142 March, 2021 Int J Agric & Biol Eng Open Access at https://www.ijabe.org Vol. 14 No.

Field study and regression modeling on soil water distribution with

mulching and surface or subsurface drip irrigation systems

Mohamed A. Mattar1,3*, Ahmed A. Al-Othman1, Hosam O. Elansary2,

Ahmed M. Elfeky1, Akram K. Alshami1
(1. Agricultural Engineering Department, College of Food and Agriculture Sciences, King Saud University, Riyadh 11451, Saudi Arabia;
2. Plant Production Department, College of Food and Agriculture Sciences, King Saud University, Riyadh 11451, Saudi Arabia;
3. Agricultural Engineering Research Institute (AEnRI), Agricultural Research Centre, Giza, Egypt)

Abstract: The soil water status was investigated under soil surface mulching techniques and two drip line depths from the soil
surface (DL). These techniques were black plastic film (BPF), palm tree waste (PTW), and no mulching (NM) as the control
treatment. The DL were 15 cm and 25 cm, with surface drip irrigation used as the control. The results indicated that both the
BPF and PTW mulching enhanced the soil water retention capacity and there was about 6% water saving in subsurface drip
irrigation, compared with NM. Furthermore, the water savings at a DL of 25 cm were lower (15-20 mm) than those at a DL of
15 cm (19-24 mm), whereas surface drip irrigation consumed more water. The distribution of soil water content (θv) for BPF
and PTW were more useful than for NM. Hence, mulching the soil with PTW is recommended due to the lower costs and
using a DL of 15 cm. The θv values were derived using multiple linear regression (MLR) and multiple nonlinear regression
(MNLR) models. Multiple regression analysis revealed the superiority of the MLR over the MNLR model, which in the
training and testing processes had coefficients of correlation of 0.86 and 0.88, root mean square errors of 0.37 and 0.35, and
indices of agreement of 0.99 and 0.93, respectively, over the MNLR model. Moreover, DL and spacing from the drip line had
a significant effect on the estimation of θv.
Keywords: palm tree waste mulching, plastic film mulching, soil water distribution, regression models
DOI: 10.25165/j.ijabe.20211402.5200

Citation: Mattar M A, Al-Othman A A, Elansary H O, Elfeky A M, Alshami A K. Field study and regression modeling on
soil water distribution with mulching and surface or subsurface drip irrigation systems. Int J Agric & Biol Eng, 2021; 14(2):
142–150.

because it uses less water than DI because of reduced evaporation

1 Introduction from the soil surface[3,4]. SDI is often used to control the quantity
In arid regions, such as Saudi Arabia, the absence of an of water applied without having any significant environmental
effective water management system results in high water loss damage as a result of flow elimination and deep penetration[5].
during irrigation[1]. The climate is also hyper arid, with these Overall, this method reinforces the production of crops by
factors having a big effect on the limited water supplies and decreasing water waste[6-8]. SDI is more effective than DI because
agriculture[2]. Therefore, drip irrigation has been adopted, as it is it provides water directly to the root zone[9]. However, this
assumed to be the foremost efficient and valuable source for system’s performance can be affected depending on the space
stabilizing the utilization of water in comparison with other between the emitters and the lined depth of the drip lines[7].
systems. Surface drip irrigation (DI) often decreases water loss Some precautionary measures should be taken in the SDI
due to less water evaporation and deep percolation[1]. Despite system for crop production. According to Enciso et al.[7], there
these advantages, there are also several disadvantages of applying should be suitable spacing between emitters, taking into account
DI, including the possibility of damage, direct exposure of the drip the depth and type of drip lines during the design process to
lines to the sun, and the occurrence of salinity. Thus, subsurface enhance irrigation efficiency and increase productivity[10].
drip irrigation (SDI) has been proposed as a more useful method Magwenzi[11] stated that the depth of the drip lines can vary from
10 to 20 cm. Several studies have been conducted on SDI in
Received date: 2019-06-13 Accepted date: 2020-10-14 different crops. Najafi[12] studied the effect of drip line depths of
Biographies: Ahmed A. Al-Othman, Associate Professor, research interests: 15 and 30 cm on tomato crop yield. He found that the best yield
irrigation systems engineering and water management, Email: othmana@ was obtained at a drip line depth of 15 cm. In contrast, Zhuge et
ksu.edu.sa; Hosam O. Elansary, Associate Professor, research interests: al.[13] found that a drip line depth of 30 cm provided better root
irrigation water management, Email: [email protected]; Ahmed M. Elfeky,
distribution and nutrient absorption in tomato cultivation than at
Researcher Assistant, research interests: irrigation systems engineering and water
management, Email: [email protected]; Akram K. Alshami, depths of 20 and 40 cm. Patel and Rajput[14] found that a drip line
Researcher Assistant, research interests: irrigation systems engineering and water depth of 10 cm was best with 60.7 cm of irrigation water applied
management, Email: [email protected]. when planting onions in sandy, loamy soil. Al-Harbi et al.[15] also
*Corresponding author: Mohamed A. Mattar, Associate Professor, research found that using a drip line of 35 cm improved growth and reduced
interests: irrigation systems engineering and water management. Agricultural
evaporation from the soil surface. A study by Douh and
Engineering Department, College of Food and Agriculture Sciences, King Saud
University, Riyadh 11451, Saudi Arabia. Tel.: +966 11 4676024, Email: Boujelben[16] showed that the soil water content was recorded up to
[email protected]. a depth of 80 cm from the surface in drip lines of 15 cm used to
 March, 2021 Mattar M A, et al. Field study and regression modeling on soil water distribution with mulching and drip irrigation systems Vol. 14 No. 2 143

produce eggplant and the water saving was 23.2% higher than that water evaporation in an arid environment. Therefore, this study
of the DI system. Moreover, the soil profile of the water aimed to (1) explore the status of the volumetric soil water content
distribution was more constant in the SDI system than the DI (θv) under different drip line depths and soil surface mulching using
system. Çolak et al.[17] revealed that evaporation losses were black plastic film (BPF) and palm tree waste (PTW), (2) develop
lower in SDI receiving slightly less water than the DI in eggplant. empirical models using multiple regression analyses to estimate the
In addition, Al-Ghobari and Dewidar[18] reported that the soil water θv values for soil profiles under an SDI for different mulching to
contents in SDI were higher than those in DI during the tomato study important parameters affecting θv.
growth stages. Further, Pisciotta et al.[19] used drip lines at a depth
of 35 cm to prevent damage from tillage during the cultivation of
2 Materials and method
grapevines. 2.1 Field conditions and experimental design
In the field of agriculture, several techniques other than SDI, The experiment was conducted at the Agricultural Research
such as soil surface mulching, have been explored to improve water Farm in Dirab, Riyadh (24.4195°N, 46.65°E, and 552 m a.s.l.)
absorption[20]. These techniques are commonly used to avoid during the summer months of 2018. Climatic data were recorded
evaporation loss from the soil surface and enhance crop growth as monthly averages during the experimental period (June, July,
environments to increase crop yields[6,21-25]. One soil surface August, and September) as follows: 36.6°C, 37.4°C, 35.8°C, and
mulching technique that has been used in recent years is crop straw, 34.6°C, respectively, for air temperatures; 10.1%, 11.8%, 13.7%,
which minimizes soil water evaporation, ameliorates the soil’s and 13.5%, respectively, for relative humidity; 5.6 m/s, 6.1 m/s,
physical and chemical properties, and promotes biological 6.6 m/s, and 5.6 m/s, respectively, for maximum wind speed; and
activity[26-29]. In maize cultivation, Li et al.[23,24] found that straw 24.7 MJ/m2·d, 24.8 MJ/m2·d, 23.4 MJ/m2·d, and 23.2 MJ/m2·d,
mulching saved about 35% of water during the growth stages. respectively, for solar radiation. Moreover, there was no rainfall
Plastic film mulching is a well-evolved technique that is currently during these months.
being used for cultivation, especially in areas where irrigation water The field experiments were planned and implemented, as
is scarce[30]. It has been shown to enhance thermal conditions and shown in Figure 1. The DI system was designed using a water
increase water storage in the upper soil layers[31,32], improving crop tank, pump, pressure gauges, pressure regulators, flow meter, ball,
growth and water productivity[33,34]. In some cases, an SDI relief, solenoid valves, main lines, sub-mains, and drip lines. A
system mulched with plastic film is used to increase vegetable 5000 L water tank was used in the irrigation network. The drip
production[35], for example, Zotarelli et al.[36] and Baghani et al.[37]. lines (16 mm inner diameter and 1 mm thickness) were placed at
Wang et al.[38] and Ma et al.[39] reported that using plastic film the soil surface (i.e., DI) and two depths from the soil surface (DL)
mulching increased the yield of maize and potato in semiarid of 15 cm and 25 cm. The emitters had a discharge rate of 4 L/h at
regions by maintaining soil moisture in the upper layers. 150 kPa operating pressure with a spacing of 30 cm on the drip line.
A combination of SDI and mulching can possibly reduce soil Two soil surface mulches, BPF and PTW, were applied, as well as

Figure 1 Experimental layout

 144 March, 2021 Int J Agric & Biol Eng Open Access at https://www.ijabe.org Vol. 14 No. 2

no mulching (NM). A randomized complete block design was Three samples of irrigation water were also analyzed to determine
used with three replications (Figure 1). Before conducting the their chemical properties, which were an electrical conductivity
experiments, three soil samples were collected from different value of 4.5 dS/cm, pH 7.48, and total dissolved solids of
depths to determine physical and chemical properties (see Table 1). 2880 mg/kg.
Table 1 Physical and chemical properties of soil samples from the experimental site
Particle size distribution/% Physical properties Chemical properties
Soil depth
Soil texture
/cm -1
Clay Silt Sand θi/% FC/% WP/% ρb/g·cm -3
EC/dS·m pH CaCO3 Na K P
Plot 1
0-25 3.2 22.5 74.3 Loamy sand 1.22 14.58 3.04 1.51 1.47 7.85 25.54 61 116 31.1
25-50 3.2 22.5 74.3 Loamy sand 1.36 15.99 3.39 1.41 2.4 7.73 27.04 181 115 21.8
Plot 2
0-25 3.2 15 81.8 Loamy sand 1.22 14.86 3.05 1.52 3.4 7.8 25.37 237 109 12.5
25-50 1.95 16.25 81.8 Loamy sand 1.15 15.15 2.45 1.40 2.37 7.87 24.75 139 110 9.3
Plot 3
0-25 4.45 16.25 79.3 Loamy sand 1.37 17.57 3.05 1.50 3.09 7.81 24.75 330 81 24.9
25-50 0.7 12.5 86.8 Sand 0.93 14.81 2.06 1.40 2.05 7.91 23.34 218 70 34.2
Note: Plot 1: drip line depth = 0 cm; Plot 2: drip line depth = 15 cm; Plot 3: drip line depth = 25 cm; θi: Initial water content; FC: field capacity; WP: Wilting point; ρb:
Soil bulk density; EC: Electrical conductivity.

2.2 Irrigation water 2.4 Multiple regression analysis

An irrigation schedule was applied using soil sensors to The multiple linear regression (MLR) and multiple nonlinear
continuously monitor the θv and supply a suitable amount of regression (MNLR) models were derived using the data acquired
irrigation water to the planned irrigation treatment. Its purpose was from the field experiment. The dependent variable is estimated
to maximize the irrigation efficiency by applying the appropriate when the values of the independent variables are obtained from the
amount of water needed to replenish the soil water to the desired linear and nonlinear combinations in a multiple regression model.
level at field capacity (FC). However, no crops were planted in this Based on the results of the MLR and MNLR, the effects of the
study. Therefore, the depth of water applied to reach the FC (Dw) influencing factors on θv can be better explained. The
was calculated for a soil depth (D) of 50 cm at increments of 10 cm corresponding model expressions can be described as follows:
using Equation (1): MLR model:
Dw = ∑ i =1 Di ( FCi − θ vi ) Yˆ = a0 + a1 X 1 + a2 X 2 + a3 X 3 + ... + am X m
n
(1) (4)
where, Dw is the depth of water applied to reach the field capacity, where, Yˆ is the predicted value of the dependent variable, ai (i =
mm; n is the number of sensors; Di is the soil depth at the ith sensor, 0, …, m) are the estimated regression coefficients for the linear
mm; FCi is the field capacity of the soil at the ith sensor; and θvi is the relation, and Xi (i = 1, …, m) are the independent variables.
soil water content at the ith sensor. MNLR model:
2.3 Monitoring soil water content Yˆ = b0 × X 1b1 × X 2 b2 × X 3b3 × ... × X m bm (5)
Three EasyAG probes (Sentek Sensor Technologies, Stepney, where, bi (i = 0, …, m) are the estimated regression coefficients for
Australia), including several sensors (electrical capacitance) were the nonlinear relation.
used and installed in each plot to monitor the values of θv at a soil The regression models were developed, and the mulching types
depth from 10 to 50 cm. These probes were placed as follows: the (MT) that were categorized as 0 for NM, 2 for BPF, and 3 for PTW,
first was placed directly at the emitter, the second was placed at a DL, S, and D were considered as independent variables, whereas θv
15 cm spacing from the drip line (S), and the third was at S of 30 cm, was the dependent variable. In the present study, 135 data points
as shown in Figure 1. were used to derive all regression models. The models were
The electrical capacitance from the probe recorded a scaled developed using 60% of the data for the training process, and 20% of
frequency reading (SF) in the field to provide a θv using a calibration the data were used to test the models. The remaining data were
equation, as follow: used for model validation. During the validation process, the
( F − FS ) performance of the developed regression models were checked
SF = A (2)
( FA − FW ) under NM, BPF, and PTW. The statistical characteristics of θv data
1 for each process are provided in Table 3. The SPSS Statistics
⎛ SF − C ⎞ B software package (version 16.0; SPSS; Chicago, IL, USA) was used
θv = ⎜ ⎟ (3)
⎝ A ⎠ to perform the regression analysis.
where, FA, FS, and FW are frequency readings in the air, soil, and Statistical performance evaluation criteria were calculated to
water, respectively, and A, B, and C are constants obtained from test the goodness of fit of the regression models during the training,
calibration procedure of the probes (Table 2). testing, and validation processes. The criteria utilized were the
Table 2 Constants of Equation (3) for three sensors after coefficients of correlation (r), root mean square errors (RMSE), and
calibration index of agreement (IA). r measured the degree of correlation
between the measured and calculated θv values with values close to
Sensor location A B C R2
1.0 indicating a perfect agreement, and it was calculated as follows:
Directly on drip line 60.619 0.109 −71.356 0.942
∑
N
( M i − M )(Ci − C )
At spacing of 15 cm from drip line 476.132 0.014 −485.695 0.844 r= i =1
(6)
∑ i =1 (M i − M )2 ∑ i =1 (Ci − C )2
N N
At spacing of 30 cm from drip line 507.365 0.011 −513.789 0.751
 March, 2021 Mattar M A, et al. Field study and regression modeling on soil water distribution with mulching and drip irrigation systems Vol. 14 No. 2 145

where, Mi is the measured value; Ci is the calculated value; N is the DL of 15 cm. Therefore, the use of SDI with BPF or PTW
number of data; M is the average measured value; and C is the mulching retained soil moisture and decreased the amount of water
average calculated value; RMSE expresses the error in the same units required[44]. Thus, it is preferable to use PTW mulching at a DL of
as the variable[40] and measures the difference between the 15 cm, as there are fewer additional costs.
calculated and measured values[41] by:

∑
N
( M i − Ci ) 2
RMSE = i =1
(7)
N
IA represents the ratio between the mean square error and the
potential error, as defined by Willmot[42]:
∑
N
( M i − Ci ) 2
IA = 1 − i =1
(8)
∑
N
i =1
(| Ci − M | + | M i − M |) 2
IA has a range from 0 to 1, with a perfect fit at 1[40].
Table 3 Statistical parameters of θv data used for the training,
testing, and validation processes
Xm Xa Xn Sx Kx Csx

Training process 15.60 14.17 12.28 0.723 −0.291 −0.010 Note: NM: no mulching; BPF: black plastic film; PTW: palm tree waste.
Testing process 15.36 13.98 12.45 0.708 −0.123 −0.030 Figure 2 Water applied under different mulching types at different
Validation process drip line depths from the soil surface (DL)
NM 14.81 13.75 12.58 0.784 −0.799 −0.410 3.2 Effect of mulching type on soil water content
BPF 15.32 14.05 13.10 0.820 −1.134 0.585 Regardless of the DL treatments, Tables 5 shows that there
PTW 15.29 14.13 12.97 0.771 −1.132 −0.028 were significant effects of MT on θv values at different S. The
Note: Xm: maximum value; Xa: mean value; Xn: minimum value; Sx: standard average θv values for the BPF treatment were the highest, followed
deviation; Kx: kurtosis coefficient; Csx; skewness coefficient; NM: no mulching; by the PTW treatment in both drip systems (Figure 3). For the DL
BPF: black plastic film; PTW: palm tree waste.
of 0 cm with BPF mulching at S of 0, 15, and 30 cm, the θv values
2.5 Statistical analysis increased by 4.08%, 1.33%, and 1.76%, respectively, compared
An analysis of variance was performed using CoStat version with the NM treatment, while there were increases of 2.39%,
6.303 (CoStat Version 6.303 Copyright 1998-2004) to determine the 0.89%, and 0.84%, respectively, with PTW mulching (Figure 3).
effects of DL and MT on the measured θv. The treatment means For a DL of 15 cm, the BPF and PTW treatments at an S of 0 cm
were separated using a least significant difference test at p = 0.05. had θv values that were 3.65% and 1.38% higher, respectively, than
3 Results and Discussion those of the NM treatment, while there were increases of 1.02%
and 0.58%, respectively, at an S of 15 cm and increases of 1.58%
3.1 Applied water and 0.75%, respectively, at an S of 30 cm (Figure 3). For a DL of
The BPF treatment achieved water savings of 10.55%, 6.79%, 25 cm, there were increases of 2.74% and 1.13%, respectively, in
and 7.02% at DL of 0 cm, 15 cm, and 25 cm, respectively, the θv values at an S of 0 cm for the BPF and PTW treatments
compared with the NM treatment, which had 93.5 mm, 69 mm, and compared with the NM treatment. Additionally, the
73 mm of applied water, respectively, while the PTW treatment corresponding values of the BPF and PTW treatment increased by
achieved water savings of 5.96%, 4.94%, and 5.26%, respectively 3.97% and 2.91%, respectively, at an S of 15 cm and by 3.15% and
(Figure 2). The water applied under DI, at a DL of 0 cm, with any 2.20%, respectively, at an S of 30 cm (Figure 3). The use of BPF
type of mulching was conserved less than that applied under SDI, or PTW mulching substantially limits the flux of water vapor from
at a DL of 15 or 25 cm, where soil surface evaporation rates were the soil surface into the air[23,24,45,46]. Although the purchase cost
high[4,43]. Additionally, as shown in Figure 2, using BPF or PTW of BPF is high, it can be replaced by PTW, which is available at no
above a DL of 25 cm required the application of more water than a additional cost.

a. DL=0 cm b. DL=15 cm c. DL=25 cm

Note: NM: no mulching; BPF: black plastic film; PTW: palm tree waste.
Figure 3 Average soil water content (θv) under different spacing from drip line (S) and mulching types at different drip line depths from the
soil surface (DL)
 146 March, 2021 Int J Agric & Biol Eng Open Access at https://www.ijabe.org Vol. 14 No. 2

3.3 Effect of the depth of the drip line on soil water content distribution throughout the soil profile. DL of 15 and 25 cm had the
Regardless of the mulching treatments, the DL significantly best uniformity of θv distribution contour lines at different S. When
affected (p<0.01) the average θv values at different S (Table 4), with S increased, the θv bulb’s spread decreased, in agreement with
a DL of 25 cm having the highest θv values, in contrast to a DL of Assouline[49], Grabow et al.[50], Badr[51], Shirahatti et al.[52], and
0 cm. The average θv values for a DL of 25 cm were highest at S of Nasrabad et al.[53]. Since the horizontal movement of water is
0, 15, and 30 cm under the NM treatment, which increased by 3.17%, limited in sandy soils, there should be close spacing between the
2.18%, and 2.63% at a DL of 15 cm, respectively, and increased by emitters[54]. In addition, the vertical movement was higher than the
5.42%, 4.29%, and 4.75%, respectively, at a DL of 0 cm (Figure 3). horizontal movement under the SDI system[43,55,56].
These findings are consistent with findings by Mokh et al.[47], who There were significant effects of interactions between the MT
observed that θv values increased with increasing DL. The θv values and DL on the θv values at an S of 15 cm (Table 4). For DL of 0, 15,
in the BPF and PTW treatments under different DL showed a similar and 25 cm at a D of 0-20 cm, the θv values of BPF were increased by
trend (Figure 3). For the BPF treatment, a DL of 25 cm produced 0.96%, 3.05% and 5.56%, respectively, compared with NM, while
the highest θv values, which were 2.26% (at an S of 0 cm), 5.17% (at the θv values of PTW were 1.25%, 2.91%, and 5.41%, higher
an S of 15 cm), and 4.22% (at an S of 30 cm) higher than the θv respectively, in agreement with Wang et al.[57] and Liu et al.[58].
values at a DL of 15 cm and 4.06%, 7.01%, and 6.18% higher, Ma et al.[39] found that using plastic films significantly increased
respectively, than the θv values at a DL of 0 cm. For the PTW the θv by 12.9% at a D of 0-20 cm, compared with the traditional
treatment, the θv values at a DL of 25 cm were 2.92%, 4.54%, and method. The use of either BPF or PTW mulching maintains water
4.10% higher than those at a DL of 15 cm at S of 0, 15, and 30 cm, evaporation and encourages the movement of water into the topsoil
respectively, while there were increases of 4.12%, 6.38%, and layers, which enhances θv during the initial growth stage of crops[44].
6.16%, respectively, comparison with a DL of 0 cm. A small The surface soil layer has a lower θv (i.e., not completely wet) for
difference in average θv values was observed between the DL of SDI than DI, thereby reducing the soil evaporation in SDI[48]. At
15 cm and DL of 25 cm. Therefore, a DL of 15 cm should be the active root depth (0-40 cm) for most crops, the NM at an S of
selected because of the lower drilling costs. Installing a drip line 15 cm had average θv values of 13.87% and 14.13% for DL of 15
below the soil surface and not exposing it to the sun increases the θv and 25 cm, respectively (Figure 4). In sandy soils, SDI enhances
value, as moisture is not lost[48]. the water recovery capacity[59], while in loamy soils, a DL > 10 cm is
Table 4 Results of variance analysis of θv values under recommended to prevent wetting of the soil surface during
mulching type (MT) and drip line depth from the soil surface irrigation[60]. The average θv corresponding values for BPF and
(DL) at different spacing from the drip line (S) PTW were 14.02% and 13.97%, respectively, at a DL of 15 cm, and
14.63% and 14.54%, respectively, at a DL of 25 cm (Figures 5 and
Treatments S = 0 cm S = 15 cm S = 30 cm
6). This increased θv of BPF and PTW treatments could be
MT ** ** *
attributed to minimal soil water loss, which has an important role in
No mulching 14.57 c 13.79 b 13.33 b
crop management and growth[61,62]. Ashrafuzzaman et al.[63]
Black plastic film 15.08 a 14.09 a 13.63 a
b a
reported that water in the form of fog fell into the topsoil layer due
Palm tree waste 14.81 13.99 13.51 ab
to the vapors formed by water trapped inside the mulch. Figures
LSD0.05 0.17 0.11 0.19
4-6 show that when the DL increased, a saturation bulb of θv
DL ** ** **
Surface drip 14.52 c 13.61 c 13.15 c
distribution below emitters became more controllable moving
Subsurface drip at 15 cm depth 14.76 b 13.86 b 13.41 b
downward, especially with BPF and PTW, which also changed the
Subsurface drip at 25 cm depth 15.14 a
14.42 a
13.91 a shape of the bulb from a circular shape to an ellipse. This is in
LSD0.05 0.17 0.11 0.20 agreement with Thorbum et al.[64], who showed that SDI controls
MT × DL ns * ns the amount of water and the spread of θv toward the soil surface.
Note: ns: Non-significant; *: Significant at p<0.05; **: Significant at p<0.01. BPF at a DL of 25 cm allowed slightly more θv downward
Mean values in columns followed with different letters are significantly different movement than PTW (Figures 5 and 6). However, it is better to
based on the least significant difference test at p<0.05 (LSD0.05). use PTW mulching at a DL of 15 cm because of its lower installation
3.4 Soil water distribution cost, despite the slightly lower θv values than at a DL of 25 cm for
Figures 4-6 illustrate the effects of MT and DL on the θv PTW mulching.

13.9

Figure 4 Soil water distribution through the emitter at different drip line depths under no mulching after irrigation for 24 h
 March, 2021 Mattar M A, et al. Field study and regression modeling on soil water distribution with mulching and drip irrigation systems Vol. 14 No. 2 147

Figure 5 Soil water distribution through the emitter at different drip line depths under black plastic film mulching after irrigation for 24 h

Figure 6 Soil water distribution through the emitter at different drip line depths under palm tree waste mulching after irrigation for 24 h
3.5 Performance of regression models the training and testing processes. From the scatter plots, the MLR
Important factors affecting θv values were identified using model provided the least scattered θv estimates whereas the MNLR
regression analyses that developed the mathematical models for θv model provided the worst results for the training and testing
as given below: processes. Table 6 demonstrates the goodness of fit of each
For MLR regression model during the training and testing processes. The r
θv = 14.58+0.071MT+0.003DL – 0.0089D – 0.047S (9) and IA values of the MLR model were higher and the RMSE value
For MNLR was lower than that of the MNLR model during the training process.
θv = 15.24×MT0.004×DL0.005×D-0.007×S-0.026 (10) The r and IA values for the MLR model were 2.87% and 9.06%
Table 6 shows the standard error (SE), t-statistic (t-stat), higher, respectively, and the RMSE value was 6.87% lower than the
probability (p-value), and variance inflation factor (VIF) for the values in the MNLR model. Based on the testing dataset, the trends
independent variables (MT, DL, S, and D) in the MLR model at a of the MLR and MNLR model results were the same for the training
95% confidence interval (CI). The t-stat values of the variables dataset. The MLR model performed better than the MNLR model,
should be greater than 1.99 or less than −1.99 to confirm the with the r and IA values increasing by 3.53% and 2.64%,
goodness of fit of the regression coefficients. The t-stats of DL and respectively, and the RMSE value decreasing by 8.97%.
S were greater than 1.99 and less than −1.99, respectively. The Table 5 Standard error of regression coefficients, t-statistic,
p-values for the DL and S were small (p < 0.05). The SE of the probability, and variance inflation factor of independent
coefficients of DL and S were 0.004. Therefore, DL and S were the variables for the multiple linear regression model
most significant variables in the calculation of θv. VIF values of all
Independent variables
variables provided in Table 5 were equal to 1. There were no Intercept
correlations between variables, which are expressive inputs in the MT DL S D
MLR model. For the MNLR analysis at 95% CI, M, DL, and S were SE 0.135 0.034 0.004 0.004 0.003
more significant for θv, where CI did not include zero. The D was t-stat 108.165 2.098 7.411 −13.110 −2.949
not significantly affected because the lower bound and upper bound p-value 5.43×10-85 0.039 1.46×10-10 3.37×10-21 0.004
of CI for D were −0.018 and 0.004, respectively. VIF 1.00 1.03 1.02 1.00
Figure 7 shows the comparisons between the measured θv Note: SE: standard error; t-stat: t-statistic; p-value: probability; VIF: variance
values and θv values calculated using the MLR and MNLR models inflation factor; MT: mulching types; DL: drip line depth from the soil surface;
when plotting the regression results of Equations (9) and (10) during S: spacing from drip line; D: soil depth.
 148 March, 2021 Int J Agric & Biol Eng Open Access at https://www.ijabe.org Vol. 14 No. 2

Table 6 Statistical performance of the regression models of the NM and PTW treatments, respectively, and 8.96% and 1.29%
during the training and testing processes more accurate, respectively, for the IA values. The MLR model’s
BPF had RMSE values that were 44.34% and 19.21% more accurate
Statistical Training process Testing process
for θv compared with NM and PTW, respectively. For the PTW
parameters MLR MNLR MLR MNLR treatment, the MLR model produced r and IA values that were
r 0.861 0.837 0.879 0.849 6.84% and 7.57% more accurate, respectively, and 31.11% more
RMSE 0.366 0.393 0.345 0.379 accurate for the RMSE values than those obtained for the NM
IA 0.987 0.905 0.934 0.910 treatment. Hence, the r, IA, and RMSE values for the two models
Note: r: coefficients of correlation; RMSE: root mean square errors; IA: index of confirmed that the MNLR model performed poorly, whereas the
agreement. MLR model successfully calculated the θv values.

a. Training process

b. Testing process
Figure 7 Scatterplots comparing measured θv values and θv values
calculated by the multiple linear regression (MLR) and multiple
nonlinear regression (MNLR) models
Figure 8 shows the statistical performance values of the MLR
and MNLR models during the validation process under different MT.
The MLR model provided better agreement between the measured
and calculated θv values than the MNLR model. This finding is
reflected in the high r (0.87-0.95) and IA (0.89-0.97) values, and the
low RMSE (0.24-0.44) values for the MLR model. The MLR
model for the NM treatment had 9.32% and 6.84% higher r and IA Note: NM: no mulching; BPF: black plastic film; PTW: palm tree waste.

values, respectively, than the MNLR model, and it had a 6.04% Figure 8 Statistical criteria of coefficients of correlation (r), root
lower RMSE value. The r and IA values for the MLR model in the mean square error (RMSE), and index of agreement (IA) for the
BPF treatment increased by 2.31% and 2.51%, respectively, whereas regression models during the validation process
the RMSE value decreased by 19.94%. Moreover, r and IA values
4 Conclusions
in the PTW treatment during the validation process increased to
15.08% and 7.76%, respectively, and the RMSE value decreased to The effects of different mulching types (MT) in an subsurface
34.21% in the MLR model. drip irrigation (SDI) system on soil water content (θv) distribution
There were considerable differences between the calculated θv were investigated. The use of black plastic film (BPF) or palm
values for the different MT in the MLR and MNLR models, as tree waste (PTW) enhanced the soil's ability to hold water. In the
shown in Figure 8. The MLR model for the prediction of θv with SDI system, BPF mulching required the application of less water
the BPF mulching was the most accurate. The MLR model for the than PTW mulching. Therefore, it is recommended that SDI is
BPF treatment yielded the highest r and IA values and the lowest beneficial to the soil, and when a drip line depth from the soil
RMSE value compared with the corresponding values for the NM surface (DL) of 15 cm is installed and mulching is conducted with
and PTW treatments. For the BPF treatment, modeling with MLR PTW, there are no additional costs. This approach expands the
gave r values that were 8.91% and 1.94% more accurate than those active soil zone to the roots of crops. Therefore, the strategy
 March, 2021 Mattar M A, et al. Field study and regression modeling on soil water distribution with mulching and drip irrigation systems Vol. 14 No. 2 149

presented in this study can save water resources in Saudi Arabia’s surface and subsurface drip systems. Scientia Horticulturae, 2017; 219:
crop fields at a remarkably low cost. 10–21.
[18] Al-Ghobari H M, Dewida A Z. Integrating deficit irrigation into surface
θv values were derived as a function of MT, DL, spacing from
and subsurface drip irrigation as a strategy to save water in arid regions.
drip line, and soil depth, using multiple linear regression (MLR) Agric. Water Manag., 2018; 209: 55–61.
and multiple nonlinear regression (MNLR) analyses. The [19] Pisciotta A, Di Lorenzo R, Santalucia G, Barbagallo M G. Response of
accuracy of the MLR and MNLR models were studied using the grapevine (Cabernet Sauvignon cv) to above ground and subsurface drip
coefficients of correlation (r), root mean square errors (RMSE) and irrigation under arid conditions. Agric. Water Manage., 2018; 197:
122–131.
index of agreement (IA). Evaluation of the models found that the
[20] Hapeman C, Durham S. Plastic mulch: harmful or helpful? Agric. Res.
MLR model fitted well based on the high r and IA values and low Mag., 2003; 51: 14–16.
RMSE values obtained. The MLR model showed better [21] Zhang S, Lövdahl L, Grip H, Tong Y, Yang X, Wang Q. Effects of
performance in calculating θv values than the MNLR model. The mulching and catch cropping on soil temperature: soil moisture and wheat
MLR model for the BPF treatment was more accurate than no yield on the Loess Plateau of China. Soil Tillage Res, 2009; 102: 78–86.
[22] Bu L, Liu J, Zhu L, Luo S, Chen X, Li S, et al. The effects of mulching
mulching and PTW treatments. Overall, it is concluded that the
on maize growth, yield and water use in a semi-arid region. Agric. Water
MLR model can be used to produce accurate outcomes to predict θv Manag., 2013; 123: 71–78.
values. [23] Li R, Hou X, Jia Z, Han Q, Ren X, Yang B. Effects on soil temperature,
moisture, and maize yield of cultivation with ridge and furrow mulching in
the rainfed area of the Loess Plateau, China. Agric. Water Manag., 2013;
Acknowledgements
116: 101–109.
The authors extend their appreciation to the Deanship of [24] Li S, Wang Z, Li S, Gao Y, Tian X. Effect of plastic sheet mulch, wheat
Scientific Research at King Saud University for funding this work straw mulch, and maize growth on water loss by evaporation in dryland
through research group No (RG-1440-022). areas of China. Agric. Water Manag., 2013; 116: 39–49.
[25] Haque M A, Jahiruddin M, Clarke D. Effect of plastic mulch on crop
yield and land degradation in south coastal saline soils of Bangladesh.
[References] International Soil and Water Conservation Research, 2018; 6(4): 317–324.
[1] Al-Amoud A I. Subsurface drip irrigation for date palm trees to conserve [26] Blanco-Canqui H, Lal R. Soil and crop response to harvesting corn
water. ISHS Acta Horticulturae, 2010; 88: 103–114. residues for biofuel production. Geoderma, 2007; 141: 355–362.
[2] Al-Shayaa S, Baig M B, Straquadine G S. Agricultural extension in the [27] Jordán A, Zavala L M, Gil J. Effects of mulching on soil physical
Kingdom of Saudi Arabia: difficult present and demanding future. J. properties and runoff under semi-arid conditions in southern Spain.
Anim. Plant Sci., 2012; 22: 239–246. Catena, 2010; 81: 77–85.
[3] Ayars J E, Phene C J, Schoneman R A, Meso B, Dale F, Penland J. [28] Sharma P, Abrol V, Sharma R. Impact of tillage and mulch management
Impact of bed location on the operation of subsurface drip irrigation on economics, energy requirement and crop performance in maize–wheat
systems. Proc. 5th International Microirrigation Congress, ASABE, 1995; rotation in rainfed subhumid inceptisols, India. Eur. J. Agron., 2011; 34:
pp.68–174. 46–51.
[4] Çolak Y B, Yazar A, Gönen E, Eroğlu E Ç. Yield and quality response of [29] Jiménez M N, Pinto J R, Ripoll M A, Sánchez-miranda A, Navarro F B.
surface and subsurface drip-irrigated eggplant and comparison of net Impact of straw and rock-fragment mulches on soil moisture and early
returns. Agric. Water Manage., 2018; 206: 165–175. growth of holm oaks in a semiarid area. Catena, 2017; 152: 198–206.
[5] Zin El-Abedin T K, Mattar M A, Alazba A A. Soil wetting pattern from [30] Dong H, Li W, Tang W, Zhang D. Early plastic mulching increases stand
subsurface drip irrigation as affected by application of a polyacrylamide establishment and lint yield of cotton in saline fields. Field Crop Res.,
layer. Irrig. Drain., 2015; 64: 609–618. 2009; 111: 269–275.
[6] Dukes M D, Scholberg J M. Soil moisture controlled subsurface drip [31] Wang X, Li Z, Xing Y, Clothier B E, Dierickx W, Oster J, et al. Effects
irrigation on sandy soils. Appl. Eng. Agric., 2005; 21: 89–101. of mulching and nitrogen on soil temperature, water content, nitrate-N
[7] Enciso J, Jifon J, Wiedenfeld B. Subsurface drip irrigation of onions: content and maize yield in the Loess Plateau of China. Agric. Water
effect of emitter spacing and drip depth on yield. American Society of Manage., 2015; 161: 53–64.
Agricultural and Biological Engineer, St Joseph, MI, USA, Paper number [32] Liang S M, Cai R, Wang P J, Wang X T, Li Y S, Xu F H, et al.
052242, 2005. Improvements of emergence and tuber yield of potato in a seasonal spring
[8] Soussa H K. Effects of drip irrigation water amount on crop yield, arid region using plastic film mulching only on the ridge. Field Crops
productivity and efficiency of water use in desert regions in Egypt. Nile Res., 2018; 223: 57–65.
Basin Water Sci. Eng. J., 2010; 3: 96–109. [33] Fan Y, Ding R, Kang S, Hao X, Du T, Tong L, et al. Plastic mulch
[9] Irmak S, Djaman K, Rudnick D R. Effect of full and limited irrigation decreases available energy and evapotranspiration and improves yield and
amount and frequency on subsurface drip ‑ irrigated maize water use efficiency in an irrigated maize cropland. Agric. Water
evapotranspiration, yield, water use efficiency and yield response factors. Manage., 2016; 179: 122–131.
Irrig. Sci., 2016; 34: 271–286. [34] Wu Y, Huang F, Jia Z, Ren X, Cai T. Response of soil water, temperature,
[10] Bryla D R, Banuelos G S, Mitchell J P. Water requirements of subsurface and maize (Zea may L.) production to different plastic film mulching
drip-irrigated fava bean in California. Irrig. Sci., 2003; 22: 31–37. patterns in semi-arid areas of northwest China. Soil Tillage Res., 2017;
[11] Magwenzi O. Evaluation of irrigation efficiency in the Swaziland sugar 166: 113–121.
industry. Proc. S. Afr. Sug. Technol. Ass., 2000; 74: 151–154. [35] Coelho R D, Monteiro R O C, Chaves, S W P, Shirahige F H. Effects of
[12] Najafi P. Effects of using Subsurface drip irrigation and treated municipal subsurface drip irrigation (SDI) and plastic mulching on melon crop under
waste water in irrigation of tomato. Pakistan J. Biol. Sci., 2006; 9: protected environment. ASABE, 095565.
2672–2676. https://doi:10.13031/2013.26915, 2009.
[13] Zhuge Y P, Zhang X D, Zhang Y L, Liu M D. Tomato root response to [36] Zotarelli L, Dukes M D, Scholbert J M, Hanselman T, LeFemminella K,
subsurface drip irrigation. Pedosphere, 2004; 14: 205–212. Munoz-Carpena R. Nitrogen and water use efficiency of zucchini squash
[14] Patel N, Rajput T B S. Effect of subsurface drip irrigation on onion yield. for a plastic mulch system on a sandy soil. Scientia Hort., 2008; 116: 8–16.
Irrig. Sci. 2009; 2: 97–108. [37] Baghani J, Dehghani S H, Sadrghaiini S H. Study the effects of plastic
[15] Al-Harbi A R, Al-Omran A M, El-Adgham F I. Effect of drip irrigation mulches and different irrigation water level on qualitative and quantitative
levels and emitters depth on okra (Abelmoschus esculentus) growth. J. yield of melon in surface and subsurface drip irrigation systems. Iran J.
Appl. Sci., 2008; 8: 2764–2769. Irrig. Drain., 2010; 4: 175–181.
[16] Douh B, Boujelben A. Water saving and eggplant response to subsurface [38] Wang Y P, Li X G, Zhu J, Fan C Y, Kong X J, Turner N C, et al.
drip irrigation. Electron. J. Agric. Segm., 2010; 1: 1–12. Multi-site assessment of the effects of plastic-film mulch on dryland maize
[17] Çolak Y B, Yazar A, Sesveren S, Çolak İ. Evaluation of yield and leaf productivity in semiarid areas in China. Agric. For. Meteorol., 2016; 220:
water potential (LWP) for eggplant under varying irrigation regimes using 160–169.
 150 March, 2021 Int J Agric & Biol Eng Open Access at https://www.ijabe.org Vol. 14 No. 2

[39] Ma D, Chen L, Qu H, Wang Y, Misselbrook T, Jiang R. Impacts of Agric. Sci., 2007; 3: 747–756.
plastic film mulching on crop yields, soil water, nitrate, and organic carbon [52] Shirahatti M S, Itnal C J, Mallikarjunappa Gouda D S. Impact of
in Northwestern China: A meta-analysis. Agric. Water Manag., 2018; differential methods of irrigation on yield levels of cotton in red soils.
202: 166–173. Karnataka J. Agric. Sci., 2007; 20: 96–98.
[40] Legates D R, McCabe Jr G J. Evaluating the use of goodness-of fit [53] Nasrabad G G, Rajput T B S, Patel N. Soil water distribution and
measures in hydrologic and hydroclimatic model validation. Water simulation under subsurface drip irrigation in cotton (Gossypium hirsutum).
Resour. Res., 1999; 35: 233–241. Indian J. Agric. Sci., 2013; 83: 63–70.
[41] Licciardello F, Zema D A, Zimbone S M, Bingner R L. Runoff and soil [54] Arbat G P, Lamm F R, Abou Kheira A A. Subsurface drip irrigation
erosion evaluation by the AnnAGNPS model in a small Mediterranean emitter spacing effects on soil water redistribution, corn yield, and water
watershed. Trans. ASABE, 2007; 50: 1585–1593. productivity. Appl. Eng. Agric., 2010; 26: 391–399.
[42] Willmot C J. On the validation of models. Phys. Geogr., 1981; 2: [55] Bajracharya R M, Sharma S. Influence of drip-irrigation method on
184–194. performance and yields of cucumber and tomato. Int. J. Appl. Sci. Eng.
[43] Al-Ghobari H M, El-Marazky M A. Surface and subsurface irrigation Tech., 2005; 1: 1–7.
systems wetting patterns as affected by irrigation scheduling techniques in [56] Douh B, Boujelben A, Khila S, Bel Haj Mguidiche A. Effect of
an arid region. Afr. J. Agric. Res. 2012; 7: 5962–5976. subsurface drip irrigation system depth on soil water content distribution at
[44] Gan Y, Siddique Kadambot H M, Turner N C, Li, X G, Niu J Y, Yang C, different depths and different tines after irrigation. Larhyss J., 2013; 13:
et al. Ridge-furrow mulching systems-an innovative technique for 7–16.
boosting crop productivity in semiarid rain-fed environments. Adv. [57] Wang Y, Xie Z, Malhi S S, Vera C L, Zhang Y, Wang J. Effects of
Agron. 2013; 118: 429–476. rainfall harvesting and mulching technologies on water use efficiency and
[45] Dong Q, Yang Y, Yu K, Feng H. Effects of straw mulching and plastic crop yield in the semi-arid loess plateau, China. Agric. Water Manag.,
film mulching on improving soil organic carbon and nitrogen fractions, 2009; 96(3): 374–382.
crop yield and water use efficiency in the Loess Plateau, China. Agric. [58] Liu J, Bu L, Zhu L. Optimizing plant density and plastic film mulch to
Water Manag., 2018; 201: 133–143. increase maize productivity and water-Use efficiency in semiarid areas.
[46] Liu T, Wang B, Xiao H, Wang R,Yang B, Cao Q, et al. Differentially Agron. J., 2014; 106(4): 1138–1146.
improved soil microenvironment and seedling growth of Amorpha [59] Badr A E, Abuarab M E. Soil moisture distribution patterns under surface
fruticosa by plastic, sand and straw mulching in a saline wasteland in and subsurface drip irrigation systems in sandy soil using neutron
northwest China. Ecol Eng., 2018; 122: 126–134. scattering technique. Irrig. Sci., 2011; 31: 317–332.
[47] Mokh F, Nagaz K, Masmoudi M M, Mechlia N B. Effects of surface and [60] Rodríguez-Sinobas L, María G, Raúl S, Javier B. Evaluation of drip and
subsurface drip irrigation regimes with saline water on yield and water use subsurface drip irrigation in a uniform loamy soil. Soil Sci., 2012; 177:
efficiency of potato in arid conditions of Tunisia. J. Agric. Environ. Inter. 147–152.
Dev., 2014; 108: 227–246. [61] Zhao H, Wang R Y, Ma B L, Xiong Y C, Qiang S C, Wang C L, et al.
[48] Solomon K. Subsurface drip irrigation: product selection and Ridge-furrow with full plastic film mulching improves water use efficiency
performance. In: Jorsengen G S, Norum K N (Eds.), Subsurface drip and tuber yields of potato in a semiarid rainfed ecosystem. Field Crops
irrigation: theory, practices and applications. CATI Publication, No. Research, 2014; 161: 137–148.
9211001, 1993. [62] Li S, Li Y, Lin H, Feng H, Dyck M. Effects of different mulching
[49] Assouline S. The effects of microdrip and conventional drip irrigation on technologies on evapotranspiration and summer maize growth. Agric.
water distribution and uptake. Soil Sci. Soc. Am. J., 2002; 66: 1630–1636. Water Manag. 2018; 201: 309–318.
[50] Grabow G L, Huffman R L, Evans R, Jordan D, Nuti R C. Water [63] Ashrafuzzaman M, Abdul-Halim M, Ismail M R, Shahidullah S M,
distribution from a subsurface drip irrigation system and drip line spacing Hossain M A. Effect of plastic mulch on growth and yield of chilli
effect on cotton yield and water use efficiency in a coastal plain soil. (Capsicum annuum L.). Braz. Arch. Biol. Technol., 2011; 54: 321–330.
Trans. ASABE, 2006; 49: 1823–1835. [64] Thorburn P J, Cook F J, Bristow K L. Soil-dependent wetting from
[51] Badr M A. Spatial distribution of water and nutrients in root zone under trickle emitters: implications for system design and management. Irrig.
surface and subsurface drip irrigation and cantaloupe yield. World J. Sci., 2003; 22: 121–127.