Introduction

Groundwater serves as an important water resource supplying the demand for drinking, domestic, agricultural and industries, especially in regions of arid and semi-arid countries (Wang et al. 2020; Li et al. 2019; Ezugwu et al. 2019; Çiner et al. 2020) and it is usually perceived as safe compared to surface water resources due to its natural filtration of contaminants by the soil as water moves into the groundwater (Sener et al. 2017a, b; Wagh et al. 2018; Edokpayi et al. 2018). However, researchers have reported several cases of groundwater contamination due to various sources (He and Wu 2019). Heavy metal pollution is among the most prominent threat to the quality of groundwater resources owing to its extreme toxicity even at low concentrations (Abou Zakhem and Hafez 2015; Singh and Kamal 2016; Tiwari et al. 2017; Egbueri and Unigwe 2020). Heavy metals enter groundwater through natural processes such as weathering and dissolution of rocks and soils, ion exchange processes, volcanism extruded products, decomposition of living matter and atmospheric matter (Prasanna et al. 2012; Rezaei et al. 2018; Çiner et al. 2020; Wang et al. 2021). Mining, industrial, agricultural, and domestic refuse disposals are some of the common anthropogenic activities that influence the concentration of heavy metals in groundwater resources (Rezaei et al. 2017; Wagh et al. 2018; Giri and Singh 2019; Ahamad et al. 2020). However, the accumulation of heavy metals in ecosystem is exacerbated by anthropogenic activities (Ukah et al. 2019). Accumulated heavy metals are classified as essential (copper, chromium, cobalt, iron, manganese and zinc) and non-essential (arsenic, cadmium and lead) (Çiner et al. 2020). Some heavy metals are required in small amounts for human bodybuilding. However, excess consumption of such heavy metals can be detrimental to health (Haloi and Sarma 2012; Chanpiwat et al. 2014; Boateng et al. 2015; Vetrimurugan et al. 2018; Xiao et al. 2019).

Several numerical and statistical models developed globally are successfully used in water quality assessment. These methods have been shown to provide better insights regarding the quality and health risk statuses of any given drinking water supply (Solangi et al. 2019; Egbueri 2020). For example, the heavy metal pollution index (HPI), heavy metal evaluation index (HEI), synthetic pollution index (SPI), and water quality index (WQI), correlation analysis (CA), principal component analysis (PCA) and hierarchical cluster analysis (HCA) (Mohan et al. 1996; Prasad and Bose 2001; Edet and Offiong 2002; Prasanna et al. 2012; Subba Rao 2012; Solangi et al. 2019; Egbueri 2020; Wu et al. 2020; Li et al. 2015, 2016; Ren et al. 2021). About 40% of the groundwater samples in Nigeria is polluted by heavy metals rendering them inappropriate for human consumption (Egbueri and Unigwe 2020). Different techniques such as PIG, ERI and HCA to assess the drinking water quality of groundwater in Ojoto, Nigeria was utilised. According to the PIG classification, 20% of the groundwater samples were found to be very highly polluted and were found unsuitable for human consumption (Egbueri 2020). The HPI, HEI and Cd results revealed that the majority of groundwater samples in central Bangladesh fall under a low level of pollution and the Cd provided better insight when compared to the other indices (Bodrud-Doza et al. 2016).

Population growth, increase in agricultural activities, rapid urbanization, industrialization, and climate change are among the factors that lead to water quality problems in South Africa (Vhonani et al. 2018). In Southern Africa, about two-thirds of the country’s population depends on groundwater for their domestic purposes (Nel et al. 2009; Vetrimurugan et al. 2017). According to DWAF (2000), 65% of the total water supply in rural areas is derived from groundwater. Direct consumption of groundwater without any form of treatment exposes local communities to various contaminants that may have an adverse impact on human health. About 3.6% of deaths per year are linked to drinking water contamination in South Africa (Nel et al. 2009). Arsenic and lead are the major contaminant sources of groundwater in South Africa (Verlicchi and Grillini 2020).

The present study is focused on Maputaland coastal aquifer, South Africa. Rural communities within the area are still without an adequate supply of water resources, as a result, they solely rely on groundwater resources for their domestic water needs. Previous studies conducted in this area revealed that groundwater is highly contaminated with iron, rendering it unsuitable for drinking purpose (Demlie et al. 2014). The ecological impact of metals in beach sediments in marine protected areas shows enrichment of metals which are due to the heavy mineral-rich coastal dunes and past mining activities (Vetrimurugan et al. 2018). The concentrations of cadmium, zinc, lead, manganese, aluminium and iron exceeded the limits of World Health Organization (WHO) standards for drinking water quality of the Maputaland coastal aquifer (Mthembu et al. 2020). Statistical and indexical approaches have not sufficiently explored the extent of metal contamination in the study region. To assess heavy metal levels in drinking water and associated risks to human health, it is imperative to perform a comprehensive study. The main objectives of this study are (1) to evaluate the extent of heavy metal pollution in groundwater using HPI, and HEI (2) assess groundwater quality for drinking purposes by adopting SPI and WQI, and (3) to evaluate potential heavy metal contamination in groundwater using multivariate statistical tools. This study will provide knowledge on major pollution sources and assist water management of South Africa in combating further contamination of groundwater resources.

Methodology

Description of the Study Area

The Maputaland coastal plain of northern KwaZulu Natal extends southerly from Mtunzini to the Mozambique border in the North (Fig. 1). It belongs to the uMkhanyakude district municipality with a geographical area covering approximately 4400 km2. The climate of the study area is humid subtropical with warm summers. The annual rainfall ranges from less than 600 mm per annum inland and rises to approximately 1000 mm per annum along the coast of the study region (Porat and Botha 2008; Ramsay 1996; Watkeys et al. 1993). Land use activities in the area consist of subsistence agriculture, commercial forestry plantations and eco-tourism. The foremost water supply for drinking, domestic and agricultural activities are from groundwater.

Fig. 1
figure 1

Map showing sampling locations and geology of the study area

Full size image

Geologically, the Maputaland coastal plain is underlain by unconsolidated to semi-consolidated sediments of Cretaceous to Quaternary age. Alluvial deposits, arenite, sandstone, and siltstone cover the study area (Fig. 1). Deposits of the Zululand Group consists of Makatini, Mzinene and St. Lucia formations. These sediments are overlain by Miocene aged Uloa formation which is in turn overlain by the cross-bedded calcarenites of the Umkhwelane formation (Meyer et al. 2001). Pleistocene aged Port Durnford and Kosi Bay formations are characterised by loosely consolidated sands, silts, clays and lignite beds (Demlie et al. 2014; Ndlovu 2015). The Holocene aged Sibayi formation is characterised by high coastal dune cordons (Watkeys et al. 1993). The area is covered by sediments that are highly permeable and indorse swift recharge to the aquifers and strongly interact with the wetlands in the region (Mkhwanazi 2010). Groundwater in this area is encountered within the shallow unconfined aquifer that yields approximately 0.5 to 5.0 l/s. (Kelbe et al. 2016; Ndlovu and Demlie 2018). Generally, groundwater flows from the west to the Indian Ocean. The highest groundwater level is approximately 25 m below ground level.

Sampling and Analysis of Groundwater

Groundwater samples were collected from randomly selected 53 bore wells of the Maputaland coastal aquifer during April 2019 (Fig. 1). The wells were purged for approximately 5 min to remove stagnant water prior to sample collection. The groundwater samples were collected and stored using high-density polyethene (HDPE) bottles after filtering with a membrane filter. The 0.5 ml of nitric acid was added to the samples to prevent metal precipitation. The samples were labelled accordingly and stored in the refrigerator at a temperature of 4 °C (Sunkari et al. 2021). Groundwater samples were analysed for pH, As, Fe, Mn, Zn, Pb, Ni, Cr, Cu, Cd and Co.

The pH of the groundwater samples was determined with a multiprobe meter instrument (Aqua Probe A-700). Heavy metals present in water samples were determined by inductively coupled plasma-mass spectrometry (NexION 2000 ICP-MS) instrument using calibration standards (Merck) at the University of Zululand hydrogeochemistry laboratory. Throughout the analytical procedures, the standards of the American Public Health Association (APHA 2005) were followed. Standards and blanks were run to assess the accuracy of the analysis. ArcGIS software v10.5 was employed to plot spatial maps using the IDW (inverse distance weighted) interpolation technique.

Heavy Metal Pollution Assessment

Heavy Metal Pollution Index (HPI)

The heavy metal pollution index (HPI) provides overall quality of water with respect to heavy metals (Mohan et al. 1996). The HPI is computed by first assigning a rating or a weightage (Wi) to each heavy metal. The rating is an arbitrary value between 0 and 1 and its selection depends on the relative importance of each water quality parameter. This weightage (Wi) can be defined as inversely proportional to the recommended standard (Si) for the corresponding parameter (Mohan et al. 1996). In this present study, the standard permissible value (Si) was taken from the SANS standards (SANS 2015). The HPI was computed by the following equation (Mohan et al. 1996);

$${\text{HPI = }}\frac{{\sum\nolimits_{i = 1}^{n} {W_{i} Q_{i} } }}{{\sum\limits_{i = 1}^{n} {W_{i} } }},$$
(1)

where Qi is the sub-index of the parameter, Wi is the unit weight of the parameter, and n is the number of parameters considered. The sub-index (Qi) is calculated by the equation

$$Q_{i} = \sum\limits_{i = 1}^{n} {\frac{{\left\{ {M_{i} ( - )I_{i} } \right\}}}{{(S_{i} - I_{i} )}}},$$
(2)

where Mi is the monitored value of the parameter, Ii is the ideal value of the parameter and Si is the standard value of the parameter. In this study, the ideal values for all the metals are given as 0 µg/L since their presence in drinking water is not desired (Vetrimurugan et al. 2017). The negative sign (−) indicates the numerical difference between the two values, ignoring the algebraic sign.

Heavy Metal Evaluation Index (HEI)

The HEI yields the overall quality and presence of heavy metal in water (Edet and Offiong 2002). It was computed as follows;

$${\text{HEI = }}\sum\limits_{{i{ = 1}}}^{n} {\frac{{H_{{\text{c}}} }}{{H_{{{\text{MAC}}}} }}},$$
(3)

where HC is the monitored value and HMAC is the maximum admissible concentration (MAC) of the parameter taken from SANS (2015).

Assessment of Drinking Water Quality

Synthetic Pollution Index (SPI)

SPI is used to evaluate the degree of pollution of groundwater resources (Solangi et al. 2019), and their drinking suitability (Egbueri and Unigwe 2020). It was computed by the following equation;

$${\text{SPI = }}\sum\limits_{i = 1}^{n} {\frac{{V_{{\text{o}}} }}{{V_{{\text{s}}} }}} \times W_{i}$$
(4)
$$W_{i} = \frac{K}{{V_{{\text{s}}} }}$$
(5)
$$K = \frac{1}{{\sum\limits_{i = 1}^{n} {\frac{1}{{V_{{\text{s}}} }}} }},$$
(6)

where K, Vs, Vo, n and Wi are the constant of proportionality, each parameter’s standard SANS (2015) level, the concentration of each parameter, the total number of observed parameters, and the weight coefficient of each parameter, respectively. Based on the SPI values, water is classified into five categories such as suitable for drinking (SPI < 0.2), slightly polluted water (SPI 0.2–0.5), moderately polluted (SPI 0.5–1.0), highly polluted (SPI 1.0–3.0), and unfit for drinking (SPI > 3.0) (Solangi et al. 2019; Egbueri and Unigwe 2020; Egbueri 2020).

Water Quality Index (WQI)

The WQI was developed by Horton (1965) and it has been widely used by various research scholars to evaluate drinking water quality. Computing the WQI involves five steps; (i) assignment of weights (wi) to each water quality parameter being analysed, (ii) calculation of relative weights (Wi) of parameters using Eq. 8, (iii) quality rating calculation (qi) based on Eq. 8, (iv) determining the sub-index value (SIi) of the analysed parameters based on Eq. 9, (v) calculation of the WQI based on Eq. 10.

$$W_{i} = \frac{{w_{i} }}{{\sum\nolimits_{i = 1}^{n} {w_{i} } }}$$
(7)
$$q_{i} = \frac{{C_{i} }}{{S_{i} }} \times 100$$
(8)
$${\text{SI}}_{i} = W_{i} \times q_{i}$$
(9)
$${\text{WQI}} = \sum\limits_{i = 1}^{n} {{\text{SI}}_{i} },$$
(10)

where qi represents the parameter quality rating, Ci represents the concentration of the chemical parameter (mg/L), and Si represents the drinking water standard prescribed by SANS/WHO. Weight values (wi) ranging from 1 to 5 were ascribed for each parameter according to their relative significance in the overall quality of drinking water and their indebted effects on human health. A maximum weight of 5 was allocated to the most significant parameters while a weight of 1 was assigned to the least significant parameters. Table 1 shows the weights and relative weights assigned while computing WQI.

Table 1 Relative weight of selected WQI parameters
Full size table

Health Risk Assessment

Intake of contaminated groundwater by elevated trace metal content is known to pose threat to human health. Evaluation of non-carcinogenic health risks due to trace metal contaminated groundwater particularly in children is crucial. According to guidelines provided by the United States Environmental Protection Agency (US EPA (1989), the chronic daily intake (CDI) risks caused by ingestion of a single trace element is computed for adults and children using the equation;

$${\text{CDI}} = \frac{{{\text{C}} \times {\text{IR}} \times {\text{EF}} \times {\text{ED}}}}{{{\text{BW}} \times {\text{AT}}}},$$
(11)

where CDI represents chronic daily intake via ingestion pathway (µg/kg/day), C is the concentration of the contaminant in drinking water (µg/L). IR represents the ingestion rate (L/day: 2.2 for adults and 1.8 for children). ED signifies the exposure duration (years: 70 and 6 years for adults and children). EF represents the exposure frequency (days/years: 365 for adults and children). BW is the body weight (kg: 70 for adults while 15 kg is for children). AT signifies the average time (days: 25 550 and 2 190 days for adults and children) (Duggal et al. 2017; Mgbenu and Egbueri 2019; Egbueri 2020; Mthembu et al. 2020). The non-carcinogenic risk of a single trace element is then calculated as the hazard quotient (HQ) using the equation.

$${\text{HQ = }}\frac{{{\text{CDI}}}}{{{\text{RfD}}}},$$
(12)

where RfD signifies the reference dose of each trace element (µg/kg/day). In this study, the RfD for the different trace metals are given as 0.3 (As), 0.5 (Cd), 40 (Cu), 700 (Fe), 24 (Mn), 1.4 (Pb), 300 (Zn), 0.3 (Co), 3 (Cr), and 20 (Ni) (Duggal et al. 2017; Mgbenu and Egbueri 2019; Egbueri 2020; Mthembu et al. 2020). The hazard index (HI) representing the non-carcinogenic risk of the heavy metal is computed and is obtained by adding the HQs values of each groundwater sample as shown by the following equation;

$${\text{HI = }}\sum {{\text{HQ}}}$$
(13)

HI values less than unity indicates that the non-carcinogenic health risk is within the acceptable limit whereas when the HI values are greater than unity it indicates that they are above the acceptable limit (Egbueri and Mgbenu 2020).

Results and Discussion

Heavy Metal Contamination

Table 2 shows the results of physicochemical parameters and their comparison to drinking water standards (WHO 2011; SANS 2015). The groundwater pH varied from 5.3 to 8.3 with an average value of 7.1 resulting the nature of water from acidic to alkaline. The pH of the samples indicated that most of the samples were within the standard limits of SANS (2015) but 30% of the samples were found below 6.5 as prescribed by WHO (2011). The electrical conductivity (EC µS/cm) varied from 115 to 1194 with an average of 330. Concentrations of cadmium are below the detection limit of 4.6 µg/L (mean 0.9 µg/L) and approximately 23% of the samples exceed the SANS (2015) and WHO (2011) standard limits. Zinc ranged from 1.9 to 19 964.5 µg/L (mean 499.9 µg/L). Approximately 2% of the samples exceeded the SANS (2015) and WHO (2011) standard limits for drinking water. Iron concentrations in groundwater ranged from 2.7 to 1 848.6 µg/L with an average range of 140 µg/L. About 11% of the samples have iron concentrations higher than SANS (2015) and WHO (2011) drinking water standard limits. The high iron concentrations may be owed to the leaching process of iron-rich sediments (Demlie et al. 2014). In this study, lead contents varied from below detection limit to 26 µg/L (mean 6.6 µg/L). Approximately, 25% of the samples have lead concentrations exceeding the SANS (2015) and WHO (2011) standard limits. Hence, the samples were classified as unfit for human consumption. Manganese concentrations ranged from 0.4 to 116 µg/L (mean 19.1). About 2% of the samples have manganese concentrations above the SANS (2015) standard limit. It was also observed that arsenic, chromium, copper, cobalt and nickel concentrations are low in groundwater samples and are within the prescribed standard limits of drinking water as prescribed by SANS (2015) and WHO (2011). Furthermore, iron, lead and cadmium were recognised to be dominant heavy metal contaminants in this area. Figure 2a–c shows the spatial distribution of selected heavy metals in the study area. In this study, high iron concentrations are observed in the north-western and south-eastern parts of the area (Fig. 2a). High lead concentrations are recorded in the south-eastern, northern, central, and southwestern parts of the area (Fig. 2b). The spatial distribution of cadmium illustrates that high concentrations are recorded in the central, northern and south-western parts of the area (Fig. 2c). The relationship between groundwater pH and the metal load was computed to classify the water samples (Ficklin et al 1992; Caboi et al 1999). Figure 3 shows that 43% and 57% of the samples are classified as a near-neutral high metal and near-neutral extreme metal. The high acidic pattern plotted against pH could be due to wastewater discharge and the inputs from agrochemicals used during irrigation (Islam et al., 2020; Church, 1998).

Table 2 Statistical summary of physicochemical parameters of groundwater and water quality standard limits
Full size table
Fig. 2
figure 2

Spatial distribution of heavy metals a iron, b lead, and c cadmium

Full size image
Fig. 3
figure 3

Classification of groundwater samples (pH versus metal load)

Full size image

Heavy Metal Pollution Assessment

Heavy Metal Pollution Index (HPI)

The computed HPI ranged from 0.28 to 128.08 with a mean value of 30.65, respectively (Table 3). The critical value for HPI for drinking water is 100 values < 100 indicates low pollution levels while values of > 100 indicates high water pollution (Mohan et al 1996; Prasad and Bose 2001). In this study, 21% of the samples exceeded the critical value of 100 indicating that these samples are unfit for consumption with respect to heavy metals (Fig. 5). The HPI can be classified into three categories such as low class (< 15), medium class (15–30) and high class (> 30) (Edet and Offiong 2002; Tiwari et al. 2017). Table 4 and Fig. 4a demonstrates that the HPI was categorised into low (71%), medium (4%) and high (25%). The high HPI values were spatially located in the northern, central and south-western part of the area which could be derived from infiltrated pesticides and fertilizers from agricultural land.

Heavy Metal Evaluation Index (HEI)

The values of HEI varied from 0.08 to 10.25 with an average value of 1.83 (Table 3), respectively. The HEI can be classified as low (< 400), medium (400–800) and high pollution (> 800) (Edet and Offiong 2002). In this study, all the groundwater samples were found below the 400 (Table 4; Fig. 4a), respectively. This suggests that the groundwater has low pollution with regards to heavy metals and is safe for human consumption.

Table 3 Results of the various quality indices
Full size table
Table 4 Classification of pollution indices
Full size table
Fig. 4
figure 4

Spatial distribution of a heavy metal pollution indices and b drinking water quality indices

Full size image

Drinking Water Quality Assessment

Synthetic Pollution Index (SPI)

The SPI is employed to classify drinking water into five categories: suitable for drinking (SPI < 0.2), slightly polluted water (SPI = 0.2–0.5), moderately polluted water SPI = 0.5–1.0), highly polluted water (SPI = 1.0–3.0), and unfit for drinking (SPI > 3.0) (Solangi et al. 2019; Egbueri and Unigwe 2020; Egbueri 2020). In this study, the SPI values varied from 0.19 to 1.28 with an average value of 0.47 (Table 3). Based on the SPI classification, 2% of the samples are classified as suitable for drinking, 74% of the samples are classified as slightly polluted with heavy metals while 24% of the samples are classified as moderately polluted (Table 4; Fig. 4b).

Water Quality Index (WQI)

The water quality index results are shown in Table 1. In the present study, the WQI varied from 7.27 to 89.44 with an average of 25.91 (Table 3). The WQI can be classified into five categories such as WQI < 25 indicate excellent water quality; 25–50 indicate good water quality; 50–75 indicate poor water quality; 75–100 indicate very poor water quality, and WQI > 100 indicate unsuitable water quality (Rostami et al 2019). The interpolated spatial distribution map of the study area (Fig. 4 b) depicts that 70% of the groundwater samples fall in the excellent water category, 9% of the samples belong to good water, 19% of the samples belong to poor water quality, and 2% of the samples are unsuitable for drinking purposes. It is shown that groundwater samples that fall in the poor and very poor water category are situated in the central, northern and southern parts of the study region.

Multivariate Statistical Analysis for Identification of Pollution Source

Correlation Matrix (CM)

Correlation matrix identifies the interrelationship among parameters and their possible sources in groundwater (Table 5). Correlation coefficients r > 0.7 is considered as strong, 0.7 > r > 0.5 as moderate and < 0.5 were considered as weak (Egbueri and Mgbenu 2020). The groundwater pH showed a moderate correlation with cadmium (r = 0.57), cobalt (r = 0.55), and HPI (r = 0.55). Iron showed a strong relationship with HEI, this suggests that iron has a key role in determining groundwater quality (Table 5). Lead showed significant correlation with chromium (r = 0.90), cadmium (r = 0.92), cobalt (r = 0.94), HPI (r = 0.96 0 and HEI (r = 0.84). A high concentration of lead and cadmium may be indicative of anthropogenic sources such as agrochemicals (Arslan and Ayyildiz Turan 2015). Chromium has shown strong correlation with cadmium (r = 0.89), cobalt (r = 0.94), HPI (r = 0.99), and HEI (r = 0.71). Furthermore, cadmium show a strong positive correlation with cobalt (r = 0.97), HPI (r = 0.99), and HEI (r = 0.86). Lead, chromium, cadmium and cobalt shows a significantly strong correlation with HPI and HEI. The correlation among heavy metal pollution indices reflected good relation between HPI and HEI (r = 0.87). This indicates that HPI and HEI can be used to evaluate the risk and contamination of heavy metals in the groundwater of this area.

Table 5 Pearson correlation matrix of analysed parameters
Full size table
Fig. 5
figure 5

Comparison between pollution evaluation indices

Full size image

Principal Component Analysis (PCA)

The principal component analysis identifies sources of heavy metals in groundwater. Varimax rotation was used with Kaiser normalization. Component loadings, Eigen values, percentage of variance and cumulative percentages of the identified principal components (PCs) are shown in Table 6. Figure 6a–c shows the spatial distribution maps of the principal components that were extracted. PCA revealed three principal components that accounted for 70.18% of the total variance. PC1 which explains 37.59% of the total variance has significant positive loadings of lead, chromium, cadmium and cobalt. The presence of these heavy metals in PC1 revealed the contribution of anthropogenic sources such as the application of agricultural fertilizers and pesticides, leaching or infiltration of domestic wastes and garden refuse from the Mbazwana landfill site (Mthembu et al. 2020). The highest values of PC1 are situated in the northern, central and southern parts of the study area, especially in samples BH12, BH15-16, BH26, BH47-53 (Fig. 6a). PC2 accounted for 19.65% of total variance with significant loadings of pH and cobalt, and moderate loading of zinc. This suggests that the groundwater pH is responsible for the release of cobalt and zinc into groundwater. Moderate PC2 values are distributed throughout the study area with the highest values seen at the central part of the study area at sample BH21 (Fig. 6b). The presence of zinc is owed to the use of agricultural fertilizers (Vetrimurugan et al. 2016; Rezaei et al. 2019). PC3 explains 12.9% of the total variance with a negative moderate loading of arsenic, strong positive loading of manganese and moderate loading of nickel. The negative loading of arsenic suggests a different source of origin. The strong and moderate loadings of manganese and nickel may be due to the weathering of manganese and nickel-bearing minerals. Landfill leachate may also contribute to the occurrence of manganese in the groundwater of this area. For PC3, the distributed loadings were high on the eastern and north-western part of the study area for samples BH27-28, BH51-52 (Fig. 6c). For better understanding, three principal components were overlaid into a single map using fuzzy overlay (Fig. 6d). The highest loadings are observed in the northern and central part of the study area with moderate values spread throughout the study area. This indicates that the samples in the northern and central parts of the study area are highly influenced by anthropogenic and weathering processes. This also corresponds with the high values of heavy metals observed in these locations (Fig. 2a–c).

Table 6 Varimax rotated principal component analysis
Full size table
Fig. 6
figure 6

Spatial distribution of principal components in the study area a PC1, b PC2, c PC3, d final overlay map of principal components

Full size image

Hierarchical Cluster Analysis (HCA)

The potential sources of pollutants in groundwater of this area were further investigated by carrying out cluster analysis. R-mode cluster analysis was carried out to evaluate the sources of heavy metals in groundwater of this area. R-mode cluster analysis revealed two groups of clusters (Fig. 7a). Cluster 1 consists of pH, chromium, cadmium, cobalt, arsenic, nickel and lead. This cluster indicates the influence of both geogenic (weathering of rock minerals) and anthropogenic sources (domestic wastes and agricultural fertilisers) in the study area. Cluster 2 comprises manganese, copper, iron and zinc and is due to the influences of geogenic and anthropogenic sources. Q-mode cluster analysis was also used to determine the similarities that exist between sampling points. Two groups of clusters were identified by Q-mode cluster analysis (Fig. 7b). Cluster 1 consisted of 46 sampling locations. The samples of this cluster are mainly located throughout the entire study area and its characteristics may be linked to the rocks found in this area such as arenite, sandstone, siltstone and alluvium. Furthermore, these sample locations are characterised by elevated average concentrations of iron, zinc, copper and manganese (Fig. 7b). These high levels are associated with PC2 and PC3, respectively. This cluster is associated with pollution by mixed sources i.e. weathering of soil or rock minerals and anthropogenic sources (agrochemicals). Based on average concentrations, pH, arsenic, nickel and copper were elevated in cluster 1 than in cluster 2. Cluster 2 included 7 sampling locations. Cluster 2 samples are spatially situated in the southern, central and north-western parts of the study area. This cluster is characterised by rocks such as arenite, sandstone and alluvium. These sampling locations have the highest concentrations of zinc, iron, manganese, and lead. These high levels correspond to PC1, PC2, and PC3, respectively. This cluster is also linked to mixed sources i.e. weathering of soil or rock minerals and anthropogenic sources. According to average values, iron, manganese, zinc, lead, chromium, cadmium and cobalt were greater in cluster 2 than those in cluster 1.

Fig. 7
figure 7

Dendrogram grouping of analysed parameters from Q-mode and R-mode HCA

Full size image

Health Risk Assessment

Health risk assessment evaluated the potential health risks of ingestion of trace metal contaminated groundwater in children and adults of the study region. Table 7 outlines the statistical summary for the HQ of the selected heavy metals. The spatial variation of HQ and HI of children and adults are shown in Fig. 8a, b. The mean values of HQ followed the following decreasing trend: Co>Pb>Cd>Zn>As>Mn>Cu>Cr>Fe>Ni for both children and adults (Table 7; Fig. 8a, b). The average HQ values suggest that Co has shown the highest of the total non-carcinogenic health risk in the study region. The HI values for groundwater samples varied from 0.06 to 8.16 for children and from 0.02 to 2.14 for adults, respectively (Fig. 8a, b). HI value for certain elements is greater suggesting non-carcinogenic health risk for ingestion. Likewise, an HI value below one implies that it is within the acceptable limit. In this study, it was discovered that 43% and 26% of groundwater samples pose a non-carcinogenic health risk in children and adults, respectively.

Table 7 Statistical summary of the heavy metals’ hazard quotients (HQ) and hazard index (HI) for both children and adult
Full size table
Fig. 8
figure 8

Spatial variation of HQ and HI values in a children and b adults

Full size image

Conclusions

The present study addresses evaluating the extent of heavy metal contamination and their source of origin in groundwater of Maputaland coastal aquifer using indexed and statistical analysis. The following conclusions were made:

  • Groundwater is acidic to alkaline in nature. The dominance of heavy metals is in the order; Zn>Fe>Cu>Mn>Pb>Cr>Ni>Co>Cd>As. Of all the heavy metals, lead, cadmium and iron exceeded the WHO and SANS standard limits for drinking water in 25%, 23% and 11% of the samples.

  • The HPI varied from a minimum of 0.28 to a maximum of 128.08. Based on the HPI classification, the samples varied from low to high pollution class. About 21% of the samples have HPI values greater than the critical limit of 100, suggesting that they are critically contaminated with respect to heavy metals and are unfit for human consumption.

  • The SPI and WQI were used to evaluate the drinking water suitability. Based on SPI results, 2% of the samples were suitable while 74% and 24% were classified as slightly and moderately polluted with heavy metals, respectively.

  • The WQI revealed that 19% and 2% of the groundwater samples belong to the poor and very poor water quality category and are spatially situated on the central, northern and southern parts of the study region.

Multivariate statistical analysis of CM and PCA confirm the release of heavy metals into groundwater which is controlled by geogenic (weathering of rocks and minerals), anthropogenic sources (domestic wastes and agricultural fertilizers) and mixed sources including weathering of parent material and anthropogenic sources.

  • HQs is in the order: Co>Pb>Cd>Zn>As>Mn>Cu>Cr>Fe>Ni for both children and adult. The HI values varied from 0.06 to 8.16 for children and from 0.02 to 2.14 for adults. In this study, it was discovered that 43% of groundwater samples pose non-carcinogenic health risk in children while 26% of the samples posed health risks in adults.

  • Water in this region is moderately polluted with heavy metals, the residents are at risk of adverse health effects. Accordingly, since children's HQs and HIs are higher than adult’s, they are more likely to be at risk for non-carcinogenic health conditions.

  • This study could be useful to the local municipalities and water management authorities to monitor the water quality of the area before supply and consumption. The study also recommends that a proper mitigation plan should be enforced by the policy makers to ensure public health and to prevent further contamination of the aquifer.