Science of the Total Environment 610611 (2018) 820830

Contents lists available at ScienceDirect

Science of the Total Environment

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

Long-term ongoing impact of arsenic contamination on the

environmental compartments of a former mining-metallurgy area
B. Gonzlez-Fernndez a, E. Rodrguez-Valds b, C. Boente b, E. Menndez-Casares a,
A. Fernndez-Braa b, J.R. Gallego b,
a
Dpto. de Explotacin y Prospeccin de Minas, Universidad de Oviedo, C/Independencia, 13, 33004 Oviedo, Asturias, Spain
b
INDUROT and Environmental Technology, Biotechnology and Geochemistry Group, C/Gonzalo Gutirrez Quirs s/n, 33600 Mieres, Asturias, Spain

H I G H L I G H T S G R A P H I C A L A B S T R A C T

40-years abandoned As- and Hg-rich

waste promoted soil pollution.
As showed greater mobility and disper-
sion than Hg.
Waste stock-piled and highly polluted
soil affected water and groundwater
quality.
River sediments were severely enriched
in As and Hg.

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

Article history: Arsenic and mercury are potentially toxic elements of concern for soil, surcial and ground waters, and sedi-
Received 21 June 2017 ments. In this work various geochemical and hydrogeological tools were used to study a paradigmatic case of
Received in revised form 1 August 2017 the combined effects of the abandonment of Hg- and As-rich waste on these environmental compartments. Con-
Accepted 14 August 2017
tinuous weathering of over 40 years has promoted As and Hg soil pollution (thousands of ppm) in the surround-
Available online xxxx
ings of a former Hg mining-metallurgy site and affected the water quality of a nearby river and shallow
Editor: F.M. Tack groundwater. In particular, the high availability of As both in soils and waste was identied as one of the main
determinants of contaminant distribution, whereas the impact of Hg was found to be minor, which is explained
Keywords: by lower mobility. Furthermore, potential additional sources of pollution (coal mining, high natural backgrounds,
Arsenic etc.) discharging into the study river were revealed less signicant than the contaminants generated in the Hg-
Mercury mining area. The transport and deposition of pollutants within the water cycle has also affected several
Water pollution kilometres downstream of the release areas and the chemistry of stream sediments. Overall, the environmental
Soil compartments studies held considerable concentrations of Hg and As, as remarkably revealed by the average
Sediment
contaminant load released in the river (several tons of As per year) and the accumulation of toxic elements in
Mining
sediments (enrichment factors of As and Hg above 35).
2017 Elsevier B.V. All rights reserved.

1. Introduction

Corresponding author. Mining and industrial activities commonly promote waste accumu-
E-mail address: jgallego@uniovi.es (J.R. Gallego). lation and contaminant leaching. These processes have proved

http://dx.doi.org/10.1016/j.scitotenv.2017.08.135
0048-9697/ 2017 Elsevier B.V. All rights reserved.
 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830
especially signicant after the closure of heavy industries in the last de-
cades of the 20th century. In this regard, various studies have addressed
the effects of the uncontrolled disposal and abandonment of toxic waste
exposed to weathering and erosion on soil, sediments and (ground)wa-
ter (Hudson-Edwards et al., 1997; Lefebvre et al., 2001; Foulds et al.,
2014). Contaminants with high toxicity and complex geochemical be-
haviour, such as Potentially Toxic Elements (PTEs), particularly arsenic
(As), are relevant as they require costly remediation treatments
(Bhattacharya et al., 2007; Bundschuh et al., 2013; Smedley and
Kinniburgh, 2002). In addition, with respect to underground mining,
the excavation of wells and galleries modies hydrogeological condi-
tions and may generate mine drainage with a high contaminant load
that affects other environmental compartments, such as soils and sedi-
ments (Moreno-Jimnez et al., 2010; Chakraborti et al., 2013).
Asturias (northern Spain) is a region with a long history of mining
and metallurgical activity linked to coal and metals. Consequently, the
current landscape holds many brownelds (Gallego et al., 2016),
which are affected by both organic contaminants and PTEs such as
heavy metals and metalloids. Hg and As pollution in particular has
been detected at several sites linked to former Hg mining and metallur-
gy activity (Loredo et al., 1999, 2006). These sites have been examined
in various studies (Sierra et al., 2011; Ordez et al., 2011; Sierra et al.,
2013; Ordoez et al., 2014). However, little attention has been paid to
the overlapping and concurrent effects caused by this sort of pollution
in the environmental compartments; i.e. soil pollution caused by
waste disposal, contamination of surcial and ground waters, and ef-
fects on the sediments of the rivers crossing the areas affected. The over-
lapping effects are especially relevant given the complexity of the
hydrogeological conditions in former mining areas, the presence of
other possible sources of contamination, and the occurrence of naturally
high geochemical backgrounds. In addition, knowledge of the mecha-
nisms that regulate the release and mobility of contaminants at this
type of site is essential for the design of strategies aimed to minimise en-
vironmental risk (Haffert and Craw, 2008; Rieuwerts et al., 2014). In this
context, a paradigmatic site in terms of contamination is El Terronal,
which has been described in depth in a previous study (Gallego et al.,
2015), and it is also of particular environmental concern (even notable
atmospheric emissions of Hg have been reported in Loredo et al.,
2007). The site is located in an area affected by other sources of pollu-
tion and it is also very close to the San Tirso River, where high concen-
trations of As have been detected (Ordoez et al., 2014).
The aims of this study were the following; i) to identify and quantify
the main pollutants in the area, specically in the river, over a period
longer than a hydrological year; ii) to perform a comprehensive study
of the complex relationships between waste disposal, soil, sediments,
water run-off and groundwater pollution; iii) to evaluate the mobility
and chemical speciation of the main pollutants found; and iv) to differ-
entiate between possible sources of pollutants (underground mining
drainage, surface run-off, coal vs. Hg mining, etc.).
2. Study area
The study area (San Tirso River valley) is located around 2 km NE of
the town of Mieres, situated in the region of Asturias in northern Spain
(Fig. 1a). The climate in the zone is oceanic with mild temperatures
(yearly average 12.5 C). Regular annual precipitation reaches values
between 1000 and 1200 mm per year and potential evaporation has
been estimated at 600 mm per year, thereby implying an average effec-
tive rainfall of approx. 400 mm per year. In addition, heavy rain events
are frequent (Mndez Pazos, 2013).
Industrial and mining activities associated with Hg ores have been
going on in the valley since Roman times. However, peak production
was recorded towards the end of the 19th century, particularly in the
1960s (Dory, 1984; Luque and Gutirrez-Claverol, 2006). All the exploi-
tations were closed in 1974 because of a crash in the Hg market. In par-
ticular, in addition to the mines, the site of El Terronal hosted a high-
821
capacity pyrometallurgical plant, which was operating from the 1950s
to 1974 (the main elements are detailed in Fig. 1b; see also Gallego
et al., 2015). There is also a ruined duct (chimney) along the western
hillside of the San Tirso valley that ends in a nal stack (350 ma.s.l.),
which once channelled a current of treated gas outside the complex.
Mining activities completely altered the hydrogeology of the area. In
brief, a 200-m-deep well was drilled in El Terronal for ore extraction
(Pozo Esperanza, 235.21 ma.s.l.), whereas in the same area the Schultz
crosscut (227.42 ma.s.l.) was used to exploit the cinnabar in siliceous
breccias. Furthermore, there are a number of secondary drifts on the
western side of the San Tirso valley connected to the Schultz crosscut.
To the SW of El Terronal, there is an area named La Pea, where the
shafts called Pea (202.15 ma.s.l., 150 m deep), Unin (201 ma.s.l.,
420 m deep) and La Pea drift (202.07 ma.s.l.) were also used to extract
cinnabar. It should also be taken into account that underground coal
mining also occurred in the 20th century (some galleries still remain).
Evidence of old coal mining works is also found in the study zone, main-
ly around 2 km NE of El Terronal, where an old gallery is drained
through a pipeline, which nally discharges into the San Tirso River.
After N 40 years of abandonment, currently El Terronal has only a few
derelict buildings and stockpiles of wastethese having been irregularly
disposed of (Gallego et al., 2015). Indeed, most of the mining waste was
encapsulated in an area now covered by a highway constructed in the
2000s.
Finally, the older mining spoil heaps around La Pea (b2 km SW of El
Terronal) are hidden below paved areas. This zone was also exploited
for Hg in the 19th century and the rst half of the 20th century.
3. Geology, hydrology and sampling locations
The study area is located in the NW limit of the Carboniferous Cen-
tral basin, one of the main subdivisions of the Cantabrian Zone of the
Iberian Massif (Lotze, 1945; Julivert, 1967). The area comprises siliceous
breccias, sandstones, lutites and coal (Westphalien age), with an
approx. NW-SE direction, in an anticlinal structure and limited to the
N and S by the La Pea and La Carrera faults, respectively (Fig. 2). The
Hg and As ores are associated mainly with breccias and less importantly
with sandstones. Hg ores are formed by cinnabar and secondarily by
metacinnabar, although native Hg has also been revealed. As is present
mainly in realgar and As-rich pyrites. In the same paragenesis, minerals
such as pyrite, melkinovite, sphalerite, guadalcazarite, galena, stibnite,
marcasite and, exceptionally, native gold have been also described
(Luque and Gutirrez-Claverol, 2010). The deposits are epigenetic of hy-
drothermal origin and are strongly controlled by tectonics, as revealed
by the presence of the main accumulations wherever two or more pref-
erential fracturing directions are located, thus generating more perme-
able areas that facilitated the ow of the ore geouids (Luque, 1985).
Regarding surcial hydrology, covering 7 km2, the San Tirso River
basin can be considered a small drainage basin. The average slope is
21.57 and the total height difference is 476.53 m. The river stretches
a total of 4.5 km from the headwaters to the river mouth in the Caudal
River, and the average slope of the channel is 9.17. The characteristics
of the rainy regime, including some episodes of heavy rain per year,
and the presence of highly erodible materials imply that under certain
circumstances this river becomes a torrential channel (Mndez Pazos,
2013). Remarkably, some of the abandoned facilities of the metallurgy
industry (the derelict chimney, pipes, etc.) are currently preferential
pathways for water runoff ow (Fig. SM1). Therefore, the rainwater
passes through the chimney, where it picks up dust and soot; it then
ows through the main area of the facilities and nally ows into the
San Tirso River (Fig. SM1).
Regarding hydrogeological behaviour, breccias and sandstones are
intergranular porosity aquifers, whereas lutites and coal can be consid-
ered aquicludes and aquitards respectively. It should be noted that the
mining activity severely affected the initial hydrogeological conditions,
promoting a notable increase in permeability and modications of
822 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830

Fig. 1. a) Geographical location of study area. b) Location of mining works and metallurgical plants in the San Tirso River valley (modied from Luque and Gutirrez-Claverol, 2006). 1.
Unin shaft; 2. Pea shaft; 3. Subsidence area; 4 to 12 Mining galleries; 13. Former La Unin metallurgical plant: 14. Esperanza shaft; 15. Schultz gallery (crosscut); 16. Former
Terronal metallurgical plant; 17. Former workshops and ofces; 18. La Flecha gallery (Drift); 19. Duct; 20. Chimney; 21. Spoil heaps, 22. Housing; 23. Coal mining gallery; and24.
Extent of underground coal mining.

natural ows and inltration capacity. In fact, in unaltered conditions,
the groundwater ow would head to the San Tirso River; however,
the excavation of mining galleries connecting El Terronal and La Pea
(Fig. 3) led to a preferential groundwater pathway that currently drains
a signicant amount of the total ow towards the La Pea drift, as this is
located at a lower altitude (202.07 ma.s.l.). In the same context, inltra-
tion and recharge were modied as a result of the connection of most of
the underground mines to the surface (galleries, ventilation chimneys,
etc.). Consequently, groundwater recharge arises from rainwater inl-
tration throughout the outcropping surfaces of the aquifers and the
mining works on the western slope of the San Tirso valley. Overall, the
discharge ow heads to the San Tirso River but mainly to La Pea drift.
Currently, part of the water drained by this drift enters the sewage sys-
tem while the other part ows into the Caudal River.
4. Materials and methods
4.1. General approach
In a rst step, an initial historical study of the mining activities in the
zone was performed, specially focused on old cartographic information
and on work plans reporting the location of the galleries in the Hg and
coal mines. These old cartographic data revealed that El Terronal and
La Pea were linked up underground. In a second stage, eldwork was
carried out in order to determine the geology and hydrology of the
area, with special emphasis on the identication of anthropogenic inu-
ences affecting the hydrologic and hydrogeological conditions of the
river and surroundings. Taking into consideration all the information
gathered and the ndings of a previous study of waste disposal at El
 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830
Fig. 2. Geological map with location of water and sediment sample stations (see Section 4 for sample description).
Terronal (Gallego et al., 2015), sampling campaigns for soil, surface
water, groundwater, and sediments were designed.
4.2. Soil sampling and analytical techniques
Soil characterisation was carried out using a regular sampling net
(25 25m). Within each mapping unit (Fig. 4), ve systematically dis-
tributed 1.5-kg samples were taken from the top 20 cm of soil using an
Edelman Auger (three increases in each sampling point). The incre-
ments sampled were screened in situ through a 2-cm mesh to remove
rocks, gravel, and other large material.
Each increment was dried at room temperature in order to minimise
the loss of volatile contaminants (Hg). It was then disaggregated by a
roller to facilitate sieving through a 2-mm screen. Materials with a
grain-size over 2 mm were vigorously washed and rubbed off to recover
ne particles. After drying, an aliquot sample of the ve included in each
grid was taken for thorough mixing, and the composite samples were
then homogenised and ground to b 100 m by means of an agate mortar.
823
For multi-element analysis, 0.250-g representative subsamples were
leached by means of an Aqua regia digestion (HCl + HNO3) in an
Anton Paar 3000 microwave. The samples were diluted and ltered.
Eight trace elements (As, Cd, Cr, Cu, Hg, Ni, Pb, Zn) were quantied by
means of an Inductively Coupled Plasma Mass Spectrometer (ICP-MS
7700, Agilent Technologies) using IDA (Isotopic Dilution Analysis).
In addition, a sequential extraction (modied from Tessier et al.,
1979) was done for selected subsamples of waste/soil (see below). In
brief, the following ve fractions were obtained: exchangeable (extract-
ed in MgCl2); carbonate-bound (extracted in a buffer of CH3COONa/
CH3COOH); Fe-Mn oxide-bound (extracted in NH2OHHCl); organic
matter-bound (extracted in several steps using H2O2, HNO3 and
NH4NO3); and the residual fraction (extracted in aqua regia). In the
same set of samples, As species were extracted with 1 M H3PO4 +
0.1 M ascorbic acid in a microwave vessel. The extracts were diluted
and ltered. As species were separated in an As-separation column
(Agilent) using a mobile phase of 2 M PBS/0.2 M EDTA (pH = 6.0) in a
1260 Innity High Pressure Liquid Chromatograph (HPLC) coupled to
824 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830

Fig. 3. Geological structures and hydrogeological system of the San Tirso River valley (unscaled schematic vertical section) and sampling stations for surface waters in the study rivers.

a 7700 ICPMS (Agilent Technologies). Hg species were extracted using a
solution of 7.6% HCl and 10% 2-mercaptoethanol in an ultrasonic bath,
then separated by means of a ZORBAX Eclipse XDB C18 (2.1 mm i.d.
50 mm, 5 m) column; 0.06 M ammonium acetate, 5% methanol
and 0.1% 2-mercaptoethanol (pH = 6.8) served as mobile phase. The
extract was then centrifuged, diluted, and analysed in the same HPLC-
ICPMS system described above.
4.3. Surface water - groundwater monitoring
4.3.1. Samples
Water sampling was carried out in various stages between January
2014 and September 2015. Eighty-one samples corresponding to conti-
nental freshwater (streams), groundwater and other freshwater (run-
offs) were collected. Several criteria were used to select the location of
the sampling stations (Figs. 2 and 3):
a) Inputs to San Tirso River
- Station CMD was used to characterize the drainage of the abandoned
coal mine. This stream ows directly into the San Tirso River.
- Station TB corresponds to a tributary of the eastern area of the valley.
This zone holds the waste of the old mining and metallurgical plants
that was encapsulated more than fteen years ago, although this iso-
lation system is not currently being monitored.
- Station ET is directly related to the run-off at El Terronal and collects
most of the leachates of the site. The water quality at this site was
also monitored in the following stations:
ETCh is located at the end of the main chimney, and so it was used to
study the characteristics of the water owing through it.

Fig. 4. Detailed view of El Terronal and sampling grid.

 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830
GwET1 and GwET2 (7 and 10 m depth respectively) are two wells that
were drilled in the main platform of El Terronal.
- GwB was used to take samples of the drainage of the typical host
rock of the Hg ores (siliceous breccia exploited in the past). This sta-
tion is accessible only at low waters levels, usually in the summer.
Downstream there are two other layers of mineralised siliceous
breccias but they are not noticeable in the riverbed because they
are covered by anthropic llers.
b) Samples from the San Tirso River
- Station T1 is located in the San Tirso River, upstream of the Hg
mining- metallurgical works, and upstream of any other non-
natural disturbance of the river hydrochemistry.
- Station T2 is located immediately after the drainage of the old gallery
of the coal mine (CMD), which ows into the main river-way, and it
is useful to evaluate the effect of this discharge in the river. Stations
T3 to T5 are also in the main course of the river but downstream of El
Terronal. Station T3 is located downstream of the last input from El
Terronal, whereas T4 (100 m downstream of the outcropping brec-
cias sampled in GwB) is appropriate to evaluate the effect of the nat-
ural discharge from the breccias. Finally, T5 is located very close to
the conuence of the San Tirso and Caudal Rivers and thus it was
used to examine the nal water composition before the river mouth.
c) Samples in the vicinity of the San Tirso River mouth
- The Caudal River was sampled at three stations: C1 located upstream
of the San Tirso River mouth, C2 located in the river mouth itself, and
C3 downstream of the river mouth.
- Finally, sample LP was taken at the sewerage plant located at the
bottom of the San Tirso valley. In this area, sewer pipes collect
water run-off from the western part of the former mining area and
this is channelled to a wastewater treatment plant downstream.
4.3.2. Analytics
Samples for As and metal analyses were collected in polyethylene
bottles and then ltered by means of a 0.45-m pore size mesh; the
rst few millilitres were used for rinsing and were then discarded The
ltrate was transferred to clean polyethylene bottles and then acidied
with HNO3 to pH b 2 prior to storage below 4 C until analysis. Eight
trace elements (As, Cd, Cr, Cu, Hg, Ni, Pb, Zn) were quantied by IDA-
ICPMS, whereas As speciation was also determined by HPLC-ICPMS
(both analytical methods described above in Section 4.2). In addition,
ion analyses by means of ion chromatography (883 Basic IC plus,
Metrohm) were conducted during high and low water periods (January
2014 and September 2015). pH and conductivity were measured in the
eld using a laptop Hach Lange HQ (HQd Series Field Probe Kit).
In order to complement the study of the inuence of the drainage
from Hg and coal mining facilities on the nal composition of the river
water, we measured the 87Sr/86Sr ratio. For this purpose, a Neptune
Plus multicollector ICPMS instrument (Thermo Electron Corporation,
Bremen, Germany) was used. The instrument was operated in low-
resolution mode (m/Dm = 400). The sample introduction system
consisted of an auto-aspirating low-ow (100 ml min1) PFA nebuliser
(Savillex, AHF Analysentechnik, Tbingen, Germany) mounted on a
combined cyclonic/double-pass spray chamber made of quartz glass.
4.3.3. Gauging
The main gauging station used was T5 (Figs. 2 and 3, before the
mouth of the San Tirso River in the conuence with the Caudal River).
Triplicate measurements were done using a Salinomadd Salt Wave
Flowmeter (MADD Technologies) by one point dilution of NaCl and con-
trol of the salt cloud.
4.4. Sediments
A collection of 11 composite samples (2 kg each) of the sediments
were taken with a sediment sampler along the San Tirso River and its
825
mouth, where it ows into the Caudal river. Sampling stations were lo-
cated within a short distance of each other (see Fig. 2), and the cam-
paign was carried out from the downstream to the upstream to avoid
resuspension of sediments caused by sampling activities. All samples
were immediately stored at 4 C until analysis. Fractions below 2 mm
were prepared and analysed analogously to soil samples (see
Section 4.2). Enrichment factors (Sierra et al., 2015) were used to eval-
uate the impact of As and the other PTEs on the sediments.
5. Results and discussion
5.1. Soils and waste
Univariate statistical analysis of the results obtained after the 8-
element analyses of the composite samples of surcial soils taken in
the 34 mapping units is shown in Table 1. Taking into account the max-
imum values measured and also the average value of the 34 mapping
units, As and Hg were found to be the main contaminants at the site.
In fact, many mapping units presented anomalous values for these
metals, i.e. higher than SSLs (Soil Screening Levels), depending on soil
use (Table 1), thereby indicating a large area of affected soils (Fig. 5).
In this regard, the extremely high concentrations of Hg and As point to
a blend of natural soil aggregates with waste. In addition, but less im-
portantly, other elements, such as Cd, Pb and Cu, exceeded SSL concen-
trations in some mapping units, whereas Ni and Cr did not show any
relation with the contaminants. Multivariate statistics (hierarchical
cluster, Figure SM2) revealed a high correlation between As and Hg,
which was also associated with Cd, Pb and Cu. These observations are
consistent with the ore sources described above.
Regarding spatial distribution, the highest concentrations were de-
tected close to the metallurgical facilities (Fig. 5), which is consistent
with the presence of abandoned pure waste stockpiles.
Overall, high concentrations of As and Hg were found to be associated
with the metallurgical facilities and linked to the presence of pure wastes.
These wastes are located mainly in cells E2, F2 and G3 (Fig. 5) and thus
this area showed extremely high concentrations of As and a very low pro-
portion of natural soil. In contrast, C3 is the natural soil area where the
main water run-off accumulates to be later released into the river after
crossing the pure waste zone. Following the preceding considerations,
the main residues in the pure waste area and the soil in C3 were subject-
ed to sequential extraction and speciation analyses (Table 2).
As leachability was notably higher than Hg in two of the three main
waste studied (Table 2). In addition, the presence of As(III) and the ne
grain-size of these residues suggest higher potential toxicity and mobility
both for mechanical and chemical weathering (Gallego et al., 2015). These
observations, together with the higher contents of As in the waste, explain
why the content of this element (Table 1) was generally one order of mag-
nitude higher than that of Hg in the soil. In addition, alternative Hg speci-
ation measurements carried out by thermal desorption also revealed the
predominant presence of stable chemical forms, such as Hg bound to or-
ganic matter, HgS and Hg-FeS2 (Rumayor et al., 2017). On the other
hand, when compared with waste data, As in soils (Table 2) was less mo-
bile, and As (V) predominated, thereby indicating an oxidisation transfor-
mation in soil. In fact, the water runoff in the area sweeps along waste
particles that leach (very slowly for Hg, more rapidly for As) into the aque-
ous media and also accumulate in the soil or river sediments (see below).
Finally, underground pollution was also assessed in the bore-
holes where the GwET1 and GwET2 wells were drilled. The results
(Table SM1) suggest the presence of residues and polluted soil in
both locations up to at least 5 m below the surface.
5.2. Water
5.2.1. Hydrochemistry
Irrespective of the location, the pH was alkaline (78.5) in all the
samples from the San Tirso River (Fig. 6), in San Tirso inputs, and in
826 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830

Table 1
Univariate statistics of the eight elements analysed in the 34 mapping units. The third column shows the SSLs (Soil Screening Levels established by Spanish law) for Industrial (Ind), Urban
(Urb), and Natural Soil (NS) uses (BOPA, 2014). The fourth column indicates the number of mapping units (of the 34 studied) with concentrations exceeding SSLs.

Element Statistics (mg/kg) SSLs (mg/kg) Mapping units above SSLs

Av. Std. Dev. CV (%) Max. Min. Ind Urb NS Ind Urb NS

As 28,547 94,272 330.2 392,238 13 200 40 40 23 32 32

Cd 1.1 1.3 116.5 5.1 0.1 200 20 2 0 0 7
Cr 46 22 48.5 122 20 10,000 10,000 10,000 0 0 0
Cu 35 15 43.4 701 15 4000 400 55 0 0 5
Hg 759 1298 171.1 4498 0.8 100 10 1 22 30 33
Ni 35 15 43.1 59 2 6500 4150 65 0 0 0
Pb 150 85 56.5 421 71 800 400 70 0 1 34
Zn 343 161 46.9 598 49 10,000 4550 4550 0 0 0

samples from the Caudal River stations. The only exceptions were
found within El Terronal, where the GwET2, GwET1, ETCh and ET sta-
tions revealed acidic or slightly acidic pH values (ranging between
5.5 and 5.8 for GwET2, between 6.3 and 6.9 for the others), and
also CMD water. However, all values in the river remained relatively
constant throughout the year. Given these observations, it can be
concluded that the inuence of the inputs on the pH is negligible
(Fig. 6).
As regards electrical conductivity (EC), this parameter presents a
clear seasonal variation in the San Tirso River, such that the highest
values correspond to the dry periods (Fig. 6). In addition, the conductiv-
ity trend seemed to increase markedly downstream of CMD. Within El
Terronal, EC reached the highest values in well GwET2 (20802270
S/cm), whereas the surface runoff waters circulating inside the facili-
ties (ET) and in the old chimney (ETCh) showed values ranging from
540 to 930 S/cm and 370 to 590 S/cm, respectively. The lowest values
measured (98280 S/cm) corresponded to groundwater from the
mineralised breccias (GwB). Finally, the average EC of the Caudal River
was around 450 S/cm.
As regards hydrochemistry, the San Tirso River water can be consid-
ered Ca(Mg)-bicarbonate type in high water seasons, whereas in low
waters it becomes sulfated in T2 and the other downstream stations
(Fig. 7). On the other hand, water in wells GwET1 and GwET2, water
run-off (ET and ETCh) at El Terronal, and CMD are sulfated. Finally, the
drainage of the ore host rock (GwB) is Ca-bicarbonate, and water in
the main river (Caudal) after the San Tirso River mouth is also Ca-
bicarbonate.
In fact, in low waters (high conductivities and Ca(Mg) SO4 type for
T2 to T5), the hydrochemistry of the river water seems to be mainly in-
uenced by the CMD, and in a lesser degree by El Terronal discharge and
all the valley run-off. However, some effects on the river sediments
should not be discarded (see below). On the contrary, in high waters, di-
lution implies a lower impact of CMD and a subsequent decrease in the
EC (at least in T2, Fig. 6.) and classication as Ca(Mg)HCO3 type (see T2
samples in Fig. 7).
5.2.2. Metals and metalloids
The low impact of Cd, Cr, Cu, Ni, Pb and Zn on soils implies that the
evaluation of the PTE content in the water samples should focus on Hg
and As. However, dissolved Hg was signicant only inside El Terronal;
i.e. in the ETCh (from 11 to 387 g/l) and ET (19 to 129 g/l) stations
and less importantly in the shallow groundwater sampled in the
GwET1 and GwET2 wells (between 1 and 5 g/l). Hg below the limit
of detection in the San Tirso River is consistent with the low Hg mobility
described above. Therefore, as regards water, Hg pollution is limited
mainly to the abandoned site.
With regard to As concentrations, most samples were clearly above
international standards. Spanish law (BOE, 2015) establishes that the
average concentration per year must be b50 g/l in rivers, b 10 g/l in
groundwater, and b 25 g/l in other surface waters. In this context, the

Fig. 5. As (mg/kg) distribution based on an ordinary kriging with logarithmic transformation and no trend elimination.
 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830 827

Table 2
Chemical speciation and estimation of As and Hg leachability calculated after sequential extraction. High leachability is attributed to the sum of extractable and carbonate-bound fractions
as dened in Section 4.1, whereas low leachability is ascribed to the sum of the other three sequential extraction fractions.

Sample % As Leachability % As species % Hg leachability % Hg species

High Low As (III) As (V) As (org) High Low Hg (inorg.) Hg (org)

a
Waste As-rich soot 14.1 85.9 85.1 14.9 0.0 0.9 99.1 100.0 0.0
Stuppa 14.7 85.3 80.3 19.7 0.0 1.1 98.9 100.0 0.0
Flue dusta 0.0 100.0 0.0 100.0 0.0 3.5 96.5 100.0 0.0
Soilb 1.8 98.2 2.0 98.0 0.0 3.6 96.4 100.0 0.0
a
Data taken from Gallego et al., 2015.
b
Average values of ve samples taken in cell C3.

San Tirso River revealed As concentrations that generally fell below 50
g/l in the sampling stations upstream of El Terronal (T1 and T2). In con-
trast, all the stations downstream (T3, T4 and T5) showed high or ex-
tremely high values (hundreds of times above the limit) (Fig. 8).
These ndings suggest that high concentrations of As are related to
the mining and metallurgical waste piles that have been exposed to
weathering since closure of the mine. As minerals, primarily oxides,
present in solid wastes (Gallego et al., 2015) are washed away by runoff
and then physically (sediments, see below) or chemically (dissolved)
enter the riverbed.
Regarding river ow (gauging station: T5), during periods of low
water the content of As was higher than in the rainy season (an accept-
able correlation is shown in Fig. 9). This observation is explained by the
dilution effect caused by a higher runoff from areas not affected by min-
ing. It must be pointed out that the average pollutant load released to
the river can be assessed by means of data on ow and As concentra-
tions. In our case, this rough estimation revealed a value exceeding 6
tons of As per year.
On the basis of the concentrations and ows to the San Tirso River
(Table 3), the following conclusions can be drawn:
- The values measured for the tributary (TB) and the coal mine drain-
age (CMD) water are below the legal limit of 50 g/l. However, al-
though traceable As is present, thereby revealing the high
background in the area (as occurred with station T1 in the river). Al-
though the TB and CMD ows are not very high, CMD slightly affects
the quality of the river water, as previously shown in Figs. 6 and 7.
- Groundwater runoff owing through the mineralised siliceous brec-
cia (GwB) exceeds 3000 g/l of As in low waters. This observation
may imply a signicant As supply to the San Tirso River; however,
the ow is very low (Table 3). Indeed, when comparing As concen-
trations in the river upstream of the breccia (T3) with downstream
values (T4) (see Fig. 8), no signicant differences are found. It
must be remarked again that this low amount of groundwater ow
draining into the San Tirso River is attributable to a notable modi-
cation in the groundwater ow direction caused by mining activity;
i.e. the ooding of the mine galleries currently promotes groundwa-
ter ow mainly towards La Pea drift.
- Regarding surface runoff circulating throughout El Terronal, ET pre-
sents high As concentrations (N 30,000 g/l in low waters) This ob-
servation is clearly in agreement with the fact that this area
collects the water runoff that washes the waste piles. Data obtain-
ed inside El Terronal also reveal high As- contents; i.e. ETCh above
1000 g/l, which is consistent since a signicant amount of As- and
Hg-enriched soot and condensate remains in the duct. Also, the shal-
low groundwater sampled in wells GwET1 and GwET2 also show
concentrations well above the maximum allowable limit (N 100
and 250,000 g/l of As respectively for GwET1 and GwET2). The ex-
treme values of GwET2 are expected, as this well was drilled at the
end of the course followed by water runoff, very close to the waste
stockpiles.

Fig. 6. Temporal variation of pH (top) and electrical conductivity (bottom) in the San Tirso
River sampling stations and its main inputs. Note that CMD, ET and GwB were not always
sampled due to difcult access.
A mass balance calculation in the river is not possible given that not
all the run-offs from the valley can be determined. In this regard, during
the sampling campaigns on the eastern side of the valley, evidences of
mining waste exposure were found in spite of all the mining wastes in
this area should have been conned. Regarding the samples taken in
the vicinity of the San Tirso River mouth, those corresponding to sewage
(LP) present a relatively constant As content irrespective of the season
(between 125 and 170 g/l). The source of As in this water can be attrib-
uted to the runoff from the western hillside of the San Tirso valley and,
remarkably, to the drainage water of La Pea drift, which is the main
discharge of groundwater ooding the mine galleries in the study
area, as explained in Section 3. This (ground)water is channelled to sew-
erage networks and therefore is the only output of the study area cur-
rently being treated in a wastewater plant.
On the other hand, the concentration of As at sampling point C1
(Caudal River upstream of the San Tirso River mouth) was lower than
the maximum permitted limit (50 g/l), whereas in C2 and C3 it far
exceeded (N100 g/l in low waters) the limit. C2 and C3 As-contents
are attributed to the As-rich water of the tributary of the Caudal River
(San Tirso River), in spite of the difference in ows (average ow of Cau-
dal River is 12 m3/s). It could also be hypothesised that the As content in
828 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830
Fig. 7. Piper diagrams of the water analysed.
the Caudal River is increased by the shallow groundwater circulating
through spoil heaps located under paving in La Pea area, although
this was not proven in this study.
5.2.3. Strontium isotopes
The 87Sr/86Sr ratio for the samples taken from the San Tirso River (T2
and T5) and CMD water showed slight deviations (0.7108 0.0008).
The sample corresponding to groundwater from the mineralised brec-
cias (GwB) showed a decrease in this ratio (0.7094 0.0001). The dif-
ference between the value of the GwB sample and those of T2, T5 and
CMD could be explained as an effect of the lithology interacting with
groundwater. In fact, according to Clark and Fritz (1997), the abundance
of 87Sr is directly linked to the geochemistry of potassium; i.e., K-rich
Fig. 8. Temporal evolution of As concentration in the San Tirso River during the
hydrological year 20142015 (Sep-14, Jun-15 and Sep-15 low waters; Dec-14 and Mar-
15, high waters). As(V) was the only species found in all the samples, with the
exception of T1, where As(III) accounted for below 2% of the total As concentration. Dot
line indicates the legal limit of 50 g/l.
rocks will have high 87Sr contents and this could be reected in the
87
Sr/86Sr ratio of water in equilibrium with K-rich materials. In our
case, the waters corresponding to river and CMD samples were in con-
tact with lutitic rocks, whose K content is higher than that of the
mineralised siliceous breccias corresponding with GwB.
Although a greater number of samples would be required to draw
reliable conclusion, these ndings conrm that the isotopic composition
of the water of the San Tirso River is inuenced by the groundwater
drainage of the coal mine (the rst main input in the high course of
the river). However, the drainage of El Terronal site or the GwB station
does not imply a change in the isotopic signature. In addition, the simi-
larity of 87Sr/86Sr values between stations T2 and T5 also rule out the
Fig. 9. Correlation between ow and concentration of As in the San Tirso River at station
T5.
 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830 829

Table 3
As content and ows in the main inputs to the San Tirso River.
Sample station As (g/l) Flow (l/s)
High waters Low waters High waters
CMD 8 20 4 1
TB 3 20 5 0.1
ET 9300 33,500 5 0.5
GwB 400 3400 b0.1
presence of signicant unknown inputs of deeper groundwater or sur-
cial discharges that may have been excluded from this study.
5.3. Sediments
The distribution of PTE concentration in the sediments of the San
Tirso River shows a similar pattern to that found for water, with a re-
markable increase after the discharge from El Terronal (Table 4). In
this case, enrichment factors were calculated to evaluate the signi-
cance of the impact; a conservative element (Cr) with no relationship
to the waste and soil pollution at El Terronal was selected as a reference
to study the enrichment of As (40 times the average), Hg (99 times) and
Pb (13 times).
These results reveal the marked impact of As- and Hg-rich solids and
leachates in the San Tirso River on sediment quality. In contrast, the ef-
fects on the sediments of the Caudal River are negligible.
6. Conclusions
The mining and metallurgical activity developed in the past in the
valley of the San Tirso River (Mieres, Spain) has extensively contami-
nated soil, water, and sediments. This area is a paradigmatic example
of the combined effect of the abandonment (N 40 years) of mining/
industrial facilities, inappropriate waste disposal, and modication of
the natural geology and hydrology of the site on several environmental
compartments.
In the study area, there is a signicant accumulation of Hg in the for-
mer browneld (waste stockpiles). The subsequent impact on soil has
not implied high Hg-concentration in the water of the San Tirso River.
This observation is attributed to the low geochemical mobility of Hg.
However, mechanical transport has promoted marked Hg enrichment
in river sediments. In contrast, the exceptionally high As contents in
waste and As mobility implies the continuous release of this contami-
nant in dissolved formsa process that has remarkably affected the
water and sediments of the San Tirso River, and consequently those of
the Caudal River. The San Tirso River transports N 6 tons of As each
year; most of this can be attributed to water runoff at El Terronal, al-
though minor sources such as a nearby coal mine drainage, high natural
backgrounds, drainage of former Hg-mining galleries by permeable
Low waters
strata, and the runoff along the whole valley should also be considered.
As-concentrations in the river show a clear seasonal variation, with
lower values in period of high water ow. This observation can be ex-
plained by dilution effects. In this context, the hydrochemical composi-
b0.1 tion of the river varies between calcium magnesium bicarbonate in
times of high water ow, and sulfated calcium magnesium in low
water ow. This variation is due to mainly to the dilution effect; for in-
stance, the sulfated water from the coal mines has a greater inuence
when the river ow is low. Overall, the pollution of the San Tirso River
derives mainly from surface runoff and shallow groundwater from El
Terronal and its surroundings, whereas the contribution of deeper
groundwater is less signicant. This is explained because the mines of
El Terronal and La Pea are connected underground, in such a way
that groundwater drainage takes place mainly through the La Pea gal-
lery. In this context, the impact of As on the sewage systems indicates
the need to control the discharge of (ground) water from the western
part of the valley (La Pea area).
On the basis of the former considerations, it is concluded that the re-
mediation of soil, water and sediment pollution of the San Tirso River is
urgent, irrespective of the development of a detailed risk assessment.
The rst measure could be the elimination of the waste stockpiles as
these are the main source of As; however, this approach should require
a previous and systematic study of physic-chemical parameters and the
quantication of waste volumes, together with the implementation of
prevention measures such as a previous isolation of the waste (see
Conesa and Schulin, 2010 and Palanques et al., 2014 for similar cases).
The control of water drainage and soil remediation will be subsequently
required, together with treatment of polluted sediments in the San Tirso
River.
Acknowledgments
This work was partially supported by the research project LIFE+ In-
novative and Demonstrative Arsenic Remediation Technologies for Soils
(I + DARTS). We would like also to thank the Environmental Assay Unit
of the Scientic and Technical Services of the University of Oviedo for
their technical support. Carlos Boente obtained a grant from the
Formacin del Profesorado Universitario program, nanced by the
Ministerio de Educacin, Cultura y Deporte de Espaa.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.scitotenv.2017.08.135.

Table 4
As concentration in the sediments of the San Tirso (SD-1 to SD-9) and Caudal (SD-10 and SD-11) riverbeds. Enrichment factors were calculated as follows: EFPTE = (PTE / Cr)downstream /
(PTE / Cr)upstream, which is the proportion between the upstream and downstream concentration ratios of the PTE and the normalizer (Cr in this case).

Elements San Tirso River Caudal River

Upstream El Terronal Downstream El Terronal Enrichment Upstream STirso R. mouth Downstream STirso R. mouth Enrichment
(mg/kg)a (mg/kg)b Factor (SD-10, mg/kg) (SD-11, mg/kg) Factor

As 65 3895 39.7 11 15 1.1

Cd 0.1 0.2 1.3 b0.1 b0.1 ND
Cr 41 62 1.0 18 23 1.0
Cu 28 37 0.9 11 12 0.9
Hg 2.0 320.3 103.9 0.3 0.4 1.1
Ni 31 45 1.0 14 17 0.9
Pb 15 232 10.2 8 9 0.9
Zn 51 94 1.2 38 49 1.0
a
Average values of samples SD-1 to SD-4.
b
Average values of samples SD-5 to SD-9.
830 B. Gonzlez-Fernndez et al. / Science of the Total Environment 610611 (2018) 820830
References
Bhattacharya, P., Welch, A.H., Stollenwerk, K.G., McLaughlin, M.J., Bundschuh, J.,
Panaullah, G., 2007. Arsenic in the environment: biology and chemistry. Sci. Total En-
viron. 379:109120. http://dx.doi.org/10.1016/j.scitotenv.2007.02.037.
BOE, Boletn Ocial del Estado, 219, September 12, 2015. Water quality and environmen-
tal criteria (Spain). https://www.boe.es/boe/dias/2015/09/12/pdfs/BOE-A-2015-
9806.pdf, Accessed date: June 2017.
BOPA, Boletn Ocial del Principado de Asturias, 91, April 21, 2014. Generic reference
levels for heavy metals in soils from Principality of Asturias, Spain. https://
sede.asturias.es/bopa/2014/04/21/2014-06617.pdf, Accessed date: June 2017.
Bundschuh, J., Bhattacharya, P., Nath, B., Naidu, R., Jack, N., Luiz, R.G., Lena, Q.M., Kyoung-
Woong, K., Jiin-Shuh, J., 2013. Arsenic ecotoxicology: the interface between
geosphere, hydrosphere and biosphere. J. Hazard. Mater. 262:883886. http://
dx.doi.org/10.1016/j.jhazmat.2013.08.019.
Chakraborti, D., Mahmudur, R.M., Murrilla, M., Das, R., Siddayya, Patil S.G., Sarkar, A.,
Dadapeer, H.J., Yendigeri, S., Ahmed, R., Das, K.K., 2013. Environmental arsenic con-
tamination and its health effects in a historic goldmining area of the Mangalur green-
stone belt of Northeastern Karnataka, India. J. Hazard. Mater. 262:10481055. http://
dx.doi.org/10.1016/j.jhazmat.2012.10.002.
Clark, I.D., Fritz, P., 1997. Environmental Isotopes in Hydrogeology. Lewis Publishers, New
York.
Conesa, H.M., Schulin, R., 2010. The CartagenaLa Union mining district (SE Spain): a re-
view of environmental problems and emerging phytoremediation solutions after f-
teen years research. J. Environ. Monit. 12:12251233. http://dx.doi.org/10.1039/
c000346h.
Dory, A., 1984. Le Mercure dans les Asturies. Revue Universelle des Mines, de la
Mtallurgie, des Travaux publiques, des Sciences et des Arts appliques l'industrie.
XXXII (3me srie) pp. 145210.
Foulds, S.A., Brewer, P.A., Macklina, M.G., Haresignb, W., Betsona, R.E., Rassner, S.M.E.,
2014. Flood-related contamination in catchments affected by historical metal mining:
an unexpected and emerging hazard of climate change. Sci. Total Environ. 476477:
165180. http://dx.doi.org/10.1016/j.scitotenv.2013.12.079.
Gallego, J.R., Esquinas, N., Rodrguez-Valds, E., Menndez-Aguado, J.M., Sierra, C., 2015.
Comprehensive waste characterization and organic pollution co-occurrence in a Hg
and As mining and metallurgy browneld. J. Hazard. Mater. 300:561571. http://
dx.doi.org/10.1016/j.jhazmat.2015.07.029.
Gallego, J.R., Rodrguez-Valds, E., Esquinas, N., Fernndez-Braa, A., Af, E., 2016. Insights
into a 20-ha multi-contaminated browneld megasite: an environmental forensics
approach. Sci. Total Environ. 563-564:683692. http://dx.doi.org/10.1016/
j.scitotenv.2015.09.153.
Haffert, L., Craw, D., 2008. Mineralogical controls on environmental mobility of arsenic
from historic mine processing residues, New Zealand. Appl. Geochem. 23:
14671483. http://dx.doi.org/10.1016/j.apgeochem.2007.12.030.
Hudson-Edwards, K., Macklin, M., Taylor, M., 1997. Historic metal mining inputs to Tees
River sediment. Sci. Total Environ. 194195:437445. http://dx.doi.org/10.1016/
S0048-9697(96)05381-8.
Julivert, M., 1967. La ventana del Ro Monasterio y la terminacin meridional del Manto
del Ponga. Trab. Geol. 1, 5976.
Lefebvre, R., Hockley, D., Smolensky, J., Glinas, P., 2001. Multiphase transfer processes in
waste rock piles producing acid mine drainage 1: conceptual model and system char-
acterization. J. Contam. Hydrol. 52:137164. http://dx.doi.org/10.1016/S0169-
7722(01)00156-5.
Loredo, J., Ordez, A., Gallego, R.J., Baldo, C., Garca-Iglesias, J., 1999. Geochemical character-
isation of mercury mining spoil heaps in the area of Mieres (Asturias, northern Spain).
J. Geochem. Explor. 67:377390. http://dx.doi.org/10.1016/S0375-6742(99)00066-7.
Loredo, J., Ordez, A., lvarez, R., 2006. Environmental impact of toxic metals and met-
alloids from the Mun Cimero mercury-mining area (Asturias, Spain). J. Hazard.
Mater. A136:455467. http://dx.doi.org/10.1016/j.jhazmat.2006.01.048.
Loredo, J., Soto, J., lvarez, R., Ordez, A., 2007. Atmospheric monitoring at abandoned
mercury mine sites in Asturias (NW Spain). Environ. Monit. Assess. 130:201214.
http://dx.doi.org/10.1007/s10661-006-9389-0.
Lotze, F., 1945. Zur Gliderung der Varisziden der Iberischen Meseta. Geotekt. Forsch. 6,
7892.
Luque, C., 1985. Las mineralizaciones de mercurio de la Cordillera Cantbrica. Doctoral
thesis. University of Oviedo (unpublished results).
Luque, C., Gutirrez-Claverol, M., 2006. La minera del mercurio en Asturias. Rasgos
histricos.
Luque, C., Gutirrez-Claverol, M., 2010. Riquezas geolgicas de Asturias.
Mndez Pazos, A., 2013. Peligrosidad asociada a la escombrera El Terronal: Anlisis del re-
lieve y de la solubilidad de metales pesados mediante extracciones secuenciales. Uni-
versity of Oviedo (unpublished results).
Moreno-Jimnez, E., Manzano, R., Esteban, E., Pealosa, J., 2010. The fate of arsenic in soils
adjacent to an old mine site (Bustarviejo, Spain): mobility and transfer to native ora.
J. Soils Sediments 10:301312. http://dx.doi.org/10.1007/s11368-009-0099-4.
Ordez, A., Alvarez, R., Charlesworth, S., De Miguel, E., Loredo, J., 2011. Risk assessment
of soils contaminated by mercury mining, Northern Spain. J. Environ. Monit. 13:
128136. http://dx.doi.org/10.1039/c0em00132e.
Ordoez, A., Silva, V., Galn, P., Loredo, J., Rucandio, I., 2014. Arsenic input into the catch-
ment of the river caudal (Northwestern Spain) from abandoned Hg mining works:
effect on water quality. Environ. Geochem. Health 36:271284. http://dx.doi.org/
10.1007/s10653-013-9566-3.
Palanques, A., Grimalt, J., Belzunces, M., Estrada, F., Puiga, P., Guilln, J., 2014. Massive ac-
cumulation of highly polluted sedimentary deposits by river damming. Sci. Total En-
viron. 497498:369381. http://dx.doi.org/10.1016/j.scitotenv.2014.07.091.
Rieuwerts, J.S., Mighanetara, K., Braungardt, C.B., Rollinson, G.K., Pirrie, D., Azizi, F., 2014.
Geochemistry and mineralogy of arsenic in mine wastes and stream sediments in a
historic metal mining area in the UK. Sci. Total Environ. 472:226234. http://
dx.doi.org/10.1016/j.scitotenv.2013.11.029.
Rumayor, M., Gallego, J.R., Rodrguez-Valds, E., Daz-Somoano, M., 2017. An assessment
of the environmental fate of mercury species in highly polluted brownelds by
means of thermal desorption. J. Hazard. Mater. 325:17. http://dx.doi.org/10.1016/
j.jhazmat.2016.11.068.
Sierra, C., Menndez-Aguado, J.M., Af, E., Carrero, M., Gallego, J.R., 2011. Feasibility study
on the use of soil washing to remediate the As-Hg contamination at an ancient min-
ing and metallurgy area. J. Hazard. Mater. 196:93100. http://dx.doi.org/10.1016/
j.jhazmat.2011.08.080.
Sierra, C., lvarez-Saiz, J.R., Gallego, J.R., 2013. Nanoltration of acid mine drainage in an
abandoned mercury mining area. Water Air Soil Pollut. 224:1734. http://dx.doi.org/
10.1007/s11270-013-1734-7.
Sierra, C., Ordez, C., Saavedra, A., Gallego, J.R., 2015. Element enrichment factor calcula-
tion using grain-size distribution and functional data regression. Chemosphere 119:
11921199. http://dx.doi.org/10.1016/j.chemosphere.2014.10.024.
Smedley, P.L., Kinniburgh, D.G., 2002. A review of the source, behaviour and distribution
of arsenic in natural waters. Appl. Geochem. http://dx.doi.org/10.1016/S0883-
2927(02)00018-5.
Tessier, A., Campbell, P.G.C., Bisson, M., 1979. Sequential extraction procedure for the spe-
ciation of particulate trace metals. Anal. Chem. 51:844851. http://dx.doi.org/
10.1021/ac50043a017.