18 ± 0.36 c 0.92 ± 0.34 c 0.61 ± 0.22 d 0.34 ± 0.14 e 1.56 ± 0.49 b 0.58 ± 0.26 d 0.29 ± 0.15 e 0.3 ± 0.08 e 2.30 ± 0.5 a 1.80 ± 0.3 b 0.70 ± 0.1 d 0.55 ± 0.1 d
0.17 0.15 0.12 0.11 0.19 0.17 0.13 0.11 0.22 0.19 0.14 0.11
Rangeland
Forest
± ± ± ± ± ± ± ± ± ± ± ±
0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.03 0.02 0.02
Bd (Mg m− 3) b b c c a b c c a a bc c
1.23 1.34 1.51 1.48 1.25 1.36 1.42 1.44 1.20 1.33 1.45 1.46
± ± ± ± ± ± ± ± ± ± ± ±
0.03 0.04 0.03 0.03 0.04 0.04 0.02 0.01 0.04 0.04 0.02 0.01
Clay (%) c b a a c b a a c b a a
23 27 27 25 22 25 22 23 21 24 25 18
± ± ± ± ± ± ± ± ± ± ± ±
3a 4a 5a 5a 8a 10 a 8a 5a 6a 7a 4a 4a
Values within the same column followed by different letter(s) are statistically different using Duncan's multiple range test at P = 0.05.
of these land uses are placed at areas with 2–25% slopes. As the PI model represents the relative soil productivity determined by the soil properties of the first 100 cm soil (Mandal et al., 2010), and also deeper rooting depth is needed for the growth of forest root systems, it is not estimated for forest soils. Moreover, soil erosion rate of the forest areas is lower than soil loss tolerance limit. The average PI values, representing in Table 4, were 0.39 and 0.36 for cropland sites and were 0.41 and 0.43 for rangeland sites at 2–12% and 12–25% slope, respectively. In this study area, the productivity of soils are associated with medium productivity class, according to Lobo et al. (2005), and OM and WHC are recognized as a more effective factors in reducing PI values. The T value, determined based on the relationship between the eroded soil depth and the soil productivity and R = 0.1%, ranges between 7.2 and 7.5 Mg ha− 1 yr− 1 for croplands and 6.3 and 6.7 Mg ha− 1 yr− 1 for rangeland at 2–12% and 12–25% slope, respectively. The T value depended on many factors, especially soil vulnerability to erosion (Duan et al., 2012). The average soil vulnerability to erosion ranges from − 0.0072 to − 0.0063 cm− 1 for croplands and from − 0.0078 to −0.0089 cm− 1 for rangelands. A greater absolute value of V implies a more responsive of the soil productivity to erosion (i.e., a faster reduction in soil productivity), and resulted in decreases of T value (Duan et al., 2012). Subsurface horizons frequently have lower productivity (lower OM and WHC, higher BD) than topsoil horizons. Thus, productivity decreases when soil is removed (Duan et al., 2011). The more ratio of topsoil fertility to subsoil fertility, the more soil vulnerability to erosion, thus, rangeland soil was recognized as a more vulnerable compared to cropland soil resulted in lower T value for rangeland. Wei et al. (2009) believe that in semi-arid regions tolerable soil erosion for rangeland are lower than those for cultivated land due to shallow topsoil depth and the slow rates of soil formation.
Fig. 4. Classified erosion risk based on soil productivity loss.
soil erosion is comparatively lower than the tolerance limits, which does not call for implementation soil conservation measures. Of the whole watershed area about 19.4% was considered to be associated with high risk of productivity loss due to erosion rate in excess of more than two times the tolerable erosion rate, and 9.1%, and 5.5% suffers low and medium risk, respectively (Table 5). Therefore, proper soil conservation practices should be implemented in this area to prevent soil degradation and abandonment resulted in extremely severe productivity loss. The annual soil loss rate for the sub-basins under forest land use (27% of watershed area) was generally much lower than 1 Mg ha− 1 yr− 1, so no risk was considered. In case of rangeland, about 76.5% of the areas exhibit an ETI associated with no risk, and approximately 10.4%, 6%, and 7.1% had low, moderate, and high risk, respectively. As expected, erosion rates on rangelands were lower compared with those on cultivated lands (USDA, 2000), may be due to their undisturbed soils and adequate proper vegetation cover during
3.3. Erosion tolerance index The erosion tolerance index (ETI), was calculated for each subbasin, and thus, a total of 144 ETI values for the whole watershed were obtained and was classified into four risk classes includes no risk (> 1), low (0.75–1), medium (0.5–0.75), and high (< 0.5). The ETI results are shown in Fig. 4. The soil productivity loss risk map showed that most of the area of the watershed (66%) has not any risk of soil erosion where
Table 5 Spatial coverage of erosion risk class based on productivity loss. Land use
Slope
Table 4 Soil loss tolerances, T values (R = 10% and t = 100 years). Cropland Land use
Slope
PI0
Cropland
2–12% 12–25% 2–12% 12–25%
0.39 0.36 0.41 0.43
Rangeland
± ± ± ±
0.01 0.02 0.04 0.03
V
T value
−0.0072 −0.0063 −0.0078 −0.0089
7.2 7.5 6.7 6.3
± ± ± ±
Rangeland
0.27 0.32 0.54 0.43
Watershed
2–12% 12–25% > 25% 2–12% 12–25% > 25% –
Productivity loss risk class (area, %) H M L N
Sum
11.7 28.0 0.0 0.5 6.6 0.0 19.4
45.6 54.4 0.0 39.6 57.9 2.5 –
5.9 3.1 0.0 3.5 0.0 2.5 5.5
0.0 11.0 0.0 0.0 10.4 0.0 9.1
28.0 12.3 0.0 36.2 40.3 0.0 66.0
H: high erosion risk, M: medium erosion risk, L: low erosion risk, N: no erosion risk.
PI0: Initial productivity index (Eq. (7)), V: soil vulnerability (Eq. (8)).
6
Catena 156 (2017) 1–9
H. Ghafari et al.
The potential of dissolved P loss into water depends on soil P concentration and the P sorption capacity of the soil. Hence, the potential of the risk of dissolved P loading to stream with runoff is very low, while the sediment-attached P delivered from the sub-basins are more dominant compared to dissolved P loss with runoff in this area. The spatial coverage of estimated erosion environmental risk classes based on SPI are illustrated in figure Fig. 6a. At the entire watershed, about 70% of watershed area were classified as no risk class, while only 4% of the area was displayed as a high risk of sediment and phosphorus delivery (hot spots), and 11% and 15% of the area reflects low and medium classes, respectively, as shown in Table 6. Therefore, about 60 km2 of the watershed area require conservation measures and should be prioritized in terms of offsite impacts of erosion. Minimizing the sediment delivery from the sub-basins closer to the reservoir can be suggested to prevent catastrophic offsite erosion hazards (Sean and Nejadhashemi, 2013; Verstraeten et al., 2003). For croplands, only 15% of the area falls in no risk class where SPI value is lower than 2 Mg ha− 1 yr− 1. The low and medium risk classes occupy about 28% (16% at the low slope, 12% at the high slope) and 45% (19% at the low slope and 26% at the high slope) of the cropland areas, respectively. The most dangerous situation (high class), found in correspondence to croplands on the steep slopes (12–25%) that covers about 12% of the cropland area. Khormali et al. (2009) reported the severe erosion in the cultivated land resulting in an increase of the soil susceptibility to erosion due to deforestation. A large part of the rangelands (94.5%) classified as no risk area and about 5.5% of the rangelands areas classified in low risk classes. Based on SPI classifications, none of the sub-basins under rangeland falls in medium and high risk classes. In general, the low and medium classes need conservation applications to maintain a sustainable productivity and environment health, while the last class (high), is very dangerous because it can be
all seasons (Wei et al., 2009). The spatial distribution of erosion risk classes for croplands was about 40% no risk, 11% low, 9% moderate, and 40% high. The highest erosion risk are found where steep lands are used mainly for agriculture (Table 5), because agricultural operations gives rise to loss of organic matter (Vagen et al., 2006), which directly affects soil structure degradation in long-term (Caravaca et al., 2004). In this area, priority must be given for afforestation of steep lands and maximize plant coverage by conservation tillage, proper rotations and cover crop practices in the croplands. 3.4. Sediment-phosphorus index The proposed sediment-phosphorus index (SPI), as an index of environmental erosion hazard, was calculated for each sub-basin using Eqs.uation (5)–(11) include sub-basin-level sediment yield, runoff (estimated by MUSLE), and SDR, as well as soil test phosphorus at the surface layer (0–20 cm soil depth) (Fig. 5). The calculated SDR varied from 2 to 94% and spatial distribution map of the SDR values indicates that sub-basins closer to the watershed outlet have higher SDR (94%), meaning that almost the total sediment loaded from these subbasins was delivered to the watershed outlet. The observed SDR variations between sub-basins are likely due to physiographic characteristics, land cover, and especially distance to watershed outlet (Sean and Nejadhashemi, 2013). The calculated sediment yield using MUSLE varies from 0 to 80 Mg ha− 1 yr− 1, while only the proportion of sediment can leaves the sub-basin and reaches watershed outlet depend on SDR. Runoff yield values exhibited the spatial variation as sediment yield, ranging from 0 to 23.42 mm. Because sediment yield is estimated by MUSLE as a function of surface runoff volume and peak runoff rate (Williams, 1995). Soil phosphorus content in the surface layers was lower than critical environmental threshold, ranging from 2 to 13 ppm, with the highest for croplands and lowest at the forest and rangelands.
Fig. 5. Sediment yield, runoff yield, SDR, and soil phosphorus maps.
7
Catena 156 (2017) 1–9
H. Ghafari et al.
Fig. 6. Classified erosion risk map based on sediment and phosphorus delivery risk under the base scenario (a), and when ETI ≥ 1 in all sub-basins (i.e. no onsite erosion risk) (b).
essentially the same as those achieved based on offsite effects. Because soil loss rate is recognized as an important factor governing soil loss tolerance in terms of its onsite impacts, while the contribution of sediment loading from sub-basins to watershed outlet was more important than actual erosion rate with regard to offsite effects. Regarding offsite effects, the highest erosion risk was relative to subbasins with highest SDR value that generally located near the watershed outlets. But the identified high erosion risk sub-basins based on soil productivity represent spatial distribution throughout the watershed. The presented methodology has two important advantages: The first one is that the degree of soil vulnerability to soil erosion, which depends on the difference between the productivity of the surface soil compared to the subsoil layers, plays an important role in determining tolerable erosion for preservation of soil productivity. The more desirable the chemical and physical conditions of the topsoil are compared to the subsoil, the greater soil vulnerability to erosion will be and, hence, tolerable erosion will decline. The second advantage is that our method has a comprehensive look at the process of soil erosion (detachment, transport, and deposition), and controls the offsite effects of erosion simultaneously with the onsite ones. Because soils that have high erosion tolerance due to their productivity may also have the high potential for transferring sediments and contaminants to surface waters (high sediment delivery ratios). Under these conditions, production and transfer of sediments gains importance. We believe that each onsite and offsite effect of soil erosion requires separate indices to evaluation and, consequently, needs different management practices to control. In all, this method takes a special look at controlling the effects of offsite erosion. The managers and decision-makers to deal with soil erosion needs to consider both onsite and offsite effects of erosion because dealing with the both effects of soil erosion needs different measures at different locations. Therefore, this study emphasized that decision should be made not only based on productive soil loss rates, but also contribution of sediment delivery to water bodies. The outputs of these studies can be used for planning of conservation practices in the study area.
Table 6 Spatial coverage of erosion risk class based on sediment and phosphorus delivery. Land use
Cropland
Rangeland
Watershed
Slope
2–12% 12–25% > 25% 2–12% 12–25% > 25% –
Sediment and P transport risk class (area, %) H M L N
Sum
0.0 12 0.0 0.0 0.0 0.0 4
46 54 0.0 40 57.5 2.5 –
19 26 0.0 0.0 0.0 0.0 15
16 12 0.0 5 0.5 0.0 11
11 4 0.0 35 57 2.5 70
H: high erosion risk, M: medium erosion risk, L: low erosion risk, N: no erosion risk.
degradative in long-term if no intervention are done and soil loss level is maintained constant in the future. The spatial distribution of high risk sub-basins derived from SPI method is different from those obtained by ETI method. Because soil loss rate is an important factor governing soil loss tolerance, in terms of its on-site impacts (Lal, 1998), while the contribution of sediment loading from sub-basins to watershed outlet is more important than actual erosion rate with regard to offsite effects (Sean and Nejadhashemi, 2013). To clarify this issue, after setting the calculated soil loss tolerance (T value) for each sub-basin based on productivity, erosion risk map was reproduced using the SPI (Fig. 6b). The results showed that environmental risk of erosion still remains at some subbasins located near the outlet of the watershed. This emphasizes the new index (SPI) should be used together with the onsite effects indicator (ETI), as a useful complement to sustainable watershed management. In terms of environmental risk, tolerable erosion based on soil productivity may not be adequate for effective sediment reduction and phosphorus load to watershed outlet especially when T value and SDR are high. The spatial variability of soil properties, land use, and topography within a watershed resulting in spatial variability in soil erosion rate and sediment delivery ratio as well as T value, the spatial evaluation approach should be used to assess soil erosion risk. To achieve this, spatially distributed models such as SWAT are particularly useful tools.
Acknowledgements The authors acknowledge the National Cartographic Center of Iran, Jihad Agriculture and Meteorological organizations of Golestan province for the provision of data.
4. Conclusion This study used an approach for identification and prioritization of high erosion risk area that considers both onsite and offsite effects of soil erosion for use in the decision making framework. Soil productivity index and sediment-phosphorus index were used as two main criteria to reflect the onsite and offsite effects of erosion. SWAT, as a spatially model, was used to achieve the aim. The conclusion was that the identified areas with high erosion risk based on onsite effects are not
References Abbaspour, K.C., Johnson, C.a., van Genuchten, M.T., 2004. Estimating uncertain flow and transport parameters using a sequential uncertainty fitting procedure. Vadose Zo. J. 3, 1340–1352. Abbaspour, K.C., Yang, J., Maximov, I., Siber, R., Bogner, K., Mieleitner, J., Zobrist, J., Srinivasan, R., 2007. Modelling hydrology and water quality in the pre-alpine/alpine Thur watershed using SWAT. J. Hydrol. 333, 413–430.
8
Catena 156 (2017) 1–9
H. Ghafari et al.
Lal, R., 2001. Soil degradation by erosion. Land Degrad. Dev. 12 (6), 519–539. Lal, R., 2003. Soil erosion and the global carbon budget. Environ. Int. 29, 437–450. Lemunyon, J.L., Gilbert, R.G., 1993. The concept and need for a phosphorus assessment tool. J. Prod. Agric. 6, 483–496. Li, L., Du, S., Wu, L., Liu, G., 2009. An overview of soil loss tolerance. Catena 78, 93–99. Lobo, D., Lozano, Z., Delgado, F., 2005. Water erosion risk assessment and impact on productivity of a Venezuela soil. Catena 64 (2–3), 297–306. Mallarino, A.P., Stewart, B.M., Baker, J.L., Downing, J.D., Sawyer, J.E., 2002. Phosphorus indexing for cropland: overview and basic concepts of the Iowa phosphorus index. J. Soil Water Conserv. 57, 440–447. Mandal, D., Sharda, V.N., 2011. Assessment of permissible soil loss in India employing a quantitative bio-physical model. Curr. Sci. 100, 383–390. Mandal, D., Dadhwal, K.S., Khola, O.P.S., Dhayni, B.L., 2006. Adjusted T values for conservation planning in Northwest Himalayas of India. J. Soil Water Conserv. 61 (6), 391–397. Mandal, D., Sharda, V.N., Tripathi, K.P., 2010. Relative efficacy of two biophysical approaches to assess soil loss tolerance for Doon valley soils of India. J. Soil Water Conserv. 65 (1), 42–49. McCormack, D.E., Young, K.K., Kimberlin, L.W., 1982. Current criteria for determining soil loss tolerance. In: Schmidt, B.L., Allmaras, R.R., Mannering, J.V., Papendick, R.I. (Eds.), Agronomy Society of America Special Publication. vol. 45 Agronomy Society of America, Madison, Wisconsin. Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D., Veith, T.L., 2007. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans. ASABE 50, 885–900. Nelson, D.W., Sommers, L.E., 1982. Total carbon organic carbon and organic matter. In: Page, A.C. (Ed.), Methods of Soil Analysis, Part 2. Chemical and Microbiologically Properties. ASA, pp. 539–579. Olsen, S.R., Sommers, L.E., 1982. Phosphorus. In: Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. American Society of Agronomy, Madison, WI, USA, pp. 403–430. Pierce, F.J., 1991. Soil Management for Sustainability, Soil Water Conserv Soc.In: Lal, R., Pierce, F.J. (Eds.), Erosion Productivity Impact Prediction, pp. 35–52 Ankeny, IA. Pierce, F.J., Larson, W.E., Dowdy, R.H., Graham, W., 1983. Productivity of soils: assessing long-term changes due to erosion. J. Soil Water Conserv. 38, 39–44. Pimentel, D., Kounang, N., 1998. Ecology of soil erosion in ecosystems. Ecosystems 1, 418–426. Pimentel, D., 1997. Soil erosion. Environment 39, 4–5. Pimentel, D., Burgess, M., 2013. Soil erosion threatens food production. Agriculture 3, 443–463. Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., Crist, S., Shpritz, L., Fitton, L., Saffouri, R., Blair, R., 1995. Environmental and economic costs of soil erosion and conservation benefits. Science 267, 1117–1123. Rijsberman, F., Wolman, M., 1985. Effect of erosion on soil productivity an international comparison. J. Soil Water Conserv. 40 (4), 349–354. Romano, N., Santini, A., 2002. Water retention and storage: field. In: Dane, J.H., Topp, G.C. (Eds.), Methods of Soil Analysis: Part 4. Physical Methods. SSSA Book Ser. 5 SSSA, Madison, WI. p, pp. 721–738. Santhi, C., Arnold, J.G., Williams, J.R., Dugas, W.A., Srinivasan, R., Hauck, L.M., 2001. Validation of the SWAT model on a large river basin with point and nonpoint sources. J. Am. Water Res. Assoc. 37, 1169–1188. Sean, A.W., Nejadhashemi, A.P., 2013. Spatial and temporal variabilities of sediment delivery ratio. Water Resour. Manag. 27, 2483–2499. Skidmore, E.L., 1982. Soil loss tolerance. In: Determinants of Soil Loss Tolerance. American Society of Agronomy, ASA Special Publication 45pp. 87–94. Soil Science Society of America, 1997. Glossary of Soil Science Terms. (Madison, WI). Sparovek, G., De Maria, I.C., 2003. Multiperspective analysis of erosion tolerance. Sci. Agric. 60, 409–416. Sparovek, G., Jong van Lier, Q., 1997. Definition of tolerable soil erosion values. Rev. Bras. Cienc. Solo. Viçosa 21, 467–471. Sudhishri, S., Kumar, A., Singh, J.K., Dass, A., Nain, A.S., 2014. Erosion tolerance index under different land use units for sustainable resource conservation in a Himalayan watershed using remote sensing and geographic information system (GIS). Afr. J. Agric. Res. 9 (41), 3098–3110. USDA, 2000. Summary Report: 1997 National Resources Inventory (Revised December 2000). USDA, Washington, DC. Vagen, T.G., Shepherd, K.D., Walsh, M.G., 2006. Sensing landscape level change in soil fertility following deforestation and conversion in the highlands of Madagascar using Vis-NIR spectroscopy. Geoderma 133, 281–294. Verheijen, F.G.A., Jones, R.J.A., Rickson, R.J., Smith, C.J., 2009. Tolerable versus actual soil erosion rates in Europe. Earth Sci. Rev. 94, 23–38. Verstraeten, G., Poesen, J., de Vente, J., Koninckx, X., 2003. Sediment yield variability in Spain: a quantitative and semi-qualitative analysis using reservoir sedimentation rates. Geomorphology 50 (4), 327–348. Walling, D.E., 2000. Linking land use, erosion and sediment yields in river basins. Hydrobiologia 410, 223–240. Wei, H., Nearing, M.A., Stone, J.J., Guertin, D.P., Spaeth, K.E., Pierson, F.B., Nichols, M.H., Moffet, C.A., 2009. A new splash and sheet erosion equation for rangelands. Soil Sci. Soc. Am. J. 73, 1386–1392. Williams, J.R., 1995. The EPIC model. In: Singh, V.P. (Ed.), Computer Models of Watershed Hydrology. Water Resources Publications, Highlands Ranch, CO, pp. 909–1000. Zarea, A., Koocheki, A., Nassiri, M., 2007. Trend analysis of yield, production and cultivated area of cereal in Iran during the last 50 years and prediction of future situation. Agron. Res. Iran 4 (1), 49–69 (in Persain, with english abstract).
Ahmadi, Ilkhchi A., Hajabbassi, M.A., Jalalian, A., 2003. Effects of converting range to dry-farming land on runoff and soil loss and quality in Dorahan, Chaharmahal & Bakhtiari Province (in persian). Iran. J. Sci. Tech. Agric. Natural Resour. 6 (4), 103–114. Alewell, C., Egli, M., Meusburger, K., 2015. An attempt to estimate tolerable soil erosion rates bymatching soil formation with denudation in Alpine grasslands. J. Soils Sediments 15, 1383–1399. Alin, S.R., O'Reily, C.M., Cohen, A.S., Dettman, D.L., Palacios-Fest, M.R., Mckee, B.A., 2002. Effects of land use change on aquatic biodiversity: a view from the Paleorecord at Lake Tanganyika, East Africa. Geology 30, 1143–1146. Arnold, J.G., Srinivasan, R., Muttiah, R.S., Williams, J.R., 1998. Large area hydrologic modeling and assessment part I: model development. J. Am. Water Res. Assoc. 34 (1), 73–89. Bazzoffi, B., 2009. Soil erosion tolerance and water runoff control: minimum environmental standards. Reg. Environ. Chang. 9, 169–179. Bhattacharyya, P., Bhatt, V.K., Mandal, D., 2008. Soil loss tolerance limits for planning of soil conservation measures in Shivalik–Himalayan region of India. Catena 73, 117–124. Bilotta, G.S., Brazier, R.E., 2008. Understanding the influence of suspended solids on water quality and aquatic biota. Water Res. 42 (12), 2849–2861. Caravaca, F., Lax, A., Albaladejo, J., 2004. Aggregate stability and carbon characteristics of particle-size fractions in cultivated and forested soils of semiarid Spain. Soil Tillage Res. 78, 83–90. Clark, E.H., 1987. Soil erosion: Offsite environmental effects. In: Harlin, J.M., Bernardi, G.M. (Eds.), Soil Loss: Processes, Policies, and Prospects. Westview, New York, pp. 59–89. Cole, G.W., Higgins, J.J., 1985. A probability criterion for acceptable soil erosion. Trans. ASAE 28, 1921–1926. Correll, D.L., 1998. The role of phosphorus in the eutrophication of receiving waters: a review. J. Environ. Qual. 27 (2), 261–266. Dahms, D., Favilli, F., Krebs, R., Egli, M., 2012. Soil weathering and accumulation rates of oxalate-extractable phases from alpine chronosequences of up to 1 Ma in age. Geomorphology 151–152, 99–113. de Vente, J., Poesen, J., Verstraeten, G., Van Rompaey, A., Govers, G., 2008. Spatially distributed modelling of soil erosion and sediment yield at regional scales in Spain. Glob. Planet. Chang. 60 (3–4), 393–415. Delgado, F., Lopez, R., 1998. Evaluation of soil degradation impact on the productivity of Venezuelan soils. Adv. GeoEcology 31, 133–142. den Biggelaar, C., Lal, R., Wiebe, K., 2003. The global impact of soil erosion on productivity II: effects on crop yields and production over time. Adv. Agronomy 81, 49–95. Du, S., Chen, A., Liu, G., 2013. Determination of purple soil loss tolerance based on soil productivity in southwest China. J. Soil Water Conserv. 68 (2), 146–152. Duan, X., Shi, X., Li, Y., Rong, L., Fen, D., 2017. A new method to calculate soil loss tolerance for sustainable soil productivity in farmland. Agron. Sustain. Dev. 37, 2. Duan, X., Xie, Y., Feng, Y.J., Yin, S.Q., 2009. Study on the method of soil productivity assessment in black soil region of Northeast China. Agric. Sci. China 8, 472–481. Duan, X., Xie, Y., Liu, B., Liu, G., Feng, Y., GAO, X., 2012. Soil loss tolerance in the black soil region of Northeast China. J. Geogr. Sci. 22 (4), 737–751. Duan, X., Xie, Y., Ou, T., Lu, H., 2011. Effects of soil erosion on long-term soil productivity in the black soil region of northeastern China. Catena 87, 268–275. Eggermont, H., Verschuren, D., 2003. Impact of soil erosion in disturbed tributary drainages on the benthic invertebrates fauna of Lake Tanganyika. East Afr. Biol. Conserv. 113, 99–109. Emadodin, I., Bork, H.R., 2011. Degradation of soils as a result of long-term humaninduced transformation of the environment in Iran: an overview. J. Land Use Sci. 7 (2), 203–219. Emadodin, I., Narita, D., Bork, H.R., 2012. Soil degradation and agricultural sustainability: an overview from Iran. Environ. Dev. Sustain. 14 (5), 611–625. FAO, UNDP, UNEP, 1994. In: Land degradation in south Asia: its severity, causes and effects upon the people. World Soil Resources Report no. 78, FAO, Rome, Italy. . Gee, G.W., Or, D., 2002. Particle-size analysis. In: Dane, J.H., Topps, G.C. (Eds.), Methods of Soil Analysis: Part 4. Physical Methods: Soil Science Society of America Book Series No. 5. ASA and SSSA, Madison, WI, pp. 255–293. Harrey, C.A., Pimentel, D., 1996. Effects of soil and wood depletion on biodiversity. Biodivers. Conserv. 5, 1121–1130. Humphreys, G.S., Wilkinson, M.T., 2007. The soil production function: a brief history and its rediscovery. Geoderma 139, 73–78. Jenny, H., 1941. Factors of Soil Formation, a System of Quantitative Pedology. McGrawHill, New-York. Kazemi, H., Sarvestani, Z.T., Kamkar, B., Shataei, S., Sadeghi, S., 2015. Ecological zoning for wheat production at province scale using geographical information system. Adv. Plants Agric. Res. 2 (1), 00035. Kelley, D.W., Nater, E.A., 2000. Historical sediment flux from three watersheds into Lake Pepin, Minnesota, USA. J. Environ. Qual. 29, 561–568. Kerr, J., 1997. The economics of soil degradation: from national policy to farmers’ needs. In: Soil Erosion at Multiple Scales: Principles and Methods for Assessing Causes and Impacts. CABI Publishing, Wallingford, UK, pp. 21–38. Khormali, F., Ajami, M., Ayoubi, S., Srinivasarao, C., Wani, S.P., 2009. Role of deforestation and hillslope position on soil quality attributes of loess-derived soils in Golestan province. Iran. Agric. Ecosyst. Environ. 134, 178–189. Lal, R., 1985. Soil erosion and sediment transport research in tropical Africa. Hydrol. Sci. J. 30, 239–256. Lal, R., 1998. Soil erosion impact on agronomic productivity and environment quality. Crit. Rev. Plant Sci. 4, 319–464.
9