sustainability

Article
Analysis of Runoff and Sediment Losses from a
Sloped Roadbed under Variable Rainfall Intensities
and Vegetation Conditions
Chunfeng Jia 1,2 , Baoping Sun 1, *, Xinxiao Yu 1 and Xiaohui Yang 3
1 School of Soil and Water Conservation, Beijing Forestry University, Beijing 100083, China;
jiachunfeng@bjfu.edu.cn (C.J.); yuxinxiao111@bjfu.edu.cn (X.Y.)
2 Shanxi Environmental Protection Institute of Transport (Co., LTD.), Taiyuan 030006, China
3 Chinese Academy of Forestry, Beijing 100864, China; xiaohuiyang@caf.ac.cn
* Correspondence: sunbaoping@bjfu.edu.cn




Received: 9 January 2020; Accepted: 5 March 2020; Published: 8 March 2020


Abstract: Vegetation plays an important role in reducing soil erosion. By exploring the allocation and
coverage of different types of vegetation, we can improve management practices that can significantly
reduce soil erosion. In this experiment, we study runoff and sediment losses on a shrub-grass planted,
grass planted, and bare slope under different rainfall intensities. Results showed that the runoff
generation time for the three subgrade types decreased as rainfall intensity increased (p < 0.05).
The slopes planted with either grass or shrub-grass were able to effectively delay runoff generation.
As rainfall intensity increased, the runoff amount increased for all treatments, with runoff in the bare
slope increasing the most. The runoff reduction rate from the shrub-grass slope ranged from 54.20%
to 63.68%, while the reduction rate from the slope only planted with grass ranged from 38.59% to
55.37%. The sediment yield from the bare slope increased from 662.66 g/m2 (15 mm/h) to 2002.95 g/m2
(82 mm/h) with increasing rainfall intensity in the plot. When compared with the bare slope, both
the shrub-grass and planted grass slopes were able to retain an additional 0.9 g/m2 to 4.9 g/m2 of
sediment, respectively. An accurate relationship between rainfall intensity, sloped vegetation types,
and runoff reduction rate was obtained by regression analysis and validated. These results can
provide a reference for improving soil and water conservation via improved vegetation allocation on
a sloped roadbed.

Keywords: precipitation; shrub-grass; soil erosion; soil and water conversation

1. Introduction
The process of modernization, construction, and land use change can have a negative impact on the
environment unless the processes by which changes are made are well-understood and sustainable [1].
When public transportation systems such as roads and railways are improved, the construction of
slopes via excavation and fill can change the original landform, increasing discharge and erosion from
sloped surfaces, with consequences for both natural and human systems. The speed and peak flow of
runoff caused by rainfall in hardened areas is much higher than that in non-hardened areas, which
results in frequent ponding in roads, flooding, and other storm water management problems [2–6].
The subgrade slope of the roadbed and the stability of the slope can be increased, which can ultimately
protect and improve local conditions. Adding vegetation to slopes is also beneficial by reducing
sediment loads associated with runoff, and has the added effect of regulating water resources and
conserving groundwater [7]. These practices can also help with ecological restoration and have
gradually become an important method of protecting railway subgrade slopes [1]. Many highway or
railway subgrade surfaces have concrete skeletons, or sashes, and greening to reduce runoff. Subgrade

Sustainability 2020, 12, 2077; doi:10.3390/su12052077 www.mdpi.com/journal/sustainability

Sustainability 2020, 12, x FOR PEER REVIEW 2 of 10
Sustainability 2020, 12, 2077 2 of 11

subgrade slopes [1]. Many highway or railway subgrade surfaces have concrete skeletons, or sashes,
and greening
greening typestovary,reduce
butrunoff. Subgrade
better results greening types
at landscape vary,been
scale have but better
achieved results at landscape
by planting grassscale
and
have been
adding achieved
irrigation by planting
systems, as well grass
as byand adding
using pureirrigation
grass. systems, as well as by using pure grass.
Festuca elata
Festuca elata Keng
Keng exex E.
E. Alexeev
Alexeev (Festuca)
(Festuca) isis aa cold-season
cold-season turf turf grass
grass that
that has
has been
been proven
proven to to have
have
good greening effects [8] and is widely used in greening construction
good greening effects [8] and is widely used in greening construction [9]. Photinia × fraseri Dress is [9]. Photinia × fraseri Dress is
commonly used
commonly used in in the
the construction
construction of of green
green projects,
projects, while
while thethe more
more widely-used
widely-used plant plant species
species are are
more suitable in east and Southwest China [10]. As a semi-deciduous
more suitable in east and Southwest China [10]. As a semi-deciduous shrub, Ligustrum japonicum is shrub, Ligustrum japonicum is
often used
often used forfor greening
greening and and is is well-adapted
well-adapted to to winter
winter conditions
conditions [11].[11]. In In recent
recent years,
years, most
most studies
studies
of greening vegetation have focused on the effects of pollutant removal
of greening vegetation have focused on the effects of pollutant removal and the relationships among and the relationships among
sediment production, rainfall, and slope. Vincent et al. [12] showed that
sediment production, rainfall, and slope. Vincent et al. [12] showed that the removal rate of ammonia the removal rate of ammonia
nitrogen by
nitrogen by grass
grass cover
cover waswas 268%
268% higher
higher than
than that
that of
of bare
bare land.
land. Anton
Anton et et al.
al. and
and Truman
Truman et et al.
al. [13,14]
[13,14]
report that these herbaceous plants have better heavy metal absorption,
report that these herbaceous plants have better heavy metal absorption, enrichment, and removal enrichment, and removal
efficiency in
efficiency in the
the rain.
rain. By
By using
using runoff plots, Blanco
runoff plots, Blanco et et al.
al. [9]
[9] found
found thatthat the
the runoff generation and
runoff generation and soil
soil
erosion modulus
erosion moduluswere werelinearly
linearlycorrelated
correlatedwith with rainfall
rainfall and and average
average rainfall
rainfall intensity.
intensity. Susana
Susana et al.et[15]
al.
[15] used runoff plots and found that roadbed slope engineering and
used runoff plots and found that roadbed slope engineering and greening can effectively prevent greening can effectively prevent
soil erosion.
soil erosion.
In order
In order to to better
better understand
understand the the effect
effect of
of grass-planting
grass-planting or or shrub-grass
shrub-grass plantingplanting on on reducing
reducing
runoff and soil erosion and increasing soil water infiltration, this study investigated
runoff and soil erosion and increasing soil water infiltration, this study investigated the rainfall yield the rainfall yield
and sediment
and sediment yieldyield using
using runoff plots for
runoff plots for aa sloped
sloped system
system withwith three different treatments
three different treatments and and five
five
different rainfall intensities. The objectives of this study were to: (i) explore
different rainfall intensities. The objectives of this study were to: (i) explore the law of runoff and the law of runoff and
sediment yield
sediment yield under
under different
different rainfall
rainfall intensities,
intensities, and (ii) evaluate
and (ii) evaluate whichwhich typestypes of of planting
planting and and
vegetation allocation
vegetation allocation havehave the the best
best soil
soil and
and water
water conservation
conservation benefits.
benefits. In In this
this experiment,
experiment, we we study
study
runoff and sediment losses on a shrub-grass planted, grass-planted,
runoff and sediment losses on a shrub-grass planted, grass-planted, and bare slope under different and bare slope under different
rainfall intensities.
rainfall intensities. ThisThis contrast
contrast provides
provides aa theoretical
theoretical reference
reference for for further
further exploring
exploring the the regulation
regulation
and control ability of subgrade slope vegetation on runoff and
and control ability of subgrade slope vegetation on runoff and sediment erosion rates, as sediment erosion rates, as well
well as as
reasons that planning slope greening may be effective in both highway
reasons that planning slope greening may be effective in both highway and railway construction. and railway construction.

2. Materials and Methods

2.1. Site Description

2.1. Site Description
The selected observation plot was located in the subgraded section of Jianning Qi Railway in
(31◦ 53’42” N,
Nantong City, Jiangsu Province (31°53’42” 121◦ 21’47”E).
N,121°21’47” E).The
Thesubgrade
subgradeslope
slopeisis7.0
7.0mmhigh
highwith
witha
aslope
slopeofof30°.◦
30 Each
. Eachexperimental
experimentalplotplotwas
wasmade
madeofofthree
threesquare
square concrete
concrete frames
frames of of each railway
subgrade.
subgrade. A A total
totalof
ofnine
nineplots
plotswere
wereselected
selectedforfor
this study.
this TheThe
study. boundary
boundaryof the
of experimental
the experimentalplot plot
was
brick and concrete
was brick smear,smear,
and concrete with runoff and sediment
with runoff collection
and sediment tanks below
collection tanks the sample
below the site. Thesite.
sample specific
The
specifications (reference standard runoff plot) of subgrade and sand sink are shown in Figure
specific specifications (reference standard runoff plot) of subgrade and sand sink are shown in Figure 1. 1.

Figure 1. Diagram of the sloped roadbed plot.

2.2. Soil Characteristics

2.2. Soil Characteristics
The
The test
test soil
soilwas
wasaayellow-brown
yellow-brownearth from
earth fromJiangsu Province.
Jiangsu TheThe
Province. thickness of the
thickness of overlying soil
the overlying
was 3
soil was about 30 cm. Before planting and greening, the density of the soil was measured to be 1.50.
about 30 cm. Before planting and greening, the density of the soil was measured to be 1.50 g/cm
Sustainability 2020, 12, 2077 3 of 11
Sustainability 2020, 12, x FOR PEER REVIEW 3 of 10
Sustainability 2020, 12, x FOR PEER REVIEW 3 of 10

g/cmparticle
The
g/cm 3.. The
3 The particle size composition was analyzed by a Nicomp (USA) 380N3000 laser particle size
size composition
particle was analyzed
size composition by a Nicomp
was analyzed (USA) 380N3000
by a Nicomp laser particle
(USA) 380N3000 laser size analyzer,
particle size
analyzer,
and the and the
results areresults
shown are
in shown
Figure in
2. Figure
analyzer, and the results are shown in Figure 2. 2.

Figure 2. Distribution
Figure 2.
Figure 2. Distribution of
Distributionof the
ofthe particle
the particle sizes
particle sizes of
sizes of soil
of soil used
soil used in
used in the
in the experiment.
the experiment.
experiment.

2.3. Vegetation Characteristics

2.3. Vegetation Characteristics
Grass seeds were sown on the grass slope and allowed to grow for about 120 days until the
Grass seeds were sown on the grass slope and allowed to grow for about 120 days until the
coverage (VFC) reached 95%, according to the grass square method [16]. The plots with grass cover
coverage (VFC) reached 95%, according to the grass square method [16]. The plots with grass cover
were
were planted
were planted withF.F.
plantedwith
with F.elata
elataKeng
elata Keng
Keng
exex
E. E.
Alexeev.
ex E. Alexeev.
Alexeev.
The plots
The
The
with
plots
plots
shrub-grass
with
with
werewere
shrub-grass
shrub-grass
planted
planted
were planted
F. elata
withwith
with F.F. Keng
elata
elata
ex E. Alexeev
Keng ex E. and L.and
Alexeev japonicum.
L. Two rows
japonicum. Two of trees
rows of were
trees planted
were acrossacross
planted a plotaofplot
1 moflength.
1 m Six trees
length. Six
Keng ex E. Alexeev and L. japonicum. Two rows of trees were planted across a plot of 1 m length. Six
were
trees planted
were in each
planted in row,
each for a total
row, for aoftotal
12 trees/m
of 12 per plot.
trees/m per The distance
plot. The between
distance the single
between the slope
single
trees were planted in each row, for a total of 12 trees/m per plot. The distance between the single
protection units was
slope protection
protection unitsabout 2.5 m (Figure
was about
about 2.5 m 3).
m (Figure
(Figure 3).
slope units was 2.5 3).

Figure 3.
Figure Diagram of
3. Diagram of the
the experimental
experimental plots: The (a–c)
plots: The (a–c) correspond
correspond to
to the
the shrub-grass,
shrub-grass, planted-grass
planted-grass
and bare slope, respectively.
respectively.
and bare slope, respectively.
2.4. Experimental Design
2.4. Experimental
Experimental Design
2.4. Design
The experimental area
The experimental
experimental was
area was located
was located
located in in
in aaa section
section
section ofof subgrade
of subgrade
subgrade underunder
under aaa real
real road.
real road. After the
road. After
After the subgrade
subgrade
The area the subgrade
fill was
was completed
fill was completedand andthe therunoff
runoff plot
plotwaswas built to meet
built to meetthe the
meet requirements,
requirements, a stainless-steel rainfall tube
fill completed and the runoff plot was built to the requirements, aa stainless-steel
stainless-steel rainfall
rainfall
was set
tube was up
was set to measure
set up
up to rainfall
to measure
measure rainfall in real
rainfall in time
in real and
real time to
time andcalculate
and to the
to calculate average
calculate the hourly
the average rainfall
average hourly intensity
hourly rainfall (mm/h)
rainfall intensity
intensity
tube
of the
(mm/h) ofmonitoring
of the site.
the monitoring In order
monitoring site. to
site. Inbetter
In order control
order to the
to better single
better control variable,
control the the
the single treatments
single variable,
variable, the with similar
the treatments
treatments rainfall
with
(mm/h) with
during
similar the observation
rainfall during period
the were
observation selected
period as the
were effective
selected observational
as the effective data. There
observational were different
data. There
similar rainfall during the observation period were selected as the effective observational data. There
amount of rainfall
were different
different on days
amount of with initial
rainfall on soil
days moisture content.
with initial
initial soilAs a result,content.
moisture the soil moisture content
As aa result,
result, the was
soil
were amount of rainfall on days with soil moisture content. As the soil
close to the
moisture content soil holding
content waswas close capacity
close toto the at
the soilthe beginning
soil holding
holding capacity of the
capacity at experiment
at the
the beginning (30%
beginning of ± 1%),
of the and
the experiment the
experiment (30%observations
(30% ±±1%),
1%),
moisture
reflected
and the the runoff andreflected
observations sediment theload of the
runoff andselected
sediment plotsload
underof specified
the rainfall
selected plotsconditions.
under specified
and the observations reflected the runoff and sediment load of the selected plots under specified
Theconditions.
rainfall number of rainfall intensity selected during the observation period is shown on Table 1.
rainfall conditions.
The types of underground
The number
number rainfallsubgrade
of rainfall intensity surfaces were
selected during
during divided into three period
the observation
observation types: isbare slope,
shown onaTable
grass 1.
Table slope,
The
The of intensity selected the period is shown on 1. The
and
types a shrub-grass
of underground slope. There
subgrade were three
surfaces werereplicate
divided plots
into for each
three type.
types: The
bare overall
slope, a rainfall
grass on
slope, each
and
types of underground subgrade surfaces were divided into three types: bare slope, a grass slope, and
subgrade type slope.
was 80There± 4 mm, with rainfall intensities of 15, 28,type.
40, 63, andoverall
82 mm/h, respectively
aa shrub-grass
shrub-grass slope. There were
were three
three replicate
replicate plots for
plots for each
each type. TheThe rainfall
overall rainfall on each
on each
being
subgrade testedtypein each
was subgrade.
80 ± 4 mm, with rainfall intensities of 15, 28, 40, 63, and 82 mm/h, respectively
subgrade type was 80 ± 4 mm, with rainfall intensities of 15, 28, 40, 63, and 82 mm/h, respectively
being tested
being tested in in each
each subgrade.
subgrade.
The runoff
The runoff generation
generation time time was
was recorded
recorded with with aa stopwatch
stopwatch duringduring the the rainfall
rainfall event.
event. After
After aa
single rainfall, an (L) event was measured, taking into account the specifications
single rainfall, an (L) event was measured, taking into account the specifications of the collecting tank of the collecting tank
and the depth of the runoff. After sediment deposition in the sedimentation
and the depth of the runoff. After sediment deposition in the sedimentation tank, most of the runoff tank, most of the runoff
Sustainability 2020, 12, 2077 4 of 11

Table 1. Significance test (Friedman Method) of the runoff generation time of the three tested different
vegetation types. SS-square of the standard deviation; df-degrees of freedom; MS-mean square.

The Difference of
SS df MS F-Value p-Value
Source
Between groups 38,568.133 2 5112.29 4.911 0.047
Within the group 94,227.600 12 7852.300 - -
Total 132,795.733 14 - - -

The runoff generation time was recorded with a stopwatch during the rainfall event. After a single
rainfall, an (L) event was measured, taking into account the specifications of the collecting tank and
the depth of the runoff. After sediment deposition in the sedimentation tank, most of the runoff flowed
out through the outlet at the bottom of the collection tank, but some evaporated. The wet sediment
was brought back to the laboratory, dried at 105 ◦ C, and its dry weight determined.

2.5. Runoff Reduction Rate

The runoff reduction rate, Q (%), was calculated as follows:

Q1 − Q2
Q(%) = × 100% (1)
Q1

where Q is the runoff reduction rate, Q1 is the total amount of runoff produced by bare slope, L; Q2 is
the runoff produced by either the planted grass or shrub-grass slopes, L.

2.6. Statistical Analysis

All statistical analyses were performed using the software package SPSS 16.0 (IBM, New York,
NY, USA). Descriptive statistics were used to calculate the mean and standard deviations for each set
of replicates. All data were tested for normal distribution and homogeneity of the variance analysis
to ensure that the data met the requirements of the variance analysis. A two-way ANOVA was
used to analyze differences in runoff generation time, runoff and sediment, with treatment type and
precipitation as independent factors. The least significant difference method was used to compare the
runoff generation time and runoff when necessary.

3. Results and Discussion

3.1. Runoff Generation Time

There were three treatments in the study and three plots in each treatment. Compared with a bare
slope, the slopes planted with grass and shrub-grass respectively showed a delaying effect on runoff
generation time. The runoff generation time of the three treatments under different rainfall intensities
is shown in Figure 4. This was due to the existence of surface vegetation, which increased the surface
roughness and prolonged the runoff path, and was also able to increase the soil infiltration intensity via
the root system [14,17,18]. In addition, the crown of the shrub-grass impacted the rainfall interception
effect, weakening the splash impact of raindrops on surface soil particles [19], enabling the soil to
maintain a strong infiltration rate for a long time, thus delaying the initial runoff generation time.
Sustainability 2020, 12, 2077 5 of 11
Sustainability 2020, 12, x FOR PEER REVIEW 5 of 10

180
Bare slope
160 Grass planting slope
Shrub-grass

140
Runoff generation time (min)

120

100

80

60

40

20

0
15 28 40 63 82
Precipitation (mm/h)
Figure 4.
Figure 4. Mean
Mean (±(± SD)
SD) runoff
runoff generation
generation time (min) under different rainfall intensities and three
three
different vegetation
different vegetation types.
types.

It
Ascan be also
rainfall seen from
intensity the figure
increased, thethat
runoff as rainfall
generation intensity increases, from
time decreased the generation timeand
2 h 20’, 30’57” for the
2h
bare,
33’ in planted
15 mm/hgrass, to (82and
mm/h),shrub-grass
to 50’47”,slopes
1 h 34’showed
and 7’10” a decreasing
in 82 mm/htrend, of the with the generation
planted-grass, times
shrub-grass
significantly
and bare slope, decreasing above
respectively. rainfallintensity
Rainfall intensities hadgreater than 28impact
a significant mm. The runoff
on the runoffgeneration
generation time in
time
the planted
for all three grass and shrub-grass
treatments. The runoff was longer time
generation than ofof the
the slope
bare slope (p = 0.047). and
with shrub-grass Compared
plantedto the
grass
bare slope, the delay
was significantly time associated
different from that of with
theusing a planted
bare slope (p < grass
0.01). slope was 38’28”.
The runoff generationThe time
delayoftimethe
associated with using
vegetated slope a shrub-grasshigher
was significantly slope than
was on average
that of the ofbare54’34”.
slopeThe (p <delay time on
0.5). These the shrub-grass
phenomena were
slopes
mainlywere due to41.2% higher
the fact thatthan
the that of the grass
infiltration slopes.
capacity This
of the showed
soil was close thattoa vegetation-covered
the soil at the start of slope
the
can effectively
experiment. prevent the
Vegetation suchformation
as shrubsof and runoff,
grassesprolonging the time
can effectively of runoff
improve formation.
the physical and Previous
chemical
studies
properties haveofalsosoil,found similar
increase soilresults
space,[20].
and increase soil water infiltration capacity [7]. However,
As rainfall
different vegetation intensity
types increased,
had differentthe runoff
impactsgeneration
on the time decreased
physical from 2 hproperties
and chemical 20’, 30’57” ofandthe2 hsoil.
33’
in 15 mm/h to (82 mm/h), to 50’47”, 1 h 34’ and 7’10” in 82 mm/h
Some studies have found that grasses provide better results than shrubs in reducing runoff of the planted-grass, shrub-grass and
bare slope, respectively.
generation time [21]. Rainfall intensity had a significant impact on the runoff generation time for
all three
Thistreatments.
was likely due Theto runoff generation
the existence time ofvegetation,
of surface the slope with whichshrub-grass
increasedand soil planted grass was
surface roughness,
significantly
lengthened the different
runofffrom
path, that
and of increased
the bare slope soil (p < 0.01). The
infiltration runoffthrough
capacity generation thetime
rootofsystem
the vegetated
[19]. In
slope was the
addition, significantly
crown ofhigher than that of
the shrub-grass the barethe
changed slope (p < 0.5). These
interception effect phenomena
of rainfall by were mainly due
weakening the
to the fact
splash that of
impact theraindrops
infiltrationon capacity
the surface of the
soilsoil
[22],was close toa the
allowing highersoilinfiltration
at the startforof the experiment.
a longer period
Vegetation
of time, thus such as shrubs
delaying the and
runoff grasses can effectively
generation time. improve the physical and chemical properties of
soil, increase soil space, and increase soil water infiltration capacity [7]. However, different vegetation
3.2. Total
types hadRunoff
different impacts on the physical and chemical properties of the soil. Some studies have
foundThe thattotal
grassesrunoffprovide
on the better
three results than shrubs
subgrade in reducing
types under runoff
different generation
rainfall time [21].
intensities is shown in
This was likely due to the existence of surface vegetation, which
Figure 5. When rainfall was held constant, increasing rainfall intensity caused the runoff of increased soil surface roughness,
all types
lengthened the runoff path, and increased soil infiltration capacity
of subgrade to increase. The total amount of runoff on the bare slope was significantly larger than through the root system [19].
In
that on the planted-grass and shrub-grass slopes (p < 0.001), indicating that the combined effectthe
addition, the crown of the shrub-grass changed the interception effect of rainfall by weakening of
planting grass and shrub-grass was able to reduce runoff (42.7%) from the slopes. Compared with
the shrub-grass slope, runoff in the planted-grass slope was not significantly different for the different
Sustainability 2020, 12, 2077 6 of 11

splash impact of raindrops on the surface soil [22], allowing a higher infiltration for a longer period of
time, thus delaying the runoff generation time.

3.2. Total Runoff

The total runoff on the three subgrade types under different rainfall intensities is shown in Figure 5.
When rainfall was held constant, increasing rainfall intensity caused the runoff of all types of subgrade
to increase. The total amount of runoff on the bare slope was significantly larger than that on the
planted-grass and shrub-grass slopes (p < 0.001), indicating that the combined effect of planting grass
and shrub-grass was able to reduce runoff (42.7%) from the slopes. Compared with the shrub-grass
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 10
slope, runoff in the planted-grass slope was not significantly different for the different rainfall intensities
(p = 0.196).
rainfall This indicated
intensities thatThis
(p = 0.196). the runoff wasthat
indicated closely associated
the runoff was with each
closely rainfall intensity
associated with each from the
rainfall
planted
intensitygrass
fromslope. This was
the planted mainly
grass because
slope. This the
wasflow velocity
mainly becauseon the
thesloped surface was
flow velocity lowsloped
on the when
there was vegetation coverage and the rainfall intensity was low. Additionally, the
surface was low when there was vegetation coverage and the rainfall intensity was low. Additionally, shear force of the
sloped flow on the soil surface was weak and flow velocity was low, while the runoff
the shear force of the sloped flow on the soil surface was weak and flow velocity was low, while the generation time
spent
runoffflowing
generationthrough timethespent
wholeflowing
surface was long.the
through Thiswhole
increased the infiltration
surface was long. effect
This of the soil and
increased the
reduced the runoff and sediment production, weakening soil erosion.
infiltration effect of the soil and reduced the runoff and sediment production, weakening soil erosion.

70
Bare slope
Grass planting slope
60 Shurb-grass

50
Runoff (L)

40

30

20

10

0
15 28 40 63 82
Rainfall intensity (mm/h)
Figure 5.
Figure 5. Mean
Mean (±
(± SD)
SD) runoff
runoff for
for the
the three
three different subgrade types
different subgrade types under
under different
different rainfall
rainfall intensities.
intensities.

Themean
The meanrunoff
runoffvalues
valuesforforthethe grass
grass slope
slope were
were 13.10
13.10 L, 22.58
L, 22.58 L, 25.84
L, 25.84 L, 27.26
L, 27.26 L, andL,33.20
andL33.20
underL
under
the fivethe five intensities,
rainfall rainfall intensities,
respectively (p < 0.05) (p
respectively < 0.05)
(Table (Table
2). The 2). The
average average
values values
of the of the were
total runoff total
runoff
10.66 L, were
16.96 10.66 L, L,
L, 17.91 16.96
20.53L, L,
17.91
and L, 20.53
26.41 L, and
L for 26.41 L for slope,
the shrub-grass the shrub-grass
respectively (p < respectively
slope, 0.05). The mean(p <
0.05). The
value mean
of total valuewas
runoff of total
29.35runoff wasL,29.35
L, 41.46 42.08L,L,41.46 L,L,
51.54 42.08
57.67L, L
51.54 L, 57.67
for the bareLslope,
for the bare slope,
respectively.
respectively.
The differenceThe
leveldifference level was also
was also significant (p < significant (p < 0.05).
0.05). This showed This
that showed
rainfall that rainfall
intensity intensity
has a significant
has a significant
impact on the totalimpact
runoffonofthethetotal runoff
three of the three
subgrade types.subgrade
This wastypes. Thisrainfall
because was because
was notrainfall
only was
the
not only
cause the cause
of runoff, butofalso
runoff, but alsoofthe
the source source
runoff. A of runoff.
large A large
number numberhave
of studies of studies
foundhave found
that the that
higher
the rainfall
the higher the rainfallthe
intensity, intensity,
larger the themiddle
larger the middleof
diameter diameter of raindrops,
raindrops, which is more which is more conducive
conducive to runoff
to runoff generation
generation [23]. When [23]. Whenintensity
rainfall rainfall intensity
and otherand other conditions
conditions are fixed,are fixed,
runoff andrunoff and sediment
sediment volume
volumedepend
mainly mainly ondepend on rainfall.
rainfall.

Table 2. Significance test (Friedman Method) amounts on the runoff from three different vegetation
types. SS—standard deviation squared; df—degrees of freedom; MS—mean squared value.

The Difference of Source SS df MS F-Value p-Value

Between groups 1846.586 2 923.293 13.558 0.001
Within the group 817.169 40 68.097 4.106 0.093
Total 2663.755 44 - - -
Sustainability 2020, 12, 2077 7 of 11

Table 2. Significance test (Friedman Method) amounts on the runoff from three different vegetation
types. SS—standard deviation squared; df—degrees of freedom; MS—mean squared value.

The Difference of Source SS df MS F-Value p-Value

Between groups 1846.586 2 923.293 13.558 0.001
Within the group 817.169 40 68.097 4.106 0.093
Total 2663.755 44 - - -

Shrub-grass slopes had the greatest impact on total runoff, with average reduction rates of 58.92%.
On grass slopes, the runoff reduction rate was 45.81%. The combination of planting and irrigating
grass can play a large role in increasing soil and water conservation, regulating and storing rainwater
runoff, and increasing soil infiltration in subgrade slopes, such as those used in railways or highways.
Table 3 shows a linear relationship between runoff reduction rate and rainfall intensity. The runoff
reduction rate associated with the planted-grass and shrub-grass slopes decreased as rainfall intensity
increased. This is partially due to the effect of the canopy interception, coupled with its associated
effect on soil infiltration. When rainfall intensity is high, the interception effect is reduced, and the
existence of a biological crust on the soil will affect the runoff generation time of the planted-grass
and shrub-grass slopes. The existence of vegetation on a slope makes the infiltration rate of soil lower
than the slope [24–26]. The combination of these two factors decreases the runoff reduction rate when
rainfall intensity is high. The runoff was extremely significantly influenced by rainfall intensity, which
is consistent with the results of Polyakov et al. [27]. The runoff generation time in each experimental
area was related to the water content of the soil. If the soil moisture content was monitored in real time
and the same water content was used, the relationship between rainfall intensity and runoff generation
time could be better quantified.

Table 3. Total runoff volume and reducing rate.

Bare Slope Shrub-Grass Reducing Rate Reducing Rate of

Rainfall Grass Runoff
Runoff Combined Runoff of Grass Shrub-Grass
Density Volume/L
Volume/L Volume/L Slop/% Combined Slop/%
15 mm·h−1 29.35 13.1 10.66 55.37 63.68
28 mm·h−1 41.46 22.58 16.96 45.54 59.09
40 mm·h−1 42.08 25.84 17.91 38.59 57.44
63 mm·h−1 51.54 27.26 20.53 47.11 60.16
82 mm·h−1 57.67 33.2 26.41 42.43 54.20
Average 44.42 24.40 18.49 45.81 58.92

3.3. Sediment Yield

Under different rainfall intensities, the sediment yield from the three subgrade surfaces is as
shown in Figure 6. As rainfall intensity increased, the amount of sediment produced by the bare slope
increased exponentially (R2 ≤ 0.98), especially at intensities above 60 mm. The sediment yield from
planted-grass and shrub-grass slopes was significantly lower than that from the bare slope, and the
difference in sediment yield between the shrub-grass and planted-grass slopes was very significant
(p < 0.01). For lower rainfall intensities, the sediment yield from the grass and shrub-grass slopes did
not increase significantly with rainfall intensity (p > 0.05), but were not significantly different from the
bare slope (p > 0.05).
increased exponentially (R ≤ 0.98), especially at intensities above 60 mm. The sediment yield from
planted-grass and shrub-grass slopes was significantly lower than that from the bare slope, and the
difference in sediment yield between the shrub-grass and planted-grass slopes was very significant
(p < 0.01). For lower rainfall intensities, the sediment yield from the grass and shrub-grass slopes did
not increase significantly with rainfall intensity (p > 0.05), but were not significantly different from
Sustainability 2020, 12, 2077 8 of 11
the bare slope (p > 0.05).

2200 6 5
y=0.28x2 -6.86x+701.21 a y=0.0007x2 -1.02x+1.49 b y=0.0006x2 -0.013x+1.12 c
2000
R2 =0.99 R2 =0.90 R2 =0.77
1800 P=0.0002 5 4
P=0.03 P=0.11
Sediment yield (g)

1600
4 3
1400
1200
3 2
1000
800 2 1
600
400 1 0
10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90
Rainfall intensity (mm/h) Rainfall intensity (mm/h) Rainfall intensity (mm/h)

Sedimentyield
Figure6.6.Sediment
Figure yieldunder
underdifferent
differentrainfall
rainfallintensities
intensitiesand
anddifferent
differentvegetation
vegetationtypes.
types.The
The(a–c)
(a–c)
were bare slope, grass planting and shrub-grass planting, respectively.
were bare slope, grass planting and shrub-grass planting, respectively.

As rainfall intensity increased, the sediment yield from the bare slope increased sharply from
As rainfall intensity increased, the sediment yield from the bare slope increased sharply from
662.66 g/m2 2 (15 mm/h) to 2002.95 g/m2 2 (82 mm/h), while the sediment concentration from the planted
662.66 g/m (15 mm/h) to 2002.95 2g/m (82 mm/h), while the sediment concentration from the planted2
grass slope was only 1.5−4.9 g/m , and the sediment yield from the shrub-grass slope was 0.9−4.5 g/m .
grass slope was only 1.5−4.9 g/m2, and the sediment yield from the shrub-grass slope was 0.9−4.5 g/m2.
The difference between the shrub-grass and grass slopes was not significant, and the soil conservation
The difference between the shrub-grass and grass slopes was not significant, and the soil conservation
effect was obvious. This was mainly because the vegetation reduced or even prevented splash erosion,
thereby being able to reduce runoff.

3.4. Effects of Rainfall Intensity on Runoff Reduction Rate

There is a significant correlation between rainfall intensity and runoff reduction rate. María et
al.’s research showed that the relationship between runoff, sediment and rainfall intensity was a power
function [28]. Gessesse et al.’s study on land use patterns with vegetative cover showed that the fitting
effect between rainfall and runoff by a curtain function was better [29]. Previous studies suggested that
runoff generation rate increased as a power function with rainfall intensity and slope [30–33]. It has
been found that the relationship of runoff and rainfall intensity is a power function for slopes of 10◦ ,
15◦ , and 20◦ [34]. A regression analysis of rainfall intensity, and runoff reduction rate produced the
following equation:
Q(%) = (−0.791lnI − 0.893) × 100% (R2 = 0.90)

In the formula above, Q denotes the runoff reduction rate (%) and I denote rainfall intensity
(mm/min).
The simulated and measured runoff reduction rates under varying rainfall intensities and
treatments were compared with the measured values using the above formula. It can be seen from
the table that the reduction rate obtained by the simulation was close to the measured value, and the
absolute error between the simulated value and the measured value ranged from 1.0%−6.6%, which
indicated that the accuracy of the simulated value obtained by the equation was good. The absolute
error between the average value of the simulated value and the average value of the measured value
was 5.9%, which indicated that the fitting effect of the above formula was satisfactory. This was
mainly because the yellow-brown soil itself was porous and had good infiltration. Infiltration directly
determines the amount of runoff. The presence of vegetation cover mainly affects runoff generation by
affecting the raindrop kinetic energy and infiltration. With yellow-brown soil, a small vegetated area
had a great impact on the production of sediment, while using the shrub-grass combination had a small
additional impact on the production of sediment [35–38]. The effects of vegetation cover and rainfall
intensity on runoff and sediment were the same. The effects of rainfall duration and soil bulk density on
runoff and sediment in different soils were significantly different, especially on yellow-brown soil [39].

4. Conclusions
The runoff generation times from a bare slope, planted-grass slope, and shrub-grass slope were
negatively correlated with rainfall intensity. The greater the rainfall intensity, the shorter the initial
Sustainability 2020, 12, 2077 9 of 11

runoff generation time, and the effect of rainfall intensity on runoff yield was very significant. The effect
of planted grass and shrub-grass on the runoff generation time was significant difference. As rainfall
intensity increased, the runoff from the three vegetation types tended to increase, and the rainfall
intensity had a significant effect on the total runoff amount. The total runoff from the planted-grass
slope and the shrub-grass slope was significantly different from that of a bare slope. The total runoff
from shrub-grass and grass slopes was reduced compared to a bare slope, but the difference between
the treatments was not significant, although the runoff on the shrub-grass slope was slightly higher.
This analysis shows that slightly highly runoff reduction effects could be achieved by vegetating sloped
subgrade areas, especially when combining of shrub and grass. As rainfall intensity increased, the
sediment the bare slope increased exponentially, especially when the rainfall intensity was higher than
60 mm. The difference in sediment yield between the shrub-grass, grass, and the bare slope was very
significant, showing that greening had a very significant and positive effect on reducing soil erosion.

Author Contributions: C.J. wrote the paper; B.S. conceived and designed the experiments; X.Y. and X.Y. analyzed
the data. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by “National Natural Science Foundation of China, grant number 41430747”
and “National Science and Technology Support Project, grant number 2015BA0804219”.
Acknowledgments: We thank MogoEdit (http://www.mogoedit.com) for linguistic assistance during the
preparation of this manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Liu, Z.Q.; Jia, G.D.; Yu, X.X. Variation of water uptake in degradation agroforestry shelterbelts on the North
China Plain. Agric. Ecosyst. Environ. 2020, 1, 106697. [CrossRef]
2. Christian, V.; Agus, F.; Alamban, R. Runoff and sediment losses from 27 upland catchments in Southeast Asia:
Impact of rapid land use changes and conservation practices. Agric. Ecosyst. Environ. 2008, 128, 225–238.
3. Pan, C.Z.; Zhou, P.; Shang, G. Runoff hydraulic characteristics and sediment generation in sloped grassplots
under simulated rainfall conditions. J. Hydrol. 2006, 331, 178–185. [CrossRef]
4. Chaplot, V. Impact of DEM mesh size and soil map scale on SWAT runoff, sediment, and NO3-N loads
predictions. J. Hydrol. 2010, 312, 207–222. [CrossRef]
5. Imeson, A.C.; Prinsen, H.A.M. Vegetation patterns as biological indicators for identifying runoff and sediment
source areas for semi-arid landscapes in Spain. Agric. Ecosyst. Environ. 2004, 104, 333–342. [CrossRef]
6. Jia, C.F.; Sun, B.P.; Yu, X.X.; Yang, X.H. Evaluation of vegetation restoration along an expressway in a cold,
arid, and desertified area of China. Sustainability 2019, 11, 2313. [CrossRef]
7. Liu, X.; Fan, D.; Yu, X.; Liu, Z.; Sun, J. Effects of simulated gravel on hydraulic characteristics of overland
flow under varying flow discharges. Sci. Rep. 2019, 9, 19781. [CrossRef] [PubMed]
8. Blanco, C.H.; Gantzer, C.J.; Anderson, S.H. Grass barrier and vegetative filter strip effectiveness in reducing
runoff, sediment, nitrogen, and phosphorus loss. Soil Sci. Soc. Am. J. 2014, 68, 1670. [CrossRef]
9. Peter, F.; Auerswald, K. Effectiveness of grassed waterways in reducing runoff and sediment delivery from
agricultural watersheds. J. Environ. Qual. 2003, 32, 927–936.
10. Li, X.H.; Yang, J.; Zhao, C.Y. Runoff and sediment from orchard terraces in southeastern China. Land Degrad.
Dev. 2014, 25, 184–192. [CrossRef]
11. Ashok, M.; Kar, S.; Singh, V.P. Determination of runoff and sediment yield from a small watershed in
sub-humid subtropics using the HSPF model. Hydrol. Process. 2007, 21, 3035–3045.
12. Vincent, C.; Jean, P. Sediment, soil organic carbon and runoff delivery at various spatial scales. Catena 2009,
88, 46–56.
13. Polyakov, V.O.; Nearing, M.A.; Nichols, M.H. Long-term runoff and sediment yields from small semiarid
watersheds in southern Arizona. Water Resour. Res. 2010, 46, 204–216. [CrossRef]
14. Anton, L.A.; Sampurno, B.C.; Kukuh, S. Runoff and sediment yield from rural roads, trails and settlements
in the upper Konto catchment, East Java, Indonesia. Geomorphology 2012, 87, 28–37.
15. Truman, C.C.; Potter, T.L.; Nuti, R.C.; Franklin, D.H.; Bosch, D.D. Antecedent water content effects on runoff
and sediment yields from two Coastal Plain Ultisols. Agric. Water Manag. 2011, 98, 1196. [CrossRef]
Sustainability 2020, 12, 2077 10 of 11

16. Susana, B.; Angeles, G.M.; Jamal, B. Plant spatial pattern predicts hillslope runoff and erosion in a semiarid
mediterranean landscape. Ecosystems 2007, 10, 987–998. [CrossRef]
17. Norton, J.B.; Sandor, J.A.; White, C.S. Runoff and sediments from hillslope soils within a Native American
agroecosystem. Soil Sci. Soc. Am. J. 2007, 71, 476–483. [CrossRef]
18. Mantzos, N.; Karakitsou, A.; Hela, D.; Patakioutas, G.; Leneti, E.; Konstantinou, I. Persistence of oxyfluorfen
in soil, runoff water, sediment and plants of a sunflower cultivation. Sci. Total Environ. 2014, 472, 767–777.
[CrossRef]
19. Zhang, G.; Liu, G.; Wang, G. Effects of canopy and roots of patchy distributed artemisia capillaris on runoff,
sediment, and the spatial variability of soil erosion at the plot scale. Soil Sci. 2012, 177, 409–415. [CrossRef]
20. You, J.Y.; Quan, H.D.; Yuan, Y.F.; Peng, X.D.; Zhao, L.S.; Yang, J. Effects of rainfall intensity on runoff and
sediment yields on bare slopes in a karst area, SW China. Geoderma 2018, 330, 30–40.
21. Silburn, D.M.; Glanville, S.F. Management practices for control of runoff losses from cotton furrows under
storm rainfall. I. Runoff and sediment on a black Vertosol. Aust. J. Soil Res. 2002, 40, 1–20. [CrossRef]
22. Zhang, Y.E.; Wang, D.D.; Liu, Z.Q.; Jia, G.D.; Yu, X.X.; Chen, L.H. Assessment of leaf water enrichment of
Platycladus orientalis using numerical modeling with different isotopic models. Ecol. Indic. 2020, 1, 106069.
[CrossRef]
23. Montgomery, D.R. Soil erosion and agricultural sustainability. Proc. Natl. Acad. Sci. USA 2007, 104,
13268–13272. [CrossRef] [PubMed]
24. Burguet, M.; Guzmán, G.; De Luna, E.; Taguas, E.V.; Gomez, J.A. Evaluation of disruption of sediment
connectivity and herbicide transport across a slope by grass strips using a magnetic iron oxide tracer.
Soil Tillage Res. 2018, 180, 268–281. [CrossRef]
25. Yang, D.; Kanae, S.; Oki, T.; Koike, T.; Musiake, K. Global potential soil erosion with reference to land use
and climate changes. Hydrol. Process. 2003, 17. [CrossRef]
26. De las Heras, M.M.; Nicolau, J.M.; Merino-Martín, L.; Wilcox, B.P. Plot-scale effects on runoff and erosion
along a slope degradation gradient. Water Resour. Res. 2010, 46. [CrossRef]
27. Lesschen, J.P.; Schoorl, J.M.; Cammeraat, L.H. Modelling runoff and erosion for a semi-arid catchment using
a multi-scale approach based on hydrological connectivity. Geomorphology 2009, 109, 174–183. [CrossRef]
28. Wang, J.; Huang, J.; Wu, P.; Zhao, X. Application of neural network and grey relational analysis in ranking
the factors affecting runoff and sediment yield under simulated rainfall. Soil Res. 2016, 54, 291. [CrossRef]
29. Sérgio, A.P.; Malvar, M.C.; Diana, C.S.V.; Lee, M.; Jan, J.K. Effectiveness of hydro mulching to reduce runoff
and erosion in a recently burnt pine plantation in central Portugal. Land Degrad. Dev. 2016, 27. [CrossRef]
30. Polyakov, V.O.; Lal, R. Soil organic matter and CO2 emission as affected by water erosion on field runoff
plots. Geoderma 2008, 143, 216–222. [CrossRef]
31. María, J.M.; Ramón, B.; Luis, J.; Raquel, P.-R. Effect of vegetal cover on runoff and soil erosion under light
intensity events. Rainfall simulation over USLE plots. Sci. Total Environ. 2007, 378, 161–165.
32. Gessesse, B.; Bewket, W.; Achim, B. Model-based characterization and monitoring of runoff and soil erosion
in response to land use/land cover changes in the modjo watershed, Ethiopia. Land Degrad. Dev. 2015, 26,
711–724. [CrossRef]
33. Deasy, C.; Quinton, J.N.; Silgram, M.; Bailey, A.P.; Jackson, B.; Stevens, C.J. Mitigation options for sediment
and phosphorus loss from winter-sown arable crops. J. Environ. Qual. 2009, 38, 2121. [CrossRef] [PubMed]
34. Cerdà, A.; Keesstra, S.D.; Rodrigo-Comino, J.; Novara, A.; Pereira, P.; Brevik, E.; Giménez-Morera, A.;
Fernández-Raga, M.; Pulido, M.; di Prima, S.; et al. Runoff initiation, soil detachment and connectivity are
enhanced as a consequence of vineyards plantations. J. Environ. Manag. 2017, 202, 268–275. [CrossRef]
35. Uchida, T.; Asano, Y. Spatial variability in the flowpath of hillslope runoff and streamflow in a meso-scale
catchment. Hydrol. Process. 2010, 24, 2277–2286. [CrossRef]
36. Murphy, T.; Dougall, C.; Burger, P.; Carroll, C. Runoff water quality from dryland cropping on Vertisols in
Central Queensland, Australia. Agric. Ecosyst. Environ. 2013, 180, 21–28. [CrossRef]
37. Helmers, M.J.; Zhou, X.; Asbjornsen, H.; Kolka, R.; Tomer, M.D.; Cruse, R.M. Sediment removal by prairie
filter strips in row-cropped ephemeral watersheds. J. Environ. Qual. 2012, 41, 1531–1539. [CrossRef]
Sustainability 2020, 12, 2077 11 of 11

38. Brazier, R.E.; Beven, K.J.; Freer, J.; Rowan, J.S. Equifinality and uncertainty in physically based soil erosion
models: Application of the glue methodology to WEPP-the water erosion prediction project-for sites in the
UK and USA. Earth Surface Process. Landf. 2015, 25, 825–845. [CrossRef]
39. Pimentel, D.; Kounang, N. Ecology of soil erosion in ecosystems. Ecosystems 1998, 1, 416–426. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).