Macroalgae in biomonitoring of metal pollution in the Bay of Bengal coastal waters of Cox’s Bazar and surrounding areas

Although coastal water marine algae have been popularly used by others as indicators of heavy metal pollution, data within the Bay of Bengal for the estuarine Cox’s Bazar region and Saint Martin’s Island has remained scarce. Using marine algae, the study herein forms an effort in biomonitoring of metal contamination in the aforementioned Bangladesh areas. A total of 10 seaweed species were collected, including edible varieties, analyzed for metal levels through the use of the technique of EDXRF. From greatest to least, measured mean metal concentrations in descending order have been found to be K > Fe > Zr > Br > Sr > Zn > Mn > Rb > Cu > As > Pb > Cr > Co. Potential toxic heavy metals such as Pb, As, and Cr appear at lower concentration values compared to that found for essential mineral elements. However, the presence of Pb in Sargassum oligocystum species has been observed to exceed the maximum international guidance level. Given that some of the algae species are cultivated for human consumption, the non-carcinogenic and carcinogenic indices were calculated, shown to be slightly lower than the maxima recommended by the international organizations. Overall, the present results are consistent with literature data suggesting that heavy metal macroalgae biomonitoring may be species-specific. To the best of our knowledge, this study represents the first comprehensive macroalgae biomonitoring study of metal contamination from the coastal waters of Cox’s Bazar and beyond.

www.nature.com/scientificreports/ making the location highly attractive, typically with millions of tourists visiting the area annually, with multiple businesses associated with the tourism industry present along the coastline area 34 .
Over the period January 2018 to April 2019, ten naturally growing seaweed species samples were collected (Table 1), each in triplicate, obtained from different sites in the Cox's Bazar and Saint Martin region of Chittagong (Fig. 1). Species and taxonomy identification followed standard morphological features, size, shape, color, etc. The samples were thoroughly rinsed to remove adhered sediments and other substances. To mitigate against the  Table 1. Identity and labeling of the 10 seaweed species collected. "R", "B", and "G" indicate red, brown, and green seaweed respectively. www.nature.com/scientificreports/ effects of natural decay and consequent influence on metals analysis, the samples in labeled plastic bags were quickly transported to the lab for subsequent processing.
Sample preparation. The procedure described by Jolly et al. was adopted in preparing the samples for analysis, the latter via the technique of Energy Dispersive X-ray Fluorescence (EDXRF), in the Atmospheric and Environmental Chemistry Laboratory of the Atomic Energy Centre, Dhaka 35 . In specific terms, the preparation procedures consisted of cutting seaweed samples into small pieces, washed with tap water, and subsequently rinsed with deionized water, to then be left to dry at room temperature, followed by placement in an oven at 60 °C until a constant dry weight was obtained. The dried seaweeds were then ground to obtain a fine homogenous powder using an agate mortar and pestle, with 0.1 g masses being pelletized using a hydraulic press pellet maker model (Specac Ltd., UK) by applying a pressure of approximately 3 tons. The dimension of the prepared pellets was 7 mm in diameter and 1 mm thick. This procedure was carried out for each triplicate of the 10 seaweed species. The pellets were then stored in clean glass Petri dishes, then held in a vacuum desiccator for subsequent measurement of the concentrations of the various metals of concern.
Sample analysis and method validation of the level of concern metals. The elemental concentration was analyzed by energy-dispersive X-ray fluorescence (EDXRF) spectroscopy, as described in our previous study 36 . In brief, a 109 Cd point source with an X-ray beam (at 22.4 keV) was applied to excite the prepared samples and produce characteristic X-ray detected by the Si(Li) detector (Canberra™) which has a resolution of 175 eV at 5.9 keV, amplified by the spectroscopy amplifier and processed by the multichannel analyzer MCA (6 K + channel). All the peak areas were integrated by AXIL and PRO/QXAS software provided by the International Atomic Energy Agency (IAEA), Vienna, Austria.
With EDXRF a direct comparison method for elemental concentration measurement, calibration curves must be constructed based on similar matrices. Improving the sensitivity of readings and nullifying the matrix effects, the calibration curves were constructed via the use of the commercially available standard reference material (SRM) Orchard leaf/NIST 1571. The average peak areas of the EDXRF irradiated SRM pellets (at least three pellets) prepared using a similar configuration to that of the algae samples were then plotted in terms of the presence of the elements as a function of atomic number 36 . Validation of the calibration curve constructed for elements present in the standards was performed via analysis of another standard reference material, Spinach/ NIST 1570a, again prepared in the same way as the algae sample 35 . All the obtained values were similar to certified values, the percentage relative error in evaluated elements being < 10%, assuring validation of the method. Comparison between the experimental and certified values is provided in Table 2.
Samples were positioned in the EDXRF spectrometer according to the defined geometry and then excited using a 109 Cd point source, providing 22.4 keV X-rays. To obtain good counting statistics, each sample was irradiated for a sufficient duration, ranging from 2000 to 5000 s. The collected spectra were analyzed using the aforementioned software, obtained from the IAEA. In this study, the standard addition method was used to obtain the metal concentration. The method involves the addition of known quantities of various analytes to the specimen. This method requires a linear calibration throughout the range of addition of various analytes. To determine each of the elements in the obtained spectrum, use was made of the calibration curves, also acknowledging an absence of inter-elemental effects; determination of metal concentrations in the analyzed algae sample was made via Eq. (1) as follows: with I i the characteristic X-ray net intensity (in cps), C i the metal concentration (in μg.g −1 ), S i the sensitivity of each analyzed element i (cps·g −1 ·cm 2 ), and A the absorption factor, equal to 1 for the samples, prepared in a thin-film geometry.
Determination of the limit of detection. The Minimum Detection Limit (MDL) depends on the counting statistics of the measurement and is a statistical process 37 . The MDL is obtained from the ratio of the amount of an element (in ppm) yielding an X-ray intensity equal to 3σ of the background under the peak in an interval (1) www.nature.com/scientificreports/ equal to the full width at half maximum (FWHM) of the peak and the concentration of the corresponding elements determined by using the calibration procedure 38 , and calculated using relation (2): Channel × FWHM of the relevant element. The calculated MDL for the analyzed elements is displayed in Table 3.

Metal pollution index (MPI).
The heavy metal burden of the 10 seaweed species was estimated on the basis of a metal pollution index (MPI), calculated using the following formula 39 : with M n the mean concentration of heavy metal n (mg kg −1 dry weight). The metals included in the analyses were As, Pb, and Cr.
Health risk assessment. Non-carcinogenic risk. A health risk assessment for an average adult was conducted following Ref. 28 , based on the method of the US Enviromental Protection Agency (EPA) 40 . The targeted hazard quotient (THQ) and hazard index (HI) was calculated based on the exposed dose (ED), the latter calculated using the following Eq. (4): with C i the mean concentration of heavy metals in seaweed (mg kg −1 ), D i the daily seaweed intake (5.2 g person -1 day −1 ), E d the average duration of exposure (70 years), B w the average body weight of the consumer (70 kg), and A t the lifetime of the consumer (70 years), the values being from recognized reference data.
The THQ, characterizing the non-carcinogenic risk to an exposed individual, is defined as the ratio of the exposed dose of a particular metal to the corresponding reference dose (R f D) and can be determined by the following Eq. (5): where R f D is the recommended oral reference dose for certain metals. Lastly, hazard index (HI) is calculated following Eq. (6): Following the US EPA 40 guidelines, a HI < 1 is regarded to offer no potential health risk. where CSF is the cancer slope factor. The CSF values of carcinogenic metals are displayed in Table 4. The total cancer risk (CR t ) was determined as the sum of the CR from the studied heavy metals 43 , as displayed in Eq. (8): Given that seaweed generally contains 80-90% water 44 , a median value of 85% was used for the calculations. An ordinary one-way analysis of the variance (ANOVA) was conducted to compare the concentrations of the 13 observed elements among the 10 seaweed species, followed by Tukey's multiple comparisons test. The same analyses were used to compare the concentration of the elements among types (Chlorophyta, Rhodophyta, and Phaeophyceae) by grouping seaweed species of the same color. Statistical significance was set to 0.05. All the analyses and graphs were performed in GraphPad Prism (version 8.4.3 for Windows).

Results and discussion
A total of 13 elements were determined and validated by the EDXRF technique for the 10 macroalgae species. Overall, the mean concentration of trace elements and heavy metals in the 10 species decreased in the descending order K > Fe > Zr > Br > Sr > Zn > Mn > Rb > Cu > As > Pb > Cr > Co ( Table 5). The mean elemental concentration and descriptive statistics per macroalgae species are displayed in Table S1. As expected, K provides the greatest concentration among the 10 species (mean of 4.2 × 10 4 mg kg −1 ), being significantly greater than the second most prevalent element (Fe, with a mean of 1.9 × 10 3 mg kg −1 ). Previous studies have shown the prevalence of K to be in the same order. For instance, in Ulva spp. (Ireland), Chondrus crispus (Denmark), and Fucus spiralis (Spain), the mean K respective concentrations were 1.2 × 10 4 , 3.3 × 10 4 , and 4.0 × 10 4 mg kg −145-47 . Seaweeds naturally contain high K content, at of the order of 2% of their dry weight 48,49 , depending on the species and environmental conditions. Co concentrations have been found to be the lowest in value, at a mean of 0.26 mg kg −1 (ranging from 0.18 to 0.33 mg kg −1 across the species); lower Co content has been reported in commercial seaweed of Asian origin (mean of 0.10 mg kg −1 ) and European origin (mean of 0.03 mg kg −1 ) 50 , constituting a naturally occurring micronutrient in seaweed 51 . Previous studies in aqueous solutions have demonstrated the leaching of elements such as Ca and Mg from grasses/algae, directly measured via inductively coupled plasma mass spectrometry (ICP-MS) 52 . In the present study, macroalgae samples were not analyzed in aqueous solutions, thus leaching and loss of metallic elements from the samples being unlikely to have occurred.
One-way ANOVA analysis shows a significant difference in the concentration of elements among macroalgae species (p = < 0.05), except for Co (p = 0.3669) (Fig. 2a). Regarding the determined hazardous heavy metals, Pb, As, and Cr; the highest mean concentrations of Pb were found in B6 (10.63 mg kg −1 ), followed by R4 (4.50 mg kg −1 ), and B5 (4.24 mg kg −1 ) (Fig. 2b). For arsenic, two species, B5 (11.89 mg kg −1 ) and B6  www.nature.com/scientificreports/ (10.75 mg kg −1 ) had significantly higher concentrations than the rest (Fig. 2c). Lastly, R4 (3.64 mg kg −1 ), R3 (1.94 mg kg −1 ), and B6 (1.90 mg kg −1 ) presented the highest levels of Cr (Fig. 2d). Interestingly, macroalgae species from the same genus (Hypnea; R1 and R2) showed similar bioaccumulation of heavy metals. B5 and B7 (both genius Padina), showed considerable differences. This suggests that despite similar mechanisms of uptake, total nutrient/metal uptake may be species-specific for some seaweed genera. Despite being consumed as food or utilized as animal feed, there is no current legislation in Bangladesh that determines the maximum levels of heavy metals in seaweed. However, some international norms are available. Pb concentration in B6 surpassed the maximum levels in seaweeds (5 mg kg −1 ) recommended by the French High Council for Public Health 53 and The Center for the Study and Development of Algae (CEVA) 28 . Moreover, the maximum levels of Pb in leafy vegetables (which may be consumed at similar levels to that of seaweed), according to the FAO 54 , is much lower (at 0.3 mg kg −1 ). To the best of our knowledge, maximum As and Cr levels in seaweed have not been addressed in international regulations. It should be noted that Sargassum sp. along with four additional species in the present study (R2, R4, G8, and G9) have been cultured and consumed in fresh or dried form for decades in Bangladesh 32 , thus presenting a potential route for heavy metal exposure to the population. Nevertheless, seaweed cultivation in Cox's Bazar remains artisanal and undeveloped 31 .
The concentrations of the three potentially toxic heavy metals that have been considered herein, Cr, As, and Pb, have been found comparable or lower than literature data elsewhere (Table 6). For instance, the mean concentration of Pb in 12 algal species in China was 1.89 mg kg −1 (ranging from 0.77 to 4.21 mg kg −1 ) 24 . In Lebanon, concentrations were even lower, with a mean of 1.04 mg kg −155 . Cr and As were particularly high in Greece and China, with mean concentrations of 9.38 and 18.33 mg kg −1 respectively 1,24,56 . Conversely, much lower concentrations were reported from a market survey in Italy 50 . The mean concentrations of Cr, As, and Pb were 0.14, 1.42, and 0.13 mg kg −1 , respectively. This may be attributed to the high hygiene standards through the food supply chain and also during seaweed culture intended for human consumption. In markets from the Canary Islands (of Spain) reported Pb concentrations in seaweeds of Asian and European origin have ranged from 0.12 to 0.004 mg kg −1 and from < LOQ to 0.05 mg kg −1 , respectively 57 . The importation of potentially contaminated edibles as a source of heavy metal exposure has been discussed by others, including in regard to the consumption of seaweed. Particular examples include toxic industrial discharges in India, with toxic heavy metal pollution in areas of plant cultivation being a particular consequence 58 . A recent study carried out along the Palk Bay coast of southeast India has observed highly elevated Pb concentrations in many seaweed species, surpassing 10 mg kg −1 in many cases. The variation in concentrations was found to depend on the sampling season. The main sources of contamination have been linked to ship washing activities, seafood processing, domestic sewage, and effluent www.nature.com/scientificreports/ discharges. In the case of boat maintenance procedures, the application of antifouling paint has been noted, the particles of these containing metallic-based biocides that can detach from the marine coatings 59 . The concentrations of most of the remaining trace elements (Mn, Cu, Zn, Br, Rb, and Sr) have been found to be in the range ~ 5 to 75 mg kg −1 (Fig. 3). Additionally, the mean Fe ranged from 1487 mg kg −1 in P. tetrastromatica to 3352 mg kg −1 in H. pannosa. Most of these elements are micronutrients involved in natural algae metabolism 60 . The concentration of micronutrients found in the present study is mostly of the same order of magnitude as that reported by Roleda et al. 61 . Seaweeds in particular are considered a great source of Fe 62 , which could increase the proportion of absorbed iron in functional meals 63 . Sr is known to be related to cell wall polysaccharides found in some macroalgae, such as alginates in most Phaeophyceae, and generally exhibit low concentrations 64 . Although Rb is not widely studied, it is suggested to be related to the geochemistry of the coastal environments 65 . Algae are noted to use the bromine and chlorine present in the environment to biosynthesize halogenated secondary metabolites 66 , many of the halogenated compounds being found to be brominated.
Concerning seaweed type, most of the studied metals show insignificant differences except for Mn, Fe, As, Br, Zr, and Pb ( Table 7). The results of the Kruskal-Wallis test and literature suggest that the seaweed type may not be as significant as the species in determining the bioaccumulation of metals in the seaweed. For instance, Rubio et al. analyzed 20 metals, finding no significant differences between Phaeophyta and Rhodophyta for the majority of the studied metals, the exceptions being Cr, Cu, Fe, Li, Mn, Mo, Sr, V, and Zn 67 . In contrast, Filippini et al. performed a similar analysis with 21 metals comparing Phaeophyta, Rhodophyta, and Chlorophyta, finding significant differences in the majority of studied metals, the exceptions being for Pb, Hg, Mn, Co, Ti, and Sb 50 . Some comparative www.nature.com/scientificreports/ studies have pointed to Phaeophyta being the most efficient algae in accumulating metals 68 , while others have pointed to Rhodophyta 24,50,69 . These results suggest that the influence of algae type in metal bioaccumulation may be limited, while some specific characteristics such as surface area and growth rates may be more important 70 . As apparent in Table 6, the specific macroalgae tissue evaluated in recent studies has not generally been specified 55,56 . Of note is that heavy metal and trace element concentrations have been observed to vary significantly in macrophytes from salt marshes, specifically between roots, shoots, and leaves 36 . Sáez et al. studied the bioaccumulation of metals in the thallus (blade, stipe, and holdfast) of the kelp Lessonia trabeculata 71 , finding metal-specific affinity in certain parts. A further variable not taken into account has been the life stage of the selected organisms. Future investigations should focus on determining the influence of the particular tissue and of seaweed age on the concentration of heavy metals. Algae-based heavy metal monitoring may require the use of different species that have better bioaccumulation affinity for different metals. For instance, the highest As concentrations in 12 macroalgae species from China were found in three species from the genus Sargassum 24 . Accordingly, in the present study, S. oligocystum was the species that exhibited the second highest As concentration (10.6 mg kg −1 ), slightly below P. tetrastromatica (11.9 mg kg −1 ), and significantly greater than the third-highest (C. racemose; 2.27 mg kg −1 ). Similarly, in the present study the highest Cr concentrations were found in species from the genus Gelidium, also as reported by Pan et al. 72 . The consistency of our results with previous studies supports the necessity of determining genius-specific macroalgae for heavy metal monitoring. Based on our results, we suggest the genus Gelidium for Cr, Padina for Cu and As, and Sargassum for Mn, As, and Pb in monitoring. Future studies may determine the uptake routes for different heavy metals in macroalgae based on their botanical characteristics.
The results of the carcinogenic and non-carcinogenic health risk assessment are displayed in Table 8, while MPI values of the 10 seaweed species are summarized in Fig. 4. The six metals, Cr, Pb, Cu, Zn, Co, and As, were selected for their potentially hazardous nature at relatively high concentrations and oral R f D data availability. The overall Hazard index was 0.993, which means the evaluated metals together may not pose a serious health risk to human health. However, Cd and Hg were not evaluated in the present study. These heavy metals could potentially contribute to a targeted hazard quotient sufficient to reach a HI > 1, thus representing a moderate to high risk for adverse human health effects. In the case of the carcinogenic risk, the CR t was calculated as 0.436.  However, the HI calculated for the Malaysian population considerably exceeds the health hazard limit (HI = 4.38), although that study presented a lower CR t than the present study (0.29) 5 . Although many studies concerning specific populations suggest heavy metal consumption from seaweed may not pose a sizeable health hazard 67,73 , in the present regard of Cox's Bazar the health authorities should look to monitoring heavy metal in consumable seaweeds, taking HI and CR t as indicators of their potential toxicity in areas of high seaweed consumption. In considering the calculated MPI of high toxicity heavy metals (Cr, Pb, and As) among the various seaweed species, it is suggested that B6 (S. oligocystum) and R4 (G. pusillum) may be the species that present the highest risk of heavy metal ingestion in Cox's Bazar. The toxicological implications of exposure to heavy metals are widely understood 74 . As an instance, lead poisoning generally links with anemia, affecting three enzymes associated with heme synthesis. In extremely high exposure scenarios, the neurological system can be critically affected 75 . Hexavalent chromium [Cr(VI)] and As are two well-established heavy metals giving rise to carcinogenic effects 76,77 . Despite the critical health effects associated with the exposure to some of the evaluated heavy metals, seaweed consumption by members of the public remains relatively low, likely insufficient to manifest in severe health effects. Of clinical cases concerning heavy metal tainted foodstuffs, these have mostly been attributed to the consumption of contaminated drinking water 78 .

Conclusions
Marine macroalgae are regarded as potential biomonitors of heavy metal and trace element pollution. In the present study, the concentrations of 13 elements were determined in 10 species of macroalgae from the coastal area of Cox's Bazar and Saint Martin's Island. The elemental bioaccumulation affinity per species and family type was investigated. Results indicate that some species may serve as better biomonitors than others for certain elements. Based on the agreement with the available literature, we suggest the genus Sargassum for Mn, As, and Pb monitoring, Gelidium for Cr, and Padina for Cu, and As. Additionally, since many of the species investigated in the present study are cultivated for human consumption, the relevant hazard indices were determined. The index value remains marginally below the limits recommended. However, since other important toxic heavy Table 8. Carcinogenic and non-carcinogenic health risk assessment from seaweed consumption in adults in Bangladesh. Oral R f D and CSF values were obtained from Kamuda et al. 80 , and Kortei et al. 42 , and Shams et al. 43 respectively.

Metal
Oral R f D Ci ED CSF THQ CR