Distribution, Concentration, and Ecological Risk Assessment of Trace Metals in Surface Sediment of a Tropical Bangladeshi Urban River

Abstract

Trace metal contamination in sediments is a global concern. This study aimed to assess the contamination level of trace metals, their sources, and ecological risk in surface sediments of Karnaphuli River—a tropical urban river in Bangladesh. Forty-five sediment samples were analyzed by atomic absorption spectrophotometry (AAS) for Cu, Fe, Zn, Pb, Cr, Cd, and Ni metals along with physicochemical parameters like pH and organic matter (OM). The pollution status and potential ecological risk were assessed by using the geo-accumulation index (I_geo), contamination factor (CF), and potential ecological risk index (PERI). Source identification of trace metals was performed by correlation analysis, cluster analysis, and principal component analysis (PCA). The results show that the range of Cu, Fe, Zn, Pb, Cr, Cd, and Ni concentrations were 0.62–1.61 mg/kg, 23.95–85.70%, 0.52–1.89 mg/kg, 7.99–12.90 mg/kg, 33.91–65.47 mg/kg, 0.77–1.17 mg/kg, and 2.73–5.36 mg/kg, respectively. The concentrations of Fe, Cd, and Cr were above the permissible limits while the contamination factor (CF) and geo-accumulation index (I_geo) values revealed that Fe and Cd were the most dominant pollutants. Cluster analysis and PERI exhibited significant anthropogenic intrusions of trace metals. A significant positive correlation between Fe-Cr, Cr-Ni, Fe-Ni, and Pb-Cd shows their common anthropogenic source and influences. PERI also revealed that Cr, Fe, and Cd have a significant contribution with a moderate to considerable potential threat.

1. Introduction

The persistence, toxicity, and bio-accumulative nature of heavy metals cause dangerous effects on the aquatic ecosystem [1,2,3]. A huge amount of toxic trace metals has entered the aquatic environments through anthropogenic activities and natural actions which cause metal contamination [4,5]. In aquatic environments, trace metals stay in two forms, dissolved and accumulated, but a large portion of trace metals remain accumulated in sediments [6,7,8,9]. Therefore, sediments have been broadly considered as environmental sinks for the assessment of trace metal pollution in the river [1]. These metals can be accumulated in sediments because of natural processes like geological weathering and erosion of rocks; therefore, industrial discharge, agricultural activities, leaching from garbage and atmospheric deposition also play a great role [10,11,12]. These toxic metals accumulate in the aquatic organisms; finally, they enter the human body from the food chain, which poses different carcinogenic and non-carcinogenic health effects [4,13].

The government of Bangladesh has stepped up efforts on the blue economy. The coast and its marine resources play a significant role in the construction of this worthy blue economy. However, marine resources are polluted by the ingestion of various toxic substances such as heavy metals, microplastics, polycyclic aromatic hydrocarbons (PAHs), pesticides, insecticides, etc. Developing countries like Bangladesh are facing great threats associated with trace metal contaminations [14,15]. These trace metals mainly originate from different industrial (leather, pulp and paper, battery, mining, and food processing) activities, domestic wastes, agrochemicals, and unplanned urbanization [16,17]. The discharged untreated industrial wastes and agrochemicals are continuously increasing the level of the metals and worsening the sediment and water quality in the rivers of Bangladesh [18,19]. Karnaphuli River is one of the largest rivers in Chittagong, the industrial city of Bangladesh, located near the seaport area [16]. In Bangladesh, the upstream of Karnaphuli River passes through Kaptai lake, and its downstream finally meets with the Bay of Bengal. Different land-use activities along the Karnaphuli River continuously affected the water and sediment quality over time [2,20]. The river ecosystems are disturbed by the ingestion of various toxic substances such as trace metals, pesticides, insecticides, etc. The river receives a huge number of untreated effluents from industries like dying, cotton, steel mills, tanneries, ship breaking, pulp and paper, food, fertilizer, pharmaceuticals, steel, power stations, oil refineries, etc. Higher amounts of trace metals such as copper (Cu), cadmium (Cd), Iron (Fe), lead (Pb), chromium (Cr), zinc (Zn), and nickel (Ni) are discharged into the Karnaphuli River. Those trace metals are deposited in sediments and finally accumulate in aquatic organisms. These industrial effluents are considered a threat to increasing trace metal pollution both for freshwater and marine ecosystem and pose a serious risk to the surrounding dwellers’ health [21,22].

Very little research regarding trace metal pollution in the Karnaphuli River Basin has been conducted. For example, Ahmed et al. [21] studied trace metal bioaccumulation in selected fishes of Karnaphuli River and concluded there are potential health risks for children and adults. Ali et al. [22] reported trace metal contamination potentials of fish species named Tenualosa ilisha, Sillaginopsis panijus, Otolithoides pama, Setipinna phasa, Pampus chinensis, Harpadon nehereus, Polynemus paradiseus, and Gudusia chapra. They found higher risk levels of As and Pb compared to Cd and Cr. People who consume contaminated fish are at high risk of chronic cancer [23,24]. It has been reported that the water quality of Karnaphuli River is degrading due to deteriorated physio-chemical properties and higher concentrations of trace metals like Fe, Mn, Cu, Cr, Cd, Zn, and Pb [23,25,26]. Increased developmental and industrial activities along the river are the key reasons for the increasing concentration of these trace metals. Karnaphuli River sediments are highly contaminated with As, Cr, Cd, and Pb. They may generate an adverse effect on its ecosystem because of the developmental and industrial growth in recent years along the bank of the river [2,15]. Moreover, there is an enormous impact of seasonal variation on Karnaphuli River sediments. A study revealed that seasonal water tides and water discharge (pre-monsoon, monsoon, and post-monsoon) from industries have an impact on the mobilization of metals [27].

Until now, most of the earlier studies have emphasized the trace metal pollution of water, sediment, and fishes in the downstream of the river near the port area. A thorough study of trace metals contamination in sediment considering different flow regions of the Karnaphuli River is still scarcely investigated. This study intends to close the research gap in the earlier studies. Therefore, it is very essential to determine the pollution status of trace metals and their associated potential ecological risk of Karnaphuli River sediment in different flow regions. The present work was focused to assess the trace metals contamination profile by spatial distributions of Cu, Cr, Pb, Fe, Zn, Ni, and Cd in Karnaphuli River sediment from Baraichari to Patenga road. Potential ecological risk and the distribution of trace metals were also studied by different pollution indexes and multivariate statistical techniques.

2. Materials and Methods

2.1. Description of the Study Area

Karnaphuli River originates from the Lushai Hills in Mamit District, Mizoram state, India. The river passes through 180 km at Rangamati in Bangladesh and then runs about 170 km through the Chittagong port city which finally meets the Bay of Bengal [15,20]. The upstream of the Karnaphuli River flows through the hilly regions, the middle of the river is surrounded by mainly ghat and agricultural areas while the Chittagong Economic Zone (CEPZ) and different industries are located near the downstream. Geologically, the Karnaphuli River contains a deep layer of tertiary rocks which is covered with sedimentary deposits layers and overlying with mud and sand [28]. The maximum temperature during summer reaches 32.3 °C and the minimum temperature in winter is 13 °C. Annual rainfall in the area was about 3000 mm and approximately 95% of the rainfall happens within the months of May to September [28,29]. As it is the largest river in Chittagong City, that is why people of this city use the river water for different aspects like drinking and household purposes. For the convenient source of water, numerous industries have been developed along the river [2,15]. Among them, a large number of industries are mainly situated in the region between Kalurghat to Patenga, for this study most of the sampling sites were taken between these areas. As we collected water and sediment samples from the same sampling sites at the same time, we use Figure 1 from another published work by Hasan et al. [30].

2.2. Sediment Sample Collection and Preparation

Forty-five sediment samples were taken from fifteen sampling sites with 3 replicates from the Karnaphuli River (Figure 1). Approximately 200 g of surface sediment samples from each site were taken from the depth of 0 to 15 cm using an Ekman dredge sampler (stainless steel grab sampler). The sandy sediment samples were ignored carefully while sampling. The grain size of collected sediment samples was within 0.002 mm to 0.0156 mm (Wentworth scale). The positions of sampling sites were taken by a global positioning system (GPS) tool. The samples were stored in clean zip-locker polyethylene bags and kept in air-dried conditions maintaining a 4 °C temperature. Samples were transferred to the laboratory of Bangladesh Agricultural Research Institute (BARI), Dhaka immediately. Plastic trays were used for spreading the samples and then dried at normal temperature for 8 days. The samples were crushed with a ceramic-coated grinder and sieved through a nylon sieve (10 mesh) and kept in labeled polypropylene containers at ambient temperature before analysis [31].

2.3. Analysis of Sediment Physicochemical Parameter

The pH of sediments was measured at a sediment-to-water ratio of 1:2.5 [32]. A representative portion of the 10 g sediment sample was taken and placed into the beaker by using a spoon. A total of 2.5 times the volume of water (25 mL) was added to the beaker, and the suspension was shaken by a magnetic stirrer for about 10 min. After 30 min the suspension was shaken another 2 min again. The pH of the suspension was recorded after stabilization of the suspension by a pH meter [33].

The sediment organic carbon (OC) was analyzed by using the Walkey and Black wet oxidation method and for an oxidizing agent, potassium dichromate was used. The filtrate was titrated with ferrous ammonium sulfate in the presence of a diphenylamine indicator to a dull green endpoint. The organic carbon content in sediment was determined by Equation (1) as follows [34]:

$$\% \mathrm{C}=\left(\mathrm{B}-\mathrm{T}\right) \mathrm{N} \times 0.003\times 1.33\times \frac{100}{\mathrm{W}\left(2\mathrm{g}\right)}$$

(1)

where B and T are the volume (in mL) of ferrous ammonium sulfate solution (needed for blank titration) and the volume (in mL) of ferrous ammonium sulfate (required for sediment samples), respectively; N is the normality of ferrous sulfate used. The amount of organic matter (OM) in sediment samples was calculated by multiplying the content of organic carbon by 1.73 (Van Bemmelen factor) [35].

Quality assurance (QA) and quality control (QC) verified that data were delivered regularly with the least error for doing the study. All of the tests were carried out for the three replicates to get rid of any errors and only average values were taken. All laboratory equipment was cleaned with distilled water and soaked in HNO₃ (10%) for at least 24 h to avoid contamination [3]. Analytical blank samples and spike samples for each trace metal were used to ensure quality assurance and control. The AAS was calibrated depending on the standard laboratory measures. The wavelengths (nm) of AAS were 217.0, 324.8, 228.8, 232.0, 357.9, 248.3, and 213.9 for Pb, Cu, Cd, Ni, Cr, Fe, and Zn respectively. The detection limits of AAS (mg/L) were 0.001 for measuring all the trace metals.

2.4. Digestion of Samples and Determination of Trace Metals

One gram (1 g) of sediment was digested with HNO₃ and HClO₄ (5:1 volume) for determining the trace metal as per Misra and Chaturved [36]. Digestion was executed at a temperature of 190 °C for 1.5 h. The samples were taken into a 50 mL volumetric flask and the solution was made with distilled water after cooling. The trace metal concentrations like Fe, Cd, Cr, Cu, Zn, Pb, and Ni were analyzed by atomic absorption spectrophotometer (VARIAN model AA2407). Each sample was tested three times to get accurate results and the data stated in mg kg⁻¹.

2.5. Trace Metal Pollution Assessment in Sediment

2.5.1. Geo-Accumulation Index (I_geo)

The I_geo index helps to assess the metal contamination in sediments, by comparison with the current concentration. The I_geo index for the metals can be determined by using the equation below [37,38]:

$${ \mathrm{I}}_{\mathrm{geo}}={ \log }_2 \left[\frac{{\mathrm{C}}_{\mathrm{n}}}{1.5{ \mathrm{B}}_{\mathrm{n}}}\right]$$

(2)

where C_n is the metal concentration level in the sediment, B_n is the background value of a given metal [39], and the factor 1.5 is used to account the possible variations in the background values. Muller [40] stated seven grades for the classification of the geo-accumulation index which are I_geo < 0 (uncontaminated); 0 < I_geo < 1 (uncontaminated to moderately uncontaminated); 1 < I_geo < 2 (moderately contaminated); 2 < I_geo < 3 (moderately to heavily contaminated); 3 < I_geo < 4 (heavily contaminated); 4 < I_geo < 5 (heavily to extremely contaminated); 5 < I_geo (extremely contaminated).

2.5.2. Contamination Factor (CF)

The sediment contamination level by metal is expressed as a contamination factor (CF) which is recommended by Hakanson [41] and calculated by Equation (3):

$$\mathrm{CF}=\frac{{\mathrm{C}}_{\mathrm{n}} \mathrm{sample}}{{\mathrm{B}}_{\mathrm{n}} \mathrm{shale}}$$

(3)

where C_n is the concentration level of metals in the target area and B_n is the background value of the given metal in the shale [38]. CF being classified in four classes for describing the contamination level are CF < 1 (low contamination); 1 ≤ CF < 3 (moderate contamination); 3 ≤ CF < 6 (considerable contamination); CF > 6 (very high contamination) [32].

2.5.3. Potential Ecological Risk Index (PERI)

The potential risk of individual metal $({\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}})$ and potential ecological risk index (PERI) were suggested by Hakanson [41] to determine the risk from the contamination of trace metals of different aquatic organisms. PERI can be calculated via the following equations:

$${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}={\mathrm{T}}_{\mathrm{R} }^{\mathrm{i}}\times { \mathrm{C}}_{\mathrm{f}}^{\mathrm{i}}$$

(4)

$$\mathrm{PERI}={\displaystyle \sum }_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$$

(5)

where ${\mathrm{C}}_{\mathrm{f}}^{\mathrm{i}}$ is the pollution of a single element factor; ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ is the potential ecological risk of a single element; PERI is the sum of ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$, and ${\mathrm{T}}_{\mathrm{R}}^{\mathrm{i}}$ is the biological toxic factor of a single element. It is determined for Zn = 1, Cr = 2, Cu = 5, Pb = 5, Ni = 5, and Cd = 30 [42,43]. Five classes of ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ and four classes of PERI are shown in

Table 1. Classification of PERI and potential ecological risk for a single regulator.

Class	Risk for Single Regulator	Pollution Degree	Potential Ecological Risk (PERI)
1	${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ < 40	Low	PERI > 95
2	40 ≤ ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ ≤ 80	Moderate	95 ≤ PERI ≤ 190
3	80 ≤ ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ ≤ 160	Considerable	190 ≤ PERI ≤ 380
4	160 ≤ ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$ ≤ 320	High	PERI ≥ 380
5	320 ≤ ${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$	Very High

[44,45,46].

2.6. Statistical Analysis

The statistical package SPSS 25 was used to statistically analyze the data. As the Shapiro–Wilk test showed the normal distribution of data, Pearson’s bivariate correlation was carried out to evaluate the inter-element relationship in sediments. Cluster analysis of trace metals and principal component analysis (PCA) were performed by ORIGIN (2018 version) software to determine the sources of contamination.

2.7. Geostatistical Method

The inverse distance weighted (IDW) method was used to show the spatial distribution of trace metals in the surface sediment. The IDW technique calculated an average value for unsampled locations using values from nearby sampled locations. ArcGIS 10.4.1 was used for representing the study map and illustrating the geospatial distribution of trace metals in the sediments of the Karnaphuli River.

3. Results and Discussions

3.1. Physicochemical Parameter of Sediments

The results of the examined parameters such as pH and OM are given in

Table 2. Trace metal concentrations (mg/kg ± SD) in the surface sediments of the Karnaphuli River (n = 3).

Sample Station	pH	OM (%)	Cu	Fe (%)	Zn	Pb	Cr	Cd	Ni
S1	6.38 ± 0.15	3.30 ± 0.40	1.07 ± 0.61	58.27 ± 19.93	0.93 ± 0.50	11.50 ± 1.60	45.09 ± 4.69	1.05 ± 0.13	3.75 ± 0.39
S2	7.95 ± 0.62	1.70 ± 0.11	0.62 ± 0.06	27.03 ± 7.24	0.52 ± 0.04	10.44 ± 2.31	37.41 ± 1.89	0.92 ± 0.21	3.08 ± 0.21
S3	7.05 ± 0.27	3.70 ± 0.88	0.87 ± 0.09	85.70 ± 7.29	0.74 ± 0.06	10.59 ± 1.01	51.11 ± 7.15	0.95 ± 0.10	4.29 ± 0.60
S4	6.67 ± 0.14	3.60 ± 1.03	1.18 ± 0.24	78.16 ± 9.55	1.01 ± 0.17	7.99 ± 1.55	45.74 ± 3.05	0.81 ± 0.19	3.84 ± 0.28
S5	5.87 ± 0.20	4.60 ± 1.11	1.53 ± 0.09	58.65 ± 11.93	1.89 ± 0.27	8.53 ± 1.32	65.47 ± 8.19	0.77 ± 0.12	5.36 ± 0.68
S6	7.33 ± 0.41	4.10 ± 1.22	1.61 ± 0.33	49.40 ± 4.81	1.38 ± 0.28	10.20 ± 1.33	40.06 ± 12.07	0.94 ± 0.10	3.38 ± 1.02
S7	7.59 ± 0.17	2.90 ± 0.98	1.31 ± 0.22	53.91 ± 4.12	1.10 ± 0.17	11.43 ± 3.03	48.55 ± 7.60	1.02 ± 0.29	3.85 ± 0.91
S8	7.47 ± 0.30	3.00 ± 0.88	1.32 ± 0.04	53.18 ± 8.67	1.09 ± 0.03	12.90 ± 2.36	46.96 ± 4.11	1.17 ± 0.21	4.10 ± 0.44
S9	7.40 ± 0.16	3.60 ± 0.87	1.09 ± 0.17	45.88 ± 2.15	1.04 ± 0.12	11.78 ± 1.17	40.57 ± 5.21	0.99 ± 0.13	3.73 ± 0.46
S10	7.53 ± 0.10	4.00 ± 1.34	1.54 ± 0.13	23.95 ± 1.07	1.22 ± 0.19	10.12 ± 0.39	40.72 ± 2.97	0.93 ± 0.05	3.13 ± 0.22
S11	8.30 ± 0.31	3.70 ± 1.14	1.14 ± 0.13	26.70 ± 3.37	0.84 ± 0.13	11.84 ± 0.88	40.32 ± 4.34	1.06 ± 0.11	3.56 ± 0.23
S12	8.05 ± 0.14	3.30 ± 1.20	1.10 ± 0.11	25.66 ± 5.27	0.94 ± 0.09	8.60 ± 0.95	35.62 ± 0.84	0.77 ± 0.09	3.06 ± 0.12
S13	7.71 ± 0.02	5.40 ± 1.45	1.33 ± 0.19	42.37 ± 8.87	1.53 ± 0.26	10.41 ± 0.64	35.91 ± 1.94	0.93 ± 0.08	2.94 ± 0.12
S14	7.62 ± 0.03	3.30 ± 1.11	1.12 ± 0.08	29.63 ± 3.66	0.93 ± 0.05	10.57 ± 1.86	34.17 ± 2.87	0.96 ± 0.17	2.93 ± 0.29
S15	7.56 ± 0.21	4.00 ± 1.23	1.12 ± 0.07	28.31 ± 8.49	0.89 ± 0.08	8.89 ± 1.12	33.91 ± 1.57	0.77 ± 0.13	2.73 ± 0.06
Range	5.87–8.30	1.70–5.40	0.62–1.61	23.95–85.70	0.52–1.89	7.99–12.90	33.91–65.47	0.77–1.17	2.73–5.36
Mean ± SD	7.36 ± 0.65	3.61 ± 0.83	1.20 ±0.26	45.79 ± 19.41	1.07 ± 0.33	10.39 ± 1.40	42.77 ± 8.25	0.94 ± 0.12	3.58 ± 0.68

. The correlation result of trace metals with pH and OM can evaluate the influence of the trace metals’ presence in river sediment. The pH level at sampling sites S1, S4, and S5 revealed low levels of pH (6.38, 6.67, and 5.87). These sites showed the acidic condition of the surface sediment. The acidic condition was mainly due to the substantial amount of organic matter that exists in the study area [47]. The pH level of sediment at sites S2, S11, and S12 had high values of 7.95, 8.05, and 8.30, respectively, which were alkaline in nature. Specific sampling sites were in alkaline conditions due to the intense deposition of the calcareous materials for different anthropogenic sources and seawater intrusion into the Karnaphuli River (Figure 1). The organic matter content in the sediment ranged from 1.70% to 5.40%. This study revealed that S13 had the highest average OM percentage (5.4%), while S2 had the lowest percentage of OM. The sampling sites between Kalurghat to Patenga are rich with OM compared to other sampling sites due to the discharge of effluent from paper mills, textile industries, asbestos industries, and paint industries [48,49].

Figure 2 portrays the correlation of pH and OM with trace metals of sediment. The concentration of Cu, Fe, Ni, Cr, and Zn showed a negative correlation with pH values among them Fe, Ni, and Cr presented a significant negative correlation [50]. It indicates that pH may have potential effects on the distribution of these trace metals in the sediment of the river. For example, S11 and S12 had lower concentrations of Fe, Ni, and Cr due to the higher pH of sediment at these sites. There is a small positive correlation between Cd and Pb and pH, which is depicted in S8. Cu and Zn were powerfully correlated (p < 0.01 for Zn and p < 0.05 for Cu) with the OM content in the sediment. In contrast, Fe, Cr, and Ni were weakly correlated, presenting that the OM content in these sites powerfully controlled the distribution of Cu and Zn. The significant correlation between Cu and Zn and OM meant that the OM reduces the mobility of these heavy metals and is associated with their accumulation in sediment. These findings agreed with the results of Yohannes et al. [51]. The insignificant correlation of Cr, Ni, and Fe with organic matter depict that the mentioned trace metals might be less bioavailable in the sediments due to their remobilizing trend in the oxidizing state, whereas Pb and Cd show a weak negative correlation with OM which is supported by the results of Shehzad et al. [52].

3.2. Concentration of Trace Metal in Sediment

The result of trace metal concentrations in Karnaphuli River sediments revealed that the studied metal concentration in sediment was in the order of Fe > Cr > Pb > Ni > Zn > Cu > Cd. The Fe concentrations were low as 23.95% at S10 and as high as 85.70% at S3 with an average concentration of 45.79%. According to Jain et al. [53], Fe is the most abundant metal in all sediments because it is a common element in the Earth’s crust. As the Karnaphuli River originates and flows through hilly regions, it carries sediment enriched with Fe because of natural processes like weathering. The sedimentary deposits of the Karnaphuli River, especially in hilly regions, are layered with banded iron formations which are coated with chert and are generally non-oolitic. Banded iron formations (banded hematite quartzite and banded magnetite quartzite) bear an anomalously high content of Fe. The extreme erosion of this type of sedimentary rock, its transportation of it with river flows, and its settlement in sediment increases the Fe concentration of the Karnaphuli River sediment. The concentration of Fe is higher than any other river in Bangladesh because of the abundance of ship-breaking industries and dockyards on the bank of the Karnaphuli River. Mamun et al. [54] stated that the Karnaphuli River is highly polluted by Fe and Al. The high concentration of Fe at S3 was in the Bazar area where wastewater was directly discharged into the river from this area. The average value of the Cr concentration was 42.77 mg/kg and the highest concentration of Cr was at S5 (65.47 ± 8.19 mg/kg) while the lowest was at S15 (33.91 ± 1.57 mg/kg). The highest concentration of Cr was in the Kalurghat area due to the presence of many industries around this location. The concentration of Cr is high at this point because the leather industry, pulp and paper industry, shoe industry, and food industry discharge their untreated wastewater in the Karnaphuli River from this point. Cr was one of the toxic trace metals, it is the least mobile metal among the toxic metals and is greatly responsible for the chronic effects on the human body. The average concentration of Cr was exceeding the maximum permissible limit of 26 mg/kg proposed by USEPA [55]. Cd concentration ranged from 0.77–1.17 mg/kg with an average value of 0.94 mg/kg. The highest concentration of Cd was at S8 (1.17 ± 0.21 mg/kg) and the lowest was at S5 (0.77 ± 0.12 mg/kg) and S12 (0.77 ± 0.09 mg/kg). The average concentration of Pb was 10.39 mg/kg and the highest was at S11 (11.84 ± 0.88 mg/kg) while the lowest was at S4 (7.99 ± 1.55 mg/kg). The highest concentration of Pb at S11 may be due to the industrial area where many manufacturing activities are operated. The Cu concentration ranged from 0.62–1.61 mg/kg with an average value of 1.20 mg/kg. The highest concentration of Cu was reported at S6 (1.61 ± 0.33 mg/kg) and the lowest concentration was at S2 (0.62 ± 0.06 mg/kg). Finally, the average concentration of Zn was 1.07 mg/kg.

The concentrations of studied trace metals (Fe, Cr, Pb, Ni, Cd, Zn, and Cu) at the sampling sites S1, S2, S3, and S7 may be influenced by several anthropogenic activities as well as natural sources. Sampling sites S5, S6, S8, S9, S13, and S14 were located near industrial areas, semi-urbanized areas, port areas, and power station areas, and the concentration of trace metals in these sites is mostly attributed to anthropogenic sources. The Karnaphuli River directly receives untreated wastewater from different factories like leather, textiles, batteries, pulp and paper, domestic sewage, urban runoff, and wastes from the construction of residential and commercial areas. The highest percentage of Fe distribution was noticed at all sampling sites, which shows that above 90% of the trace metal pollution came from Fe because of natural weathering processes and erosion of Fe-enriched rocks as well as residential, commercial, and industrial activities. Apart from Fe, the largest amounts of trace metal pollution at S8 and S11 came from Cd and Pb. Intensive industrial and commercial activities at S8 and S11 may be responsible for such a quantity of Cd and Pb entering the river. The comparatively highest concentrations of Zn, Cr, Cu, and Ni were noticed at S5, S6, and S13 because of the construction, shipping breaking, and port activities near the sites.

The spatial distribution of trace metals for all sites of Karnaphuli River is presented in Figure 3. The map illustrates the distribution of trace metals in the river, the highest concentrations are shown by the red color and the yellow color indicates the lowest concentrations.

From the map, it is seen that the concentrations of Cu, Zn, Ni, and Cr were higher in S5 and S6, Pb was higher in S7 and S8, and Fe was higher in S3 and S4. This indicates that S5 (Kalurghat area) and S6 (industrial area) had many intense industrial and commercial activities nearby which contributed to the concentration of Zn, Cu, Cr, and Ni in the river. For power plant and port activities, Pb showed higher concentration in S8, S9, and S11. However, Cd showed a different distribution. Cd was higher in S1 and S8. S1 showed higher concentration because of intensive use of phosphate fertilizers in agricultural land and S8 showed for refining, manufacturing, and power plant activities. From the trace metal distribution pattern in the surface sediment, it is evident that for most metals the high concentration values were gathered in the S5, S6, S8, S9, and S11 regions. The high concentration values may be due to the massive anthropogenic and industrial activities along these sampling sites of the Karnaphuli River.

3.3. Comparison of Trace Metals in Sediment of Karnaphuli River with Sediment Guidelines and Previous Studies

The comparison of trace metal concentrations from this study with various international guidelines gives better viewpoints of the state of metal toxicity in the sediment. Another comparison of elemental concentrations with the average shale value (continental crust value) revealed the absence of pollution, except for Fe and Cd. For Fe, the average shale value is 4.72 (%) and the recorded value of Fe was 45.79 (%). The concentrations of Cd (0.94 mg/kg) exceed the average shale value (0.30 mg/kg). The average concentration of Cr exceeds the maximum permissible limits and threshold effect level (TEL) proposed by USEPA (1999) and MacDonald et al. (2000) [55,56] (

Table 3. Comparison of trace metals in sediment (mg/kg) with different international guidelines and other studies.

River (Locations)	Cr	Ni	Cd	Pb	Cu	Zn	Fe (%)	References
Karnaphuli River (Bangladesh)	42.77	3.58	0.94	10.39	1.20	1.07	45.79	Present study
Karnaphuli River (Bangladesh)	70.06	NA	1.51	38.33	NA	NA	NA	Ali et al. [2]
Old Brahmaputra River (Bangladesh)	6.6	12.8	0.48	7.6	6.2	52.7	NA	Bhuyan et al. [18]
Feni River (Bangladesh)	35.28	33.27	NA	6.47	NA	NA	NA	Islam et al. [57]
Halda River (Bangladesh)	8.84	15.97	0.04	8.80	5.90	79.58	NA	Bhuyan et al. [27]
Passur River (Bangladesh)	19.36	20.61	NA	6.91	15.83	NA	21,306.03 (mg/kg)	Shil et al. [58]
Meghna River (Bangladesh)	31.74	76.1	0.23	9.47	NA	79.02	1281.42 (mg/kg)	Hasan et al. [59]
Dhaleshwari River (Bangladesh)	27.39	NA	2.08	15.79	37.45	NA	NA	Ahmed et al. [60]
Perak River (Malaysia)	NA	NA	2.94	28.86	24.67	55.38	35.07	Salam et al. [3]
DRS River (China)	NA	NA	0.29	25.2	24.6	72.5	3.65	Lin et al. [61]
Tamagawa River (Japan)	NA	NA	0.15	14.4	28.77	72.7	4.01	Shikazono et al. [12]
Sediment Guidelines
ASV	90	68	0.30	20	45	95	4.72	Turekian and Wedepohl [38]
TRV	26	16	0.60	31	16	110	30	USEPA SQG [55]
TEL	37.3	35	0.59	18	35.7	123	NA	MacDonald et al. [56]

ASV: Average shale values; TRV: Toxicity reference value; TEL: Threshold effect levels.

). According to Ali et al. [2], the average concentration for Cr, Cd, and Pb on downstream of the Karnaphuli River was 70.06 mg/kg, 1.51 mg/kg, and 38.33 mg/kg respectively. Where this study was conducted on the upstream to downstream of Karnaphuli River, the average concentrations for Cr, Cd, and Pb were 42.77 mg/kg, 0.94 mg/kg, and 10.39 mg/kg, respectively. However, the average concentrations of the studied trace metals for the downstream flow region were higher than the average concentration of trace metals for the whole river.

According to Ali et al. [2] the average concentrations for Cr, Cd, and Pb on the downstream of the Karnaphuli River were 70.06 mg/kg, 1.51 mg/kg, and 38.33 mg/kg, respectively. Where this study was conducted on the upstream to downstream of Karnaphuli River, the average concentrations for Cr, Cd, and Pb were 42.77 mg/kg, 0.94 mg/kg, and 10.39 mg/kg, respectively. However, the average concentrations of the studied trace metals for the downstream flow region were higher than the average concentration of trace metals for the whole river.

3.4. Risk Assessment of Trace Metals

The pollution index extensively used to assess the trace metal pollution in the surface sediments is the geo-accumulation index (I_geo) [40,62]. Based on I_geo results the trace metals pollution intensity order was Fe > Cd > Pb > Cr > Ni > Cu > Zn. I_geo values of Cu, Zn, Pb, Cr, and Ni were lower than 0 which is categorized as uncontaminated, whereas the I_geo values of Fe and Cd were categorized as moderately to heavily contaminated and moderately contaminated, respectively (Figure 4). Among the targeted trace metals Fe showed the highest accumulation I_geo values at S3 (3.60), which was reported to class 4 (moderately to heavily contaminated). Cd showed the second-highest accumulation value of a maximum of 1.90 (S3) which stated that the samples from S3 were moderately contaminated. I_geo values ranged from −6.77 to −5.38 for Cu, 1.76 to 3.60 for Fe, −8.09 to −6.23 for Zn, −1.90 to −1.21 for Pb, −1.99 to 1.04 for Cr, 0.77 to 1.90 for Cd, and −5.22 to −4.25. The index of geo-accumulation (I_geo) shows that the Karnaphuli River is not contaminated with Cr, Zn, Pb, Ni, and Cu because their I_geo values are lower than 0.

The contamination factor (CF) values of Cd, Fe, Cr, Pb, Ni, Zn, and Cu were compared with the different levels of contamination degrees suggested by Islam et al. [32]. Total contamination factors followed the order of Fe > Cd > Pb > Cr > Cu > Ni. The highest CF value of Fe was at S4 (16.36) and the lowest was at S10 (5.07) which is categorized as very high contamination. The reported CF values for Cd ranged between 2.57 to 3.90 among all the stations. The highest value was revealed at S8 (3.90) and the least was at both S5 and S15 (2.57) (Figure 5).

The contamination factors of Fe and Cd were high in the Karnaphuli River due to the natural as well as the anthropogenic sources. The discharge of effluent from power/desalination plants, printed circuit board (e-waste), electroplating and refining industries, and the ship braking industries are responsible for the increased concentration of Fe and Cd. Cd is readily absorbed and rapidly translocated trace metal and has dangerous effects on aquatic organisms as well as the human body [63,64]. Potential ecological risk for single trace metals in sediments were in the order of Cd > Pb > Cr > Ni > Cu > Zn. Cd showed a higher ecological risk due to the excessive use of phosphate fertilizers in the agricultural field along the river and waste disposal from the city [65]. A total of 73% of sampling sites showed a moderate pollution degree. Among them, S7, S8, S9, and S11 had the highest PERI (

Table 4. Potential ecological risk factor and the potential ecological risk index (PERI) of trace metals in sediments of the Karnaphuli River.

Sample Station	Potential Ecological Risk Factor (${\mathrm{E}}_{\mathrm{R}}^{\mathrm{i}}$)						Risk Index (PERI)	Pollution Degree
	Cu	Zn	Pb	Cr	Cd	Ni
S1	0.12	0.01	2.88	1.00	105	0.28	109.28	Moderate
S2	0.07	0.01	2.61	0.83	92	0.23	95.74	Moderate
S3	0.10	0.01	2.65	1.14	95	0.32	99.20	Moderate
S4	0.13	0.01	2.00	1.02	81	0.28	84.44	Low
S5	0.17	0.02	2.13	1.45	77	0.39	81.17	Low
S6	0.18	0.01	2.55	0.89	94	0.25	97.88	Moderate
S7	0.15	0.01	2.86	1.08	102	0.28	106.38	Moderate
S8	0.15	0.01	3.23	1.04	117	0.30	121.73	Moderate
S9	0.12	0.01	2.95	0.90	99	0.27	103.25	Moderate
S10	0.17	0.01	2.53	0.90	93	0.23	96.85	Moderate
S11	0.13	0.01	2.96	0.90	106	0.26	110.25	Moderate
S12	0.12	0.01	2.15	0.79	77	0.23	80.30	Low
S13	0.15	0.02	2.60	0.80	93	0.22	96.78	Moderate
S14	0.12	0.01	2.64	0.76	96	0.22	99.75	Moderate
S15	0.12	0.01	2.22	0.75	77	0.20	80.31	Low

). Ecological entities and aquatic organisms along these sites are at great risk compared to other sites.

3.5. Identification of Trace Metals Pollution Sources

Different multivariate analyses such as cluster analysis (CA), Pearson correlation, and principal component analysis (PCA) were used for recognizing the sources of trace metals and interpreting their spatial variations. Cluster analysis (CA) was used to cluster similar sampling sites and to identify exact areas of contamination [57]. Figure 6 displays that all sampling sites on the river are classified into two statistically significant clusters. Cluster 1 contains two sub-clusters. Cluster 1(a) includes seven sites (S1, S5, S6, S7, S8, S9, S13) with the highest trace metal concentrations which are surrounded by semi-urban, commercial, power plant, port, and industrial areas. Cluster 1(b) consists of Sites S3 and S4 which had high metal concentrations because these sites are located near the Bazar and commercial area, where untreated wastewater and domestic sewage were discharged directly to the river. Cluster 2 includes sites (S2, S10, S11, S12, S14, S15) that are less effected compared to cluster 1.

Metal to metal relations may display the sources and pathways of the metals present in the sediment. Correlation analysis was performed with Pearson’s correlation coefficients to investigate the degree of correlation among trace metals and is depicted in

Table 5. Pearson correlation analysis of trace metals of sediments in the Karnaphuli River.

	Cu	Fe	Zn	Pb	Cr	Cd	Ni
Cu	1
Fe	0.015	1
Zn	0.841 **	0.144	1
Pb	−0.101	−0.044	−0.234	1
Cr	0.283	0.657 **	0.483	−0.062	1
Cd	−0.028	0.067	−0.214	0.964 **	−0.010	1
Ni	0.223	0.676 **	0.431	0.028	0.974 **	0.069	1

Bold values represent correlation with significance. ** Correlation is significant at the 0.01 level (2-tailed).

. A strong positive correlation was noticed among the trace metal pairs in the sediment which were Cu–Zn, Fe–Cr, Fe–Ni, Pb–Cd, and Cr–Ni. These strong correlations pair may be indicated the same sources of these trace metals as well as similar geochemical features [3]. From the results, a solid proof of mutual dependence of these trace metals in the sediment was expected.

Principal component analysis (PCA) is a multivariate statistical technique that was performed to decrease the number of correlated variables and form a smaller number of uncorrelated variables by finding a combination of the original variables. It also helps to find out the relationship between the concentration of trace metals in sediments and their sources. Kaiser–Meyer–Olkin (KMO) values and Bartlett sphericity tests supported the reliability of PCA for the data set in this study [66]. A total of three PCs were extracted with eigenvalues higher than 1.0, describing 87.18% of the total variance (

Table 6. Component matrix of a two-factor model with strong to moderate loadings in sediment.

	PC1	PC2	PC3
pH	−0.42559	−0.11714	0.20314
OM	0.27934	−0.35917	0.30190
Cu	0.29395	−0.23626	0.49318
Fe	0.32270	0.31929	−0.22262
Zn	0.30524	−0.23902	0.39386
Pb	−0.16313	0.47887	0.45225
Cr	0.42922	0.27313	−0.09142
Cd	−0.12766	0.49136	0.46849
Ni	0.40859	0.32634	−0.07548
Eigen value	3.91542	2.28297	1.64747
Percentage of variance	43.50%	25.37%	18.31%
Cumulative percentage	43.50%	68.87%	87.18%

and Figure 7).

The first principal component (PC1) was dominated by Cr, Fe, and Ni, explaining 43.50% of the total variance. A strong correlation between Fe, Cr, and Ni shows that their major concentrations are entered into the environment through human-oriented activities. The source of Fe and Cr may possibly be the refining and steel industry, excessive use of ferrous sulfate fertilizers, leather industry, ship-breaking industries, and dockyards, as well as natural processes like weathering and erosion of iron-enriched rocks. Ni was released into the river from the battery and food processing industry [52]. The second component (PC2) consisted of Cd and Pb, accounting for 25.37% of the total variance. It supports the correlation analysis which was performed earlier. The sources of Cd and Pb would be the discharge of gasoline from boats, battery industries, power/desalination plants, electroplating and refining industries, and intensive use of cadmium fertilizers [52,64,65,66]. The third principal component (PC3) explained 18.31% of the total variation and was dominated by Cu, Zn, and OM. OM has a positive relation with Cu and Zn in Figure 2 which supports the PC3 results. The site scores from PC1 and PC2 show the same result observed in cluster analysis for the sites S1, S7, and S8 from cluster 1(a) and sites S10, S11, S12, S14, and S15 from cluster 2 which indicates the concentrations of trace metals in these sites may be from common anthropogenic sources. The concentrations of trace metals in S2, S3, S4, S5, S6, S9, and S13 are possibly from a mix of natural and anthropogenic sources.

4. Conclusions

This research revealed that among the targeted trace metals, Fe, Cd, and Cr exceeded the recommended values for trace metals in sediment. Though Cu, Zn, Pd, and Ni metals are still not at an alarming level; if this level of contamination continues, it might create an adverse effect on the Karnaphuli Riverine ecosystem in the near future. Most of the targeted trace metals showed the highest concentrations at S7, S8, S9, and S11 because these sites are embedded in dying and cotton industries, steel mills, tanneries, ship breaking, pulp and paper, food, fertilizer, pharmaceuticals, steel, power station, and oil refinery industries. As Karnaphuli River meets the Bay of Bengal in downstream, so, discharged trace metals from these industries cause the deterioration of marine ecosystems. Many precious maritime resources are being damaged which hampers our blue economy and impedes achieving Sustainable Development Goal (SDG) 14 of the 2030 Agenda for conservation and sustainable use of the oceans, seas, and marine resources, explicitly considering coastal areas. Hence this study recommended that the point sources of trace metals along the river should be closely monitored and the relevant regulatory bodies should impose restrictions on the discharge of trace metals and urban sewage from different sources.