- Email: [email protected]
Modeling Runoff from Semi-Arid Agricultural Lands in Northwest Iran
Pedosphere 24(5): 595–604, 2014 ISSN 1002-0160/CN 32-1315/P c 2014 Soil Science Society of China Published by Elsevier B.V. and Science Press
Modeling Runoff from Semi-Arid Agricultural Lands in Northwest Iran∗1 A. R. VAEZI∗2 Department of Soil Science, Agriculture Faculty, University of Zanjan, Zanjan 45371-38791 (Iran) (Received September 17, 2013; revised June 30, 2014)
ABSTRACT Transformation of rainfall into runoff over an area is a very complex process which exhibits both temporal and spatial variability; runoff in a defined area can be affected by factors such as topography, vegetation, rainfall characteristics and soil properties. This study was conducted to develop an empirical model using the rainfall characteristics and soil properties for predicting runoff from dry-farming lands in a semi-arid agricultural area in Hashtroud, Northwest Iran. Runoff plots (1.83 m × 22.1 m) in triplicate were installed in thirty-six sloped dry-farming lands in the study area. Runoff under natural rainfalls was measured in each plot during a 2-year period. The results showed that runoff for 41 runoff-producing rainstorm events with duration longer than 30 min was largely associated with a rainfall index obtained by multiplying the positive square root of rainfall depth (h0.5 ) by the logarithm of the maximum 30-minute intensity (LogI30 ) (R2 = 0.81). Runoff significantly varied among the plots (P < 0.001), which was considerably related to the effective soil properties (R2 = 0.74), i.e., soil permeability (Per) and aggregate stability (AS). A multiple linear regression model was developed between runoff and the rainfall index (h0.5 logI30 ) and the effective soil properties (AS and Per). Evaluation of the model using 34 runoff-producing rainstorm events that occurred during the next two years resulted in high values of the efficiency coefficient and R2 (0.88 and 0.91, respectively), which revealed that the model developed in this study could be used in predicting runoff from the dry-farming lands in the semi-arid regions. Key Words:
aggregate stability, dry-farming lands, natural rainfall, rainfall depth, rainfall index, rainfall intensity, soil permeability
Citation: Vaezi, A. R. 2014. Modeling runoff from semi-arid agricultural lands in Northwest Iran. Pedosphere. 24(5): 595–604.
INTRODUCTION Runoff generation is the original driving force of soil erosion (Le Bissonnais et al., 2005) and nutrient movement from soil surface (Zeng et al., 2008), particularly in arid and semi-arid regions. Almost 39% of land surface of Iran (642 797 km2 ) has a semi-arid climate, with an annual precipitation between 200 and 500 mm. In these regions, about 33% of the annual precipitation losses as surface flows (Alizadeh, 2003). In many places, agriculture is mostly performed under rainfed condition and so crop production is wholly dependent on storage of rain water in the soil. Prevention of runoff generation in these regions is essential to conserving soil productivity and supplying water for crop production. There is a strong demand to develop an accurate and easily used model that can appropriately simulate the process of runoff generation (Lin and Wang, 2007). Many such models have recently been developed (Ma et al., 2009). Nevertheless, modeling runoff ∗1 Supported
by the University of Zanjan, Iran. author. E-mail: [email protected].
∗2 Corresponding
in semi-arid regions is also a challenging task because many of the hydrological models developed for humid areas are tuned to a saturation excess mechanism and not to the infiltration excess mechanism that often dominates in dry regions (Faur`es et al., 1995). The process of rainfall transformation into runoff over a catchment is very complex, highly nonlinear, with both temporal and spatial variability (ASCE, 2000). Runoff occurs more commonly in arid and semiarid regions when rainfall intensities are higher than infiltration capacity of surface soil. Over the last two decades, a large body of knowledge has been built up about the processes of runoff generation in the semiarid regions (Yair and Lavee, 1985; Abrahams et al., 1988; Mart´ınez-Mena et al., 1998). These studies show that the runoff-controlling factors in semi-arid catchments are different from those which regulate the hydrology of wetter environments. Runoff generation in semi-arid regions is dominated by an infiltration excess mechanism with a short time to final infiltration rates and a fast response due to steep hillslopes with shallow
596
A. R. VAEZI
soils and lack of vegetation (Greenbaum et al., 2006). Also, surface conditions such as soil crusting and rock pavement (Sol´e-Benet et al., 1997) are the most relevant factors in these regions. Runoff generation in a watershed can be influenced by the physical and hydrological parameters of the watershed, topography, vegetation cover, soil properties and rainfall characteristics. Under the same conditions of topography and climate in the watersheds, runoff generation in different hillslopes is solely affected by the rainfall characteristics and soil properties. Many studies have been performed on the relationship between runoff and rainfall characteristics (Jacquin and Shamseldin, 2006; Al-Qurashi et al., 2008; Bahat et al., 2009). Studies on the relationship between runoff and soil properties have been mainly associated with the influences of antecedent soil moisture (Wei et al., 2007; Mart´ınez-Murillo et al., 2013), rock fragment (Cerd` a, 2001) and soil management systems (G´omez et al., 2009) on the runoff generation. More studies done in Iran have focused on the effect of land use change on runoff generation (Saadati et al., 2006), application of the hydrological models in predicting runoff (Rostamian et al., 2008), effect of geomorphologic properties of the watersheds on runoff generation (Abdollahi et al., 2003) and effect of soil particles and surface gravel on runoff generation (Javadi et al., 2004). However, there is no quantita-
Fig. 1
tive study to model runoff in dry-farming lands of the semi-arid regions in Iran. Therefore, the objectives of this study were to quantify the influences of the rainfall characteristics (intensity, depth, etc.) and soil physicochemical properties on the runoff generation and to develop an empirical model to predict runoff in dryfarming lands of the semi-arid regions. MATERIALS AND METHODS Study area The study was carried out in Hashtroud Township, located in the southern part of East Azarbijan Province, Northwest Iran, from March 2005 to March 2007. The study area was an agricultural zone of 900 km2 in area (30 km × 30 km) between 37◦ 18 49 and 37◦ 35 0 N, and 46◦ 46 5 and 47◦ 6 5 E (Fig. 1) in the Gharrangouchai Watershed. The Gharrangouchai Watershed is the largest watershed in Hashtroud, with an area of 2 655 km2 and a length of 190 km. The latitude in the study area ranges from 1 570 to 1 660 m. The climate is semi-arid with an average annual precipitation of 322 mm and a mean annual temperature of 13 ◦ C. Precipitations mostly occurred in the winter, late autumn and early spring. Agricultural soils are mostly located in 3◦ –9◦ slopes and cover about 70% of the land surface. The soils, weakly developed in the profile and often classified as chestnut soils, have a low organic
Location of the study area, rainfall gauge stations and plots used for measuring surface runoff.
MODELING RUNOFF FROM AGRICULTURAL LANDS
597
matter content (11 g kg−1 ) and are mainly calcareous with a moderate amount of total carbonates (Hakimi, 1986). They are usually cultivated up to down slopes to wheat under rainfed condition. Field observations showed that cultivation in slope direction was a main factor in producing surface runoff and so declining crop productivity in dry farming lands of the study area.
a commercial tipping bucket rain gauge which provides a switch-closure output for each 0.01 inch of rainfall. The maximum 30-min intensity (I30 ) (Wischmeier and Smith, 1978) was calculated using the rain gauge data recorded in the 2-year period.
Runoff plots
Some soil properties including independent properties (mineral particles, gravel, organic matter, lime and potassium) and dependent properties (initial and final infiltration rate and aggregate stability) were investigated in order to determine the influences of soil properties on runoff. Soil samples (0–30 cm depth) were taken randomly from three locations within each plot before plowing to determine the independent properties. Then, the samples of each plot were mixed, dried, grounded to pass a 2-mm sieve and stored in sealed polyethylene bags in a cool and dry place until the chemical analysis in the laboratory. The particle size distribution including coarse sand (0.1–2 mm), very fine sand (0.05–0.1 mm), silt (0.002–0.05 mm) and clay (< 0.002 mm) was determined by the Robinson’s pipette method (Avery and Bascomb, 1982). Gravel (2–8 mm) was determined by the weighting method (Gee and Bauder, 1986). The total soil organic carbon was measured by the Walkley-Black wet dichromate oxidation method (Nelson and Sommers, 1982) and converted to organic matter through multiplying it by 1.724. To determine the lime content, the calcium carbonate equivalent (total carbonates) was measured as the loss in mass of soil upon addition of HCl solution (Goh et al., 1993). The aggregate stability and infiltration rate were determined in the plots immediately after plowing. The mean weight diameter (MWD) of the water-stable aggregates was determined using the wetsieving method (Angers and Mehuys, 1993) by placing 100 g aggregates with diameter larger than 6 mm on the top of a sieve set, which was then moved up and down in a water cylinder for one minute. The initial and final steady infiltration rates were determined by measuring the one-dimensional water flow into the soil per unit time by a double-ring infiltrometer (Bouwer, 1986) in four to six replications for each plot. The initial infiltration rate was determined using the fall of water level in the inner ring (30 cm in diameter) 30 s after the start of the measurement. When the amount of water entering into the soil was fairly constant over time for three consecutive measurements, the final infiltration rate was calculated and assumed as soil permeability. The infiltration measurements were carried out at the end of the dry season (in July 2005) in order to exclude the influence of different initial moisture
To install runoff plots, the study area was divided into thirty six grids with a dimension of 5 km × 5 km (Fig. 1). In each grid, a dry-farming land located in a south 5.14◦ slope and under fallow condition was selected. Length of the lands varied from 38.4 to 87.2 m, with an average of 52.13 m. The lands were plowed in slope direction in accordance with the common plow direction in the area to facilitate runoff generation and harrowed to make a uniform slope in February 2005. Three unit plots with 1.83-m width, 22.1-m length (Wischmeier and Smith, 1978) and 1.2-m spacing were installed with 10 m distance from slope boundary in each land on March 2005. Runoff-collecting installations consisting of gutter pipes, pipes and tanks were established at the lower parts of the plots. After each natural rainfall producing runoff, total volume of the tank content (runoff and sediment) in a total of 108 plots was measured in the field. The tank content was then mixed thoroughly and a 0.5-kg sample was taken for determination of runoff volume. The runoff volume was determined after filtering the sample in the lab. Runoff generation (m3 ) of each plot was determined by multiplying the total volume of the tank content by the runoff volume proportion in the sample. Annual surface runoff was also computed from summation of total surface runoff produced under different natural rainstorms during the 2-year study period. Mean runoff produced in each plot under all rainstorms was also calculated to determine the relationship between runoff and rainfall characteristics. Determination of rainfall characteristics Rainfall duration and depth data were taken from five rainfall gauge stations located in the study area. Four standard rainfall gauges located in grids 2, 10, 26 and 30 of the study area were used to manually measure the depth of rainfall during the study period (Fig. 1). These rain gauges consisted of a measuring tube inside an overflow canister and a screened collection funnel on top. An automatic rain gauge, belonging to the Irrigation Office of Hashtroud, Iran, located in grid 17, was also used to determine rainfall intensity. The automatic rain gauge consisted of
Determination of soil properties
598
A. R. VAEZI
contents as described by Turner and Sumner (1978).
Runoff production
Statistical analysis and computation
Runoff generation in the plots largely varied from 0.04 to 1.22 mm due to the variation of the rainfall characteristics (Table I). The runoff coefficient (runoff/rainfall ratio) was on average 0.06 mm mm−1 for all the plots, which confirmed the high infiltration capacity of the soils during the rainfall events (Augeard et al., 2005). Botha et al. (2003) also showed that in the semi-arid areas with fine textured soils, runoff can vary between 8% and 49% of the annual rainfall depending on the prevailing conditions. As shown in Fig. 2, there was a weak correlation between runoff and rainfall intensity (R2 = 0.41). This result was contrary to that of Morin and Cluff (1980), who showed that the rainfall intensity was one of the most important factors influencing runoff in the semiarid areas. Significant correlations were observed between runoff and the rainstorm depth (R2 = 0.68) as well as the maximum 30-min intensity of rainstorms (R2 = 0.70). Nevertheless, runoff generation was strongly related to the rainfall index which was obtained by multiplying the positive square root of rainfall depth by logarithm of the maximum 30-minute intensity (R2 = 0.81). In other words, the rainfall index can be stated as a better index to explain runoff generation potential of the rainstorms in the study area. Findings of Dunj´ o et al. (2004) also showed that runoff generation in different land uses significantly correlated with the rainfall depth (R2 of 0.77–0.96). According to Wischmeier and Smith (1978), the maximum 30minute intensity along with the kinetic energy is one of the most important indices of the rainstorms to estimate soil loss in the runoff plots. Mean annual surface runoff produced in the dryfarming lands varied from 3.39 to 11.92 mm year−1 , with an average of 8.09 mm year−1 (Table III). Runoff significantly varied among the dry-farming lands in the study area (P < 0.001). Since there was no significant difference in the runoff-producing rainstorm depth among the rain gauge stations, the spatial distribution of the rainstorms was uniform in the study area. Thus, differences in runoff generation among the lands were directly related to their soil properties.
Data were assessed for normality using the Kolmogorov-Smirnov test. Differences in runoff among the plots in the area were analyzed using the parametric Duncan’s test. The most effective characteristics of rainfalls in runoff generation were determined by the determination coefficients (R2 ) between mean runoff and rainfall characteristics. Soil properties influencing runoff was determined by the correlation coefficient (r) using Pearson’s method (Sokal and Rohlf, 1981). A stepwise multiple regression analysis was utilized to determine the relationship between runoff and the effective soil properties. Finally, a linear model was developed using the effective soil properties and the most effective rainfall characteristics. To evaluate the capability of the model, comparisons between the measured and estimated runoff were done with R2 and the 1:1 line. Also, the model efficiency coefficient (E), which is commonly used as a measure of model performance in hydrology, was determined according to Nash and Sutcliffe (1970) as follows: n (Qpre,i − Qobs,i )2 E = 1 − i=1 (1) n (Qobs,i − Qobs )2 i=1
where n is the total number of observations; Qobs,i is the ith observed runoff depth; Qobs is the mean of observed (measured) runoff depth and Qpre,i is the ith predicted runoff depth. RESULTS AND DISCUSSION Rainfall characteristics Ninety seven natural rainfalls occurred in the study area during the 2-year study period. Table I shows the mean characteristics of the rainfall events from March 2005 to March 2007. Forty one rainstorms produced runoff and sediment in the runoff plots in the study area (Table I). The duration of the rainstorms was higher than 30 min and their intensity varied from 2.11 to 13.78 mm h−1 , with an average of 4.82 mm h−1 . The mean depths of the storms in the rain gauge stations located in grids 2, 10, 17, 27 and 30 were 7.22, 6.59, 6.98, 7.14 and 6.84 mm, respectively. As shown in Table II, there was no significant difference in the rainstorm depth among different rain gauge stations (F = 0.027, P = 0.994). Thus, the spatial distribution of the rainstorms was uniform and runoff variations in all the plots were directly associated with soil properties.
Soil properties As shown in Table IV, the soils of the study area were mainly clay loam having 36.7% sand, 31.6% silt and 32.0% clay, on average. They had a low organic matter content (1.1%) and were calcareous (limy) with 13% lime. Mean values of gravel in the soils were 10%. Soil aggregates were mainly granular with a mean di-
MODELING RUNOFF FROM AGRICULTURAL LANDS
599
TABLE I Characteristics of the runoff-producing rainstorms and mean runoff produced during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran Rainstorm event No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Rainstorm characteristics
Mean runoff
Duration
Depth
Intensity
Maximum 30-min intensity
h 1.1 1.4 3.4 1.0 1.3 1.1 6.9 0.7 1.5 0.7 0.7 1.1 1.2 0.9 1.6 2.1 1.3 0.5 0.8 1.4 0.6 0.6 4.0 0.8 1.7 3.2 1.6 1.5 1.2 1.8 2.4 1.4 1.4 1.9 1.8 0.5 3.5 0.5 0.5 2.1 1.8
mm 2.5 3.6 13.7 2.7 4.8 3.7 17.8 2.8 8.4 2.0 2.5 4.2 11.9 12.4 8.1 12.5 10.4 3.5 1.9 15.3 4.0 2.4 9.3 5.3 4.2 6.7 12.7 4.2 3.3 5.6 8.1 4.0 3.4 4.8 6.8 4.1 18.7 4.6 2.0 14.3 8.1
2.2 2.7 4.0 2.7 3.7 3.4 2.6 4.0 5.6 2.8 3.4 3.6 10.1 13.8 5.1 5.9 8.0 7.0 2.5 11.1 6.1 4.1 2.3 6.3 2.5 2.1 7.9 2.8 2.6 3.6 3.4 2.9 2.5 2.6 3.8 8.2 5.3 9.2 4.0 6.6 4.6
mm h−1 3.0 3.2 15.2 3.0 4.8 5.4 7.6 5.4 8.4 3.8 4.8 5.0 21.8 22.8 12.5 13.0 12.2 7.0 3.6 22.4 6.8 4.6 4.4 8.2 5.2 4.2 14.5 5.0 4.0 6.0 7.4 4.2 4.0 7.6 6.6 8.2 13.0 9.2 4.0 12.4 9.6
ameter of 5 mm. The water-aggregate stability of the soils was very low with the mean weight diameter (MWD) values ranging between 0.27 and 1.91 mm. The soil permeability (final infiltration rate) values varied between 1.4 and 5.8 cm h−1 , with an average of 3.5 cm h−1 . The initial infiltration rate was in the range between 60 and 81.3 cm h−1 . Statistical distributions of the soil properties were normal. Soil properties affecting runoff Runoff significantly correlated with coarse sand, silt, organic matter, lime, aggregate stability, initial
mm 0.05 0.11 0.80 0.10 0.23 0.18 0.62 0.31 0.41 0.08 0.23 0.30 0.98 1.22 0.70 0.56 0.72 0.36 0.06 0.91 0.58 0.22 0.60 0.13 0.08 0.09 0.68 0.17 0.15 0.37 0.50 0.35 0.19 0.27 0.39 0.04 0.75 0.20 0.04 0.82 0.58
infiltration rate and soil permeability (Table V). The soil permeability was the most important soil property controlling runoff in the study area (r = −0.78, P < 0.001). In fact, infiltration excess of rainfall was the predominant mechanism for runoff generation in the runoff plots (Buda et al., 2009). Findings of G´ omez et al. (2001) also showed that approximately 50% of variability of runoff in fallow plots can be explained by the final infiltration rate. Many have noted that mechanisms closely related to the runoff generation is the infiltration capacity of the soils. Indeed, infiltration capacity, which was determined based on the final infi-
600
A. R. VAEZI
TABLE II Analysis of variance on the depths of the runoff-producing rainstorms during the period from March 2005 to March 2007 at the rain gauge stations in the runoff plots installed in the dryfarming lands of Northwest Iran Location of the rain gauge station in the study areaa)
Mean rainstorm depth
Grid Grid Grid Grid Grid
mm 7.15 6.77 6.98 7.08 6.82
2 10 17 26 30
F value
P value
0.027
0.994
a) The
study area was divided into thirty six grids with a dimension of 5 km × 5 km.
ltration rate, was the most important factor controlling runoff in the soils (Morin and Cluff, 1980; Roth, 2004). The aggregate stability had better correlations with the initial infiltration rate (r = 0.61, P < 0.01) than with the soil permeability (r = 0.13). With an increase in the aggregate stability, runoff strongly decreased (r = −0.47, P < 0.01). Findings of Cerd` a (1996) and Barth`es and Roose (2002) showed that the aggregate stability is a relevant indicator of runoff
in semi-arid environments. In soils with high percentages of the water-stable aggregates, the soil surface can hardly be disrupted by raindrop impacts, resulting in reduced soil surface sealing. Therefore, rainwater infiltration into soil increased particularly at the initial stage of rainfall, which accords with the result of Lal and Shukla (2004), who noted that soils having a poor structure lead to surface sealing of pores and crusting and consequently lower infiltration and greater runoff. Augeard et al. (2007) also showed that soil surface sealing drastically reduces infiltration in bare soils exposed to rainfall and subsequently affects runoff generation and soil erosion. Runoff model The multiple regression analysis showed that mean runoff was significantly correlated with the final infiltration rate and aggregate stability (R2 = 0.74, P < 0.001) (Table VI). With an increase in these soil properties, the runoff generation considerably decreased. A linear regression relationship between runoff (mm) and the effective soil properties was extracted from the multiple regression analysis as follows: Runoff = 0.728 − 0.066Per − 0.088AS
(R2 = 0.74)(2)
Fig. 2 Relationships between runoff and rainfall characteristics, including rainfall depth (h), rainfall intensity (I), rainfall maximum 30-minute intensity (I30 ) and rainfall index obtained by multiplying the positive square root of rainfall depth (h0.5 ) by the logarithm of I30 (logI30 ), during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran.
MODELING RUNOFF FROM AGRICULTURAL LANDS
601
TABLE III Runoff generated for the rainstorms during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran Plot No.
Runoff Mean
Plot No. Standard deviation
Runoff Mean
year−1
1 2 3 4 5 6 7 8 9 10 11 12
9.38 9.50 9.14 8.81 11.38 4.20 6.93 5.43 10.47 10.84 11.92 5.99
Plot No. Standard deviation
Runoff Mean
Standard deviation
10.35 6.59 4.92 11.06 9.12 8.65 6.67 3.39 6.87 7.20 8.13 8.19
mm year−1 0.51 0.14 0.59 0.25 0.33 0.40 0.33 0.19 0.35 0.41 1.00 0.55
year−1
mm 0.37 0.27 0.49 0.45 0.26 0.44 0.49 0.57 0.36 0.65 0.16 0.24
13 14 15 16 17 18 19 20 21 22 23 24
9.94 9.48 10.02 4.61 6.00 9.51 9.96 7.34 6.57 6.56 8.74 7.32
TABLE IV Some physicochemical properties of soils in the runoff plots installed in the dry-farming lands of Northwest Iran Soil property
Mean
Standard deviation
Coarse sand (%) Very fine sand (%) Silt (%) Clay (%) Gravel (%) Organic matter (%) Lime (%) Mean weight diameter of the waterstable aggregates (mm) Initial infiltration rate (cm h−1 ) Final infiltration rate (cm h−1 )
18.9 17.8 31.5 31.8 9.9 1.1 12.7 1.13
5.3 3.2 7.1 5.7 2.4 0.2 5.2 0.44
81.33 3.5
14.37 1.2
where Per is the soil permeability (final infiltration rate, cm h−1 ) and AS is the aggregate stability (mean
mm 0.71 0.44 0.06 0.52 0.27 0.42 0.50 0.24 0.24 0.59 0.21 0.53
25 26 27 28 29 30 31 32 33 34 35 36
weight diameter of the water-stable aggregates, mm). The aggregate stability and final infiltration rate were strongly affected by some independent soil properties (mineral particles, organic matter and lime). Contrary to silt, coarse sand, organic matter and lime positively affected the soil permeability and consequently reduced runoff in the soils. The effect of coarse sand in reducing runoff agrees with Santos et al. (2003), who found that due to presence of macropores in the sandy soils, the rate of water entering the soil is higher and generation of runoff is lower as compared with the fine-textured soils. Although silt particles considerably decreased the soil permeability, they did not affect the aggregate stability. Despite the fact that clay particles can be absorbed on sand and silt fractions and enhance stability of aggregates (Lal and Shukla, 2004), their effects on runoff were not considerable due to the negligible role in the soil permeability. The effect of
TABLE V Correlation matrix between runoff and some physicochemical properties of the soils in the runoff plots installed in the dry-farming lands of Northwest Iran Variablea)
CS
VFS
Silt
Clay
Gravel
OM
Lime
AS
Per
IR
Runoff
CS VFS Silt Clay Gravel OM Lime AS Per IR Runoff
1 0.22 −0.74*** −0.18 0.03 0.27 −0.01 −0.17 0.76*** −0.06 −0.56**
1 −0.20 −0.50** −0.01 −0.31* −0.56** −0.67** −0.05 −0.34* 0.23
1 −0.40* 0.02 −0.23 0.17 −0.12 −0.55** −0.06 0.42*
1 −0.06 0.21 0.03 0.70*** −0.07 0.41* −0.09
1 0.16 −0.03 −0.09 0.09 −0.36* −0.09
1 0.05 0.29* 0.54** 0.44** −0.56**
1 0.48** 0.29* 0.32* −0.39*
1 0.13 0.61** −0.47**
1 0.19 −0.78***
1 −0.38*
1
*, **, *** Significant at P < 0.05, P < 0.01 and P < 0.001, respectively. a) CS = coarse sand; VFS = very fine sand; OM = organic matter; AS = aggregate stability (mean weight diameter of the water-stable aggregates, wet-sieving method); Per = soil permeability (final infiltration rate); IR = initial infiltration rate.
602
A. R. VAEZI
TABLE VI Multiple regression analysis of the relationship between mean runoff and soil properties in the runoff plots installed in the dry-farming lands of Northwest Iran (R2 = 0.74, P < 0.001) Model variablea)
Constant Per As a) AS
Unstandardized coefficients
Standardized coefficient
Model coefficient
Standard error
0.728 −0.066 −0.088
0.036 0.008 0.021
−0.728 −0.369
t level
P level
20.014 −8.084 −4.101
0.001 0.001 0.001
= aggregate stability (mean weight diameter of the water-stable aggregates); Per = soil permeability (final infiltration rate).
gravel on runoff also was not considerable, contrary to Mathys et al. (2005), who indicated that infiltration rate is increased by the gravel cover of soil surface in the Black Marls of the French Alps. Organic matter has been recognized as an important binding and bridging agent in enhancing the soil structural stability and infiltration capacity (Yu et al., 2006) and in consequence reducing runoff (Fern´ andez et al., 2006; Zhang et al., 2007). Lime was also shown to be an important factor controlling runoff in the soil. As shown by Vitro et al. (2011) in soils of semi-arid regions, aggregation is affected by carbonate contents. In fact, calcium linkage is a mechanism in the formation of water-stable aggregates and the increase in infiltration rate (Pepper and Morrissey, 1985). A linear empirical model was developed using the linear relationships between runoff (mm) and the effective soil properties, Eq. 2, and the most effective rainfall characteristics (rainfall index) obtained by multiplying the positive square root of rainfall depth (mm) (h0.5 ) by the logarithm of the maximum 30-minute intensity (mm h−1 ) (logI30 ) as follows: Runoff = (1.852 − 0.168Per − 0.224AS)× (0.198h0.5 logI30 − 0.056)
(R2 = 0.81) (3)
Fig. 3 Estimated v.s. observed runoff under the thirty-four rainstorm events during the period from March 2007 to March 2009 in the runoff plots installed in the dry-farming lands of Northwest Iran.
duration rainstorm events, which are known to have high rainfall depths and maximum 30-min intensity. The model efficiency coefficient was high (0.88), which showed that the model predictions could be acceptable in the study area. Use of the model could determine the susceptibility of different soils to runoff generation under different rainfalls in the study area and other regions of similar climate, topography and soil.
Model evaluation CONCLUSIONS To evaluate the model, field runoff measurements in the runoff plots in the study area continued for two years from March 2007 to March 2009. Thirtyfour rainstorm events lasting more than 30 min caused runoff for all the plots in the study area and runoff was also estimated by the model developed for each rainstorm event. As shown in Fig. 3, there was a relatively high correlation between the estimated and observed runoff (R2 = 0.91). In fact, 91% of the runoff variation in the soils under natural rainfalls could be explained by the model developed. The model slightly overestimated runoff, particularly for some large rainstorm events (Fig. 3). The negligible overestimation of runoff by the model was observed in some intensive short-
Only rainfalls with duration longer than 30 min could produce runoff in a two-year period in the runoff plots installed in the study area. Runoff generation by the rainstorms was largely associated with the rainfall index obtained by multiplying the positive square root of rainfall depth by the logarithm of the maximum 30-minute intensity. Water infiltration rate and aggregate stability were the most important properties of the soils which remarkably affected the runoff generation. A multiple linear regression model was developed between runoff and the rainfall index and the effective soil properties (i.e., water infiltration rate and aggregate stability), with R2 = 0.81. Model evaluation using
MODELING RUNOFF FROM AGRICULTURAL LANDS
thirty-four rainstorm events which generated runoff in the next two years showed high efficiency coefficient and R2 values of 0.88 and 0.91, respectively, which suggested that the model predictions were acceptable in the study area. Therefore, the model could be reliably used to predict runoff in the fallow sloped dry-farming lands in the study area and other similar semi-arid areas. REFERENCES Abdollahi, K. H., Sadeghi, S. H. R. and Farajzadeh-Asl, M. 2003. Runoff modeling based on geomorphological properties using GIS in Khanmirza Watershed, Iran (in Persian). M.S. Thesis, Tarbiat Modares University. Abrahams, A. D., Parsons, A. J. and Luk, S. H. 1988. Hydrologic and sediment responses to simulated rainfall on desert hillslopes in southern Arizona. Catena. 15: 103–117. Alizadeh, A. 2003. Applied Hydrology (in Persian). 3rd Ed. Asetan-e Ghods-e Razavi Publication, Mashhad. Al-Qurashi, A., McIntyre, N., Wheater, H. and Unkrich, C. 2008. Application of the Kineros2 rainfall-runoff model to an arid catchment in Oman. J. Hydrol. 355: 91–105. Angers, D. A. and Mehuys, G. R. 1993. Aggregate stability to water. In Cartner, M. R. (ed.) Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton. pp. 651–657. ASCE Task Committee on Application of the Artificial Neural Networks in Hydrology (ASCE). 2000. Artificial neural networks in hydrology. I: Preliminary concepts. J. Hydraul. Eng-ASCE. 5: 115–123. Augeard, B., Assouline, S., Fonty, A., Kao, C. and Vauclin, M. 2007. Estimating hydraulic properties of rainfall-induced soil surface seals from infiltration experiments and X-ray bulk density measurements. J. Hydrol. 341: 12–26. Augeard, B., Kao, C., Chaumont, C. and Vauclin, M. 2005. Mechanisms of surface runoff genesis on a subsurface drained soil affected by surface crusting: A field investigation. Phys. Chem. Earth. A/B/C. 30: 598–610. Avery, B. W. and Bascomb, C. L. 1982. Soil Survey Laboratory Methods. Soil Survey Technical Monograph No. 6. Rothamsted Experimental Station, Harpenden. Bahat, Y., Grodek, T., Lekach, J. and Morin, E. 2009. Rainfallrunoff modeling in a small hyper-arid catchment. J. Hydrol. 373: 204–217. Barth` es, B. and Roose, E. 2002. Aggregate stability as an indicator of soil susceptibility to runoff and erosion; validation at several levels. Catena. 47: 133–149. Botha, J. J., van Rensburg, L. D., Anderson, J. J., Hensley, M., Macheli, M. S., van Staden, P. P., Kundhlande, G., Groenewald, D. G. and Baiphethi, M. N. 2003. Water Conservation Techniques on Small Plots in Semi-Arid Areas to Enhance Rainfall Use Efficiency, Food Security, and Sustainable Crop Production. WRC Report No. 1176/1/03. Water Research Commission, South Africa. Bouwer, H. 1986. Intake rate: Cylinder infiltrometer. In Klute, A. (ed.) Methods of Soil Analysis. Part I. Physical and Mineralogical Methods. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 825–844. Buda, A. R., Kleinman, P. J. A., Srinivasan, M. S., Bryant, R. B. and Feyereisen, G. W. 2009. Factors influencing surface runoff generation from two agricultural hillslopes in central Pennsylvania. Hydrol. Process. 23: 1295–1312.
603
Cerd` a, A. 1996. Soil aggregate stability in three Mediterranean environments. Soil Technol. 9: 133–140. Cerd` a, A. 2001. Effects of rock fragment on soil infiltration, interrill runoff and erosion. Eur. J. Soil Sci. 52: 1–10. Dunj´ o, G., Pardini, G. and Gispert, M. 2004. The role of land use-land cover on runoff generation and sediment yield at a microplot scale, in a small Mediterranean catchment. J. Arid Environ. 57: 99–116. Faur` es, J. M., Goodrich, D. C., Woolhiser, D. A. and Sorooshian, S. 1995. Impact of small-scale spatial rainfall variability on runoff modeling. J. Hydrol. 173: 309–326. Fern´ andez, C., Vega, J. A., Fonturbel, T., P´erez-Gorostiaga, P., Jim´ enez, E. and P´ erez, J. R. 2006. Runoff and soil erosion after rainfall simulations in burned soils. Forest Ecol. Manag. 234: S191. Gee, G. W. and Bauder, J. W. 1986. Particle-size analysis. In Klute, A. (ed.) Methods of Soil Analysis. Part I. Physical and Mineralogical Methods. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 383–411. Goh, T. B., Arnaud, R. J. St. and Mermut, A. R. 1993. Carbonates. In Cartner, M. R. (ed.) Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton. pp. 177–185. G´ omez, J. A., Nearing, M. A., Gir´ aldez, J. V. and Alberts, E. E. 2001. Analysis of sources of variability of runoff volume in a 40 plot experiment using a numerical model. J. Hydrol. 248: 183–197. G´ omez, J. A., Sobrinho, T. A., Gir´ aldez, J. V. and Fereres, E. 2009. Soil management effects on runoff, erosion and soil properties in an olive grove of Southern Spain. Soil Till. Res. 102: 5–13. Greenbaum, N., Ben-Zvi, A., Haviv, I. and Enzel, Y. 2006. The hydrology and paleohydrology of the Dead Sea tributaries. In Enzel, Y., Agnon, A. and Stein, M. (eds.) New Frontiers in Dead Sea Paleoenvironmental Research. Geological Society of America, Inc., Boulder. pp. 63–93. Hakimi, A. 1986. The Briefly Study of Soil Science in Hashtrood (in Persian). In Soil and Water Research Institute, Agriculture Ministry, Iran (ed.) Research Report No. 767. Soil and Water Research Institute, Agriculture Ministry, Tehran. pp. 2–15. Jacquin, A. P. and Shamseldin, A. Y. 2006. Development of rainfall-runoff models using Takagi-Sugeno fuzzy inference systems. J. Hydrol. 329: 154–173. Javadi, P., Rohipour, H. and Mahboubi, A. A. 2004. Role of gravel cover on runoff and erosion rate using rainfall simulator. Iran. J. Rangeland Desert Res. (in Persian). 3: 279– 287. Lal, R. and Shukla, M. K. 2004. Principle of Soil Physics. Marcel Dekker, Inc., New York. Le Bissonnais, Y., Daroussin, J., Jamagne, M., Lambert, J. J., Le Bas, C., King, D., Cerdan, O., L´eonard, J., Bresson, L. M. and Jones, R. 2005. Pan-European soil crusting and erodibility assessment from the European soil geographical database using pedotransfer rules. Adv. Environ. Monit. Model. 2: 1– 15. Lin, G. F. and Wang, C. M. 2007. A nonlinear rainfall-runoff model embedded with an automated calibration method— Part 1: The model. J. Hydrol. 341: 186–195. Ma, T., Zhou, C. H. and Cai, Q. G. 2009. Modeling of hillslope runoff and soil erosion at rainfall events using cellular automata approach. Pedosphere. 19: 711–718. Mart´ınez-Mena, M., Albaladejo, J. and Castillo, V. M. 1998. Factors influencing surface runoff generation in a Mediterranean semi-arid environment: Chicamo watershed, SE Spain. Hydrol. Process. 12: 741–754.
604
Mart´ınez-Murillo, J. F., Nadal-Romero, E., Reg¨ u´ es, D., Cerd` a, A. and Poesen, J. 2013. Soil erosion and hydrology of the western Mediterranean badlands throughout rainfall simulation experiments: A review. Catena. 106: 101–112. Mathys, N., Klotz, S., Esteves, M., Descroix, L. and Lapetite, J. M. 2005. Runoff and erosion in the Black Marls of the French Alps: Observations and measurements at the plot scale. Catena. 63: 261–281. Morin, J. and Cluff, C. B. 1980. Runoff calculation on semi-arid watersheds using a Rotadisk Rainulater. Water Resour. Res. 16: 1085–1093. Nash, J. E. and Sutcliffe, J. V. 1970. River flow forecasting through conceptual models part I—A discussion of principles. J. Hydrol. 10: 282–290. Nelson, D. W. and Sommers, L. E. 1982. Total carbon, organic carbon, and organic matter. In Page, A. L. (ed.) Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 539–579. Pepper, R. G. and Morrissey, J. G. 1985. Soil properties affecting runoff. J. Hydrol. 79: 301–310. Rostamian, R., Jaleh, A., Afyuni, M., Mousavi, S. F., Heidarpour, M., Jalalian, A. and Abbaspour, K. C. 2008. Application of a SWAT model for estimating runoff and sediment in two mountainous basins in central Iran. Hydrolog. Sci. J. 53: 977–988. Roth, C. H. 2004. A framework relating soil surface condition to infiltration and sediment and nutrient mobilization in grazed rangelands of northeastern Queensland, Australia. Earth Surf. Proc. Land. 29: 1093–1104. Saadati, H., Gholami, S. H. A, Sharifi, F. A. S. and Ayoubzadeh, S. A. 2006. An investigation of the effect of land use change on simulating surface runoff using SWAT mathematical model (case study: Kasilian catchment area). Iran. J. Nat. Resour. (in Persian). 59: 301–313. Santos, F. L., Reis, J. L., Martins, O. C., Castanheira, N. L. and Serralheiro, R. P. 2003. Comparative assessment of infiltration, runoff and erosion of sprinkler irrigated soils. Biosyst.
A. R. VAEZI
Eng. 86: 355–364. Sokal, R. R. and Rohlf, F. J. 1981. Biometry: The Principals and Practice of Statistics in Biological Research. 4th Ed. Freeman W. H. & Co Ltd, New York. Sol´ e-Benet, A., Calvo, A., Cerd` a, A., L´ azaro, R., Pini, R. and Barbero, J. 1997. Influences of micro-relief patterns and plant cover on runoff related processes in badlands from Tabernas (SE Spain). Catena. 31: 23–38. Turner, D. P. and Sumner, M. E. 1978. The influence of initial soil moisture content on field measured infiltration rates. Water SA. 4: 18–24. Virto, I., Gartzia-Bengoetxea, N. and Fern´ andez-Ugalde, O. 2011. Role of organic matter and carbonates in soil aggregation estimated using Laser Diffractometry. Pedosphere. 21: 566–572. Wei, L. H., Zhang, B. and Wang, M. Z. 2007. Effects of antecedent soil moisture on runoff and soil erosion in alley cropping systems. Agr. Water Manage. 94: 54–62. Wischmeier, W. H. and Smith, D. D. 1978. Predicting Rainfall Erosion Losses—A Guide to Conservation Planning. Agriculture Handbook No. 537. U.S. Department of Agriculture, Washington D.C. Yair, A. and Lavee, H. 1985. Runoff generation in arid and semiarid zones. In Anderson, M. G. and Burt, T. P. (eds.) Hydrological Forecasting. John Wiley & Sons, Chichester. pp. 183–220. Yu, D. S., Shi, X. Z. and Weindorf, D. C. 2006. Relationships between permeability and erodibility of cultivated Acrisols and Cambisols in subtropical China. Pedosphere. 16: 304–311. Zeng, S. C., Su, Z. Y., Chen, B. G., Wu, Q. T. and Ouyang, Y. 2008. Nitrogen and phosphorus runoff losses from orchard soils in South China as affected by fertilization depths and rates. Pedosphere. 18: 45–53. Zhang, G. S., Chan, K. Y., Oates, A., Heenan, D. P. and Huang, G. B. 2007. Relationship between soil structure and runoff/soil loss after 24 years of conservation tillage. Soil Till. Res. 92: 122–128.
Modeling Runoff from Semi-Arid Agricultural Lands in Northwest Iran∗1 A. R. VAEZI∗2 Department of Soil Science, Agriculture Faculty, University of Zanjan, Zanjan 45371-38791 (Iran) (Received September 17, 2013; revised June 30, 2014)
ABSTRACT Transformation of rainfall into runoff over an area is a very complex process which exhibits both temporal and spatial variability; runoff in a defined area can be affected by factors such as topography, vegetation, rainfall characteristics and soil properties. This study was conducted to develop an empirical model using the rainfall characteristics and soil properties for predicting runoff from dry-farming lands in a semi-arid agricultural area in Hashtroud, Northwest Iran. Runoff plots (1.83 m × 22.1 m) in triplicate were installed in thirty-six sloped dry-farming lands in the study area. Runoff under natural rainfalls was measured in each plot during a 2-year period. The results showed that runoff for 41 runoff-producing rainstorm events with duration longer than 30 min was largely associated with a rainfall index obtained by multiplying the positive square root of rainfall depth (h0.5 ) by the logarithm of the maximum 30-minute intensity (LogI30 ) (R2 = 0.81). Runoff significantly varied among the plots (P < 0.001), which was considerably related to the effective soil properties (R2 = 0.74), i.e., soil permeability (Per) and aggregate stability (AS). A multiple linear regression model was developed between runoff and the rainfall index (h0.5 logI30 ) and the effective soil properties (AS and Per). Evaluation of the model using 34 runoff-producing rainstorm events that occurred during the next two years resulted in high values of the efficiency coefficient and R2 (0.88 and 0.91, respectively), which revealed that the model developed in this study could be used in predicting runoff from the dry-farming lands in the semi-arid regions. Key Words:
aggregate stability, dry-farming lands, natural rainfall, rainfall depth, rainfall index, rainfall intensity, soil permeability
Citation: Vaezi, A. R. 2014. Modeling runoff from semi-arid agricultural lands in Northwest Iran. Pedosphere. 24(5): 595–604.
INTRODUCTION Runoff generation is the original driving force of soil erosion (Le Bissonnais et al., 2005) and nutrient movement from soil surface (Zeng et al., 2008), particularly in arid and semi-arid regions. Almost 39% of land surface of Iran (642 797 km2 ) has a semi-arid climate, with an annual precipitation between 200 and 500 mm. In these regions, about 33% of the annual precipitation losses as surface flows (Alizadeh, 2003). In many places, agriculture is mostly performed under rainfed condition and so crop production is wholly dependent on storage of rain water in the soil. Prevention of runoff generation in these regions is essential to conserving soil productivity and supplying water for crop production. There is a strong demand to develop an accurate and easily used model that can appropriately simulate the process of runoff generation (Lin and Wang, 2007). Many such models have recently been developed (Ma et al., 2009). Nevertheless, modeling runoff ∗1 Supported
by the University of Zanjan, Iran. author. E-mail: [email protected].
∗2 Corresponding
in semi-arid regions is also a challenging task because many of the hydrological models developed for humid areas are tuned to a saturation excess mechanism and not to the infiltration excess mechanism that often dominates in dry regions (Faur`es et al., 1995). The process of rainfall transformation into runoff over a catchment is very complex, highly nonlinear, with both temporal and spatial variability (ASCE, 2000). Runoff occurs more commonly in arid and semiarid regions when rainfall intensities are higher than infiltration capacity of surface soil. Over the last two decades, a large body of knowledge has been built up about the processes of runoff generation in the semiarid regions (Yair and Lavee, 1985; Abrahams et al., 1988; Mart´ınez-Mena et al., 1998). These studies show that the runoff-controlling factors in semi-arid catchments are different from those which regulate the hydrology of wetter environments. Runoff generation in semi-arid regions is dominated by an infiltration excess mechanism with a short time to final infiltration rates and a fast response due to steep hillslopes with shallow
596
A. R. VAEZI
soils and lack of vegetation (Greenbaum et al., 2006). Also, surface conditions such as soil crusting and rock pavement (Sol´e-Benet et al., 1997) are the most relevant factors in these regions. Runoff generation in a watershed can be influenced by the physical and hydrological parameters of the watershed, topography, vegetation cover, soil properties and rainfall characteristics. Under the same conditions of topography and climate in the watersheds, runoff generation in different hillslopes is solely affected by the rainfall characteristics and soil properties. Many studies have been performed on the relationship between runoff and rainfall characteristics (Jacquin and Shamseldin, 2006; Al-Qurashi et al., 2008; Bahat et al., 2009). Studies on the relationship between runoff and soil properties have been mainly associated with the influences of antecedent soil moisture (Wei et al., 2007; Mart´ınez-Murillo et al., 2013), rock fragment (Cerd` a, 2001) and soil management systems (G´omez et al., 2009) on the runoff generation. More studies done in Iran have focused on the effect of land use change on runoff generation (Saadati et al., 2006), application of the hydrological models in predicting runoff (Rostamian et al., 2008), effect of geomorphologic properties of the watersheds on runoff generation (Abdollahi et al., 2003) and effect of soil particles and surface gravel on runoff generation (Javadi et al., 2004). However, there is no quantita-
Fig. 1
tive study to model runoff in dry-farming lands of the semi-arid regions in Iran. Therefore, the objectives of this study were to quantify the influences of the rainfall characteristics (intensity, depth, etc.) and soil physicochemical properties on the runoff generation and to develop an empirical model to predict runoff in dryfarming lands of the semi-arid regions. MATERIALS AND METHODS Study area The study was carried out in Hashtroud Township, located in the southern part of East Azarbijan Province, Northwest Iran, from March 2005 to March 2007. The study area was an agricultural zone of 900 km2 in area (30 km × 30 km) between 37◦ 18 49 and 37◦ 35 0 N, and 46◦ 46 5 and 47◦ 6 5 E (Fig. 1) in the Gharrangouchai Watershed. The Gharrangouchai Watershed is the largest watershed in Hashtroud, with an area of 2 655 km2 and a length of 190 km. The latitude in the study area ranges from 1 570 to 1 660 m. The climate is semi-arid with an average annual precipitation of 322 mm and a mean annual temperature of 13 ◦ C. Precipitations mostly occurred in the winter, late autumn and early spring. Agricultural soils are mostly located in 3◦ –9◦ slopes and cover about 70% of the land surface. The soils, weakly developed in the profile and often classified as chestnut soils, have a low organic
Location of the study area, rainfall gauge stations and plots used for measuring surface runoff.
MODELING RUNOFF FROM AGRICULTURAL LANDS
597
matter content (11 g kg−1 ) and are mainly calcareous with a moderate amount of total carbonates (Hakimi, 1986). They are usually cultivated up to down slopes to wheat under rainfed condition. Field observations showed that cultivation in slope direction was a main factor in producing surface runoff and so declining crop productivity in dry farming lands of the study area.
a commercial tipping bucket rain gauge which provides a switch-closure output for each 0.01 inch of rainfall. The maximum 30-min intensity (I30 ) (Wischmeier and Smith, 1978) was calculated using the rain gauge data recorded in the 2-year period.
Runoff plots
Some soil properties including independent properties (mineral particles, gravel, organic matter, lime and potassium) and dependent properties (initial and final infiltration rate and aggregate stability) were investigated in order to determine the influences of soil properties on runoff. Soil samples (0–30 cm depth) were taken randomly from three locations within each plot before plowing to determine the independent properties. Then, the samples of each plot were mixed, dried, grounded to pass a 2-mm sieve and stored in sealed polyethylene bags in a cool and dry place until the chemical analysis in the laboratory. The particle size distribution including coarse sand (0.1–2 mm), very fine sand (0.05–0.1 mm), silt (0.002–0.05 mm) and clay (< 0.002 mm) was determined by the Robinson’s pipette method (Avery and Bascomb, 1982). Gravel (2–8 mm) was determined by the weighting method (Gee and Bauder, 1986). The total soil organic carbon was measured by the Walkley-Black wet dichromate oxidation method (Nelson and Sommers, 1982) and converted to organic matter through multiplying it by 1.724. To determine the lime content, the calcium carbonate equivalent (total carbonates) was measured as the loss in mass of soil upon addition of HCl solution (Goh et al., 1993). The aggregate stability and infiltration rate were determined in the plots immediately after plowing. The mean weight diameter (MWD) of the water-stable aggregates was determined using the wetsieving method (Angers and Mehuys, 1993) by placing 100 g aggregates with diameter larger than 6 mm on the top of a sieve set, which was then moved up and down in a water cylinder for one minute. The initial and final steady infiltration rates were determined by measuring the one-dimensional water flow into the soil per unit time by a double-ring infiltrometer (Bouwer, 1986) in four to six replications for each plot. The initial infiltration rate was determined using the fall of water level in the inner ring (30 cm in diameter) 30 s after the start of the measurement. When the amount of water entering into the soil was fairly constant over time for three consecutive measurements, the final infiltration rate was calculated and assumed as soil permeability. The infiltration measurements were carried out at the end of the dry season (in July 2005) in order to exclude the influence of different initial moisture
To install runoff plots, the study area was divided into thirty six grids with a dimension of 5 km × 5 km (Fig. 1). In each grid, a dry-farming land located in a south 5.14◦ slope and under fallow condition was selected. Length of the lands varied from 38.4 to 87.2 m, with an average of 52.13 m. The lands were plowed in slope direction in accordance with the common plow direction in the area to facilitate runoff generation and harrowed to make a uniform slope in February 2005. Three unit plots with 1.83-m width, 22.1-m length (Wischmeier and Smith, 1978) and 1.2-m spacing were installed with 10 m distance from slope boundary in each land on March 2005. Runoff-collecting installations consisting of gutter pipes, pipes and tanks were established at the lower parts of the plots. After each natural rainfall producing runoff, total volume of the tank content (runoff and sediment) in a total of 108 plots was measured in the field. The tank content was then mixed thoroughly and a 0.5-kg sample was taken for determination of runoff volume. The runoff volume was determined after filtering the sample in the lab. Runoff generation (m3 ) of each plot was determined by multiplying the total volume of the tank content by the runoff volume proportion in the sample. Annual surface runoff was also computed from summation of total surface runoff produced under different natural rainstorms during the 2-year study period. Mean runoff produced in each plot under all rainstorms was also calculated to determine the relationship between runoff and rainfall characteristics. Determination of rainfall characteristics Rainfall duration and depth data were taken from five rainfall gauge stations located in the study area. Four standard rainfall gauges located in grids 2, 10, 26 and 30 of the study area were used to manually measure the depth of rainfall during the study period (Fig. 1). These rain gauges consisted of a measuring tube inside an overflow canister and a screened collection funnel on top. An automatic rain gauge, belonging to the Irrigation Office of Hashtroud, Iran, located in grid 17, was also used to determine rainfall intensity. The automatic rain gauge consisted of
Determination of soil properties
598
A. R. VAEZI
contents as described by Turner and Sumner (1978).
Runoff production
Statistical analysis and computation
Runoff generation in the plots largely varied from 0.04 to 1.22 mm due to the variation of the rainfall characteristics (Table I). The runoff coefficient (runoff/rainfall ratio) was on average 0.06 mm mm−1 for all the plots, which confirmed the high infiltration capacity of the soils during the rainfall events (Augeard et al., 2005). Botha et al. (2003) also showed that in the semi-arid areas with fine textured soils, runoff can vary between 8% and 49% of the annual rainfall depending on the prevailing conditions. As shown in Fig. 2, there was a weak correlation between runoff and rainfall intensity (R2 = 0.41). This result was contrary to that of Morin and Cluff (1980), who showed that the rainfall intensity was one of the most important factors influencing runoff in the semiarid areas. Significant correlations were observed between runoff and the rainstorm depth (R2 = 0.68) as well as the maximum 30-min intensity of rainstorms (R2 = 0.70). Nevertheless, runoff generation was strongly related to the rainfall index which was obtained by multiplying the positive square root of rainfall depth by logarithm of the maximum 30-minute intensity (R2 = 0.81). In other words, the rainfall index can be stated as a better index to explain runoff generation potential of the rainstorms in the study area. Findings of Dunj´ o et al. (2004) also showed that runoff generation in different land uses significantly correlated with the rainfall depth (R2 of 0.77–0.96). According to Wischmeier and Smith (1978), the maximum 30minute intensity along with the kinetic energy is one of the most important indices of the rainstorms to estimate soil loss in the runoff plots. Mean annual surface runoff produced in the dryfarming lands varied from 3.39 to 11.92 mm year−1 , with an average of 8.09 mm year−1 (Table III). Runoff significantly varied among the dry-farming lands in the study area (P < 0.001). Since there was no significant difference in the runoff-producing rainstorm depth among the rain gauge stations, the spatial distribution of the rainstorms was uniform in the study area. Thus, differences in runoff generation among the lands were directly related to their soil properties.
Data were assessed for normality using the Kolmogorov-Smirnov test. Differences in runoff among the plots in the area were analyzed using the parametric Duncan’s test. The most effective characteristics of rainfalls in runoff generation were determined by the determination coefficients (R2 ) between mean runoff and rainfall characteristics. Soil properties influencing runoff was determined by the correlation coefficient (r) using Pearson’s method (Sokal and Rohlf, 1981). A stepwise multiple regression analysis was utilized to determine the relationship between runoff and the effective soil properties. Finally, a linear model was developed using the effective soil properties and the most effective rainfall characteristics. To evaluate the capability of the model, comparisons between the measured and estimated runoff were done with R2 and the 1:1 line. Also, the model efficiency coefficient (E), which is commonly used as a measure of model performance in hydrology, was determined according to Nash and Sutcliffe (1970) as follows: n (Qpre,i − Qobs,i )2 E = 1 − i=1 (1) n (Qobs,i − Qobs )2 i=1
where n is the total number of observations; Qobs,i is the ith observed runoff depth; Qobs is the mean of observed (measured) runoff depth and Qpre,i is the ith predicted runoff depth. RESULTS AND DISCUSSION Rainfall characteristics Ninety seven natural rainfalls occurred in the study area during the 2-year study period. Table I shows the mean characteristics of the rainfall events from March 2005 to March 2007. Forty one rainstorms produced runoff and sediment in the runoff plots in the study area (Table I). The duration of the rainstorms was higher than 30 min and their intensity varied from 2.11 to 13.78 mm h−1 , with an average of 4.82 mm h−1 . The mean depths of the storms in the rain gauge stations located in grids 2, 10, 17, 27 and 30 were 7.22, 6.59, 6.98, 7.14 and 6.84 mm, respectively. As shown in Table II, there was no significant difference in the rainstorm depth among different rain gauge stations (F = 0.027, P = 0.994). Thus, the spatial distribution of the rainstorms was uniform and runoff variations in all the plots were directly associated with soil properties.
Soil properties As shown in Table IV, the soils of the study area were mainly clay loam having 36.7% sand, 31.6% silt and 32.0% clay, on average. They had a low organic matter content (1.1%) and were calcareous (limy) with 13% lime. Mean values of gravel in the soils were 10%. Soil aggregates were mainly granular with a mean di-
MODELING RUNOFF FROM AGRICULTURAL LANDS
599
TABLE I Characteristics of the runoff-producing rainstorms and mean runoff produced during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran Rainstorm event No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Rainstorm characteristics
Mean runoff
Duration
Depth
Intensity
Maximum 30-min intensity
h 1.1 1.4 3.4 1.0 1.3 1.1 6.9 0.7 1.5 0.7 0.7 1.1 1.2 0.9 1.6 2.1 1.3 0.5 0.8 1.4 0.6 0.6 4.0 0.8 1.7 3.2 1.6 1.5 1.2 1.8 2.4 1.4 1.4 1.9 1.8 0.5 3.5 0.5 0.5 2.1 1.8
mm 2.5 3.6 13.7 2.7 4.8 3.7 17.8 2.8 8.4 2.0 2.5 4.2 11.9 12.4 8.1 12.5 10.4 3.5 1.9 15.3 4.0 2.4 9.3 5.3 4.2 6.7 12.7 4.2 3.3 5.6 8.1 4.0 3.4 4.8 6.8 4.1 18.7 4.6 2.0 14.3 8.1
2.2 2.7 4.0 2.7 3.7 3.4 2.6 4.0 5.6 2.8 3.4 3.6 10.1 13.8 5.1 5.9 8.0 7.0 2.5 11.1 6.1 4.1 2.3 6.3 2.5 2.1 7.9 2.8 2.6 3.6 3.4 2.9 2.5 2.6 3.8 8.2 5.3 9.2 4.0 6.6 4.6
mm h−1 3.0 3.2 15.2 3.0 4.8 5.4 7.6 5.4 8.4 3.8 4.8 5.0 21.8 22.8 12.5 13.0 12.2 7.0 3.6 22.4 6.8 4.6 4.4 8.2 5.2 4.2 14.5 5.0 4.0 6.0 7.4 4.2 4.0 7.6 6.6 8.2 13.0 9.2 4.0 12.4 9.6
ameter of 5 mm. The water-aggregate stability of the soils was very low with the mean weight diameter (MWD) values ranging between 0.27 and 1.91 mm. The soil permeability (final infiltration rate) values varied between 1.4 and 5.8 cm h−1 , with an average of 3.5 cm h−1 . The initial infiltration rate was in the range between 60 and 81.3 cm h−1 . Statistical distributions of the soil properties were normal. Soil properties affecting runoff Runoff significantly correlated with coarse sand, silt, organic matter, lime, aggregate stability, initial
mm 0.05 0.11 0.80 0.10 0.23 0.18 0.62 0.31 0.41 0.08 0.23 0.30 0.98 1.22 0.70 0.56 0.72 0.36 0.06 0.91 0.58 0.22 0.60 0.13 0.08 0.09 0.68 0.17 0.15 0.37 0.50 0.35 0.19 0.27 0.39 0.04 0.75 0.20 0.04 0.82 0.58
infiltration rate and soil permeability (Table V). The soil permeability was the most important soil property controlling runoff in the study area (r = −0.78, P < 0.001). In fact, infiltration excess of rainfall was the predominant mechanism for runoff generation in the runoff plots (Buda et al., 2009). Findings of G´ omez et al. (2001) also showed that approximately 50% of variability of runoff in fallow plots can be explained by the final infiltration rate. Many have noted that mechanisms closely related to the runoff generation is the infiltration capacity of the soils. Indeed, infiltration capacity, which was determined based on the final infi-
600
A. R. VAEZI
TABLE II Analysis of variance on the depths of the runoff-producing rainstorms during the period from March 2005 to March 2007 at the rain gauge stations in the runoff plots installed in the dryfarming lands of Northwest Iran Location of the rain gauge station in the study areaa)
Mean rainstorm depth
Grid Grid Grid Grid Grid
mm 7.15 6.77 6.98 7.08 6.82
2 10 17 26 30
F value
P value
0.027
0.994
a) The
study area was divided into thirty six grids with a dimension of 5 km × 5 km.
ltration rate, was the most important factor controlling runoff in the soils (Morin and Cluff, 1980; Roth, 2004). The aggregate stability had better correlations with the initial infiltration rate (r = 0.61, P < 0.01) than with the soil permeability (r = 0.13). With an increase in the aggregate stability, runoff strongly decreased (r = −0.47, P < 0.01). Findings of Cerd` a (1996) and Barth`es and Roose (2002) showed that the aggregate stability is a relevant indicator of runoff
in semi-arid environments. In soils with high percentages of the water-stable aggregates, the soil surface can hardly be disrupted by raindrop impacts, resulting in reduced soil surface sealing. Therefore, rainwater infiltration into soil increased particularly at the initial stage of rainfall, which accords with the result of Lal and Shukla (2004), who noted that soils having a poor structure lead to surface sealing of pores and crusting and consequently lower infiltration and greater runoff. Augeard et al. (2007) also showed that soil surface sealing drastically reduces infiltration in bare soils exposed to rainfall and subsequently affects runoff generation and soil erosion. Runoff model The multiple regression analysis showed that mean runoff was significantly correlated with the final infiltration rate and aggregate stability (R2 = 0.74, P < 0.001) (Table VI). With an increase in these soil properties, the runoff generation considerably decreased. A linear regression relationship between runoff (mm) and the effective soil properties was extracted from the multiple regression analysis as follows: Runoff = 0.728 − 0.066Per − 0.088AS
(R2 = 0.74)(2)
Fig. 2 Relationships between runoff and rainfall characteristics, including rainfall depth (h), rainfall intensity (I), rainfall maximum 30-minute intensity (I30 ) and rainfall index obtained by multiplying the positive square root of rainfall depth (h0.5 ) by the logarithm of I30 (logI30 ), during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran.
MODELING RUNOFF FROM AGRICULTURAL LANDS
601
TABLE III Runoff generated for the rainstorms during the period from March 2005 to March 2007 in the runoff plots installed in the dry-farming lands of Northwest Iran Plot No.
Runoff Mean
Plot No. Standard deviation
Runoff Mean
year−1
1 2 3 4 5 6 7 8 9 10 11 12
9.38 9.50 9.14 8.81 11.38 4.20 6.93 5.43 10.47 10.84 11.92 5.99
Plot No. Standard deviation
Runoff Mean
Standard deviation
10.35 6.59 4.92 11.06 9.12 8.65 6.67 3.39 6.87 7.20 8.13 8.19
mm year−1 0.51 0.14 0.59 0.25 0.33 0.40 0.33 0.19 0.35 0.41 1.00 0.55
year−1
mm 0.37 0.27 0.49 0.45 0.26 0.44 0.49 0.57 0.36 0.65 0.16 0.24
13 14 15 16 17 18 19 20 21 22 23 24
9.94 9.48 10.02 4.61 6.00 9.51 9.96 7.34 6.57 6.56 8.74 7.32
TABLE IV Some physicochemical properties of soils in the runoff plots installed in the dry-farming lands of Northwest Iran Soil property
Mean
Standard deviation
Coarse sand (%) Very fine sand (%) Silt (%) Clay (%) Gravel (%) Organic matter (%) Lime (%) Mean weight diameter of the waterstable aggregates (mm) Initial infiltration rate (cm h−1 ) Final infiltration rate (cm h−1 )
18.9 17.8 31.5 31.8 9.9 1.1 12.7 1.13
5.3 3.2 7.1 5.7 2.4 0.2 5.2 0.44
81.33 3.5
14.37 1.2
where Per is the soil permeability (final infiltration rate, cm h−1 ) and AS is the aggregate stability (mean
mm 0.71 0.44 0.06 0.52 0.27 0.42 0.50 0.24 0.24 0.59 0.21 0.53
25 26 27 28 29 30 31 32 33 34 35 36
weight diameter of the water-stable aggregates, mm). The aggregate stability and final infiltration rate were strongly affected by some independent soil properties (mineral particles, organic matter and lime). Contrary to silt, coarse sand, organic matter and lime positively affected the soil permeability and consequently reduced runoff in the soils. The effect of coarse sand in reducing runoff agrees with Santos et al. (2003), who found that due to presence of macropores in the sandy soils, the rate of water entering the soil is higher and generation of runoff is lower as compared with the fine-textured soils. Although silt particles considerably decreased the soil permeability, they did not affect the aggregate stability. Despite the fact that clay particles can be absorbed on sand and silt fractions and enhance stability of aggregates (Lal and Shukla, 2004), their effects on runoff were not considerable due to the negligible role in the soil permeability. The effect of
TABLE V Correlation matrix between runoff and some physicochemical properties of the soils in the runoff plots installed in the dry-farming lands of Northwest Iran Variablea)
CS
VFS
Silt
Clay
Gravel
OM
Lime
AS
Per
IR
Runoff
CS VFS Silt Clay Gravel OM Lime AS Per IR Runoff
1 0.22 −0.74*** −0.18 0.03 0.27 −0.01 −0.17 0.76*** −0.06 −0.56**
1 −0.20 −0.50** −0.01 −0.31* −0.56** −0.67** −0.05 −0.34* 0.23
1 −0.40* 0.02 −0.23 0.17 −0.12 −0.55** −0.06 0.42*
1 −0.06 0.21 0.03 0.70*** −0.07 0.41* −0.09
1 0.16 −0.03 −0.09 0.09 −0.36* −0.09
1 0.05 0.29* 0.54** 0.44** −0.56**
1 0.48** 0.29* 0.32* −0.39*
1 0.13 0.61** −0.47**
1 0.19 −0.78***
1 −0.38*
1
*, **, *** Significant at P < 0.05, P < 0.01 and P < 0.001, respectively. a) CS = coarse sand; VFS = very fine sand; OM = organic matter; AS = aggregate stability (mean weight diameter of the water-stable aggregates, wet-sieving method); Per = soil permeability (final infiltration rate); IR = initial infiltration rate.
602
A. R. VAEZI
TABLE VI Multiple regression analysis of the relationship between mean runoff and soil properties in the runoff plots installed in the dry-farming lands of Northwest Iran (R2 = 0.74, P < 0.001) Model variablea)
Constant Per As a) AS
Unstandardized coefficients
Standardized coefficient
Model coefficient
Standard error
0.728 −0.066 −0.088
0.036 0.008 0.021
−0.728 −0.369
t level
P level
20.014 −8.084 −4.101
0.001 0.001 0.001
= aggregate stability (mean weight diameter of the water-stable aggregates); Per = soil permeability (final infiltration rate).
gravel on runoff also was not considerable, contrary to Mathys et al. (2005), who indicated that infiltration rate is increased by the gravel cover of soil surface in the Black Marls of the French Alps. Organic matter has been recognized as an important binding and bridging agent in enhancing the soil structural stability and infiltration capacity (Yu et al., 2006) and in consequence reducing runoff (Fern´ andez et al., 2006; Zhang et al., 2007). Lime was also shown to be an important factor controlling runoff in the soil. As shown by Vitro et al. (2011) in soils of semi-arid regions, aggregation is affected by carbonate contents. In fact, calcium linkage is a mechanism in the formation of water-stable aggregates and the increase in infiltration rate (Pepper and Morrissey, 1985). A linear empirical model was developed using the linear relationships between runoff (mm) and the effective soil properties, Eq. 2, and the most effective rainfall characteristics (rainfall index) obtained by multiplying the positive square root of rainfall depth (mm) (h0.5 ) by the logarithm of the maximum 30-minute intensity (mm h−1 ) (logI30 ) as follows: Runoff = (1.852 − 0.168Per − 0.224AS)× (0.198h0.5 logI30 − 0.056)
(R2 = 0.81) (3)
Fig. 3 Estimated v.s. observed runoff under the thirty-four rainstorm events during the period from March 2007 to March 2009 in the runoff plots installed in the dry-farming lands of Northwest Iran.
duration rainstorm events, which are known to have high rainfall depths and maximum 30-min intensity. The model efficiency coefficient was high (0.88), which showed that the model predictions could be acceptable in the study area. Use of the model could determine the susceptibility of different soils to runoff generation under different rainfalls in the study area and other regions of similar climate, topography and soil.
Model evaluation CONCLUSIONS To evaluate the model, field runoff measurements in the runoff plots in the study area continued for two years from March 2007 to March 2009. Thirtyfour rainstorm events lasting more than 30 min caused runoff for all the plots in the study area and runoff was also estimated by the model developed for each rainstorm event. As shown in Fig. 3, there was a relatively high correlation between the estimated and observed runoff (R2 = 0.91). In fact, 91% of the runoff variation in the soils under natural rainfalls could be explained by the model developed. The model slightly overestimated runoff, particularly for some large rainstorm events (Fig. 3). The negligible overestimation of runoff by the model was observed in some intensive short-
Only rainfalls with duration longer than 30 min could produce runoff in a two-year period in the runoff plots installed in the study area. Runoff generation by the rainstorms was largely associated with the rainfall index obtained by multiplying the positive square root of rainfall depth by the logarithm of the maximum 30-minute intensity. Water infiltration rate and aggregate stability were the most important properties of the soils which remarkably affected the runoff generation. A multiple linear regression model was developed between runoff and the rainfall index and the effective soil properties (i.e., water infiltration rate and aggregate stability), with R2 = 0.81. Model evaluation using
MODELING RUNOFF FROM AGRICULTURAL LANDS
thirty-four rainstorm events which generated runoff in the next two years showed high efficiency coefficient and R2 values of 0.88 and 0.91, respectively, which suggested that the model predictions were acceptable in the study area. Therefore, the model could be reliably used to predict runoff in the fallow sloped dry-farming lands in the study area and other similar semi-arid areas. REFERENCES Abdollahi, K. H., Sadeghi, S. H. R. and Farajzadeh-Asl, M. 2003. Runoff modeling based on geomorphological properties using GIS in Khanmirza Watershed, Iran (in Persian). M.S. Thesis, Tarbiat Modares University. Abrahams, A. D., Parsons, A. J. and Luk, S. H. 1988. Hydrologic and sediment responses to simulated rainfall on desert hillslopes in southern Arizona. Catena. 15: 103–117. Alizadeh, A. 2003. Applied Hydrology (in Persian). 3rd Ed. Asetan-e Ghods-e Razavi Publication, Mashhad. Al-Qurashi, A., McIntyre, N., Wheater, H. and Unkrich, C. 2008. Application of the Kineros2 rainfall-runoff model to an arid catchment in Oman. J. Hydrol. 355: 91–105. Angers, D. A. and Mehuys, G. R. 1993. Aggregate stability to water. In Cartner, M. R. (ed.) Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton. pp. 651–657. ASCE Task Committee on Application of the Artificial Neural Networks in Hydrology (ASCE). 2000. Artificial neural networks in hydrology. I: Preliminary concepts. J. Hydraul. Eng-ASCE. 5: 115–123. Augeard, B., Assouline, S., Fonty, A., Kao, C. and Vauclin, M. 2007. Estimating hydraulic properties of rainfall-induced soil surface seals from infiltration experiments and X-ray bulk density measurements. J. Hydrol. 341: 12–26. Augeard, B., Kao, C., Chaumont, C. and Vauclin, M. 2005. Mechanisms of surface runoff genesis on a subsurface drained soil affected by surface crusting: A field investigation. Phys. Chem. Earth. A/B/C. 30: 598–610. Avery, B. W. and Bascomb, C. L. 1982. Soil Survey Laboratory Methods. Soil Survey Technical Monograph No. 6. Rothamsted Experimental Station, Harpenden. Bahat, Y., Grodek, T., Lekach, J. and Morin, E. 2009. Rainfallrunoff modeling in a small hyper-arid catchment. J. Hydrol. 373: 204–217. Barth` es, B. and Roose, E. 2002. Aggregate stability as an indicator of soil susceptibility to runoff and erosion; validation at several levels. Catena. 47: 133–149. Botha, J. J., van Rensburg, L. D., Anderson, J. J., Hensley, M., Macheli, M. S., van Staden, P. P., Kundhlande, G., Groenewald, D. G. and Baiphethi, M. N. 2003. Water Conservation Techniques on Small Plots in Semi-Arid Areas to Enhance Rainfall Use Efficiency, Food Security, and Sustainable Crop Production. WRC Report No. 1176/1/03. Water Research Commission, South Africa. Bouwer, H. 1986. Intake rate: Cylinder infiltrometer. In Klute, A. (ed.) Methods of Soil Analysis. Part I. Physical and Mineralogical Methods. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 825–844. Buda, A. R., Kleinman, P. J. A., Srinivasan, M. S., Bryant, R. B. and Feyereisen, G. W. 2009. Factors influencing surface runoff generation from two agricultural hillslopes in central Pennsylvania. Hydrol. Process. 23: 1295–1312.
603
Cerd` a, A. 1996. Soil aggregate stability in three Mediterranean environments. Soil Technol. 9: 133–140. Cerd` a, A. 2001. Effects of rock fragment on soil infiltration, interrill runoff and erosion. Eur. J. Soil Sci. 52: 1–10. Dunj´ o, G., Pardini, G. and Gispert, M. 2004. The role of land use-land cover on runoff generation and sediment yield at a microplot scale, in a small Mediterranean catchment. J. Arid Environ. 57: 99–116. Faur` es, J. M., Goodrich, D. C., Woolhiser, D. A. and Sorooshian, S. 1995. Impact of small-scale spatial rainfall variability on runoff modeling. J. Hydrol. 173: 309–326. Fern´ andez, C., Vega, J. A., Fonturbel, T., P´erez-Gorostiaga, P., Jim´ enez, E. and P´ erez, J. R. 2006. Runoff and soil erosion after rainfall simulations in burned soils. Forest Ecol. Manag. 234: S191. Gee, G. W. and Bauder, J. W. 1986. Particle-size analysis. In Klute, A. (ed.) Methods of Soil Analysis. Part I. Physical and Mineralogical Methods. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 383–411. Goh, T. B., Arnaud, R. J. St. and Mermut, A. R. 1993. Carbonates. In Cartner, M. R. (ed.) Soil Sampling and Methods of Analysis. Lewis Publishers, Boca Raton. pp. 177–185. G´ omez, J. A., Nearing, M. A., Gir´ aldez, J. V. and Alberts, E. E. 2001. Analysis of sources of variability of runoff volume in a 40 plot experiment using a numerical model. J. Hydrol. 248: 183–197. G´ omez, J. A., Sobrinho, T. A., Gir´ aldez, J. V. and Fereres, E. 2009. Soil management effects on runoff, erosion and soil properties in an olive grove of Southern Spain. Soil Till. Res. 102: 5–13. Greenbaum, N., Ben-Zvi, A., Haviv, I. and Enzel, Y. 2006. The hydrology and paleohydrology of the Dead Sea tributaries. In Enzel, Y., Agnon, A. and Stein, M. (eds.) New Frontiers in Dead Sea Paleoenvironmental Research. Geological Society of America, Inc., Boulder. pp. 63–93. Hakimi, A. 1986. The Briefly Study of Soil Science in Hashtrood (in Persian). In Soil and Water Research Institute, Agriculture Ministry, Iran (ed.) Research Report No. 767. Soil and Water Research Institute, Agriculture Ministry, Tehran. pp. 2–15. Jacquin, A. P. and Shamseldin, A. Y. 2006. Development of rainfall-runoff models using Takagi-Sugeno fuzzy inference systems. J. Hydrol. 329: 154–173. Javadi, P., Rohipour, H. and Mahboubi, A. A. 2004. Role of gravel cover on runoff and erosion rate using rainfall simulator. Iran. J. Rangeland Desert Res. (in Persian). 3: 279– 287. Lal, R. and Shukla, M. K. 2004. Principle of Soil Physics. Marcel Dekker, Inc., New York. Le Bissonnais, Y., Daroussin, J., Jamagne, M., Lambert, J. J., Le Bas, C., King, D., Cerdan, O., L´eonard, J., Bresson, L. M. and Jones, R. 2005. Pan-European soil crusting and erodibility assessment from the European soil geographical database using pedotransfer rules. Adv. Environ. Monit. Model. 2: 1– 15. Lin, G. F. and Wang, C. M. 2007. A nonlinear rainfall-runoff model embedded with an automated calibration method— Part 1: The model. J. Hydrol. 341: 186–195. Ma, T., Zhou, C. H. and Cai, Q. G. 2009. Modeling of hillslope runoff and soil erosion at rainfall events using cellular automata approach. Pedosphere. 19: 711–718. Mart´ınez-Mena, M., Albaladejo, J. and Castillo, V. M. 1998. Factors influencing surface runoff generation in a Mediterranean semi-arid environment: Chicamo watershed, SE Spain. Hydrol. Process. 12: 741–754.
604
Mart´ınez-Murillo, J. F., Nadal-Romero, E., Reg¨ u´ es, D., Cerd` a, A. and Poesen, J. 2013. Soil erosion and hydrology of the western Mediterranean badlands throughout rainfall simulation experiments: A review. Catena. 106: 101–112. Mathys, N., Klotz, S., Esteves, M., Descroix, L. and Lapetite, J. M. 2005. Runoff and erosion in the Black Marls of the French Alps: Observations and measurements at the plot scale. Catena. 63: 261–281. Morin, J. and Cluff, C. B. 1980. Runoff calculation on semi-arid watersheds using a Rotadisk Rainulater. Water Resour. Res. 16: 1085–1093. Nash, J. E. and Sutcliffe, J. V. 1970. River flow forecasting through conceptual models part I—A discussion of principles. J. Hydrol. 10: 282–290. Nelson, D. W. and Sommers, L. E. 1982. Total carbon, organic carbon, and organic matter. In Page, A. L. (ed.) Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. 2nd Ed. American Society of Agronomy, Inc. and Soil Science Society of America, Inc., Madison. pp. 539–579. Pepper, R. G. and Morrissey, J. G. 1985. Soil properties affecting runoff. J. Hydrol. 79: 301–310. Rostamian, R., Jaleh, A., Afyuni, M., Mousavi, S. F., Heidarpour, M., Jalalian, A. and Abbaspour, K. C. 2008. Application of a SWAT model for estimating runoff and sediment in two mountainous basins in central Iran. Hydrolog. Sci. J. 53: 977–988. Roth, C. H. 2004. A framework relating soil surface condition to infiltration and sediment and nutrient mobilization in grazed rangelands of northeastern Queensland, Australia. Earth Surf. Proc. Land. 29: 1093–1104. Saadati, H., Gholami, S. H. A, Sharifi, F. A. S. and Ayoubzadeh, S. A. 2006. An investigation of the effect of land use change on simulating surface runoff using SWAT mathematical model (case study: Kasilian catchment area). Iran. J. Nat. Resour. (in Persian). 59: 301–313. Santos, F. L., Reis, J. L., Martins, O. C., Castanheira, N. L. and Serralheiro, R. P. 2003. Comparative assessment of infiltration, runoff and erosion of sprinkler irrigated soils. Biosyst.
A. R. VAEZI
Eng. 86: 355–364. Sokal, R. R. and Rohlf, F. J. 1981. Biometry: The Principals and Practice of Statistics in Biological Research. 4th Ed. Freeman W. H. & Co Ltd, New York. Sol´ e-Benet, A., Calvo, A., Cerd` a, A., L´ azaro, R., Pini, R. and Barbero, J. 1997. Influences of micro-relief patterns and plant cover on runoff related processes in badlands from Tabernas (SE Spain). Catena. 31: 23–38. Turner, D. P. and Sumner, M. E. 1978. The influence of initial soil moisture content on field measured infiltration rates. Water SA. 4: 18–24. Virto, I., Gartzia-Bengoetxea, N. and Fern´ andez-Ugalde, O. 2011. Role of organic matter and carbonates in soil aggregation estimated using Laser Diffractometry. Pedosphere. 21: 566–572. Wei, L. H., Zhang, B. and Wang, M. Z. 2007. Effects of antecedent soil moisture on runoff and soil erosion in alley cropping systems. Agr. Water Manage. 94: 54–62. Wischmeier, W. H. and Smith, D. D. 1978. Predicting Rainfall Erosion Losses—A Guide to Conservation Planning. Agriculture Handbook No. 537. U.S. Department of Agriculture, Washington D.C. Yair, A. and Lavee, H. 1985. Runoff generation in arid and semiarid zones. In Anderson, M. G. and Burt, T. P. (eds.) Hydrological Forecasting. John Wiley & Sons, Chichester. pp. 183–220. Yu, D. S., Shi, X. Z. and Weindorf, D. C. 2006. Relationships between permeability and erodibility of cultivated Acrisols and Cambisols in subtropical China. Pedosphere. 16: 304–311. Zeng, S. C., Su, Z. Y., Chen, B. G., Wu, Q. T. and Ouyang, Y. 2008. Nitrogen and phosphorus runoff losses from orchard soils in South China as affected by fertilization depths and rates. Pedosphere. 18: 45–53. Zhang, G. S., Chan, K. Y., Oates, A., Heenan, D. P. and Huang, G. B. 2007. Relationship between soil structure and runoff/soil loss after 24 years of conservation tillage. Soil Till. Res. 92: 122–128.