Physics and Chemistry of the Earth 104 (2018) 66–75

Contents lists available at ScienceDirect

Physics and Chemistry of the Earth

journal homepage: www.elsevier.com/locate/pce

Projected land use changes impacts on water yields in the karst mountain T
areas of China
Yanqing Langa,b, Wei Songa,∗, Xiangzheng Denga,c
a
Key Laboratory of Land Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101,
China
b
College of Architecture and Urban Planning, Chongqing Jiaotong University, Chongqing 400041, China
c
Center for Chinese Agricultural Policy, Chinese Academy of Sciences, Beijing 100101, China

A R T I C L E I N F O A B S T R A C T

Keywords: Human-induced land use changes over short time scales have significant impacts on water yield, especially in
Karst mountain area China because of the rapid social economic development. As the biggest developing country of the world, China's
Projected land use change economy is expected to continuously grow with a high speed in the next few decades. Therefore, what kind of
Water yield land use changes will occur in the future in China? How these changes will influence the water yields? To
Ecosystem services
address this issue, we assessed the water yields in the karst mountain area of China during the periods of
1990–2010 and 2010–2030 by coupling an Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST)
model and a Conversion of Land Use and its Effects (CLUE) model. Three different land use scenarios i.e. natural
growth, economic development, and ecological protection, were developed in 2030 using the CLUE model. It
was concluded that, given land use changes between 1990 and 2010, total water yields in the karst mountain
area are characterized by a trend towards fluctuating reduction. However, total water yields of 2030 in the
economic development scenario revealed an increase of 1.25% compared to the actual water yields in 2010. The
economy development in karst mountain areas of China in the future has a slight positive influence on water
yields.

1. Introduction
Ecosystem services play a vital role in the sustainable development
of human societies and social economies (Daily, 1997; Marquès et al.,
2013; Song and Deng, 2015; Song et al., 2015), and form the basis of
survival and development (Costanza, 1997; Deng et al., 2013). Eco-
system services not only maintain the stability of global ecological
systems, but also provide a variety of resources useful for humans, in-
cluding food and raw materials (Bennett et al., 2009; Song and Deng,
2017). These concepts have therefore become a hot research topic, at
the forefront of ecology and related fields (Han et al., 2016; Liu et al.,
2017a; Zhao et al., 2017).
The supply of water is one important component of ecosystem ser-
vices that is strongly connected to regional natural and economic
conditions (Jie et al., 2015; Liquete et al., 2011). On the one hand,
water yields influence the regional levels of this resource, essential for
human survival and development (Tao et al., 2016; Wang and
Blackmore, 2009; Zhang et al., 2010); therefore, mismatches between
water supply and demand will restrict the sustainable development of a
regional economy (Deng and Zhao, 2015; Li et al., 2013; Wu et al.,
2017). Because water yields are also closely related to regional natural
geographic conditions and human activities (Sun et al., 2016; Yang
et al., 2016), the latter will directly impact the volume of this resource
(Smith, 1997). In contrast, anthropogenic activities also influence
yields and the availability of this resource by affecting climate, land
use, and water quality (Ming et al., 2016; Liu et al., 2017b; Pervez and
Henebry, 2015).
In light of ever increasing demands for water and the impacts of
climate change on this resource, supply has tended to decrease as an
ecosystem service while shortages have become increasingly marked
(Cademus et al., 2014). As a result, conflicts caused by water usage
have intensified, greatly threatening the sustainable development of
impacted regions (Cheng et al., 2017; Yi et al., 2016). Studies that
address water yields have increased as a result, including those that
evaluate regional characteristics and influencing factors with the aim of
developing effective approaches to mitigate this resource crisis, im-
prove utility, ensure safety, and strengthen management (Benda et al.,
2016; Little and Lara, 2010; Mark and Dickinson, 2008).
Previous assessment-based studies have tended to focus on temporal
and spatial variations in water yields at the watershed scale as well as

∗
Corresponding author.
E-mail address: songw@igsnrr.ac.cn (W. Song).

https://doi.org/10.1016/j.pce.2017.11.001
Received 15 August 2016; Received in revised form 6 September 2017; Accepted 2 November 2017
Available online 06 November 2017
1474-7065/ © 2017 Elsevier Ltd. All rights reserved.
Y. Lang et al.
Fig. 1. Geographical location of the study area.
on factors that influence these changes. Gang et al. (2017) noted that
the water yields of South Korean forests more-or-less doubled between
1971 and 2010 as the result of changes in forest stock and climate while
a spatially homogeneous supply capacity was maintained. In related
work, Keeler and Polasky (2014) evaluated the spatial yield char-
acteristics of two watersheds in Minnesota, while Zhang and Wenhua
(2012) assessed the Xitiaoxi River and showed that the spatial dis-
tribution of regional water yield was similar to that of rainfall. Climate
and land use changes are both known to be key influences on water
yield; these factors were shown to be important in studies by
Lópezmoreno et al. (2010) and Lu et al. (2013) who both noted that the
latter exerts less influence on regional water yields than the former.
Similarly, Pan et al. (2015) noted the significant influence of climate
and land use changes on the regional water supply of the Yellow River
catchment area. These works showed that, depending on timescale,
these factors had both positive and negative effects on regional water
supply. Specifically, climate change mainly affects water supply by al-
tering precipitation and potential evapotranspiration; thus, the ecolo-
gical degradation and underlying surface changes caused by this factor
as well as variations in land use are thought to be the main forces
driving water yield changes within the area of the Yellow River
catchment. Yu et al. (2015) assessed the water yield of Dali County,
China, and showed that precipitation, the expansion of built-up areas,
and the implementation of land use policies were the three dominant
factors impacting water yield changes, while Principe (2012) explored
the temporal and spatial relationship between climate, land use change,
and water yield by estimating watershed yields.
In addition to the assessment of water yield and the discussion of
influencing factors, increasing importance has been attached to un-
derstanding underlying mechanisms. In this context, Fricke (2014)
evaluated the impacts on the water supply of the city of Urumqi using a
regional yield simulation based on social and economic models that
took into account different climate and land use change scenarios. Si-
milarly, Hosseini et al. (2012) showed that land use changes influenced
water balance by causing gradual increases in surface runoff and con-
tinual decreases in groundwater flow. Bossa et al. (2014) later
67
Physics and Chemistry of the Earth 104 (2018) 66–75
investigated the influence of climate and land use changes on the
supply of water resources by applying scenario simulations and calcu-
lating contribution rates to understand the relative impacts of these
factors on yields.
The karst mountain areas of China tend to be fragile geological
environments (Tong et al., 2016). Since the implementation of the
‘grain-for-green’ policy in China in the 1990s, the ecological environ-
ment of karst mountain areas has improved, but water supply remains a
significant problem that restricts social and economic development. The
karst mountain areas of China faced double pressure from both the
economic development and the ecological environment protection.
Therefore, it is crucial to assess how has the water yield changed under
these pressures in the past and future.
To answer the question, we evaluated the impacts of future devel-
opment scenarios on water yields and explore the influence of past and
current land use changes via coupled use of the Integrated Valuation of
Ecosystem Services and Tradeoffs (InVEST) and the Conversion of Land
Use and its Effects (CLUE) models. This approach enables us to provide
reference data that will underpin future sustainable development in
these regions. The key aims of this research are: 1) To document tem-
poral and spatial variations in land use change within karst mountain
areas; 2) To quantitatively evaluate the water yields of these areas and
the influence of land use change given constant climatic conditions (via
the InVEST model), and; 3) To evaluate the influence of land use change
on water yields given different scenarios for development (coupling the
InVEST and CLUE models).
2. Study region
The region evaluated in this study is located in south and southwest
China (22°12′-28°22′ N and 104°29′-110°38′ E) and encompasses the
majority of Guizhou and Guangxi provinces, a total area of ca.
214,100 km2 between zero and 3000 m above sea level. The average
annual rainfall in this region ranges between 900 mm and 1850 mm,
while average annual temperatures range between 13 °C and 25 °C.
This region is a typical Chinese karst mountain area (Fig. 1),
Y. Lang et al. Physics and Chemistry of the Earth 104 (2018) 66–75

Fig. 2. Driving forces of land use change simulation.

68
Y. Lang et al.
comprising small and medium-sized undulating mountainous land that
gradually increases in elevation from southeast-to-northwest; the karsts
in this area also gradually change from plateau-peak-cluster types to
hilly-plain-hoodoo physiognomies along this transect. Water systems in
this region include the Sanchahe, Yachihe, Qingshuijiang, Nanpanjiang,
Hongshuihe, Liujiang, Yujiang, and Qianjiang rivers as well as a nu-
merous smaller tributaries and channels.
3. Methods and data sources
3.1. Water yield assessment
We adopted the InVEST model to assess the water yields in the karst
mountain area of China (Ranganathan, 2011). The InVEST model has
been utilized across a range of disparate research fields including bio-
diversity, environmental ecology, water (e.g., hydropower) and land
resources (Hamel et al., 2015; Leh et al., 2013); InVEST visualizes
quantified ecosystem service functions in the form of maps given a low
level of input data, but generates high volumes of output data and
simplifies complex issues, thereby increasing problem-solving ability.
The model comprises multiple components, including soil conservation,
biodiversity, carbon reserve, water yield, and nutrition interception
modules. In the context of this study, the water yield module embedded
within the InVEST model provides an estimation method based on
water balance (Budyko, 1975). The detailed descriptions of water yield
module including the parameter settings can be found in Lang et al.
(2017).
3.2. Simulating land use changes
We adopted the CLUE model to forecast the land use changes in
karst mountain areas of China in 2030. CLUE model was developed by
the scientists of Wageningen University (Verburg et al., 2002). The
CLUE model is based on the empirically quantitative relationship be-
tween land use and the underlying factors that drive changes in this
variable, including dynamically changing competition among different
types. This model is therefore based on the assumptions that land use
changes within a region are driven by demand, and that their dis-
tribution patterns will always remain in dynamic balance with this
variable, the natural environment, and socioeconomic conditions. Thus,
the CLUE model addresses the competitive relationship between dif-
ferent land use types using system theory via synchronous simulations
and can be divided into a non-space requirement and spatial distribu-
tion modules. The first of these denotes the number of land use types
under different scenarios, while the latter assigns demand data in these
cases to appropriate spatial positions within the region under study via
binary logistic regression according to a total probability distribution.
This process enables simulation of the spatial distribution of land use
change.
We used a binary logistic regression method to calculate event
probabilities, including the independent variable as the predicted value.
This approach can be used to assess the quantitative relationships be-
tween the spatial distributions of each land use type as well as to de-
termine likely underlying driving forces, and is computed as follows:
Table 1
Areas of different land use types (km2) in the karst mountain areas of China from 1990 to 2010.
Year Cultivated land Forested land
1990 56,180.34 121,291.23
1995 53,810.92 122,818.78
2000 56,227.80 121,034.22
2005 56,122.66 12,1927.71
2010 55,509.58 121,888.68
Physics and Chemistry of the Earth 104 (2018) 66–75
Pi ⎞
Log ⎛⎜ ⎟ = β0 + β1 X1, i + β2 X2, i + ⋅⋅⋅⋅⋅⋅+βn Xn, i
⎝ 1 − Pi ⎠ (1)
where Pi is the probability of a certain land use type i appearing within
each grid unit (i.e., values range between 0 and 1), X denotes the un-
derlying variable, and β is the regression coefficient of each influencing
factor. However, as different underlying variables will all exert dis-
tinctive influences and effects controlling land use changes, we in-
corporated a digital elevation model (DEM), values for population and
gross domestic product (GDP), and distances from the nearest river,
national highway, provincial highway, railway, urban residential area,
and rural residential area, as well as soil and karst physiognomy types
within our analysis (Fig. 2).
3.3. Data sources
Data input to the water yield module embedded within the InVEST
model included soil depth and data on land use, rainfall, annual average
reference evapotranspiration, and vegetation-effective water content,
as well as the biophysical list, and area coverage. We obtained soil
depth data from the World Soil Database, and land use data encom-
passing the years 1990, 1995, 2000, 2005, and 2010 from the remote
sensing (RS) monitoring database for current land use in China at a
1:100,000 scale. These data are mainly based on landsat TM/ETM RS
images. The rainfall data used in this study are annual averages for the
period between 1990 and 2010 provided by the China Meteorological
Administration; these data was spatially interpolated using the Kriging
method in the software ARCGIS to obtain 1000 m grids. The biophysical
list comprises a land type name, code, root-limited depth, and evapo-
transpiration coefficient; root-limited depths were obtained from the
World Soil Database while the evapotranspiration coefficient is from
the InVEST manual. The land use change driving force data required by
the CLUE model were obtained from the Resources and Environment
Scientific Data Center, Chinese Academy of Sciences.
4. Results
4.1. Land use changes in karst mountain areas from 1990 to 2010
The dominant land use type in karst mountain areas is forest
(Table 1), accounting for ca. 56% of the total. At the same time, cul-
tivated land encompasses 25% of the total area, while grassland covers
15%, and water areas, construction land, and barren land collectively
comprise 2%. Forested land is mainly distributed spatially in the central
and southwestern region of karst mountain areas, including the cities of
Hechi and Baise, while grassland mainly covers the northwest of the
region and cultivated land mainly occurs in the northwest and south-
east (Fig. 3). Construction land is distributed sporadically throughout
the whole region (Fig. 3).
Between 1990 and 2010, the total area of cultivated land and
grassland in karst mountain areas decreased, while the coverage of
forested land, water area, construction land, and barren land increased
(Table 1 and Fig. 3). Specifically, the area of cultivated land initially
decreased before increasing again; a reduction of 670.76 km2 (1.19%)
in this land use type was seen between 1990 and 2010. In contrast, the
area of grassland decreased steadily over this time period, an overall
Grassland Water areas Construction land Barren land
32,300.57 1850.00 2484.39 23.39
32,992.32 1800.39 2676.99 30.53
32,209.32 1964.97 2670.96 22.65
31,335.96 1945.12 2776.93 21.55
30,788.52 2317.35 3597.95 27.84
69
Y. Lang et al.
Fig. 3. Land use of the karst mountain areas in 1990 (a) and conversions from 1990 to 2010 (b and c).
loss of 1512.02 km2 (4.68%), while the area of forested land slightly
increased (597.45 km2; 0.49%) and the area of water increased by
467.35 km2 (25.62%). The areas of construction and barren land both
fluctuated between 1990 and 2010; the area under construction in-
creased by 1113.56 km2 (44.82%) while that of barren land increased
by 4.45 km2 (19.02%).
Changes in land use between 1990 and 2010 in karst mountain
areas were dominated by the mutual transfers of cultivated land to
70
Physics and Chemistry of the Earth 104 (2018) 66–75
forested land, forested land to grassland, and cultivated, forested, and
grassland to construction land (Fig. 3). Over this time period, grassland
was the largest area to be transferred-out (2356.50 km2), mainly into
forested land (73.24%), cultivated land (16.44%), and construction
land (6.36%). Most of this change took place within Qiannan Autono-
mous Prefecture in the center of the karst mountain areas, as well as in
the cities of Anshun and Bijie in the northwest, in the city of Guiyang
and in Qiandongnan Autonomous Prefecture at the north, and in the
Y. Lang et al.
cities of Laibin and Guigang in the southeast (Fig. 3). The second largest
area of transferred-out land was forested, 1617.13 km2, mainly into
grassland (45.09%), cultivated land (25.38%), and construction land
(15.91%). Forested land was mainly transferred-out within the Qiannan
Autonomous Prefecture and in the Hechi region within the center of the
karst mountain areas, as well as in the southwestern city of Baise. A
total of 1515.06 km2 of cultivated land was transferred-out over this
time period, mainly into construction land (49.72%) and forested land
(31.27%) in the city of Nanning in the south, in Qiannan Autonomous
Prefecture in the center of this region, and in the city of Liupanshui in
the northwest. Transferred-out areas of water, grassland, and barren
land were all relatively small, 30.19 km2, 54.04 km2, and 3.56 km2,
respectively; this land use change mostly took place within the city of
Chongzuo in the south of the region.
The forested land encompassed the largest transferred-in area, a
total of 2214.58 km2, mainly from grassland (77.93%) and cultivated
land (21.39%). This change in land use mostly took place in Qiannan
Autonomous Prefecture in the center of the karst mountain areas, as
well as in the city of Guigang in the southeast, and in the cities of Bijie
and Anshun in the northwest (Fig. 3). The area of construction land was
the second largest to be transferred-in, 1167.60 km2, in this case mainly
from cultivated land (64.52%), forested land (22.04%), and grassland
(12.83%). This change in land use mostly took place within the city of
Nanning in the south, the city of Baise in the southwest, the city of
Guiyang in the north, and the city of Liupanshui in the northwest. Si-
milarly large areas of cultivated land and grassland were also trans-
ferred-in, 844.30 km2 of the first type, mainly from forested land
(48.61%) and grassland (45.89%) in Qiannan Autonomous Prefecture
and the city of Hechi in the center of the region, the city of Chongzuo in
the south, and the city of Guiyang in the north. Almost the same area of
grassland was transferred-in, 844.45 km2, in this case mainly from
forested land (86.35%) and cultivated land (12.22%) in Qiannan Au-
tonomous Prefecture and within the city of Hechi in the center, the city
of Anshun in the northwest, and the city of Laibin in the southeast.
Relatively small areas of water and barren land were transferred-in over
the time period considered in this study, just 497.54 km2 and 8.01 km2,
respectively, mainly within the city of Hechi in the central part of the
region.
4.2. Temporal and spatial variation of water yields in the karst mountain
area
The meteorological data used in this paper is an average of
1990–2010 which eliminated the impact of climate change on regional
water yields. Therefore, land use change became the sole factor that can
influence the water yields from 1990 to 2000 in our research. The re-
sults show that, between 1990 and 2010, water yields in the karst
mountain area tended to fluctuate but decrease overall; annual total
water yields of 212,221,542.00 mm, 212,096,643.66 mm,
212,257,650.37 mm, 212,210,433.23 mm, and 212,210,433.23 mm
were recorded in 1990, 1995, 2000, 2005, and 2010, respectively. The
total water yield of the karst mountain area decreased by
11,108.77 mm between 1990 and 2010 (Fig. 4). Annual water yields
within the karst mountain area fluctuated between 400 mm and
1700 mm and gradually increased from northwest-to-southeast (Fig. 4).
Regions with high water yields included the cities of Guigang and
Liuzhou in the southeast where annual averages exceeded 1200 mm; in
contrast, minimum annual water yields (below 800 mm) were recorded
for the city of Bijie in the northwest as well as for the cities of Guiyang
and Zunyi in the north. The average annual water yield of the rest areas
are between 800 and 1200 mm.
Construction land was characterized by the largest annual average
water yield (ca. 1300 mm) of all the types considered in this study
(Table 2), although cultivated land, grassland, and barren land also had
relatively large annual average yields, fluctuating around 1000 mm.
The annual average water yield of barren land exhibited a relatively
71
Physics and Chemistry of the Earth 104 (2018) 66–75
large amplitude of variation that fluctuated between 1000 mm and
1200 mm, while forested land and barren land both had annual average
water yields of less than 1000 mm. The annual average water yield of
forested land was slightly smaller of these two (Table 2). The annual
average water yield of construction land are the highest among the
different land use types, followed by barren land, grassland, cultivated
land, water areas, and forested land.
Due to the effects of land use change, the cities of Chongzuo and
Nanning in the south, the city of Liuzhou in the east, the city of Baise in
the southwest, Qiannan Autonomous Prefecture in the center, and the
cities of Guiyang and Zunyi in the north all recorded increasing annual
average water yields (Fig. 5). Specifically, the annual average water
yield of the city of Nanning increased by 1.82 mm, while that of the city
of Liuzhou increased by 0.79 mm, and those of the five other regions
increased by between zero and 0.5 mm. In contrast, the cities of Gui-
gang and Laibin in the southeast, as well as Bijie, Liupanshui, and
Anshun and Qianxinan Autonomous Prefecture in the northwest, and
the city of Hechi in the center all exhibited decreasing annual average
water yields, the reduction of Guigang is the largest by 2.25 mm; yields
for the cities of Tongren and Anshun also decreased by 1.69 mm and
1.54 mm, respectively, while those for the city of Bijie and Qianxinan
Autonomous Prefecture fell by 1.36 mm and 1.37 mm, respectively.
Yield values for the cities of Hechi, Liupanshui, and Laibin as well as for
Qiandongnan Autonomous Prefecture all decreased by less than
0.5 mm.
4.3. Projected land use changes of the karst mountain areas in 2030
We adopted the CLUE model to forecast the land use changes in
2030. The basic land use data that were utilized to design the scenarios
and to simulate the spatial changes in 2030 were two land use maps of
karst mountain areas in 2005 and 2010, and 11 natural and socio-
economic driving factors (Fig. 2). Three scenarios were designed in-
cluding a natural growth scenario, an economic development scenario
and an ecological protection scenario. For natural growth scenario, land
use change was allowed to vary according to the rate observed between
2005 and 2010, while the likely situation in the karst mountain area in
2030 was simulated using 2005 land use data. For economic develop-
ment scenario, development of the economy was allowed to dominate,
so areas of construction and cultivated land substantially increased,
while forested and grassland areas decreased (Table 3). For ecological
protection scenario, protection of the ecological environment dom-
inates, so areas of forested land and grassland were allowed to increase
at high rates, while those of cultivated land decreased, and construction
land slowly increased. The amount of different land use types in 2030
were designed in Table 3.
Compared with the land use situation in 2010, the area of forested
land in the center of the study region increased significantly in a si-
mulated natural growth scenario in 2030. At the same time, the areas of
construction land in the north, south, and southeast also increased to
some extent, while the areas of grassland and cultivated land decreased
(Fig. 6). This contrasts with land use changes in an economic devel-
opment scenario; in this case, areas of cultivated land and construction
land in the south, southeast, north, and northwest all increased sig-
nificantly compared with 2010, while the areas of grassland and
forested land decreased sharply. Finally, in an ecological protection
scenario, the area of forested land within the karst mountain area in-
creased significantly, while the areas of grassland and construction land
increased slightly, and the area of cultivated land decreased.
4.4. Water yield changes under projected land use changes from 2010 to
2030
Water yields of three scenarios were assessed by InVEST model.
Simulated distributions of the total water yields were similar to the
actual spatial situation, gradually increasing from northwest-to-
Y. Lang et al. Physics and Chemistry of the Earth 104 (2018) 66–75

Fig. 4. Water yield (mm) of the karst mountain areas in 1990 (a), 1995 (b), 2000 (c), 2005 (d) and 2010 (e).

Table 2 The forecasted changes of annual water yield from 2010 to 2030 are
Changes of annual average water yields from 1990 to 2010 in the karst mountain areas more drastic than that from 1990 to 2010 (Table 4). From 2010 to
(mm). 2030, in natural growth scenario, the cities of Anshun, Liupanshui and
Land use 1990 1995 2000 2005 2010
Qianxinan Autonomous Prefecture in the northwest, the cities of
Guiyang and Zunyi in the north, the city of Guigang in the southeast,
Cultivated land 1041.19 1037.85 1040.84 1041.30 1041.67 the city of Nanning in the south all recorded increasing annual average
Forested land 953.59 952.85 954.60 954.81 953.19 water yields. Specifically, the annual average water yield of the cities of
Grassland 1056.89 1063.22 1053.52 1052.91 1058.53
Water areas 1003.37 1007.83 997.26 1000.48 967.64
Guiyang, Liupanshui and Zunyi increased by 5.05 mm, 8.42 mm and
Construction land 1303.08 1299.13 1300.52 1290.51 1260.88 5.46 mm separately, and those of the four other regions increased by
Barren land 1007.77 1096.84 1007.77 1007.77 1118.61 between 0 and 3 mm. In contrast, the other regions all showed de-
Average 996.79 996.20 996.95 996.74 996.74 creasing trends in annual average water yields. The decrease of
Qiannan Autonomous Prefecture is the largest with a value of
11.41 mm. Except for Anshun, Chongzuo, Guigang, Liuzhou and Nan-
southeast (Fig. 7). Thereinto, total water yields were highest in the
ning, the changes of annual water yields from 2010 to 2030 are all more
economic development scenario (214,871,610.47 mm) in 2030, fol-
drastic than them from 1990 to 2010 (Table 4).
lowed by natural growth (211,625,912.02 mm) of the three simulations
In the economic development scenarios, all cities of karst mountain
considered in this study. At the same time, water yields in the ecological
areas recorded increasing annual average water yields. The increment
protection scenario in 2030 were lowest (211,231,280.47 mm). Because
of Liupanshui is the largest one with a value of 32.11 mm (Table 4). In
of significant differences in land use, total water yields of 2030 in
addition, the variations of annual water yield of all cities are more
natural growth and ecological protection scenarios were both lower
drastic than that in the period of 1990–2010.
than the actual total water yields in 2010, decreasing by
In the ecological protection scenario, the cities of Guiyang and
584,521.21 mm (0.28%) and 979,152.75 mm (0.46%), respectively. In
Zunyi in the north, the cities of Liupanshui, Qianxinan Autonomous
contrast, total water yields of 2030 in the economic development sce-
Prefecture in the northwest, and the cities of Guigang in the southeast
nario revealed an increase of 26,611,77.25 mm (1.25%) compared to
all recorded increasing annual average water yields. Thereinto, the
actual total water yields, which indicates that this scenario is more
increment of Liupanshui is the largest by 10.15 mm. The annual
conducive to increasing regional water yields.
average water yields of the rest regions all decreased. Specifically, the

72
Y. Lang et al.
decrease of Laibing is the largest by 14.20 mm. Except for Anshun,
Guigang, Nanning and Qianxinan Autonomous Prefecture, the changes
of annual water yields are all more drastic than them in the period of
1990–2010 (Table 4).
Annual average water yields of the various land use types con-
sidered in this study combined with area variations in land use under
different scenarios revealed that in the natural growth scenario, in-
creases in forested land area occur at the expense of cultivated and
grassland and lead to lower total water yields. Similarly, increases in
forested, construction land, and grassland areas occurred at the cost of
construction land in the ecological protection scenario; in this case,
forested land increments were far higher than those of either con-
struction land or grassland and resulted in the highest total water
yields. In the economic development scenario, increases in construction
land and cultivated land occurred at the expense of forested and
grassland areas and again resulted in increased total water yield rates.
5. Discussion and conclusions
This study assessed the influence of land use changes on water
yields in the karst mountain area. Results show that the average water
yield between 1990 and 2010 was 991.12 mm, higher than values re-
ported in similar studies on the Sancha River Basin, Guizhou Province
(Lang et al., 2017), the upper reaches of the Heihe River Basin (Geng
et al., 2014), and the Three Rivers source area (Pan et al., 2013). At the
same time, however, results also reveal complex temporal and spatial
variation in regional water yields within the karst mountain area. One
key reason for this is the fact that this region encompasses a huge area,
spanning most of Guizhou and Guangxi provinces, and is characterized
by large variations in both natural and socioeconomic conditions.
Variations in the influence of natural conditions on water yields is
mainly the result of regional rainfall; the karst region in Guizhou Pro-
vince has a higher elevation and is located further inland than its
counterpart in Guangxi Province that occurs at a lower elevation, closer
to the coast. This difference translates to large variations in rainfall
Table 3
Proportions of designed land use types in three scenarios (%).
Cultivated land Forested land
Natural growth 25.29 57.48
Economic development 30.00 53.00
Ecological protection 20.00 60.00
73
Physics and Chemistry of the Earth 104 (2018) 66–75
Fig. 5. Changes of annual average water yield (mm) from
1990 to 2010 in the karst mountain areas.
which directly affects water yield. In contrast, the influence of different
socioeconomic conditions on water yields is mainly due to differences
in land use; natural geographic conditions have inevitably enabled a
lower degree of economic development in Guizhou Province compared
to Guangxi Province, resulting in different changes in land use. There is
a higher proportion of construction and cultivated land in more de-
veloped regions, while less developed areas tend to have higher pro-
portions of forested land and grassland. These differences in land use
directly influence evapotranspiration rates and root depth, and thus
regional water yields.
Uncontrollable climate change is the most obvious and marked of
the two factors influencing water yields. Since the early years of the
20th century, natural disasters have occurred frequently within this
region of China, and the number of extreme weather events continues
to increase, significantly impacting both the ecological environment
and social stability. Climate change regulation is complex and requires
long time spans to have a noticeable impact. Regional land planning, in
contrast, is much easier to control and evaluate; thus, via planning,
policies, and other approaches, governments can effectively regulate
variation in regional economic development and land use. This is cor-
roborated by the results of our study; in the economic development
simulated scenario, water yields in the karst mountain area rose to a
certain extent. This suggests that regional economic development and
ecological services can be simultaneously taken into account. In actual
planning and management processes, human activities and economic
development should be guided reasonably and appropriately, however,
and compulsory mechanisms should be adopted to control regional
development trends. This approach means that both the regional eco-
logical environment can be protected, while economic development is
also ensured, promoting regional sustainable development.
The water yield module embedded within the InVEST model was
used in this study to evaluate changes in this parameter between 1990
and 2010 as well as to investigate land use and yield changes in the
karst mountain area of China. The results of this study show that
forested land, construction land, and water areas in the karst mountain
Grassland Water areas Construction land Barren land
12.97 1.52 2.72 0.02
10.98 1.00 5.00 0.02
15.50 1.00 3.48 0.02
Y. Lang et al. Physics and Chemistry of the Earth 104 (2018) 66–75

Fig. 6. Forecasted land use changes under scenarios of natural growth (a), economic development (b) and ecological protection (c) in the karst mountain areas of China.

Fig. 7. Water yields under the scenarios of natural growth (a), economic development (b) and ecological protection (c) in the karst mountain areas.

Table 4 have tended to increase, while those of cultivated land and grassland
Changes of annual water yields in karst mountain areas in periods of 1990–2010 and have tended to decrease. Considering changes in land use, annual water
2010–2030 (mm). yields in the karst mountain area varied between 400 mm and 1700 mm
1990–2010 Scenarios during 2010–2030
and gradually increased from northwest-to-southeast. Water yields in
the south of this region were higher over the time period of this study
natural economic ecological than in the north. Thus, between 1990 and 2010, total water yields in
growth development protection the karst mountain area have tended to decrease as a result of land use
Anshun −1.54 0.15 11.01 −0.91
changes; overall, these yields have fluctuated but decreased by a rela-
Baise 0.27 −7.36 3.76 −9.76 tively small amount, 11,108.77 mm. The water yields of different land
Bijie −1.36 −4.75 5.42 −6.97 use types are obviously different, roughly conforming to the decreasing
Chongzuo 0.51 −0.12 25.24 −1.27 sequence of construction land, barren land, grassland, cultivated land,
Guigang −2.25 1.67 17.12 1.25
water areas, and forested land. In the context of land use change, water
Guiyang 0.07 5.05 24.98 6.28
Hechi −0.05 −4.75 7.09 −6.80 yields in the karst mountain area varied slightly, while the processes
Laibing −0.13 −7.20 11.18 −14.20 underlying these changes were relatively complex.
Liuzhou 0.80 −0.00 18.11 −4.21 Defining different land use scenarios and assessing water yields in
Liupanshui −0.24 8.42 32.11 10.15 this context, we assessed variation in water yields given natural growth,
Nanning 1.82 0.16 19.12 −1.12
Qiandongnan −0.07 −0.12 12.43 −0.94
economic development, and ecological protection from 2010 to 2030.
Qiannan 0.48 −11.41 4.76 −14.32 Results show that water yields in 2030 under natural growth and eco-
Qianxinan −1.37 2.11 14.29 1.33 logical protection scenarios are both lower than the actual situation (in
Tongren −1.69 −3.70 4.36 −6.90 2010), while yields in an economic development scenario reached
Zunyi 0.16 5.46 23.50 5.71
214,871,610.47 mm, higher than reality. In the three scenarios, the
variations in annual water yields during the period of 2010–2030 are

74
Y. Lang et al.
more drastic than them in the period of 1990–2010. We therefore
conclude that increases in construction and cultivated land areas in an
economic development scenario would be most beneficial to improve
regional water yields.
Acknowledgements
This research was supported by the National Key Research Program
of China (Grant number 2015CB452702) and the National Natural
Science Foundation of China (Grant numbers 41671177 and
41501192).
References
Benda, L., Miller, D., Barquin, J., McCleary, R., Cai, T.J., Ji, Y., 2016. Building virtual
watersheds: a global opportunity to strengthen resource management and con-
servation. Environ. Manag. 57, 722–739.
Bennett, E.M., Peterson, G.D., Gordon, L.J., 2009. Understanding relationships among
multiple ecosystem services. Ecol. Lett. 12, 1394–1404.
Bossa, A.Y., Diekkrüger, B., Agbossou, E.K., 2014. Scenario-based impacts of land use and
climate change on land and water degradation from the meso to regional scale. Water
6, 3152–3181.
Budyko, M.I., 1975. Ecological factors of evolution. Zhurnal Obshchei Biol. 36, 36–47.
Cademus, R., Escobedo, F.J., McLaughlin, D., Abd-Elrahman, A., 2014. Analyzing trade-
offs, synergies, and drivers among timber production, carbon sequestration, and
water yield in pinus elliotii forests in southeastern USA. Forests 5, 1409–1431.
Cheng, H., Hu, Y.A., Zhao, J.F., 2017. Meeting China's water shortage crisis: current
practices and challenges. Environ. Sci. Technol. 43, 240–244.
Costanza, R., 1997. The value of the World's ecosystem services and natural capital.
World Environ. 387, 253–260.
Daily, G.C., 1997. Nature's services: societal dependence on natural ecosystems. Natures
Serv. Soc. Dependence Nat. Ecosyst. 1, 220–221.
Deng, X., Li, Z., Huang, J., Shi, Q., Li, Y., 2013. A revisit to the impacts of land use
changes on the human wellbeing via altering the ecosystem provisioning services.
Adv. Meteorol. 2013, 183–186.
Deng, X., Zhao, C., 2015. Identification of water scarcity and providing solutions for
adapting to climate changes in the Heihe River basin of China. Adv. Meteorol. 2015,
1–13.
Fricke, K., 2014. Analysis and Modelling of Water Supply and Demand under Climate
Change, Land Use Transformation and Socio-economic Development. Springer
International Publishing.
Gang, K., Lim, C.H., Kim, S., Lee, J., Son, Y., Lee, W.K., 2017. Effect of national-scale
afforestation on forest water supply and soil loss in South Korea, 1971–2010.
Sustainability 9, 1017.
Geng, X., Wang, X., Yan, H., Zhang, Q., Jin, G., 2014. Land use/land cover change in-
duced impacts on water supply service in the upper reach of Heihe River Basin.
Sustainability 7, 366–383.
Hamel, P., Chaplin-Kramer, R., Sim, S., Mueller, C., 2015. A new approach to modeling
the sediment retention service (InVEST 3.0): case study of the Cape Fear catchment,
North Carolina. USA. Sci. Total Environ. 524–525, 166–177.
Han, Z., Song, W., Deng, X., 2016. Responses of ecosystem service to land use change in
Qinghai province. Energies 9, 303–318.
Hosseini, M., Ghafouri, A.M., Amin, M.S.M., Tabatabaei, M.R., Goodarzi, M., Kolahchi,
A.A., 2012. Effects of land use changes on water balance in Taleghan catchment, Iran.
J. Agric. Sci. Technol. 14, 1159–1172.
Jie, X., Yu, X., Na, L., Hao, W., 2015. Spatial and temporal patterns of supply and demand
balance of water supply services in the dongjiang lake basin and its beneficiary areas.
J. Resour. Ecol. 6, 386–396.
Keeler, B.L., Polasky, S., 2014. Land-use change and costs to rural households: a case
study in groundwater nitrate contamination. Environ. Res. Lett. 9, 74002–74011.
Lang, Y., Song, W., Zhang, Y., 2017. Responses of the water-yield ecosystem service to
climate and land use change in Sancha River Basin, China. Phys. Chem. Earth Parts
A/b/c (in press). https://doi.org/10.1016/j.pce.2017.06.003.
Leh, M.D.K., Matlock, M.D., Cummings, E.C., Nalley, L.L., 2013. Quantifying and map-
ping multiple ecosystem services change in West Africa. Agric. Ecosyst. Environ. 165,
6–18.
Li, Z., Deng, X., Huang, J., Zhang, R., Huang, J., 2013. Critical studies on integrating land-
use induced effects on climate regulation services into impact assessment for human
well-being. Adv. Meteorol. 2013, 78–89.
Liquete, C., Maes, J., Notte, A.L., Bidoglio, G., 2011. Securing water as a resource for
society: an ecosystem services perspective. Ecohydrol. Hydrobiol. 11, 247–259.
Little, C., Lara, A., 2010. Ecological restoration for water yield increase as an ecosystem
75
Physics and Chemistry of the Earth 104 (2018) 66–75
service in forested watersheds of south-central Chile. Bosque 3, 175–178.
Liu, Y., Song, W., Deng, X., 2017a. Spatiotemporal patterns of crop irrigation water re-
quirements in the Heihe River basin, China. Water 9, 616. http://dx.doi.org/10.
3390/w9080616.
Liu, Y., Song, W., Mu, F., 2017b. Changes in ecosystem services associated with planting
structures of cropland: a case study in Minle County in China. Phys. Chem. Earth
Parts A/b/c (in press). https://doi.org/10.1016/j.pce.2016.09.003.
Lópezmoreno, J.I., Vicenteserrano, S.M., Morantejeda, E., Zabalza, J., Lorenzolacruz, J.,
Garcíaruiz, J.M., 2010. Impact of climate evolution and land use changes on water
yield in the Ebro basin. Hydrol. Earth Syst. Sci. Discuss. 15, 311–322.
Lu, N., Sun, G., Feng, X., Fu, B., 2013. Water yield responses to climate change and
variability across the North–South Transect of Eastern China (NSTEC). J. Hydrol.
481, 96–105.
Mark, A.F., Dickinson, K.J., 2008. Maximizing water yield with indigenous non-forest
vegetation: a New Zealand perspective. Front. Ecol. Environ. 6, 25–34.
Marquès, M., Bangash, R.F., Kumar, V., Schuhmacher, M., Sharp, R., 2013. The impact of
climate change on water provision under a low flow regime: a case study of the
ecosystems services in the Francoli river basin. J. Hazard. Mater. 263, 224–232.
Ming, C., Qin, X., Zeng, G., Jian, L., 2016. Impacts of human activity modes and climate
on heavy metal “spread” in groundwater are biased. Chemosphere 152, 439–445.
Pan, T., Wu, S., Liu, Y., 2015. Relative contributions of land use and climate change to
water supply variations over Yellow River source area in tibetan plateau during the
past three decades. PLos One 10, e0123793.
Pan, T., Wu, S.H., Dai, E.F., Liu, Y.J., 2013. Spatiotemporal variation of water source
supply service in Three Rivers Source Area of China based on InVEST model. Chin. J.
Appl. Ecol. 24, 183–189.
Pervez, M.S., Henebry, G.M., 2015. Assessing the impacts of climate and land use and
land cover change on the freshwater availability in the Brahmaputra River basin. J.
Hydrol. Regional Stud. 3, 285–311.
Principe, J.A., 2012. Exploring climate change effects on watershed sediment yield and
land cover-based mitigation measures using swat model, RS and Gis: case of cagayan
River Basin, Philippines. ISPRS - Int. Archives Photogram. Remote Sens. Spatial Inf.
Sci. XXXIX-B8, 193–198.
Ranganathan, J., 2011. Natural Capital: Theory and Practice of Mapping Ecosystem
Services. Oxford University Press.
Smith, E.J., 1997. The balance between public water supply and environmental needs.
Water & Environ. J. 11, 8–13.
Song, W., Deng, X., 2015. Effects of urbanization-induced cultivated land loss on eco-
system services in the north China plain. Energies 8, 5678–5693.
Song, W., Deng, X., 2017. Land-use/land-cover change and ecosystem service provision in
China. Sci. Total Environ. 576, 705–719.
Song, W., Deng, X., Yuan, Y., Wang, Z., Li, Z., 2015. Impacts of land-use change on valued
ecosystem service in rapidly urbanized North China Plain. Ecol. Model. 318,
245–253.
Sun, S., Sun, G., Cohen, E., McNulty, S.G., Caldwell, P.V., Duan, K., Zhang, Y., 2016.
Projecting water yield and ecosystem productivity across the United States by linking
an ecohydrological model to WRF dynamically downscaled climate data. Hydrol.
Earth Syst. Sci. 20, 935–952.
Tao, J., Xiaoyu, Q., Huang, L., 2016. Changes in grain production and the optimal spatial
allocation of water resources in China. J. Resour. Ecol. 7, 28–35.
Tong, X., Wang, K., Brandt, M., Yue, Y., Liao, C., Fensholt, R., 2016. Assessing future
vegetation trends and restoration prospects in the karst regions of southwest China.
Remote Sens. 8 357–357.
Verburg, P.H., Soepboer, W., Veldkamp, A., Limpiada, R., Espaldon, V., Mastura, S.S.,
2002. Modeling the spatial dynamics of regional land use: the CLUE-S model.
Environ. Manag. 30, 391–405.
Wang, C.H., Blackmore, J.M., 2009. Resilience concepts for water resource systems. J.
Water Resour. Plan. Manag. 135, 528–536.
Wu, Y., Yang, Y., Feng, Z., Sun, T., 2017. Exploitation, utilization status, and development
prospects of water resources in ningbo city. J. Resour. Ecol. 8, 205–211.
Yang, X., Zhou, Z., Jing, L.I., Xin, F.U., Xingmin, M.U., Ting, L.I., 2016. Trade-offs be-
tween carbon sequestration, soil retention and water yield in the Guanzhong-
Tianshui Economic Region of China. J. Geogr. Sci. 26, 1449–1462.
Yi, L., Wei, X., Zhang, W., Chao, W., Wang, P., 2016. Life cycle assessment of water supply
alternatives in water-receiving areas of the South-to-North Water Diversion Project in
China. Water Res. 89, 9–19.
Yu, J., Yuan, Y., Nie, Y., Ma, E., Li, H., Geng, X., 2015. The temporal and spatial evolution
of water yield in Dali county. Sustainability 7, 6069–6085.
Zhang, C., Wenhua, L.I., 2012. Water yield of Xitiaoxi River basin based on InVEST
modeling. J. Resour. Ecol. 3, 50–54.
Zhang, Y., Min, C., Zhou, W., Zhuang, C., Ouyang, Z., 2010. Evaluating Beijing's human
carrying capacity from the perspective of water resource constraints. J. Environ. Sci.
22, 1297–1304.
Zhao, Z., Wu, X., Zhang, Y., Gao, J., 2017. Assessment of changes in the value of eco-
system services in the Koshi River Basin, central high himalayas based on land cover
changes and the ca-markov model. J. Resour. Ecol. 8, 67–76.