Journal of Environmental Management 273 (2020) 111148

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: http://www.elsevier.com/locate/jenvman

Research article

Assessment of water resources pollution associated with mining activity in

a semi-arid region
Caroline S. Santana *, Diango M. Montalva�n Olivares , Vinnícius H.C. Silva , Francisco H.
M. Luzardo , Fermin G. Velasco , Raildo M. de Jesus
Santa Cruz State University, Department of Exact and Technological Sciences, Jorge Amado Highway, Ilh�eus, Bahia, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Mining, although relevant and indispensable for human socioeconomic development, is considered one of the
Heavy metals most polluting anthropogenic activities. Water resources are the most vulnerable environmental compartment to
ICP OES the direct impacts of mining, especially in the semi-arid regions. In these regions, mining activity constitutes an
Pollution indices
important challenge in the management of water resources; since its impacts can be maximized by the adverse
Ecological risk
meteorological conditions. This study aimed to assess the level of contamination in water resources of three
important mining areas in a semi-arid region, where approximately 70,000 people live. The concentrations of
eleven heavy metals in sediment, surface and groundwater samples were determined by ICP OES. The results
obtained for water samples indicated significant contamination by Cd, Pb, and U, based on the limits established
by Brazilian and international regulatory legislation. In the case of sediment samples, higher concentrations of
Cr, Cu, Ni, and V were observed. Pollution indices (PLI, CF and Igeo) revealed moderate to extreme contamination
mainly along the Jacar�e and Contas rivers. The Pearson correlation, Principal Component, and Hierarchical
Cluster analyses were performed to identify patterns in the distribution of elements and common sources of
pollution. The results suggested that the concentrations of Co, Cr, Cu, Fe, Mn, Ni, and V were mainly related to
mining activities and, to a lesser extent, natural sources. In the case of Cd and Pb, contamination may have an
important contribution from fertilizers use, whereas Zn has a mixed source of both lithogenic and anthropogenic
origin. On the other hand, the high concentrations of U, specifically in groundwater samples, were associated
with geogenic causes. Although the potential ecological risk values indicated a low ecological risk; other sedi­
ment quality indices (TEL, PEL, ΣTU, and PEL-Q) revealed that there is 25% of the probability that the content of
heavy metals in water resources near the mining areas induces adverse toxic effects on aquatic organisms.

1. Introduction Mining, although relevant and indispensable for human socioeco­

nomic development, accelerates natural processes and increases the
Water resources are threatened not only by climate change but also likelihood of releasing toxic elements at higher rates in adjacent areas
by increasing levels of pollution, which has become a serious environ­ (Satapathy et al., 2009). The generation of large amounts of solid waste
mental problem on a global scale (Satapathy et al., 2009). In the current and effluents, derived from mining activities, represents a potential
scenario, where there is a growing demand for freshwater, uncontami­ source of pollution of soils and water bodies since the trace metals
nated and disease-free water supplies are increasingly scarce (Burritt contained in them can be dispersed by the action of natural factors (wind
and Christ, 2018). Trace elements such as heavy metals can occur in and/or rain) or accidently (García-Lorenzo et al., 2012; Lottermoser,
both surface and groundwaters as a result of the impact of several 2007). For these reasons, mining is considered as one of the human
anthropogenic activities (e.g. agriculture, energy producing, industry, activities with the greatest negative impact on the environment (Para­
manufacturing, mining, etc.). The contamination of water resources by guassú et al., 2019).
heavy metal poses a real risk to aquatic environments and human health Water resources are the most vulnerable environmental compart­
due to the toxicity, persistence, and biological accumulation of these ment to the direct impacts of mining. According to Burritt and Christ
pollutants (Lo
�pez et al., 2019; Reis et al., 2019; Yang et al., 2018). (2018), this economic activity is one of the most water-intensive

* Corresponding author.
E-mail address: [email protected] (C.S. Santana).

https://doi.org/10.1016/j.jenvman.2020.111148
Received 14 March 2020; Received in revised form 9 June 2020; Accepted 27 July 2020
Available online 4 August 2020
0301-4797/© 2020 Elsevier Ltd. All rights reserved.
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

industries, in which water is a crucial resource and, typically, is the Concentration values were compared not only with the Brazilian law but
primary medium that carries pollutants into the wider environment. also with international reference values established by the United States
Eventual discharges of contaminated water from mines and acid Environmental Protection Agency (U.S. EPA), World Health Organiza­
drainage formation play an important role in the release and mobility of tion (WHO) and the Initiative for Responsible Mining Assurance (IRMA).
various chemical contaminants. (Angelovi�cova � and Fazeka�sov� a, 2014; As most heavy metal ions are deposited in the sediment by adsorption,
Camizuli et al., 2014). These processes favor the transport of heavy hydrolysis, co-precipitation, and other physical and chemical processes;
metals and are one of the severest and most widespread causes of water this environmental compartment was also studied in the main aquatic
resource pollution (Jarvis and Younger, 2000). Geochemical impacts, ecosystems. Powerful statistical analysis tools such as Pearson Correla­
generated by mining activities, can affect the quality of surface and tion, Principal Component Analysis (PCA), and Hierarchical Cluster
groundwater at a regional scale even decades after mine closure (Baeten Analysis (HCA) were used to identify the relationships between the el­
et al., 2018). ements and possible common sources of pollution. In addition, different
In semi-arid regions, water resources are very sensitive to seasonal pollution and risk indices were calculated to evaluate the level of
climate variability, therefore freshwater resources are scarce. Ground­ contamination and the possible ecological risk in these aquatic
water is the most important and often the only natural resource for water ecosystems.
supply (Wheater et al., 2010). In these regions, mining activity consti­ It is important to note that such extensive research on water re­
tutes a major challenge in the management of water resources; since its sources in the region has not been carried out before; therefore, the data
impacts can be maximized by the adverse meteorological conditions. and analyses reported in this study will certainly be useful to help
The ore piles generally placed outdoors, undergo greater reactivity, evaluate, discuss and reduce the impact on water quality derived from
which contributes to greater mobility of the toxic substances (Perlatti mining activities.
et al., 2015; Navarro et al., 2008). In these regions, the transport of
pollutants occurs both through runoff in surface waters and through the 2. Materials and methods
underground flow (Zeng and Wu, 2009). For this reason, the geochem­
istry of lake and river sediment has the capacity to register changes in 2.1. Study area
the water resources at the time of their accumulation (Liu et al., 2014).
The diversity of geological environments gives Brazil one of the This study was conducted in the iron and uranium mining areas
biggest mineral potentials in the world. The country has a rich reserve of belonging to the municipality of Caetit�e and in the vanadium mining
important metals such as bauxite, iron, manganese, niobium, and nickel. area in Marac� as. As mentioned before both municipalities are located in
The state of Bahia (Northeast, Brazil) is the federative unit with the the semi-arid region of the state of Bahia (Northeast, Brazil). The region
largest number of municipalities located in the semi-arid region is characterized by an annual average temperature that varies between
involved in mining activities. This is the case of the municipalities of 22 and 23 � C. However during the summers (December to March),
Caetit� e and Maraca�s, located in the central-south and central-eastern temperatures can reach up to 35 � C. Since the dry season lasts most of
region of Bahia, 636 km and 250 km, respectively, of Salvador (capital the year (April to October), there are prolonged drought events and
of Bahia). The extraction and current processing of uranium in Brazil are short periods of intense rainfalls in the summer, as a consequence the
carried out in Caetit�e since in this municipality there is the only uranium annual precipitation in the region is approximately 800 mm/year. Pre­
mine in operation in Latin America (INB, 2020). In addition, this mu­ cisely, these climatic conditions are responsible for the rivers in the
nicipality stands out for its amethyst and iron reserves, which make region intermittent courses and their beds remain dry for most of the
Bahia the third-largest iron producer in the country. On the other hand, year, except in some sections that retain small portions of water.
in the municipality of Maraca �s, vanadium has been exploited since In physiographic terms, the mining area studied is situated within
2014. This mine was projected to produce between 10,000 and 15,000 the basins of the Sa ~o Francisco and Contas rivers, which are two of the
tonnes of vanadium pentoxide per year and is one of the lowest-cost most important basins of the entire northeast region of Brazil. The
producers of the material in the vanadium market (Largo Resources uranium mine is specifically located in the Riacho da Vaca basin, which
Ltd, 2020). occupies an area of 94.5 km2 and comprises some important rivers such
In recent years, Brazil has witnessed five failures in tailings dams as the Fundo and Sa ~o Joa ~o rivers, which are indirect tributaries of the
associated with bauxite, iron, gold and tin mining. The events in Contas River. The iron mine situated in the southern part of the mu­
Mariana and Brumadinho (Minas Gerais/Brazil) in 2015 and 2019, nicipality of Caetit� e is inserted in the Carnaíba de Dentro river basin,
respectively, were classified as true environmental disasters. These ac­ which flows in the NE-NOW direction, occupying an area of 2700 km2
cidents caused the contamination of the main rivers in that region, the and belongs to the Sa ~o Francisco river basin. The Grande River, which is
interruption of the supply of potable water to residents, the destruction part of this basin, passes near the iron mining area and flows into the
of towns and the death of more than 260 people (Do Carmo et al., 2017; Ceraíma reservoir, the main water supply point in the region. Finally,
World information service on energy, 2019). the vanadium mine, in the municipality of Marac� as, is located north of
Based on the facts that: (1) the mining sites studied are located near the widest section of the Contas River and 4 km east of one of its trib­
inhabitant areas and the Contas River, one of the most important in the utaries, the Jacar�e River.
Brazilian semi-arid region, and (2) the regional population suffers from According to Fo et al. (2008), the geological characteristics of the
freshwater scarcity and it depends only on artesian wells and limited region favored the formation of fissured bedrock aquifers. This type of
sources of surface water for domestic use, irrigation and animal feed aquifer is characterized by a network of well connected and widely
(Margulis et al., 2002); it can be stated that mining activity in the region dispersed fractures, thus resulting in a highly permeable zone and good
implies a significant risk to the integrity of the water resources and aquifer storage. For that reason, groundwater constitutes the main
human health. supply of fresh water for more than 70,000 people, who live in these
In this context, the aim of this study was to assess the impacts of mining areas and depend on these water resources for agricultural, do­
mining on the water resources in an area of high vulnerability, which is mestic, and industrial activities.
essential to assure water quality, and therefore environmental safety. To Caetit�e region is located in the Lagoa Real granite-gneissic complex,
fulfill the purpose of this work, the concentration of eleven toxic ele­ which has a significant number of albite-rich (NaAlSi3O8) rocks that
ments: Cd, Co, Cr, Cu, Fe, Mn, Ni, Pb, U, V, and Zn in both surface and contain uranium mineralization. Lobato et al. (2015) pointed out that,
underground waters located in the area of influence of three important the uranium deposits are associated with rocks dominated by plagio­
semi-arid mining regions of the state of Bahia were determined by clase (albite � oligoclase), which have high levels of sodium enrichment
Inductively Coupled Optically Emission Spectrometry (ICP OES). and Fe2þ-dominant phases. Additionally, uraninite crystals are mainly

2
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

associated with other minerals such as andradite [Ca3Fe2(SiO4)3], bio­ information on the sampling points can be found in Table S1 in the
tite [K(Mg,Fe) 3(AlSi3O10) (F,OH) 2] and calcite (CaCO3) (Lobato et al., supplementary material.
2015; Machado, 2008). The orebodies contain a total reserve of ca. 112, The water and sediment samples were collected and preserved ac­
000 metric tons of U3O8 with a uranium content of up to 0.7% (INB, cording to the criteria of EMBRAPA (2011) and the sample collection
2020). The southern portion of the municipality of Caetit�e hosts and preservation guide of CETESB (2011). Surface water samples were
important iron deposits of iron with a total reserve of 370 million tons. collected at a depth of up to 30 cm from the water slide and stored in
Hematite (Fe2O3) and itabirite (banded-quartz hematite) are the main 500 mL polyethylene bottles, previously washed with detergent and
minerals exploited in this region, due to their high Fe content: 66% and deionized water. In each case, the pH of the water samples was
32%, respectively. These minerals also contain fractions of aluminum measured with pH test strips. To properly preserve the samples and
and manganese oxides, as well as, silicates, which are higher in itabirite minimizes any physical modification and/or chemical reaction; samples
minerals (SiO2: 51%; Al2O3: 1.1% and MnO: 0.36%) (ENRC – Eurasian were acidified by adding 3 mL L 1 of concentrated nitric acid (HNO3).
Natural Resources Corporation, 2012). Other minerals such as amethyst, Sediment samples were collected at the depositional margins of the
biotite and rhodonite [CaMn3Mn(Si5O15)] are also found in this region. rivers with the help of a plastic spatula and stored in airtight bottles.
Amethyst, in fact, is mined in the district of Brejinho das Ametistas, Both water and sediment samples were put in an ice bath and kept below
south of Caetit�e. 20 � C until the analysis.
Geologically, Maraca�s region forms part of the Capivara complex,
which is located in the boundary between Gavi~ ao and Jequi�e blocks.
This complex is an arched structure 2.5 km long and 1 km wide (Brito 2.3. Sample preparation and analysis by ICP OES
et al., 2018). Anorthosite, gabbro, norite and other mafic to ultramafic
rocks, which are iron-rich igneous rocks, predominates in this region Although acidification with nitric acid preserves water samples for
and are associated with chromite (FeCr2O4) and vanadiferous titano­ up to six months, all samples were digested and measured within one
magnetites minerals. Brito et al. (2018) reported that the chromium (III) month after sample collection. For the estimation of the heavy metal
oxide (Cr2O3) content in chromite minerals is 32%, while for Fe it is content in both water and sediment samples the aqua regia digestion was
19%. The vanadium oxide (V2O5) content in the vanadian-tittanian used following the standard procedure ISO 11466 (1995). This method
magnetite minerals ranges from 0.33% to 0.98%, which is considered is a widely used analytical procedure and is standardized for water and
a high content of this metal. Actually, the total reserve (23 million tons) sediment analysis (S�
�can�car et al., 2000). The ISO standard 11,466 de­
and the quality (based on the high content of V2O5) of the ore found in scribes the use of an assembled reflux digestion system with
the region distinguish the Maraca �s mine worldwide, which produce in temperature-controlled heating apparatus and a substantial amount of
average more than 14.6 tons/year (Largo Resources Ltd, 2020). aqua regia (HNO3/HCl) and sample (Santoro et al., 2017).
Before the acid digestion, the surface water and groundwater sam­
ples were filtered using a microfiltration membrane (pore size 0.45 μm)
2.2. Sample collection and preservation as recommended by the CETESB (2011) and EMBRAPA (2011). This
pretreatment procedure is capable of sieving out colloidal particles and
Samples of surface water and sediment were collected from the main suspended material at low pressure value (1–2 atm). A representative 10
water bodies in the region (Contas, Caldeira ~o, Jacar�
e, Fundo, and Sa ~o mL aliquot of each water sample was transferred to a digestion tube to
Joa
~o rivers), as well as from the Ceraíma reservoir and small lakes in the which 0.20 mL of nitric acid (HNO3, Merck) and 0.10 mL of hydrochloric
vicinity of the mines. Groundwater samples were collected from artesian acid (HCl, Merck) were added, as recommended by (ISO 11466, 1995;
wells in the main communities located around the mines. As a result, a United States Environmental Protection Agency - USEPA, 1994a). The
total of 48 samples: 27 surface water (SW), 8 groundwater (GW), and 13 samples were subsequently heated in a digester block at 85 � C for 3 h
sediments (SD) were collected during the autumn, which corresponds to and kept under reflux. After acid digestion, the samples were cooled and
the end of the dry season. The 29 sampling locations were georeferenced transferred to 50 mL Falco tubes, where 15 mL of ultrapure Milli-Q
with a digital device Global Positioning System (GPS) (Fig. 1). The main deionized water was added. Finally, the samples were centrifuged at

Fig. 1. Map of the studied area and sampling points.

3
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

2000 rpm for 5 min in a centrifuge (Solab Científica, model SL-700). areas.
In the case of sediment samples, they were initially dried in an oven Since background concentration values are essential to interpret the
at 85 � 2 � C to reduce the water content. A portion of each sediment geochemical data, choosing the right ones constitutes an important
sample was ground in a porcelain mortar and then sieved through a 100- challenge. However, the lack of data from a specific area makes it
mesh sieve (Bertel Ltda, 300 x 2”, inox 304). 0.25 g subsamples were difficult to correctly evaluate the level of sediment pollution. In this
placed in a digestion tube and 2.5 mL of concentrated HNO3 and 1.5 mL work, the average concentration of heavy metals in the sediments re­
of 37% concentrated HCl were added The digestion tubes were sealed ported in previous studies conducted in Brazilian river basins were
and allowed to stand at 20 � C for 16 h (USEPA, 1994b; International selected as background values for both CF and Igeo calculations. The
Organization For Standardization - ISO 11466). Then, the samples were values for Cr (18.0 mg kg 1), Cu (6.80 mg kg 1), Pb (5.70 mg kg 1), and
placed in the digester block and heated at 130 � C for 2 h. Afterward, the Zn (13.0 mg kg 1) were taken from Filho et al. (2015), while Fe (1.63%),
samples were cooled, completed with 15 mL of Milli-Q deionized water Mn (92.5 mg kg 1), and Ni (9.40 mg kg 1) from Sabadini-Santos et al.
and centrifuged at 2000 rpm for 5 min. (2009). Co (2.00 mg kg 1) and V (8.40 mg kg 1) were taken from
The quantification of the elements of interest (Cd, Co, Cr, Cu, Fe, Mn, Sodrzeieski et al. (2018). These works were chosen because they
Ni, Pb, U, V, and Zn) in the collected samples was performed using an established background levels for heavy metals in sediments for regions
Inductively Coupled Plasma Optical Emission Spectrometer (model 710 with similar geological and climatic characteristics to that of our study
ES Varian, Australia) under the operating conditions reported in and for rivers not influenced by anthropogenic activities.
Table S2. The ICP OES was calibrated using two analytical grade The contamination factor (CF) is one of the most commonly used
chemicals solutions: a mono-elemental standard Merck (Germany) and a quantitative indices to assess the pollution levels in a given environ­
multi-elemental high-purity standard (USA). In all cases, linearity was mental compartment. The CF is defined as the ratio between the con­
considered satisfactory since R2 values were greater than 0.9994. The centration of the metal analyzed in the sample (CS) and the
limit of quantification (LOQ) for each element was determined from ten concentration of the same metal reported as a background reference
analytical blanks, which were prepared using ultrapure Milli-Q deion­ value (CBG) as shown:
ized water (resistivity: 18.2 MΩ cm 1) under the same analytical con­
CS
ditions as the samples. The spectral lines, as well as the analysis range for CF ¼ (2.1)
CBG
each metal analyzed, are shown in Table S3. Samples that exceeded the
upper limit of the calibration curve, for some metal, were appropriately According to the CF value, four possible levels of pollution are
diluted when required. All measurements were made in triplicate. defined: low contamination (CF < 1); moderate contamination (1� CF
To ensure the reliability of the metal analysis on the spectrometer, < 3); severe contamination (3 � CF < 6) and extreme contamination
three standard reference materials (NIST 1643f: Trace metals in water, (CF � 6).
NIST 8704: Buffalo River sediment, and IAEA-158: Trace metals and methyl The pollution load index (PLI) allows to assess the level of contam­
mercury in marine sediment) were analyzed. Subsamples of these refer­ ination considering the simultaneous presence of n pollutants. It is
ence materials were prepared and measured under the same conditions calculated from CF values as shown in eq. (2.2). PLI value equal to zero
as the water and sediment samples. The results of this quality test did not indicates a non-polluted area, PLI equal to one indicates the presence of
show a significant difference between the certified and measured con­ baseline level of pollutants and a value greater than one suggests pro­
centration of the selected metals according to the t-Student and Z-score gressive contamination of the area (Liu et al., 2017):
tests at a significance level of 95% (see Table S4). pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
The percentage recovery of the selected elements in the water ranged PLI ¼ n CF1 � CF2 � … � CFn (2.2)
from 95% to 116%, which are within the permitted values (Gray et al., The geoaccumulation index (Igeo) enables the assessment of
2015). However, for sediment samples the values ranged between 59% contamination by also comparing the metal concentrations with those
and 109% with Co (59%), V (71%) and Fe (77%) being the elements reported in the background, but using a different mathematical
with the lowest recoveries. Since the concentration of a particular heavy approach and introducing the factor 1.5, which is the correction factor
metal determined after aqua regia digestion is considered as the of background matrix due to lithospheric effects (Ali et al., 2016):
pseudo-total concentration, an underestimation of the total amount of
CS
metal in an environmental matrix may occur (Santoro et al., 2017). Igeo ¼ log2 (2.3)
Therefore, unsatisfactory recovery rates may be obtained (Krause et al., 1:5⋅CBG
1995). Difference up to 50% may be found between total and aqua regia Igeo values can be interpreted as: Igeo � 0 ─ uncontaminated; 0 � Igeo
extractable content, because the strength of the solvent is unable to < 1 ─ low to moderately contaminated; 1 � Igeo < 2 ─ moderately
completely dissolved silicate lattices and refractory crystalline oxides to contaminated; 2 � Igeo < 3 ─ moderately to heavily contaminated; 3 �
which some metals such as Cd, Co, Fe and others may be bound (S ��can�car Igeo < 4 ─ heavily contaminated; 4 � Igeo� 5 ─ heavily to extremely
et al., 2000). In general, the lower the recovery, the greater the associ­ contaminated and Igeo � 5 ─ extremely contaminated.
ation of metals with silicate minerals (Silva et al., 2016). However, total
content and aqua regia extractions can provide comparable results, since 2.4.2. Assessment of the ecological risk
the difference is covered by the relative standard deviation (RDS), i.e. Two types of Sediment Quality Guidelines (SQG) for freshwater were
the variation of the measurement results, which typically varies between used: the Threshold Effect Level (TEL) and the Probable Effect Level
1% and 15% (Santoro et al., 2017). (PEL) to aid in the interpretation of sediment quality and evaluate the
extent to which the concentration of sediment contaminant can
2.4. Assessment method of contamination in sediments adversely affect the aquatic organisms. TEL refers to the concentration
below which adverse effects would be rarely expected; whereas PEL
2.4.1. Estimation of pollution indices indicates the chemical concentration above which adverse effects occur
Sediment is not only a major carrier of contaminants but also a po­ frequently (Saleem et al., 2013). Based on these two guidelines values
tential source of contaminants in aquatic systems; therefore this envi­ three main ranges of adverse biological effects were defined:
ronmental compartment has been widely used as an ecological indicator minimal-effects range (<TEL); possible-effects range (TEL-PEL) and
to estimate the quality of the aquatic ecosystems (Liu et al., 2017). The probable-effects range (>PEL). The values adopted for TEL and PEL are
contamination factor (CF), pollution load index (PLI), and the index of shown in Table 1. They were taken from the Environmental National
geoaccumulation (Igeo) were used to estimate the anthropogenic Council of Brazil (CONAMA 344, 2004). In addition, the PEL quotient
contribution of the metals analyzed in the sediments of the studied

4
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

(PEL-Q) was determined for the n heavy metals analyzed (see Eq. (2.4)) 3. Results and discussion
to evaluate the possible biological effects derived from the combination
of toxic elements. 3.1. Metal concentration in surface and groundwaters
X
n
ðCi =PELi Þ
PEL Q¼ (2.4) The water samples exhibited a slightly acid average pH value of 6.5,
i
n which is within the interval of values accepted by the CONAMA 357
(2005) legislation. The pH values for each sampling point are shown in
where Ci is the concentration of the element i in the sediments and PELi Tables S5 and S6 (Supplementary Material). The lowest pH value was 5
is the guidelines values for each element (Soliman et al., 2015). PEL-Q is and was found at sampling points: 4 (SW-04 and GW-04) and 16 (SW-16
classified in the following grades: <0.10; 0.11–1.50; 1.51–2.30 and > and GW-16), which are the closest to the vanadium and uranium mines,
2.31, which denote 10%, 25.5%, 50% and 76% probability that toxicity respectively. The level of acidity observed in these bodies of water may
of heavy metals in sediments will have adverse biological effects for the be associated with eventual discharges of mine water, which are highly
aquatic communities (amphipod, benthic species, mollusks, etc.). acidic, and the acid drainage of solid wastes. These processes can change
Additionally, the potential acute toxicity of pollutants was calculated as the chemical condition of rivers and streams (Jarvis and Younger, 2000).
the numerical sum of the toxic units (ΣTUs), which are defined as the The general results of heavy metal concentration in surface and
ratio of a given concentration to PEL value (Saleem et al., 2013). groundwater are shown in Table 2. The same information specified for
To further assess the risk posed by heavy metals in sediments of the each sampling point can be found in Tables S5 and S6 in the Supple­
aquatic systems studied, the Potential Ecological Risk (PERI) was mentary Material. The average concentration of studied metals in sur­
calculated as the sum of individual potential ecological risk (Er) for all face waters followed a decreasing order of Fe > Mn > Zn > Pb > U > Cu
the analyzed elements, as proposed by Hakanson (1980): > V > Cr > Co > Ni > Cd. For groundwater, there were variations in the
X
n X
n order of heavy metal concentration. Uranium ranked third, displacing
PERI ¼ Eri ¼ Tri ⋅CF i (2.5) Zn and Pb, while Cd rose two positions.
The quality of surface and groundwater was evaluated considering
i i

the guidelines values established by the Brazilian legislation CONAMA

where CF is the contamination factor and Tri is the toxic-response factor
357 (2005) and international laws such as World Health Organization -
for a given metal (see Table 2). For Co, the same value as Cu was used,
WHO (2017), Initiative for responsible mining assurance (IRMA) (2018)
whereas for Fe and Mn the one of Zn was used (Hakanson, 1980; Hilton
and USEPA (2019). In the case of CONAMA law, the reference values for
et al., 1985; Ullah et al., 2019; Wei et al., 2016). Table 1 also provides
freshwater classes II and III were adopted. Class II freshwater is intended
the classification standards for the potential ecological risk of heavy
for human consumption (after conventional treatment), protection of
metals.
aquatic communities, irrigation, fishing, and recreation. On the other
hand, class III can be used for human consumption (after advanced
2.5. Statistical analysis
treatment) and the consumption of farm animals.
As can be seen, the concentration values for Co, Cr, Ni, V, and Zn
In order to understand the dynamics of the distribution of metals and
obtained in the surface waters of the region did not exceed the reference
identify potential sources of heavy metals, a robust statistical analysis
values established by the environmental laws. The highest concentration
was carried out. Multivariate analyses such as Pearson’s correlation
values of Co, Cr, Ni, and V were found near human settlements (SW-18,
coefficient, Principal Components Analysis (PCA) and Hierarchical
SW-22, and SW-29.2) and in some water resources near the uranium and
Cluster Analysis (HCA) were performed on the data set using the STA­
vanadium mines. Interestingly, the highest values of Zn were observed
TISTICA 10 software. Correlation analysis at the significance levels of p
at SW-25 and SW-27 which may be associated with runoff from farmed
< 0.05 and p < 0.01 was used to find the relationships between the
areas. In the case of the groundwater, the concentrations of these metals
various types of heavy metals, as well as to determine possible common
also remained within the permitted values, with GW-01, GW-02, GW-16,
sources of pollution. PCA was applied to better explain the correlation
and GW-21 being the samples with the highest concentrations.
structure of the variables and to make data easy to explore by reducing
93% of the samples analyzed showed Cu concentrations lower than
its dimensionality. The main composition of each principal component
the guidelines values. However, the content of this metal in samples SW-
(PC) was assumed to be given by those variables with loadings greater
18 and SW-22 was almost twice the limit allowed for freshwater class III
than the absolute value of 0.50. In the case of HCA, the Euclidean dis­
recognized by CONAMA, which means that these water resources are
tance was used as a measure of similarity between sampling points.
inappropriate even for consumption animals. The groundwater sample
These statistical tools allowed us to distinguish the natural and
(GW-21) collected in the same region as SW-22 also exhibited a Cu
anthropogenic origins of heavy metals in water and sediment samples,
concentration value above the limit for freshwater class II, indicating
improving the interpretation of the data.
that it should not be used either for human consumption or for crop
irrigation.
Table 1 As already noted, Fe and Mn were the most abundant elements in
Guidelines values for sediment and grades of potential ecological risk.
both surface and groundwater (CONAMA, 2009). Concentrations above
Cr Cu Ni Pb V Zn the Fe and Mn limits reported by CONAMA (class II) and IRMA were
TEL 37.3 35.7 18 35 n/a 123 observed in 93% and 41%, respectively, of surface water samples. In
PEL 160 108 42 112 57* 315 addition, six samples exceeded the limit value of Fe in freshwater class
TRV 26 16 16 31 n/a 110 III, and three for Mn. The highest concentrations of Fe (97,200 μg L 1)
Tr 2 5 5 5 2 1
and Mn (846 μg L 1) were observed in SW-29.2, which is 15 km from the
Grades of Ecological Risk
Low Eir <40 PERI <150 iron mine and is, therefore within its indirect area of influence. In
Moderate 150 � PERI <300
addition, this point is near the amethyst mine in the south of Caetit�e. The
40 � Eir < 80
exploitation of this semiprecious stone, which owes its violet tone to
Considerable 80 � Eir < 160 300 � PERI <600
impurities of Fe and other transition metals as Mn, could probably be the
High 160 � Eir < 320 PERI �600
main reason why the concentration of Fe and Mn found in this sample
Very high Eir �320 –
were 15 and 5 times greater than the average concentration in the
TRV - Toxicity Reference Value (USEPA, 1999); n/a: Not available; *:taken from region.
USEPA (2015) In general, high contents of Fe and Mn were found near the uranium

5
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

Table 2
Heavy metal concentrations in the water resources of the studied region and guideline values.
Metals Surface water (μg L 1) Groundwater (μg L 1
) Guidelines values (μg L 1)

Range Mean � Range Mean � CONAMA 357 World Health Organization - Initiative for responsible mining USEPA
S.D. S.D. (2005) WHO (2017) assurance (IRMA) (2018) (2019)

Class Class
II III

Cd < LOQ – 3 1.3 � 0.2 < LOQ – 1.1 � 0.3 1 10 3 5 5

2.9
Co 0.6–13 2.2 � 0.3 0.7–1.2 0.9 � 0.1 50 200 – – –
Cr < LOQ – 11 2.2 � 0.4 < LOQ – 1.4 � 0.5 50 50 50 50 100
4.6
Cu < LOQ – 24 3.5 � 1.1 < LOQ – 2.4 � 1.2 9 13 2000 1000 1300
9.6
Fe 161–97,200 6342 � < LOQ – 283 � 12 300 5000 – 300 –
799 283
Mn 6.2–846 178 � 36 2–186 44 � 22 100 500 – 100 –
Ni < LOQ – 12 2.1 � 0.5 < LOQ – 0.8 � 0.5 25 25 70 40 –
3.6
Pb 6.5–40 19 � 2 < LOQ – 15 � 3 10 33 10 10 15
26
U < LOQ – 54 5.1 � 2.6 < LOQ – 80 � 15 20 20 30 – 30
132
V 1.3–20 3.6 � 0.7 1.1–3.9 2.1 � 0.3 100 100 – – –
Zn 17–78 38 � 3 17–77 36 � 7 180 5000 – 3000 5000

S.D.: Standard Deviation; LOQ: Limit of Quantification.

mine area (SW-13 and SW-16) and in human settlements in the Caetit�e 2003).
region (SW-18, SW-19, SW-22, and SW-25). This could be associated The mean values of Pb in surface water (19 μg L 1) and groundwater
with the presence in the region of granite-gneissic rocks containing iron (15 μg L 1) exceeded the permitted values established by CONAMA
rich minerals such as biotite and andradite. However, it has been re­ (class II), World Health Organization - WHO (2017), Initiative for
ported that wastes from uranium mine activities contain high concen­ responsible mining assurance (IRMA) (2018), and USEPA (2019) (see
tration of oxides of iron (FeOT) and pyrite (FeS2). The oxidative process Table 2). Fig. 2 shows that 78% of the surface waters samples surpassed
of residual pyrite causes the production of acid drainage that leaches Fe these limits and three samples (SW-02, SW-07, and SW-24) showed a
and other metals from the solid phase, thus increasing the risk of concentration above the limit for class III, whereas 100% of the
contamination of the adjacent soil, surface, and underground waters groundwater samples in which Pb was detected (7 samples) exhibited
(Fernandes et al., 2004). Besides mining activities, other anthropogenic higher values than those recommended.
sources, such as the use of pesticides and urban effluents may be The variation observed in the concentration of Pb throughout the
responsible for the high content of these metals in regional water studied region indicates diffuse contamination, suggesting that mining
resources. may not be the only source of contamination. It has been reported that
In the case of groundwater analysis, only the GW-02 sample located solid wastes derived from uranium extraction contain, among other
in the Maraca �s region, had a concentration of Mn above the established metals, a high concentration of 210Pb, which can be easily transported by
limit (CONAMA, class II), which may be related to the weathering of the both the liquid effluents and the acid drainage, thus reaching soils and
rocks, since it tends to cause the leaching of this metal in aquifers water resources near mines (Fernandes et al., 2004). However, the use of
(Postawa and Hayes, 2013). pesticides, the dumping of solid waste and traffic emission are important
60% of surface water samples and 50% of the groundwater samples sources of Pb emissions. In the case of groundwater, the two samples
exhibited Cd concentrations greater than 1 μg L 1, which is the reference with the highest concentration of Pb (GW-02 and GW-21) are far from
value for freshwater class II established by CONAMA, and two of them: mining areas, which could be an indication that the presence of Pb in
SW-24 and SW-29.2 exceeded also the guideline value established by groundwaters may be of a different origin not related to mining.
World Health Organization - WHO (2017). One of the main sources of Fig. 3 displays the concentration of U in the different locations
water contamination by Cd is mining, mainly of non-ferrous metals. studied. Uranium was detected only in five (19%) surface water samples,
However, the deposition of solid waste in the soil, the discharge of from which 60% (3 samples) presented concentrations above the
sewage and agricultural activities constitute also important sources of CONAMA value for classes II and III. Additionally, SW-02 and SW-23
contamination by this metal (Muniz and Filho, 2006). The impact of one also exceeded the limits established by the World Health Organization
or more of these anthropogenic activities would explain the high con­ - WHO (2017) and USEPA (2019). On the other hand, all samples of
centrations of Cd in the water resources studied, especially in the sam­ groundwater in which U was detected (88%) showed high concentra­
ples close to the mining areas (SW-07, SW-24, SW-29.2, GW-12, and tions values, being the mean value: 80 μg L 1, twice the CONAMA
GW-21), as well as in agricultural areas (SW-02, SW-10, GW-01, and (2005) recommended values.
GW-02). Cadmium is not only a highly toxic element but also a human Since Caetit�e is a region where uranium occurs naturally and
carcinogen; therefore, its presence in high concentrations in water re­ Maraca �s is located in a geological area of contact between two blocks,
sources poses a serious risk to both aquatic ecosystems and human high concentration of uranium is expected to occur (Da Silva et al., 2013;
health. Dardenne and Schobbenhaus, 2003; De Queiroz, 2016). Besides, the
Of all the heavy metals studied, the contents of Pb and U in the water prevailing climate conditions in this area favor uranium transport in the
samples analyzed were the most worrying, due to their toxicity and form of uranyl ion (UOþ22 ) by forming carbonated complexes. Since most
radiobiological effects. Lead poisoning is related to various diseases of surface waters have pH values between 6.5 and 8, the solubility of
the nervous system, kidney damage and behavioral disturbances in uranium tends to increase (Fo et al., 2008). Based on these facts, the high
children. Both Pb and U are classified as possible human carcinogens U concentrations detected in surface and underground waters probably
and are associated with blood, bones, lung and stomach cancers (J€ arup, have a natural origin, related to chemical transformations and

6
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

Fig. 2. Lead concentrations in surface (SW) and groundwater (GW) in the analyzed samples.

Table 3
Uranium concentration values in water resources in uranium natural occurrence
areas reported in the literature.
Country Type Concentration Region Reference
(μg∙L 1)

China Surface water Average: 190.4 Guangdong Wang et al.

Province (2019)
(Xiazhuang
uranium
deposit)
India Groundwater �312 State of Coyte et al.
Rajasthan (2019)
Iraq Surface water 0.52–2.38 State of An- Abojassim
Najaf (“Abu- et al.
Shakir” (2019)
Uranium Mine)
Japan Surface water �0.22 Fukushima (20 Somboon
km from the et al.
Fukushima (2019)
nuclear plant)
United Surface Average: 3.8 Navajo nation, Hoover
States water/ (12.5% above the Southwest U.S. et al.
Fig. 3. Uranium concentration in surface (SW) and groundwater (GW) in the Groundwater allowed value) (Uranium (2017)
analyzed samples. Mining Area)
Brazil Surface 0.95–25.6 Caetit�
e (Near Silva et al.
water/ the uranium (2011)
dissolution of ore in these rocks. Groundwater mine)
The average concentrations of U obtained in surface and ground­ Surface 12–98 Caetit�
e This study
water/
water were compared with the concentrations reported in regions rich in
Groundwater
natural uranium deposits, regions affected by nuclear accidents such as Surface 9.9–132 Marac�
as
Fukushima (Japan) and other mining areas (see Table 3). As can be seen, water/
a high concentration of this element in surface and groundwater is Groundwater
commonly found in both mining areas and regions with a natural
occurrence of uranium. However, compared to the previous study by
(see Fig. 4).
Silva et al. (2011), which was also conducted in the Caetit�e region; a
The most abundant elements in the sediments were Fe (15,317 mg
considerable increase in uranium concentrations is revealed.
kg 1) and Mn (297 mg kg 1), as expected. Sediment samples collected at
Jacar�e River (SD-03 and SD-05) and in the confluence with Contas River
3.2. Metal concentration in sediments (SD-06) exhibited the highest concentrations of Fe and Mn. In particular,
in SD-05 the concentrations for both metals were five times the average
The values of heavy metal concentrations in the sediments specified concentrations, which could be a reflection of the impact of anthropo­
for each sampling point are shown in Table S7 (Supplementary Mate­ genic activities, specifically mining, due to the proximity of Jacar�
e River
rial). The average concentrations of the analyzed metals showed the to the vanadium mine.
following decreasing order: Fe > Mn > Zn > Cr > V > Ni > Pb > Cu > Co

7
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

also showed a concentration of Cu higher than TEL. Interestingly, the

concentration of V in both locations was approximately three times the
average concentration, and more than eight times the concentration in
SD-03, which is located before vanadium mine.
Since the measured concentration values for Cd and U s were below
the limit of quantification, these metals were not considered in the
analysis of the sediment samples. This result can be better understood
considering that: (1) the recovery rate of Cd in sediment samples can be
very low, ranging from 15% to 65% (S ��can�car et al., 2000) and (2) ura­
nium is among the elements that may not be detected by using the aqua
regia digestion procedure (Krause et al., 1995). If we also take into ac­
count that Fe and Mn were the most abundant elements in sediments
samples and that uranium deposits in Caetit�e region are associated with
albite, biotite and other plagioclase feldspar (Machado, 2008); low
extraction efficiencies of Cd and U were expected, because metals bound
to silicate lattice or crystalline refractory oxides (aluminum, iron and
manganese oxides), which are not completely dissolved by the aqua regia
Fig. 4. The box and whiskers plots show the distribution of the concentration
digestion, are more challenging to quantify for some methods as ICP OES
of the analyzed metals in sediments. Fe and Mn values were rescaled by (Santoro et al., 2017).
dividing by 1000 and 10, respectively. Boxes depict 25th, 50th (median) and For all sediment samples collected, the contamination factor (CF)
75th percentiles, while the whiskers indicate the minimum and maximum was calculated, which revealed that 90% of the most contaminated
values. Average values (●). samples are found in the Jacar� e and Contas rivers, located approxi­
mately 6.8 km from the vanadium mine. CF values for Cr, Cu, Ni, and V
In the case of Co, Pb, and Zn, no sample showed concentrations that at these critical points are shown in Fig. 5. As can be noted, SD-05 and
exceeded neither the TEL (Threshold Effect Limit) value nor the TRV SD-06 exhibited a severe level of contamination by Cr and Cu, as well as
(Toxicity Reference Value) (see Table 1). The highest concentrations for extreme contamination for V. The other locations fell into the category
these metals were found again in SD-03, SD-05, and SD-06, and also in of moderate to severe pollution for Cr, Cu, and Ni; and extreme to severe
the Contas River (SD-10), as well as in areas near the uranium and iron for V. In the Caetit�e region, only the Sample SD-13 showed a general
mines (SD-13, SD-15, and SD-27). level of moderate to severe contamination for these elements.
On the other hand, for Cr, Cu, Ni, and V some locations exhibited Similar results revealed the analysis of the Index of Geoaccumulation
concentrations higher than the values recommended by CONAMA 344 (Igeo). 31% of the samples, all located in the Caetit�e region, can be
(2004) and USEPA (2015), which could cause some kind of adverse ef­ classified as non-contaminated (Igeo � 0). In general, higher values were
fect on aquatic organisms. The samples collected in the Jacar�e and observed for Fe, Ni, V, and Zn in the vanadium mining region, which
Contas Rivers (SD-05, SD-06, SD-10, and SD-11) exceeded the values of present a moderate to heavy contamination (2 � Igeo < 3). Moderate
TEL and TRV for Cr and Ni. SD-10 and SD-11 fell in the range between pollution (1 � Igeo < 2) was observed for Co, Cr, Cu, and Pb in the same
TEL and PEL, indicating that occasionally adverse ecological effects may area. While for Mn no level of contamination was detected in any case.
occur. In the case of SD-05 and SD-06, the Cr, Ni, and V concentrations The PLI values, calculated by equation (2.2), ranged from 0.0 to 3.6. As
were higher than their respective PEL (Probable Effect Limit) values, and expected, higher PLI values were observed near the vanadium mine (SD-

Fig. 5. Contamination Factors (CF) for Cr, Cu, Ni and V in sediments.

8
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

05 and SD-06). Multivariate PCA and HCA were performed to further understand
Table 4 provides a comparison between the concentrations of heavy and analyze the sources of heavy metals in the water and sediment
metals found in sediments and those reported in other studies in Brazil samples of the studied area. The PCA for water samples reduced the
and worldwide. The highest concentrations of Cr, Ni, V, and Zn found in number of variables to two principal components (see Fig. 6), which
the sediments of this region are above the world average. Although they explains 63.63% of the data variance and showed eigenvalues greater
do not exceed the values reported for the Hindon River (India), Korotoa than one: PC1 (4.86) and PC2 (2.14).
River (Bangladesh) and Pearl River estuary (China), which are known In more detail, PC1 accounts for 44.2% of the variance and exhibited
worldwide for being extremely polluted due to industrial and urban the highest loadings of Co (0.80), Cr (0.90), Cu (0.74), Mn (0.81), Ni
effluents discharges; it should be noted that the maximum concentration (0.87), and V (0.84), as well as a moderate loading of Fe (0.56). This
values of Co, Cr, Ni, Pb, V, and Zn found in this study exceed between 1.2 principal component could best be related to mining activity. PC2 rep­
and 8.7 times those reported for other important rivers in Brazil. Un­ resents 19.4% of the total variance and showed the highest loading of Cd
doubtedly, this constitutes a warning signal because it indicates a ( 0.79) and moderate values for Pb ( 0.54), therefore it can also be
negative impact on the quality of these water resources that: (1) are associated with an anthropogenic source but of a different origin from
located in a semi-arid region where the availability of freshwater re­ that related to mining. The weak loadings obtained for U (PC1: 0.32,
sources is decreasing; (2) are used for domestic and agricultural activ­ PC2: 0.07) and Zn (PC1:0.18, PC2: 0.01) clearly indicates that these
ities by local population. elements are not associated with either PC1 or PC2, suggesting that both
metals certainly originate from a natural source.
The PCA analysis was also applied to the sediment samples, resulting
3.3. Statistical analysis of heavy metal content in water and sediments
samples

Pearson correlation analysis was used in both water and sediment

samples to obtain relevant information on common sources and the
analogous behavior of heavy metals analyzed in this study. The Pearson
correlation matrices corresponding to metals in surface and ground­
water (Table S8) and sediments (Table S9) are shown in the Supple­
mentary Material.
In surface and groundwater, significantly positive correlations (0.84
� r � 0.89, p < 0.01) were found between Cr–Cu–Ni–V, thus indicating
that might have common sources. Likewise, Co–Fe–Mn showed signifi­
cant positive correlations between them (0.71 � r � 0.92, p < 0.01) and
weaker correlation (0.50 � r � 0.60, p < 0.01) with the elements of the
first group. The Cd–Pb pair also exhibited a strong positive correlation
(r ¼ 0.83, p < 0.01), which suggests similarities in the behavior of these
pollutants. On the other hand, U and Zn did not show any significant
correlation with the other metals. These metals likely come from
another anthropogenic source of pollution or from natural sources,
especially uranium, as already discussed.
In the case of sediments, similar results were obtained. The group of
elements (Cr–Cu–Fe–Mn–Ni–V) showed strong positive correlation fac­
tors (0.721 � r � 0.989), reinforcing the idea of a similar origin. Ac­
cording to Venkatramanan et al. (2015), hydrated oxides of Fe and Mn
may be associated with heavy metals, since these metals (Fe–Mn)
constitute potential sinks for heavy metals (Cr, Cu, Ni, V) in aquatic
systems, particularly under oxidation conditions (pH 6–9). This is a
Fig. 6. Principal Component Analysis (PCA) results for analyzed metals in
physicochemical process of adsorption and co-precipitation. For Co, Pb
water samples.
and Zn not significant correlations were observed.

Table 4
Heavy metal concentration range in sediments compared to studies conducted in Brazil and worldwide.
1
Countries Concentration (mg∙kg )

Co Cr Cu Ni Pb V Zn

This study (NE/Brazil) 0.7–11 2.7–96 3.2–38 0.4–61 3–19 1.7–104 7.6–115
ao Francisco estuary (NE/Brazil)a
S~ – 10–82 1–26 2–27 4–16 – 1–57
Dilúvio Stream (SE/Brazil)b 0–2 1–6 2–11 0.5–3.5 2–15 6–12 5–35
Tapacur�a river (NE/Brazil)c – 0.04–6 1–57 – 0.0–1.3 – 4–55
Hindon River (India)d – 42–250 9.4–195 – 5.1–59 – 3.9–85
Korotoa River (Bangladesh)e – 55–183 35–118 37–163 36–83 – –
Pearl River estuary (China)f – – – 39–76 1.9–88 90–136 –
World averageg 25 90 45 20 20 60 95
a
Sabadini-Santos et al. (2009).
b
Sodrzeieski et al. (2018).
c
Aprile and Bouvy (2008).
d
Suthar et al. (2009).
e
Islam et al. (2015).
f
Liu et al. (2011).
g
Zhou et al. (2019).

9
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

again in two principal components accounting for 80.26% of the total

variability and with eigenvalues greater than one: PC1 (5.78) and PC2
(1.44). Based on the information evaluated from this analysis, HCA was
performed and the result is shown as a dendrogram (Fig. 7). As can be
observed, two main groups of elements can be identified in sediment
samples. Cluster 1 is mainly composed of metals whose origin may be
associated with natural sources and mining activities (Co, Cr, Cu, Fe,
Mn, Ni, V, and Zn), whereas cluster 2 includes only Pb, indicating it has
an anthropogenic origin different from that of cluster 1. It is worth
remembering that neither Cd nor U was detected in sediment samples.
Fig. 8 shows the Pearson correlation coefficients between metal
contents in the surface water and in the sediments for those sampling
points where these two matrices were collected. For this analysis, only
those metals that were detected in both sediments and surface waters
(Co, Cr, Cu, Fe, Mn, Ni, Pb, V, and Zn) were considered. 92% of the
sampling points exhibited coefficients close to the unit (r~0.9998), thus
indicating a strong correlation between the metal content in both
environmental compartments and that their presence may be attributed
to common sources. Sampling point 02 showed a less strong correlation
than the others (r ¼ 0.83, p > 0.01), which suggests that there are slight Fig. 8. Pearson correlation coefficient for metal content in water vs. metal
differences in the source of pollution for sediments and waters. content in sediment.

concentrations for these metals that exceeded the world average con­
3.4. Identification of the sources of heavy metals
centration and the guidelines established by the environmental laws. It
is no coincidence that the most contaminated points were precisely these
Based on the results of the Pearson correlation, the PCA and HCA,
locations, which are the closest to mining areas. Heavy metals trans­
three main sources can be identified: (1) mining activities and, to a lesser
ferred, through sorption processes, to the sediments of the Jacar�e River
extent, natural sources may be responsible for the concentrations of Co,
can be spread in south-southeast direction towards the Contas River.
Cr, Cu, Fe, Mn, Ni, and V in both waters and sediments; (2) anthropo­
Likewise, other sampling points near the uranium mine exhibited high
genic activities, but from a different source, such as agriculture may be
concentrations for some of these metals. In general, the high values
related to Cd and Pb concentrations, and (3) geogenic causes are asso­
obtained from pollution indices reflect the impact of mining on these
ciated with the content of U (in water) and Zn.
aquatic systems. In the case of the iron mine, these impacts were not so
Concerning the first group of elements (Co, Cr, Cu, Fe, Mn, Ni and,
evident by the rivers being extremely dry.
V), it is believed that their presence in waters and sediments is associ­
Since high concentrations of Cd and Pb were found in farming areas
ated with natural sources and anthropogenic activities. According to the
(sampling points: 2, 7, 10, 25, and 27); the content of these metals in
literature, vanadium deposits may be associated with platinum group
aquatic systems was associated with the agricultural source. It is well
elements, such as Ni, and with other transition metals such as Co, Cr, and
documented that these pollutants are closely related to the intensive use
Cu (De Paula and Abrah~ ao, 1991; Pereira et al., 2018). The strong cor­
of pesticides and fertilizers; therefore, they are considered marker ele­
relations obtained for the Cu–Cr–Ni–V cluster support the idea of a
ments of agricultural activities (Marrugo-Negrete et al., 2017; Wang
common natural origin. However, these metals together with Fe and Mn
et al., 2015). Based on the correlations obtained and on two important
are contained in the solid mining wastes, can be transported in ionic
facts: (1) Brazil constitutes the fourth-largest consumer of fertilizer and
form through the acid drainage, leaving them available in the environ­
(2) agriculture is a widely extended economic activity in the study re­
ment and, consequently, contaminating the water resources adjacent to
gion; it was concluded that the source of contamination by Cd and Pb is
the mining areas (Fungaro, 2005).
directly related to the application of these agricultural inputs. This does
As shown in this study, the Jacar�e and Contas rivers exhibited
not imply that other anthropogenic activities such as mining and in­
dustrial and domestic wastes (e.g. plastics, electronic devices, paint
pigments, motor-vehicle exhaust) can contribute to increasing the level
of these pollutants in water resources. Although Zn did not show any
strong correlation with these metals and it seems to be have a lithogenic
origin, it is worth pointing out that the application of phosphate fertil­
izers also increased the level of Zn in agricultural soils. For that reason,
the infuence of agriculture on the content of Zn in the aquatic systems
studied should not be ignored.
The geological characteristics of the study area, which is rich in
uranium deposits, justify the availability of this metal in water re­
sources, specifically in groundwater (Conceiça ~o et al., 2002; Dardenne
and Schobbenhaus, 2003). As already discussed, high concentrations of
U in surface and groundwater are common in uranium natural occur­
rence regions. Therefore, higher concentrations were expected near the
uranium mine (sampling points: 14, 16 and 23) and in the confluence of
the Jequi� e and Gavi~ao geological blocks (sampling point: 2). However,
due to uranium is also present in mining waste; the possibility that its
concentration in the waters may be increased by this economic activity
should not be disregarded.
Fig. 7. Hierarchical Cluster Analysis (HCA) results for analyzed metals in
sediment samples.

10
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

3.5. Potential ecological risk (PERI) assessment

As a closure of this study, sediment quality guide indicators (ΣTU and

PEL-Q), as well as the potential ecological risk were assessed to support
the discussion carried out throughout the work. The sum of the toxic
units (ΣTU) exhibited higher values (~3.3) in the samples of Jacar�e and
Contas rivers and smaller values in the Caetit� e region. The PEL-Q index
(Eq. (2.4)) ranged from 0.02 to 0.67 (mean value 0.22), thus indicating
that the combination of these pollutants has a 25% probability of being
toxic and causing mortality of aquatic organisms, low species richness
and liver lesions in some species of fish (Long et al., 1995).
The order of the average potential ecological risk for each metal (Er)
was: Cu > Co > Pb > Ni > Cr > V > Zn > Mn > Fe. The values for all the
metal were less than 40; therefore they pose a low potential risk. PERI
values ranged from 3.03 to 97.82, with an average value of 31.45. All
sampling points showed values of PERI <150, indicating a low overall
ecological risk. The calculated values are presented in Table S10 of the
Supplementary Material.
The joint analysis of the ΣTU PEL-Q and PERI indices (Fig. 9) allowed Fig. 9. Values of ΣTU, PEL-Q and PERI indices in sediment samples.
us to have a more general view of the heavy metal contamination in the
study region. As can be noted, mining has a considerable impact on the reason, further studies that continue to monitor and assess the level of
water resources of this semi-arid region. The highest values for the three pollution of water resources in this semi-arid region are crucial to pre­
indices in SD-05, SD-06, SD-11 confirm that the vanadium mine is serving access to safe freshwater sources.
contributing to the increase of heavy metals in the Jacar�e and Contas To finish, the authors want this study to be a useful tool for the au­
rivers. Similarly, the sampling points SD-13 and SD-27 near the uranium thorities responsible for environmental management, especially when
and iron mines show affectation, although to a lesser extent. assessing the impact of economic activities such as mining and agri­
These indices provide some understanding about the quality of the culture on water resources in the semi-arid region.
sediments and offer valuable information to the decision makers on the
pollution status of the aquatic systems. Although no significant ecolog­ CRediT authorship contribution statement
ical risk was observed for any of the analyzed locations; it is important to
highlight that PERI values close to 100 were observed in the Jacar�e and Caroline S. Santana: Formal analysis, Investigation, Methodology,
Contas rivers. If we take into account that the mining activity in the Writing - original draft, Writing - review & editing. Diango M. Mon­
region began only 5 years ago, that it intends to extend for more than 20 talva�n Olivares: Conceptualization, Writing - review & editing, Inves­
years and that in this short period an impact on the quality of these tigation. Vinnícius H.C. Silva: Formal analysis, Validation. Francisco
aquatic systems is already perceived; the results obtained in this study H.M. Luzardo: Supervision, Funding acquisition. Fermin G. Velasco:
constitute a “wake-up” call to carry out actions to protect and preserve Project administration, Funding acquisition. Raildo M. de Jesus:
these water resources. Resources.

4. Conclusion Declaration of competing interest

The results of this study provided valuable information on heavy The authors declare that they have no known competing financial
metal contamination in water resources located in a semi-arid mining interests or personal relationships that could have appeared to influence
region. The diagnosis of surface and groundwater samples indicated as the work reported in this paper.
main pollutants the Cd, Pb, and U, which exhibited concentrations
higher than the reference values established by both Brazilian and in­ Acknowledgment
ternational public organisms. In the case of sediment samples, higher
concentrations of Co, Cr, Cu, Fe, Mn, Ni, V, and Zn were observed. The The authors thank the Coordination for the Improvement of Higher
calculated pollution indices (CF, Igeo, and PLI) indicated a severe to Education Personnel (Coordenaça ~o de Aperfeiçoamento de Pessoal de
extreme contamination level for Cu, Cr, Ni, and V, as well as a moderate Nível Superior - CAPES), for providing the doctoral scholarship. Like­
level for Co, Fe, Mn, and Zn. The sampling points along the Jacar�e and wise, we thank the Center for Research in Radiation Sciences and
Contas rivers were the most polluted and according to the sediment Technologies (Center for Research in Sciences and Technologies of
quality guide indicators (TEL, PEL, ΣTU, and PEL-Q), there is 25% of the Radiaço~es - CPqCTR) of the State University of Santa Cruz (UESC), for
probability that heavy metal content in these locations induces adverse the logistical support provided for the collection and analysis of samples.
toxic effects on aquatic organisms.
Based on the statistical analyses performed (Person correlation, PCA Appendix A. Supplementary data
and HCA); three main sources of pollution were identified. The con­
centrations of Co, Cr, Cu, Fe, Mn, Ni, and V were mainly related to Supplementary data to this article can be found online at https://doi.
mining activities and, to a lesser extent, natural sources. In the case of Cd org/10.1016/j.jenvman.2020.111148.
and Pb, contamination may have an important contribution from fer­
tilizers use, whereas Zn has a mixed source of both lithogenic and
References
anthropogenic origin. Finally, the high concentrations of U, specifically
in groundwater samples, were associated with geogenic causes. CONAMA (National Environment Council), 2005. Resolution 357. Ministry of the
Although the PERI values indicated a low ecological risk; it should Environment, Brazil.
not be forgotten that we are dealing with an extremely vulnerable semi- CONAMA (National Environment Council), 2009. Resolution Nº 420. Ministry of the
Environment, Brazil, p. 249.
arid region, where there is water scarcity and that the Contas River is United States Environmental Protection Agency - USEPA, 1994b. Method 3051:
one of the main supplies of freshwater in the state of Bahia. For that Microwave Assisted Digestion of Sediments and Sludges, Soils and Oils.

11
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

United States Environmental Protection Agency - USEPA, 1999. Screening Level Hoover, J., Gonzales, M., Shuey, C., Barney, Y., Lewis, J., 2017. Elevated arsenic and
Ecological Risk Assessment Protocol for Hazardous Waste Combustion Facilities. uranium concentrations in unregulated water sources on the Navajo Nation, USA.
Appendix E: Toxicity Reference Values. US Environmental Protection Agency, Exp. Health 9 (2), 113–124. https://doi.org/10.1007/s12403-016-0226-6.
Washington, DC. INB, 2020. Nuclear Industries of Brazil). INB Caetit�e - Uranium Concentrate Unit [online].
United States Environmental Protection Agency - USEPA, 2015. Sediment Benchmarks Rio de Janeiro: Nuclear Industries of Brazil [viewed: October, 09th, 2019]. Available
for Aquatic Life. from: http://www.inb.gov.br/A-INB/Onde-estamos/Caetite.
United States Environmental Protection Agency - USEPA, 2019. Regional Screening Level Initiative for responsible mining assurance (IRMA), 2018. IRMA Water Quality Criteria
Summary Table. Disponível em: https://www.epa.gov/risk/regional-screening- by End-Use Tables. IRMA, Washington.
levels-rsls-generic-tables. International Organization For Standardization - ISO 11466, 1995. Soil Quality -
Abojassim, A.A., Mohammed, H.A.U., Najam, L.A., El-Taher, A., 2019. Uranium isotopes Extraction of Trace Metals Soluble in Aqua Regia.
concentrations in surface water samples for Al-Manathera and Al-Heerra regions of Islam, M.S., Ahmed, M.K., Raknuzzaman, M., Habibullah-Al-Mamun, M., Islam, M.K.,
An-Najaf, Iraq. Environ. Earth Sci. 78 (5), 132. https://doi.org/10.1007/s12665- 2015. Heavy metal pollution in surface water and sediment: a preliminary
019-8134-2. assessment of an urban river in a developing country. Ecol. Indicat. 48, 282–291.
Ali, M.M., Ali, M.L., Islam, M.S., Rahman, M.Z., 2016. Preliminary assessment of heavy https://doi.org/10.1016/j.ecolind.2014.08.016.
metals in water and sediment of Karnaphuli River, Bangladesh. Environ. J€
arup, L., 2003. Hazard of heavy metal contamination. Br. Med. Bull. 68, 167–182.
Nanotechnol. Monitor. Manag. (5), 27–35. https://doi.org/10.1016/j. https://doi.org/10.1093/bmb/ldg032.
enmm.2016.01.002. Jarvis, A.P., Younger, P.L., 2000. Broadening the scope of mine water environmental
Angelovi�cov� a, L., Fazeka�sov�
a, D., 2014. Contamination of the soil and water impact assessment: a UK perspective. Environ. Impact Assess. Rev. 20, 85–96.
environment by heavy metals in the former mining area of Rud� nany (Slovakia). Soil https://doi.org/10.1016/S0195-9255(99)00032-3.
Water Res. 9, 18–24. Krause, P., Erbsl€oh, B., Niederges€ aβ, R., Pepelnik, R., Prange, A., 1995. Comparative
Aprile, F.M., Bouvy, M., 2008. Distribution and enrichment of heavy metals in sediments study of different digestion procedures using supplementary analytical methods for
at the Tapacur� a River Basin, northeastern Brazil. Braz. J. Aquat. Sci. Technol. 12 (1), multielement-screening of more than 50 elements in sediments of the river Elbe.
1–8. https://doi.org/10.14210/bjast.v12n1.p1-8. Fresenius’ J. Anal. Chem. 353, 3–11. https://doi.org/10.1007/BF00322882.
Baeten, J., Langston, N., Lafreniere, D., 2018. A spatial evaluation of historic iron mining Liu, B.L., Hu, K., Jiang, Z.L., Yang, J., Luo, X.M., Liu, A.H., 2011. Distribution and
impacts on current impaired waters in Lake Superior’s Mesabi Range. Ambio 47, enrichment of heavy metals in a sediment core from the Pearl River Estuary. Environ
231–244. https://doi.org/10.1007/s13280-017-0948-0. Earth Sci 62, 265–275. https://doi.org/10.1007/s12665-010-0520-8.
Brito, R.S.C., Pinto dos Santos, T., Silva Knuppel, S., Feitosa, M.P., Carvalho, J.A.A., Liu, W., Wu, J.L., Zeng, H., Ma, L., 2014. Geochemical evidence of human impacts on
Almeida, D.Y., 2018. Geology and Cr and Fe-Ti-V Mineralization of the Capivara deep Lake Fuxian, southwest China. Limnologica 45, 1–6. https://doi.org/10.1016/
Complex. Marac� as-Bahia. 4th Brazilian Geology Congress. https://doi.org/10.18285/ j.limno.2013.09.003.
geonomos.v25i2.1080. Liu, X., Li, D., Song, G., 2017. Assessment of heavy metal levels in surface sediments of
Burritt, R.L., Christ, K.L., 2018. Water risk in mining: analysis of the Samarco dam estuaries and adjacent coastal areas in China. Front. Earth Sci. 11 (1), 85–94.
failure. J. Clean. Prod. 178, 196–205. https://doi.org/10.1016/j. https://doi.org/10.1007/s11707-016-0569-0.
jclepro.2018.01.042. Lobato, L.M., Pimentel, M.M., Cruz, S.C., Machado, N., Noce, C.M., Alkmim, F.F., 2015.
Camizuli, E., Monna, F., Scheifler, R., Amiotte-Suchet, P., Losno, R., Beis, P., Bohard, B., U–Pb geochronology of the Lagoa Real uranium district, Brazil: implications for the
Chateau, C., Alibert, P., 2014. Impact of trace metals from past mining on the aquatic age of the uranium mineralization. J. S. Am. Earth Sci. 58, 129–140. https://doi.org/
ecosystem: a multi-proxy approach in the Morvan (France). Environ. Res. 134, 10.1016/j.jsames.2014.12.005.
410–419. https://doi.org/10.1007/s12665-019-8158-7. Long, E.R., Macdonald, D.D., Smith, S.L., Calder, F.D., 1995. Incidence of adverse
Do Carmo, F.F., Kamino, L.H.Y., Junior, R.T., De Campos, I.C., Silvino, G., De Castro, K.J. biological effects within ranges of chemical concentrations in marine and estuarine
S.X., Mauro, M.L., Rodrigues, N.U.A., Miranda, M.P.S., Pinto, C.E.F., 2017. Fund~ ao sediments. Environ. Manag. 19 (1), 81–97. https://doi.org/10.1007/BF02472006.
tailings dam failures: the environment tragedy of the largest technological disaster of L�opez, R., Hallat, J., Castro, A., Miras, A., Burgos, P., 2019. Heavy Metal Pollution in
Brazilian mining in global context. Perspect. Ecol. Conserv. 15 (3), 145–151. https:// Soils and Urban-Grown Organic Vegetables in the Province of Sevilla, Spain.
doi.org/10.1016/j.pecon.2017.06.002. Biological Agriculture & Horticulture. https://doi.org/10.1080/
CETESB (Environmenta Agency of the state of S~ ao Paulo), 2011. National Guide for 01448765.2019.1590234, 2019.
Sample Collection and Preservation: Water, Sediment, Aquatic Communities and Lottermoser, B.G., 2007. Mine Wastes Characterization, Treatment, Environmental
Liquid Effluents (S~ ao Paulo). Impacts. Springer, Berlin. 2ed.
CONAMA (National Environment Council), 2004. Resolution Nº 344. Official Diary of the Machado, G.D.S., 2008. Geology of Thesouthern Portion of the Lagoa Real Complex.
Union, Brasília (Ministry of the Environment, Brazil). Caetit�
e-BA. Geosciences Institute of the Federal University of Bahia. https://doi.org/
Conceiç~ ao, E.G., Souza, M.C.B., Franklin, M.R., Fernandes, H.M., 2002. Preliminary 10.25249/0375-7536.200737S4128146. Bahia, Brazil.
study of leaching of exhausted ore piles at the Uranium Concentration Unit (URA/ Margulis, S., Hughes, G., Gambrill, M., Azevedo, L.G.T., 2002. Brazil: Water Quality
INB), with a view to assessing the environmental impact. Proceedings of Management. Insertion of Environmental Issues on the Water Sector Agenda, vol. 62.
International Nuclear Atlantic Conference. CD-Room, Rio de Janeiro, p. 6p. World Bank, Brasília.
Coyte, R., Singh, A., Furst, K.E., Mitch, W., Vengosh, A., 2019. Co-occurrence of geogenic Marrugo-Negrete, J., Pinedo-Hern� andez, J., Díez, S., 2017. Assessment of heavy metal
and anthropogenic contaminants in groundwater from Rajasthan, India. Sci. Total pollution, spatial distribution and origin in agricultural soils along the Sinú River
Environ. 688, 1216–1227. https://doi.org/10.1016/j.scitotenv.2019.06.334. Basin, Colombia. Environ. Res. 154, 380–388. https://doi.org/10.1016/j.
Dardenne, M.A., Schobbenhaus, C., 2003. Mineral deposits of Brazil: distribution in time, envres.2017.01.021.
provinces and metallogenic epochs. Geol. Tecton. Miner. Resour. Brazil IV, 421–422. Muniz, D.H.F., Filho, E.C.O., 2006. Heavy metals from mining waste and their effects on
EMBRAPA (Brazilian Agricultural Research Corporation), 2011. Manual of sampling health and the environment. Health Sci. 4, 83–100. https://doi.org/10.5102/ucs.
procedures and physical-chemical analysis of water. Embrapa Florestas. v4i1.24.
ENRC – Eurasian Natural Resources Corporation, 2012. Deposit Geology and the Fe Ore Navarro, M.C., P�erez-Sirvent, C., Martínez-S� anchez, M.J., Vidal, J., Tovar, P.J., Bech, J.,
Project in Caetit�e. Bahia. www.enrc.com. 2008. Abandoned mine sites as a source of contamination by heavy metals: a case
Fernandes, H.M., Franklin, M.R., Leoni, G.M., Almeida, M., 2004. A case study on the study in a semi-arid zone. J. Geochem. Explor. 96, 183–193. https://doi.org/
Uranium tailings dam of Pocos de Caldas uranium mining and milling site. IAEA- 10.1016/j.gexplo.2007.04.011.
TECDOC—1403. Paraguassú, L., Leite, M.G.P., Moreira, F.W.A., Mendonça, F.P.C., Eskinazi-Sant’Anna, E.
Filho, F.J. de P., Marins, R.V., de Lacerda, L.D., Aguiar, J.E., Peres, T.F., 2015. M., 2019. Impacts of mining in artificial lake of Iron Quadrangle-MG: past marks and
Background values for evaluation of heavy metal contamination in sediments in the changes of the present. Environmental Earth Sciences 167 (78), 1–10. https://doi.
Parnaíba River Delta estuary, NE/Brazil. Mar. Pollut. Bull. (91), 424–428. https:// org/10.1007/s12665-019-8158-7.
doi.org/10.1016/j.marpolbul.2014.08.022. De Paula, F.C.F., Abrah~ ao, J.R.S., 1991. Geochemical prospecting for vanadium and
Fo, F.L.S., Da Silva, L.F., Santos, R.R., Marques Jr., A., 2008. Contamination Evaluation chromium in the Iramaia sheet, Bahia State, Brazil. J. Geochem. Explor. (4),
of Fissural Aquifers through Geochemical Modeling and Structural Survey in 125–135. https://doi.org/10.1016/0375-6742(91)90079-A.
Uranium Mining Area in the Semi-arid Region of Bahia, pp. 1–18. Groundwater. Pereira, T.L., Wallner-Kersanach, M., Costa, L.D.F., Costa, D.P., Baisch, P.R.M., 2018.
Fungaro, D.A., 2005. Acid mine drainage treatment uses zeolites synthesized from coal Nickel, vanadium, and lead as indicators of sediment contamination of marina,
ash, 18. Filter Media Magazine. refinery, and shipyard areas. Environ. Sci. Pollut. Res. 25, 1719–1730. https://doi.
García-Lorenzo, M.L., P�erez-Sirvent, C., Martínez-S� anchez, M.J., Molina-Ruiz, J., 2012. org/10.1007/s11356-017-0503-3.
Trace elements contamination in an abandoned mining site in a semi-arid zone. Perlatti, F., Ferreira, T.O., Roberto, F.A. da C., Romero, R.E., Sartor, L.R., Otero, X.R.,
J. Geochem. Explor. 113, 23–35. https://doi.org/10.1016/j.gexplo.2011.07.001. 2015. Trace metal/metalloid concentrations in waste rock, soils and spontaneous
Gray, P.J., Mindak, W.R., Cheng, J., 2015. Inductively coupled plasma-mass plants in the surroundings of an abandoned mine in semi-arid NE-Brazil. Environ
spectrometric determination of arsenic, cadmium, chromium, lead, mercury, and Earth Sci 74, 5427–5441. https://doi.org/10.1007/s12665-015-4556-7.
other elements in food using microwave assisted digestion version 1.1. Elemental Postawa, A., Hayes, C., 2013. Best Practice Guide on the Control of Iron and Manganese
Analysis Manual. US Food and Drug Administration. in Water Supply. IWA Publishing, Londres.
Hakanson, L., 1980. An ecological risk index for aquatic pollution control.a De Queiroz, T.D.A., 2016. Geological Mapping, Petrography, Lithogeochemical and
sedimentological approach. Water Res. 14, 975–1001. https://doi.org/10.1016/ Geophysical Aspects of the Marac� as Region (FOLHA SD.24-V-D-I). Dissertation
0043-1354(80)90143-8. presented to the Geosciences Institute of the Federal University of Bahia, Bahia,
Hilton, J., Davison, W., Ochsenbein, U., 1985. A mathematical model for analysis of Brazil.
sediment coke data. Chem. Geol. 48 (1–4), 281–291. https://doi.org/10.1016/0009- Reis, M.M., Santos, L.D.T., da Silva, A.J., de Pinho, G.P., Montes, W.G., 2019. Metal
2541(85)90053-1. contamination of water and sediments of the vieira river, montes claros, Brazil. Arch.
Environ. Contam. Toxicol. 1–10 https://doi.org/10.1007/s00244-019-00666-1.

12
C.S. Santana et al. Journal of Environmental Management 273 (2020) 111148

Largo Resources, 2020. Marac� as Menchen Mine [online]. Ontario. Largo Resources Ltd. discharges. J. Hazard Mater. 171, 1088–1095. https://doi.org/10.1016/j.jhazm
[viewed: January, 19th, 2020] Available from: https://www.largoresources. at.2009.06.109.
com/operations/maracas-menchen-mine/ Ullah, R., Muhammad, S., Jadoon, I.A.K., 2019. Potentially harmful elements
Sabadini-Santos, E., Knoppers, B.A., Oliveira, E.P., Leipe, T., Santelli, R.E., 2009. contamination in water and sediment: evaluation for risk assessment and provenance
Regional geochemical baselines for sedimentary metals of the tropical S~ ao Francisco in the northern Sulaiman fold belt, Baluchistan, Pakistan. Microchem. J. 147,
estuary, NE-Brazil. Mar. Pollut. Bull. (58), 601–634. https://doi.org/10.1016/j. 1155–1162. https://doi.org/10.1016/j.microc.2019.04.053.
marpolbul.2009.01.011. United States Environmental Protection Agency - USEPA, 1994a. Method 200.8:
Saleem, M., Iqbal, J., Shah, M.H., 2013. Study of seasonal variations and risk assessment determination of trace elements in waters and wastes by inductively coupled plasma
of selected metals in sediments from Mangla Lake, Pakistan. J. Geochem. Explor. – mass Spectrofotometry. ReVision 5, 4. EMMC Version.
125, 144–152. https://doi.org/10.1016/j.gexplo.2012.12.006. Venkatramanan, S., Chung, S.Y., Ramkumar, T., Selvam, S., 2015. Environmental
Santoro, A., Held, A., Linsinger, T.P.J., Perez, A., Ricci, M., 2017. Comparison of total monitoring and assessment of heavy metals in surface sediments at Coleroon River
aqua regia extractability of heavy metals in sewage sludge: the case of study of a Estuary in Tamil Nadu, India. Environ. Monit. Assess. 187, 505. https://doi.org/
certified reference material. Trends Anal. Chem. 89, 34–40. https://doi.org/ 10.1007/s10661-015-4709-x.
10.1016/j.trac.2017.01.010. Wang, Y., Yang, L., Kong, L., Liu, E., Wang, L., Zhu, J., 2015. Spatial distribution,
Satapathy, D.R., Salve, P.R., Katpatal, Y.B., 2009. Spatial distribution of metals in ecological risk assessment and source identification for heavy metals in surface
ground/surface waters in the Chandrapur district (Central India) and their plausible sediments from Dongping Lake, Shandong, East China. Catena 125, 200–205.
sources. Environ. Geol. 56 https://doi.org/10.1007/s00254-008-1230-3, 1323–135. https://doi.org/10.1016/j.catena.2014.10.023.
S�
�can�car, J., Mila�ci�c, R., Horvat, M., 2000. Comparison of various digestion and Wang, Z., Qin, H., Liu, X., 2019. Health risk assessment of heavy metals in the soil-water-
extraction procedures in analysis of the heavy metals in sediments. Water Air Soil rice system around the Xiazhuang uranium mine, China. Environ. Sci. Pollut. Control
Pollut. 118, 87–99. https://doi.org/10.1023/A:1005187602820. Ser. 26 (6), 5904–5912. https://doi.org/10.1007/s11356-018-3955-1.
Da Silva Barbosa, N., Teixeira, W., Leal, L.R.B., de Menezes Leal, A.B., 2013. Crustal Wei, G., Shu-hua, Y., Jian-xin, X., Dong-yu, X., Li, G., Hong, H., Bo, G., 2016. Pollution
evolution of the western sector of the arqueano block Gavi~ ao, S~
ao Francisco craton, characteristics and ecological risk assessment of vanadium in sediments of the three
based on evidence U-Pb, Sm-Nd and Rb-Sr. USP geology. Sci. Ser. 13 (4), 63–88. gorges reservoir (chongqing -yichang section). Environ. Sci. J. Integr. Environ. Res.
Silva, J.G.B., Geraldo, L.P., Yamazaki, I.M., 2011. Assessment of the total uranium 37 (9), 3333–3339. https://doi.org/10.13227/j.hjkx.2016.09.011.
concentration in surface and underground water samples from the Caetit� e region, Wheater, H.S., Mathias, S.A., Li, X. (Eds.), 2010. Groundwater Modelling in Arid and
Bahia, Brazil, 7. Multidisciplinary Scientific Magazine of the University Center of the Semi-arid Areas. Cambrigde University Press, pp. 1–4. https://doi.org/10.1017/
Educational Foundation of Barretos, pp. 17–21, 2. CBO9780511760280.
Silva, Y.J.A.B.D., Nascimento, C.W.A.D., Silva, Y.J.A.B.D., Biondi, C.M., Silva, C.M.C.A. World Health Organization - Who, 2017. Guidelines for Drinking-Water Quality, fourth
C., 2016. Rare earth element concentrations in Brazilian benchmark soils. Brazil. J. ed. (Switzerland).
Soil Sci. 40 https://doi.org/10.1590/18069657rbcs20150413. World information service on energy (WISE) – Uranium project, 2019. Chronology Of
Sodrzeieski, P.A., de Andrade, L.C., Tiecher, T., Camargo, F.A. de O., 2018. Physico- Major Tailings Dam Failure [online]. Peter Diehl, Arnsdorf, Germany [viewed:
chemical variability and heavy metal pollution of surface sediment in a non- October 6th, 2019]. Available from: http://www.wise-uranium.org/mdaf.html.
channeled section of Dilúvio Stream (Southern Brazil) and the influence of Yang, Q., Li, Z., Lu, X., Duan, Q., Huang, L., Bi, J., 2018. A review of soil heavy metal
channeled section in sediment pollution. Environ. Mag. Water 14 (1), 1–13. https:// pollution from industrial and agricultural regions in China: pollution and rsik
doi.org/10.4136/ambi-agua.2285. assessment. Sci. Total Environ. 642, 690–700. https://doi.org/10.1016/j.
Soliman, N.F., Nasr, S.M., Okbah, M.A., 2015. Potential ecological risk of heavy metals in scitotenv.2018.06.068.
sediments from the Mediterranean coast, Egypt. J. Environ. Health Sci. Eng. 13 (70), Zeng, H.A., Wu, J.L., 2009. Sedimentary records of heavy metal pollution in fuxian lake,
1–12. https://doi.org/10.1186/s40201-015-0223-x. yunnan province, China: intensity, history and sources. Pedosphere 19 (5), 562–569.
Somboon, S., Inoue, K., Fukushi, M., Tsuruoka, H., Shimizu, H., Kasar, S., Area, H., https://doi.org/10.1016/S1002-0160(09)60150-8.
Kavasi, N., Sahoo, S.K., 2019. Distribution of uranium in Japanese river waters Zhou, Y., Gao, L., Xu, D., Gao, B., 2019. Geochemical baseline establishment,
determined with inductively coupled plasma mass spectrometry. J. Radioanal. Nucl. environmental impact and health risk assessment of vanadium in lake sediments,
Chem. 319 (3), 1307–1314. https://doi.org/10.1007/s10967-018-6176-7. China. Sci. Total Environ. 660, 1338–1345. https://doi.org/10.1016/j.
Suthar, S., Nema, A.K., Chabukdhara, M., Gupta, S.K., 2009. Assessment of metals in scitotenv.2019.01.093.
water and sediments of Hindon River, India: impact of industrial and urban

13