3.1. Heavy-Metal Concentration in Fish and Shrimp
The studied fish and shrimp specimens included a wide range of heavy-metal concentrations (
Table 2). The loads of six selected heavy metals (Pb, Cd, Cr, Ni, Cu, and Zn) followed the decreasing order of Zn (40.8 ± 9.7 μg/g) > Cu (17.8 ± 7.1 μg/g) > Pb (6.2 ± 1.8 μg/g) > Ni (0.4 ± 0.3 μg/g) > Cd (0.06 ± 0.02 μg/g). The concentration of Cr was below the detection level in all of the experimented samples. Among the metals, Pb concentrations exceeded the recommended guidelines. The findings, however, revealed significant variations in the four measured metal contents among the four species at a 99% confidence interval (
p < 0.01). It has been demonstrated through experimentation that several biotic and abiotic factors, such as the fish habitat, eating behavior, age, sex, body weight, physiological state, water temperature, pH, etc., can affect the variation in heavy metal accumulation in fish. Benthic or bottom-dwelling fish typically accumulate more metals than their surface-dwelling counterparts because they take up metals from both water and sediments [
12,
20,
21,
22]. The level of heavy-metal contamination increases with fish size and age and is higher in females than in males and higher in predators than in herbivores [
22]. Once more, the accumulations are lowest in the muscle and are highest in the liver and gills. The levels also vary depending on the study area because higher metal levels have been seen in regions with growing populations, industries, traffic, and agricultural activities.
The findings of this study were compared to earlier research on the presence of heavy metals in marine and estuarine fish and shellfish from different regions of Bangladesh, as shown in
Table 2. In this study, the average metal level was high in
P. monodon, followed by
S. panijus,
P. fuscus, and
L. bata. The maximum loads of Zn (49.98 µg/g), Cu (42.33 µg/g), and Pb (7.24 µg/g) were recorded in
P. monodon, whereas Ni (0.81 µg/g) and Cd (0.08 µg/g) were higher in
L. bata and
P. fuscus, respectively. The highest levels of metals were recorded in omnivores (
P. monodon) and carnivores
(P. fuscus), and the lowest levels were in herbivores
(L. bata). Researchers have discovered that the feeding habits of fish have a significant impact on the amount of metal they absorb. For instance, carnivorous fish have higher metal concentrations than herbivores and omnivores. In addition, the fish habitat may potentially be a factor in the metal concentrations in fish. Hossain et al. [
26] documented higher metal concentrations in benthic fish than in demersal fish. The variation in metal loads in the fish and shellfish of this study might be due to the different ecological needs, metabolism, and feeding patterns of the examined fishes [
27,
28,
29]. The toxicity or bioavailability of metals is influenced by physical–chemical factors, chemical species, and the living things themselves. In a slightly acidic environment (pH = 5.0), metal solubilities were higher, and they sharply rose when the pH was below 4. The toxicity of metals is also said to rise with rising temperatures. The rate of oxygen consumption rose by about 34% with a temperature increase of 4 degrees Celsius (from 20 to 24 degrees C), which enhanced copper toxicity by around 7% [
28].
From the comparison in
Table 2, it was observed that the examined fish (
L. bata,
P. fuscus, and
S. panijus) had higher concentrations of Pb than other fish in Bangladesh [
12,
20,
21,
22]. Although Sultana et al. [
1] recorded higher Pb in
P. monodon than the present study, Pb was found to be lower in other shellfish, e.g.,
Macrobrachium rosenbergii,
Metapenaeus dobsoni,
P. sculptilis, and
P. versicolor [
5,
22]. However, Pb, in all of the examined fish and shrimp of this study, exceeded the guideline for Bangladesh (0.3 µg/g for fish and 0.5 µg/g for shellfish) [
24]. Pb is responsible for many health problems, such as neurotoxicity, nephrotoxicity, etc. [
1]. However,
P. monodon is a bottom-dwelling shellfish and grazes on sediments. Therefore, sediments could be a major source of Pb in fish and shellfish [
18].
The content of Cd in fish and shellfish muscle in this study was similar to the findings of Tahity et al. [
12] and Sultana et al. [
1]. In contrast, the Cd load in the present studied species was found to be lower than the findings of Islam et al. [
21] and Baki et al. [
5]. However, the load of Cd in all of the species did not cross the tolerance level or guidelines for Bangladesh (
Table 1). Cd is a chronic-toxicity-producing metal that naturally exists in the ecosystem in a very low quantity. Aquatic ecosystems may have Cd from industrial waste, smelting or electroplating, and the use of fertilizer in agricultural fields. Excess Cd may cause renal failure and the softening of bones as a result of long-term or high-dose-contamination exposure, and prostate cancer may develop in response to high levels of Cd [
29,
30].
In our study, the concentration of Zn was higher than those of other metals in the examined fish and shellfish muscle (
Table 2). However, the Zn concentration in this study was lower than the findings of Hossain et al. [
20]. Further, this did not exceed the permissible limit in fish and shellfish muscle. Zn is an essential micronutrient for living organisms and acts as a catalyst for about 300 enzymes in aquatic creatures. Therefore, a relatively high level of Zn is required to balance certain biological functions. Zn is involved in most metabolic pathways in humans, and its insufficiency can lead to loss of appetite, the inhibition of growth, skin changes, and immunological abnormalities [
31].
In this research, the Cu level in fish and muscle exceeded the guideline value for Bangladesh. In addition, the values of Cu were lower than the findings of Hossain et al. [
20]. Due to its presence in several enzymes and its necessity for hemoglobin synthesis, it is known as an essential element [
32]. High levels of copper (Cu) can cause acute toxicity. In this study, the highest level of Cu was found in the body of
Penaeus monodon (42.33 µg/g), and the lowest concentration was found in the body of
Labeo bata (2.93 µg/g). However, Cu in the aquatic environment can originate from dyeing and tanning industries, textiles, photography, paints and inks, the battery industry, and surface runoff from upstream agricultural fields [
14,
33,
34]. The concentration of Ni in the environment is very low, and it causes a variety of adverse health effects, such as lung inflammation, fibrosis, emphysema, and tumors. The highest concentration of Ni was found in the body of
Labeo bata (0.81 mg/kg), and the lowest concentration was reported in the body of
Sillaginopsis penigus (0.02 µg/g).
3.3. Source Identification
Strong correlations between specific heavy metals may indicate similar types of pollution and/or release from the same source of contamination, mutual influence, and equivalent behavior when being transported to the aquatic system [
1,
12,
16,
22]. A correlation matrix among the metals from fish species is presented in
Table 5. The correlation matrix analyzes whether there are any interrelationships among the elements. The correlation can be positive or negative and significant or highly significant (
Table 5). The presence of a correlation between metals indicates that they have originated from similar sources, and no correlation implies different sources or diverse sources. A negative correlation was observed between Ni and Pb (
r = −0.816); positive correlations were found between Cu and Pb (
r = 0.623), Zn and Cu (
r = −0.871), and Zn and Cu (
r = 0.498); and very weak or no correlations were calculated between Pb and Cd, Ni and Zn, and Pb and Zn. The correlations among Ni, Pb, Zn, and Cu indicate they originated from anthropogenic sources. The study area and its catchment are mainly used for agriculture activities, poultry farming, fish farming, small industries such as food processing and packaging, still workshops, ship and boat building, oil refineries, cement industries, electroplating, etc. (Islam et al., 2018 [
14]; Tahity et al., 2022 [
12]). Cd and Ni are generally used in nickel–cadmium batteries, foils, alloys, and electroplating (Islam et. al., 2018 [
14]). The battery industry employs Cd and generates and discharges wastes with high Cd levels. Pb, on the other hand, has a long history of being employed as an anti-corrosive element in paint for steel mills and boats. Pb is also released via engine oil leakages from boats and ships. The burning of fuels also possibly contributes to the increased level of Pb. Although naturally available, advective transport through the river and atmospheric deposition might have contributed to the levels of Cu and Zn.
PCA identifies a compressed set of factors that affect the variance in an entire data set [
36]. Hence, PCA was performed for the qualitative analysis of the clustering behavior [
37]. The results included the corresponding loading values, along with the particular eigenvalues, the variance in 100% of the data, and the cumulative variance for each factor, and two elements evolved with the eigenvalues, which were greater than 1 (
Table 6).
Figure 3 and
Table 2 show the PCA outcomes with two grouping factors and a variance of 89.66%; the results show that PC1 provided the highest eigenvalue, 2.36, which was dominant, and a highly significant factor group contributing 47.26% of the total variance with the loading of Zn (
r = 0.533). The findings specified that industrial and agricultural chemicals were probable sources [
38,
39]. Meanwhile, PC2 contributed to 42.40% of the total variance in the loads of Pb (
r = 0.645) and Cu (0.535). The source of Pb and Cu might be urban or industrial wastes and other non-geogenic activities and electroplating actions, respectively [
22].
Cluster analysis is mostly used for the presentation of similar groups of the sampling site that have special variability. Similar groups of sites are presented as a group of clusters, and dissimilar sites are plotted in another group of clusters to identify specific areas in terms of contamination [
39,
40,
41]. A hierarchical agglomerative cluster analysis was performed on the normalized data set using Ward’s method with Euclidean distance as a measurement of similarity (
Figure 4). The cluster analysis resulted in a dendrogram with a significant cluster at (Dlink/Dmax) × 100 = 20. Two major groups were identified: Pb, Cu, and Zn in cluster 1 and Cd and Ni in cluster 2, demonstrating their close association and possibly indicating the same sources of origin.