You are on page 1of 10

Chemosphere 60 (2005) 542–551

Assessment of heavy metal pollution in surface soils of

urban parks in Beijing, China
Tong-Bin Chen a,*, Yuan-Ming Zheng a, Mei Lei a, Ze-Chun Huang a,
Hong-Tao Wu a, Huang Chen a, Ke-Ke Fan b, Ke Yu c,
Xiao Wu b, Qin-Zheng Tian b
Center for Environmental Remediation, Institute of Geographic Sciences and Natural Resources Research,
Chinese Academy of Sciences, 11A Datun Road, Beijing 100101, PR China
Middle School Affiliated to People’s University of China, Beijing 100081, PR China
Computing Laboratory, Oxford University, Parks Road, Oxford OX1 3QD, UK

Received 29 March 2004; received in revised form 23 December 2004; accepted 24 December 2004
Available online 10 February 2005


Assessing the concentration of potentially harmful heavy metals in the soil of urban parks is imperative in order to
evaluate the potential risks to residents and tourists. To date, little research on soil pollution in ChinaÕs urban parks has
been conducted. To identify the concentrations and sources of heavy metals, and to assess the soil environmental qua-
lity, samples were collected from 30 urban parks located in the city of Beijing. Subsequently, the concentrations of Cu,
Ni, Pb and Zn in the samples were analyzed. The investigation revealed that the accumulations of Cu and Pb were read-
ily apparent in the soils. The integrated pollution index (IPI) of these four metals ranged from 0.97 to 9.21, with the
highest IPI in the densely populated historic center district (HCD). Using multivariate statistic approaches (principal
components analysis and hierarchical cluster analysis), two factors controlling the heavy metal variability were
obtained, which accounted for nearly 80% of the total variance. Nickel and Zn levels were controlled by parent material
in the soils, whereas Cu, Pb and, in part, Zn were accounted for mainly by anthropogenic activities. The findings pre-
sented here indicate that the location and the age of the park are important factors in determining the extent of heavy
metal, particularly Cu and Pb, pollution. In addition, the accumulation of Zn did not appear to reach pollution levels,
and no obvious pollution by Ni was observed in the soils of the parks in Beijing.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Beijing; Heavy metal; Pollution index; Soil; Urban park

1. Introduction

The mobilization of heavy metals into the biosphere

by human activity has become an important process in
Corresponding author. Tel.: +86 10 64889080; fax: +86 10 the geochemical cycling of these metals. This is acutely
64889303. evident in urban areas where various stationary and
E-mail address: (T.-B. Chen). mobile sources release large quantities of heavy metals

0045-6535/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
T.-B. Chen et al. / Chemosphere 60 (2005) 542–551 543

into the atmosphere and soil, exceeding the natural emis- the Southeast Suburb. Recently we have undertaken
sion rates (Nriagu, 1989; Bilos et al., 2001). In contrast an intensive investigation on heavy metal concentrations
to agricultural soils, urban soils, especially that in parks in agricultural and forestry soils of the city (Zheng et al.,
and residential areas which is not used for food crops, 2003). However, the heavy metal concentrations and
may also have a direct influence on public health since subsequent soil pollution levels in BeijingÕs urban parks
it can be easily transferred into human bodies (De Mig- remain unknown. The concentrations and pollution lev-
uel et al., 1997; Mielke et al., 1999; Madrid et al., 2002). els of soil Cu, Ni, Pb and Zn in urban parks were as-
In particular, the ingestion of dust and soil has been sessed to evaluate the environmental quality of the
widely regarded as one of the key pathways by which soils in the urban parks and the potential risks to resi-
children are exposed to the heavy metals and metalloids dents and tourists.
from paint, leaded gasoline, vehicles and local industry
(Meyer et al., 1999; Rasmussen et al., 2001).
Several researchers have indicated the need for a bet-
2. Materials and methods
ter understanding of urban soil pollution (De Kimple
and Morel, 2000; Manta et al., 2002), and, indeed,
2.1. Study area
increasing research has focused on heavy metals in
urban soils (Kelly et al., 1996; Chen et al., 1997; Mielke
Beijing city, located in the Northern China, is not
and Reagan, 1998). Heavy metals in urban soils may
only a quickly developing city, but is also an ancient city
come from various human activities, such as industrial
with over 1000 years of history and more than 8 million
and energy production, construction, vehicle exhaust,
urban residents. The city consists of fourteen adminis-
waste disposal, as well as coal and fuel combustion
trative districts and four counties. In seven central
(Komai, 1981; Ikeda and Yoda, 1982; Ritter and Rine-
districts, namely Chaoyang, Chongwen, Dongcheng,
fierd, 1983; Chon et al., 1995; Wong and Mak, 1997;
Fengtai, Haidian, Xicheng and Xuanwu, there are
Martin et al., 1998; Li et al., 2001). These activities send
approximately 40 main parks (Fig. 1). The 30 parks
heavy metals into the air and the metals subsequently
investigated in this study are major resort or scenic sites
are deposited into urban soil as the metal containing
in Beijing that frequently welcome a large number of
dust falls. Sakagami et al. (1982) reported that there
visitors, including local residents and tourists from
was a close relationship between heavy metal concentra-
throughout the world. Some of the parks, such as
tions in soils and those in the dust falls. Heavy metals in
the Forbidden City, Summer Palace, Winter Palace
the soils can also generate airborne particles and dusts,
and Heaven Temple, were built hundreds of years ago
which may affect the air environmental quality (Chen
(Table 1).
et al., 1997; Bandhu et al., 2000; Cyrys et al., 2003; Gray
The seven districts listed above are the majority of
et al., 2003).
the commercial, industrial and traditional residential
Previous studies indicated that the extent of heavy
areas of Beijing. The districts of Chongwen, Dongcheng,
metal pollution in urban areas varied across time (Pfeif-
Xicheng and Xuanwu comprise the older historic center
fer et al., 1991) and location (Albasel and Cottenie,
district (HCD), while the districts of Haidian in the
1985), and that increased levels of heavy metals in urban
northern region, Chaoyang in the eastern region and
soil was related to the intensity of human activities and
Fengtai in the southern region are the newer districts
traffic volume (Zheng et al., 2002). The investigation of
and have experienced substantial development over the
soil heavy metal concentrations in parks and green areas
past 50 years. This study examined parks both within
in Seville, Spain, indicated that the concentrations of Pb,
the HCD as well as in the other newer regions (Fig. 1;
Zn and particularly Cu in the soil often exceeded the
Table 1). According to the Beijing Municipal Statistical
acceptable limits for residential, recreational and institu-
Bureau (2001), the population densities in 2000 were
tional sites (Madrid et al., 2002). However, differences
27 332 persons/km2 in the HCD and 3337 persons/km2
among cities including population density and industrial
outside the HCD. Moreover, although the area of the
activities could have a large impact on the findings of
HCD comprised only 8.2% of Beijing city, it accounted
individual studies. Moreover, little information is avail-
for 37.2% of the traffic burden (Jin, 2000). Thus, the
able on heavy metal concentrations in soils of urban
HCD is a much more densely populated and high traffic
parks located in older cities with large populations.
Beijing, the capital of China, is one of the oldest and
most densely populated cities in the world. During the
past two decades, research on heavy metal concentra- 2.2. Sampling and analysis
tions in the natural and agricultural soils of rural areas
has been conducted in Beijing, including surveys of Depending on the area of the parks, 14–30 sub-sam-
the background concentrations of soil heavy metals ples of the topsoil (0–5 cm) were collected in each park
(CNEMC, 1990) and heavy metal contamination in and mixed thoroughly to get a representative sample.
544 T.-B. Chen et al. / Chemosphere 60 (2005) 542–551

Fig. 1. Sampling sites of soils taken from the urban parks. Full names of the parks listed in Table 1.

Table 1
Brief description of the parks investigated in Beijing
Name of parks Abbr. name District History (years) Location (inside or outside the HCD) Area (ha)
Forbidden City GUG Dongcheng 596 Inside 72
Heaven Temple TTP Chongwen 582 Inside 270
Zhongshan Park ZSP Dongcheng 581 Inside 24
Xuanwu Park XWP Xuanwu 538 Inside 7.4
Solar Temple RTP Chaoyang 472 Outside 1
Terra Temple DTP Dongcheng 472 Inside 42.7
Lunar Temple YTP Xicheng 472 Inside 8.1
Taoranting Park TRT Xuanwu 307 Inside 59
Winter Palace YMY Haidian 258 Outside 350
Jingshan Park JSP Xicheng 251 Inside 23
Summer Palace YHY Haidian 238 Outside 290.8
Beihai Park BHP Xicheng 206 Inside 71
Beijing Zoo BJZ Xicheng 94 Inside 90
Longtanhu Park LTH Chongwen 50 Inside 120
Honglingjin Park HLJ Chaoyang 44 Outside 37.5
Liuyin Park LYP Dongcheng 44 Inside 19
Qingnianhu Park QNH Dongcheng 44 Inside 17
Yuyuantan Park YYT Haidian 42 Outside 140.7
Zizhuyuan Park ZZY Haidian 37 Outside 14
Guanyuan Park GYP Xicheng 20 Inside 15
Chaoyang Park CYP Chaoyang 18 Outside 320
Shuangxiu Park SXP Xicheng 18 Inside 6.4
Tuanjiehu Park TJH Chaoyang 16 Outside 13.8
Rendinghu Park RDH Xicheng 16 Inside 9
Fengtai Park FTP Fengtai 16 Outside 20
Daguan Yuan DGY Xuanwu 16 Inside 13
Beijing Amusement Park BJA Chongwen 15 Inside 40
Wanfangting Park WFT Fengtai 12 Outside 10.6
World Park SJP Fengtai 9 Outside 46.7
Minzu Park MZP Chaoyang 8 Outside 40
T.-B. Chen et al. / Chemosphere 60 (2005) 542–551 545

The sample was air-dried, ground, passed through a 3. Results

nylon sieve of 100 meshes, and digested with HNO3
and H2O2 using the Method 3050B suggested by Descriptive statistical results are reported in Table 2,
USEPA (1996). The concentrations of Cu, Ni, Pb and including the geometric mean and the geometric stan-
Zn in the digestion solution were analyzed with a flame dard deviation. Mean values were used when heavy
atomic absorption spectrometer (Vario 6, Jena Co. Ltd., metal concentrations had normal distributions and geo-
Germany). Standard reference materials, GSS-1 soils, metric mean values were used when heavy metal concen-
obtained from Center of National Standard Reference trations had lognormal distributions. We found that Ni
Material of China, were inserted for quality assurance levels were comparable with the BC, while Cu, Pb and
and quality control (QA/QC) procedures. Satisfactory Zn exhibited higher concentrations than the BC, partic-
recoveries were obtained for Cu (91–95%), Ni (103– ularly Cu and Pb, which had about 2- and 1-fold higher
112%), Pb (95–108%) and Zn (95–104%). levels, respectively. The highest concentrations of Cu
and Pb, 457.5 and 207.5 mg/kg, found in GUG, the old-
2.3. Statistical treatment est park, were 18- and 6-fold higher than the BC. There-
fore, it was evident that Cu and Pb pollution existed in
This study employed hierarchical cluster analysis the park soils samples.
(HCA) and principal components analysis (PCA) of
the Cu, Ni, Pb, and Zn concentrations in the soils. Be- 3.1. Hierarchical cluster analysis (HCA)
cause the concentrations of the soil heavy metals varied
greatly, the raw data were standardized before HCA. Based on the HCA results, the samples were classified
Statistical calculations were also made within each clus- into six clusters using a criteria value of rescaled distance
ter to allow for further comparison between clusters. between 5 and 10 (Fig. 2). There were 13 parks in cluster
To assess the soil environment quality, a pollution I, 11 parks in cluster II, two parks in clusters III and V,
index (PI) of each metal and an integrated pollution and only one park in clusters IV and VI, respectively
index (IPI) of the four metals were attributed to each (Fig. 1; Table 3).
park. The PI was defined as the ratio of the heavy metal The mean concentrations of heavy metals in cluster I
concentration in the study to the geometric means of were higher than those in cluster II (Table 3). Means
background concentration (BC) of the corresponding comparison of heavy metal indicated that there were sig-
metal of Beijing (Chen et al., 2004). The PI of each metal nificant differences between the two clusters for Pb, Ni
was calculated and classified as either low (PI 6 1), mid- and Cu. Notably, 12 of the 13 parks in cluster I were in-
dle (1 < PI 6 3) or high (PI > 3). The IPI of the four side the HCD, while nine of the 11 parks in cluster II
metals for each park was defined as the mean value of were outside the HCD (Fig. 1). The location of the parks
the metalÕs PI, and was then classified as low (IPI 6 1), appears to affect the heavy metal concentrations in the
middle (1 < IPI 6 2) or high (IPI > 2). soil samples greatly.
The distribution of data was tested with the Shapiro– The samples in clusters III, IV and VI, which were
Wilk method (p < 0.05). All statistical treatments men- characterized with longer histories, had much higher
tioned above were performed using SPSS and Origin concentrations of Pb compared with those in clusters I
Pro software. and II. There were significant differences between the

Table 2
Heavy metal concentrations in the soils of urban parks in Beijing
Metal concentration (mg/kg) Distribution type
Min. Max. Mean S.D. Geom. M. Geom. S.D.
Soils from the investigation of urban parks (Soil samples = 30)
Cu 24.1 457.5 71.2 74.7 59.6 1.63 Lognormal
Ni 6.10 37.2 22.2 8.70 21.4 1.49 Normal
Pb 25.5 207.5 66.2 44.2 55.8 1.76 Lognormal
Zn 25.7 196.9 87.6 31.2 82.5 1.44 Lognormal
Background concentrations in the soils of Beijing (Chen et al., 2004)
Cu 6.00 37.9 19.7 6.30 18.7 1.41 Lognormal
Ni 11.0 59.3 27.9 7.90 26.8 1.34 Lognormal
Pb 11.5 38.2 25.1 5.10 24.6 1.28 Lognormal
Zn 27.9 119.8 59.6 16.3 57.5 1.30 Lognormal
546 T.-B. Chen et al. / Chemosphere 60 (2005) 542–551

higher than the mean concentrations of all parks.

GUG was built as an imperial palace about 600 years
ago. It is a popular park located in the HCD with the
highest population and traffic density in the city, and
accommodates a large quantity of visitors every year.
All these factors are likely to accelerate accumulation
of heavy metals in the soils.
Therefore, it seemed that the location of the urban
park was a very important factor in determining the soil
heavy metal concentrations since the samples collected
from inside the HCD had a much greater accumulation
of heavy metals. The history of park may also contribute
to the increased concentrations of heavy metals in the
There were no significant differences between the Zn
concentrations of any two clusters, suggesting that
human activities have less different influence on Zn levels
in the soils.

3.2. Principal components analysis (PCA)

The results of the PCA indicated that Cu, Ni, Pb and

Zn concentrations could be reduced to two components,
Fig. 2. Dendrogram of hierarchical cluster analysis of heavy
metal concentrations in urban park soils of Beijing.
which accounted for nearly 80% of the total variance for
the data (Table 4). The communalities of variables ran-
ged from 75% for Pb to 85% for Ni. All the elements
were well represented by two components.
Pb concentrations of cluster III and those of clusters I The initial component matrix indicated that Cu and
and II. There were only one park in clusters IV and VI Pb were associated, displaying high values in the first
and no comparison could be conducted between cluster component (F1). Nickel and Zn showed greater values
IV or VI and other clusters. in the second component (F2) and were also partially
Cluster V was unique because of its low concentra- represented in F1. The rotation of the matrix eliminated
tion of Ni. The Pb concentration was not significantly ambiguities. As shown in Table 4, F1 included Cu and
different from that of cluster II, but was significantly dif- Pb, F2 included Ni, and Zn was distributed in both F1
ferent from those of clusters I and III. The samples in and F2. This implies that some of the Cu, Pb and Zn
this cluster were taken from the parks at the southern in the soils may originate from similar pollution sources,
border of the HCD. Therefore, the concentrations of such as the deposition of aerosol particles emitted by
the metals, with the exception of Ni, were similar to that vehicular traffic (Artaxo et al., 1999; Cyrys et al., 2003;
in cluster II. Gray et al., 2003), and that the parent materials of the
Cluster VI consisted of GUG with the highest Pb and soils may control the concentrations of Ni, and Zn in
Cu concentrations, which were about 6- and 3-fold part.

Table 3
Statistics of the clusters in hierarchical cluster analysis (HCA)
Cluster No. of samples Means of park history (years) Soil heavy metal concentration (mg/kg)
Cu Ni Pb Zn
Mean S.D. Mean S.D. Mean S.D. Mean S.D.
I 13 197.2 62.9a 14.8 27.1a 4.41 63.1a 26.7 83.0 17.5
II 11 99.5 46.5b 12.7 17.6b 4.25 41.1b 11.6 78.7 23.8
III 2 222.0 74.4ab 23.5 19.5b 6.37 147.0c 13.6 102.5 20.3
IV 1 472.0 74.7 – 37.2 – 136.6 – 80.8 –
V 2 161.0 62.7ab 3.66 6.10 – 37.8b 1.8 124.3 102.6
VI 1 596.0 457.5 – 36.6 – 207.5 – 148.5 –
a, b and c showed that there are significant differences between heavy metal concentrations at 0.05 levels.
T.-B. Chen et al. / Chemosphere 60 (2005) 542–551 547

Table 4
Total variance explained and component matrices for the heavy metals
Total variance explained
Factor Initial eigenvalues Extraction sums of squared Rotation sums of squared loadings
Total % of Cumulative Total % of Cumulative Total % of Cumulative
variance (%) variance (%) variance (%)
1 2.140 53.51 53.51 2.140 53.51 53.51 2.064 51.59 51.59
2 1.126 28.15 81.66 1.126 28.15 81.66 1.203 30.07 81.66
3 0.473 11.83 93.49
4 0.260 6.51 100.00
Component matrix
Element Component matrix Rotated component matrix
1 2 1 2
Cu 0.917 0.013 0.878 0.265
Ni 0.487 0.780 0.254 0.884
Pb 0.870 0.007 0.835 0.246
Zn 0.553 0.720 0.729 0.540
Extraction Method: Principal Component Analysis (PCA).

3.3. Statistical results Table 5

Statistical results of pollution index (PI) of heavy metals in the
According to the location, all parks investigated were urban park soils of Beijing
classified into two groups: the parks inside the HCD PI Number of samples
(group I) or the ones outside the HCD (group II). The Min Max Mean Low Middle High
Cu concentration for group I was significantly greater
Cu 1.29 24.47 3.81 0 14 16
than that for group II (p < 0.01).
Ni 0.23 1.39 0.83 20 9 0
Based on the history of their establishment, the parks Pb 1.04 8.43 2.69 0 21 9
were classified into two groups: the parks with a history Zn 0.45 3.42 1.52 4 25 1
< 100 years (class A) and the parks with a history > 100
years (class B). There were significant differences be-
tween two classes (p < 0.01).
There were significant correlations between the his- The PIs of Cu and Pb were much higher since all of
tory of the parks and the concentrations of Cu (correla- the 30 samples contained medium to high PIs, ranging
tion coefficients: 0.580, p < 0.01) and Pb (correlation from 1.29 to 24.47 for Cu and from 1.04 to 8.43 for
coefficients: 0.628, p < 0.001). Pb. These data indicate that the Cu and Pb pollution
All results listed above made the basis of discussion is widespread in the urban parks of Beijing. Further-
later. more, there were 16 parks with high PIs for Cu and nine
parks with high PIs for Pb, which accounted for about
one-half and one-third of all of the parks investigated
3.4. Assessment of the environmental quality for the in this study, respectively. The maximal PIs were 24.47
soils of urban parks for Cu and 8.43 for Pb. Thus, it is very likely that many
of the urban parks in Beijing are highly polluted with Cu
The PIs, calculated according to the BC of heavy met- and Pb.
als, varied greatly across the different metals (Table 5). The IPIs of all parks varied from 0.97 to 9.21 with an
Nickel and Zn exhibited lower values, ranging from average of 2.21 (Table 6). There was only one park with
0.23 to 1.39 and from 0.45 to 3.42, respectively. For Ni, an IPI < 1, 15 parks with an IPI between 1 and 2, and 14
the mean PI was 0.83 and all of the samples had low or parks with an IPI > 2. In particular, the GUG demon-
mid-level PIs, indicating that the concentration of Ni in strated an IPI of 9.21, indicating the presence of serious
the soil samples were comparable with the BC of Beijing heavy metal pollution. Approximately one-half of the
and there was no obvious pollution of Ni in the park soil parks investigated had high pollution levels. Thus the
samples. The mean PI for Zn was 1.52, but only one sam- soil quality of BeijingÕs urban parks has clearly been im-
ple was classified as high PI, indicating an absence of pacted, particularly those parks visited by large crowds
problematic Zn pollution of soils in Beijing parks. of people, such as GUG, BHP and YYH.
548 T.-B. Chen et al. / Chemosphere 60 (2005) 542–551

Table 6
Integrated pollution index (IPI) of heavy metals in the urban
park soils of Beijing
High (IPI > 2.0) Middle Low
(1.0 < IPI 6 2.0) (IPI 6 1.0)
Park IPI Park IPI Park IPI
GUG 9.21 QNH 1.99 SJP 0.97
BHP 3.54 BJZ 1.85
RTP 3.09 SXP 1.73
GYP 2.76 FTP 1.67
JSP 2.71 WFT 1.67
YHY 2.69 LTH 1.66
DTP 2.44 YMY 1.65 Fig. 3. Mean concentrations of heavy metals in the soils:
ZSP 2.39 CYP 1.55 classified in terms of the location of parks inside or outside the
DGY 2.29 HLJ 1.53 HCD. Concentrations of Ni and Zn are arithmetic mean values
XWP 2.24 MZP 1.52 within the normal distribution. Others are geometric mean
TTP 2.18 RDH 1.50 values within the lognormal distribution.
YTP 2.18 TRT 1.47
BJA 2.11 TJH 1.40 Table 7
LYP 2.01 YYT 1.37 Numbers of the parks attributed with high PIs or IPIs classified
ZZY 1.03 with location and history
Numbers Location History
4. Discussion Inside Outside <100 P100
the HCD the HCD yrs yrs
In this study, the assessment of the soil environment Total numbers 18 12 18 12
quality suggests that heavy metal pollution in the park No. of high PIs 14 2 5 11
soils may be related to the location and the history of for Cu
the parks. Different locations are associated with differ- No. of high PIs 7 2 2 7
ences in the density of traffic and human activities. The for Pb
age of the parks affected the heavy metal deposition No. of high IPIs 12 2 4 10
found in the soils samples. It is also true that the older
parks are more well-known and attract greater numbers
of tourists, which is obviously accompanied by a greater tant factors in determining the accumulation of heavy
amount of traffic. metals in urban soils (Albasel and Cottenie, 1985). The
research performed in Seville, Spain (Madrid et al.,
4.1. Location of the parks and the heavy metal 2002) indicated that the concentrations of Pb, Zn and,
concentrations particularly, Cu in the soils from major parks and green
areas closer to the historic center often exceeded the
According to the location, all parks investigated were acceptable limits for residential, recreational and institu-
classified into two groups: the parks inside the HCD tional sites. The results of another study found that the
(group I) or the ones outside the HCD (group II). The soils from the center of Rostock, Germany had higher
mean values of the four heavy metals in the parks of metal contents than did the soils from the outskirts of
group I were all much greater than those in the parks the town (Kahle, 2000). Therefore, the higher concentra-
of group II (Fig. 3). The Cu concentration for group I tions of heavy metals in the HCD of Beijing may be
was significantly greater than that for group II. In addi- interpreted as being directly related to heavy human
tion, 14 of 16 parks with high PIs for Cu were inside the activities, such as traffic. The location of a park is clearly
HCD, while seven of nine parks with high PIs for Pb one important factor determining the extent of heavy
were inside the HCD (Table 7). Of the14 parks with high metal accumulation, particularly of Cu and Pb, in urban
IPIs, only two were outside the HCD. Therefore, in this park soil.
study park location was related to the amount of heavy Based on location, parks with similar development
metal accumulation found in the soil samples, particu- history (<50 years) were classified into two groups:
larly for Cu and Pb. parks inside or outside the HCD. Comparison of the
The HCD of Beijing is the oldest urban region of the means showed that the parks inside the HCD had mark-
city. The traffic intensity in HCD is 4.1 times greater edly higher Cu concentrations than those outside the
than that in the districts outside the HCD (Jin, 2000). HCD. Large differences in heavy metal concentrations
Commercial and traffic activities are known to be impor- in the soils, especially of Cu and Pb, were seen among
T.-B. Chen et al. / Chemosphere 60 (2005) 542–551 549

the soil samples from parks at different locations with a 4.3. Analysis of the pollutant sources
similar development history. It is important to note that
TRT and BJA, which are located in regions with a The concentrations of Cu and Pb in the soils samples
shorter urbanizing history (at the southern HCD border from most of the urban parks of Beijing exceeded the
and near the Fengtai district), had lower heavy metal corresponding BC and, indeed, serious soil pollution
concentrations than did other parks in the HCD. was found in some parks. Anthropogenic sources may
contribute to this elevation of Cu and Pb concentrations
4.2. History of the parks and the extent of pollution in the soils. In fact, Cu and Pb soil pollution appears to
be readily affected by anthropogenic factors (Artaxo
Based on the history of their establishment, the parks et al., 1999; Martin, 2001; Gray et al., 2003).
were classified into two groups: the parks with a history Atmospheric deposition of heavy metals, including
< 100 years (class A) and the parks with a history > 100 Cu and Pb, is considered to be a significant factor in soil
years (class B). As shown in Fig. 4, the mean concentra- pollution (Martin, 2001; Nicholson et al., 2003). In
tions of Cu and Pb in class B were approximately 1.5 woodlands this pollution is increased by 33–259%,
and 2 times as high as those in class A, respectively. depending on the metal. Lead deposition in particular
There were significant differences between two classes. is greatly influenced by vehicle emissions and the intro-
In addition, among the 12 parks in class B, 10 had high duction of Pb into gasoline (Blake and Goulding,
IPIs overall, 11 parks had high PIs for Cu, and 7 had 2002). This study focused on a region with a high density
high PIs for Pb (Table 7). These findings indicate that of traffic and human activities, associated with higher
most of the parks in class B have been polluted with levels of airborne heavy metals. This in turn may affect
heavy metals, particularly with Cu and Pb. Further- the soil environmental quality via subsequent atmo-
more, the parks in class B accounted for the majority spheric deposition. Although petrol with lead additives
of the parks with high PIs or IPIs overall. There were has been banned in Beijing for several years, lead pollu-
also significant correlations between the history of the tion may continue to affect the soil environment for
parks and the concentrations of Cu and Pb. years to come.
Vehicle emissions and other human activities may The normal activity and deterioration of vehicles on
contribute to the accumulation of heavy metals, such the roads can emit heavy metals into the air, especially
as Cu and Pb, in soils (Pfeiffer et al., 1991; Blake and Cu (Ritter and Rinefierd, 1983; Martin et al., 1998).
Goulding, 2002; Zheng et al., 2002). In this study, the Thus, the various traffic densities can influence the
older and more tourist-appealing parks appear to suffer amount of Cu emitted by local vehicles. Application of
the highest amount of human traffic, which in turn may Pb- and Cu-containing house paints may also contribute
elevate Cu and Pb pollution levels of soils in the parks. to Pb and Cu pollution levels (Alloway and Ayres, 1997;
Other reports have found that the extent of pollution Wells et al., 2000). The age of the dwelling is also impor-
varied with age and was more serious in the historic area tant in determining the heavy metal concentrations in
of the city (Pfeiffer et al., 1991; Madrid et al., 2002). the surroundings (Tong, 1998; Rieuwerts et al., 1999).
Taken with our findings, it is evident that the history Therefore, the high concentrations of Cu and Pb in the
of the parks should be another factor related to the in- older parks might be related to the discharge of heavy
crease of the heavy metal concentrations in the soils. metals contained in the paints in the region.
In general, influences between air and soil pollution
are mutual. Just as the atmosphere can transfer a large
amount of heavy metals into urban soils through precipi-
tation (Carey et al., 1980; Ritter and Rinefierd, 1983; Pa-
tel et al., 2001), soil dust can also contribute to the
concentrations of heavy metals in the air (Chen et al.,
1997). Consequently, airborne particles and soil dust
containing elevated heavy metal concentrations may en-
ter and harm the human body through breathing and
ingestion (Chen et al., 1997). Substantial evidence indi-
cates that blood Pb concentration is correlated with that
in oneÕs environment (McMichael et al., 1985; Bellinger
et al., 1990).
Fig. 4. Mean concentrations of heavy metals in the soils: In the present study, soil concentration of Zn was
classified in terms of the history of the parks. Concentration of also found to be correlated with the level of human
Ni is an arithmetic mean value within the normal distribution. activities (Artaxo et al., 1999; Cyrys et al., 2003; Gray
Others are geometric mean values within the lognormal et al., 2003). Although the Zn concentration in the urban
distribution. park soils may be influenced by anthropogenic activities,
550 T.-B. Chen et al. / Chemosphere 60 (2005) 542–551

it appears to be largely related to the parent materials of Alloway, B.J., Ayres, D.C., 1997. In: Chemical Principles of
the soils since Zn was distributed in F1 and F2 in the re- Environmental Pollution. Blackie Academic and Profes-
sults of PCA. Zinc accumulation in park soil was rela- sional, London, p. 17.
tively low compared to that of the soil background Zn Artaxo, P., Oyola, P., Martinez, R., 1999. Aerosol composition
and source apportionment in Santiago de Chile. Nuclear
concentration in Beijing soils. With respect of the pollu-
Instruments and Methods B 150, 409–416.
tion levels of Cu and Pb, the pollution extent of Zn was Bandhu, H.K., Sanjiv, P., Garg, M.L., Singh, B., Shahi, J.S.,
much lower. Mehta, D., Swietlicki, E., Dhawan, D.K., Mangal, P.C.,
We did not find Ni pollution in the soil samples of the Singh, N., 2000. Elemental composition and sources of air
urban parks. The soil concentration of Ni is mainly pollution in the city of Chandigarh, India, using EDXRF
attributable to the parent materials in the soil (Pierce and PIXE techniques. Nuclear Instruments and Methods B
et al., 1982; Chen et al., 1999; Zupan et al., 2000). In this 160, 126–138.
study, the Ni concentrations in the park soil samples Beijing Municipal Statistical Bureau (compiled), 2001. Beijing
were comparable with the background Ni concentration Statistical Yearbook. China Statistics Press, Beijing,
in Beijing soils. p. 72.
Bellinger, D., Leviton, A., Slowman, J., 1990. Antecedents and
correlates of improved cognitive performance in children
exposed in utero to low levels of lead. Environmental Health
5. Conclusion
Perspectives 89, 5–11.
Bilos, C., Colombo, J.C., Skorupka, C.N., Presa, M.J.P., 2001.
This investigation of soil samples from 30 urban Source, distribution and variability of airborne trace metals
parks in Beijing revealed a clear accumulation of Cu in La Plata City area, Argentina. Environmental Pollution
and Pb. All of the parks had middle or high PIs of Cu 111, 149–158.
and Pb, indicative of pollution from the two metals, par- Blake, L., Goulding, K.W.T., 2002. Effects of atmospheric
ticularly in parks located inside the historic central areas deposition, soil pH and acidification on heavy metal
and parks with hundreds of years of history, such as concentrations in soils and vegetation of semi-natural
GUG, BHP, RTP, YHY, and JSP. Therefore, park loca- ecosystems at Rothamsted Experimental Station, UK. Plant
and Soil 240, 235–251.
tion and age were important factors related to the accu-
Carey, A.E., Gowen, J.A., Forehand, T.J., Tai, H., Wie-
mulation of heavy metals, particularly of Cu and Pb, in
rsma, G.B., 1980. Heavy metal concentrations in soils of
the soils. Zinc concentration in the soil samples exceeded five United States Cities, 1972 Urban Soils Monitoring
background concentrations, but the soil pollution of Zn Program. Pesticides Monitoring Journal 13, 150–154.
was low. The Ni concentrations in the urban park soils Chen, T.B., Wong, W.J.C., Zhou, H.Y., Wong, M.H., 1997.
samples were similar to the soil background concentra- Assessment of trace metal distribution and contamination in
tions of Beijing, thus a Ni pollution problem was not surface soil of Hong Kong. Environmental Pollution 96,
evident. All but one of the 30 parks had middle or high 61–68.
IPIs. Therefore, the soil quality in most of the urban Chen, T.B., Zheng, Y.M., Chen, H., Zheng, G.D., 2004.
parks in Beijing has deteriorated. These findings indicate Background concentrations of soil heavy metals in Beijing.
Chinese Journal of Environmental Science 25, 117–122 (in
that more attention should be paid to heavy metal pol-
lution of the park soils in Beijing.
Chen, M., Ma, L.Q., Harris, W.G., 1999. Baseline concentra-
tions of 15 trace elements in Florida surface soils. Journal of
Environmental Quality 28, 1173–1181.
Acknowledgment China National Environmental Monitoring Center (CNEMC),
1990. The Background Concentrations of Soil Elements in
The research was supported by National Science China. Chinese Environment Science Press, Beijing, p. 346
Fund for Distinguished Young Scholars (Grant No. (in Chinese).
40325003), the Natural Science Foundation of Beijing Chon, H.T., Ahn, J.S.G., Jung, M.C., 1995. Metal contamina-
(Grant No. 6990002) and the Scientific Innovation Pro- tion of soils and dusts in Seoul metropolitan city, Korea.
gram of Chinese Academy of Sciences (Grant No. Environmental Geochemistry and Health 17, 23–37.
KZCX2-04-01). Opinions in the paper do not constitute Cyrys, J., Stolzel, M., Heinrich, J., Kreyling, W.G., Menzel, N.,
Wittmaack, K., Tuch, T., Wichmann, H.E., 2003. Elemental
an endorsement or approval by the funding agencies and
composition and sources of fine and ultrafine ambient
only reflect the personal views of the authors. particles in Erfurt, Germany. The Science of the Total
Environment 305, 143–156.
De Kimple, C.R., Morel, J.F., 2000. Urban soil management: a
References growing concern. Soil Science 165, 31–40.
De Miguel, E., Llamas, J.F., Chacon, E., Berg, T., Larssen, S.,
Albasel, N., Cottenie, A., 1985. Heavy metal contamination Royset, O., Vadset, M., 1997. Origin and patterns of
near major highways, industrial and urban areas in Belgian distribution of trace elements in street dust: unleaded petrol
grassland. Water Air and Soil Pollution 24, 103–109. and urban lead. Atmospheric Environment 31, 2733–2740.
T.-B. Chen et al. / Chemosphere 60 (2005) 542–551 551

Gray, C.W., McLaren, R.G., Roberts, A.H.C., 2003. Atmospheric Patel, K.S., Shukla, A., Tripathi, A.N., Hoffmann, P., 2001.
accessions of heavy metals to some New Zealand pastoral soils. Heavy metal concentrations of precipitation in east Madhya
The Science of the Total Environment 305, 105–115. Pradesh of India. Water Air and Soil Pollution 130, 463–
Ikeda, A., Yoda, K., 1982. Soil pollution by heavy metals in 468.
Sakai City. Japanese Journal of Ecology 32, 241–249. Pfeiffer, E.M., Freytag, J., Scharpenseel, H.W., 1991. Investi-
Jin, D.X., 2000. Methods to resolve the problem of traffic in the gation of heavy metal and arsenic pollution of soils and
city zone of Beijing. Beijing City Planning & Construction 1, plants in the urban area of Manila, Philippines. Mitteilun-
31–34 (in Chinese). gen der Deutschen Bodenkundlichen Gesellschaft 66, 1169–
Kahle, P., 2000. Heavy metals in Rostock garden soils. Journal 1172.
of Plant Nutrition and Soil Science 163, 191–196. Pierce, F.J., Dowdy, R.H., Grigal, D.F., 1982. Concentrations
Kelly, J., Thornton, I., Simpson, P.R., 1996. Urban geochem- of six trace metals in some major Minnesota soil series.
istry: A study of the influence of anthropogenic activity on Journal of Environmental Quality 11, 412–416.
the heavy metal content of soils in traditionally industrial Rasmussen, P.E., Subramanian, S.K., Jessiman, B.J., 2001. A
and non-industrial areas of Britain. Applied Geochemistry multi-element profile of housedust in relation to exterior
11, 363–370. dust and soils in the city of Ottawa, Canada. The Science of
Komai, Y., 1981. Heavy metal contamination in urban soils: I. the Total Environment 267, 125–140.
Zinc accumulation phenomenon in urban environments as Rieuwerts, J.S., Farago, M., Bencko, V., Cikrt, M., 1999.
clues of study. Bulletin of the University of Osaka Prefec- Heavy metal concentrations in and around households near
ture 33, 7–15. a secondary lead smelter. Environmental Monitoring and
Li, X.D., Poon, C.S., Pui, S.L., 2001. Heavy metal contamina- Assessment 58, 317–335.
tion of urban soils and street dusts in Hong Kong. Applied Ritter, C.J., Rinefierd, S.M., 1983. Natural background and
Geochemistry 16, 1361–1368. pollution levels of some heavy metals in soils from the area
Madrid, L., Diaz-Barrientos, E., Madrid, F., 2002. Distribution of Dayton, OH. Environmental Geology 5, 73–78.
of heavy metal contents of urban soils in parks of Seville. Sakagami, K.I., Eamada, R., Kurobe, T., 1982. Heavy metal
Chemosphere 49, 1301–1308. contents in dust fall and soil of the National Park for
Manta, D.S., Angelone, M., Bellanca, A., Neri, R., Sprovieri, Nature Study in Tokyo. Mitteilungen der Deutschen
M., 2002. Heavy metals in urban soils: a case study from the Bodenkundlichen Gessellscaft 33, 59–66.
city of Palermo (Sicily), Italy. The Science of the Total Tong, S.T.Y., 1998. Indoor and outdoor household dust
Environment 300, 229–243. Contamination in Cincinnati, OH, USA. Environmental
Martin, A.C., Rivero, V.C., Marin, M.T.L., 1998. Contamina- Geochemistry and Health 20, 123–133.
tion by heavy metals in soils in the neighborhood of a United States Environmental Protection Agency (USEPA),
scrapyard of discarded vehicles. The Science of the Total 1996. Method 3050B: Acid digestion of sediments, sludges
Environment 212, 142–152. and soils (revision 2).
Martin, L., 2001. Urban land use influences on heavy metal Wells, E.C., Terry, R.E., Parnell, J.J., Hardin, P.J., Jackson,
fluxes and surface sediment concentrations of small lakes. M.W., Houston, S.D., 2000. Chemical analyses of ancient
Water Air and Soil Pollution 126, 363–383. anthrosols in residential areas at Piedras Negras,
McMichael, A.J., Baghurst, P.A., Robertson, E.F., Vimpani, Guatemala. Journal of Archaeological Science 27, 449–
G.V., Wigg, N.R., 1985. The port Pirie study: blood lead 462.
concentrations in early childhood. Medical Journal of Wong, J.W.C., Mak, N.K., 1997. Heavy metal pollution in
Australia 143, 499–503. children playgrounds in Hong Kong and its health impli-
Meyer, I., Heinrich, J., Lippold, U., 1999. Factors affecting cations. Environmental Technology 18, 109–115.
lead, cadmium and arsenic levels in housedust in a smelter Zheng, Y.M., Chen, T.B., Wu, H.T., Chen, H., Zhou, J.L.,
town in Eastern German. Environment Research 81, 32–44. Luo, J.F., Huang, Z.C., 2003. The spatial structure and
Mielke, H.W., Gonzales, C.R., Smith, M.K., Mielke, P.W., contents distribution of Ni in soils of Beijing suburb. Acta
1999. The urban environment and childrenÕs health: soils as Geographica Sinica 58, 470–476 (in Chinese).
an integrator of lead, zinc and cadmium in New Orleans, Zheng, Y.M., Yu, K., Wu, H.T., Huang, Z.C., Chen, H., Wu,
Louisiana, USA. Environment Research 81, 117–129. X., Tian, Q.Z., Fan, K.K., Chen, T.B., 2002. Lead
Mielke, H.W., Reagan, P.L., 1998. Soil is an important concentrations of soils in Beijing urban parks and their
pathway of human lead exposure. Environmental Health pollution assessment. Geographical Research 21, 418–424
Perspectives 106, 217–229. (in Chinese).
Nicholson, F.A., Smith, S.R., Alloway, B.J., Carlton-Smith, C., Zupan, M., Einax, J.W., Kraft, J., Lobnik, F., Hudnik, V.,
Chambers, B.J., 2003. An inventory of heavy metals inputs 2000. Chemometric characterization of soil and plant
to agricultural soils in England and Wales. The Science of pollution. Part 1: multivariate data analysis and geostatis-
the Total Environment 311, 205–219. tical determination of relationship and spatial structure of
Nriagu, J.O., 1989. A global assessment of natural sources of inorganic contaminants in soil. Environmental Science and
atmospheric trace metals. Nature 338, 47–49. Pollution Research 7, 89–96.