Geospatial mapping, source identification and human health risk assessment of heavy metals in soils of Gyumri (Armenia)

Abstract

Gyumri was destroyed by a devastating earthquake in 1988. Today the city is in the reconstruction stage, and noticeable traces of earthquake are a significant pollution source of the urban environment by heavy metals (HM). To assess HM pollution levels, to identify their possible sources and evaluate the potential impact to human health soils survey of Gyumri was done. Totally, 443 soil samples were collected, and the contents Fe, Ti, Mn, Co, Cu, As, Zn, Hg, Pb, Cd, Ag, Ba, and Mo have been determined by X-ray fluorescence spectrometry (Olympus Innov-X-5000 (USA)). Geospatial mapping and multivariate geostatistical analysis showed that there exist a significant spatial correlation between pollution sources and hot spots of studied elements. According to the Principal component analysis, four groups were generated explaining 73.4% of the total variance. PC1 including Cu, Zn, Ba, Pb and Mo, PC3: Ag and Cd, and PC4: Cu, Co and As were identified as an anthropogenic group. Risk assessment showed that observed contents of HM pose a non-carcinogenic risk to children health. The riskiest element was Pb which HI>1 in 1.6% of city territory. The results of this study highlight the need for further medico-ecological investigations and development of risk reduction measures.

Full text

15th International Conference on Environmental Science and Technology
Rhodes, Greece, 31 August to 2 September 2017
CEST2017_00489
Geospatial mapping, source identification and human health
risk assessment of heavy metals in soils of Gyumri (Armenia)
Tepanosyan G.1,*, Sahakyan L.1, Kafyan M.1 And Saghatelyan A.1
1Department of Environmental Geochemistry, The Center for Ecological-Noosphere Studies of the National Academy of
Sciences, Yerevan 0025, Abovian-68, Republic of Armenia
*corresponding author:
e-mail: gevorg.tepanosyan@cens.am
Abstract
Gyumri was destroyed by a devastating earthquake in
1988. Today the city is in the reconstruction stage, and
noticeable traces of earthquake are a significant pollution
source of the urban environment by heavy metals (HM).
To assess HM pollution levels, to identify their possible
sources and evaluate the potential impact to human health
soils survey of Gyumri was done. Totally, 443 soil samples
were collected, and the contents Fe, Ti, Mn, Co, Cu, As,
Zn, Hg, Pb, Cd, Ag, Ba, and Mo have been determined by
X-ray fluorescence spectrometry (Olympus Innov-X-5000
(USA)). Geospatial mapping and multivariate geostatistical
analysis showed that there exist a significant spatial
correlation between pollution sources and hot spots of
studied elements. According to the Principal component
analysis, four groups were generated explaining 73.4% of
the total variance. PC1 including Cu, Zn, Ba, Pb and Mo,
PC3: Ag and Cd, and PC4: Cu, Co and As were identified
as an anthropogenic group. Risk assessment showed that
observed contents of HM pose a non-carcinogenic risk to
children health. The riskiest element was Pb which HI>1 in
1.6% of city territory. The results of this study highlight
the need for further medico-ecological investigations and
development of risk reduction measures.
Keywords: Heavy metals, urban soils, pollution, mapping,
geostatistical analysis
1. Introduction
In urban areas, human activities change the soil
environment and their properties (Johnson et al., 2011).
Acting as a geochemical sink for various hazardous
contaminants urban soils can impact the health of the urban
population (Filippelli et al., 2012). Among soil
contaminants, particular attention is given to heavy metals
(HM) (Gu et al., 2016; Peña-Fernández et al., 2014; Wong
et al., 2006) which are known to cause different disorders
when entering into the human organism (US EPA, 1989).
For this reason human health risk assessment was done in
different parts of the world (Gu et al., 2016; Peña-
Fernández et al., 2014) and the results obtained (Lee et al.,
2006; Lu et al., 2011; Rapant et al., 2010) indicated that
the model proposed by US EPA (Pepper et al., 2011; US
EPA, 1989) is applicable for the soils. Gyumri -
Armenia’s second biggest city, destroyed by the Spitak
Earthquake in 1988. After that, during last 25 years the
remnants of the disaster (collapsed buildings, debris and
other garbage dumps, etc.) and post-earthquake industrial
and social activities, as well as significantly mixed
anthropogenic (asphalt, concrete, gravel layers, etc.) and
natural horizons completely change the environmental
status of the city. Therefore, Gyumri soils survey was done
(2013) and the goal of this research was to assess HM
pollution levels, to identify their possible sources and
evaluate the potential impact to human health.
2. Methods and materials
2.1. Sampling and chemical analysis
Pedogeochemical survey of Gyumri was done on a scale
1:25000 (16 samples per/sq.km) in August 2013. Totally,
443 soil samples collected (Fig.1). To establish Gyumri
soils HM local background (LB) 33 rural soils samples
from the adjacent villages were collected. A bulk sample
was generated by mixing of 3-5 randomly collected
subsamples. In the laboratory, samples were air-dried,
homogenized and sieved (2 mm), milled according to ISO-
11464 (BSI Standards Limited., 2006). The total contents
of 13 elements (Ti, Mn, Fe, Co, Cu, Zn, As, Mo, Ba, Hg,
Pb, Ag, Cd) determined by XRF spectrometry (US EPA
Method 6200, 2007). To ensure the quality assurance and
quality control of the analysis, standard samples (NIST
2710a and NIST 2711a, USA), a blank (SiO2) of NIST
USA and 5.2% lab duplicate samples were analyzed.
Precision for studied HM ranges 0.9-17.4%, and accuracy
was < 20%. locations of industrial units in the city of
Gyumri.
2.2. Statistical analysis and geospatial mapping
Descriptive statistics of datasets are summarized in Table
2. Principal Component Analysis (PCA) was performed to
group studied HM and identifies their potential sources.
Geospatial mapping was done to illustrate the spatial
distribution of HM pollution levels, to juxtaposition them
with the location of possible sources and to identify riskiest
areas. Colour surface maps were created by IDW method
(the power - 2, the number of neighboring samples - 8)
using ArcGIS 10.1.
2.3. Pollution levels and human health risk assessment

2
CEST2017_00489
To study city soils poly-elemental pollution levels the
summary pollution index (Zc) (Johnson et al., 2011;
Perelman and Kasimov, 1999) calculated using the
following formulae:
 , (1),
   

 , (2),
where Kc is anomaly coefficient, Ci is the concentration of
studied HM in the sample, Cb is the LB values, n is the
number of elements with Kc>1. Zc was classified as low
level (Zc<16), mean i.e. moderately hazardous level
(16<Zc<32), high i.e. hazardous level (32<Zc<128), and
very high i.e. extremely hazardous level (Zc>128)
(Perelman and Kasimov, 1999).
Figure 1. Spatial distribution of soil sampling points and
Non-carcinogenic risk assessment was done, and as a
preferential exposure pathway of HM for humans, soil
ingestion was chosen. In this study non-carcinogenic risk
was assessed for Ag, As, Ba, Cd, Co, Cu, Fe, Hg, Mn, Mo,
Pb and Zn due to the lack of quantitative information
concerning elements toxicity level and impact to the
human health. Non-carcinogenic risk (RAIS, 2017)
calculated by the following formula:
 
  (3)
where: C is the element concentration in soil (mg/kg), IRS
is ingestion rate: in the case of children 200 mg/day, for
adults 100 mg/day; ED is exposure duration: for children 6
years and for adults 26 years; EF is exposure frequency:
350 days/year (RAIS, 2017), BW is average body weight:
15 kg for children and 70 kg for adults (US EPA, 1989).
Non-carcinogenic hazard quotient per element was
calculated by the formula (3).
   
 (4)
where: RfD is a reference dose for a corresponding
element (Table 1). The sum of HQ values represents a
Hazard Index HI=ΣHQi. HI<1 indicates the absence of
significant non-carcinogenic health risk, whereas HI>1
denotes a possibility of adverse health effects (RAIS, 2017;
US EPA, 1989).
Table 1. Defined oral reference doses (RfD, mg/kg-1 day-1)
for evaluated risk elements
Heavy metals
RfD
Data source
As
0.0003
(RAIS, 2017)
Cr
0.003
Hg
0.00016
Pb
0.0035
(WHO, 2008)
Cu
0.04
(RAIS, 2017)
Zn
0.3
Mo
0.005
Ni
0.02
Co
0.0003
Mn
0.024
Ba
0.2
V
0.00504
3. Results and discussion
3.1. Heavy metals contents
Descriptive statistics of the contents of HM (Ti, Mn, Fe,
Co, Cu, Zn, As, Mo, Ba, Hg, Pb, Ag, and Cd) in Gyumri
soils, LB values and Predicted Empirical Global Soil
(PEGS2) reference values (De Caritat et al., 2012) are
given in Table 2. For all HM (Table 2) skewness is
different from 0, indicating nonnormality of the data
distribution. Moreover, the comparison of mean and 5%
trimmed mean showing the presence of outliers and
extreme values in the case of Cu, Zn, Pb, Ag, and Cd.
Besides Cu, all these elements showed a significant value
of CV: 119.1%, 180.4%, 222.1% and 119.8% for Zn, Pb,
Ag and Cd, respectively. Excesses of mean contents vs.
PEGS2 and vs. LB observed for Cu, Zn, Pb, Ag and Cd
4.1, 4.4, 4.8, 5.4 և 6.1 times, and 1.4, 2.1, 3.5, 1.9 and 6.0
time, correspondingly. The latter inferred the
anthropogenic origin of these HM. The mean values of Ti,
Mn, Fe, Co, As, Hg and Ba vs. LB ranges 0.8-1.1,
suggesting that the differences in comparison with PEGS2
values of Mn, Fe, Co, As, Hg, Ba, and Mo can be the
outcome of local peculiarities of origin and spatial
distribution of these elements.
3.2. Principal component analysis of heavy metals in
Gyumri soils
According to the PCA results first four principal
components (PC) showed >1 eigenvalues (PC1-3.4; PC2-
3.3; PC3-1.7 and PC4-1.1) and explained 73.4 % of the
total variance.PC1 (26.5% of the total variance) showed
strong (>0.7) positive loading for Zn, Pb, Ba, moderate
(>0.5) positive loading for Cu, Mo. Negative moderate
loading of Hg in PC1 inferred other sources of origin that .

CEST2017_00489
Table 2. Descriptive statistics, local background (LB) and Predicted Empirical Global Soil (PEGS2) reference values of
studied heavy metals (mg/kg)
Elements
Parameters
Mean
5%
Trimmed
Mean
Median
SD
Min
Max
Skew
CV,
%
PEGS2
LB
Ti
3011.8
2983.1
2957.0
511.6
1352.0
9117.0
4.2
17.0
-
3381
Mn
640.8
635.6
626.0
101.6
335.0
1068.0
0.8
15.9
370.0
805.5
Fe
31439.1
31353.8
30916.0
3956.6
13118.0
45881.0
0.3
12.6
17066.0
32379
Co
5.1
5.1
5.03
0.73
1.6
8.3
0.6
14.3
9.0
5
Cu
53.7
50.1
47.9
26.8
25.1
289.0
4.7
49.9
13.0
39
Zn
207.7
180.0
162.4
247.3
42.8
4071.0
9.9
119.1
47.0
100
As
0.75
0.74
0.75
0.20
0.2
3.5
7.8
27.2
5.0
0.68
Ba
519.6
505.1
505.0
170.5
214.0
3077.0
11.1
32.8
353.0
526.5
Pb
80.9
60.2
50.8
145.9
5.4
1714.0
7.9
180.4
17.0
23.2
Ag
0.14
0.10
0.10
0.30
0.0002
3.3
8.4
222.1
0.025
0.07
Cd
0.67
0.57
0.41
0.80
0.002
7.0
3.1
119.8
0.11
0.11
Mo
0.47
0.45
0.47
0.27
0.08
3.2
4.7
58.1
0.31
2
Hg
0.07
0.07
0.07
0.04
0.01
0.38
2.1
52.3
-
0.07
those in the case of Zn, Pb, Ba, Cu and Mo. The
visualization of PC scores (Fig. 2) showed that high values
(>0.5) of PC1 are spatially allocated near industrial units
and in densely populated part of the city suggesting that
PC1 is an anthropogenic group. The latter is also
confirmed by high values CV of some elements included in
PC1. For Pb, Zn and Cu, similar results were observed in
the case of Yerevan (Tepanosyan et al., 2016) and other
studies worldwide (Chabukdhara and Nema, 2013; Guo et
al., 2012; Sun et al., 2010; Yuan et al., 2014).PC2 (25.1%
of total variance) showed strong positive loadings for Fe,
Ti, Mn, and Co indicating they may have a natural origin,
which is also confirmed by low values of CV. However,
Co with Cu (moderate positive loading) and As (strong
positive loading) formed also PC4 (8.5% of total variance).
The latter can be explained by the presence of Gyumri
bicycle plant which operated before the earthquake. This
plant is known by its workshops of metal plating and paint-
and-lacquer coating, assemblage and woodworking,
pressing which were potential sources of Co, Cu and As.
Thus, although the contents of Co in city soils have mainly
natural origin, there still exist some local sites where the
fingerprint of historical pollution observed. PC3 (13.3% of
total variance) showed strong positive loadings for Ag and
Cd. The detailed inspection of the PC3 map (Fig. 2)
revealed that significant areas of high scores are spatially
associated with industrial units located in the eastern and
south parts of the city. Th latter suggests that PC3 is an
anthropogenic group which confirmed by the high values
of CV of Ag and Cd.
3.3. Pollution levels and human health risk
A low level (Zc<16) of summary pollution index (Fig. 3)
was observed in 75.2% of the city (33.4sq.km, 317 soil
samples). Мean pollution level covers 22.3% (9.9sq.km,
102 soil samples) of the area and is spatially distributed in
the central part of the city. High level of pollution have a
local character, point-like shape and are surrounded by
mean pollution level. High level occupies only 2.5%
(1.1.sq.km. 24 soil samples) of the city area. Extremely
high pollution level has not been found. It should be
mentioned that mean and high levels poly-element
pollution are found mainly near industrial units and
densely populated parts of Gyumri. Human health risk
assessment showed that in the case of adults HI<1.
However, children’s health non-carcinogenic risk (Fig. 3)
observed in the whole area of the city indicating adverse
health effect for children. Mean HQ values of studied HM
decrease in the order of
Fe>Mn>Pb>Co>Cu>Ba>As>Cd>Hg>Mo>Ag>Zn. The
HQ>1 observed only for Cu (1 sampling site) and Pb (17
sampling site).
4. Conclusions
The study revealed that the excesses of mean vs. LB and
vs. PEGS2 observed for Cu, Zn, Pb, Ag and Cd ranges1.4-
6.0, and 4.1-6.1, respectively, indicating that the
concentrations of these HM influenced by anthropogenic
and industrial activities. Also, based on the PCA results,
three groups of HM identified as having an anthropogenic
origin. According to the Zc values, low level of pollution
observed in 75.2% of Gyumri area. The areas of medium
and high pollution levels lie near industrial units and in
densely populated parts of the city. Human health risk
assessment revealed that in whole city HM in soils serves
as risk factors for children’s health. The riskiest element
was Pb. Therefore, more detailed investigations and risk
reduction measures needed.

CEST2017_00489
Figure 2. Spatial distribution of PC scores in the city of Gyumri
Figure 3. Spatial distribution of soils HM pollution and health risk levels in the city of Gyumri
References
BSI Standards Limited., 2006. BS ISO 11464:2006 Soil quality.
Pretreatment of samples for physio-chemical analysis.
Chabukdhara, M., Nema, A.K., 2013. Heavy metals assessment
in urban soil around industrial clusters in Ghaziabad, India:
Probabilistic health risk approach. Ecotoxicol. Environ. Saf.
87, 57–64. doi:10.1016/j.ecoenv.2012.08.032

CEST2017_00489
De Caritat, P., Reimann, C., Bastrakov, E., NGSA Project Team
and GEMAS Project Team, 2012.. Comparing results from
two continental geochemical surveys to world soil
composition and deriving Predicted Empirical Global Soil
(PEGS2) reference values. Earth Planet. Sci. Lett. 319–320,
269–276. doi:10.1016/j.epsl.2011.12.033
Filippelli, G.M., Morrison, D., Cicchella, D., 2012. Urban
Geochemistry and Human Health. Elements 8, 439–444.
doi:10.2113/gselements.8.6.439
Gu, Y.-G., Gao, Y.-P., Lin, Q., 2016. Contamination,
bioaccessibility and human health risk of heavy metals in
exposed-lawn soils from 28 urban parks in southern China’s
largest city, Guangzhou. Appl. Geochemistry 67, 52–58.
doi:10.1016/j.apgeochem.2016.02.004
Guo, G., Wu, F., Xie, F., Zhang, R., 2012. Spatial distribution
and pollution assessment of heavy metals in urban soils from
southwest China. J. Environ. Sci. 24, 410–418.
doi:10.1016/S1001-0742(11)60762-6
Johnson, C.C., Demetriades, A., Locutura, J., Ottesen, R.T.,
2011. Mapping the Chemical Environment of Urban Areas.
Lee, S.-W., Lee, B.-T., Kim, J.-Y., Kim, K.-W., Lee, J.-S., 2006.
Human risk assessment for heavy metals and as
contamination in the abandoned metal mine areas, Korea.
Environ. Monit. Assess. 119, 233–44. doi:10.1007/s10661-
005-9024-5
Lu, Y., Yin, W., Huang, L., Zhang, G., Zhao, Y., 2011.
Assessment of bioaccessibility and exposure risk of arsenic
and lead in urban soils of Guangzhou City, China. Environ.
Geochem. Health 33, 93–102. doi:10.1007/s10653-010-9324-
8
Peña-Fernández, a., González-Muñoz, M.J., Lobo-Bedmar,
M.C., 2014. Establishing the importance of human health risk
assessment for metals and metalloids in urban environments.
Environ. Int. 72, 176–185. doi:10.1016/j.envint.2014.04.007
Pepper, I.L., Gerba, C.P., Brusseau, M.L., 2011. Environmental
and Pollution Science. Academic Press.
Perelman, A.I., Kasimov, N.S., 1999. Landscape geochemistry.
RAIS, 2017. Risk Exposure Models for Chemicals User’s Guide
[WWW Document]. Risk Assess. Inf. Syst. URL
https://rais.ornl.gov/tools/rais_chemical_risk_guide.html
(accessed 1.1.17).
Rapant, S., Fajčíková, K., Khun, M., Cvečková, V., 2010.
Application of health risk assessment method for geological
environment at national and regional scales. Environ. Earth
Sci. 64, 513–521. doi:10.1007/s12665-010-0875-x
Sun, Y., Zhou, Q., Xie, X., Liu, R., 2010. Spatial, sources and
risk assessment of heavy metal contamination of urban soils
in typical regions of Shenyang, China. J. Hazard. Mater. 174,
455–462. doi:10.1016/j.jhazmat.2009.09.074
Tepanosyan, G., Sahakyan, L., Belyaeva, O., Saghatelyan, A.,
2016. Origin identification and potential ecological risk
assessment of potentially toxic inorganic elements in the
topsoil of the city of Yerevan, Armenia. J. Geochemical
Explor. 167, 1–11. doi:10.1016/j.gexplo.2016.04.006
US EPA, 1989. Risk Assessment Guidance for Superfund
(RAGS), Volume I: Human Health Evaluation Manual
(HHEM). Part A. Baseline risk assessment. EPA/540/1-
89/002.
US EPA Method 6200, 2007. Field portable x-ray fluorescence
spectrometry for the determination of elemental
concentrations in soil and sediment [WWW Document]. URL
http://yosemite.epa.gov/r9/sfund/r9sfdocw.nsf/3dc283e6c5d6
056f88257426007417a2/e599199dc919b049882576a300616
943/$file/attachment
1.pdf#_ga=1.260838029.1899728909.1418825614
WHO, 2008. Guidelines for drinking-water quality - Volume 1:
Recommendations. Third Ed. Inc. first Second addenda 668.
Wong, C.S.C., Li, X., Thornton, I., 2006. Urban environmental
geochemistry of trace metals. Environ. Pollut. 142, 1–16.
doi:10.1016/j.envpol.2005.09.004
Yuan, G.-L., Sun, T.-H., Han, P., Li, J., Lang, X.-X., 2014.
Source identification and ecological risk assessment of heavy
metals in topsoil using environmental geochemical mapping:
Typical urban renewal area in Beijing, China. J. Geochemical
Explor. 136, 40–47. doi:10.1016/j.gexplo.2013.10.002