Role of mafic and ultramafic rocks in drinking water quality and its potential health risk assessment, Northern Pakistan

Water is vital to human life and therefore, an adequate supply of clean and safe water is a determining factor for the health status of exposed population (Shah et al. 2012; Khan et al. 2013a, c, 2015; Chappells et al. 2014; Lotter et al. 2014; Singh et al. 2014). Unfortunately, the water can become contaminated with metals from natural (weathering of rocks and ore deposits) and anthropogenic (mining, agriculture, industry and domestic wastewater) sources (Khan et al. 2011, 2013a, c, 2014). Some metals are essential for normal human body function, including calcium (Ca), magnesium (Mg), sodium (Na), potassium (K), cobalt (Co), copper (Cu) and zinc (Zn). However, high intake of these metals may produce toxicity. Whereas other trace metals (TM), e.g., lead (Pb), chromium (Cr), nickel (Ni), arsenic (As) and cadmium (Cd), are extremely toxic even in very low concentrations to human health and the environment (Khan et al. 2013b, 2014, 2015; Hussain et al. 2014; Shah et al. 2014; Gul et al. 2015). Metal contaminated drinking water and food is the most direct route for human exposure (Shah et al. 2012; Khan et al. 2013b, 2014) and intake of such contaminated water may produce chronic and carcinogenic effects in the exposed human population (Muhammad et al. 2010, 2011; Khan et al. 2013c).

Contamination not only spoils water quality, but also threaten ecosystems, human health, social prosperity and economic development (Arsovski et al. 1991). Users are dependent on the quality of water; if the quality is poor, the exposed population must pay an additional cost for treatment or incur the property damage or health risk (Koc 2010). Therefore, water quality monitoring and assessment has one of the highest priorities in environmental protection policy (Simeonov et al. 2002; Khan et al. 2015). Regular water quality monitoring provides helpful information about the status of drinking water (Rode & Suhr 2007; Muhammad et al. 2011; Shah et al. 2012).

Statistical analyses have proven useful for comparing and interpreting environmental data such as water quality. These analyses allow the identification of the possible contamination sources that influence the water system. These analyses also provide a valuable tool for reliable management of water resources and rapid solution to contamination problems (Krishna et al. 2009; Khan et al. 2013a; Kelepertzis 2014). The study area is dominantly composed of mafic and ultramafic rocks and could be the potential source of TM contamination into the water and environment (Shah et al. 2010; Muhammad et al. 2011; Khan et al. 2013a). Therefore, this study has been carried out to assess the drinking water and evaluate the potential human health risk assessment. Further, using statistical analyses the study will identify the contamination sources of drinking water in the Swat district.

Groundwater (dug wells n = 42, tube-wells n = 21 and springs n = 21) and surface water (streams n = 15 and Swat River n = 18) samples were randomly collected in autumn season from selected sites of the Mingora and Kabal areas in the Swat district. From each site, clean polyethylene bottles (1 L) were filled in triplicates (Figure 1). Samples were properly marked and transported to the laboratory and stored in refrigerator at 4 °C for major cationic metals (MCC) and TM analyses.

Basic parameters including pH, electrical conductivity (EC) and total dissolved solids (TDS) were measured on the spot using a Consort C931 electrochemical analyzer. Water was filtered through Whatman (0.45 μm) filter paper and acidified with nitric acid (HNO₃) from each sampling site. This acidification with HNO₃ preserves the TM in water samples due to a reduction in microbial activity, precipitation and sorption losses to container walls (APHA 2005). All filtered and acidified water samples were analyzed for MCC (Na, K, Ca and Mg) and TM (Fe, Cu, Pb, Zn, Ni, Cr, Co and Mn) under standard optimum conditions on a Perkin Elmer 700 atomic absorption spectrophotometer equipped with a heated graphite atomizer. All samples were analyzed in triplicates and the reproducibility was accepted at a 95% confidence level. However, the average value was used for data interpretation. Chemicals and reagents used were of analytical grade (MERCK). These analyses were performed in the Geochemistry Laboratory of National Centre of Excellence in Geology, University of Peshawar, Pakistan.

Statistical manipulations including range, mean and standard deviation were calculated using Excel 2007 (Microsoft Office). Univariate and multivariate statistical analyses (one-way analysis of variance (ANOVA) procedure, correlation analysis and principal component analysis (PCA)) were performed using the SPSS 21 (SPSS Inc., Chicago, IL, USA).

Water pH naturally depends on CO₂, carbonate and bicarbonate equilibrium (APHA 1998). The usual range of pH in drinking water has no direct effects on human health but may exert indirect effects by changing the solubility of metals and pathogen survival. A bitter taste of drinking water may be associated with high pH (Muhammad et al. 2010). In the study area, pH values were circum-neutral ranging from 6.95 to 7.10 and 6.92 to 7.20 in groundwater and surface water, respectively (Table 1). All sources of water have nearly equal mean pH values. Drinking water pH in the study area was found to be lower than those reported by Khan et al. (2013a) for water in the upper Swat. EC reflects the amount of TDS in water and depends on the geology of the area, size and type of watershed and other sources of ions including contaminants (APHA 1998). EC values ranged from 189 to 1,515 μS/cm in groundwater and from 71 to 465 μS/cm in surface water (Table 1). The TDS values depend on the concentrations of carbonates, bicarbonates, chloride, sulfate, nitrate, Na, K, Ca and Mg (Muhammad et al. 2010). In the study area, TDS values were in the ranges 78–808 mg/L and 37–248 mg/L for groundwater and surface water, respectively (Table 1). Groundwater revealed higher TDS values as compared with surface water. These higher values in groundwater may be attributed to residence time and underground geology of the area (Shah et al. 2012). The TDS mean values in groundwater of the study area were found to be lower than those reported by Shah et al. (2012) in groundwater or subsurface water. Water pH, EC and TDS values were within the permissible limit of the World Health Organization (WHO 2011), except for 7% of samples collected from dug wells in the Mingora area which showed higher EC values.

Calcium is needed for normal body functions including blood clotting, transmission of nerve impulses and regulation of heart rhythm (Muhammad et al. 2010). Magnesium and Ca deficiency in drinking water may be associated with cardiovascular diseases (Yang et al. 2006). The concentrations of Ca were in the ranges 1–126 mg/L and 2–58 mg/L in groundwater and surface water, respectively (Table 1). Higher concentrations of Ca are generally released to water from deposits of limestone, dolomite, gypsum and gypsiferous shale (APHA 1998).

Similarly, Mg concentrations were in the ranges 1–99 mg/L and 5–13 mg/L in groundwater and surface water, respectively (Table 1). In the study, groundwater showed multifold higher Mg concentrations as compared with surface water. These multifold higher concentrations were may be attributed to the dissolution of Mg rich mafic and ultramafic bed rocks (Shah et al. 2010, 2012). The concentrations of Ca and Mg were within the permissible limits set by WHO (2011), except for 13% of instances where Mg concentrations (50 mg/L) exceeded these limits. These higher concentrations of Mg can be attributed to the percolation of water through Mg rich ultramafic rocks of the mélange zone.

The concentrations of Na were in the ranges 3–223 mg/L and 2–13 mg/L in groundwater and surface water, respectively (Table 1). This could be a localized phenomenon with input from the granitic rock which has higher amounts of sodium feldspar. However, most of the water samples from the area have quite low concentrations of Na, but not low enough to cause health problems, such as mental apathy, low blood pressure, fatigue, depression, and dehydration (Robert & Mari 2003). Latorre & Toro (1997) reported that on average, the adult daily intake is <0.1% of K through water ingestion. Like other MCC, K is also necessary for normal body functions. Similarly, K concentrations were in the ranges 1–28 mg/L and 1–26 mg/L in groundwater and surface water, respectively (Table 1). This could be attributed to the presence of schistose rocks containing greater amounts of minerals like muscovite and potash-feldspar. However, most of the water samples contained quite low concentrations (<4 mg/L) of K, but not low enough to cause heart problems, hypertension, muscle weakness, bladder weakness, kidney diseases and asthma (Marijic & Toro 2000; Aparna 2001a). The concentrations Na and K were within the permissible limits set by WHO (2011) except for 2% of the samples that showed higher (12 mg/L) K concentrations.

TMs are pollutants that may be harmful to human health owing to their toxicity, persistence and bioaccumulative nature in the environment (Khan et al. 2013b, 2014). Iron is one of the essential elements for human body and needed in hemoglobin, myoglobin and a number of enzymes. However, in excess Fe is also toxic and can cause human health issues including diarrhea, vomiting, liver, kidney and blood problems, with subsequent effects on the cardiovascular and central nervous systems (Goldhaber 2003). Among TM, Fe concentrations were in the ranges 29–800 μg/L and 73–479 μg/L in groundwater and surface water, respectively (Table 1).

High intake of Mn can cause toxicity in the nervous system, muscle tremors, dizziness, liver disease, cancer, oedema, fibroid tumors and colitis, while its deficiency may cause hypoglycemia, joint discolorations, asthma, migraine, osteoporosis and gastrointestinal disorder (Mergler 1999a, b). Mn concentrations were in the ranges 20–2,198 μg/L and 31–87 μg/L in groundwater and surface water, respectively (Table 1). The abnormally high Mn concentration (2,198 μg/L) in one sample of a dug well could be due to contamination from a huge solid waste dump site in the vicinity. In the study area, Mn concentrations were found to be higher than those reported by Muhammad et al. (2011). Therefore, the solid waste should be removed and the dug well water regularly monitored.

Excessive intake of Cu in drinking water can lead to several health problems, while its normal concentration is necessary for health including the immune system and artery strength (Kidd 2003). In this study, Cu concentrations were in the ranges 17–129 μg/L and 26–113 μg/L in groundwater and surface water, respectively (Table 1).

Pb concentrations were in the ranges 2–9 μg/L and 3–8 μg/L in groundwater and surface water, respectively (Table 1), and Zn concentrations were in the ranges 47–1,913 μg/L and 57–890 μg/L in groundwater and surface water, respectively (Table 1). Concentrations of Fe, Mn, Cu, Pb and Zn were within the permissible limits set by the WHO (2011). However, 4% and 2% of the water samples showed higher concentration than their limits for Fe (300 μg/L) and Mn (400 μg/L), respectively.

Results of the study revealed that groundwater samples collected from sampling sites (MW3, MW6, MW13, MW14, MW2, MW4 and MW5) in the Mingora-Shangla mélange zone horizon (Figure 1) are enriched in Mg, Cr and Ni, while all the other TM are within safe limits (WHO 2011). This enrichment of Mg, Cr and Ni can be attributed to the percolation of water and scavenging of Mg, Cr and Ni from ultramafic rocks enriched in Mg, Cr and Ni (Arif & Jan 1993; Shah et al. 2012).

Information about sex, age, literacy rate, livelihood, health and drinking water of local people was collected in the field. The study population is dependent on groundwater (bore well, dug well, hand pump and tube well) and surface water (stream) sources for drinking purpose. Therefore, samples from these sources were collected and evaluated for exposure assessment (viz., ADI) and chronic risk assessment (viz., HQ).

Average daily intake (DIM) values of the study area revealed the highest (6.28 × 10¹μg/kg-day) for Mn, with the lowest (1.70 × 10⁻¹μg/kg-day) for Ni in groundwater (Table 2). Other TM ADI values were found in between the two extremes. The highest ADI values of Mn in groundwater were attributed to solid waste dump alongside Mingora city. The ADI values of the study area were noted to be higher than those reported by Khan et al. (2013a) for drinking water in upper Swat. These higher ADI values for Mn, Fe, Cr in drinking water may be attributed due to mafic, ultramafic bed rocks and the solid waste dump in the vicinity.

The DIM values were further evaluated for the chronic risk assessment or HQ. The HQ value was found highest (1.94) for Cr, whereas lowest (<0.01) for Zn (Table 2). The chronic risk of TM depends on type, variety, toxicity, consumption rate and concentration (Kapaj et al. 2006). These higher HQ values in drinking water may be due to their higher concentration. These HQ values were noted to be higher than those reported by Muhammad et al. (2011) for drinking water of the mafic and ultramafic rocks in the Kohistan region, Pakistan but lower than drinking water of Thiva area, Greece (Kelepertzis 2014). Therefore, the exposed population may be at chronic risk of Cr and Ni contamination.

Statistical comparison using one-way ANOVA for different water sources revealed significant variations at the level of p < 0.05, which means that these sources contribute differently to the mean water contamination of Mg, Ni, Cr and Mn (Table 1). The Mingora-Shangla mélange zone horizon may be contributing to high contamination levels in the groundwater sources of the study area. The Pearson correlation matrices between the selected physico-chemical parameters in groundwater and surface water are shown in Tables 3 and 4, respectively. Many physico-chemical parameters showed strong significant correlations in groundwater and surface water. High correlation at a significant level were found between the metal pairs including Mg-Ni (r = 0.734) and Mg-Cr (r = 0.840) showing a common source of these metals. Mn showed high correlation with Na (r = 0.757) and Co (0.526) suggesting their geochemical association. In surface water, the metal pairs such as Na-Ca (r = 0.880), K-Ca (r = 0.684) showed higher correlation (Table 4).

The PCA results for metals concentration in groundwater are shown in Table 5. The PCA method resulted in a reduction of initial dimension of dataset to five components having eigenvalues higher than 1.00 (before and after rotation), and explained 73.56% of the data variation in groundwater. The PC1, PC2, PC3, PC4 and PC5 components accounted for the 21.07, 19.43, 12.06, 10.54 and 10.45% of the total variance. TMs including Ni and Cr were found to be strongly associated with Mg in the first component (PC1) suggesting the influence of the olivine and pyroxene minerals of ultramafic rock on groundwater. High association was observed by EC with TDS and K due to inorganic dissolved solids and ionic species (Jonnalagadda & Mhere 2001). PC3 includes Mn, Co, Na, and Pb with high loadings and demonstrates the contribution of carbonate bed rock minerals such as talc carbonate schist. The natural or background contribution of metals to groundwater is also represented by PC4 and PC5. The five components of PCA in surface water explained 88.48% of the data variation (Table 6). The PC1, PC2, PC3, PC4 and PC5 components accounted for the 36.75, 16.08, 14.03, 10.89 and 10.74% of the total variance in surface water. MCC including Na, Fe and Ca were found to be associated with TDS and EC in the first component (PC1) suggesting the strong influence of carbonate bedrocks such as talc carbonate schist in surface water. Iron was found to be associated with Co (PC2); Cr with Ni (PC4), and Cu, and Mg with the K (PC3) suggesting the mafic and ultramafic bed rocks contributed to the surface water contamination. This study is in good agreement with previous studies reported by Krishna et al. (2009), Muhammad et al. (2010, 2011) and (Kelepertzis 2014).

Results of the study showed that physico-chemical parameters were found to be within the drinking water permissible limit set by the WHO, except for Mn, Fe, Ni, Cr and Co in a small number of samples. The ADI and HQ through drinking water consumption were found in the order of Mn > Zn > Fe > Cr > Ni > Cu > Pb and Cr > Fe > Mn > Pb > Ni >Cu > Zn, respectively. Statistical analyses revealed that different water sources contribute significantly (p < 0.05) to the mean contamination. In the Mingora site, dug wells water showed higher level of TM (Ni and Cr). Statistical analyses showed that the main contamination sources are geogenic. Enrichment of these potentially hazardous metals in the water can be attributed to the ultramafic rocks of the mélange zone in the area. High intake of these metals through drinking water could have chronic effects on human health which need to be explored by carrying out epidemiological research in the target area. These remote areas in northern parts of Pakistan need special attention as far as metal contaminations through geogenic sources are concerned. Further studies need to consider the spatial and temporal variations in potentially toxic metal contaminants in greater detail as a matter of urgency.

The efforts of Mr Tariq Khan, Laboratory Assistant, are highly acknowledged. We are very grateful to the Director of the National Centre of Excellence in Geology, University of Peshawar, Pakistan, for the financial support for fieldwork and laboratory studies. We are grateful to the editor and reviewers for their valuable time and comments on this manuscript. All authors have declared no conflict of interest.