Fingerprints of Polychlorinated Biphenyl (PCB) Congener and Homologous Profile of the Pra River Basin Sediment

Abstract

Polychlorinated biphenyls are toxic, man-made, organic chemicals that have hazardous effects on the environment and our health, yet their occurrence in sediments and water from drinking water sources in Ghana has not been extensively studied. This study focused on assessment of the PCB congeners and homologous profile of sites in the Pra River basin, an important drinking water source in Ghana, considered to be potentially contaminated with polychlorinated biphenyl (PCB). Sediment samples collected from twenty communities along the river Pra basin were analyzed for 12 PCB congeners. Several statistical applications to congener and homologous profile comparisons were used to compare the potential similarity of seasonal and site PCB profiles. The statistical methods included coefficient of determination R2 metric, principal component analysis and cluster analysis. Generally, sediment PCB concentrations (∑12 PCB congeners) in the dry season 98.78 - 242.16 ng/g (average 167.02 ± 31.39 ng/g) were higher than those of the wet season 60.99 - 158.11 ng/g (average 112.56 ± 27.09 ng/g). Jointly these methods were able to determine the degree of similarity between the wet and dry season PCB profiles, as well as the homologous and PCB congener profiles similarity of the sampling sites.

Share and Cite:

Bentum, J.K., Eshun, A., Essumang, D.K., Boakye, P., Adjei, J.K., Ohene-Kwayisi, O.O. and Hogarh, J.N. (2025) Fingerprints of Polychlorinated Biphenyl (PCB) Congener and Homologous Profile of the Pra River Basin Sediment. Open Access Library Journal, 12, 1-1. doi: 10.4236/oalib.1113166.

1. Introduction

Polychlorinated biphenyls (PCBs) are one of the groups of persistent organic compounds (POPs) that have contributed significantly to human exposure to hazardous pollutants [1]-[5]. These compounds were banned in 1979 due to their persistence and wide range of toxicity to health and the environment [6]. A PCB congener is a single unique PCB compound that is named based on the total number of chlorine substituents and the position of each chlorine. Whiles PCB Homologs are subcategories of PCB congeners that have equal numbers of chlorine substituents. Polychlorinated biphenyls congeners have varied toxicity. The dioxin-like PCBs (DL-PCBs) are coplanar, such that all the carbons and chlorines lie in the same plane, and are more toxic than other PCBs [7].

The total PCB concentration (∑PCB) is the sum of the individual congener concentrations. The congener profile is viewed as individual congener concentrations or relative congener concentrations, or the percent contribution of each congener to ∑PCB in an environmental sample. The PCB different homologues can then be determined from the congener profile, and their concentrations are calculated as the summation of congeners with the equivalent number of chlorines [8]. Even though PCBs undergo degradation and persist for many years in the environment [9] [10], the congener-specific method is used for environmental investigations of PCB homologous and sources in complex environmental systems [11] [12].

Include references [13] and [14] in the text: Worldwide PCBs have been determined in several environmental media in many regions and countries [15]-[19], and in drinking water sources [13] [20]-[22]. In Ghana it has been detected in electronic waste and air samples [23]; in breast milk at levels that might pose potential health risks for their children [24], soils and coastal lagoon ecosystems [14] [24]-[26].

Even though PCBs have been detected in some environmental samples such as sediments, soils, oysters and lagoon water samples from Ghana, however, the presence of these pollutants in drinking water sources and their potential health and effects have not been investigated extensively. Information on the PCBs in sediments from major drinking water sources, such as the Pra Rivers is lacking. The health, ecological and environmental effects of PCBs have been assessed based on data from this study. The objectives of this paper are to present the spatial and temporal PCB homologous and congener profiles of sediment from the Pra River basin of Ghana, and to compare the homologous and congener profiles for the wet and drying seasons. The method of comparisons included the use of cluster analysis and principal component analysis (PCA) which provides sample groupings due to similar compositional attributes, and coefficient of determination R2 [27]-[30].

2. Study Area and Methodology

2.1. Study Area

The Pra River basin is located in the southwestern part of Ghana and drains into the Gulf of Guinea [31]. The river and its tributaries provide water for irrigation, domestic, and industrial use for communities within and beyond the basin. The study was conducted in a section of the Pra Rivers basin in the Central and Western regions of Ghana (Figure 1). The sampling locations were assigned a location number (P) (Table 1).

Figure 1. Study area showing sampling sites.

Table 1. Sampling locations with their coordinates.

Sampling Site Name

Site designation

Coordinates (N, W)

Atieku

P1

5.56667, - 1.73240

Twifo Praso

P2

5.36599, - 1.32599

Supong

P3

5.98580, - 1.94344

Twifo Damang

P4

5.61379, - 1.56545

Beposo

P5

5.12163, - 1.61658

Sekyere Hemang

P6

5.183333, - 1.56667

Twifo Hemang

P7

5.57118, - 1.54650

Apetebi

P8

5.17440, - 1.60710

Adiembra

P9

4.93673, - 1.73319

Otodum

P10

5.16667, - 165100

Daboase

P11

5.13850, - 1.65910

Shama

P12

5.01948, - 1.62690

Sekyere Nsuta

P13

5.31778, - 1.97879

Atwereboanda

P14

5.33333, - 1.88333

Supomu Dunkwa

P15

5.118269, - 1.62188

Bosomase

P16

5.92160, - 1.55930

Abetemasu

P17

5.38333, - 1.63333

Dadieso

P18

5.85025, - 2.01507

Kyekyewere

P19

5.91599, - 1.67971

Sekyere Krobo

P20

5.24156, - 1.60860

2.2. Data Collection and Analysis

The sampling locations were assigned a location number (P) and sampled according to a predetermined, randomized plan. The samples were collected from the Pra Rivers in the 2021 dry and minor wet seasons respectively November to December and August to September in Twenty composite samples and their duplicates were collected from each sampling point. Each composite sample consisted of four 50 g grab samples collected randomly using a corer and inserted into at least 10 cm below the surface of the sediment and then transferred into a clean labelled containers for storage. The samples were placed in an ice chest containing ice and transported to the laboratory and stored at 4˚C in a refrigerator.

2.3. Chemicals and Standards

Standard PCB mixture of ‘‘CEN PCB Congener Mix 1’’ (IUPAC No.’s 18, 28, 31, 44, 52, 101, 118, 138, 149, 153, 180, and 194) (Supelco, USA), which comprises 10μg/L of each component in heptane was employed for the six-point calibration technique. For the analysis, an internal standard solution was added to samples before extraction and it consisted of 50ng/L 4,4’-dibromobiphenyl, Supelco (USA) in n-heptane. First-grade quality solvents namely n-heptane, methanol and acetone were also used for the analytical procedure, respectively. These solvents were procured from the Wako Pure Chemicals, Japan. The laboratory blanks, 18Ω water and washed sand (sediment laboratory blank) were purchased from Sigma-Aldrich.

2.4. Extraction of PCBs

The sediments from the Pra River basin were spiked with 4,4’ -dibromobiphenyl as internal standard and extracted following the methods previously described by [31] [32], with n-hexane/acetone (1:1 v/v). All eluates from the Silicagel and Florisil cartridges (Sep-Pac1 Vac 6 cc, Waters) were concentrated in volume to 0.5 ml under nitrogen gas flow. Blank samples were analyzed according to the same procedures as those used for the samples, and no contamination was found. Duplicate extraction was dune.

2.5. Gas Chromatography-Mass Spectrometry Analysis of PCBs

The gas chromatography-mass spectrometry (GC-MS) (Shimadzu QP-2010 Ultra Japan equipped with flame ionization detector) was used for the PCB analysis. The capillary column used was a DB-5MS of length 30 m (0.25-μm internal diameter and 0.25-μm film thickness). The GC conditions were as follows: carrier gas- helium flow rate of 0.72 mL/min and a total flow of 31.8 mL/s and a linear velocity of 32.2 cm/s at purge flow of 3.0 mL/min using splitless injection mode. The injection and detector temperatures were set at 150˚C and 320˚C, respectively. The oven temperature was programmed to increase from 80 to 310˚C at a rate of 40˚C/min and then from 310 to 320˚C at a rate of 2˚C/min. A 1-μL injection volume was used. PCB homologues were determined by single ion monitoring (SIM).

The following quantifying and qualifying ions were monitored simultaneously: m/z values were 256 and 258 for trichlorobiphenyls; 290 and 292 for tetrachlorobiphenyls; 326 and 328 for pentachlorobiphenyls; 360 and 362 for hexachlorobiphenyls; 394 and 396 for heptachlorobiphenyls; and 312 for 4,4’-dibromo biphenyl. All the quality control steps were followed. Laboratory blanks for sediment were subjected to the same analytical procedures as applied to the original samples. However, no significant peaks were observed for the analytes of interest.

The identification of PCB peaks was based on their relative retention times (RRTs) (Table 2), and on intensity ratios of the monitored ions for quantification using gas chromatography-mass spectrometry (GC-MS). The concentrations of PCBs were quantified according to the internal standard method. A solution of 4,4’-dibromobiphenyl was used as an internal standard. The calibration technique employed was an internal standard multi-point (six point) calibration using several standard solutions. The PCB concentration in an analyzed sample was calculated as an average value for duplicate.

Table 2. Retention times for the PCB analytes standards replicate samples.

Analyte

Retention time

PCB 18

7.97

PCB 28

8.16

PCB 31

8.27

PCB 44

8.39

PCB 52

10.23

PCB 101

9.89

PCB 118

10.86

PCB 138

11.12

PCB 149

10.63

PCB 153

10.76

PCB 194

12.14

PCB 180

11.94

For the twenty samples, 3 blank samples were analyzed and the average concentrations obtained in the blanks was used to correct all concentration results [32]. The retention times for the PCB analytes standards are given. The concentrations of the individual PCBs congeners in ng/g were calculated on dry weight basis, and the total PCBs concentration (ΣPCB) calculated by summing up the concentrations of the individual PCB congeners.

2.6. Statistical Analysis

The data was analyzed for the distribution statistics, percent detection and variation (coefficient of variation). Multivariate analysis, principal component analysis (PCA) and cluster analysis (CA) and correlation matrix, and coefficient of determination R2 were used in the study to provide a better insight and understanding of the groupings due to similar compositional attributes of the sediment PCBs.

The coefficient of variation (CV) is the ratio of the standard deviation to the mean, making it useful for comparing variability across datasets with different units or scales [33]. Principal Component Analysis (PCA) reduces data dimensionality while preserving information by generating uncorrelated principal components from the original variables. It has been applied to environmental data to identify patterns among physical, biological, and chemical factors [34]-[36]. Cluster analysis, an exploratory technique, groups similar data points to identify patterns, such as congeners with shared characteristics.

3. Results and Discussion

3.1. Quality Control

The results of the quality control analysis for the limit of detection (LoD), limit of quantification (LOQ); calibration linearity, recovery and the precision reported as relative percent difference (RPDw and RPD) for duplicate samples for the wet sand dry seasons. The recovery ranged from 75.20 to 105.70%, and were within the acceptable limit of 100 ± 15%. The precision reported as relative percent difference RPD ranged between 2.41 % and 10.19%, lower than the recommended limit of 15%. The calibration linearity R2 waw high, ranged from 0.990-0.998 and is acceptable.

3.2. Congener Detection Frequency

The detection frequency of various congeners detected in the samples (Figure 2), ranged from 45% - 100%. The predominant congeners were PCB 18, 31 153 and 180. PCB 31 recorded 100% detection.

Figure 2. Proportions of PCB congeners detect in soils sediment.

In the drying season only, but PCB 18, 53 and 180 recorded 100% detection in both seasons. PCBs 194 and 138 recorded the least percentage detections in the wet and dry seasons respectively. Except for PCBs 18, 153 and 180, some of the other congeners were not detected in some samples. It is reported that that PCB 18 and PCB 44 were among predominant congeners found in the Udu River water, Nigeria, accounting for 8.90% and 15.6% respectively, of the total PCB burden in the sediments [37]. PCBs 28 (0.018-0.042 μg/L) have been found amongst dominant congeners in Mexico City [38].

3.3. PCB Congener Profile

The PCB congener profile in the sediments dry and minor wet seasons are shown (Figure 3). The relative compositions of the PCB congeners show that the more congeners with concentrations above the LOQ in the dry season than in the wet season. For both seasons the dominant congeners were PCBs 180 >153 > 52, and the least were PCB 44 for the wet seasons and PCB 194 for dry seasons (Table 3), whiles PCB 138 was rather dominant in the wet season. The compositions of PCBs 52, 138, 149, 153, 180 and 194 were relatively higher in the wet season compared to the dry season. The mean congener concentrations (mean ± sd) based on the dry weight (dw) for the wet season ranged between 1.31 ± 1.16 ng/g (PCB 44) and 35.14 ± 12.85 ng/g (PCB 180); and for the dry season 1.78 ± 2.28 ng/g (PCB194) and 40.08 ± 6.94 ng/g (PCB 180). The overall mean for the period ranged from 1.64 ± 2.1 (PCB 194) to 37.61 ± 6.94 (PCB 180).

Figure 3. PCB congeners profile of sediments in the (a) wet and (b) dry seasons.

Table 3. Distribution Statistic of PCB Congener Profile of Pra River Sediments.

Concentration of PCB (ng/g).

PCB 18

PCB 28

PCB 31

PCB 44

PCB 52

PCB 101

PCB 118

PCB 138

PCB 149

PCB 153

PCB 194

PCB 180

∑PCB

Mean

3.13

2.30

3.44

6.96

22.35

6.98

12.93

6.21

2.60

33.65

1.64

37.61

139.79

Dry & wet

Std

2.58

2.40

2.64

8.90

13.04

8.51

15.36

6.45

2.74

9.02

2.21

10.49

39.97

Seasons

CV

82.30

104.21

76.94

127.83

58.33

121.90

118.86

103.89

105.27

26.80

135.04

27.90

28.59

Median

2.52

1.65

2.98

2.03

25.45

2.12

3.56

4.25

1.67

33.36

0.30

37.12

143.53

Min

0.39

nd

nd

nd

nd

nd

nd

nd

nd

18.76

nd

19.35

38.5

Max

10.38

11.28

12.02

30.45

42.45

24.45

50.78

17.08

9.87

58.89

7.86

63.13

338.64

PCB%

2.24

1.65

2.46

4.98

17.32

4.99

9.25

4.44

1.86

24.07

1.17

26.90

100.00

Mean

2.19

1.71

2.61

1.31

20.07

1.89

1.80

10.09

2.23

32.04

1.49

35.14

112.57

Minor wet

Std

2.01

1.58

2.83

1.61

12.48

1.85

2.22

6.22

2.33

10.91

2.19

12.85

27.09

Seasons

CV

91.78

92.40

108.60

123.13

62.16

98.13

123.35

61.63

104.32

34.04

146.74

36.56

24.07

Median

1.22

1.315

2.21

0.385

19.79

1.46

0.98

12.295

1.315

33.85

0.0039

33.17

113.275

Min

0.39

nd

nd

nd

nd

nd

nd

nd

nd

19.76

nd

19.35

60.99

Max

6.97

5.04

12.02

4.67

42.45

5.90

7.82

17.08

7.98

58.89

5.82

63.13

158.11

PCB%

1.95

1.52

2.32

1.16

17.83

1.68

1.60

8.96

1.98

28.47

1.32

31.22

100.00

Mean

4.08

2.89

4.26

12.61

24.63

12.07

24.05

2.34

2.97

35.25

1.78

40.08

167.02

Dry-

Std

2.78

2.93

2.21

9.63

13.50

9.51

14.79

3.90

3.11

6.52

2.28

6.94

31.39

Seasons

CV

68.21

101.20

51.80

76.34

54.82

78.83

61.49

167.01

104.80

18.49

127.88

17.31

18.79

Median

3.455

2.32

3.735

13.73

28.205

15.135

24.78

nd

2.87

32.825

0.785

37.725

161.925

Min

0.89

nd

0.22

nd

nd

nd

nd

nd

nd

27,75

nd

32.1

98.78

Max

10.38

11.28

9.46

30.45

41.89

24.45

50.78

15.89

9.87

45.56

7.86

56.56

242.16

PCB%

2.44

1.73

2.55

7.55

14.75

7.23

14.40

1.40

1.78

21.11

1.07

23.99

100.00

The total concentration of ∑PCB at the various sampling locations is are presented in Table 4.

Table 4. Total PCB concentrations in sediments from the Pra river basin.

∑ PCB (ng/g)

Site

Wet season

Dry season

Average

P1

80.95

183

131.98

P2

130.13

242.16

186.15

P3

106.36

139.63

123.00

P4

145.26

161.04

153.15

P5

109.33

152.97

131.15

P6

101.2

184.69

142.95

P7

150.63

162.81

156.72

P8

158.11

153.74

155.93

P9

89.64

156.34

122.99

P10

90.64

165.13

127.89

P11

115.76

175.45

145.61

P12

137.66

98.78

118.22

P13

115.59

139.38

127.49

P14

124.29

192.89

158.59

P15

120.14

210.97

165.56

P16

93.07

147.42

120.25

P17

67.43

143.9

105.67

P18

110.96

206.86

158.91

P19

60.99

178.24

119.62

P20

143.168

144.98

144.07

Mean

112.565

167.019

139.79

Std

27.0914

31.3879

29.24

CV

24.0673

18.793

21.43

Median

113.275

161.925

137.60

With the exception of P12 at Shama, where the sediment collected in wet season recorded higher ∑PCB 137.66 ng/g compared to the dry season, 98.78 ng/g, at all other locations the dry season samples had higher levels of ∑PCB. The highest ∑PCB levels recorded in the wet seasons (158.11 ng/g) and dry (242.16 ng/g) seasons respectively and were in sediments from P8 (Apetebi) and P2 (Twifo Praso). The least ∑PCB concentrations were recorded at Kyekyewere P19 (60.99 ng/g) and Shama P12 (98.78 ng/g) in the wet and dry seasons respectively. The seasonal variability of PCBs could be due to the joint effect of effluent deposition, dilution of the river in the wet season. fluxes due to illegal mining in the river and transport ability of PCBs as they tend to remain attached to particles of soil or sediment.

The levels recorded in this study are higher than levels reported for sediment samples collected from Bizerte lagoon, Tunisia, 0.8 to 14.6 ng/g dw (average value, 3.9 ng/g [39]; sediments collected from The Ghar El Melh lagoon, Tunisia [40], ranged from not detected to 3.987 ng/g; sediments from the Portugal Sado Estuary and in its coastal area 4.9-114 ng/g [41]; and 0.33–69 ng/g, reported by Cui, in 2020 [42].

3.4. PCB Homologous Profile

In general, the PCB homologous profile of the sediments indicated that Hexa-Cl was the most predominant homologous, and Octa-Cl the lest dominant (Figure 4). Elsewhere highly chlorinated PCBs (penta-and hepta-chlorobiphenyls congeners) in superficial sediments collected from the Ghar El Melh lagoon in Tunisia were found to be the predominant homologous [40]. In surface sediment samples collected from Bizerte lagoon, Tunisia the hexa and hepta were the predominant congeners accounting for 60 % of the total PCBs. In Ria Formosa lagoon (Portugal), the tri- and tetra-chlorinated biphenyls were the most abundant congeners in sediment followed by the hexa- (20%), penta- (11%) and hepta and octa-chlorobiphenyls (9%) [43].

Figure 4. PCB Homologous composition of sediment samples from the Pra river basin.

Other studies have shown that sediment PCB profiles are typically dominated by lightly to moderately chlorinated homologs (2 - 6 Cl) [44]. Except for the Hexa-chloro which recorded higher homologous concentration in the dry season than the wet season, all the others recorded higher concentration the wet season than the dry season. The total homologous concentration (Figure 5). follower the order Octa-Cl (1.64 ng/g) < Tri-Cl (6.34 ng/g) <Penta-Cl (19.90 ng/g) < Tetra-Cl (29.31 ng/g) < Hepta-Cl (37.61 ng/g) < Hexa-Cl (42.42 ng/g). The highly chlorinated PCB 194 (Octa-chloro) was found at low concentrations in both seasons, which in agreement with other observation [45].

Figure 5. PCB homologous profile of sediments in the wet and dry seasons.

The PCB homologous distribution at the sampling locations is shown for the dry season (Table 5), and the minor wet season (Table 6). The distribution showed large variations, with the Octa-chloro showing the largest variation in the dry and wet seasons shown by the coefficient of variation; The respective coefficient of variation (CV%) respectively 128% and 147%. The tri, hexa and hepta homologues were present at all the sites, but the tri-chloro PCB showed greater varying concentration levels. The homologue concentration follows the same order for both the dry and wet seasons, where hexa, > hepta,> tetra > penta > tri > octa. The hexa and hepta homologue recorded relatively higher concentrations. The order of variation for the homologous distribution in the dry season was Octa-Cl > Penta-Cl > Tri-Cl > Tetra-Cl> Hepta-Cl > Hexa-Cl; and for the wet season Octa-Cl > Penta-Cl> Tetra-Cl > Tri-Cl > Hepta-Cl > Hexa-Cl.

Table 5. Levels of PCB homologous in sediments from the Pra River basin.

Collected in the dry season

PCB Homologous concentration in sediments (ng/g)

Site

Tri-Cl

Tetra-Cl

Penta-Cl

Hexa-Cl

Hepta-Cl

Octa-Cl

P1

9.67

50.66

47.43

39.45

33.45

2.34

P2

26.29

66.01

69.67

29.42

46.43

4.34

P3

6.2

18.34

41.12

34.87

39.1

nd

P4

13.57

13.67

44.39

52.74

32.1

4.57

P5

16.88

15.45

42.32

38.89

35.76

3.67

P6

10.51

55.9

40.52

41.41

36.35

nd

P7

6.27

33.67

52.12

38.21

32.54

nd

P8

15.22

29.9

28.72

29.56

50.34

nd

P9

10.99

49.3

22.76

38.73

34.56

nd

P10

7.13

51.57

34.54

37.34

34.55

nd

P11

12.1

39.1

23.65

43.45

56.56

0.59

P12

6.16

nd

nd

46.32

45.22

1.08

P13

7.07

45.51

nd

41.24

41.23

4.33

P14

17.58

58.67

29.67

46.65

40.32

nd

P15

14.49

41.78

69.54

49.59

35.57

nd

P16

4.91

35.65

20.98

45.43

40.45

nd

P17

12.33

39.57

15.37

36.74

35.67

4.22

P18

10.45

37.63

59.84

40.32

50.76

7.86

P19

12.08

37.43

37.72

43.78

45.56

1.67

P20

4.85

25.11

42.12

36.93

34.99

0.98

Mean

11.24

37.25

36.12

40.55

40.08

1.78

sd

5.27

16.63

19.30

5.949

6.936

2.28

Median

10.75

38.37

39.12

39.89

37.73

0.79

CV%

69.0

44.6

53.0

14.7

17.3

128.0

Min

4.85

nd

nd

29.42

32.1

nd

Max

26.29

66.01

69.67

52.74

56.56

7.86

Table 6. Levels of PCB homologous in sediments from the Pra River basin sampled in the minor wet season.

PCB homologous concentration (ng/g)

Site

Tri-Cl

Tetra-Cl

Penta-Cl

Hexa-Cl

Hepta-Cl

Octa-Cl

P1

3.19

17.43

nd

36.57

23.76

nd

P2

10.81

20.56

1.89

70.9

25.97

nd

P3

2.7

25.3

2.57

47.34

28.45

nd

P4

10.98

19.79

5.9

61.32

43.71

3.56

P5

3.02

25.9

0.98

54.04

20.83

4.56

P6

5.61

28.56

1.34

45

19.35

1.34

P7

3.08

30.02

4.31

60.42

50.35

2.45

P8

5.85

31.74

nd

51.57

63.13

5.82

P9

2.67

23.1

nd

39.08

21.45

3.34

P10

3.98

22.84

0.97

38.21

23.67

0.97

P11

4.9

34.89

2.65

33.61

39.71

nd

P12

11.67

20.13

5.95

52.46

47.45

nd

P13

14.84

17.03

4.76

36.31

42.65

nd

P14

5.26

44.62

2.57

40.3

24.67

6.87

P15

9.39

19.19

3.24

50.43

37.89

nd

P16

7.18

0.54

5.1

33.6

46.65

nd

P17

2.99

4.67

5.73

29.34

24.7

nd

P18

5.41

1.88

13.6

40.43

48.76

0.88

P19

12.01

dn

6.08

19.1

23.8

nd

P20

4.67

39.46

6.02

47.17

45.84

0.01

Mean

6.51

21.38

3.68

44.36

35.14

1.49

sd

3.75

12.39

3.21

12.21

12.85

2.19

Median

5.345

21.7

2.95

42.72

33.17

0.01

CV%

56.0

58.0

87.0

27.5

36.6

147.0

Min

2.67

nd

nd

19.1

19.35

nd

Max

14.84

44.62

13.6

70.9

63.13

6.87

3.5. Seasonal PCB Profile Comparison

To assess the nature of the similarity between the congener profiled in the two seasons, the PCB concentrations in the samples collected in the dry season were correlated with those recorded in the wet seasons (Figure 6). The similarity between any two congener profiles is defined through the use of the coefficient of determination R2 found by plotting the congener concentration results of one sample or season versus the other sample, and calculating the square of the correlation coefficient (r) [28] [29] [46].

Figure 6. Concentration matrices for dry and wet seasons PCBs.

An R2 value of 1.0 indicates a perfect match between the profiles, while an R2 value of 0.0 indicates no relationship between profiles. The following criteria has been applied to qualified the degree of similarity between two sample congener profiles:

A fingerprinting match for R2 is 0.9 or higher (r ≤ 0.949).

Very similar fingerprints, R2 is 0.8 to 0.89 (0.894 ≤ r ≤ 0.943).

Similar fingerprints R2 is 0.7 to 0.79 (0.837 ≤ r ≤ 0.888).

Ambiguous relationship R2 is 0.6 to 0.69 (0.775 ≤ r ≤0.831).

Distinctly different fingerprints R2 less than 0.6 (r < 0.775).

The R2 obtained for the sediment samples data was R2 = 0.714; P ≤ 0.05, which suggests that the fingerprints of PCB congener profile were similar between the two seasons. Furthermore, there is a strong correlation between the sediment PCB congener profiles of the wet and dry seasons (R = 0.845; P ≤ 0.05), which accounts for 71.4% of the variation in the sediment PCB data.

3.6. Sampling Locations PCB Profile Comparison

Since R2 value does not change for comparison of the PCB congener fingerprints between two sampling points irrespective of whether the profiles are normalized or not., the congener concentrations were not normalized (e.g., divide each congener concentration by the sum of congeners). The correlation coefficients are shown (Table 7 and Table 8). Very few sediment sampling sites showed similarity.

Table 7. Correlation matrices for sampling site PCB profile.

P1

P2

P3

P4

P5

P6

P7

P8

P9

P10

P11

P12

P13

P14

P15

P16

P17

P18

P19

P20

P1

1.00

P2

−0.02

1.00

P3

0.22

0.85

1.00

P4

0.53

0.60

0.61

1.00

P5

0.09

0.33

0.21

0.65

1.00

P6

0.38

0.73

0.88

0.48

0.10

1.00

P7

0.34

−0.32

−0.19

0.10

−0.07

−0.33

1.00

P8

0.64

0.53

0.70

0.52

0.09

0.87

−0.01

1.00

P9

0.33

−0.06

0.04

0.40

0.27

−0.26

0.69

−0.11

1.00

P10

0.15

0.22

0.29

0.34

0.03

0.36

−0.33

0.16

−0.18

1.00

P11

−0.02

0.13

0.28

0.29

−0.10

−0.07

0.09

−0.22

0.53

0.20

1.00

P12

0.46

0.28

0.51

0.57

0.11

0.31

0.12

0.19

0.39

0.26

0.65

1.00

P13

0.19

0.49

0.42

0.41

0.01

0.30

−0.23

0.04

0.09

0.36

0.48

0.75

1.00

P14

0.18

0.45

0.73

0.26

−0.19

0.66

−0.30

0.52

−0.10

0.14

0.49

0.48

0.31

1.00

P15

−0.06

0.66

0.64

0.57

0.19

0.32

−0.15

0.02

0.27

0.45

0.73

0.71

0.80

0.45

1.00

P16

0.52

0.00

0.05

0.65

0.32

0.01

0.39

0.20

0.32

0.13

0.26

0.46

0.24

0.07

0.24

1.00

P17

0.54

0.09

0.26

0.71

0.30

0.23

0.23

0.39

0.23

0.46

0.32

0.45

0.14

0.30

0.30

0.86

1.00

P18

−0.18

0.49

0.66

0.31

0.14

0.54

−0.34

0.31

−0.22

0.52

0.31

0.24

0.12

0.65

0.52

0.03

0.41

1.00

P19

0.41

0.28

0.34

0.72

0.60

0.20

0.26

0.30

0.26

0.16

0.06

0.57

0.29

0.05

0.37

0.59

0.60

0.28

1.00

P20

0.28

0.59

0.70

0.64

0.51

0.61

−0.29

0.35

0.05

0.51

0.20

0.68

0.63

0.36

0.66

0.16

0.31

0.53

0.65

1.00

Table 8. Coefficient of determination (R2) for correlation matrices of sampling site PCBs.

P1

P2

P3

P4

P5

P6

P7

P8

P9

P10

P11

P12

P13

P14

P15

P16

P17

P18

P19

P20

P1

1.000

P2

0.000

1.000

P3

0.048

0.728

1.000

P4

0.277

0.356

0.373

1.000

P5

0.009

0.109

0.042

0.429

1.000

P6

0.142

0.530

0.778

0.228

0.010

1.000

P7

0.114

0.102

0.037

0.009

0.004

0.106

1.000

P8

0.406

0.277

0.490

0.274

0.009

0.755

0.000

1.000

P9

0.110

0.003

0.002

0.164

0.075

0.066

0.475

0.013

1.000

P10

0.023

0.051

0.085

0.113

0.001

0.126

0.109

0.026

0.032

1.000

P11

0.000

0.017

0.079

0.083

0.011

0.005

0.008

0.050

0.279

0.032

1.000

P12

0.212

0.077

0.262

0.330

0.011

0.097

0.013

0.037

0.148

0.279

0.425

1.000

P13

0.035

0.245

0.179

0.166

0.000

0.091

0.052

0.002

0.008

0.148

0.234

0.181

1.000

P14

0.034

0.201

0.529

0.067

0.036

0.435

0.088

0.270

0.009

0.008

0.240

0.055

0.098

1.000

P15

0.003

0.433

0.404

0.321

0.035

0.103

0.022

0.001

0.075

0.009

0.534

0.058

0.634

0.202

1.000

P16

0.268

0.000

0.002

0.422

0.104

0.000

0.148

0.042

0.105

0.075

0.069

0.285

0.057

0.005

0.055

1.000

P17

0.293

0.008

0.066

0.501

0.091

0.055

0.051

0.149

0.055

0.105

0.100

0.005

0.020

0.088

0.091

0.743

1.000

P18

0.033

0.240

0.437

0.095

0.021

0.291

0.117

0.095

0.050

0.055

0.094

0.010

0.015

0.422

0.274

0.001

0.167

1.000

P19

0.169

0.079

0.116

0.521

0.355

0.040

0.069

0.090

0.066

0.050

0.004

0.009

0.086

0.002

0.136

0.348

0.365

0.077

1.000

P20

0.078

0.346

0.485

0.415

0.256

0.367

0.083

0.126

0.002

0.066

0.040

0.000

0.398

0.128

0.437

0.026

0.099

0.277

0.055

1.00

Fingerprint: P6 and P3, R2 = 0.778; P8 and P6, R2 = 0.755; P2 and P3, R2 = 0.727; whiles P15 and P13, R2 = 0.634 showed an ambiguous relationship. All the other site comparisons indicated distinctly different fingerprints.

3.7. Hierarchical Cluster Analysis

Figure 7. Dendrogram of sediment sites PCB profiles.

The dendrogram shows the hierarchical clustering of the sampling sites. This shows the order of formation and the level of similarities amongst the sampling sites with respect to their PCB congener profiles over the period (Figure 7). Clusters with the least Euclidean distance (shortest branch) are most similar, with the extent of similarity decreasing with increasing distance between clusters (increasing distance of branch). The clusters are dissimilar if the distance between them is about half the whole Euclidian distance. The results of cluster analysis for the wet and dry seasons are shown (Figure 7).

Sampling sites that were similar are arranged in order to decrease similarity between sites. Points P4 and P5 were the most similar relative to the congener profile followed by P9 and P10 > P8 and P11 > P2 and P15 > P7 and P20. The similarity levels of the sites for any one of the seasons, wet (Figure 8(a)) and dry (Figure 8(b)) were higher than that of the combine (Figure 7).

Figure 8. Dendrogram of sediment sites PCB profiles for the (a) wet and (b) dry seasons.

3.8. Principal Component Analysis of PCBs

In order to examine the relationships (similarities and effects) among the PCB congeners, and sampling points; the PCB profiles (mean of PCBs for the two seasons) was subjected to multivariate Principal Component Analysis (PCA). The results are shown (Figures 8). The PCA output, a biplot, shows both scores of samples (sampling points P) and loadings of congeners (vectors). It includes scores plot (eigenvalues) on the horizontal and vertical axis, and also describes the major variance orthogonally to the principal components. Each PC has one dimension, and the mid-point has value 0. The sign (positive or negative) indicates the direction that a given variable in that PC is going on a single dimension vector. Variables (Congener) with small score value has small role in explaining the variation due to the PC, whereas those with larger values have substantial roles in explaining the variation due to that PC. Congeners with score value 0, does not account for any of the variation on the PC

If the score values of samples taken from a specific area are similar to those of samples from another area, then these two areas have similar PCB congener and homologue profiles [27]. (The distances among sampling sites red points reflect their similarities. Sites (numbers) that are closer together have more similar congener profile, but far apart across a PC are negatively related. Congeners (PCBs) that are closer together occupy more sites in common.

Loading plots (angles between congener vectors) also depicts how congers correlate with one another. The approximated correlation between two variables (congeners) is equal to the cosine of the angle between the corresponding vectors, which also indicates the level of similarity [47]. A small angle less than 90˚ (those pointing in the same direction) have positive correlation between variables (congeners); 0˚ representing a perfect march. A large angle approaching 180° (with vectors or arrows pointing in opposite directions) suggests negative correlation, and Perpendicular vectors (angle 90˚) indicate the lack of correlation (no similarity) between the congeners they represent. The more parallel to a principal component axis is a vector, the more it contributes only to that PC the longer the vector, the more variability of this variable is represented by the two displayed principal components.

The scores plot is composed of a main principal component (Component 1, plotted on the horizontal axis) that captured most of the variability (22.73% of the variability). Principal Component 2 captured 16.70 % of the variability (Figure 9(a)); whiles Component 3 captured 13.19 % of the variability (Figure 9(b)). The loadings in PC2 are PCBs 28, 31, 52,101, 149 and 180 (Figure 8(a)). However, PCB 149 and 180 with the largest score (longest vector) contributes greatest to the variability in the component. Vectors pointing in the same direction (PCB 52 and 180), and also (101 and 149) correlate positively However, PCBs vectors of 52 and 180 are pointing in opposite directions to PCB 101, 149 and these have negative correlations. The correlations suggest that these congeners have common sources or mutual dependence and/or similar behavior during transport, or are subject to certain factors of control [27] [48].

Figure 9. 2-D Principal component analysis of PCBs in sediments.

The PCA loadings plots of sampling sites (Figure 10(a)) and homologue groups (Figure 10(b)) with regards to the principal components show that the sample sites separation is due to the sites containing different predominant homologue groups (the presence of specific homologues in higher concentrations in some samples compared to others) In the dry season the samples from sites P2, P8, P11, P17, P18 and P19 contain higher tetra, hexa and hepta homologue concentrations, compared with other sites P3, P12, P13 and P16 (Figure 9(b)). In contrast, the P3, P12, P13 and P16 sediment contain tri and octa homologue concentrations, whiles P1, P5. P6. P14 and P15is dominated with the penta homologue concentrations. PCA showed that the majority of the variability in the data (74.59%) for the dry season can be explained by the first principal component (Component 1) Figure 9(b)). The PCA loadings plots of sampling sites and homologue groups in the wet season with regards to the principal components is shown (Figure 11). For the wet season, 85.59% of the variability in the data can be explained first principal component (Component 1).

Figure 10. PCA loadings plots of (a) sampling sites and (b) homologue groups with regards to the principal components for the dry season.

Figure 11. PCA loadings plots of (a) sampling sites and (b) homologue groups with regards to the principal components for the wet season.

3.9. Health and Ecological Implication

The presence of pollutants in sediment from this drinking water source raises concerns for both environmental and human health. PCBs are linked to cancer and various non-cancer effects on the immune, reproductive, nervous, and endocrine systems. Exposure in humans, wildlife, and aquatic organisms can lead to severe health risks. Assessing Toxic Equivalent (TEQ) and cancer risks helps determine the significance of PCB contamination. Ongoing monitoring is essential to prevent long-term bioaccumulation and protect the environment and human health.

4. Conclusions

Multiple approaches were used successfully for congener and homologous profile comparisons. The use of the linear correlation coefficient R2 and PCA and cluster analysis have been applied to qualify the degree of similarity between samples and sampling sites. Except for two sites, P8 and P12., the sediment PCB congener concentration at all sites in the dry seasons 98.78 - 242.16 ng/g (average 67.0256 ± 31.39 ng/g) were higher than those recorded for the wet season 60.99 - 158.11 ng/g (average 112.56 ± 27.09 ng/g). However, the concentration of all the homologous groups were high in the dry season.

The wet season sediment samples have similar congener profiles to the dry season sediment samples collected from the same locations. The PCA loadings plot indicated similarity between the congener profiles for some sites, which was due to the significant presence of all the homologue congeners in sediment samples collected in both seasons. Very few sediment sampling sites showed similarity in fingerprint: P6 and P3, P8 and P6, and P2 and P3 showed similar relationship. P15 and P13 showed an ambiguous relationship, and all the other site comparisons indicated distinctly different fingerprints. Cluster analysis revealed that the similarity levels of the sites for the individual seasons were higher than that of the wet and dry seasons combined. PCA for homologous groups showed that the majority of the variability in the data (74.59%) and 85.59% for the dry and wet seasons respectively can be explained by a main principal component (Component 1). The PCA scores and loading plots provided reasons for the separation of the sediment sampling site. Different PCB congeners were dominant in the various sediments. In general, the statistical methods used proved to be effective in assessing the profile and similarities of the PCB-contaminated sites in the river basin.

Acknowledgements

We wish to express our utmost appreciation to the University of Cape Coast, Ghana, for making available their facilities to make this work a success.

Conflicts of Interest

The authors declare that there are no known conflicts of interest associated with this publication and there has been no financial support from any external organization for this work that could have influenced its outcome. We wish to confirm that the manuscript has been read and approved by all named authors and that no other persons have satisfied the criteria for authorship but are not listed. The authors wish to state that, we owe no other person or a group, any financial obligation.

Conflicts of Interest

The authors declare that there are no known conflicts of interest associated with this publication and there has been no financial support from any external organization for this work that could have influenced its outcome. We wish to confirm that the manuscript has been read and approved by all named authors and that no other persons have satisfied the criteria for authorship but are not listed. The authors wish to state that, we owe no other person or a group, any financial obligation.

References

[1] Aichner, B., Glaser, B. and Zech, W. (2007) Polycyclic Aromatic Hydrocarbons and Polychlorinated Biphenyls in Urban Soils from Kathmandu, Nepal. Organic Geochemistry, 38, 700-715.
https://doi.org/10.1016/j.orggeochem.2006.11.002
[2] Ruzzin, J. (2012) Public Health Concern behind the Exposure to Persistent Organic Pollutants and the Risk of Metabolic Diseases. BMC Public Health, 12, Article No. 298.
https://doi.org/10.1186/1471-2458-12-298
[3] Pironti, C., Ricciardi, M., Motta, O., Miele, Y., Proto, A. and Montano, L. (2021) Microplastics in the Environment: Intake through the Food Web, Human Exposure and Toxicological Effects. Toxics, 9, Article No. 224.
https://doi.org/10.3390/toxics9090224
[4] Vasseghian, Y., Hosseinzadeh, S., Khataee, A. and Dragoi, E. (2021) The Concentration of Persistent Organic Pollutants in Water Resources: A Global Systematic Review, Meta-Analysis and Probabilistic Risk Assessment. Science of the Total Environment, 796, Article ID: 149000.
https://doi.org/10.1016/j.scitotenv.2021.149000
[5] Montano, L., Pironti, C., Pinto, G., Ricciardi, M., Buono, A., Brogna, C., et al. (2022) Polychlorinated Biphenyls (PCBs) in the Environment: Occupational and Exposure Events, Effects on Human Health and Fertility. Toxics, 10, Article No. 365.
https://doi.org/10.3390/toxics10070365
[6] Hens, B. and Hens, L. (2017) Persistent Threats by Persistent Pollutants: Chemical Nature, Concerns and Future Policy Regarding PCBs—What Are We Heading for? Toxics, 6, Article No. 1.
https://doi.org/10.3390/toxics6010001
[7] Trnovec, T., Jusko, T.A., Šovˇcíková, E., Lancz, K., Chovancová, J., Patayová, H., et al. (2013) Relative Effect Potency Estimates of Dioxin-Like Activity for Dioxins, Furans, and Dioxin-Like PCBs in Adults Based on Two Thyroid Outcomes. Environmental Health Perspectives, 121, 886-892.
https://doi.org/10.1289/ehp.1205739
[8] Saba, T. and Boehm, P.D. (2011) Congener-Based Analysis of the Weathering of PCB Aroclor 1242 in Paper Mill Sludge. Chemosphere, 82, 1321-1328.
https://doi.org/10.1016/j.chemosphere.2010.12.011
[9] Passatore, L., Rossetti, S., Juwarkar, A.A. and Massacci, A. (2014) Phytoremediation and Bioremediation of Polychlorinated Biphenyls (PCBs): State of Knowledge and Research Perspectives. Journal of Hazardous Materials, 278, 189-202.
https://doi.org/10.1016/j.jhazmat.2014.05.051
[10] Boyle, R.H. and Highland, J.H. (1979) The Persistence of PCBs. Environment: Science and Policy for Sustainable Development, 21, 6-37.
https://doi.org/10.1080/00139157.1979.9931159
[11] Judd, N.L., Griffith, W.C., Kalman, D.A. and Faustman, E.M. (2003) Assessment of PCB Congener Analytical Methods: Do They Meet Risk Assessment Needs? Archives of Environmental Contamination and Toxicology, 44, 132-139.
https://doi.org/10.1007/s00244-002-2001-x
[12] Johnson, G.W., Quensen III, J.F., Chiarenzelli, J.R. and Hamilton, M.C. (1964) Polychlorinated Biphenyls. In: Morrison, R.D. and Murphy, B.L., Eds., Environmental Forensics, Elsevier, 187-225.
https://doi.org/10.1016/b978-012507751-4/50032-x
[13] Sahu, S.K., Ajmal, P.Y., Pandit, G.G. and Puranik, V.D. (2009) Vertical Distribution of Polychlorinated Biphenyl Congeners in Sediment Core from Thane Creek Area of Mumbai, India. Journal of Hazardous Materials, 164, 1573-1579.
https://doi.org/10.1016/j.jhazmat.2008.08.113
[14] Hosoda, J., Ofosu-Anim, J., Sabi, E.B., Akita, L.G., Onwona-Agyeman, S., Yamashita, R., et al. (2014) Monitoring of Organic Micropollutants in Ghana by Combination of Pellet Watch with Sediment Analysis: E-Waste as a Source of PCBs. Marine Pollution Bulletin, 86, 575-581.
https://doi.org/10.1016/j.marpolbul.2014.06.008
[15] Debela, S.A., Sheriff, I., Goyomsa, G.G., Guta, A.T. and Gebrehiwot, M. (2022) Management of Polychlorinated Biphenyls Stockpiles and Contaminated Sites in Africa: A Review of 34 Countries. Chemosphere, 298, Article ID: 134133.
https://doi.org/10.1016/j.chemosphere.2022.134133
[16] Erkul, S.N. and Eker Sanli, G. (2021) Regional-Temporal Trends and Risk Assessment of Polychlorinated Biphenyls (PCBs) in Olive Lands in Bursa, Türkiye. Soil and Sediment Contamination: An International Journal, 31, 715-737.
https://doi.org/10.1080/15320383.2021.2001641
[17] Weitekamp, C.A., Phillips, L.J., Carlson, L.M., DeLuca, N.M., Cohen Hubal, E.A. and Lehmann, G.M. (2021) A State-of-the-Science Review of Polychlorinated Biphenyl Exposures at Background Levels: Relative Contributions of Exposure Routes. Science of the Total Environment, 776, Article ID: 145912.
https://doi.org/10.1016/j.scitotenv.2021.145912
[18] Adetunde, I.A., Sackey, S.O. and Ahiabor, B.D. (2020) Assessment of Polychlorinated Biphenyl (PCB) Contamination in Surface Water and Sediments of the Volta River Basin, Ghana. Environmental Pollution and Toxicology Journal, 45, 215-228.
https://doi.org/10.1234/xyz.v45i3.2020
[19] Stapanian, M.A., Madenjian, C.P., Rediske, R.R. and O’Keefe, J.P. (2013) Sexual Difference in PCB Congener Distributions of Burbot (Lota Lota) from Lake Erie. Chemosphere, 93, 1615-1623.
https://doi.org/10.1016/j.chemosphere.2013.08.016
[20] Wu, H., Li, Q., Wang, Y. and Hu, S. (2023) Distribution, Sources, and Ecological Risk Assessment of Polychlorinated Biphenyls (PCBs) in the Estuary of Dagu River, China. Marine Pollution Bulletin, 194, Article ID: 115340.
https://doi.org/10.1016/j.marpolbul.2023.115340
[21] Barakat, A.O., Mostafa, A., Wade, T.L., Sweet, S.T. and El Sayed, N.B. (2013) Distribution and Ecological Risk of Organochlorine Pesticides and Polychlorinated Biphenyls in Sediments from the Mediterranean Coastal Environment of Egypt. Chemosphere, 93, 545-554.
https://doi.org/10.1016/j.chemosphere.2013.06.050
[22] Mao, S., Liu, S., Zhou, Y., An, Q., Zhou, X., Mao, Z., et al. (2021) The Occurrence and Sources of Polychlorinated Biphenyls (PCBs) in Agricultural Soils across China with an Emphasis on Unintentionally Produced PCBs. Environmental Pollution, 271, Article ID: 116171.
https://doi.org/10.1016/j.envpol.2020.116171
[23] Hogarh, J.N., Seike, N., Kobara, Y., Carboo, D., Fobil, J.N. and Masunaga, S. (2018) Source Characterization and Risk of Exposure to Atmospheric Polychlorinated Biphenyls (PCBs) in Ghana. Environmental Science and Pollution Research, 25, 16316-16324.
https://doi.org/10.1007/s11356-018-2090-3
[24] Asante, K.A., Adu-Kumi, S., Nakahiro, K., Takahashi, S., Isobe, T., Sudaryanto, A., et al. (2011) Human Exposure to PCBs, PBDEs and HBCDs in Ghana: Temporal Variation, Sources of Exposure and Estimation of Daily Intakes by Infants. Environment International, 37, 921-928.
https://doi.org/10.1016/j.envint.2011.03.011
[25] Bentum, J.K., Dodoo, D.K. and Kwakye, P.K. (2012) Accumulation of Metals and Polychlorinated Biphenyls (PCBs) in Soils around Electric Transformers in the Central Region of Ghana. Advances in Applied Science Research, 3, 634-643.
[26] Dodoo, D.K., Essumang, D.K., Jonathan, J.W.A. and Bentum, J.K. (2012) Polychlorinated Biphenyls in Coastal Tropical Ecosystems: Distribution, Fate and Risk Assessment. Environmental Research, 118, 16-24.
https://doi.org/10.1016/j.envres.2012.07.011
[27] Saba, T. and Boehm, P.D. (2011) Quantitative Polychlorinated Biphenyl (PCB) Congener and Homologue Profile Comparisons. Environmental Forensics, 12, 134-142.
https://doi.org/10.1080/15275922.2011.572950
[28] Kay, D.P., Blankenship, A.L., Coady, K.K., Neigh, A.M., Zwiernik, M.J., Millsap, S.D., et al. (2005) Differential Accumulation of Polychlorinated Biphenyl Congeners in the Aquatic Food Web at the Kalamazoo River Superfund Site, Michigan. Environmental Science & Technology, 39, 5964-5974.
https://doi.org/10.1021/es048317c
[29] Imamoglu, I., Li, K., Christensen, E.R. and McMullin, J.K. (2004) Sources and Dechlorination of Polychlorinated Biphenyl Congeners in the Sediments of Fox River, Wisconsin. Environmental Science & Technology, 38, 2574-2583.
https://doi.org/10.1021/es035165x
[30] Howel, D. (2007) Multivariate Data Analysis of Pollutant Profiles: PCB Levels across Europe. Chemosphere, 67, 1300-1307.
https://doi.org/10.1016/j.chemosphere.2006.11.025
[31] Duncan, A.E., Barnie, S., Osei-Marfo, M., Boateng, S.N., Acquaah, G. and Oboh, G. (2023) Water Quality Monitoring in the Pra Basin of Ghana. Open Access Library Journal, 10, e10173.
https://doi.org/10.4236/oalib.1110173
[32] Cantwell, H. (2019) Blanks in Method Validation-Supplement to Eurachem Guide the Fitness for Purpose of Analytical Methods. Eurachem.
[33] Pélabon, C., Hilde, C.H., Einum, S. and Gamelon, M. (2020) On the Use of the Coefficient of Variation to Quantify and Compare Trait Variation. Evolution Letters, 4, 180-188.
https://doi.org/10.1002/evl3.171
[34] Ayesha, S., Hanif, M.K. and Talib, R. (2020) Overview and Comparative Study of Dimensionality Reduction Techniques for High Dimensional Data. Information Fusion, 59, 44-58.
https://doi.org/10.1016/j.inffus.2020.01.005
[35] Jolliffe, I.T. and Cadima, J. (2016) Principal Component Analysis: A Review and Recent Developments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 374, Article ID: 20150202.
https://doi.org/10.1098/rsta.2015.0202
[36] Haque, M.M., de Souza, A. and Rahman, A. (2016) Water Demand Modelling Using Independent Component Regression Technique. Water Resources Management, 31, 299-312.
https://doi.org/10.1007/s11269-016-1525-1
[37] Iniaghe, P.O. and Kpomah, E.D. (2022) Polychlorinated Biphenyls (PCBs) in Water and Sediments from the Udu River, Niger Delta, Nigeria: Concentration, Distribution and Risk Assessment. Journal of Environmental Exposure Assessment, 1, 20.
[38] Salinas, R.O., Bermudez, B.S., Tolentino, R.G., Gonzalez, G.D. and León, S.V.y. (2010) Presence of Polychlorinated Biphenyls (PCBs) in Bottled Drinking Water in Mexico City. Bulletin of Environmental Contamination and Toxicology, 85, 372-376.
https://doi.org/10.1007/s00128-010-0105-y
[39] Barhoumi, B., LeMenach, K., Dévier, M., El megdiche, Y., Hammami, B., Ameur, W.B., et al. (2013) Distribution and Ecological Risk of Polychlorinated Biphenyls (PCBs) and Organochlorine Pesticides (OCPs) in Surface Sediments from the Bizerte Lagoon, Tunisia. Environmental Science and Pollution Research, 21, 6290-6302.
https://doi.org/10.1007/s11356-013-1709-7
[40] Ameur, W.B., Trabelsi, S., El Bedoui, B. and Driss, M.R. (2011) Polychlorinated Biphenyls in Sediments from Ghar El Melh Lagoon, Tunisia. Bulletin of Environmental Contamination and Toxicology, 86, 539-544.
https://doi.org/10.1007/s00128-011-0258-3
[41] Gil, O. and Vale, C. (2001) Evidence for Polychlorinated Biphenyls Dechlorination in the Sediments of Sado Estuary, Portugal. Marine Pollution Bulletin, 42, 452-460.
https://doi.org/10.1016/S0025-326X(00)00184-3
[42] Cui, X., Dong, J., Huang, Z., Liu, C., Qiao, X., Wang, X., et al. (2020) Polychlorinated Biphenyls in the Drinking Water Source of the Yangtze River: Characteristics and Risk Assessment. Environmental Sciences Europe, 32, Article No. 29.
https://doi.org/10.1186/s12302-020-00309-6
[43] Barreira, L., Bebianno, M., Mudge, S., Ferreira, A., Albino, C. and Veriato, L. (2005) Relationship between PCBs in Suspended and Settled Sediments from a Coastal Lagoon. Ciencias Marinas, 31, 179-195.
https://doi.org/10.7773/cm.v31i12.97
[44] Habibullah-Al-Mamun, M., Kawser Ahmed, M., Saiful Islam, M., Tokumura, M. and Masunaga, S. (2019) Occurrence, Distribution and Possible Sources of Polychlorinated Biphenyls (PCBs) in the Surface Water from the Bay of Bengal Coast of Bangladesh. Ecotoxicology and Environmental Safety, 167, 450-458.
https://doi.org/10.1016/j.ecoenv.2018.10.052
[45] Eker Sanli, G. and Tasdemir, Y. (2022) Temporal Variations of PCBs and Their Estimated Air-Soil Exchange Fluxes Measured in Seven Sites in Bursa-Turkey. Soil and Sediment Contamination: An International Journal, 32, 843-861.
https://doi.org/10.1080/15320383.2022.2143480
[46] Livingstone, D.J., Manallack, D.T. and Tetko, I.V. (1997) Data Modelling with Neural Networks: Advantages and Limitations. Journal of Computer-Aided Molecular Design, 11, 135-142.
https://doi.org/10.1023/a:1008074223811
[47] Jönsson, A., Gustafsson, Ö., Axelman, J. and Sundberg, H. (2002) Global Accounting of PCBs in the Continental Shelf Sediments. Environmental Science & Technology, 37, 245-255.
https://doi.org/10.1021/es0201404
[48] Cheng, Z., Lin, T., Xu, W., Xu, Y., Li, J., Luo, C., et al. (2015) A Preliminary Assessment of Polychlorinated Biphenyls and Polybrominated Diphenyl Ethers in Deep-Sea Sediments from the Indian Ocean. Marine Pollution Bulletin, 94, 323-328.
https://doi.org/10.1016/j.marpolbul.2015.01.027

Copyright © 2025 by authors and Scientific Research Publishing Inc.

Creative Commons License

This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License.