Environment International 37 (2011) 435–442

Contents lists available at ScienceDirect

Environment International
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n v i n t

Health risks from large-scale water pollution: Trends in Central Asia

Rebecka Törnqvist a,⁎, Jerker Jarsjö a, Bakhtiyor Karimov b
a
Department of Physical Geography and Quaternary Geology, Svante Arrhenius väg 8C, Stockholm University, SE-106 91 Stockholm, Sweden
b
Institute of Water Problems of Uzbekistan Academy of Sciences, Fayzulla Khojaev Street 25A, Tashkent 700125, Uzbekistan

a r t i c l e i n f o a b s t r a c t

Article history: Limited data on the pollution status of spatially extensive water systems constrain health-risk assessments at
Received 17 March 2010 basin-scales. Using a recipient measurement approach in a terminal water body, we show that agricultural
Accepted 10 November 2010 and industrial pollutants in groundwater–surface water systems of the Aral Sea Drainage Basin (covering the
Available online 4 December 2010
main part of Central Asia) yield cumulative health hazards above guideline values in downstream surface
waters, due to high concentrations of copper, arsenic, nitrite, and to certain extent dichlorodiphenyltri-
Keywords:
Aral Sea
chloroethane (DDT). Considering these high-impact contaminants, we furthermore perform trend analyses of
Health risk their upstream spatial–temporal distribution, investigating dominant large-scale spreading mechanisms. The
Pollution ratio between parent DDT and its degradation products showed that discharges into or depositions onto
Irrigation surface waters are likely to be recent or ongoing. In river water, copper concentrations peak during the spring
Surface water season, after thawing and snow melt. High spatial variability of arsenic concentrations in river water could
Groundwater reflect its local presence in the top soil of nearby agricultural fields. Overall, groundwaters were associated
with much higher health risks than surface waters. Health risks can therefore increase considerably, if the
downstream population must switch to groundwater-based drinking water supplies during surface water
shortage. Arid regions are generally vulnerable to this problem due to ongoing irrigation expansion and
climate changes.
© 2010 Elsevier Ltd. All rights reserved.

1. Introduction
Groundwater and surface waters that constitute an integral part of
the continental water cycle can transport and spread contaminants
from spatially limited industrial areas or mining areas to extensive
downstream regions. This is in addition to spreading of more diffuse
contamination, such as from pesticides distributed over agricultural
fields. Water resources are affected by man-made pollution world-
wide to such a degree that restoration to pristine conditions is not
achievable. In this context, health-risk assessments (e.g., WHO, 2006;
U.S. EPA, 1989, 2006; Ma, 2000; Fryer et al., 2006; Törnqvist et al.,
2008) and ecological impact assessments (e.g., Schlüter et al., 2005)
provide means for rational identification of unacceptable pollution
levels and appropriate remediation target levels.
For site-specific tailoring of effective remediation methods, one
must also assess how downstream pollution levels can change in
response to different plausible upstream remediation measures, e.g.,
source removal, land use change, and various water treatments. In
addition, climatic and other changes in ambient conditions may
influence pollution spreading. Hence, remediation assessments need
to be underpinned by scientific understanding of dynamic flow and
transport processes on relatively large scales. The current scientific
⁎ Corresponding author. Tel.: +46 8 6747833; fax: +46 8 164818.
E-mail address: rebecka.tornqvist@natgeo.su.se (R. Törnqvist).
understanding (e.g., Malmström et al., 2000, 2008; Jarsjö et al., 2005;
Shibuo et al., 2007; Jarsjö and Bayer-Raich, 2008; Destouni et al.,
2010a) is hampered, however, by the lack of clear field datasets and
syntheses of large-scale contaminant spreading that can be used
for independent testing of various predictive models. In particular,
observed changes need to be large enough to stand out from random
fluctuations caused, for example, by weather conditions and geo-
logical heterogeneity.
An example of severe human-induced changes in the continen-
tal cycling of water and dissolved substances is provided by the
deepening environmental, health and socioeconomic crisis in Central
Asia (traditionally defined as the former Soviet Union states of
Uzbekistan, Turkmenistan, Tajikistan, and Kyrgyzstan), which to
large extent coincides with the extensive Aral Sea Drainage Basin
(ASDB) that also includes southernmost Kazakhstan and northern
Afghanistan (Fig. 1). The crisis is mainly due to the overexploitation of
water resources caused by gigantic irrigation programs in the lower
arid region of ASDB during the Soviet era (Micklin, 2003; Glantz,
1999). The overexploitation has led to a drastic drop in the river
discharge of the two main rivers into the Aral Sea causing its des-
iccation, see Fig. 1. Groundwater flows have not decreased as much as
surface water flows and have therefore become increasingly important
for the overall water budget in the Aral Sea vicinity in recent years
(Jarsjö and Destouni, 2004; Gascoin and Renard, 2005; Shibuo et al.,
2006, 2007). The intensive agriculture in the region is also associated
with extensive use of fertilizers and pesticides, which has polluted

0160-4120/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.envint.2010.11.006
436 R. Törnqvist et al. / Environment International 37 (2011) 435–442

Fig. 1. The Aral Sea drainage basin (black solid line) with the Amu Darya river delta (dotted line) south of the sea, in which Mejdurechye Reservoir is located. Location of sampling
sites for surface water and groundwater, main irrigated areas (outlined in grey; Global Map of Irrigation Areas, Siebert et al., 2005), the extent of the Aral Sea in 2005 (filled blue) and
in 1960 are shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

water and soils. From a scientific viewpoint, the large magnitude of
observed environmental change within the ASDB provides unique
opportunities for testing hypotheses regarding large-scale effects of
various contaminant transport processes.
Elevated concentrations of organochlorine pesticides and toxic
metals, including arsenic, have been observed in human blood, milk,
hair and urine of the population living in the ASDB (Atanyiazova et al.,
2001; Jensen et al., 1997; Erdinger et al., 2004). Maternal and infant
mortality, delay in growth and puberty of children, liver and digestive
system disorders, allergic problems and diseases related to bacterio-
logical status of water and soils have significantly increased in this
region compared to 1960, prior to the ecological decline (Elpiner,
1999; Zetterström, 1999). Locally produced animal foodstuff has been
found to provide an important exposure pathway for persistent
organic pollutants in Karakalpakstan in Uzbekistan (Muntean et al.,
2003). Furthermore, consumption of local, untreated water is
common in the lower ASDB. In Karakalpakstan for instance, only 20
to 30% of the rural population have access to piped water and 25% of
the population use water from irrigation channels as their key water
source (Small et al., 2003). It is plausible that pollutant exposure
through drinking water can considerably contribute to main health
problems. For example, several compounds have been found to
exceed World Health Organisation (WHO) drinking water standards
in groundwaters and surface waters of the drainage basin, such as
nitrogen compounds, copper, lead, chromium and uranium (Crosa
et al., 2006a; Froebrich et al., 2006; Kawabata et al., 2006; Friedrich,
2009).
We here investigate health hazards related to water-borne, basin-
scale spreading of multiple persistent pollutants, including pesti-
cides and toxic metals. When such large scales are considered, a main
difficulty is that the spatial coverage of surface water and groundwa-
ter measurement points becomes limited. Not least the groundwater
system is relatively unavailable to detailed measurement. This
difficulty is addressed by interpretation of novel measurements
in the downstream Mejdurechye reservoir that due to the last
decade's acute water scarcity in effect has replaced the Aral Sea as a
terminal water body and main recipient of pollutants from major
groundwater–surface water systems of the ASDB. We identify the
most hazardous contaminants (from a human health perspective)
from this hydrological focal point and synthesise data on their
distribution in upstream water systems of the largest sub-basin of the
ASDB, namely the Amu Darya drainage basin. We investigate in
particular spatial and temporal trends along hydrological pathways,
and differences between different hydrological systems (river water,
reservoir water and groundwater). More generally, the analyses aim
at revealing large-scale effects of physical and chemical processes that
govern the spreading of persistent toxic contaminants.
2. Site description and methods
2.1. Site description
The ASDB is an endorheic basin of 1,874,000 km2, which comprises
two principal rivers that discharge into the Aral Sea; the 2400 km long
Amu Darya river and the 2500 km long Syr Darya river, see Fig. 1. Amu
Darya river discharge constitutes a major part (about 70%) of the
renewable water resources within the ASDB. Meltwater from snow
and glaciers in the Pamir mountains is a main source for the river
discharge, which causes flow peaks between April and September
(Froebrich, 2004). The average Amu Darya discharge before the large-
scale expansion of the irrigation system was 78 km3 year−1 (Crosa
et al., 2006a). However, because the Amu Darya river basin contains
most of the irrigated areas in the ASDB, today's river discharge varies
between 0 and 1 km3 year−1 (Oberhänsli et al., 2009).
Through extensive use of fertilizers, pesticides and herbicides,
toxic chemicals have been spread over the agricultural fields, including
 R. Törnqvist et al. / Environment International 37 (2011) 435–442 437

dichlorodiphenyltrichloroethane (DDT), γ-hexachlorocyclohexane
(γ-HCH), arsenic and copper. Surplus irrigation water evaporates
or returns to the rivers, which results in downstream accumulation
of agrochemicals and salt in the rivers (Crosa et al., 2006b). Fur-
thermore, during the Soviet era significant mineral deposits were
found within the ASDB. For instance, metals recovered from today's
mining industry in Uzbekistan include cadmium, copper, lead and
uranium (Clark and Naito, 1998). As a consequence of the well-
developed mining industry, a main pollution source in ASDB is acid
mine drainage, which generally contains high metal concentrations,
including also elevated arsenic concentrations (see e.g. Smedley and
Kinniburgh, 2002; Duker et al., 2005).
The Amu Darya river delta (Fig. 1) is a recipient of water-borne
pollutant flows from the main part of ASDB due to its physical location
far downstream in the basin. The delta is situated in an arid climate
zone. Temperatures range from 45 °C in summers to −30°C in winters
(Nezlin et al., 2004), the average precipitation is 133 mm year−1 and
the potential evaporation varies between 1200 to 1600 mm year−1
(Schutter and Dukhovny, 2003). It is thus entirely dependent on
runoff routed through rivers and canals from upstream areas. From a
water balance — vulnerability perspective, it therefore shares
characteristics with, e.g., the Nile delta in Egypt, and the Eufrat–Tigris
marshes in Iraq. The Amu Darya delta and associated irrigation canal
systems cover approximately 40,000 km2, which is almost double the
size of the Nile delta. The Amu Darya river is, since about a decade,
water depleted before reaching even the former shoreline of the Aral
Sea. The Mejdurechye reservoir, having an area of about 180 km2 and
a maximum depth of 4 m (see the Supplementary information for a
more detailed description), is one of the largest water bodies in the
delta and has in effect become a new terminal water body of the delta.
2.2. Sampling and experimental methods
Concentration measurements of 13 different substances were
conducted in the Mejdurechye reservoir waters during sampling
campaigns in July and October 2007, and in May and July 2008.
Sampling procedures, measurement locations (Fig. S1), analysis
methods (Table S2), instruments (Table S2), detection limits (Table
S2) and further references are given in the Supplementary information.
2.3. Interpretation methods
2.3.1. Hazard factors (HFs) and cumulative HFs (CHFs)
We express health risks through hazard factors (HFs), by dividing
measured concentration levels with a reference concentration, such
that HFs above 1 indicate potential health risks. The reference
concentration is a substance-specific concentration below which the
risk for adverse health effects is very low, given a lifetime intake of
water at the reference concentration (Fryer et al., 2006; WHO, 2006).
The HF quantifications given here are based on drinking water
guideline values from WHO (2006), see Table 1. Guideline values from
the U.S. Environmental Protection Agency (U.S. EPA, 2006) can differ
from the WHO-values, due to quantification uncertainties and
different assumptions regarding for instance body weight (WHO
assumes 60 kg and U.S. EPA assumes 70 kg) and the relative source
contribution for substances that mainly are likely to come from food
(WHO assumes 10% and U.S. EPA assumes 20%). For chromium,
fluoride, mercury, and γ-HCH, the two standards differ by a factor two
or more (Table 1). We then report to which extent resulting
differences in HF impact interpretations and conclusions.
Possible health hazards of several co-existing dissolved contami-
nants can be quantified through the addition of the mixture's
substance-specific HFs. This yields a cumulative hazard factor
(CHF = ∑HF), which with its assumed additive toxic effects provides
a relevant first-order estimate of aggregated toxicity, at least in the
absence of more detailed data on interactions between the compo-
nents of the considered mixture (WHO, 2006; U.S. EPA, 1989). CHFs
were calculated for the mixture of substances measured at each of the
four sampling campaigns performed during 2007 and 2008 in
Mejdurechye reservoir.
2.3.2. Statistical analysis of trends along the Amu Darya river
For the substances that predominantly contributed to elevated
CHFs in the Mejdurechye Reservoir, statistical analyses were
conducted of available monthly river monitoring data from the period
between 1998 and 2002 at Termez, Tuyamuyn, Nukus and Kziljar
gauging stations (Nasrulin and Zahidova, 2002; Fig. 1). The analysed
dataset comprised in total four substances and 680 observations. In
a series of statistical tests the substance-specific data were further
grouped into (1) four categories according to location (e.g., all
Termez-data in one category), (2) a maximum of 20 categories
according to location and year (e.g., all Termez-data from 2002 in
one category), and (3) a maximum of 16 categories according to
location and season (categorising December to February as winter,
March to May as spring, June to August as summer, and September
to November as autumn). A total of 84 category pairs were tested
for whether or not the difference in the category mean values was
statistically significant, using the Smith–Satterthwaite two sample
two tail test for populations with unknown and unequal variances at a
95% confidence level. The test approximates a t-distribution with a

Table 1
Guideline values for drinking water quality from WHO (2006) and U.S. EPA (2006) as well as main adverse health effects related to excess intake of the considered substances (Howd
and Fan, 2008; WHO, 2006). When the two standards differ by a factor two or more for a certain substance, the guideline value is marked in bold.

WHO U.S. EPA Health effects (Howd and Fan, 2008; WHO, 2006)
[μg L−1] [μg L−1]

Nitrate, NO-3 50 000 44 000 Shortness of breath and blue baby syndrome for infants
Nitrite, NO-2 200 (long-term exposure) 3300 Shortness of breath and blue baby syndrome for infants
3000 (short-term exposure)
Chromium tot, Cr 50 100 Allergic problems
Arsenic, As 10 10 Skin damage and circulatory system problems; increased risk of lung, bladder or skin cancer
Fluoride, F- 1500 4000 Bone fluorosis
Copper, Cu 2000 1300 Gastrointestinal effects, liver and kidney damage
Lead, Pb 10 15 Children: delay in physical and mental development; Adults: kidney damage
Cadmium, Cd 3 5 Kidney damage
Mercury, Hg 6 2 Kidney damage
Manganese, Mn 4000 – Neurological effect
Nickel, Ni 70 – Allergic effect
DDT 1 – Liver cancer
DDD 1 – Liver cancer
DDE 1 – Liver cancer
γ-HCH 2 0.2 Liver and kidney problems
438 R. Törnqvist et al. / Environment International 37 (2011) 435–442
degree of freedom estimated from observed sample variances
(Johnson, 2000).
2.3.3. Analysis of HFs in different water systems of the basin
The following sites and water systems were considered in the
basin analysis (see Fig. 1 for locations): (i) reservoir water samples
from two measurement points in Tuyamuyn Hydroengineering
Complex (THC) (conducted in April 2003, mean values of three to
five measurements per substances are reported for nitrate, nitrite,
copper, cadmium, chromium, manganese, nickel and lead; Crosa et al.,
2006b), (ii) river water samples from Darganata (conducted in April
2003, mean values of three to five measurements per substances are
reported for nitrate, nitrite, copper, cadmium, chromium, manganese,
nickel and lead; Crosa et al., 2006b), Termez, Tuyamuyn downstream
of THC, Nukus, and Kziljar (average of 60 measurements per
substance between 1998 to 2002 for nitrate, nitrite, fluoride, arsenic,
mercury, copper, chromium, manganese, nickel and lead; Nasrulin
and Zahidova, 2002), (iii) drainage water samples from Bukhara
(conducted in April 2003, mean values of three to five measurements
per substance of nitrate and nitrite; Crosa et al., 2006b) and Kungrad
and Nukus–Chimbay–Kegeyeli–Bozatau, the latter hereafter abbrevi-
ated to Nukus (conducted in the winters of 1999 and 2000, mean
values of occasional measurements of nitrate and nitrite; Kurbanbaev
et al., 2002), and (iv) 14 groundwater sampling locations within a
distance of 45 km from the Amu Darya river, including Chalish,
Dusenbay, Yangibazar, and Buzkala (mean values of three measure-
ments per location and substance of copper, chromium and lead;
Froebrich et al., 2006). The samples were taken in formations with
hydrogeological characteristics that can allow considerable water
withdrawals for, e.g., drinking water supply. In addition, we compare
resulting main trends in HFs of groundwater and surface water from
the Amu Darya drainage basin with results synthesised from the Syr
Darya drainage basin from groundwater or surface water measure-
ments in Amanotkel (river water; Bosch et al., 2007), Zhalangash
(groundwater; Bosch et al., 2007) and Kazalinsk (river water and
groundwater; Chiba et al., 2006).
3. Results and discussion
3.1. Trends along the Amu Darya river
In the terminal surface water recipient of water-borne pollution in the Amu Darya
river basin, the Mejdurechye reservoir, the four recent (2007–2008) measurement
campaigns showed that the concentrations of cadmium and nitrite from time to time
exceeded WHO's health-risk based guideline values (Table S1). Copper, fluoride, γ-HCH
and arsenic concentrations were just below their respective guideline values, whereas
nitrate, mercury, lead and chromium concentrations were well below their guideline
values. Furthermore, during the 2007–2008 campaigns, DDT and its degradation products
were detected for the first time since at least 2002. The CHF for the Mejdurechye reservoir,
expressing combined toxic effects of the measured substances, was approximately 1 in
the two 2007-campaigns, 0.03 in May 2008, and 9 in July 2008. Different combinations
of substances contributed to elevated CHFs (i.e., CHF ≥ 1) on different measurement
occasions; in July 2007 copper, fluoride and arsenic dominated, whereas nitrite, fluoride
and copper dominated in October 2007, and cadmium and nitrite dominated in July 2008
(see further Table S1 in Supplementary information for detailed Mejdurechye measure-
ment results). The WHO long-term exposure reference concentration was used for nitrite
when calculating CHFs.
For four of the top five substances contributing to the Mejdurechye CHFs in 2007 and
2008 (arsenic, copper, fluoride and nitrite), monthly measurement data from the stations
Termez (Te), Tuyamuyn (Tu), Nukus (Nu) and Kziljar (Kz) located along an extensive
upstream stretch of Amu Darya river were available from a previous period (1998–2002).
Considering these four high HF substances, Table 2 shows the percentage of significantly
different ensemble mean HF values between categories, when categorising data according
to location (rows 1, 2 and 3), year (row 4), and season (row 5) of measurement.
The difference in mean arsenic HFs between locations (1) Te–Tu, (2) Te–Nu, (3)
Te–Kz, (4) Tu–Nu, (5) Tu–Kz, and (6) Nu–Kz, was found to be statistically significant in
four out of these six cases, yielding a percentage of 67, as shown in column 1, row 1, of
Table 2. The two location pairs that did not show statistically different mean values
were (1) Te–Tu and (6) Nu–Kz. This result can be understood by considering
underlying ensemble mean HFs and standard deviation values for Te, Tu, Nu and Kz,
plotted in Fig. 2a. The figure shows a considerable overlap in range (mean ± one
standard deviation) between the upstream Te and Tu, stations, as well as between the
Table 2
The proportion (in percentage) of significant differences in means for HFs for individual
substances (As, Cu, F− and NO− 2) between (1) the stations within the measurement
period 1998–2002, (2) the stations within years, (3) the station within seasons, (4) the
years within locations, and (5) the seasons within stations.
Arsenic, Copper, Fluoride, F- Nitrite,
As Cu NO-2
Variation between locations:
(1) Within the measurement period 67 0 33 50
(2) Within years 32 16 11 5
(3) Within seasons 19 19 16 19
(4) Variation between years:
Within locations 12 0 28 0
(5) Variation between seasons:
Within locations 6 31 0 19
downstream Nu and Kz stations, and a lack of overlap between all other possible station
combinations, which yields the statistical outcome presented in Table 2. There is hence
a clear upstream–downstream difference in mean arsenic HFs (Fig. 2a). Nitrite HFs also
show a considerable spatial variation (Table 2, column 4). Table 2 furthermore shows
that the spatial variations of arsenic and nitrite HFs are strong relative to annual
(Table 2, row 4) and seasonal (Table 2, row 5) variations.
The overall results for copper showed no significant differences in mean HF values
between measurement locations (Table 2, column 2, row 1). Some spatial differences
exist within individual years and seasons (column 2, rows 2 and 4). However, the most
pronounced variation in copper HFs is the seasonal one, yielding statistically significant
differences in 31% of the considered cases (column 2, row 5 in Table 2). Fig. 2b shows
that relatively low copper HFs during summer and autumn at the four measurement
points contribute to this statistical outcome. Finally, fluoride HFs show both annual
(column 3, row 4 in Table 2) and spatial (column 3, row 1 in Table 2) variations. Fig. 2c
shows that there is an underlying trend of annually increasing fluoride HFs between the
years 1999 and 2002.
3.2. HFs in different water systems of the basin
Fig. 3 shows comparisons between mean HFs for river and reservoir water in the
Amu Darya drainage basin (shown to the right of the vertical lines in Fig. 3) on the one
hand, and drainage water (Fig. 3a) and groundwater (Fig. 3b and c), on the other hand
(shown to the left of the vertical lines in Fig. 3). The measurement stations of the river
and reservoir waters are ordered considering surface water flow directions, from the
station furthest upstream (Termez) to the station furthest downstream (Mejdurechye
Reservoir 3). In analogy, the drainage water stations are ordered from upstream Bukhara
to downstream Kungrad (Fig. 3a, left side) and the groundwater stations are ordered
from upstream Chalish to downstream Dusenbay (Fig. 3b and c, left side). The mean
groundwater flow directions were taken from estimates of Johansson et al. (2009).
The HFs for nitrate and nitrite in river water were below one and did not indicate a
health risk, whereas drainage water from Nukus had a nitrite HF close to 3 (Fig. 3a).
Furthermore, the Mejdurechye reservoir water is associated with significantly higher
nitrite HFs (close to 4; Fig. 3a, right side) than upstream river waters of Amu Darya. The
HFs for fluoride, arsenic, mercury and copper were less than one in the sampled river
and reservoir waters of the lower Amu Darya drainage basin (Fig. 3b, right side). The
highest HFs in the river-reservoir system were observed for fluoride (HF = 0.4) and
copper (HF = 0.2) in the Mejdurechye reservoir water. The arsenic HF was also slightly
higher in the Mejdurechye reservoir than in upstream river waters. Of the four
substances shown in Fig. 3b, copper is the only one that reportedly has been subject to
measurement in groundwater. The results show that copper HFs in groundwater are
about two orders of magnitude higher than in river water and one order of magnitude
higher than in reservoir water (Fig. 3b, left side). The temporal and spatial variations of
copper HFs in the river water of the Amu Darya delta are much smaller than that
compared with the standard deviation values of Fig. 2b, which indicates that there is a
systematic difference between copper HFs in groundwater and river water. Fig. 3b also
shows that the copper HFs in groundwater increase along the groundwater flow
direction, from Chalish to Dusenbay.
Fig. 3c (left side) shows that chromium, lead, manganese and nickel concentrations
in the THC reservoir waters and the upstream river water at Darganata are hazardously
high with HFs between 1 and 5. HFs for chromium in reservoir waters are however
lower than 1 if the higher U.S. EPA guideline value is used. This is the only substance for
which the choice of reference concentration standards affects whether the HF value
exceeds 1 or not. Notably, Fig. 3c (right side) demonstrates exceptionally high HFs for
chromium (up to 280) and lead (up to 65) in groundwater samples from Chalish,
Yangibazar, Dusenbay and Buzkala. These groundwater values are about four orders of
magnitude higher than mean HFs for chromium and lead at the river measurement
stations between the THC and the Mejdurechye reservoir. Moreover, although we in
Fig. 3c for illustrative purposes plot results from only 4 out of 14 groundwater sampling
locations of Froebrich et al. (2006), the other 10 sampling locations (in the Amu Darya
delta within a distance of 45 km from the Amu Darya River) had equally high
groundwater HFs for chromium and lead (intervals of 120–280 and 20–70,
 R. Törnqvist et al. / Environment International 37 (2011) 435–442 439

a)

b)

c)

Fig. 2. Hazard factor (HF) ensemble mean (filled circles) and standard deviation (error bars) for: a) arsenic in Termez, Tuyamuyn, Nukus and Kziljar during 1998–2002, b) copper in
Termez, Tuyamuyn, Nukus and Kziljar grouped according to season, and c) fluoride in Termez, Tuyamuyn, Nukus and Kziljar grouped according to years. See Fig. 1 for sample site
location.

respectively). The 10 sampling locations also had as high groundwater HFs for copper
(interval of 1–4.5) as the 4 locations illustrated in Fig. 3b.
A synthesis of groundwater (gw)–river water (rw) measurement results from the
adjacent Syr Darya drainage basin (Bosch et al., 2007; Chiba et al., 2006) furthermore
showed that the highest HFs were associated with groundwater (like in the Amu Darya
river basin). Specifically, of the 13 substances considered here, the following had
relatively high HFs in water systems of the Syr Darya drainage basin: arsenic, 0.9 (gw)
and 0.2 (rw); fluoride, 0.7 (gw) and 0.4 (rw); nitrate 4.9 (gw) and 0.1 (rw); and
cadmium 0.7 (gw) and 0.6 (rw). Three of these substances were also included among
the top HF substances in surface waters of Amu Darya during 2007–2008 (arsenic,
fluoride and cadmium; see section 3 of the Supplementary information). Notably, the
groundwater concentrations of these three substances are unknown in the Amu Darya
river basin due to lack of measurements (or, at least, lack of publically available
measurement results).
3.3. Possible sources and large-scale spreading mechanisms of key substances
Water samples from the 2007–2008 measurement campaigns in Mejdurechye con-
tained concentrations of the banned pesticides γ-HCH, DDT and degradation products
of DDT (dichlorodiphenyldichloroethane (DDD) and dichlorodiphenyldichloroethylene
440 R. Törnqvist et al. / Environment International 37 (2011) 435–442

a)

b)

c)

Fig. 3. Mean HFs in: a) drainage water and river water for nitrite and nitrate, b) groundwater, reservoir and river water for fluoride, arsenic, mercury and copper, and c) groundwater
reservoir and river water for cadmium, chromium, manganese, nickel and lead. Reservoir water sample locations are underlined. Values below the horizontal solid lines indicate
samples with concentrations below the detection limits (DL) and thereby of low HFs. Locations are categorised according to upstream to downstream mean surface waters and
groundwater flows, i.e., from left to right in the figure. Abbreviations: Res. THC = Tuyamuyn Hydrological Complex Reservoir, Res. Mejdu. = Mejdurechye Reservoir. See Fig. 1 for
sample site location.

(DDE)), which is notable since levels of these pesticides have been below the detection
limit in Mejdurechye reservoir measurements since 2002. These elevated concentrations
constitute a break from the trend of significant decrease that has prevailed in the entire
Amu Darya delta since at least 1987 (Karimov et al., 2005). The average ratio of parent DDT
concentration and sum concentration of the degradation products DDE and DDD
(hereafter denoted rD; this ratio hence becomes lower as the plume ages and degrades)
was equal to 4.8 for the 2007 and 2008 campaigns in the Mejdurechye reservoir
(Supplementary information, Table S1).
The time span between the DDT ban in the Soviet Union (in 1973; Zhulidov et al.,
2000) and the 2007–2008 measurement campaigns is longer than the half-life of
DDT in soil (15 years; Howard, 1991). Neglecting differences in transport properties
between DDT, DDD and DDE (e.g., assuming equal retardation), sources of old DDT from
before the ban would yield rD b 1 in water, which is significantly lower than observed.
Hence, if this simplifying assumption would be valid, the detected DDT could not
possibly result from subsurface transport of old DDT. The simplifying assumption can
be relaxed by the consideration of differences in retardation between DDT and its
degradation products, with the partition coefficient Kd being higher for DDT (log
Kd = 4.8) than for DDE (log Kd = 2.7) and DDD (log Kd = 2.9) (Walters and Aitken,
2001), implying higher retardation and slower transport of DDT than DDE and DDD.
In comparison with the equal retardation case, these retardation differences can
contribute to decreases in rD at downstream observation points, since low DDT
concentrations would prevail for longer times due to its higher retardation. Whereas
 R. Törnqvist et al. / Environment International 37 (2011) 435–442
this reasoning is relevant in many cases, there are exceptions for which retardation
differences can contribute to increases in rD. Such events (that hence can mask the
true age of old spills; making them appear more recent) are however limited in space
and/or time, see the Supplementary information for details. Notably, the Mejdurechye
reservoir receives pollutants from large upstream areas and has been monitored during
relatively long periods of time.
Overall, it therefore appears unlikely that old DDT can explain the high rD found in
the Mejdurechye reservoir. A plausible alternative explanation would be that the DDT
leakage is relatively recent. Since the half-life of DDT is short in surface water (1 year;
Howard, 1991), it cannot have been in contact with surface water for any prolonged
periods of time, before detection in the Mejdurechye reservoir. This limits plausible
DDT sources to illegally used stockpiles, recent leakage from storage sites and
atmospheric deposition. Similarly high rD-values have for instance been reported in the
Pearl River Delta waters in South China, indicating ongoing DDT use after China's 1983
ban (Guo et al., 2009).
In the ASDB, main sources of copper contamination are located at the ground
surface, or in the near-surface environment. Sources include copper mine tailings and
agricultural fields, where fertilizers (of e.g., copper sulphate) and pesticides (e.g.,
copper hydroxide; Zhang et al., 2004) are used. In the most downstream river gauging
station of Amu Darya (the Kziljar station; located close to the Mejdurechye reservoir
inlet), copper HFs peak during the spring, when precipitation is high. It is primarily
elevated copper concentrations in March that contribute to the statistically significant
correlation between season and copper concentration. A plausible explanation for this
correlation is that copper may accumulate in upper soil layers when conditions are dry,
and mobilise during early high-precipitation events of the growing season, as a result of
leaching and/or soil erosion. Zhang et al. (2004) for instance referred to this mechanism
as a plausible explanation for observed peaks of dissolved copper in drainage water
from agricultural fields in St. Lucie County, Florida. It can also explain some ob-
servations of Bordalo et al. (2001) in Bangpakong River, Thailand. The lower Amu Darya
region contrasts to the study areas of Zhang et al. (2004) and Bordalo et al. (2001)
through its lower winter temperatures (below 0 °C), its more extensive irrigation, and
its arid climate with precipitation events mainly limited to winter and early spring
(Sorrel et al., 2007). The present results specifically show that copper concentrations
peak after thawing and snow melt, and that concentrations decline in the summer
season despite considerable water re-routing to irrigated fields. Nitrite concentrations
of the lower Amu Darya river were also higher during spring than summer; co-
variations between copper and nitrogen compounds also exist under different ambient
conditions as reported by, e.g., Zhang et al. (2004) and Bordalo et al. (2001).
Surface water systems are generally not affected by natural arsenic, unless there
are geothermal sources (Smedley and Kinniburgh, 2002). Our results show that, in the
presence of superficial arsenic sources of human origin (from agriculture and possibly
mining), arsenic can leach to surface waters to such an extent that considerably
contributes to CHFs in drinking water. Since previous studies on arsenic leaching from
cotton fields additionally have shown that vertical arsenic migration is very limited
even under irrigation (Reedy et al., 2007), it is likely that leaching into surface water
systems constitutes a main transport pathway for anthropogenic arsenic (in contrast to
natural arsenic). The high spatial variability of arsenic concentrations in river water
along Amu Darya is furthermore consistent with findings summarized in Smedley and
Kinniburgh (2002) of spatially limited arsenic contamination in surface waters around
known superficial sources (mining tailings). High arsenic values in downstream ASDB
river water stations could therefore signal its presence in the top soil of nearby
agricultural fields.
4. Conclusions
Surface water results show that depending on time and location of
drinking water intake, the cumulative health hazard factor (CHF) of
the considered contaminants varies from insignificant to approxi-
mately 10. Concentrations of copper, lead, nickel, manganese and
chromium were detected at concentrations exceeding the WHO
drinking water guideline values in the Amu Darya delta. Although
average concentrations of cadmium were below the guideline value, it
was exceeded at individual measurement occasions. The known
adverse health effects (Table 1; WHO, 2006) of an excess intake of
these contaminants have already been observed in the population of
the river basin (Elpiner, 1999; Zetterström, 1999). It is therefore
reasonable to assume that poor drinking water quality has contrib-
uted to the situation. Since low-risk surface waters still exist today
within the Amu Darya delta at some locations (at least temporarily) it
is plausible that water regulation schedules could be modified to
optimize water quality with regard to aggregated toxicity at critical
drinking water intake points.
In contrast to relatively low-risk surface water systems of the Amu
Darya delta, its groundwater in formations that (from a hydrogeolo-
gical viewpoint) can support larger-scale drinking water supply
441
contain substance concentrations that are orders of magnitude higher
than WHO drinking water guideline values. This implies that such
groundwater-based drinking water can be associated with severe
health risks, clearly above corresponding surface water risks. In
addition, in the Amu Darya delta, pressures are high on switching
from surface water-based drinking water supplies to groundwater-
based supplies (Rakhmatullaev et al., 2010), as a consequence of
surface water shortage that already has led to the desiccation of the
Aral Sea. Current results suggest that such a change of water supply
systems (i.e., from surface water sources to groundwater sources)
could, apart from possibly decreasing groundwater levels due to
insufficient recharge, imply a considerable worsening of health risks.
The surface water shortage is furthermore expected to become more
severe in the near future due to continued irrigation expansion in the
upstream part of the Amu Darya drainage basin, and ongoing climatic
changes (Dukhovny et al., 2007; Milly et al., 2005; Shibuo et al., 2007;
Destouni et al., 2010b). Analogously, polluted groundwater may
increasingly become the only remaining water source as in many
arid areas of the world, since surface water discharges are expected
to decrease globally in arid areas, in response to projected global
irrigation increase (e.g., due to population growth and increasing
living standards; Rockström et al., 2009) and projected climatic
changes (Intergovernmental Panel on Climate Change, 2007).
Acknowledgements
This study was funded by the Swedish International Development
Cooperation Agency (SIDA). The work has been carried out within the
Bert Bolin Centre for Climate Research, which is supported by a
Linnaeus grant from the Swedish Research Council (VR) and the
Swedish Research Council for Environment, Agricultural Sciences and
Spatial Planning (Formas). We gratefully acknowledge constructive
discussions with Jochen Froebrich, Centre for Water and Climate
(CWK), Wageningen University & Research centre — Alterra, The
Netherlands, including interpretation feedback on past groundwater
measurement campaigns in the Amu Darya delta region.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
doi:10.1016/j.envint.2010.11.006.
References
Atanyiazova OA, Baumann RA, Liem AKD, Mukhopadhyay UA, Vogelaar EF, Boersma ER.
Levels of certain metals, organochlorine pesticides and dioxins in cord blood,
maternal blood, human milk and some commonly used nutrients in the
surroundings of the Aral Sea (Karakalpakstan, Republic of Uzbekistan). Acta
Paediatr 2001;90:801–8.
Bordalo AA, Nilsumranchit W, Chalermwat K. Water quality and uses of the
Bangpakong river (Eastern Thailand). Water Res 2001;35(15):3635–42.
Bosch K, Erdinger L, Ingel F, Khussainova S, Utegenova E, Bresgen N, et al. Evaluation of
the toxicological properties of ground- and surface-water samples from the Aral
Sea Basin. Sci Total Environ 2007;374:43–50.
Chiba M, Shinohara A, Sekine M, Hiraishi S. Drinking water quality from the aspect of
element concentrations. J Radioanal Nucl Chem 2006;269(3):519–26.
Clark AL, Naito K. Risks and opportunities for foreign investment in the mineral sectors
of the Central Asian Republics Kazakstan, Kyrgyzstan, Tajikistan and Uzbekistan.
Resour Policy 1998;24(2):105–14.
Crosa G, Froebrich J, Nikolayenko V, Stefani F, Galli P, Calamari D. Spatial and seasonal
variations in the water quality of the Amu Darya River (Central Asia). Water Res
2006a;40:2237–45.
Crosa G, Stefani F, Bianchi C, Fumagalli A. Water Security in Uzbekistan: implication of
Return Waters on the Amu Darya Water Quality. Environ Sci Pollut Res 2006b;13
(1):37–42.
Destouni G, Persson K, Prieto C, Jarsjö J. General quantification of catchment-scale
nutrient and pollutant transport through the subsurface to surface and coastal
waters. Environ Sci Technol 2010a;44(6):2048–55.
Destouni G, Asokan SM, Jarsjö J. Inland hydro-climatic interaction: effects of human
water use on regional climate. Geophys Res Lett 2010b;37:L18402. doi:10.1029/
2010GL044153.
Duker AA, Carranza EJM, Hale M. Arsenic geochemistry and health. Environ Int 2005;31
(5):631–41.
442 R. Törnqvist et al. / Environment International 37 (2011) 435–442
Dukhovny V, Umarov P, Yakubov H, Madramootoo CA. Drainage in the Aral Sea basin.
Irrig Drainage 2007;56:S91-S100.
Elpiner L. Public health in the Aral Sea coastal region and the dynamics of change in the
ecological situation. In: Glantz M, editor. Creeping environmental problems and
sustainable development in the Aral Sea basin. Cambridge: Cambridge University
Press; 1999. p. 128–56.
Erdinger L, Eckl P, Ingel F, Khussainova S, Utegenova E, Mann V, et al. The Aral Sea
disaster — human biomonitoring of Hg, As, HCB, DDE, and PCBs in children living in
Aralsk and Akchi, Kazakhstan. Int J Hyg Environ Health 2004;207:541–7.
Friedrich J. Uranium contamination of the Aral Sea. J Mar Syst 2009;76(3):322–35.
Froebrich J. Water management aspects of Amu Darya. In: Nihoul JCJ, Zavialov PO,
Micklin PP, editors. Dying and Dead Seas. Climatic Versus Anthropic Causes.
Dordrecht: Kluwer Academic Publisher; 2004. p. 49–76.
Froebrich J, Ikramova M, Razakov R. Improved groundwater management strategies at
the Amu Darya River. In: Baba A, Howard KWF, Gunduz O, editors. Groundwater
and ecosystems. Dordrecht: Springer; 2006. p. 153–65.
Fryer M, Collins CD, Ferrier H, Colvile RN, Nieuwenhuijsen MJ. Human exposure
modelling for chemical risk assessment: a review of current approaches and
research and policy implications. Environ Sci Policy 2006;9:261–74.
Gascoin S, Renard P. Hydrological balance modelling of the southern Aral Sea between
1993 and 2001. Hydrol Sci J 2005;50(6):1119–35.
Glantz M. Creeping environmental problems and sustainable development in the Aral
Sea basin. Cambridge: Cambridge University Press; 1999.
Guo Y, Yu HY, Zeng EY. Occurrence, source diagnosis, and biological effect assessment of
DDT and its metabolites in various environmental compartments of the Pearl River
Delta, South China: a review. Environ Pollut 2009;157(6):1753–63.
Howard PH. Handbook of environmental fate and exposure data for organic chemicals
Vol. II: pesticidesChelsea, Michigan: Lewis Publishers; 1991.
Howd RA, Fan AM. Risk assessment for chemicals in drinking water. New Jersey: John
Wiley & Sons, Inc.; 2008.
Intergovernmental Panel on Climate Change. In: Pachauri RK, Reisinger A, editors.
Climate Change 2007: Synthesis Report, Contribution of Working Groups I, II and III
to the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change; 2007. Geneva, Switzerland, 104 pp. (available electronically at www.ipcc.
ch, accessed 2010-03-17).
Jarsjö J, Destouni G. Groundwater discharge into the Aral Sea after 1960. J Mar Syst
2004;47(1-3):109–20.
Jarsjö J, Bayer-Raich M, Ptak T. Monitoring groundwater contamination and delineating
source zones at industrial sites: uncertainty analyses using integral pumping tests.
J Contam Hydrol 2005;79(3–4):107–34.
Jarsjö J, Bayer-Raich M. Estimating plume degradation rates in aquifers: effect of
propagating measurement and methodological errors. Water Resour Res 2008;44:
W02501. doi:10.1029/2006WR005568.
Jensen S, Mazhitova Z, Zetterström R. Environmental pollution and child health in the
Aral Sea region in Kazakhstan. Sci Total Environ 1997;206:187–93.
Johansson O, Aimbetov I, Jarsjö J. Variation of groundwater salinity in the partially
irrigated Amudarya River delta, Uzbekistan. J Mar Syst 2009;76(3):287–95.
Johnson RA. Miller & Freund's probability and statistics for engineers. 6th ed. Upper
Saddle River: Prentice Hall; 2000.
Karimov B, Lieth H, Kurambaeva H, Matsapaeva I. The problems of fishermen in the
southern Aral Sea region. Mitig Adapt Strateg Glob Chang 2005;10:87-103.
Kawabata Y, Yamamoto M, Aparin V, Ko S, Shiraishi K, Nagai M, et al. Uranium pollution of
water in the western part of Uzbekistan. J Radioanal Nucl Chem 2006;270(1):137–41.
Kurbanbaev RE, Matsapaeva IV, Konstantinova LG. The role of collecting drainage
systems in the formation of chemical composition of lake water in the Amu Darya
lower reaches. Water Resour 2002;29(3):346–50.
Ma H. The incorporation of stochasticity in risk analysis and management: a case study.
Stoch Environ Res Risk Assess 2000;14:195–206.
Malmström ME, Destouni G, Banwart SA, Strömberg BHE. Resolving the scale-
dependence of mineral weathering rates. Environ Sci Technol 2000;34(7):1375–8.
Malmström ME, Berglund S, Jarsjö J. Combined effects of spatially variable flow and
mineralogy on the attenuation of acid mine drainage in groundwater. Appl
Geochem 2008;23(6):1419–36.
Micklin P. The Aral Sea crisis. In: Nihoul JCJ, Zavialov PO, Micklin PP, editors. Dying and
Dead Seas. Climatic Versus Anthropic Causes. Dordrecht: Kluwer academic
Publisher; 2003. p. 99-123.
Milly PCD, Dunne KA, Vecchia AV. Global pattern of trends in streamflow and water
availability in a changing climate. Nature 2005;438:347–50.
Muntean N, Jermini M, Small I, Falzon D, Furst P, Migliorati G, et al. Assessment of
dietary exposure to some persistent organic pollutants in the Republic of
Karakalpakstan of Uzbekistan. Environ Health Perspect 2003;111(10):1306–11.
Nasrulin A, Zahidova M. The experience of creation of the GIS System of ecological
monitoring in Aral Sea Basin. J Prob Mech 2002;4:38–43.
Nezlin NP, Kostianoy AG, Lebedev SA. Interannual variations of the discharge of Amu
Darya and Syr Darya estimated from global atmospheric precipitation. J Mar Syst
2004;47(1–4):67–75.
Oberhänsli H, Weise SM, Stanichny S. Oxygen and hydrogen isotopic water
characteristics of the Aral Sea, Central Asia. J Mar Syst 2009;76(3):310–21.
Rakhmatullaev S, Huneau F, Kazbekov J, Le Coustumer P, Jumanov J, El Oifi B, et al.
Groundwater resources use and management in the Amu Darya River Basin
(Central Asia). Environ Earth Sci 2010;59(6):1183–93.
Reedy RC, Scanlon BR, Nicot JP, Tachovsky JA. Unsaturated zone arsenic distribution and
implications for groundwater contamination. Environ Sci Technol 2007;41(20):
6914–9.
Rockström J, Falkenmark M, Karlberg L, Hoff H, Rost S, Gerten D. Future water availability
for global food production: the potential of green water for increasing resilience to
global change. Water Resour Res 2009;45:W00A12. doi:10.1029/2007WR006767.
Schlüter M, Savitsky AG, McKinney DC, Lieth H. Optimizing long-term water allocation
in the Amudarya River delta: a water management model for ecological impact
assessment. Environ Model Softw 2005;20(5):529–45.
Schutter J, Dukhovny VA, editors. South Prearalie — New Perspectives, report. Scientific-
Information Center of the Interstate Coordination Water Commission of Central
Asia, Tashkent; 2003. available electronically at http://www.cawater-info.net/
library/eng/nato_eng.pdf. accessed 2010-02-17.
Shibuo Y, Jarsjö J, Destouni G. Bathymetry–topography effects on saltwater-fresh
groundwater interactions around the shrinking Aral Sea. Water Resour Res
2006;42:W11410. doi:10.1029/2005WR004207.
Shibuo Y, Jarsjö J, Destouni G. Hydrological responses to climate change and irrigation in the
Aral Sea drainage basin. Geophys Res Lett 2007;34:L21406. doi:10.1029/2007GL031465.
Siebert S, Döll P, Hoogeveen J, Faures JM, Frenken K, Feick S. Development and
validation of the global map of irrigated areas. Hydrol Earth Syst Sci 2005;9(5):
535–47.
Small I, Falzon D, van der Meer JBW, Ford N. Safe water for the Aral Sea Area. Could it get
any worse? Eur J Public Health 2003;13(1):87–9.
Smedley PL, Kinniburgh DG. A review of the source, behaviour and distribution of
arsenic in natural waters. Appl Geochem 2002;17(5):517–68.
Sorrel P, Popescu SM, Klotz S, Suc JP, Oberhänsli H. Climate variability in the Aral Sea
basin (Central Asia) during the late Holocene based on vegetation changes. Quat
Res 2007;67(3):357–70.
Törnqvist R, Norström A, Kärrman E, Malmqvist PA. A framework for planning of
sustainable water and sanitation systems in peri-urban areas. Water Sci Technol
2008;58(3):563–70.
U.S. EPA. Risk Assessment Guidance for Superfund. EPA/540/1-89/002. Office of
Emergency and Remedial Response. Human Health Evaluation Manual (Part A)
Washington, DC: U.S. Environmental Protection Agency; 1989.
U.S. EPA. 2006 Edition of the Drinking Water Standards and Health Advisories EPA 822-
R-06-013. Washington, DC: Office of Water, U.S. Environmental Protection Agency;
2006.
Walters GW, Aitken MD. Surfactant-enhanced solubilization and anaerobic biodegra-
dation of 1, 1, 1 Trichloro-2, 2-bis(p-chlorophenyl)-ethane (DDT) in contaminated
soil. Water Environ Res 2001;73(1):15–23.
WHO. Guidelines for drinking water Quality Vol. 1 Recommendations. 3rd ed. Geneva:
World Health Organisation; 2006.
Zhang MK, He ZL, Calvert DV, Stoffella PJ. Spatial and temporal variations of water
quality in drainage ditches within vegetable farms and citrus groves. Agric Water
Manag 2004;65:39–57.
Zetterström R. Child health and environmental pollution in the Aral Sea region in
Kazakhstan. Acta Paediatr Suppl 1999;88(429):49–54.
Zhulidov AV, Headley JV, Pavlov DF, Robarts RD, Korotova LG, Vinnikov YY, et al.
Riverine fluxes of the persistent organochlorine pesticides hexachlorcyclohexane
and DDT in the Russian Federation. Chemosphere 2000;41:829–41.