2.1 Standards and reagents

This present research utilized analytical-grade potassium permanganate (KMNO₄) and tin (II) chloride dehydrate (SnCl₂·2H₂O) procured from Merck (Germany). As for the standard compounds (≥99.5 % purity) for As, Pb, and Cd (in 2 % HNO₃) as well as the total Hg (in 10 % HNO₃) and chloramphenicol (≥98 %), they were purchased from Sigma Aldrich (USA). Using deionised water from Sartorius arium® pro ultrapure water system (Gottingen, Germany), the standards and sample solutions were prepared. For performing the sample digestion process, HNO₃ (65 %), HCL (37 %), and H₂O₂ (30 %), purchased from QRëC (Malaysia) were used.

2.2 Study areas and species identification

Shrimp samples were collected from three different stages of supply chain in Johor Bahru, Johor, Malaysia involving seafood landing sites (Mersing and Pontian), major markets (A and B), as well as selected retail grocers. The grocers were situated within 2 km radius from the two aforementioned major markets (4 grocers for each surrounding area). The Global Positioning System (GPS) locations for all the sampling sites are provided in Table 1.

While P. silasi was the species identified at seafood landing sites Mersing and Pontian, L. vannamei was the one available at the major markets A and B as well as the selected surrounding retail grocers. Morphological taxonomic identification of the shrimps was performed by a fishery expert from Universiti Malaysia Terengganu, followed by comparing the DNA sequences of Cytochrome oxidase 1 gene with that of the GenBank database (https://blast.ncbi.nlm.nih.gov). The molecular identification was performed by a certified DNA analyst at Research Laboratory Division, Institute of Tropical Aquaculture and Fisheries (AKUATROP), Universiti Malaysia Terengganu.

2.3 Sampling

Each sample (300 g of pooled shrimps) was purchased from all the landing sites and markets over a period of six sampling intervals (September, October and December 2020 as well as March, April and May 2021). Each collected sample was transported to the laboratory in a cold box filled with ice cubes, and immediately kept at −20 °C prior to analysis. Taking into account the better homogeneity as well as the readiness of guidelines for inferring toxicity (Mahat et al., 2018), the dry weight of the shrimp sample was used in this study. The sample was thawed at room temperature (about 26 °C) on a clean tissue paper, followed by removing their exoskeletons. For enabling fast drying process, the flesh was cut into small pieces (about 2 cm each) and dried in an oven at 60 °C until the constant dry weight was achieved. Next, the sample was pulverized into a fine powdery form using kitchen grinder and stored in an airtight plastic container at room temperature (about 26 °C) until further analysis.

2.4 Analyses of metal contaminants

The chemical analyses were conducted in a certified (MS ISO/IEC 17025:2017) commercial laboratory at Universiti Teknologi Malaysia, Johor Bahru. Analysis of metal contaminants was performed using the laboratory in-house methods (TM/UTM/CHEM/SF), adapted from the AOAC 999.10 standard methods, Association of Official Analytical Collaboration, AOAC (2008). Prior to chemical analysis using a graphite furnace-atomic absorption spectrometer (GF-AAS) and flow injection mercury/hydride system (FI-MHS), the sample was digested using a microwave digester (Titan-MPS, Perkin Elmer). About 0.5 g of pulverized and dried sample of shrimps was weighed directly into a digestion vessel, followed by the addition of 5.0 mL of concentrated HNO₃ 65 % (w/w) and 2 mL of hydrogen peroxide (30 %). The sealed vessel was then placed in a microwave digestion system. Upon completion, the obtained solution was transferred into a volumetric flask (25 mL) and further diluted with deionized water to a final total volume of 25 mL.

An atomic absorption spectrometer (Analyst PinAAcle 900 T, Perkin-Elmer) equipped with a graphite furnace and an auto sampler (GF-AAS) was used for quantitating the respective concentrations of As, Pb, and Cd, while flow injection mercury/hydride system (FI-MHS) (FIAS 100, Perkin Elmer) was utilized for the total Hg. The argon gas flow was maintained at 250 L/min with the sample volume of 20 µL, the temperature programming of the furnace started with the drying process (about 30 s at 120 °C) and increased to 160 °C after 30 s. The pyrolysis was achieved at 1200 °C for 20 s. Electrode-less discharge lamp was used with 0.7 nm slit width and the observed wavelengths were: 283.3 nm (Pb), 228.8 nm (Cd) and 193.7 nm (As). The analytical conditions of FI-MHS were as follows: argon flow 90 mL/min at 20 s integration time, wavelength of 253.7 nm and a slid width of 0.7 nm.

The working standards for As and Pb were 10–100 µg/L, while Cd and total Hg were 1–4 µg/L and 1–10 μg/L, respectively. A 3 % HCl solution was used as the diluent for total Hg standard. For preparing the reducing agent (1.1 % SnCl₂), 11 g of SnCl₂·2H₂O was mixed with 10 mL deionized water and 3 mL of concentrated HCl, prior to leaving it to stand for 10 min. The solution was then transferred into a 100 mL volumetric flask and further diluted with deionized water up to the calibration mark. For preparing the carrier solution for the FI-MHS analysis, 30 mL of concentrated HCl was diluted with deionized water (up to the mark) in a 1 L volumetric flask. The KMNO₄ solution (5 %) was prepared by diluting 1.25 g of KMNO₄ with deionized water in a 25 mL volumetric flask. To stabilize the solutions for calibration standard of Hg and blank, four drops of % KMNO₄ solution were added.

2.5 Analysis of chloramphenicol

Determination of chloramphenicol for all samples was conducted by a certified (MS ISO/IEC 17025:2017) commercial laboratory in Johor Bahru using an in-house developed Direct Competitive enzyme-linked immunosorbent assay (ELISA) method (MY/STP/119). The obtained results were then compared among the different supply chain stages.

2.6 Statistical analysis

The data were analyzed using the UTM licensed IBM Statistical Package for Social Sciences (SPSS) version 27.0. Since the data violated the assumption of normality, inferential statistic was done using the non-parametric Kruskal-Wallis H and Mann-Whittney U tests. Mann-Whitney U test was used individually for comparing the concentrations of metal contaminants in shrimps sampled between the two landing sites (Mersing and Pontian) as well as the two major markets A and B. As for comparing the concentrations of metal contaminants in shrimps sampled among the major markets and their surrounding grocers (2 km radius), Kruskal-Wallis H was used. Since no significant differences were observed between the major markets and their surrounding grocers, pairwise comparison using Mann-Whittney U test was not attempted. However, pairwise comparison using Mann-Whittney U test was performed following the statistically significant differences detected by Kruskal-Wallis H while comparing the concentrations of the metal contaminants across the six different sampling intervals. Level of significance of 0.05 (α < 0.05) was utilized for determining the significant differences among groups.

2.7 Estimation of daily intake of metal contaminants and health risk assessment

Using Equation (1) as suggested by Song et al. (2009), the Estimated Daily Intake (EDI) values for the selected metal contaminants (i.e. As, Pb, Cd and total Hg) in shrimp samples evaluated here were compared with the provisional tolerable daily intakes (PTDI) suggested by the relevant authorities.

While the EDI refers to an adult daily intake of a metal contaminant via consumption of L. vannamei/P. silasi (in μg/g/day), C_metal is the concentration of a metal contaminant analyzed in such shrimps (in μg/g, wet weight), W_shrimps being the daily consumption of L. vannamei/P. silasi (in g/day), and B_w represents the body weight of an adult (in kg). W_shrimps was estimated from the data reported by Ministry of Health (MOH, 2008) i.e. 3 g/day and 9 g/day for average consumers (AC) and high consumers (HC), respectively. The average body weight of an adult for Malaysians B_w was reported to be 63 kg (Dee et al., 2019).The results obtained here were used for assessing the risk of metal contaminants through the consumption of L. vannamei/P. silasi. Therefore, the target hazard was calculated based on the US EPA Region III risk-based concentration table (US EPA, 2020) (Eq. (2)):

The exposure frequency (EF_r) is 350 days/ year, ED, W_shrimps and C_i^metal are the exposure duration (70 years), daily consumption of L. vannamei/P. silasi (in g/day) and the individual concentration of metal contaminants in L. vannamei/P. silasi (in μg/g, wet weight), respectively. In addition, RfD refers to the oral reference dose (in μg/g/day) (US EPA, 2020), Bw is the average body weight of a Malaysian adult (63 kg) and AT_n being the average exposure time for no carcinogens (ED × 365 days/year).

Because the RfD value for Pb remains unreported, there is no proof of a limit below which Pb may be permitted as non-harmful (U.S. EPA, 2004). Therefore, the equation recommended by Jović and Stanković, (2014), was used to calculate THQ for Pb as detailed below:

While THQ_i refers to the target hazard quotient of an individual metal contaminant, HI is the total hazard index for all the four metal contaminants analyzed here viz. As, Pb, Cd, and total Hg with the n value of 4.

3.1 Method validation

Table 2 (a) represents the analytical figures of merit (calibration curve, LOD and LLOQ) for the analysis of As, Pb, Cd and total Hg in L. vannamei and P. silasi samples. The calibration curves for all the four metal contaminants had the R² values of at least 0.995 with the RSD of<10 %, and hence complied fully with the prescribed criteria for accepting a calibration curve, as suggested by US FDA (2018). The LOD (signal/noise ratio 3:1) and LLOQ (signal/noise ratio 5:1) for all the analytes were calculated using the data from accepted calibration curves, ranging from 0.69 to 10.20 µg/kg and 1.04–15.44 µg/kg, respectively. From the three days of validation period, an average of R² ≥ 0.995 for the calibration curves was attained over the ranges of 10–100 µg/kg for As and Pb, 1–10 µg/kg for total Hg, and 1–4 µg/kg for Cd (Table 2 a). Taking into account that the concentrations of metal contaminants in shrimps sampled from the various sampling sites during the six sampling intervals would vary substantially, the wide linear ranges of analytical method were considered.

In this research, precision was expressed as percent coefficient of variation and accuracy as the percent difference from the nominal values, and the data are presented in Table 2 (b). It was observed that the intra-day precision (0.09–2.37 %) and accuracy (0.15–10.50 %) for the four metal contaminants were<10 %, except for the upper limit calibrator of 5 µg/kg for total Hg (10.50 %). As for the inter-day precision (0.99–5.70 %) and accuracy (0.98–9.40 %) for the metal contaminants, the values were<10 %. Considering the indications made by the US FDA, (2018) that for accepting the precision and accuracy the values must not exceed 20 % for bioanalytical validation, the values reported here fell well within the prescribed range. Moreover, the validation results also revealed that the percentage recoveries for all the four metal contaminants ranged between 89.10 and 111.30 %, and therefore, complied fully with the range of percentage recoveries (80 – 120 %) prescribed by the prevailing guideline AOAC (2016). Having considered that all the data of analytical figures of merit and validation parameters complied with the prescribed guidelines, it can be construed that the analytical methods utilized here appear robust, accurate and reliable for analyzing the metal contaminants in L. vannamei and P. silasi shrimps.

3.2 Analysis of metal contaminants

This present research investigated the concentrations of metal contaminants (viz. As, Pb, Cd and total Hg) in shrimps sampled from the two major landing sites in Johor (viz. Mersing and Pontian), as well as the two major markets (A and B) and their surrounding grocers at six different sampling intervals in Johor Bahru, Johor, Malaysia. Table 3 represents the comparison of metal contaminant concentrations in shrimps between the Mersing and Pontian landing sites as well as the two major markets (A and B) in Johor Bahru. Since the data violated the assumption of normality, they are presented as median (range) and tested using the non-parametric Mann-Whitney U test.

Significantly higher median concentrations of As (52.64 mg/kg), Cd (0.21 mg/kg), and total Hg (0.50 mg/kg) were observed for shrimp samples from the landing site of Mersing when compared with that of Pontian (As: 32.72 mg/kg; Cd: 0.08 mg/kg; total Hg: 0.19 mg/kg) (P < 0.05). Considering the relative proximity of the landing site of Mersing and major market A (≈ 122 km) as well as Pontian and major market B (≈ 42.5 km), a suitable comparison of metal contaminant concentrations was made. It was observed that significantly higher concentrations of metal contaminants were present in shrimps sampled from both of the landing sites (P < 0.05) when compared with those of the major markets A (As: 0.46 mg/kg; Cd: 0.02 mg/kg; total Hg: 0.04 mg/kg) and B (As: 0.87 mg/kg; Cd: 0.02 mg/kg; total Hg: 0.03 mg/kg), correspondingly. Notwithstanding, statistical comparison between the concentrations of metal contaminants in shrimps sampled from major markets A and B did not reveal any significant differences (P > 0.05).

Using the non-parametric Kruskal-Wallis H test, comparisons between the concentrations of metal contaminants in shrimps sampled at the major markets A and B with their surrounding grocers (within 2 km radius from each major market) were made and the data are presented in Tables 4 and 5. Since no significant difference in the concentrations of metal contaminants in shrimps between the major market A and its surrounding four grocers was observed (Table 4, P > 0.05), pairwise comparison was not performed. The same finding also prevailed for the comparison made between major market B and its surrounding grocers (Table 5, P > 0.05). Because no significant differences were observed, the proposition that the shrimps available in the surrounding grocers may originate from their respective nearby major markets, appears to be supported.

For comparing the concentrations of As, Cd and total Hg across the six sampling intervals from all the sampling sites (Figs. 1–3), Kruskal-Wallis H with pairwise comparison using Mann Whittney U test was used. Results revealed significant differences in the concentrations of these three metal contaminants across the six sampling intervals (Figs. 1–3, P < 0.05). Considering that review of literature does not reveal any specific study focusing on the variations of metal contaminants in shrimps over varying sampling intervals, an exact comparison could not be made. However, a previous study on P. viridis mussels conducted in Malaysia (Mahat et al., 2018) reported significant differences in the concentrations of the same metal contaminants among the 3 months of their sampling period. The authors attributed such variations to the temporal changes at the harvesting site due to extensive industrial and seaport activities.

Close

Fig. 1

(a-d) Concentration of As in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of As were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

Close

Fig. 1

(a-d) Concentration of As in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of As were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

Close

Fig. 2

(a-d) Concentration of Cd in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of Cd were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

Close

Fig. 2

(a-d) Concentration of Cd in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of Cd were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

Close

Fig. 3

(a-d) Concentration of total Hg in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of total Hg were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

Close

Fig. 3

(a-d) Concentration of total Hg in shrimps sampled from (a) landing sites, (b) major markets A and B, as well as (c) grocers surrounding major markets A and (d) B during the six sampling intervals. Significant differences (P < 0.05) in concentrations of total Hg were observed for all the six sampling intervals from all the sampling sites.

Export to PPT

It is imperative to mention here that although the median concentrations for metal contaminants observed in samples from all the sampling sites may be lower than the MPPs (As, Pb and Cd: 1.00 mg/kg; total Hg: 0.50 mg/kg) prescribed by the Malaysian Food Act (1983) and Regulations (2021), the ranges of the individual concentrations were higher during several sampling intervals (Tables 3–5). Except for Cd, the ranges of As and total Hg expanded beyond the prescribed MPPs and therefore, investigating its possible detrimental health implications for consuming the shrimps is a matter of public health significance. A similar indication has also been made by previous researchers whom investigated the parallel issue following the ingestion of Perna viridis mussels in Johor Bahru, Johor, Malaysia (Mahat et al., 2018). In this context, many authors have attributed the presence of metal contaminants in the aquatic organisms to the natural and anthropogenic sources, leading to their bioaccumulation and biomagnification in the organisms (Mahat et al., 2018; Ali et al., 2021). Since the areas surrounding Pontian, parts of Malacca and Mersing are undergoing rapid development and land reclamation (Wang et al., 2019) as well as being the routes of shipment, contamination of metal contaminants in aquatic organisms like shrimps is within expectation. Moreover, seepage of minerals and metals from geographical processes (Briffa et al., 2020) may also add to the quantum of metal contaminants observed in the shrimps. As for Pb, it was non-detectable in any of the shrimp samples from all the sampling sites. Hence, further studies to explain this observation may be required.

An interesting fact discovered in this present research was that, significantly varying concentrations of metal contaminants were observed between the shrimps collected from the two landing sites (Mersing and Pontian) with those of major markets A and B as well as their surrounding grocers nearby Johor Bahru. Such a situation has triggered further exploration to ascertain if the shrimps are of the different species although they may look similar. The taxonomic and molecular identifications of the shrimps sampled from all the sampling sites revealed the presence of two different species namely P. silasi (false white shrimp) and L. vannamei (white shrimp). P. silasi was the only species of ‘white shrimps’ available during sampling at both the Mersing and Pontian landing sites. On the other hand, L. vannamei was the one found in major markets A and B as well as their surrounding grocers. As such, it can be construed that the shrimps harvested from the ocean nearby Mersing and Pontian may not be the one contributing to the supply chain of shrimps sold in major markets A and B and their surrounding grocers in Johor Bahru, as investigated during the course of six different sampling intervals of this present research.

Fascinatingly, a few studies in Malaysia concluded that the consumption of shrimps from local markets and the coastal areas in Peninsular Malaysia as safe (Anandkumar et al., 2017) with no non-carcinogenic risk (Yap et al., 2019) and ‘does not pose a health risk to the consumers’ (Rajan and Ishak, 2017). However, the conclusions made by these authors can be considered as pre-matured and erroneous, considering the limitations of their studies. The limitations are the exclusion of As and Hg from the analyses (Yap et al., 2019), insufficient description of the sampling process and locations of study (Rajan and Ishak, 2017), outdated data (collected in 2007 but published in 2019) (Yap et al., 2019), as well as the absence of analytical validation data (Rajan and Ishak, 2017). In view of these limitations and because medium to long-term study investigating all the four metal contaminants in shrimps at various supply chain stages in Malaysia remains unreported, this present research that factored all these pertinent aspects in P. silasi and L. vannamei shrimps sampled in Johor merits scientific considerations.

3.3 Health risk assessment

Considering that metal contaminants in shrimps may cause detrimental human health implications, even at concentrations lower than that of prescribed MPPs following chronic exposure, health risk assessment parameters (EDI, THQ and HI) were calculated. While Table S1 represents the EDI of As, Cd, and total Hg by consuming the shrimps sampled during six sampling intervals, Table 6 depicts the associated THQ and HI values, respectively. Results revealed high EDI values for As in shrimps sampled from landing sites Mersing (0.920–29.130 μg/kg/day) and Pontian (0.290–8.480 μg/kg/day) when compared with that of PTDI value (3.570 μg/kg/ day) suggested by JECFA (1999), JECFA (2003).

As for the other sampling sites, the EDI values for As and other metal contaminants were evidently lower than the suggested PTDI values (Table S1). The fact that previous health risk assessment studies on shrimps in Malaysia (Anandkumar et al., 2017; Lee et al., 2017; Rajan & Ishak, 2017; Anual et al., 2018; Yap et al., 2019) did not include As, suitable comparison with the EDI values of As reported here with the literature could not be attempted. The EDI values for Cd and total Hg observed in this present research ranged between not available to 0.079 μg/kg/ day and 0.002 to 0.163 μg/kg/ day, respectively (Table S1). While the EDI values for Cd was within the range of 0.05–0.15 ug/kg bw/day reported by previous researchers in Selangor (Rajan and Ishak, 2017), the EDI for total Hg (0.228 μg/kg/ day) can be made out as higher than the Estimated Weekly Intake of 0.368 μg/kg bw reported by Anual et al., (2018) in Perak and Selangor.

Except for shrimps consumed from grocer A2, the calculated THQs for As for HC exceeded 1.00 for all the sampling sites. The highest ranges of THQs for HC were observed in shrimps from the landing sites Mersing (2.92–92.15) and Pontian (0.90–27.23), followed by the remaining sampling sites of major markets and their surrounding grocers. On the other hand, the ranges of THQs for AC exceeded 1.00 only for shrimps from the landing sites Mersing (0.97–30.72) and Pontian (0.30–9.08) alone (Table 6). THQ refers to the ratio of consumed dose of metal contaminant via oral ingestion and can be useful for evaluating such intake of metal contaminant from contaminated food (Mahat et al., 2018; Ali et al., 2022). The fact that the ranges of THQ (especially for those of HC) were higher than 1.00, the possibility of human population exposed to such contaminated shrimp may ‘experience deleterious effects’ cannot be excluded.

Moreover, assessment of the HI value is important to consider the combinatory influence of multiple metal contaminants that occur in aquaculture products (like shrimps). The index would throw light at explaining the risk of having non-cancer health impacts following the ingestion of contaminated food (Mahat et al., 2018). Results revealed that the ranges of HI exceeded 1.00 in all the samples, except for grocer A2, and expectedly, the highest HI ranges were evident in shrimps sampled from the landing sites Mersing (3.00–92.22) and Pontian (0.94–27.43) (Table 6). Since the ranges of HI reported here exceeded the prescribed cut-off value of 1.00, the fact that populations exposed to such contaminated shrimps may be at risk for developing non-cancer health impacts is an alarming matter of public health consideration. Interestingly, there are several studies on shrimps in Malaysia that reported the THQ and HI values of<1.00, indicating that the shrimps are safe for human consumption (Rajan and Ishak, 2017; Yap et al., 2019). However, oddly these studies did not evaluate the complete list of metal contaminants in shrimp (viz. As, Pb, Cd and total Hg) as listed in the Malaysian Food Act (1983) and Regulations (2021). Hence, the inference that they made may be impractical for making any suitable public health decision.

3.4 Analysis of chloramphenicol

Analysis of chloramphenicol was performed by a certified private laboratory using the in-house method of ELISA with the LOD of 0.03 ppb. Results did not reveal any detectable amount of chloramphenicol in any of the samples of shrimps from the landing sites Mersing and Pontian (P. silasi) as well as major markets A and B and their surrounding grocers (L. vannamei). It is acknowledged that broad-spectrum antibiotics (like chloramphenicol) are illegally used in aquaculture (Suseno et al., 2016), and their prolonged exposure in human can be carcinogenic as well as deleterious to various body systems (Chen et al., 2020). Due to such unwanted adverse effects, many countries have banned aquaculture products that contain these antibiotics. Similarly, the Malaysian law has specified that chloramphenicol shall not be present in any seafood products mean for human consumption (Food Act (1983) and Regulations (2021)).

In this context it is pertinent to mention here that shrimp exporters from Malaysia have been put on the red list of the US FDA since 2018 due to detectable amount of chloramphenicol in the consignments (Malay Mail, 2020, January 4). Although the Malaysian fishery authority had responded by indicating the possibility of transshipment element that can be the reason for such contaminated consignments (DOF, 2020), no empirical proof was provided. Element of transshipment refers to importation of shrimps without the official surveillance of DOF and exporting them to other countries (DOF, 2020).To address such an important issue, adequate sampling involving the different stages of supply chain proves necessary. Therefore, the non-detectable results of chloramphenicol in this present study that sampled shrimps sold at various supply chain stages (landing sites, major markets, and surrounding grocers) in Johor Bahru appear to support the proposition of the existence of transshipment element, as advocated by the relevant authority.

Another important aspect to consider is the bacterial cross-resistance, attributable to the chemical interaction between the toxic heavy metals with antibiotics like chloramphenicol, suggesting the pivotal role of the interaction in environments where metal and antibiotic are co-existing (Chen et al. 2015). In this context, mutation in genes encoding antibiotic resistance induced by the presence of toxic heavy metals like arsenate (an arsenic species) could be the possible explanation (Knapp et al. 2011), and hence, further studies to elucidate this aspect proves necessary.