Ecological Indicators 117 (2020) 106677

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Ecological risk assessment of the wetlands in Beijing-Tianjin-Hebei urban T

agglomeration
⁎
Zhuo Lia, Weiguo Jianga,b, , Wenjie Wangc, Zheng Chenc, Ziyan Lingb, Jinxia Lva
a
Beijing Key Laboratory for Remote Sensing of Environment and Digital Cities, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
b
Key Laboratory of Beibu Gulf Environment Change and Resources Use of Ministry of Education, Nanning Normal University, Nanning 530001, China
c
Chinese Research Academy of Environmental Sciences, Beijing 100012, China

A R T I C LE I N FO A B S T R A C T

Keywords: Wetlands are some of the most highly productive ecosystems on earth and are one of the most important en-
Wetland vironments for human survival and development. However, with the acceleration of industrialization and ur-
Hazard degree banization, wetlands have been constantly occupied and destroyed and have even disappeared in some areas, so
Vulnerability it is necessary to establish a reasonable and scientific ecological risk assessment system for wetlands. In this
Ecological risk assessment
study, 181 counties in Beijing-Tianjin-Hebei urban agglomeration (Jing-Jin-Ji) were selected as the research
Beijing-Tianjin-Hebei urban agglomeration
objects, and an index system with 22 indexes was established from two aspects: the external hazard degree and
the vulnerability of the wetland itself. The ecological risk of the wetlands during 6 periods from 1990 to 2015
was comprehensively evaluated by long-term data and multiple index factors. The results showed that (1) due to
the decrease in the hazard degree and the increase in the vulnerability, the ecological risk of the wetlands in
Jing-Jin-Ji first increased and then decreased; (2) more than 60% of the counties were at the middle or middle-
high risk level, especially in the center of urban development, which generally had higher risk level than other
regions; (3) most wetland parks were at the middle-high risk level, while wetland reserves, reservoirs and lakes
were mostly in the low risk or middle-low risk level. Therefore, in recent years, the implementation of wetland
protections and restoration policies and projects has played a significant role in reducing the ecological risk of
the wetlands. In the future, more targeted management and protection measures need to be formulated and
followed up in time.

1. Introduction 2014; Mao et al., 2018a). In recent decades, as people have gradually
realized the importance of wetlands, an increasing number of re-
Wetlands are important ecosystems with a wide range of distribu- searchers have conducted research on wetlands, including on wetland
tion (Fu et al., 2018). The ecosystem type has important functions and degradation, the evolution and driving factors, the current situation and
value in gas regulation, water supply, biodiversity maintenance, and threats, data extraction and mapping, ecological risk assessment and
human health (Hu et al., 2017; Meng et al., 2017). Wetlands are ex- other aspects (Bai et al., 2013; Cui et al., 2018; Jiang et al., 2017; Mao
tremely rich in animal and plant species and provide many basic re- et al., 2018b; Serran and Creed, 2015; Xu et al., 2019).
sources for the development of human society. In the past few centuries, An ecological risk assessment evaluates the possibility of ecosystem
however, human beings have neglected the existence of wetlands. For damage (or impacts) caused by some ecological environmental condi-
the rapid development of the social economy, wetlands have been tions (such as chemical pollution or potential hazards caused by human
exploited and utilized recklessly, and wetlands have also become activities), accidents or disasters. This concept and research started in
emission sites for the pollutants produced by various industries (Xu the 1970s and was developed on the basis of risk assessment for human
et al., 2019). Therefore, many wetlands have suffered serious damage health. Then, the methods and techniques of risk assessment gradually
from disturbance by human activities and the vulnerabilities of the became more common and tended to be standardized. In 1992, the
wetlands themselves. Many studies have shown that half of the wet- United States Environmental Protection Agency (USEPA) proposed a
lands in the world have disappeared in the last century (Davidson, definition for the ecological risk assessment. Then, in 1998, the

⁎
Corresponding author at: Beijing Key Laboratory for Remote Sensing of Environment and Digital Cities, Faculty of Geographical Science, Beijing Normal
University, Beijing 100875, China.
E-mail address: [email protected] (W. Jiang).

https://doi.org/10.1016/j.ecolind.2020.106677
Received 3 February 2020; Received in revised form 24 June 2020; Accepted 27 June 2020
1470-160X/ © 2020 Elsevier Ltd. All rights reserved.
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Ecological Risk Assessment Guidelines issued by the United States Cangzhou coastal wetlands (Chen et al., 2019; Ji et al., 2019; Mao et al.,
proposed “three steps” for ecological risk assessment, including pro- 2010; Tian et al., 2010; Guo et al., 2011; Guo and Zhang, 2019). In
blem formulation, analysis phase (exposure characterization and eco- addition, studies on ecological risk assessment are mostly concentrated
logical effect characterization) and risk characterization (Norton et al., on multiple land use (Kang et al., 2018). Therefore, this paper evaluated
1992; Ni et al., 2019; Xu et al., 2016; Chen et al., 2013). At present, the ecological risk of the entire Jing-Jin-Ji wetland system from hazard
ecological risk assessment has been widely used for many aspects and degree and vulnerability and analyzed the temporal and spatial changes
fields, such as ecological risk assessment of stormwater ponds, vulner- in the risk from 1990 to 2015.
ability assessment of estuarine ecosystems, and the risk assessment of Based on the theory of ecological risk assessment, this study con-
the pollutants in coastal areas (Tixier et al., 2011; Ribeiro et al., 2016; structed the conceptual framework of the wetland ecological risk as-
Shi et al., 2018). At the same time, based on the ecological risk as- sessment from hazard degree and vulnerability and carried out a risk
sessment concept, evaluation methods are increasingly diverse and assessment for the Jing-Jin-Ji wetlands from 1990 to 2015. Therefore,
accurate, such as the Weight of Evidence (WOE) and relative risk model the specific objectives were (1) to fully analyze the changes in hazard
(RRM). These methods are mostly used to evaluate the situation of degree among the sources of risk and the vulnerability characteristics of
multiple risk sources of risk in a region (Burton et al., 2002; Walker the wetlands through the established indicator framework; (2) to cal-
et al., 2001). In addition, an increasing number of research methods are culate the comprehensive risk value of the wetland ecology according
combined with computer and artificial intelligence algorithms, such as to the indicator system and quantitatively evaluate the risk value; and
artificial neural network models (ANNs), which can handle more (3) to analyze and discuss the changes in the ecological risk of the Jing-
complex nonlinear relationships, and the results are more accurate Jin-Ji wetlands with long-term data and multiple index elements.
(Jiang et al., 2013). The evaluation method used in this study was also
adapted from ecological risk assessment, which consider that risks
2. Methodology
should be composed of the probabilities (hazard degree) of accidents
(or danger) and the damage (risk consequence) caused by the accidents
2.1. Study area
(or danger). That is, the risk was equal to the product of the hazard
degree and the vulnerability (Jiang et al., 2017).
Jing-Jin-Ji is one of the three largest urban agglomerations in China
Wetland ecological risk assessment is based on the development of
(Fig. 1). Jing-Jin-Ji includes Beijing (the capital city of China), Tianjin
ecological risk assessment. This kind of evaluation is more focused on
(one of municipalities in China)and Shijiazhuang, Baoding, Cangzhou,
the study of the possible hazards from the main sources of risk faced by
Chengde, Handan, Hengshui, Langfang, Qinhuangdao, Tangshan,
wetlands, such as natural disasters, human factors, etc., and considered
Xingtai, Zhangjiakou in Hebei Province; the overall area of Jing-Jin-Ji is
a variety of risk sources. For example, the risks faced by wetlands have
21.7 million ha (Cui et al., 2019; Sun and Zhao, 2019). There are 181
been evaluated in terms of the industrial pollution and the natural and
districts and counties in the study area (Fig. S1, Fig. S2). With the rapid
human factors (Liang et al., 2019; Sarkar et al., 2015). In addition,
urbanization and industrialization process, Jing-Jin-Ji has become one
researchers also evaluate the ecological security and ecological health
of the pillars of growth in the economic development of China. How-
of wetlands based on the concept of ecological risk assessment (Peng
ever, this phenomenon has also brought tremendous pressure to the
et al., 2017; Wanda et al., 2015). For example, the mangroves along the
ecosystems, and environmental problems have become increasingly
coast are often vulnerable to pollution, and researchers have conducted
serious, including the degradation of wetlands and increased water
a comprehensive risk analysis of the heavy metal pollution in the region
pollution (Han et al., 2018; Chen et al., 2018).
(Shi et al., 2019). Some scholars have integrated chemical, physical and
biological stress factors to evaluate the ecological risk of wetlands
(Malekmohammadi and Rahimi Blouchi, 2014). In addition, some 2.2. Data sources
scholars have proposed a five-step method for regional ecological risk
assessment for analyzing and studying wetland risk (Xu et al., 2003). The main data used in this study include multiyear land use data-
The methods used in the above literatures include several categories. sets, meteorological data and socioeconomic data. Among them, land
One was to use a single indicator to assess ecological risk, such as use maps of Jing-Jin-Ji were selected for six periods, including 1990,
wetland area, risk quotient, geo-accumulation index. This kind of 1995, 2000, 2005, 2010 and 2015. These data were interpreted based
method was relatively simple, but the evaluation result was one-sided on the Landsat 8 OLI and GF2 remote sensing satellite data, and the
(Peng et al., 2017; Liang et al., 2019; Shi et al., 2019). The second was classification results were verified by unmanned aerial vehicle images
the use of models, such as fuzzy risk assessment model. The result was and field investigation. The classification accuracy of the second-level
more scientific, but the process was more complex (Sarkar et al., 2015). land uses was more than 90% (Liu et al., 2014). These data were ob-
The third was multi-index comprehensive evaluation methods tained from the Institute of Geographic Sciences and Natural Resources
(Malekmohammadi and Rahimi Blouchi, 2014; Xu et al., 2003). The Research, China (http://www.resdc.cn/) and resampled to 100 m. The
research method of this paper was the third one. Natural and human first-level classification of the data includes cropland, woodland,
hazard were the external pressures and risks that wetlands might be grassland, water body, built-up land and unused land (Table S1). Paddy
exposed to, and the internal vulnerability was assessed from many as- field of cropland, river, lake, reservoir and pond, tidal flat, beach of
pects, such as area, structure and function. This method was compre- water body, and marsh of unused land were included in the wetland of
hensive, scientific and easy to implement. this study (Table S2).
The concept of Beijing-Tianjin-Hebei urban agglomeration (Jing- The meteorological data were obtained from the National
Jin-Ji) was developed from the capital economic circle. With the con- Meteorological Information Center. The monthly average precipitation
tinuous development of the society, economy and culture of the area, it and temperature were collected from 33 meteorological stations in
has become a world-class urban agglomeration with some influence on Jing-Jin-Ji. And the Kriging Method was used to obtain the precipita-
East Asia and the world. In recent years, an increasing number of tion and temperature data of the whole study area from 1990 to 2015.
scholars have studied many aspects of Jing-Jin-Ji in many fields. The socioeconomic data from 1990 to 2015 were collected from
Research on the wetland ecological risk assessment has also become a Beijing Statistical Yearbook, Tianjin Statistical Yearbook, Hebei
widely discussed topic. Many scholars have carried out research on Statistical Yearbook, China Rural Statistical Yearbook, etc. Data were
ecological health, risk assessment, wetland integrity, ecological model collected that included population, Gross Domestic Product (GDP),
establishment, restoration and the effectiveness of protections for the grain prices, per unit yield of the grain crops and the area of the grain
typical wetlands in Jing-Jin-Ji, such as the Baiyangdian, Qilihai and crops.

2
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 1. The location of Jing-Jin-Ji and the distribution of the wetlands from 1990 to 2015.

2.3. Method for the wetland ecological risk assessment 2015 (Fig. 2). The idea of this conceptual model stemmed from the
“three steps” of ecological risk assessment proposed by the USEPA. The
2.3.1. Overall framework of the wetland ecological risk assessment analysis step of the model included exposure characterization and
Ecological risk assessment is used to predict and evaluate the pos- ecological effects characterization (Norton et al., 1992). Therefore, in
sibility and extent of ecosystem damage through methods such as order to achieve the target of wetland ecological risk assessment, there
mathematics and probability theory. Therefore, ecological risk assess- were two aspects to consider: the external risk sources faced by wet-
ment is a complex and comprehensive process that includes the char- lands and the effects on the interior of the wetlands after it has been
acteristics of multiple sources of risk and multiple types of impacts (Luo exposed to stressors. That is, the external hazard degree and internal
et al., 2018; Ni et al., 2019). vulnerability. Secondly, the establishment of the element layer referred
This research chose to use the conceptual model of the ecological to the conceptual model of wetland degradation risk assessment (Jiang
risk assessment to assess the ecological risk of the wetlands in Jing-Jin- et al., 2017). The risk sources were considered from the perspective of
Ji from a regional perspective on 181 counties in 6 periods from 1990 to the human and natural factors, while the vulnerability of the wetlands

Fig. 2. The overall framework of wetland ecological risk assessment.

3
Z. Li, et al. Ecological Indicators 117 (2020) 106677

themselves was determined from their area, structure and function. The the wetlands. The value profoundly affected the well-being and inter-
improvement of this paper was the selection of indexes. According to ests of human beings. And the value of wetlands was closely related to
the actual situation of the study area, a total of 22 indexes from 5 the growth and area of wetlands in that year. The ecosystem services
elements were selected for detailed and perfect analysis, and the ori- value could make people more clear and easier to understand how
ginal parameters were further optimized. Therefore, it was feasible and many functions the wetlands could provide, so it was reasonable and
stable to apply the model to Jing-Jin-Ji. easier to evaluate the function of wetland from the perspective of value
(Luo et al., 2019). According to the equivalent weighting factors of the
2.3.2. Selection of the indexes and analysis method ecosystem services value put forward by Xie et al. (Table 1), the eco-
Wetlands are greatly affected by natural factors, especially pre- system services value was converted by combining the current grain
cipitation and temperature. Lower precipitation and higher tempera- areas, grain yields and grain prices (Xie et al., 2015; Costanza et al.,
ture lead to the degradation and damage of wetlands (Muro et al., 1997). This approach considered natural and human factors to have had
2018). Therefore, precipitation and temperature were selected as in- similar effects on the agro-ecosystems and other ecosystems in the same
dexes of the natural hazard. However, due to the acceleration of ur- region. Therefore, the value of the other ecosystems could be derived
banization, human activities have increasingly impacted wetlands. In from the economic value of the food, and it was converted into 1/7 of
particular, the expansion of built-up land and the encroachment of the value of the food in the same year. It was calculated by Eq. (4) (Qian
cropland have had impacts. At the same time, the rapid growth of po- et al., 2018; Liu et al., 2019):
pulation density and GDP also have had indirect impacts on the changes 1 n ai ∙pi ∙qi
in wetlands. V=
7
∑i=1 A
(i = 1, ⋯, n)
(4)
Vulnerability was evaluated from three aspects: the area, structure
where V represents the economic value of food production produced by
and function of the wetlands. The area was characterized by the per-
cropland (yuan/hm2), which is 1/7 of the output value per unit area of
centage of the wetland area in comparison to the total area. The
main grain in the study area in that year. i represents the crop species,
structural aspect of wetland vulnerability included four landscape in-
including rice, wheat, corn and soybeans; ai, pi and qi represent the area
dexes, such as patch number, patch density, fragmentation degree and
(hm2), the national average market price (yuan/t) and yield of grain
aggregation degree. Eleven functional indexes were selected from the
crop i (t/hm2), respectively; and A represents the total area of n kinds of
four aspects: regulating services, provisioning services, supporting ser-
grain crops (hm2).
vices and cultural services.
Because of the difference of land use classification, the equivalent
Finally, the result of the risk value was equal to the product of the
weight of wetlands in the table was assigned to tidal flats, beaches and
values of hazard degree and vulnerability, it was calculated by Eq. (1)
marshes, and the equivalent weight of water bodies was assigned to
(Jiang et al., 2017):
rivers, lakes, reservoirs and ponds. So, the ecosystem services value was
Risk(R) = Hazard(H) × Vulnerability(V) (1) calculated by Eqs. (5) and (6):
m
where hazard degree of the wetland risk was a function of the degree of ESVi = ∑j =1 Ai∙Vi
natural hazard and the degree of human hazard and their respective (5)
weights. Wetland vulnerability was manifested by changes in the wet- n

land internal characteristics. Similarly, it was a function of the wetland

ESVj = ∑i=1 Aj∙Vj (6)
area, structure and function and their respective weights. Hazard de- where ESVi represents the ecosystem services value of wetland type I
gree and vulnerability were calculated by Eqs. (2) and (3), respectively. (yuan); m = 11; ESVj represents the ecosystem services value of func-
H = xn∙wn + xh∙wh (2) tion j (yuan); n = 7; Ai and Aj denote the area of wetland type i and the
total wetland area that provided ecosystem service j, respectively
V = ya ∙wa + ys∙ws + yf ∙wf (3) (hm2); and Vi and Vj denote the unit area of the ecosystem services
value of type i and ecosystem service function j, respectively (yuan/
where H denotes the hazard degree; V denotes the degree of vulner-
hm2).
ability; xn and xh denote the degree of natural hazard and human ha-
zard, respectively; ya, ys and yf denote the changes in the area, structure
2.3.4. Data processing and analysis
and function, respectively; and w indicates the weight.
When the selection and calculation of the indexes was completed,
Wetland ecological risk assessment is a comprehensive assessment
the indexes needed to be normalized. By combining the normalized
of multiple indexes and multiple factors, so it is particularly important
values of each index with its weight, the hazard value, vulnerability
to reasonably determine the weight of each index. The analysis hier-
value and wetland risk value were obtained through a series of calcu-
archy process (AHP) was put forward by Thomas Saaty in 1977 (Saaty,
lations. The 22 indexes selected in this paper were divided into positive
1980). This method constructs the system from the main goal, sub-goal
indexes and reverse indexes, so the normalization methods were as
and alternative plan. Because it is simple and easy to calculate, the
follows:
process has strong applicability and flexibility and has been applied in
various fields (Kokangül et al., 2017). The proportion of each index vij − min (vj)
1≤i≤m
weight is shown in Fig. 3. pij =
max (vj) − min (vj)
1≤i≤m 1≤i≤m (7)
2.3.3. Calculation of the hazard indexes and vulnerability indexes
max (vj) − vij
The temperature and precipitation were obtained by calculating 1≤i≤m
nij =
their annual averages. By calculating the percentage of built-up land max (vj) − min (vj)
1≤i≤m 1≤i≤m (8)
and cropland area in the total area, the impact of the urban expansion
and cropland encroachment on the wetlands was determined. The po- where pij and nij represent the normalized results of the positive indexes
pulation density and GDP data were obtained by consulting the statis- and reverse indexes, respectively; vij represents the original value of j
tical yearbooks. The four indexes selected in the structure were calcu- index in county i; and max(vj) and min(vj) represent the maximum and
lated by Fragstats software. minimum values vj in the M counties, M = 181.
It should be noted that this study used value to calculate the func- In addition, in the analysis of the results, the linear least square
tions. The amount of value was intuitively quantified from the eco- regression analysis was used for each index. The calculated slope value
nomic perspective of the amount of function and benefit provided by was considered as the direct measurement value of the index trend (Ma

4
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 3. The proportion of each index weight (Among them, + represents positive indexes and - represents reverse indexes).

Table 1
Equivalent weight factor of the ecosystem services per unit of land.
First level Second level Paddy fields Water bodies Wetlands

Provisioning service Food production 1.36 0.80 0.51

Raw materials 0.09 0.23 0.50
Water supply −2.63 8.29 2.59
Regulating service Gas regulation 1.11 0.77 1.90
Climate regulation 0.57 2.29 3.60
Environment purification 0.17 5.55 3.60
Hydrological regulation 2.72 102.24 24.23
Supporting service Soil conservation 0.01 0.93 2.31
Nutrient cycling maintenance 0.19 0.07 0.18
Biodiversity maintenance 0.21 2.55 7.87
Cultural service Landscape aesthetics 0.09 1.89 4.73

et al., 2019). The positive or negative slopes indicated that the index due to the cropland occupation. Other regions were mainly affected by
increased or decreased over many years. natural factors.
From 1990 to 2015, the change in hazard level was mainly between
middle-low, middle and middle-high level, and the number of counties
3. Result of these three levels had an increasing trend. Especially at the middle-
high level, the number of counties reached a maximum of 68 in 2015.
3.1. Spatiotemporal changes in the wetland hazard degree And it was basically in the central and southern part of the study area.
The cities with high hazard degree were Baoding, Shijiazhuang, Xingtai
The hazard degree of the wetlands was evaluated from natural and and Handan. And the development centers of these four cities were
human aspects. The results were divided into five grades by the method basically in high hazard level during 25 years. The cities with low ha-
of Natural Breaks: low, middle-low, middle, middle-high, and high, as zard degree include Chengde, Zhangjiakou and Qinhuangdao. Although
shown in Fig. 4. Overall, the hazard degree of the wetlands in Jing-Jin- the overall hazard degree decreased, but some areas still need to im-
Ji first rose and then declined from 1990 to 2015. Specifically, from plement the strategy of wetland protection as soon as possible.
0.327 in 1990 to 0.353 in 1995 and then to 0.323 in 2015. Spatially, the
hazard degree of the wetlands was low in the northwest, high in the
center and middle in the east. Most of the wetlands in the northwest 3.1.1. Spatiotemporal changes in the natural hazard degree of wetland
were at low or middle-low hazard level, and the other areas were This section calculated the changes of the natural hazard degree and
mostly at the level of middle or middle-high hazard. However, ac- the precipitation and temperature from 1990 to 2015 (Fig. 5a, Fig. S5).
cording to the results of natural and human hazard, the human pressure In general, the natural hazard degree of the wetlands in Jing-Jin-Ji
faced by the northwest wetlands was gradually increasing, most likely tended to be low in the north, high in the middle and low in the south,

5
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 4. Hazard degree level of wetlands in different counties of Jing-Jin-Ji from 1990 to 2015.

but the change rate of the natural indexes in most counties was not human hazard level in these areas was high and has been reduced, but
significant. At the same time, due to the impacts of climate change and the rate of change was also very small. The western region had a large
human activities on the climate, the precipitation in most areas of Jing- amount of natural wetland resources, and the human hazard in the
Jin-Ji showed a downward trend, while the temperature increased, western region had increased, causing the wetlands to face greater
especially in the central region, including Baoding, Shijiazhuang and threats.
Hengshui. The decrease in precipitation and the increase in tempera-
ture reduced the wetland water resources and caused the wetlands in
Jing-Jin-Ji to face greater threats. 3.2. Spatiotemporal changes in the wetland vulnerability

The vulnerability of the wetlands should be considered from three

3.1.2. Spatiotemporal changes in the human hazard degree of wetland aspects: the area, structure and function. Here, the wetland vulner-
The increase in the area of built-up land, population density, GDP ability was divided into five levels. Overall, the vulnerability of the
and the decrease in cropland made the human hazard degree obviously wetlands in Jing-Jin-Ji showed an upward trend, rising from 0.473 in
spatially different (Fig. 5b, Fig. S6). In the western and northwestern 1990 to 0.476 in 2015 (Fig. 6). The lowest value was in 1995, and a
regions, the degree of human hazard increased, while in the central, maximum was reached in 2010, and there was no obvious spatial dif-
eastern and southern regions, the degree of human hazard decreased. ference. The wetland vulnerability of most counties was in the middle-
The former phenomenon might be due to the increase of cropland area high level, and only a few counties were at the low or high vulnerability
and economic development, which had caused the wetlands to bear level. From 1990 to 2015, the high vulnerability counties were mainly
greater risks. Although there were more counties with declining hazard distributed in the southern regions, such as Lixian and Boye, Baoding,
level, this was not a positive phenomenon. In the central, eastern and and Xinji and Luancheng, Shijiazhuang. The low vulnerability counties
southern regions, the wetlands were smaller and more developed. The were mainly distributed in the eastern coastal areas, such as Changli

6
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 5. Change slope of each hazard index from 1990 to 2015. (a) is the comprehensive change of natural hazard index, as well as the change of precipitation and
temperature; (b) is the comprehensive change of human hazard index, as well as the change of built-up land, cropland, population density and GDP.

and Lulong, Qinhuangdao, and Tanghai and Leting, Tangshan. Ac- Among them, several counties with higher degrees of change were lo-
cording to the specific analysis, the ecosystem services value provided cated on the seashore. Lulong and Changli of Qinhuangdao had in-
by most wetlands was slightly decreased, the wetland area was greatly creasing trends, while Luannan and Leting of Tangshan and Ninghe of
reduced, and the structure was widely fragmented. And more was the Tianjin had decreasing trends.
damage of natural wetlands. These changes made an increase in the
high vulnerability wetlands and a decrease in the low vulnerability 3.2.2. Changes in the wetland structure
wetlands. Therefore, the policies of wetland restoration and protection The variables patch number, patch density, fragmentation degree
should consider the intrinsic problems of the wetlands. and aggregation degree were selected to describe the wetland structure
(Fig. 7b, Fig. S7). Overall, the patch number, patch density and frag-
3.2.1. Changes in the wetland area mentation degree of wetlands increased, while aggregation degree de-
The wetlands in Jing-Jin-Ji are widely distributed, but they are also creased. Especially in 2010 and 2015, the significant changes in the
fragmented. Most wetlands are located in the Bohai Bay region and in degrees of fragmentation and aggregation should be considered. The
the northern mountainous areas. By quantifying the wetland area, it patch number and patch density showed a spatial distribution of high in
was found that the wetland area did not change much from 1990 to the north and low in the south. The wetland area in the south was small
2005. However, after 2005, the area had undergone significant and its fragmentation degree was high. The aggregation degree of
changes, mainly in Bohai Bay and the southern region. Wetland area coastal wetland was higher than other areas. Whether this phenomenon
decreased by more than 20% in 2015 compared to 1990 (Fig. S3). Most was due to human factors or natural factors, it was clear that the
of these wetlands were converted into built-up land and cropland. In wetlands were gradually fragmented and dispersed.
terms of different wetland types, the area of paddy field, tidal flat,
beach and marsh decreased, while the area of reservoir and pond, lake 3.2.3. Changes in the wetland function
and river increased slightly. Among them, there were some transfor- This part calculated the value of the 11 ecosystem service functions
mations inside the wetland, for example, part of the marsh was trans- of wetlands, including the regulating services, provisioning services,
formed into reservoir and pond, and a small part of the lake was supporting services and cultural services (Fig. S8–S11). The average
transformed into tidal flat and marsh. It can be found that some natural values of the wetland function and the overall change rates of each
wetlands were transformed into artificial wetlands, and some wetlands county from 1990 to 2015 were calculated (Fig. 7c). The wetland values
that provide more functions were gradually degraded (Fig S4). in Jing-Jin-Ji first increased and then declined, with 2000 as a turning
This section also calculated the overall change rate of the wetland point, and the wetland value reached its maximum in 1995. The wet-
area in each county (Fig. 7a). It can be seen from the maps that most of land value in 2015 increased by 4.41% compared with that in 1990.
the areas decreased, but the wetland areas of some counties increased, From a functional perspective, the values of the water supply and hy-
such as counties in Chengde, Beijing, Qinhuangdao, and Cangzhou. drological regulation functions increased, while the values of other

7
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 6. Vulnerability level of wetlands in different counties of Jing-Jin-Ji from 1990 to 2015.

functions decreased. According to the results of the area and the cal- approximately equal, while there were fewer wetlands with high risk
culation method used in this study, due to the development of aqua- level. The level of the changes was also relatively stable, with most
culture in coastal areas, the increase of reservoir and pond area directly counties shifting between low and middle-low risk level and middle and
affected the increase of wetland value in the study area. At the county middle-high risk level.
scale, the wetland values in more than half of the counties increased but Spatially, the risk level of the wetlands in the study area decreased
not obvious. Some counties in Tianjin and Cangzhou increased sig- gradually from the central and southern areas to the surrounding areas,
nificantly. The counties with declining values were mostly located in while the risk level of the wetlands in the northern and coastal areas
Zhangjiakou, Beijing, Hengshui, and Langfang. were lower. Although coastal areas were faced with high hazard degree,
they were rich in wetland resources and low in vulnerability, the
comprehensive risk was not high. However, the southern region was
3.3. Spatiotemporal changes in the wetland ecological risk level different, the huge external pressure, the shortage and degradation of
its own wetland resources had left it in a state of high risk for a long
The spatial distribution of the wetland risk results was very similar time.
to that of the results of hazard degree and showed obvious spatial dif- In addition, taking 2015 as an example, the municipal districts of
ferences. The risks faced by wetlands in Jing-Jin-Ji still need to be each city were marked out. The risk of the wetlands in municipal dis-
considered. Overall, the risk of the wetlands in Jing-Jin-Ji first in- tricts was generally higher than that in other counties, especially in the
creased and then decreased (Fig. 8). The risk value reached its max- southern parts of Handan, Xingtai, and Cangzhou, etc. This phenom-
imum in 1995 and then reached its lowest level in 2005. The result in enon was related to the high hazard degree in the local wetland.
2015 was very close to the result in 1990 but with a slight decrease. In The average risk values of each city from 1990 to 2015 were also
181 counties, the number of low and middle-low risk level counties and calculated to compare the cities with higher wetland risk values
the number of middle and middle-high risk level counties were

8
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 7. Change slope of each vulnerability index from 1990 to 2015. (a), (b), (c) respectively represent the changes of wetland area, structure and function.

(Fig. 9). Among them, the areas with higher risk values included protection, were in the low, middle-low, and middle-high risk level.
Handan, Hengshui, Langfang, Shijiazhuang and Xingtai, and the wet- Specifically, most of the wetland parks were at the middle-high risk
land risk value in Chengde was the lowest among all cities. The wetland level, while most of the wetland reserves were at the low risk level and
risk level of Cangzhou, Qinhuangdao and Beijing had downward trends, most of the lakes and reservoirs were at the middle-low risk level. The
while in other cities, the risk values after the turning point in 2005 purpose of wetland park construction was not only to protect but also to
increased significantly. Compared with those in 1990, the cities with provide a place for human recreation, so these are more vulnerable to
higher wetland risk values in 2015 included Baoding, Chengde, impacts from human activities. Wetland reserves are different. They
Handan, Hengshui, Shijiazhuang, Tangshan, Zhangjiakou and Tianjin. have a very clear division among the core zone, buffer zone and de-
Some of these cities had very small wetland areas, so the protection and velopment zone. All kinds of human destruction and occupation in the
restoration of the wetlands was even more difficult for these cities. core area are prohibited, as a result, nearly half of the wetland reserves
were at a low risk level. And 70% of the types of wetlands in the lakes
and reservoirs in Jing-Jin-Ji were in the low or middle-low risk grades.
3.4. Ecological risk level of some important wetlands The former was protected because of its high hydrological regulation,
water supply and other functions. The reservoir was constructed
Based on the results of wetland risk level from 1990 to 2015, the manually and has very strict control measures, so both risks were at a
locations of some wetland parks, wetland reserves, lakes and reservoirs lower level.
in Jing-Jin-Ji were marked on the map, and the risk level of the specific
sites and patches were calculated (Fig. S12). Low risk wetland parks
and reserves gradually increased. However, the middle-high risk level 4. Discussion
still occupied a high proportion, indicating that the protection work
should be further strengthened. Lakes and reservoirs were basically 4.1. The significance of the study
distributed in the north, west and coastal areas, and their risk levels for
many years showed a downward trend. Based on the theory of regional ecological risk, this paper selected
Here, the results of 2015 were analyzed in detail (Fig. 10). Overall, 22 indexes from two aspects of the hazard degree and vulnerability to
most of these wetlands, which had received considerable attention and evaluate the ecological risk of the wetlands in Jing-Jin-Ji. The results

9
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 8. Risk levels of wetlands in different counties of Jing-Jin-Ji from 1990 to 2015. The last one marked the location of municipal district, it represents the core
component of the city's main part and the center of urban development.

showed that the wetland ecological risk in the whole region first in- increased and then decreased from 1990 to 2015, which indicated that
creased and then decreased with the decreasing degree of hazard and the wetland protections has been effective in recent years. According to
the increasing degree of vulnerability. The wetland risk value in 1990 the previous studies, although the wetland area in Jing-Jin-Ji had de-
was basically the same as that in 2015. First, the decline in the hazard clined, the momentum of recovery and growth in the later period could
degree was a favorable trend that was due in large part to the reduction not be ignored (Mao et al., 2018a). In addition, it should also be noted
of the degree of human hazard. Engineering measures, such as re- that most of the wetlands with reduced areas were natural wetlands,
turning cropland to wetlands and adjusting the structure of the da- while most of the restored wetlands were constructed wetlands. The
maged wetland ecosystems, played important roles in the decreased functions and value of these two kinds of wetlands were different.
level of the human hazard. However, according to the trend of climate Therefore, an increasing number of scholars have shown that not only
change over the past 25 years, the decrease in rainfall and the rise in the wetlands of Jing-Jin-Ji but the entire ecosystem is still facing a
temperature will still greatly threaten wetlands in the future. Second, greater threat, and the situation of its ecological security is still very
there has been a clear upward trend in vulnerability since 2000. The serious (Chen et al., 2019; Chu et al., 2017; Kang et al., 2018). This was
reduction of wetland area and the fragmentation and separation of the similar to our results. By dividing the risk level, more counties were at
wetland structure were all important factors that affected the change of the middle or middle-high risk level.
the wetlands. Therefore, in addition to controlling the external stressors In addition, several wetlands that required more attention were
as much as possible, wetland management and protection should start discussed separately, including the Baiyangdian wetland in Baoding,
from within the wetland. Methods for more quickly controlling or ex- the Yeyahu wetland in Beijing and the coastal wetlands in Tianjin. In
panding the wetland area and for improving the health of the wetlands order to verify the accuracy of the results of this study from a smaller
to increase the service functions are urgently needed. area. The Baiyangdian Wetland is a national key wetland, which has
The results mentioned that the ecological risk of the wetlands first received tremendous attention and protection in recent years. The

10
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Fig. 9. Average risk value of the wetlands in each city from 1990 to 2015.

Yeyahu Wetland is the largest wetland nature reserve in Beijing, and its As a result, in addition to the Yeyahu wetland, the risk values of the
protection effect is very significant. Tianjin's coastal wetland has a large other two wetlands gradually increased. At the same time, according to
area, which has been seriously damaged in recent years. These three previous studies, the area of the Baiyangdian wetland declined and was
wetlands are typical in Jing-Jin-Ji. Compared with other wetlands, the very vulnerable to climate factors and human activities. In recent
three wetlands are located in counties with low risk or middle-low risk decades, the precipitation in the region has continued to decrease, and
level. However, they could not be considered healthy and without risk. the water level was insufficient, so wetlands are more vulnerable to

Fig. 10. Risk level of the wetland parks, wetland nature reserves, reservoirs and lakes in 2015, and the proportion of each level.

11
Z. Li, et al. Ecological Indicators 117 (2020) 106677

threats (Bai et al., 2016; Hu et al., 2012; Wang et al., 2018). The Yeyahu 4.3. Uncertainty and future research directions
wetland have been greatly affected by human protection and restora-
tion. Therefore, its ecological health was relatively good in the past ten In fact, wetland ecological risk assessment is a very complex pro-
years (Chen et al., 2019). However, it should not be ignored that with cesses that needs to consider many factors. Therefore, there are some
the hosting of the 2022 Winter Olympic Games, some negative human errors and uncertainties in this research that need to be continuously
activities began to gradually affect the Yeyahu wetland; strategies for improved and deepened in future research. First, the selection of the
maintaining the current excellent ecological status of the Yeyahu wet- indexes was not sufficiently comprehensive. In the hazard degree, this
land are very important. The Tianjin coastal wetland showed the most paper did not consider the impact of pollution on the wetlands, such as
rapid increase in risk level of the three wetlands. In addition to the the industrial sewage discharged into the wetlands or the garbage
possible threats considered in this paper, water pollution, alien species thrown by tourists, which can cause high level of pollution for the
invasion and other issues affected the wetland throughout the study wetland. In addition, the change in wetland connectivity was also not
period (Xie et al., 2012). Therefore, the protection of the coastal wet- considered in the vulnerability assessment. Second, this paper studied
lands is an urgent problem that must be solved. wetland function by calculating the value of the ecosystem services,
which was more intuitive and easier to calculate, but there might have
been some errors. The material quality or the emergy can be selected in
4.2. Implementation of the wetland protection and restoration policy future studies. Third, the subjective factors of the researchers might
have caused some fluctuations in the research results. In future re-
People have begun to pay attention to wetlands and protect wet- search, the framework can be adjusted continuously with the different
lands only in recent decades. Since China joined the Ramsar Convention objectives, research areas and data to make it more scientific and thus
on Wetlands in 1992, people have realized that it is necessary to carry contribute more effectively to the protection and restoration of wet-
out effective management and scientific protection and restoration of lands.
wetlands. As a result, China listed “wetland protection and rational
utilization” as a priority development area of China’s Agenda 21 and
established a series of policies and plans. From the China Wetland 5. Conclusions
Conservation Action Plan that was implemented in 2000 to the
Nationwide Wetland Protection Program Plan (2002–2003) issued in In this study, the Jing-Jin-Ji wetlands were the research object, and
2003 to the creation of the Wetland Protection and Restoration System the wetland ecological risk assessment system was established to ana-
Plan (Jiang and Xu, 2019), all have shown that the protection and re- lyze the spatiotemporal changes in the wetland ecological risk during
storation of wetlands has risen to the national level; all plans have the 25-year study period. The results showed that the wetland ecolo-
achieved remarkable results in the past two decades. At present, China gical risk in Jing-Jin-Ji first increased and then decreased from 1990 to
has identified 57 internationally important wetlands, 898 national 2015 and that the risk value in 2015 was nearly equal to that in 1990, at
wetland parks and 602 wetland nature reserves. The wetland resources only 0.12% lower. Wetlands still faced a greater threat when the vul-
in these areas have been well-protected and maintained. With the es- nerability of the wetland itself was high. In the 181 studied counties,
tablishment of the Ministry of Natural Resources and the Ministry of the wetlands of most counties were still at the middle or middle-high
Ecology and Environment, these two leading departments with wetland risk level. However, it was still undeniable that the relevant policies
management responsibilities will make wetland management more made wetlands safer and healthier; for example, more wetland nature
scientific and effective in the future. reserves were at the low or middle-low risk level. Therefore, in the
Since the 1990s, Beijing, Tianjin and Hebei have gradually im- future, Jing-Jin-Ji should introduce stricter measures and control sys-
plemented measures to protect and restore wetlands. In 1997, Beijing tems, strengthen the management and protection of wetlands, and
established the first wetland nature reserve, the Yanqing Yeyahu Nature thereby aid the sustainable development and construction of the eco-
Reserve, and then developed a series of measures, such as the Beijing logical civilization in Jing-Jin-Ji.
Wetland Protection Action Plan, and implemented a series of projects
for wetland restoration and management. The wetland resources in
Tianjin are abundant, but they have been significantly reduced in re- CRediT authorship contribution statement
cent decades. Therefore, the local government has implemented the
Tianjin Wetland Protection Regulations, Tianjin Wetland Protection Zhuo Li: Data curation, Methodology, Software, Formal analysis,
and Restoration Work Implementation Plan and other policies that aim Writing - original draft, Writing - review & editing. Weiguo Jiang:
to control of all of the wetland areas and strive to restore the wetlands. Conceptualization, Funding acquisition, Resources, Writing - review &
Hebei has several important wetland nature reserves. Meanwhile, in the editing. Wenjie Wang: Supervision, Funding acquisition. Zheng Chen:
past decade, Hebei has also increased its efforts in wetland management Validation, Data curation. Ziyan Ling: Formal analysis. Jinxia Lv:
and protection and issued the Implementation Plan for Hebei Province Methodology, Data curation.
Wetland Protection and Restoration System, the Hebei Province
Wetland Protection Plan (2015–2030) and other documents, with the
main goal of protecting, restoring and expanding the wetland ecological Declaration of Competing Interest
space. The strategic position of the Jing-Jin-Ji is very important, so
strategies for balancing the contradiction between the development of The authors declare that they have no known competing financial
the social economy and the protection of the ecological environment interests or personal relationships that could have appeared to influ-
are the key issues for achieving the coordinated development plan of ence the work reported in this paper.
Jing-Jin-Ji. In the Thirteenth Five-Year Plan for forestry development,
the construction of the Jing-Jin-Ji ecological coordination circle was
proposed, which would comprehensively protect the wetland resources; Acknowledgments
increase the restoration of lake wetlands, such as Baiyangdian and
Hengshui Lake; strengthen the construction of wetland parks; and This work was supported by the National Key Research and
create an important pillar for national green growth. At present, some Development Program of China (2016YFC0503002), the National
wetlands have been effectively protected. Natural Science Foundation of China (41571077, 41731286,
U1901219).

12
Z. Li, et al. Ecological Indicators 117 (2020) 106677

Appendix A. Supplementary data Luo, Q.L., Luo, L.Y., Zhou, Q.F., Song, Y., 2019. Does China's Yangtze River Economic Belt
policy impact on local ecosystem services? Sci. Total Environ. 676, 231–241.
Ma, J., Xiao, X.M., Miao, R.H., Li, Y., Chen, B.Q., Zhang, Y., Zhao, B., 2019. Trends and
Supplementary data to this article can be found online at https:// controls of terrestrial gross primary productivity of China during 2000–2016.
doi.org/10.1016/j.ecolind.2020.106677. Environ. Res. Lett. 14.
Malekmohammadi, B., Rahimi Blouchi, L., 2014. Ecological risk assessment of wetland
ecosystems using Multi Criteria Decision Making and Geographic Information
References System. Ecol. Ind. 41, 133–144.
Mao, D.H., Luo, L., Wang, Z.M., Wilson, M.C., Zeng, Y., Wu, B.F., Wu, J.G., 2018a.
Bai, J.H., Lu, Q.Q., Wang, J.J., Zhao, Q.Q., Ouyang, H., Deng, W., Li, A.N., 2013. Conversions between natural wetlands and farmland in China: a multiscale geospatial
Landscape pattern evolution processes of alpine wetlands and their driving factors in analysis. Sci. Total Environ. 634, 550–560.
the Zoige Plateau of China. J. Mountain Sci. 10 (1), 54–67. Mao, D.H., Wang, Z.M., Wu, J.G., Wu, B.F., Zeng, Y., Song, K.S., Yi, K.P., Luo, L., 2018b.
Bai, Y., Jiang, B., Alatalo, J.M., Zhuang, C.W., Wang, X.Y., Cui, L.J., Xu, W.H., 2016. China's wetlands loss to urban expansion. Land Degrad. Dev. 29, 2644–2657.
Impacts of land management on ecosystem service delivery in the Baiyangdian river Mao, X.F., Yang, Z.F., Chen, B., Chen, H., 2010. Examination of wetlands system using
basin. Environ. Earth Sci. 75 (3), 258. ecological network analysis: a case study of Baiyangdian Basin, China. Proc. Environ.
Burton, G.A., Chapman, P.M., Smith, E.P., 2002. Weight-of-Evidence approaches for as- Sci. 2, 427–439.
sessing ecosystem impairment. Hum. Ecol. Risk Assess. 8, 1657–1673. Meng, W.Q., He, M.X., Hu, B.B., Mo, X.Q., Li, H.Y., Liu, B.Q., Wang, Z.L., 2017. Status of
Chen, S.Q., Chen, B., Fath, B.D., 2013. Ecological risk assessment on the system scale: a wetlands in China: a review of extent, degradation, issues and recommendations for
review of state-of-the-art models and future perspectives. Ecol. Model. 250, 25–33. improvement. Ocean Coast. Manag. 146, 50–59.
Chen, W., Cao, C.X., Liu, D., Tian, R., Wu, C.Y., Wang, Y.Q., Qian, Y.F., Ma, G.Q., Bao, Muro, J., Strauch, A., Heinemann, S., Steinbach, S., Thonfeld, F., Waske, B., Diekkrüger,
D.M., 2019. An evaluating system for wetland ecological health: Case study on B., 2018. Land surface temperature trends as indicator of land use changes in wet-
nineteen major wetlands in Beijing-Tianjin-Hebei region, China. Sci. Total Environ. lands. Int. J. Appl. Earth Obs. Geoinf. 70, 62–71.
666, 1080–1088. Ni, L.L., Wang, D., Singh, V.P., Wu, J.F., Wang, Y.K., Tao, Y.W., Liu, J.F., Zou, Y., He,
Chen, Z., Jiang, W.G., Wang, W.J., Deng, Y., He, B., Jia, K., 2018. The impact of pre- R.M., 2019. A hybrid model-based framework for estimating ecological risk. J.
cipitation deficit and urbanization on variations in water storage in the Beijing- Cleaner Prod. 225, 1230–1240.
Tianjin-Hebei Urban Agglomeration. Rem. Sens. 10 (1), 4. Norton, S.B., Rodier, D.J., van der Schalie, W.H., Wood, W.P., Slimak, M.W., Gentile, J.H.,
Chu, X., Deng, X.Z., Jin, G., Wang, Z., Li, Z.H., 2017. Ecological security assessment based 1992. A framework for ecological risk assessment at the EPA. Environ. Toxicol. Chem.
on ecological footprint approach in Beijing-Tianjin-Hebei region, China. Phys. Chem. 11, 1663–1672.
Earth, Parts A/B/C 101, 43–51. Peng, J., Pan, Y.J., Liu, Y.X., Zhao, H.J., Wang, Y.L., 2017. Linking ecological degradation
Costanza, R., d’Arge, R., Groot, R.D., Farber, S., Grasso, M., Hannon, B., Limburg, K., risk to identify ecological security patterns in a rapidly urbanizing landscape. Habitat
Naeem, S., O’Neill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., Belt, M.v.D., 1997. The Int. 71, 110–124.
value of the world’s ecosystem services and natural capital. Nature. 387, 253–260. Qian, D.W., Yan, C.Z., Xiu, L.N., Feng, K., 2018. The impact of mining changes on sur-
Cui, L.L., Li, G.S., Ouyang, N.L., Mu, F.Y., Yan, F., Zhang, Y.T., Huang, X.Y., 2018. rounding lands and ecosystem service value in the Southern Slope of Qilian
Analyzing coastal wetland degradation and its key restoration technologies in the Mountains. Ecol. Complexity 36, 138–148.
coastal area of Jiangsu, China. Wetlands 38 (3), 525–537. Ribeiro, D.C., Costa, S., Guilhermino, L., 2016. A framework to assess the vulnerability of
Cui, X.G., Fang, C.L., Liu, H.M., Liu, X.F., 2019. Assessing sustainability of urbanization estuarine systems for use in ecological risk assessment. Ocean Coast. Manag. 119,
by a coordinated development index for an Urbanization-Resources-Environment 267–277.
complex system: a case study of Jing-Jin-Ji region, China. Ecol. Indic. 96, 383–391. Saaty, T.L., 1980. The analytic hierarchy process. New York: McGraw Hill. International,
Davidson, N.C., 2014. How much wetland has the world lost? Long-term and recent Translated to Russian, Portuguese, and Chinese, Revised editions, Paperback (1996,
trends in global wetland area. Mar. Freshw. Res. 65, 934–941. 2000), Pittsburgh: RWS Publications.
Fu, Y.C., Zhao, J.Y., Peng, W.Q., Zhu, G.P., Quan, Z.J., Li, C.H., 2018. Spatial modelling of Sarkar, S., Parihar, S.M., Dutta, A., 2015. Fuzzy risk assessment modelling of East Kolkata
the regulating function of the Huangqihai Lake wetland ecosystem. J. Hydrol. 564, Wetland Area: a remote sensing and GIS based approach. Environ. Modell. Software
283–293. 75, 105–118.
Guo, T., Zhang, X.J., Cheng, T.X., 2011. Evaluation of wetland ecological restoration Serran, J.H., Creed, I.F., 2015. New mapping techniques to estimate the preferential loss
project based on SD: a case study on Qilihai Wetland in Tianjin. Proc. Environ. Sci. of small wetlands on prairie landscapes. Hydrol. Process. 30 (3), 396–409.
10, 2587–2593. Shi, C., Ding, H., Zan, Q.J., Li, R.L., 2019. Spatial variation and ecological risk assessment
Guo, Z.L., Zhang, M.Y., 2019. The conservation efficacy of coastal wetlands in China of heavy metals in mangrove sediments across China. Mar. Pollut. Bull. 143,
based on landscape development and stress. Ocean Coast. Manage. 175, 70–78. 115–124.
Han, J.Y., Wang, J.H., Zhao, Y., Wang, Q.M., Zhang, B., Li, H.H., Zhai, J.Q., 2018. Spatio- Shi, Y.J., Xu, X.B., Li, Q.F., Zhang, M., Li, J., Lu, Y.L., Liang, R.Y., Zheng, X.Q., Shao, X.Q.,
temporal variation of potential evapotranspiration and climatic drivers in the Jing- 2018. Integrated regional ecological risk assessment of multiple metals in the soils: a
Jin-Ji region, North China. Agric. Forest Meteorol. 256–257, 75–83. case in the region around the Bohai Sea and the Yellow Sea. Environ. Pollut. 242,
Hu, S.J., Niu, Z.G., Chen, Y.F., Li, L.F., Zhang, H.Y., 2017. Global wetlands: potential 288–297.
distribution, wetland loss, and status. Sci. Total Environ. 586, 319–327. Sun, Y., Zhao, S.Q., 2019. Spatiotemporal dynamics of urban expansion in 13 cities across
Hu, S.S., Liu, C.M., Zheng, H.X., Wang, Z.G., Yu., J.J., 2012. Assessing the impacts of the Jing-Jin-Ji Urban Agglomeration from 1978 to 2015. Ecol. Ind. 87, 302–313.
climate variability and human activities on streamflow in the water source area of Tian, J.H., Wang, J.H., Liu, Z.C., Li, H.C., Wang, X.Z., Xie, D.Y., Tian, Y.K., Liu, F.C., Yu,
Baiyangdian Lake. Journal of Geographical Sciences. 22(5), 895-905. X.J., 2010. A conceptual ecological model of Cangzhou Coastal Wetlands, Hebei
Ji, Z.H., Zhang, H., Zhang, Y., Chen, T., Long, Z.W., Li, M., Pei, Y.S., 2019. Distribution, Province, China. Proc. Environm. Sci. 2, 1002–1011.
ecological risk and source identification of heavy metals in sediments from the Tixier, G., Lafont, M., Grapentine, L., Rochfort, Q., Marsalek, J., 2011. Ecological risk
Baiyangdian Lake, Northern China. Chemosphere 237. assessment of urban stormwater ponds: literature review and proposal of a new
Jiang, B., Xu, X.B., 2019. China needs to incorporate ecosystem services into wetland conceptual approach providing ecological quality goals and the associated bioas-
conservation policies. Ecosyst. Serv. 37. sessment tools. Ecol. Ind. 11 (6), 1497–1506.
Jiang, W.G., Lv, J.X., Wang, C.C., Chen, Z., Liu, Y.H., 2017. Marsh wetland degradation Walker, R., Landis, W., Brown, P., 2001. Developing a regional ecological risk assessment:
risk assessment and change analysis: a case study in the Zoige Plateau, China. a case study of a tasmanian agricultural catchment. Hum. Ecol. Risk Assess. 7,
Ecological Indicators. 82, 316–326. 417–439.
Jiang, Y.C., Nan, Z.R., Yang, S.C., 2013. Risk assessment of water quality using Monte Wanda, E.M.M., Mamba, B.B., Msagati, T.A.M., Msilimba, G., 2015. Determination of the
Carlo simulation and artificial neural network method. J. Environ. Manage. 122, health of Lunyangwa wetland using Wetland Classification and Risk Assessment
130–136. Index. Phys. Chem. Earth, Parts A/B/C. 92, 52–60.
Kang, P., Chen, W.P., Hou, Y., Li, Y.Z., 2018. Linking ecosystem services and ecosystem Wang, X.Y., Wang, W.J., Jiang, W.G., Jia, K., Rao, P.Z., Lv, J.X., 2018. Analysis of the
health to ecological risk assessment: a case study of the Beijing-Tianjin-Hebei urban dynamic changes of the Baiyangdian Lake surface based on a complex water ex-
agglomeration. Sci. Total Environ. 636, 1442–1454. traction method. Water 10 (11), 1616.
Kokangül, A., Polat, U., Dağsuyu, C., 2017. A new approximation for risk assessment Xie, G.D., Zhang, C.X., Zhang, L.M., Chen, W.H., Li, S.M., 2015. Improvement of the
using the AHP and Fine Kinney methodologies. Saf. Sci. 91, 24–32. evaluation method for ecosystem service value based on per unit area. J. Natural
Liang, X.X., Junaid, M., Wang, Z.H., Li, T.H., Xu, N., 2019. Spatiotemporal distribution, Resour. 30 (8), 1243–1254 (in Chinese).
source apportionment and ecological risk assessment of PBDEs and PAHs in the Xie, Z.L., Liu, J.Y., Ma, Z.W., Duan, X.F., Cui, Y.P., 2012. Effect of surrounding land-use
Guanlan River from rapidly urbanizing areas of Shenzhen, China. Environ. Pollut. change on the wetland landscape pattern of a natural protected area in Tianjin,
250, 695–707. China. Int. J. Sustain. Dev. World Ecol. 19 (1), 16–24.
Liu, J.Y., Kuang, W.H., Zhang, Z.X., Xu, X.L., Qin, Y.W., Ning, J., Zhou, W.C., Zhang, S.W., Xu, T., Weng, B.S., Yan, D.H., Wang, K., Li, X.N., Bi, W.X., Li, M., Cheng, X.J., Liu, Y.X.,
Li, R.D., Yan, C.Z., Wu, S.X., Shi, X.Z., Jiang, N., Yu, D.S., Pan, X.Z., Chi, W.F., 2014. 2019. Wetlands of international importance: status, threats, and future protection.
Spatiotemporal characteristics, patterns, and causes of land-use changes in China Int. J. Environ. Res. Public Health 16 (10), 1818.
since the late 1980s. J. Geog. Sci. 24 (2), 195–210. Xu, X.B., Yang, G.S., Tan, Y., Zhuang, Q.L., Li, H.P., Wan, R.R., Su, W.Z., Zhang, J., 2016.
Liu, W., Zhan, J.Y., Zhao, F., Yan, H.M., Zhang, F., Wei, X.Q., 2019. Impacts of urbani- Ecological risk assessment of ecosystem services in the Taihu Lake Basin of China
zation-induced land-use changes on ecosystem services: a case study of the Pearl from 1985 to 2020. Sci. Total Environ. 554–555, 7–16.
River Delta Metropolitan Region, China. Ecol. Indic. 98, 228–238. Xu, X.G., Lin, H.P., Fu, Z.Y., 2003. Probe into the method of regional ecological risk
Luo, F.H., Liu, Y.X., Peng, J., Wu, J.S., 2018. Assessing urban landscape ecological risk assessment—a case study of wetland in the Yellow River Delta in China. J. Environ.
through an adaptive cycle framework. Landscape Urban Plann. 180, 125–134. Manage. 70, 253–262.

13