Seasonal dynamics of trace metal concentrations and hydrological parameters in the Balu River, Bangladesh
Md. Khairul Amin, Maiesha Nujhat, Abu Hena Mostafa Kamal, Sheikh Md. Ruhul Amin, Wahida Khatun, Joynab Akther, Mir Mohtadi Ibtesum
Article ID: 2951
Vol 5, Issue 1, 2024
DOI:
https://doi.org/10.54517/ps.v5i1.2951
Received: 20 September 2024; Accepted: 15 October 2024; Available online: 25 November 2024;
Issue release: Vol 5 No 1
VIEWS - 3378
(Abstract)
Download PDF
Abstract
This study investigates the seasonal dynamics of trace metal concentrations and hydrological parameters in the Balu River, Bangladesh, to inform sustainable water management practices and support efforts to meet environmental standards. Water level, discharge, rainfall, and groundwater data were collected from February to August 2019, alongside sediment samples analyzed for Zn, Pb, Cu, Cr, and Cd concentrations. Statistical analyses, including correlation studies, time series modeling, and Kolmogorov-Smirnov tests, revealed significant seasonal variations. Water levels and discharge rates increased dramatically during the wet season, with upstream levels rising by 862.68% and downstream by 752.92%. Trace metal concentrations showed diverse responses: Cd increased by 554.52%, Zn by 11.86%, while Pb, Cu, and Cr decreased by 44.15%, 5.16%, and 32.72% respectively. Strong correlations were observed between certain metals, particularly in wet periods, with notable relationships between Cr-Cu (r = 0.870) and Zn-Cr (r = 0.742). The integrated analysis of hydrological parameters and metal concentrations, using Gower coefficient and partial correlation analyses, suggests complex interactions between seasonal changes and pollutant dynamics. These findings highlight the need for season-specific water management strategies and more frequent monitoring to better understand and mitigate the environmental impacts of trace metal pollution in the Balu River system. The study contributes valuable insights into the interplay between hydrological conditions and trace metal behavior, essential for developing effective pollution control measures and achieving sustainable river management in Bangladesh.
Keywords
trace metal pollution; seasonal variation; hydrological parameters; urban river system; Bangladesh water quality
Full Text
1. Introduction
Heavy metal contamination has emerged as a critical threat to water quality globally, posing significant risks to ecosystems and human health [1–3]. The proliferation of industrial activities, unsustainable agricultural practices, urban wastewater discharge, and excessive traffic contribute to the escalating levels of heavy metals in our environment [2,4–6]. This pollution not only endangers the availability of safe drinking water but also threatens aquatic life and the entire food chain through bioaccumulation [7,8]. Recognizing the severity of this issue, the United Nations’ Sustainable Development Goals have explicitly targeted the improvement of water quality by reducing pollution and minimizing the release of hazardous chemicals by 2030 [9].
In developing countries like Bangladesh, river pollution has reached alarming levels, with hundreds of factories discharging toxic effluents directly into nearby waterways. The Balu River, flowing around the Dhaka metropolitan area, has become a primary receiver of industrial wastewater [10]. Despite government efforts to relocate tanneries and implement pollution control measures, progress has been slow, and the discharge of contaminated effluents continues to pose a significant threat to the river’s ecosystem [11,12].
The assessment of heavy metal concentrations in water and sediment is crucial for evaluating the overall quality of aquatic environments [13]. Sediments, in particular, play a vital role as both habitats for benthic organisms and potential reservoirs of pollutants [14,15]. The behavior of heavy metals in aquatic systems is complex, with factors such as pH and dissolved oxygen levels influencing their adsorption and release from sediments [14]. Furthermore, the bioaccumulation of these metals in fish and other aquatic organisms presents a direct pathway for human exposure through the food chain [16,17].
The heavy metals examined in this study-zinc (Zn), lead (Pb), copper (Cu), chromium (Cr), and cadmium (Cd)-were selected due to their prevalence in industrial effluents and significant toxicological impacts. Lead is a neurotoxin particularly harmful to children, while cadmium, a known carcinogen, affects kidneys and bones [18]. Copper and zinc, though essential micronutrients, can be toxic in excess, causing organ damage and mineral imbalances [19]. Chromium, especially in its hexavalent form, is carcinogenic and can cause respiratory and dermal issues [20]. In aquatic ecosystems, these metals can bioaccumulate, leading to reduced growth, impaired reproduction, and increased mortality in organisms [21,22]. The World Health Organization’s guidelines for heavy metal concentrations in sediment (Pb: 85 mg/kg, Cd: 0.8 mg/kg, Cu: 36 mg/kg, Cr: 100 mg/kg, Zn: 50 mg/kg) underscore the importance of monitoring these metals [23]. Their presence in the Balu River is of particular concern given the area’s industrial activities, especially textile and dyeing factories [24,25].
Seasonal variations significantly influence heavy metal concentrations in water bodies. Factors such as temperature, rainfall, and runoff can alter metal concentrations throughout the year [26,27]. Understanding these seasonal dynamics is essential for comprehensive environmental monitoring and effective pollution management strategies. In the case of the Balu River, the recent relocation of tanneries and the establishment of new industrial zones necessitate ongoing assessment to evaluate the effectiveness of these measures in reducing pollution levels [28–30].
The Balu River, situated near Dhaka, receives substantial pollution from various industries, including textile and dyeing factories in Gazipur [24]. These industries discharge large volumes of wastewater containing a cocktail of chemicals, including heavy metals, oils, and hazardous pollutants [25,26]. The contamination of this aquatic system not only threatens the local ecosystem but also poses risks to human health through various exposure pathways.
To address these complex environmental challenges, it is crucial to conduct comprehensive research on sediment quality and metal contamination in river systems [6–9,31,32]. Such studies can provide valuable insights into the sources, distribution, and potential risks associated with heavy metal pollution. By employing advanced analytical techniques such as correlation and cluster analysis, researchers can better identify and classify pollutant sources, enabling more targeted and effective pollution control strategies [2,14,33]. This research aims to contribute to this body of knowledge by examining the seasonal dynamics of trace metal concentrations and hydrological parameters in the Balu River, Bangladesh, with the ultimate goal of informing sustainable water management practices and supporting efforts to achieve national and international environmental standards.
2. Materials and methods
2.1. Study area and data collection
This study focuses on the Balu River in Bangladesh, which flows for about 44 km through the marshes of Beel Belai and eastern Dhaka before entering the Shitalakshya River at Demra in the Dhaka District. The Balu River basin is characterized by a tropical monsoon climate, with distinct wet (June to October) and dry (November to May) seasons. The average annual rainfall in the region is approximately 2000 mm, with about 80% occurring during the monsoon season [34]. The area experiences hot and humid summers with temperatures reaching up to 35 °C, and mild winters with temperatures rarely dropping below 10 °C [35]. Topographically, the Balu River flows through the low-lying floodplains of the Ganges-Brahmaputra Delta, with an average elevation of 4–5 m above sea level [36]. The geology of the region is dominated by Quaternary alluvial deposits, consisting primarily of sand, silt, and clay [37]. Hydrogeologically, the area is characterized by a shallow aquifer system, with groundwater levels typically ranging from 1 to 5 m below the surface, fluctuating seasonally with rainfall and river levels [38]. The alluvial nature of the soil and shallow groundwater table make the region susceptible to both surface water pollution and groundwater contamination [39].
The Balu river’s width ranges from 100 to 200 m, with a depth of 3 to 4.5 m, which varies seasonally, particularly during monsoons [40,41]. Point sources (sewage from urban areas, industrial wastewater from concrete, tanneries, paint, and textile industries) and non-point sources (agricultural runoff) all contribute to pollution [42]. Ten sampling stations were deliberately placed over a 4.7 km section of the river based on surrounding industrial activity. Sediment samples were collected from the surface layer at these sites between February (non-monsoon) and August 2019 (monsoon), and the station names and surrounding industries are included in Table 1 and Figure 1. Water level and discharge data were collected from both upstream (U/S) and downstream (D/S) locations during 2019 by the Institute of Water Modelling (IWM). Data was obtained at three-hour intervals to capture diurnal fluctuations as shown in Figure 2. Furthermore, rainfall data was collected from the Bangladesh Meteorological Department (BMD). Daily groundwater table (GWT) data was obtained from an existing well in Gulshan Station by the Bangladesh Water Development Board (BWDB).
Figure 1. Location of the study area.
Table 1. Hydrological stations with other parameters and related information.
Data Sources Parameters with availability Station name Geographical location of the sampling stations
Latitude (N) Longitude (E)
Field Data Trace Metal Concentration in Sediments (mg/kg) (01/02/2019; 01/08/2019) St-1: Crown cement ready-mix concrete 23°48′20.6″ 90°29′5.6″
St-2: National development engineers - Balu bridge site office 23°48′30.7″ 90°29′5.6″
St-3: National development engineers ready mix concrete 23°48′36.3″ 90°29′5.3″
St-4: Jalshiri golf club 23°48′39.7″ 90°29′6.7″
St-5: Bashundhara ready-mix and construction industries limited - Plant office 23°48′42.5″ 90°29′8.5″
St-6: Bashundhara ready-mix and construction industries limited 23°48′47.4″ 90°29′6.5″
St-7: Property development limited ready-mix plant 23°48′51.8″ 90°29′6.7″
St-8: International cricket stadium – Construction site 23°50′11.5″ 90°28′40.0″
St-9: Balu bridge 23°50′14.3″ 90°28′38.4″
St-10: Biswas builders limited 23°50′20.5″ 90°28′35.2″
Table 1. (Continued).
Data Sources Parameters with availability Station name Geographical location of the sampling stations
Latitude (N) Longitude (E)
BWDB
(3 Hourly) GW Table (m) (01/01/2019–31/12/2019) Gulshan Station
(Well ID: GT2626900) 23°51′06.1″ 90°26′55.0″
IWM (3 Hourly) Discharge (m3/s) (01/01/2019–31/12/2019) Balu Inflow (Station ID: 9878) 23°50′12.6″ 90°28′12.6″
Balu Outflow (Station ID: 29700) 23°43′32.0″ 90°30′01.4″
Water Level (mMSL) (01/01/2019–31/12/2019) Balu Inflow (Station ID: 9000) 23°50′12.6″ 90°28′12.6″
Balu Outflow (Station ID: 27900) 23°43′32.0″ 90°30′01.4″
BMD (Daily) Rainfall (mm) (01/01/2019–31/12/2019) Dhaka Station 23°46′49.8″ 90°22′43.3″
Figure 2. Temporal variations in trace metal concentrations, hydrological parameters, and water quality indicators for the Balu River (2019).
2.2. Hydrological analysis
Statistical analyses were performed on the collected hydrological data using R statistical software (version 4.0.3). One-sample t-tests and Wilcoxon Signed-Rank Tests were conducted to assess the significance of water level and discharge variations. The Shapiro-Wilk test was employed to evaluate the normality of data distribution. Time series analysis was performed using the ‘forecast’ package in R, with ARIMA models fitted to the data to identify trends and seasonal patterns. Model performance was evaluated using metrics such as R2, Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE).
2.3. Trace metal sampling and analysis
Sediment samples (n = 35, approximately 100 g each) were collected from ten sites along the Balu River basin in Bangladesh during both dry and wet seasons. Sampling locations, determined via mobile GPS and illustrated in Figure 1, included the river itself and nearby farmlands irrigated with its water, particularly near industrial discharge points. Samples were extracted to a depth of 55–60 cm using horizon sampling methods [43]. A stainless-steel Ekman grab sampler, cleaned between sites to prevent cross-contamination, was used to collect the top 2–3 cm of undisturbed sediment, which typically contains the highest pollutant concentrations [44,45]. Samples were immediately frozen and transferred to the Bangladesh Fisheries Research Institute laboratory in Chandpur for analysis. There, samples were thawed, homogenized, sieved through a 2 mm mesh, oven-dried at 105 ℃, and finely ground [46]. For digestion, 2 g of each sample was heated with water and concentrated H2SO4 in a Gerhardth test tube using a Gerhadt-Kieldathern digestion chamber (150 ℃ for 0.5 h, then 250 ℃ for 2 h) [47]. The digested samples were then transferred to 50 mL volumetric flasks, and metal concentrations (Zn, Pb, Cu, Cr, Cd) were determined using Atomic Absorption Spectroscopy (AAS) (Model No.120, Graphite Tube Atomizer). The following wavelengths were used: Zn (213.9 nm), Pb (283.3 nm), Cu (324.8 nm), Cr (357.9 nm), and Cd (228.8 nm). Standard curves (0–4 ppm) were used for calibration. To ensure analytical quality, the study used NIST’s SRM 3100 series via ICP-OES with an internal laboratory-developed method as a certified reference material. The recovery rates for all metals ranged from [e.g., 95% to 105%], which is within the acceptable range. The relative standard deviation for triplicate analyses was less than 5% for all metals. Additionally, blank samples were analyzed every 10 samples to check for potential contamination during the analytical process, and no significant contamination was observed.
2.4. Correlation analysis of trace metals
Correlation analyses were performed to examine the relationships between trace metal concentrations during wet and dry periods. Three different statistical methods were employed: Spearman rank correlation, Kendall’s tau, and Pearson correlation. These analyses were conducted using the ‘cor’ function in R, with significance tested at the 0.05 level. The use of multiple correlation methods allowed for a robust assessment of relationships, accounting for potential non-linearity and non-normality in the data.
2.5. Seasonal variation analysis
To assess seasonal variations in trace metal concentrations and hydrological parameters, data were segregated into dry (February) and wet (August) periods. Two-sample t-tests and Kolmogorov-Smirnov tests were performed to compare the distributions of metal concentrations between these periods. F-tests, Levene’s tests, and Bartlett’s tests were used to assess the equality of variances. These analyses were conducted using the ‘stats’ package in R. Percentage changes in concentrations and hydrological parameters between dry and wet periods were calculated to quantify seasonal variations.
2.6. Integrated analysis
The relationships between hydrological parameters and trace metal concentrations were explored using Gower coefficient analysis and partial correlation analysis. The Gower coefficient, calculated using the ‘cluster’ package in R, provided a measure of similarity between variables. Partial correlation analysis, performed using the ‘ppcor’ package, allowed for the examination of relationships between two variables while controlling for the effects of other variables. These methods provided insights into the complex interplay between hydrological conditions and trace metal dynamics in the river system.
2.7. Quality control and quality assurance
Rigorous quality control and quality assurance procedures were implemented throughout the study. For sample collection and handling, the study followed standardized protocols and collected field blanks and duplicate samples at a 10% rate. Laboratory analyses were performed in a certified facility using standard methods, with regular use of certified reference materials, method blanks, and internal standards. Data validation included ion balance calculations, conductivity balance checks, and statistical outlier detection. Multiple correlation methods were employed to ensure robust interpretation of relationships between variables. Where possible, the findings were compared with previous regional studies for external validation. These measures ensure the reliability and reproducibility of the results.
3. Results
3.1. Hydrological analysis of the Balu River
3.1.1. Statistical analysis of water levels and discharges
Statistical analysis of upstream (U/S) and downstream (D/S) water levels and discharges in the Balu River reveals significant variations and patterns (Table 2). The one-sample t-test shows high observed values (93.893 for U/S water level, 97.534 for D/S water level) compared to the critical value of 1.961, indicating statistically significant differences from hypothesized means. The Wilcoxon Signed-Rank Test further supports this, with large V values (4,244,241 for water levels, 3,927,833 and 3,676,565.5 for discharges) and Z scores exceeding 34, all suggesting significant deviations from central tendencies. The one-sample variance test reveals substantial variability, particularly in discharges, with chi-square values far exceeding the critical value of 3062.435. The Shapiro-Wilk test results (W values close to 1 and p-values < 0.0001) indicate that the data significantly deviates from a normal distribution for all parameters.
Table 2. Statistical analysis results for U/S and D/S water levels and discharges in the Balu River.
Parameters U/S water level D/S water level U/S discharge D/S discharge
One-Sample t-Test 95% confidence interval on the mean [2.108, 2.198] [2.011, 2.093] [52.661, 57.777] [54.616, 60.316]
Difference 2.153 2.052 55.219 57.466
t (Observed value) 93.893 97.534 42.327 39.538
|t| (Critical value) 1.961
Table 2. (Continued).
Parameters U/S water level D/S water level U/S discharge D/S discharge
Wilcoxon Signed-Rank Test V 4,244,241 3,927,833 3,676,565.500
Z 46.745 39.776 34.241
One-Sample Variance Test Variance 1.531 1.289 4956.194 6151.640
95% confidence interval on the variance [1.455, 1.612] [1.225, 1.358] [4711.113, 5221.005] [5847.445, 6480.324]
Chi-square (Observed value) 4455.616 3751.998 14,427,479.555 17,907,422.621
|Chi-square| (Critical value) 3062.435
DF (Degrees of Freedom) 2911
Shapiro-Wilk Test W 0.979 0.982 0.880 0.948
p-value (Two-tailed) < 0.0001
The Wilcoxon signed-rank test results for water levels and discharges in the Balu River reveal significant differences between paired observations at both upstream and downstream locations (Figure 3). All four parameters (U/S and D/S water levels and discharges) show observed Z-scores (Zobs) that far exceed the critical Z-values (Zcrit = ±1.960) at the α/2 = 0.025 significance level. The U/S water level and D/S water level both have Zobs = 46.745, indicating identical statistical significance. The U/S discharge shows the highest Zobs of 39.776, while the D/S discharge has a Zobs of 34.241. These high Z-scores, all well above the critical value, provide strong evidence to reject the null hypothesis of no difference between paired observations. The consistency in high Z-scores across all parameters suggests that there are significant, non-random variations in both water levels and discharges throughout the river system. These results imply that the Balu River experiences substantial and statistically significant fluctuations in its hydrological characteristics, which could be attributed to factors such as seasonal changes, weather patterns, or anthropogenic influences affecting the river’s flow regime.
Figure 3. Wilcoxon signed-rank test results for upstream (U/S) and downstream (D/S) water levels and discharges in the Balu River, showing observed Z-scores and critical values.
3.1.2. Descriptive statistics and time series analysis
Descriptive statistics provide further insights into the river’s hydrological characteristics (Table 3). Mean water levels are slightly higher upstream (2.15 m) than downstream (2.05 m), while mean discharge increases downstream from 55.20 m3/s to 57.45 m3/s. The range of values is considerable, with water levels varying by 4.71 m upstream and 4.45 m downstream, and discharges showing extreme fluctuations (326.48 m3/s upstream, 401.98 m3/s downstream). Notably, minimum discharge values are negative for both U/S and D/S (−41.34 m3/s and −90.44 m3/s respectively), potentially indicating reverse flow conditions or measurement errors. Standard deviations are relatively high, especially for discharges (70.38 m3/s upstream, 78.41 m3/s downstream), underscoring the river’s dynamic nature. These results collectively paint a picture of a highly variable river system with complex flow patterns, suggesting the need for careful water management strategies and further investigation into the causes of extreme fluctuations and potential reverse flows.
Table 3. Descriptive statistics for U/S and D/S water levels and discharges in the Balu River.
Parameters U/S water level D/S water level U/S discharge D/S discharge
Mean 2.15 2.05 55.20 57.45
Maximum 4.95 4.63 285.14 311.54
Minimum 0.24 0.18 −41.34 −90.44
Difference 4.71 4.45 326.48 401.98
Median 1.99 1.92 27.17 38.34
Standard deviation 1.24 1.14 70.38 78.41
The time series analysis of water levels and discharges in the Balu River reveals distinct seasonal patterns and overall increasing trends throughout 2019 for both upstream and downstream locations (Figure 4). All four parameters exhibit a general upward trajectory, with the most pronounced increases observed in discharge rates. The graphs clearly illustrate seasonal fluctuations, with peak water levels and discharges occurring around July–August, likely corresponding to the monsoon season. The model predictions (orange lines) closely follow the observed data trends, indicating a good fit. The 95% confidence intervals (dashed lines) for both the mean (gray) and observations (yellow) provide insight into the variability and uncertainty of the predictions. Notably, the confidence intervals widen towards the end of the year, suggesting increased uncertainty in long-term projections. The upstream and downstream patterns are similar, but downstream locations show slightly higher water levels and discharges, consistent with expected river dynamics. These results suggest that the Balu River experiences significant seasonal variations in water levels and discharges, with a general increasing trend over the year, potentially influenced by factors such as monsoon patterns, land use changes, or climate variations.
Figure 4. Time series analysis of water levels and discharges in the Balu River for upstream (U/S) and downstream (D/S) locations during 2019, showing observed data, model predictions, and confidence intervals.
3.1.3. Model performance and trend analysis
The model performance metrics for U/S and D/S water levels and discharges in the Balu River reveal moderate to weak predictive capabilities (Table 4). The R2 and Adjusted R2 values are low across all parameters, ranging from 0.08 to 0.21, indicating that the models explain only a small portion of the variance in the data. Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) values are notably high, especially for discharge measurements (RMSE of 67.33 and 75.44 for U/S and D/S discharges respectively), suggesting significant deviations between predicted and observed values. The Mean Absolute Percentage Error (MAPE) is particularly high for U/S Discharge (1147.20%), indicating poor model accuracy. Durbin-Watson (DW) statistics close to 0 for water levels and discharges suggest positive autocorrelation in the residuals. The Akaike Information Criterion (AIC) and Schwarz Bayesian Criterion (SBC) values are lower for water level models compared to discharge models, indicating relatively better fit for water level predictions. The high Press statistics, especially for discharges, further confirm the models’ limited predictive power.
Table 4. Model performance metrics and additional parameters for U/S and D/S water levels and discharges in the Balu River.
Parameters U/S water level D/S water level U/S discharge D/S discharge
Observations 2913.00
Sum of Weights 2913.00
DF 2910.00
R2 0.20 0.21 0.09 0.08
Adjusted R2 0.20 0.21 0.09 0.07
MSE 1.22 1.02 4533.10 5691.01
RMSE 1.11 1.01 67.33 75.44
MAPE 59.17 55.75 1147.20 512.87
DW 0.01 0.02 0.06 0.33
Cp 2.00 2.00 2.00 2.00
AIC 584.53 46.97 24,518.60 25,181.03
Table 4. (Continued).
Parameters U/S water level D/S water level U/S discharge D/S discharge
SBC 596.48 58.92 24,530.55 25,192.98
PC 0.80 0.79 0.92 0.93
Press 3559.36 2959.42 13,206,404.69 16,580,400.05
Q2 0.20 0.21 0.08 0.07
DF: Degrees of Freedom; R2: Coefficient of Determination; MSE: Mean Squared Error; RMSE: Root Mean Squared Error; MAPE: Mean Absolute Percentage Error; DW: Durbin-Watson statistic; Cp: Mallows’ Cp; AIC: Akaike Information Criterion; SBC: Schwarz Bayesian Criterion; PC: Prediction Criterion; Press: Predicted Residual Sum of Squares; Q2: Predictive Squared Correlation.
The time series analysis of water level and discharge data for both upstream (U/S) and downstream (D/S) locations in the Balu River reveals seasonal patterns and trends throughout 2019 (Table 4, Figure 5). All four parameters (U/S and D/S water levels and discharges) show a general increasing trend from January to December, with Sen’s slopes ranging from 0.006 to 0.202 (Table 5). The most pronounced increase is observed in D/S discharge (slope = 0.202). Kendall’s Tau values (0.207 to 0.334) indicate weak to moderate positive correlations between time and the measured parameters, suggesting a consistent but not strong upward trend. The graphs clearly illustrate seasonal fluctuations, with peak water levels and discharges occurring around July–August, likely corresponding to the monsoon season. However, the p-values (0.291 to 0.453) for all parameters are greater than 0.05, indicating that these trends are not statistically significant at the 95% confidence level. This suggests that while there are observable patterns and increases in water levels and discharges over the year, these changes could be due to natural variability rather than a definitive long-term trend. The high variance (Var(S)) values further support the considerable fluctuations observed in the data throughout the year.
Figure 5. M-K analysis of water levels and discharges in the Balu River for upstream (U/S) and downstream (D/S) locations during 2019, showing Sen’s slope.
Table 5. Statistical parameters and correlation analysis for U/S and D/S water levels and discharges in the Balu River.
Parameters U/S water level D/S water level U/S discharge D/S discharge
Sen’s Slope 0.006 0.006 0.183 0.202
Kendall’s Tau 0.331 0.334 0.255 0.207
S 1,403,852 1,414,997 1,083,133 879,963
Var(S) 1,796,649,599,551 1,799,449,752,747 1,363,072,147,460 1,374,017,901,114
p-value (Two-tailed) 0.295 0.291 0.354 0.453
3.2. Trace metal analysis
3.2.1. Correlation analysis of trace metals
The Pearson correlation matrix reveals distinct patterns of trace metal associations in the Balu River during wet and dry periods (Table 6, S1–S2). In the wet period, strong positive correlations are observed between Zn and Cr (r = 0.742), as well as Cu and Cr (r = 0.870), suggesting common sources or similar environmental behaviors for these metals during rainy conditions. Conversely, the dry period shows weaker correlations overall, with the strongest being between Zn and Cr (r = 0.539). Notably, Cd exhibits weak to moderate negative correlations with most other metals in both periods, indicating its unique behavior. The p-values demonstrate that many of these correlations are statistically significant (p < 0.05), particularly in the wet period. The coefficients of determination (R2) further support these findings, with the wet period showing stronger relationships, especially for the Zn-Cr and Cu-Cr pairs. These results suggest that metal dynamics in the river are significantly influenced by seasonal factors, with stronger interrelationships during wet periods, possibly due to increased runoff and mobilization of pollutants. The distinct behavior of Cd implies it may have different sources or environmental pathways compared to the other metals studied.
Table 6. Pearson correlation matrices of trace metal concentrations for wet and dry periods.
Wet period Zn Pb Cu Cr Cd
Zn 1 0.501 0.554 0.742 −0.148
Pb 0.501 1 0.141 0.401 −0.309
Cu 0.554 0.141 1 0.870 0.176
Cr 0.742 0.401 0.870 1 −0.131
Cd −0.148 −0.309 0.176 −0.131 1
Dry Period Zn Pb Cu Cr Cd
Zn 1 0.209 0.574 0.539 0.063
Pb 0.209 1 −0.058 −0.171 −0.295
Cu 0.574 −0.058 1 0.291 0.481
Cr 0.539 −0.171 0.291 1 0.212
Cd 0.063 −0.295 0.481 0.212 1
Values in bold are different from 0 with a significance level alpha = 0.05.
Table 7. Spearman correlation matrices of trace metal concentrations for wet and dry periods.
Wet period Zn Pb Cu Cr Cd
Zn 1 0.503 0.638 0.784 −0.212
Pb 0.503 1 0.286 0.620 −0.321
Cu 0.638 0.286 1 0.802 0.067
Cr 0.784 0.620 0.802 1 −0.170
Cd −0.212 −0.321 0.067 −0.170 1
Dry Period Zn Pb Cu Cr Cd
Zn 1 0.164 0.479 0.539 −0.006
Pb 0.164 1 −0.115 −0.055 −0.321
Cu 0.479 −0.115 1 0.273 0.455
Cr 0.539 −0.055 0.273 1 −0.091
Cd −0.006 −0.321 0.455 −0.091 1
Values in bold are different from 0 with a significance level alpha = 0.05.
The Spearman correlation analysis reveals distinct patterns of trace metal associations in the Balu River across wet and dry periods (Table 7, S3–S4). During the wet period, strong positive correlations are observed between Zn and Cr (rho = 0.784), Zn and Cu (rho = 0.658), and Cu and Cr (rho = 0.802), all statistically significant (p < 0.05). This suggests common sources or similar environmental behaviors for these metals during rainy conditions. In contrast, the dry period shows generally weaker correlations, with the strongest being between Zn and Cu (rho = 0.479). Notably, Cd exhibits weak to moderate negative correlations with most other metals in the wet period but shows positive correlations in the dry period, particularly with Cu (rho = 0.455). The coefficients of determination (R2) further support these findings, with higher values in the wet period, especially for Zn-Cr and Cu-Cr pairs. These results indicate that metal dynamics in the river are significantly influenced by seasonal factors, with stronger interrelationships during wet periods, possibly due to increased runoff and mobilization of pollutants. The seasonal shift in Cd’s behavior implies it may have different sources or environmental pathways compared to the other metals, which vary with hydrological conditions.
The Kendall correlation analysis reveals distinct patterns of trace metal associations in the Balu River between wet and dry periods (Table 8, S5–S6). During the wet period, strong positive correlations are observed between Zn and Cr (τ = 0.584), Cu and Cr (τ = 0.659), and Zn and Cu (τ = 0.560), all statistically significant (p < 0.05). This suggests common sources or similar environmental behaviors for these metals during rainy conditions. In contrast, the dry period shows generally weaker correlations, with the strongest being between Zn and Cr (τ = 0.467). Notably, Cd exhibits weak negative correlations with most other metals in the wet period but shows slight positive correlations in the dry period. The coefficients of determination (R2) further support these findings, with higher values in the wet period, especially for the Zn-Cr and Cu-Cr pairs. These results indicate that metal dynamics in the river are significantly influenced by seasonal factors, with stronger interrelationships during wet periods, possibly due to increased runoff and mobilization of pollutants. The seasonal shift in Cd’s behavior implies it may have different sources or environmental pathways compared to the other metals, which vary with hydrological conditions. The Kendall correlation results generally align with the previous Spearman analysis, providing robust evidence for these seasonal metal interaction patterns.
Table 8. Kendall correlation matrices of trace metal concentrations for wet and dry periods.
Wet period Zn Pb Cu Cr Cd
Zn 1 0.378 0.360 0.584 −0.200
Pb 0.378 1 0.180 0.405 −0.200
Cu 0.360 0.180 1 0.659 0.090
Cr 0.584 0.405 0.659 1 −0.135
Cd −0.200 −0.200 0.090 −0.135 1
Dry Period Zn Pb Cu Cr Cd
Zn 1 0.111 0.378 0.467 0.022
Pb 0.111 1 −0.067 −0.156 −0.244
Cu 0.378 −0.067 1 0.289 0.378
Cr 0.467 −0.156 0.289 1 −0.067
Cd 0.022 −0.244 0.378 −0.067 1
Values in bold are different from 0 with a significance level alpha = 0.05.
3.2.2. Statistical analysis of trace metal concentrations
The statistical analysis of trace metal concentrations in the Balu River reveals significant variations between dry and wet periods for certain metals (Table 9). The T-test results show significant differences (p < 0.05) for Zn, Pb, and Cd, with Zn and Cd exhibiting particularly low p-values (p < 0.0001). The 95% confidence intervals for the differences between means do not include zero for these metals, further supporting the significance of the seasonal variations. Notably, Cd shows the largest standardized difference (t-value) of 7.101, indicating a substantial change in concentration between dry and wet periods.
Table 9. Statistical analysis of trace metal concentrations in the Balu River: T-test, Kolmogorov-Smirnov, Fisher’s F-test, Levene’s test, Bartlett test, and Mann-Whitney test results.
T-test
Zn Pb Cu Cr Cd
95% confidence interval on the difference between the means [−70.463, 168.959] [−17.278, −8.913] [−25.764, 17.687] [−58.328, −9.206] [1.039, 6.558]
Difference 49.248 −13.096 −4.039 −33.767 3.799
t (Observed value) 0.864 −6.578 −0.391 −2.888 2.892
|t| (Critical value) 2.101
DF 18
p-value (Two-tailed) 0.399 < 0.0001 0.701 0.010 0.010
Kolmogorov-Smirnov test
Zn Pb Cu Cr Cd
D 0.300 0.900 0.300 0.600 0.600
p-value (Two-tailed) 0.759 0.001 0.759 0.055 0.055
Fisher’s F-test
Zn Pb Cu Cr Cd
95% confidence interval on the ratio of variances [1.304, 21.140] [0.524, 8.501] [0.238, 3.852] [0.181, 2.927] [13.344, 216.282]
Ratio 5.251 2.111 0.957 0.727 53.72
F (Observed value) 5.251 2.111 0.957 0.727 53.72
F (Critical value) 4.026
DF1 9.000
DF2 9.000
p-value (Two-tailed) 0.021 0.281 0.949 0.643 < 0.0001
Levene’s test
F (Observed value) 4.611 1.209 0.020 0.695 17.481
F (Critical value) 4.414
DF1 1.000
DF2 18.000
p-value (Two-tailed) 0.046 0.286 0.889 0.415 0.001
Bartlett’s test
Chi-square (Observed value) 5.293 1.164 0.004 0.216 22.462
Chi-square (Critical value) 3.841
DF 1.000
p-value (Two-tailed) 0.021 0.281 0.949 0.642 < 0.0001
Table 9. (Continued).
T-test
Zn Pb Cu Cr Cd
Mann-Whitney test
U 51 2 43 17 89
U (standardized) 0.000
Expected value 50.000
Variance (U) 175.000 175.000 174.868 174.868 175.000
p-value (Two-tailed) 0.971 < 0.0001 0.616 0.011 0.002
The Kolmogorov-Smirnov test results corroborate these findings, showing significant differences in the distribution of Zn, Pb, and Cd concentrations between seasons (p < 0.05) (Table 9). The F-test for equality of variances indicates significant differences in variability for Zn and Cd (p < 0.05), suggesting not only changes in mean concentrations but also in the spread of values between dry and wet periods. Levene’s test, which is less sensitive to non-normality, confirms the variance heterogeneity for Cd (p = 0.001) (Table 9). The Bartlett’s test results align with these findings, particularly highlighting the strong heterogeneity in Cd concentrations (p < 0.0001). These comprehensive statistical analyses collectively suggest that Zn, Pb, and especially Cd concentrations in the Balu River are significantly influenced by seasonal factors, potentially due to variations in runoff, precipitation patterns, or seasonal anthropogenic activities affecting the river system.
The univariate clustering analysis of trace metals in the Balu River reveals distinct patterns between dry and wet seasons (Tables 10 and 11). During the dry season, zinc (Zn) shows the highest concentration (mean: 398.0 µg/L), followed by chromium (Cr) and copper (Cu), while cadmium (Cd) has the lowest concentration (mean: 1.0 µg/L). The wet season exhibits a similar order but with generally lower concentrations, particularly for Zn (mean: 304.0 µg/L). Notably, the clustering results indicate higher variability in metal concentrations during the dry season, as evidenced by larger ranges and standard deviations. For instance, Zn’s range in the dry season (252.5–535.5 µg/L) is wider than in the wet season (235.0–373.0 µg/L). The clustering also reveals potential outliers, particularly for Cr and Cu in both seasons. These results suggest significant seasonal variations in trace metal concentrations, with the dry season characterized by higher concentrations and greater variability. This pattern could be attributed to factors such as reduced dilution during low flow periods, increased anthropogenic inputs, or changes in biogeochemical processes between seasons. The clustering analysis provides valuable insights for understanding metal pollution dynamics in the Balu River and can inform targeted environmental management strategies.
Table 10. Univariate clustering results for trace metals in the Balu River during dry season.
Dry season Zn Pb Cu Cr Cd
Class 1 2 3 1 2 3 4 1 2 3 4 1 2 3 1 2 3 4 5
Objects 6 3 1 4 3 2 1 5 1 2 2 5 4 1 1 1 6 1 1
Sum of weights 6 3 1 4 3 2 1 5 1 2 2 5 4 1 1 1 6 1 1
Within-class variance 536.2 456.6 0 1.5 0.3 0.4 0 24.4 0 0.2 194.9 26.4 122.4 0 0 0 0 0 0
Minimum distance to centroid 0.2 6.2 0 0.4 0.1 0.4 0 2 0 0.3 9.9 0.3 4.7 0 0 0 0.1 0 0
Average distance to centroid 17.9 15.8 0 1 0.4 0.4 0 3.9 0 0.3 9.9 3.4 9 0 0 0 0.1 0 0
Maximum distance to centroid 35.2 23.8 0 1.5 0.6 0.4 0 7.7 0 0.3 9.9 7.6 13.3 0 0 0 0.2 0 0
St1 St4 St7 St1 St2 St3 St5 St1 St3 St4 St6 St1 St3 St4 St1 St2 St3 St4 St7
St2 St5 St4 St9 St7 St2 St7 St10 St2 St6 St5
St3 St10 St6 St10 St5 St5 St7 St6
St6 St8 St8 St8 St10 St8
St8 St9 St9 St 9
St9 St10
Table 11. Univariate clustering results for trace metals in the Balu River during wet season.
Wet Season Zn Pb Cu Cr Cd
Class 1 2 3 4 1 2 3 4 5 1 2 3 4 1 2 3 4 5 1 2 3 4
Objects 5 3 1 1 6 1 1 1 1 5 1 2 2 4 1 2 2 1 1 2 6 1
Sum of weights 5 3 1 1 6 1 1 1 1 5 1 2 2 4 1 2 2 1 1 2 6 1
Within-class variance 823.9 1172.3 0 0 1.7 0 0 0 0 75 0 5.1 6.8 26.5 0 4.6 70 0 0 0.9 1.1 0
Minimum distance to centroid 3.3 8.3 0 0 0.3 0 0 0 0 0.1 0 1.6 1.8 1 0 1.5 5.9 0 0 0.7 0.1 0
Average distance to centroid 19.8 25.1 0 0 0.9 0 0 0 0 6 0 1.6 1.8 3.6 0 1.5 5.9 0 0 0.7 0.8 0
Maximum distance to centroid 49.5 37.6 0 0 2 0 0 0 0 15 0 1.6 1.8 7.2 0 1.5 5.9 0 0 0.7 1.4 0
St1 St4 St7 St10 St1 St2 St5 St9 St10 St1 St4 St5 St7 St1 St4 St5 St7 St10 St1 St2 St3 St4
St2 St5 St3 St2 St10 St9 St2 St8 St9 St7 St5
St3 St9 St4 St3 St3 St6
St6 St6 St6 St6 St8
St8 St7 St8 St9
St8 St10
The cumulative distribution plots in Figure 6 compare trace metal concentrations (Zn, Pb, Cu, Cr, and Cd) between dry and wet periods in the Balu River using the Two-sample Kolmogorov-Smirnov test/Two-tailed test. For zinc (Zn), lead (Pb), and copper (Cu), the distributions show some separation between dry and wet periods, indicating potential seasonal variations in concentrations. Chromium (Cr) exhibits a more pronounced difference, with the dry period curve generally higher than the wet period, suggesting higher Cr levels during dry conditions. Cadmium (Cd) shows the most striking contrast, with the wet period curve rising much more steeply than the dry period, implying significantly higher Cd concentrations during wet periods. These results suggest that trace metal concentrations in the Balu River vary seasonally, with different metals exhibiting distinct patterns between dry and wet periods. These variations may be attributed to factors such as changes in river flow, runoff patterns, or seasonal differences in anthropogenic activities affecting the river.
Figure 6. Kolmogorov-Smirnov test between dry and wet period trace metals concentrations in Balu River.
The correlation analyses of trace metal concentrations during wet and dry periods revealed distinct patterns across three different statistical methods: Spearman, Kendall, and Pearson correlations (Tables 12–14, S7–S12). During the wet period, strong positive correlations (p < 0.05) were consistently observed between Cr and Cu (Spearman: 0.802, Kendall: 0.659, Pearson: 0.870), as well as between Cr and Zn (Spearman: 0.784, Kendall: 0.584, Pearson: 0.742). Additionally, Zn showed a significant positive correlation with Cu across all methods. Pb exhibited moderate positive correlations with Zn and Cr in the Spearman analysis, but these relationships were less pronounced in the Kendall and Pearson correlations.
Table 12. Pearson correlation matrix of hydrological parameters in the Balu River.
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
Ground water 1 −0.138 0.279 0.287 0.416 0.426
Rainfall −0.138 1 0.000 0.001 −0.016 −0.024
U/S discharge 0.279 0.000 1 0.999 0.933 0.915
Table 12. (Continued).
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
D/S discharge 0.287 0.001 0.999 1 0.936 0.917
U/S water level 0.416 −0.016 0.933 0.936 1 0.998
D/S water level 0.426 −0.024 0.915 0.917 0.998 1
Values in bold are different from 0 with a significance level alpha = 0.05.
Table 13. Spearman correlation matrix of hydrological parameters in the Balu River.
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
Ground water 1 −0.118 0.137 0.137 0.165 0.164
Rainfall −0.118 1 −0.046 −0.045 −0.073 −0.077
U/S discharge 0.137 −0.046 1 0.997 0.975 0.964
D/S discharge 0.137 −0.045 0.997 1 0.974 0.963
U/S water level 0.165 −0.073 0.975 0.974 1 0.997
D/S water level 0.164 −0.077 0.964 0.963 0.997 1
Values in bold are different from 0 with a significance level alpha = 0.05.
Table 14. Kendall correlation matrix of hydrological parameters in the Balu River.
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
Ground water 1 −0.072 0.040 0.038 0.076 0.077
Rainfall −0.072 1 −0.032 −0.033 −0.049 −0.050
U/S discharge 0.040 −0.032 1 0.964 0.867 0.837
D/S discharge 0.038 −0.033 0.964 1 0.861 0.831
U/S water level 0.076 −0.049 0.867 0.861 1 0.969
D/S water level 0.077 −0.050 0.837 0.831 0.969 1
Values in bold are different from 0 with a significance level alpha = 0.05.
In contrast, the dry period demonstrated weaker and fewer significant correlations among the trace metals. The most notable relationship during the dry period was between Cu and Cd, which showed a moderate positive correlation in the Spearman (0.455) and Pearson (0.481) analyses, though this was not reflected in the Kendall correlation. Zn maintained weak to moderate positive correlations with Cu and Cr across all methods during the dry period. Interestingly, Pb showed weak negative correlations with Cu and Cr in the dry period, a pattern not observed in the wet period. These results suggest that the relationships between trace metal concentrations are more pronounced and generally more positively correlated during wet periods compared to dry periods, indicating potential differences in metal behavior or sources between these two environmental conditions.
The Gower coefficient and partial correlation analyses of hydrological parameters in the Balu River reveal both similarities and differences in their relationships (Tables 15 and 16). The Gower coefficient shows strong correlations between U/S and D/S discharge (0.964), as well as between U/S and D/S water levels (0.999), indicating high similarity in these paired measurements. Interestingly, rainfall exhibits strong correlations with water levels (0.855 and 0.854) in the Gower analysis. However, the partial correlation analysis presents a different picture, highlighting the relationships between variables when controlling for others. It shows a very strong correlation between U/S and D/S discharge (0.980) and between U/S and D/S water levels (0.998), consistent with the Gower results. Yet, it reveals weak or negative correlations for most other parameter pairs, suggesting that many apparent relationships in the Gower analysis may be indirect. Notably, the strong rainfall-water level relationship observed in the Gower analysis disappears in the partial correlation, implying that this connection may be mediated by other factors. These results underscore the complex interplay of hydrological parameters in the Balu River, where direct relationships between discharge measurements and water levels are robust, but other parameter interactions are more nuanced and potentially influenced by confounding variables.
Table 15. Gower coefficient correlation matrix of hydrological parameters in the Balu River.
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
Ground water 1 0.339 0.496 0.491 0.200 0.199
Rainfall 0.339 1 0.408 0.379 0.855 0.854
U/S discharge 0.496 0.408 1 0.964 0.347 0.345
D/S discharge 0.491 0.379 0.964 1 0.318 0.317
U/S water level 0.200 0.855 0.347 0.318 1 0.999
D/S water level 0.199 0.854 0.345 0.317 0.999 1
Values in bold are different from 0 with a significance level alpha = 0.05.
Table 16. Partial correlation coefficient matrix of hydrological parameters in the Balu River.
Variables Ground water Rainfall U/S discharge D/S discharge U/S water level D/S water level
Ground water 1 −0.168 −0.149 0.089 0.144 −0.116
Rainfall −0.168 1 −0.019 −0.019 0.205 −0.205
U/S discharge −0.149 −0.019 1 0.980 −0.256 0.253
D/S discharge 0.089 −0.019 0.980 1 0.426 −0.418
U/S water level 0.144 0.205 −0.256 0.426 1 0.998
D/S water level −0.116 −0.205 0.253 −0.418 0.998 1
Values in bold are different from 0 with a significance level alpha = 0.05.
3.3. Integrated analysis of hydrological parameters and trace metals
3.3.1. Correlation analysis of hydrological parameters
The probability plots for Ground Water Table (GWT) and Rainfall in the Balu River basin reveal distinct patterns in data distribution and their deviation from theoretical normal distribution (Figure 7). For the GWT, the observed data (pink line) closely follows the theoretical normal distribution (dashed line) between the 10th and 90th percentiles, indicating a generally normal distribution within this range. However, deviations are noticeable at the extremes, particularly below the 10th percentile (79.8 units) and above the 90th percentile (83.7 units), suggesting some skewness in the tails of the distribution. The rainfall data shows a more pronounced deviation from normality, with a clear S-shaped curve indicating non-normal distribution. This curve suggests that rainfall data is likely skewed, with a higher frequency of lower rainfall amounts and fewer extreme high rainfall events. The stark difference between GWT and rainfall distributions implies that while groundwater levels maintain a relatively stable, near-normal distribution, rainfall patterns are more variable and potentially influenced by extreme events. These findings have important implications for water resource management in the basin, indicating the need for different approaches in modeling and predicting groundwater levels versus rainfall patterns.
Figure 7. Probability plots for Ground Water Table and Rainfall in the Balu River, showing observed data distribution and theoretical normal distribution.
3.3.2. Seasonal variations in trace metals and hydrological parameters
The analysis of trace metal concentrations and hydrological parameters in the Balu River revealed significant variations between the dry (February) and wet (August) periods of 2019 (Table 17). Most notably, zinc (Zn) and cadmium (Cd) concentrations increased by 11.86% and 554.52% respectively, while lead (Pb), copper (Cu), and chromium (Cr) decreased by 44.15%, 5.16%, and 32.72%. These changes coincided with substantial increases in hydrological parameters, particularly upstream and downstream water levels (862.68% and 752.92% increase) and discharge rates. Groundwater levels showed a modest 3.22% increase. The dramatic rise in discharge and water levels, coupled with the occurrence of rainfall in the wet period, indicates a major shift in hydrological conditions. The contrasting behavior of metal concentrations suggests complex interactions between hydrological factors and metal mobilization or dilution processes. While the dilution effect is evident for Pb, Cu, and Cr, the significant increases in Zn and especially Cd concentrations warrant further investigation into potential sources or mobilization mechanisms during wet periods. These results highlight the need for more frequent sampling to better understand the seasonal dynamics of metal concentrations in relation to hydrological changes in the Balu River.
Table 17. Comparison of trace metal concentrations (mg/kg) and hydrological parameters in the Balu River between dry and wet periods (2019).
Parameters Dry period (01 February 2019) Wet period (01 August 2019) Difference
Zinc (Zn) 415.36 464.61 11.86% increase
Lead (Pb) 29.66 16.57 44.15% decrease
Copper (Cu) 78.32 74.28 5.16% decrease
Chromium (Cr) 103.19 69.43 32.72% decrease
Cadmium (Cd) 0.69 4.49 554.52% increase
Ground water (m) 80.11 82.69 3.22% increase
Rainfall (mm) 0.00 1.00 Rainfall Occurred
U/S discharge (m3/s) 0.62 261.08 Increased
D/S discharge (m3/s) −1.65 269.02 Increased
U/S water level (mMSL) 0.49 4.72 862.68% increase
D/S water level (mMSL) 0.50 4.28 752.92% increase
4. Discussion
The present study reveals significant seasonal variations in trace metal concentrations and hydrological parameters in the Balu River, Bangladesh. The findings indicate complex interactions between metal pollution and hydrological factors, which are crucial for understanding the river’s ecosystem health and managing water resources effectively.
The observed increase in zinc (Zn) and cadmium (Cd) concentrations during the wet season, coupled with decreases in lead (Pb), copper (Cu), and chromium (Cr), suggests a nuanced relationship between metal mobilization and hydrological changes. This pattern aligns with the findings of Banu et al. [38] who reported higher metal concentrations in the dry season for most metals, except for specific cases. The dramatic increase in Cd concentration (554.52%) during the wet season is particularly concerning and warrants further investigation. This surge could be attributed to increased runoff from agricultural areas or industrial effluents, as suggested by Hossain et al. [48], who found elevated Cd levels in fish from the Turag-Tongi-Balu river system.
The substantial increases in water levels and discharge rates during the wet season (over 750% for both parameters) highlight the river’s dynamic hydrology. These findings are consistent with the observations of Atiq et al. [49], who noted that pollutant concentrations in the Balu River are generally higher during the non-monsoon period due to lower water flow. However, the results show that this pattern doesn’t hold true for all metals, particularly Zn and Cd. This discrepancy underscores the complex nature of metal behavior in aquatic systems and suggests that factors beyond simple dilution effects are at play.
The correlation analyses between trace metals reveal stronger relationships during the wet period, especially for Cr-Cu, Cr-Zn, and Zn-Cu pairs. These findings support the observations of Islam et al. [50], who noted that heavy metal contamination varies according to the nature of the sites and the intensity of industrial and agricultural runoff. The weaker correlations observed during the dry period suggest that point sources of pollution may have a more dominant influence when river flow is reduced. This aligns with the conclusions of Hasan et al. [51], who found that water pollution in the Balu River is generally more severe during the dry season due to reduced dilution.
The strong positive correlations observed between Cr-Cu, Cr-Zn, and Zn-Cu pairs during the wet season provide valuable insights into the dynamic behavior of these metals in the Balu River ecosystem. The consistent correlation across Spearman, Kendall, and Pearson analyses suggests a robust relationship that may be attributed to similar sources or shared geochemical behavior. For instance, the strong Cr-Cu correlation (Spearman: 0.802, Pearson: 0.870) could indicate a common industrial source, such as tanneries or metal plating facilities, which are known to release both metals [52,53]. The Cr-Zn and Zn-Cu correlations might reflect the influence of urban runoff, as these metals are often associated with vehicular emissions and wear of galvanized structures [54,55]. The weaker correlations observed during the dry season suggest that the behavior and sources of these metals may be altered under different hydrological conditions. This seasonal variation in metal relationships could be explained by changes in pH, redox conditions, or the relative contributions of point and non-point pollution sources [56,57]. The stronger correlations during the wet season may indicate enhanced mobilization and co-transport of these metals during high-flow periods, possibly due to resuspension of contaminated sediments or increased industrial discharge [58]. These findings highlight the need for season-specific pollution management strategies and underscore the importance of considering metal interactions in environmental risk assessments.
The integrated analysis of hydrological parameters and trace metals provides valuable insights into the river’s pollution dynamics. The strong correlations between upstream and downstream discharge and water levels, as revealed by both Gower coefficient and partial correlation analyses, indicate a well-connected river system. However, the weak or negative correlations between rainfall and other parameters in the partial correlation analysis suggest that the relationship between precipitation and river pollution is not straightforward. This complexity is reflected in the findings of Roy et al. [59], who recommended maintaining appropriate distances for waste dumping sites and strict implementation of effluent treatment regulations to mitigate pollution.
The results highlight the urgent need for comprehensive water quality management strategies in the Balu River basin. The seasonal variations in metal concentrations and their relationships with hydrological parameters underscore the importance of tailored approaches for different seasons. As Sami and Sikder [60] emphasized, reviving the Balu River is not just a matter of healing the waterway but also of rejuvenating Dhaka itself. Implementing robust conservation strategies, as suggested by Bhuiyan et al. [61], such as stricter industrial discharge regulations, effective effluent treatment, and control of both point and non-point pollution sources, is crucial for improving the river’s health.
Future research should focus on more frequent sampling to capture the fine-scale dynamics of metal concentrations in relation to hydrological changes. Additionally, investigating the sources and pathways of Cd pollution, given its alarming increase during the wet season, should be a priority. The potential for microplastic pollution, as highlighted by Odora et al. [40], adds another dimension to the complexity of the Balu River’s environmental challenges. Integrating these various aspects of pollution and hydrology will be essential for developing a holistic understanding of the river ecosystem and implementing effective restoration measures. Authors should discuss the results and how they can be interpreted from the perspective of previous studies and of the working hypotheses. The findings and their implications should be discussed in the broadest context possible. Future research directions may also be highlighted.
5. Conclusion
This study provides comprehensive insights into the seasonal dynamics of trace metal concentrations and hydrological parameters in the Balu River, Bangladesh. The findings reveal significant seasonal variations in metal pollution, with zinc and cadmium showing marked increases during the wet season, while lead, copper, and chromium concentrations decreased. These changes coincided with substantial increases in water levels and discharge rates, highlighting the complex interplay between hydrological factors and metal mobilization. The dramatic increase in cadmium concentrations (554.52%) during the wet season is particularly concerning due to its potential toxic effects on aquatic life and human health.
The exceptionally high increase in cadmium and the contrasting behavior of different metals during wet and dry seasons highlight the complex dynamics of pollution in this urban river system. These findings go beyond general expectations and provide crucial, site-specific data for targeted environmental management. The results demonstrate that even in a monsoon-affected region, the patterns and magnitudes of seasonal metal pollution can vary significantly and unpredictably, necessitating continuous, detailed monitoring and adaptive management strategies.
This research has significant impacts at multiple levels. Globally, it contributes to understanding seasonal dynamics of trace metal pollution in monsoon-affected urban river systems. Nationally, it informs environmental policies and water management strategies across Bangladesh, particularly highlighting the critical issue of cadmium pollution. Locally, our findings have immediate implications for the management of the Balu River basin and surrounding urban areas of Dhaka, providing a scientific basis for targeted remediation efforts, urban planning, and public health initiatives.
These results underscore the need for season-specific water quality management strategies and more rigorous monitoring of industrial effluents and agricultural runoff. The findings support the implementation of stricter regulations on industrial discharges, improved effluent treatment practices, and better control of both point and non-point pollution sources.
Future research should focus on more frequent sampling to capture fine-scale dynamics of metal concentrations, investigation of cadmium pollution sources, and integration of microplastic pollution studies. Ultimately, the revitalization of the Balu River is crucial not only for the health of the aquatic ecosystem but also for the sustainable development of Dhaka and its surrounding areas. This study provides a foundation for evidence-based policy-making and conservation efforts aimed at improving the ecological status of this vital water resource, taking into account the complex seasonal dynamics of pollutant behavior in monsoon-affected urban river systems.
Author contributions: Conceptualization, MKA and MN; methodology, MKA; software, MKA; validation, MKA, MN and AHMK; formal analysis, MKA; investigation, MKA; resources, MKA; data curation, MN; writing—original draft preparation, MKA; writing—review and editing, MKA and MN; visualization, MMI; supervision, SMRA, WK and JA; project administration, SMRA, WK and JA; funding acquisition, AHMK. All authors have read and agreed to the published version of the manuscript.
Conflict of interest: The authors declare no conflict of interest.
References
1. Venkata Mohan S, Nithila P, Jayarama RS. Estimation of heavy metals in drinking water and development of heavy metal pollution index. J Environ Sci Heal Part A Toxic/Hazardous Subst Environ Eng. 1996; 31(2): 283–289. doi: 10.1080/10934529609376357
2. Nasehi F, Monavari M, Naderi Gh, et al. Investigation of heavy metals accumulation in the sediment and body of carp fish in Aras River. Iran J Fish Sci. 2013; 12(2): 398–410.
3. Islam MS, Ahmed MK, Raknuzzaman M, et al. Heavy metal pollution in surface water and sediment: A preliminary assessment of an urban River in a developing country. Ecol Indic. 2015; 48: 282–291.
4. Mondol MN, Chamon AS, Faiz B, Elahi SF. Seasonal variation of heavy metal concentration in water and plant samples around Tejgaon industrial area of Bangladesh. J Bangladesh Acad Sci. 2011; 35(1): 19–41.
5. Kihampa C. Heavy metal contamination in water and sediment downstream of municipal wastewater treatment plants, Dar es Salaam, Tanzania. Int J Environ Sci. 2013; 3(5): 1407–1415.
6. Zhang F, Yan X, Zeng C, et al. Influence of traffic activity on heavy metal concentrations of roadside farmland soil in mountainous areas. Int J Environ Res Public Health. 2012; 9(5): 1715–1731.
7. Ashraf W. Levels of selected heavy metals in tuna fish. Arab J Sci Eng. 2006; 31(1A): 89–92.
8. Terra BF, Araújo FG, Calza CF, et al. Heavy metal in tissues of three fish species from different trophic levels in a tropical Brazilian River. Water Air Soil Pollut. 2008; 187(1–4): 275–284.
9. United Nations. Goal 6: Clean Water and Sanitation in The Sustainable Development Goals Report. 2017; 60. Available online: https://www.unep.org/topics/sustainable-development-goals/why-do-sustainable-development-goals-matter/goal-6-clean-2 (accessed on 10 September 2024)
10. Paul R, Haq A. Challenges of water quality management: case of peripheral Rivers in Dhaka Mega City. In: Proceedings of the presentation from the world water week in Stockholm; 5–11 September 2010; Stockholm, Sweden.
11. A shocking further delay in tannery relocation. Available online: https://www.newagebd.net/article/94630/a-shocking-further-delay-in-tanneryrelocation?fbclid = IwAR2IuKp1Xm1steI4IM YRmWkqt1Sv3f0dGJOZjtUbvfcflrye8RGXVo6WVJk (accessed on 10 September 2024).
12. dark-flows-the-River-turag-1587946@www.thedailystar.net. Available online: https://www.thedailystar.net/star-weekend/environment/dark-flows-the-River-turag-1587946?amp (accessed on 10 September 2024).
13. Lee CSL, Li XD, Zhang G, et al. Heavy metals and Pb isotopic composition of aerosols in urban and suburban areas of Hong Kong and Guangzhou, South China Evidence of the long-range transport of air contaminants. Atmos Environ, 2007; 41(2): 432–447.
14. Li H, Shi A, Li M, Zhang X. Effect of pH, temperature, dissolved oxygen, and flow rate of overlying water on heavy metals release from storm sewer sediments. J Chem. 2013.
15. Duncan AE, de Vries N, Nyarko KB. Assessment of heavy metal pollution in the sediments of the River Pra and its tributaries. Water Air Soil Pollut. 2018; 229(8): 272.
16. Van-Duijn JRC. Diseases of fishes. In: The Lancet. Narendra Publishing House; 1828. Volume 9. p 651.
17. Jatav SK, Ravikant, Dhruve D, et al. Heavy Metal Toxicity in Fishes and Their Impact on Human’s Health: A Review. The Pharma Innovation Journal, vol. SP-12, no. 10, 2023, pp. 1448–1452.
18. Politis MD, Yao M, Gennings C, et al. Associations of Exposure to Urinary Metals with Kidney Injury Biomarkers among Pre-adolescent Children: A Cross-Sectional Study. In: Proceedings of the 35th Annual Conference of the International Society of Environmental Epidemiology (ISEE 2023); 17–21 September 2023; Taiwan, China.
19. Kitala K, Tanski D, Godlewski J, et al. Copper and Zinc Particles as Regulators of Cardiovascular System Function—A Review. Nutrients. 2023; 15(13): 3040. doi: 10.3390/nu15133040
20. Layek M, Khatun N, Karmakar P, et al. Toxicity of Hexavalent Chromium: Review. In: Kumar N, Walther C, Gupta DK. (editors). Chromium in Plants and Environment. Springer; 2023.
21. Naz S, Chatha AMM, Téllez-Isaías G, et al. A comprehensive review on metallic trace elements toxicity in fishes and potential remedial measures. Water. 2023; 15(16): 3017. doi: 10.3390/w15163017
22. Emon FJ, Rohani MF, Sumaiya N, et al. Bioaccumulation and bioremediation of heavy metals in fishes—A review. Toxics. 2023; 11(6): 510. doi: 10.3390/toxics11060510
23. WHO. World Health Organization. Available online: https://iris.who.int/handle/10665/38551 (accessed on 10 September 2024).
24. Saha PK, Hossain M. Geochemical and ecotoxical approach for evaluation of heavy metal pollution in the Buriganga River sediment. In: Proceedings of the Bangladesh Geotechnical Conference; 03–06 November 2010; Dhaka, Bangladesh.
25. Islam MS, Chowdhury MAH, Billah MMS, et al. Investigation of effluent quality discharged from the textile industry of Purbani Group, Gazipur, Bangladesh and it’s textile industry of Purbani Group, Gazipur, Bangladesh and it’s management. Bangladesh J Environ Sci. 2012; 23: 123–130.
26. Yahaya M, Mohammad S, Abdullahi B. Seasonal variations of heavy metals concentration in abattoir dumping site soil in Nigeria. J Appl Sci Environ Manag. 2010; 13(4): 9–13.
27. Mohiuddin KM, Otomo K, Ogawa Y, Shikazono N. Seasonal and spatial distribution of trace elements in the water and sediments of the Tsurumi River in Japan. Environ Monit Assess. 2012; 184(1): 265–279.
28. Close tanneries at Hazaribagh_The Daily Star. Available online: https://www.thedailystar.net/frontpage/close-tanneries-hazaribagh-1372036 (accessed on 10 September 2024).
29. Buriganga River. Available online: https://en.wikipedia.org/wiki/Buriganga_River (accessed on 10 September 2024).
30. Savar tannery ETP stuck by a string_The Business Standard. Available online: https://www.tbsnews.net/economy/savar-cetp-stuck-98-two-years (accessed on 10 September 2024)
31. Islam MS, Ahmed MK, Habibullah-Al-Mamun M, Hoque MF. Preliminary assessment of heavy metal contamination in surface sediments from a River in Bangladesh. Environ Earth Sci. 2015; 73(4): 1837–1848.
32. Whitehead PG, Bussi G, Peters R, et al. Modelling heavy metals in the Buriganga River System, Dhaka, Bangladesh: impacts of tannery pollution control. Sci Total Environ. 2019; 697: 134090.
33. Nürnberg HW. The voltammetric approach in trace metal chemistry of natural waters and atmospheric precipitation. Anal Chim Acta. 1984; 164(C): 1–21.
34. Abdullah F, Mamoon WB, Islam AS, et al. Hydrological responses of the Brahmaputra river basin using CMIP6 GCM projections for supporting climate resilient infrastructure design. Sustainable and Resilient Infrastructure. 2024; 1–27. doi: 10.1080/23789689.2024.2371741
35. Ramabrahmam K, and Venkatareddy K. Trend Analysis of Climate Variables and Extremes Over Maner River Basin, India. In: International Virtual Conference on Developments and Applications of Geomatics. Springer Nature Singapore; 2022. pp. 509–517.
36. Singhal P, Goel MK, Bronstert A, et al. Understanding the Flood Seasonality in the Himalayan River Basin under Changing Climate. 2024; Preprint (Version 1). doi: 10.21203/rs.3.rs-4550138/v1
37. Salviya LL, Taryana D, Masruroh H, Aristin NF. Estimation of erosion using the modified universal soil loss equation (musle) method at SUB das bale. Journal of Innovative Integration and Harmony of Social Sciences. 2023; 3(3): 313–334. doi: 10.17977/um063v3i3p313–334
38. Banu M, Jahan B, Mondal MN, Hossain MA. Toxicity Effects of Metals Bioaccumulation in Water and Fishes of the Balu River, Bangladesh. Egyptian Journal of Aquatic Biology & Fisheries. 2023; 27(3). doi: 10.21608/ejabf.2023.301087
39. Rahman MM, Islam MA, Khan MB. Status of heavy metal pollution of water and fishes in Balu and Brahmaputra rivers. Progressive Agriculture. 2016; 27(4): 444–452. doi: 10.3329/pa.v27i4.32129
40. Odora AT, Aysha S, Sultan MB, Bhuiyan MAR. Evaluating the sources of microplastic contamination and quantifying its abundance in the Balu River, Dhaka, Bangladesh. Environmental Monitoring and Assessment. 2024; 196(9): 867. doi: 10.1007/s10661-024-13051-w
41. Islam MS, Karim MR, Islam MT, et al. Abundance, characteristics, and ecological risks of microplastics in the riverbed sediments around Dhaka city. Science of The Total Environment. 2023a; 877: 162866. doi: 10.1016/j.scitotenv.2023.162866
42. Hasan MK, Happy MA, Nesha MK, Karim KR. Pollution status of Balu river due to industrial input at Dhaka, Bangladesh. Open Journal of water pollution and treatment. 2014; 1(1): 34–42. doi: 10.15764/wpt.2014.01004
43. Soil Survey Staff. Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting Soil Surveys, 2nd ed. In: Agriculture Handbook, No. 436, Natural Resources Conservation Service, United States Department of Agriculture. U.S. Government Printing Office Washington; 1999. p. 886.
44. Richardson S, Mayes WM, Faetsch S, et al. Mixtures of sediment chemical contaminants at freshwater sampling sites across Europe with different contaminant burdens. Science of The Total Environment. 2023; 891: 164280. doi: 10.1016/j.scitotenv.2023.164280
45. Risacher FF, Schneider H, Drygiannaki I, et al. A review of peeper passive sampling approaches to measure the availability of inorganics in sediment porewater. Environmental Pollution. 2023; 328: 121581. doi: 10.1016/j.envpol.2023.121581
46. Quirino DF, Lima NS, Palma MN, et al. Evaluation of Heating Times for Loss on Drying at 105℃ for Estimation of Laboratory Dry Matter in Animal Feeds. Journal of AOAC International. 2023; 106(2): 261–266. doi: 10.1093/jaoacint/qsad004
47. Binder GA, Metcalf R, Atlas Z, Daniel KG. Evaluation of digestion methods for analysis of trace metals in mammalian tissues and NIST 1577c. Analytical biochemistry. 2018; 543: 37–42. doi: 10.1016/j.ab.2017.11.016
48. Hossain MM, Jahan I, Al Nahian A, et al. Immediate health risk: Concentration of heavy metals in contaminated freshwater fishes from the river channel of Turag‐Tongi‐Balu. Environmental Toxicology. 2024; 39(1): 120–34. doi: 10.1002/tox.23959
49. Atiq AR, Rahaman MM, Sojib RM, Shil SC. Water quality changes in Balu River during non-monsoon and monsoon period. Journal of Water Resource Engineering and Management. 2015; 2: 38–49.
50. Islam MZ, Noori A, Islam R, et al. Assessment of the contamination of trace metal in Balu River water, Bangladesh. Journal of Environmental Chemistry and Ecotoxicology. 2012; 4(14): 242–249. doi: 10.5897/JECE10.025
51. Hasan MK, Khan MRI, Nesha MK, Happy MA. Analysis of water quality using chemical parameters and metal status of Balu River at Dhaka, Bangladesh. Open J. Water Pollut. Treat. 2014; 1(2): 58–74. doi: 10.15764/WPT.2014.02007
52. Oruko RO, Edokpayi JN, Msagati TA, et al. Investigating the chromium status, heavy metal contamination, and ecological risk assessment via tannery waste disposal in sub-Saharan Africa (Kenya and South Africa). Environmental Science and Pollution Research. 2021; 28: 42135–42149. doi: 10.1007/s11356-021-13703-1
53. Murtaza G, Usman M. Assessment of various heavy metals level in groundwater and soil at tannery manufacturing areas of three mega cities (Sialkot, Lahore and Karachi) of Pakistan. Desalination and Water Treatment. 2022; 266: 121–130. doi: 10.5004/dwt.2022.28640
54. Lerat-Hardy A, Coynel A, Schäfer J, et al. Impacts of highway runoff on metal contamination including rare earth elements in a small urban watershed: case study of Bordeaux Metropole (SW France). Archives of Environmental Contamination and Toxicology. 2021; 1–21. doi: 10.1007/s00244-021-00816-4
55. Charters FJ, O’Sullivan AD, Cochrane TA. Influences of zinc loads in urban catchment runoff: Roof type, land use type, climate and management strategies. Journal of Environmental Management. 2022; 322: 116076. doi: 10.1016/j.jenvman.2022.116076
56. Gómez-Alvarez A, Meza-Figueroa D, Valenzuela-García JL, et al. Behavior of Metals Under Different Seasonal Conditions: Effects on the Quality of a Mexico–USA Border River. Water, Air, & Soil Pollution. 2014; 225: 1–13. doi: 10.1007/s11270-014-2138-z
57. O’Connor AE, Canuel EA, Beck AJ. Drivers and seasonal variability of redox-sensitive metal chemistry in a shallow subterranean estuary. Frontiers in Environmental Science. 2022; 9: 613191. doi: 10.3389/fenvs.2021.613191
58. da Silva EP, Simões TR, Antoniosi Filho NR, et al. Effect of rainfall characteristics on the transport of trace metals in suspended particles during rainfall events. Journal of Hydrology. 2024; 634: 131062. doi: 10.1016/j.jhydrol.2024.131062
59. Roy S, Banna LN, Hossain M, Rahman H. Water quality of Narai canal and Balu river of Dhaka City: An impact of industrialization. Journal of the Bangladesh Agricultural University. 2014; 12(2): 285–290. doi: 10.22004/ag.econ.211242
60. Sami MG, Sikder SP. Healing Balu River: Restoration of river eco-system through revitalizing the riparian zone of Balu River, Dhaka. ResearchGate, 2020. doi: 10.13140/RG.2.2.25999.51367
61. Bhuiyan NZ., Hossain MB, Ali MM, et al. Seasonal variation of trace elements in water and sediment of the Turag and Balu Rivers, Bangladesh. Egypt. J. Aquat. Biol. Fish. 2022; 26.
Refbacks
- There are currently no refbacks.
Copyright (c) 2024 Md. Khairul Amin, Maiesha Nujhat, Abu Hena Mostafa Kamal, Sheikh Md. Ruhul Amin, Wahida Khatun, Joynab Akther, Mir Mohtadi Ibtesum
License URL:
https://creativecommons.org/licenses/by/4.0/
This site is licensed under a
Creative Commons Attribution 4.0 International License (CC BY 4.0).
.jpg)
