Evaluation of metal contamination in surface sediments and macroalgae in mangrove and port complex ecosystems on the Brazilian equatorial margin

This study evaluated metal contamination in surface sediments and macroalgae of mangroves and port complexes on the Brazilian equatorial margin. Samples were collected between August 2020 and February 2021 at seven points in a mangrove swamp under the influence of port activity and at two points without port activity. Metal concentrations in the macroalgae and sediments were determined using inductively coupled plasma‒optical emission spectrometry. All macroalgal species bioaccumulated metals, as demonstrated by their bioaccumulation factors. The geochemical contamination indices indicated that the estuarine complex was influenced by port activity as moderately contaminated by Pb, Cr, Mn, and Fe and considerably contaminated by Zn and Cu. The enrichment factor confirmed significant mineral enrichment of Zn and Cu in this environment. The concentrations of the metals in the sediment followed the order Fe > Mn > Cr > Zn > Cu > Pb at most sampling points. Cladophoropsis membranacea recorded the highest bioaccumulation values for Pb (0.44), Rhizoclonium africanum for Zn (1.08), Cr (0.55), and Fe (0.30), and Bostrychia radicans for Mn (2.22). The bioaccumulation pattern of metals in the most abundant macroalgal species followed the order Bostrychia radicans (Mn > Zn > Cu > Cr > Pb > Fe) and Rhizoclonium africanum (Zn > Mn > Cr > Cu > Pb > Fe).


Introduction
Globally, coastal regions undergo significant environmental stresses and are under pressure from various forms of anthropogenic activity. Intense urbanization, tourism, port use, and industrial activity are examples of large-scale human interventions within coastal regions.
Most mangrove extensions on the South American Eastern margin (approximately 80%) occur along the Brazilian coast (Ferreira & Lacerda, 2016). Mangroves cover large areas in Brazil, with a patchy distribution along the 6800 km coastline. The most extensive mangrove areas are in the north-Amapá State, Pará State, and Maranhão State (Diniz et al. (2019)-with wide tidal ranges and high levels of rainfall, which encourage their development. According to Neumann et al. (2015), high rates of population growth and urbanization in coastal regions are likely to persist. This rapid and disorganized urban expansion has a significant impact on mangroves.
Among the human impacts, we highlight the release of domestic and industrial sewage, mining, indiscriminate use of fertilizers and pesticides, and port activities. These activities are potentially harmful to estuaries, mangroves, and the organisms that depend on them. Potentially toxic metals (PTMs) are the most abundant contaminants resulting from these activities (Casado-Martínez et al., 2006).
Potentially toxic metals can enter the aquatic environment in the following forms: dissolved (in the water column and interstitial water of the sediments), particulate (bound to suspended material and bottom sediment) (Yahya et al., 2018), and biotic absorption, which can bioaccumulate in the trophic chain (Bakshi et al., 2018;Ghosh et al., 2021;Zhao et al., 2022), which makes it relevant to study the absorption of these elements by primary producers, such as macroalgae.
Macroalgae are benthic photosynthetic organisms closely related to nutrient cycling in the environment where they occur. They act as the basis of the food chain because they are primary producers that play a fundamental role in ecological structure, functioning, and balance while serving as an essential renewable resource within marine ecosystems (Valentin et al., 2010). In mangroves, macroalgae attach themselves to various substrates, such as roots, pneumatophores, dead branches, trunks, and muddy surfaces (Dawes et al., 1999).
The use of macroalgae as biomonitoring organisms has been increasing in metal contamination assessments (Billah et al., 2017;Chakraborty et al., 2014;Reis et al., 2016;Taylor et al., 2017). These organisms provide key information due to their high capacity to accumulate toxic substances, which allows their use as a tool for quantitatively assessing their environment. Therefore, they are called biological sponges, whose substance capture is controlled by two factors (Baoli & Congqiang, 2004). The first factor is the environment and is influenced by metal concentrations. The second is the biological factor and is influenced by a macroalgal affinity for metals.
Metals can enter macroalgae in two ways: adsorption on their extracellular walls (biosorption) or absorption inside the cell (bioaccumulation) (Veglio & Beolchini, 1997). Studies have shown that the effect of each metal on algae depends on factors such as the type of metal, concentration, and exposure time (Mamboya et al., 1999), in addition to the chelation and detoxification levels of each alga (Lombardi & Vieira, 1998).
In Brazil, studies of macroalgae as biomonitors of metal contamination have been conducted in Rio de Janeiro (Amado Filho et al., 1999;Karez et al., 1994;Lacerda et al., 1985), Pernambuco (Macedo et al., 2009), and Bahia (Brito et al., 2012). Overall, few studies have reported the biomonitoring of metal contamination using macroalgae from mangrove ecosystems. However, the accumulation patterns of metals in the mangroves of the Brazilian equatorial margin and port areas have not yet been determined.
Therefore, studying estuarine environments and mangroves on the equatorial margin of Brazil using sediment and macroalgae is of great importance because the region has the longest continuous band of mangroves in the country. This ecosystem partially includes one of the most important port areas on the north coast of Brazil and the world, with emphasis on the intensive movement of iron and manganese minerals that occurs in the region.
There are several significant scientific evidences that environmental changes brought about by anthropogenic activities are gradually affecting the biotic components due to the bioaccumulation potential of PTMs (Bakshi et al., 2018;Ghosh et al., 2021). Port activities are potential sources of PTMs in coastal ecosystems (Jahan & Strezov, 2019;Buruaem et al., 2012;Shafie et al., 2013). For this reason, it is necessary to understand the processes and the level of contamination that occurs in the mangroves where port activity is present, as these activities can compromise the conservation status of the mangroves.
The present study focused on (1) determining the metal levels (Cu, Zn, Fe, and Mn) in sediment in the study area; (2) investigating the patterns of metal accumulation in the dominant mangrove macroalgal species associated with pneumatophores and tree trunks of mangroves; and (3) determining the geoaccumulation index (I geo ), contamination factor (CF), enrichment factor (EF), and pollution load index (PLI). An assessment of metal concentrations in this region is necessary to identify the vulnerability of these ecosystems to pollution and compile baseline data for future monitoring.

Study area
The mangroves, where the entire study was conducted, are comprised of two large estuarine complexes of great importance for the Brazilian equatorial margin: the São Marcos estuarine complex (SMEC) and the São José complex (SJEC). The Itaqui Port Complex (IPC) is located at the SMEC, which has been internationally recognized for its operational capacity. The IPC comprises the Itaqui Organized Terminal (IOT), Ponta da Espera Maritime Terminal (PEMT), and Alumar Private Use Terminal (PUT Alumar). Intensive movement of solid mineral bulk consisting of iron and manganese ores occurs in the PEMT (MTPA, 2018). PUT Alumar is responsible for handling bauxite and alumina loads (MTPA, 2018). The processing of bauxite generates as a byproduct the so-called red mud consisting of iron, aluminum, sodium, and calcium oxides, as well as titanium dioxide and metals such as gallium, vanadium, and rare earth. SJEC was marked by intense urbanization, mainly at points 09 and 08. However, there was little to no industrial activity, especially in the absence of port activity. Due to the strong fluvial input in both estuarine complexes, the waters are characteristically turbid, with high concentrations of particulate matter, mainly clay and silt, originating from continental and mangrove areas.
The region's climate is tropical, marked by well-defined periods of rain and drought, with an average annual rainfall of 1896 mm/year and an average temperature of 27 °C (Online document Climate-Data (2021)). Thus, it is part of the context of humid tropical regions between 15°N and 15°S latitude, whose characteristics are high and constant precipitation (> 1,500 mm/year), temperatures > 20 °C, and low thermal variation (Nittrouer et al., 1995).

Sediment and macroalgae sampling
Sampling was conducted between August 2020 and February 2021 at nine sampling points: P01 to P07 in SMEC and P08 and P09 in SJEC (Fig. 1). The study only considered a spatial and not a seasonal approach.
At all points, the in situ water temperature (T°C), pH, salinity (Sal), total dissolved solids (TDS), dissolved oxygen (DO), and percentage of dissolved oxygen (PDO) were determined using a multiparametric probe model HANNA HI 98,194, always at low tide.
Sediment samples were collected from the mangrove soil. Samples were collected from the surface layer of the sediment (first 5 cm), always at low tide, using a plastic spoon. Three replicates of each sample were conducted in each plot of the study area (n = 54; considering two sample surveys). The sediment samples were dried in the laboratory at 60 °C for 48 h and ground for further treatment.
Macroalgae were collected from the trunks and pneumatophores in the intertidal zone at each sampling point, always at low tide. Pneumatophores were randomly harvested from a single sampling plot and cut in the mudline using a clipper (Billah et al., 2017). In the laboratory, the samples were carefully washed to remove adhered particles and identified based on marine macroalgae literature (Jha et al., 2009).

Determination of metal concentrations in macroalgae
In the laboratory, the algae samples were lyophilized at − 20 °C for 48 h, crushed, and homogenized for additional procedures. Extraction of inorganic elements from macroalgae was performed using 0.750 g (dry weight) of the crushed sample. Initially, the lyophilized samples were placed in Teflon tubes (X-press) to which concentrated HNO 3 (8 ml) and concentrated HF (2 mL) were added.
The extracts were allowed to rest overnight at room temperature and then placed in a microwave, model Mars X-press (CEM), for 40 min (15 min-ramp and 25 min-wait) at a temperature of 175 °C and power of 1600 W, adapted from EPA 3052. After cooling (30 min), H 3 BO 3 (12 mL) was added to neutralize the HF, and then the tubes were taken back to the microwave for 25 min (15 min, ramp and 10 min, hold) at 170 °C. The final cooled extract was filtered (Whatman No. 40) and measured at a final volume of 25 mL with 0.5 N HNO 3 in a volumetric flask. The elements were determined using inductively coupled plasma-optical emission spectrometry (IPC-OES 720). Precision measurements were obtained from the digestion of macroalgae and sediment in triplicate every 10 samples, and accuracy was determined from certified macroalgae and sediment standards Apple Leaves NIST SRM 1515. The recovery rates for metals in the macroalgae were 106% for Zn, 92.7% for Cu, 99.6% for Cr, and 104.5% for Fe.

Elementary extraction and determination of metals in sediment
Initially, dried and powdered sediment (0.5 g) was placed in Teflon tubes (X-press) to which HNO 3 (9 ml), HF (4 ml), and HCl (2 ml) were added. The extracts were allowed to rest overnight at room temperature and then placed in a microwave, model Mars X-press (CEM), for 40 min (15 min-ramp and 25 min-waiting) at a temperature of 175 °C and power of 1600 W. These adjustments were adapted from a method described by the United States Environmental Protection Agency (EPA 3052). After cooling (30 min), H 3 BO 3 (25 mL) was added to neutralize the HF, and the tubes returned to the microwave where they remained for 25 min (15 min, ramp and 10 min, hold) at 170 °C. After cooling for 30 min, the final extract was filtered (Whatman No. 40) and measured at a final volume of 50 mL with 0.5 N HNO 3 in a volumetric flask. The determination of metals was performed later in the IPC-OES 720 equipment, and NIST 1646a was used as the standard. The recovery rates for metals in the sediment were 98% for Zn, 92% for Cu, 111% for Pb, 95% for Cr, 94% for Mn, and 95% for Fe. All analyses were performed in triplicate.

Granulometric analysis and Pejrup diagram
An interval-programmed Sald-3101 Shimadzu laser diffraction particle analyzer (Table 1) was used to determine the particle size using an aliquot of 2 g of the crude sample.
The particle size distribution was treated using the SYSGRAN program (version 3.0), producing the following parameters: mean, median, standard deviation, asymmetry, kurtosis, normalized kurtosis, Wentworth classification, selection degree, and kurtosis classification. The statistical parameters of the particle size distributions were calculated from the particle size diameters (fi, mean, median, standard deviation, asymmetry, and kurtosis) using the equations proposed by Folk and Ward (1957).
The values obtained from the particle size analysis were used to construct a Pejrup diagram (1988). This diagram indicates the hydrodynamic conditions of the environment based on the sand, silt, and clay content present in the sampled sediment. A higher percentage of grains in the fine fraction indicates calmer hydrodynamic conditions.
The triangular diagram has four hydrodynamic sections (I to IV) that indicate hydrodynamic conditions: low (I), moderate (II), high (III), and very high (IV). In addition, constant sand content lines are suitable for textural classification. Thus, the triangle was divided into four texture classes, indicating a decrease in the sand content in the sediment. The sand percentages ranged from A to D, with A (90-100%), B (50-90%), C (10-50%), and D (0-100%).
Geoaccumulation index (I geo ), contamination factor (CF), enrichment factor (EF), and pollution load index (PLI) The geochemical background values for Zn, Cu, Pb, Cr, Mn, and Fe proposed by De Paula Filho et al. (2015) in a study carried out in the Parnaíba Delta were used to calculate the indices. The São Luís Island and Parnaíba sedimentary basins are classified into the same morphological unit type, known as the Barreiras do Tertiary Group (CPRM, 2000). This region presents a sedimentary basin within the ironrich area, which is composed of coarse to mediumgrained ferruginous sandstone blocks (Santos et al., 2019).
Fe was adopted as a reference element because of its similarity to other trace metals, uniform natural concentration, and characteristic association with thin solid surfaces (Varol, 2011). Thus, it was considered suitable as a normalizing element for the evaluated indices.

Geoaccumulation index (I geo )
This index quantifies the metal pollution in aquatic sediments (Muller, 1979) using the following formula: where CN is the concentration of the "N" element measured in the sediment, 1.5 is a correction factor for the background matrix that includes any possible variations in values due to lithogenic effects (Muller, 1979), and BN is the background value of the element (Ozkan, 2012; Zhiyuan et al., 2011). Below, seven descriptive classes were proposed for the values obtained from the geoaccumulation index (Table 2).
Contamination factor (CF) Hakanson (1980) proposed the contamination factor (CF) as a marker to assess the level of metal concentrations in soils. CF (Eq. 1) corresponds to the ratio between the concentration of metals measured in the sediments and the background value of the metal at a given location (Turekian & Wedepohl, 1961).
where for the "N" element, CF N is the ratio between the concentration of the element in the sediment (C N ) and the background value (B N ). Using the classification of Hakanson (1980), we obtain: FC < 1: Low contamination 1 ≤ FC < 3: Moderate contamination 3 ≤ FC < 6: Considerable contamination FC ≥ 6: High contamination The sum of all CF values per sampling point indicates the degree of contamination (DC) and determines the total value of sediment pollution at each specific point (Eq. 2) (Hakanson, 1980). where: DC < 8 = low degree of contamination 8 ≤ DC < 16 = moderate degree of contamination 16 ≤ DC < 32 = considerable degree of contamination DC ≥ 32 = very high degree of contamination indicating serious anthropogenic pollution Enrichment factor (EF) The enrichment factor (EF) corresponds to the standardization of a tested element with a reference element (Fe) that has low occurrence variability. The most common reference elements are Sc, Mn, Ti, and Al (Reimann & De Caritat, 2000;Sutherland, 2000).
Ms and Mb are the concentration of a given metal in the sediment and its background value, respectively. Fes and Feb are the concentration in the sediment and the background value of the reference metal Fe, respectively. The enrichment factor has seven categories (Nowrouzi & Pourkhabbaz, 2014), which are interpreted as follows: EF < 1: No enrichment EF = 1 -3: Minor enrichment EF = 3 -5: Moderate enrichment EF = 5 -10: Moderately to severe enrichment EF = 10 -25: Severe enrichment EF = 25 -50: Very severe enrichment EF > 50: Extremely severe enrichment EF values ranging between 0.5 and 1.5 suggest the contribution of the crust as a metal source, while values > 1.5 indicate anthropogenic influence (Zhang & Liu, 2002).

Pollution load index (PLI)
The pollution load index (PLI) is obtained by multiplying the previously calculated contamination factors (CFs) and subsequent extraction of the root, according to the following Eq. 4: where: PLI < 1: Not polluted PLI > 1: Polluted environment

Bioconcentration factor (BCF)
Bioconcentration is the accumulation of contaminants in aquatic organisms through nondietary absorption pathways, for example, from the soluble phase. The bioconcentration factor (BCF) has been used to assess the potential for metal bioaccumulation (Akcali & Kucuksezgin, 2011;Conti & Cecchetti, 2003). The calculation corresponds to the ratio of the metal concentration in the macroalgae to that in the sediment according to the following (Eq. 5): Ecological risk (Er) and potential ecological risk index (Ri) Hakanson (1980) developed an ecological risk assessment to assess whether concentrations of metals found in sediments have a potential effect on marine ecosystem organisms. The ecological risk assessment (Er) uses the toxicity coefficient (Tr) and the contamination factor (CF), as shown in Eq. 6. The toxicity coefficients for each metal are Cu = 5, Pb = 5, Zn = 1 and Cr = 2. In turn, the potential ecological risk index (Ri) is determined by the sum of all values per sampling point, as shown in Eq. 7.
An Er of < 40 is a low potential ecological risk, while 40 < Er < 80 is a moderate potential ecological risk, 80 < Er < 160 is a considerable potential ecological risk, 160 < Er < 320 is a high potential ecological risk and > 320 is a very high potential ecological risk. A Ri under 150 is considered a low ecological risk, 150 ≤ Ri < 300 is considered a moderate ecological risk, 300 ≤ Ri < 600 is considered a considerable ecological risk, and an RI ≥ 600 is considered a very high ecological risk (Hakanson, 1980).
Threshold effect level (TEL), probable effect level (PEL), effects range low (ERL), and effects range medium (ERM) TEL and PEL are interpretive toxicity criteria for marine and estuarine sediments that were established by Canadian legislation based on compiled biological and chemical data from laboratory studies, field measurements, and numerous models. TEL indicates the level below which there are no adverse effects on the biological community, and PEL is the level above which adverse effects are expected ( Table 3). The range above the TEL and below the PEL represents a possible adverse effect on the community (MacDonald, 1994;MacDonald et al., 1996). ERL and ERM are American criteria established by Long et al. (1995) using chemical and biological data from field studies in marine and estuarine sediments. The ERL is the concentration limit below which it is rarely toxic. The ERM, in turn, is the limit above which there is a high likelihood of toxicity, and the range between the ERL and ERM indicates possible toxicity (Long et al., 1995). The ERL and ERM limits were adopted as reference values by resolutions 344/2004 and 454/2012 of the National Environmental Council (CONAMA, Brazil) for estuarine and marine sediments (Table 3). The TEL, PEL, ERL, and ERM limits are not applicable to Al, Fe, and Mn.

Statistical analysis
The Shapiro-Wilk test was used to verify the normality of all data. ANOVA (one-way) and Kruskal-Wallis tests were used to test for significant differences. Spearman's correlation coefficient was used to understand the paired relationships between the measured variables. For all tests, p < 0.05 and p < 0.01 were considered to indicate statistical significance. Analyses were performed using SPSS v26 software.

Physical and chemical variables
The physical and chemical data obtained during sampling are summarized in Table 4. Water temperature, salinity, and dissolved oxygen showed significant differences between the sampling points (ANOVA, p < 0.05) and TDS (Kruskal-Wallis, p < 0.05). All data shown as averages are results from samples collected in triplicate. The lowest and highest mean water temperatures were observed in P09 and P02, at 28.14 and 35.60 °C, respectively. The mean salinity ranged from 17.9 to 28.6, and the lowest values were found at the sampling points belonging to the SMEC. P07 (17.9) showed statistically significant differences compared to all other sampling points (ANOVA, p < 0.05).

Granulometric analysis and Pejrup diagram
The granulometric analysis recorded mean Phi values between 2.84 and 8.28, showing that the local particle size ranged from fine sand to clay. Sedimentary grains ranged from very poorly to moderately selected, characterizing a variation in particle size between sampling points (Table 5).
Most sampling points presented platykurtic or mesokurtic kurtosis with a heterogeneous value distribution, except for P05 and P08, which recorded leptokurtic kurtosis; these data were concentrated around the mean (Table 5).
Only points P01 and P05 showed significant differences in particle size (ANOVA, p = 0.05). There was no significant correlation between the metals in the sediment and the particle size. Although there was no significant correlation between metals and grain size, correlation analysis showed a trend towards a direct correlation between metals and silt and clay grains (Phi 4-8) (Table S5). On the other hand, there was an inverse correlation trend between metals and sand grains (Phi 2-3) ( Table S5).
The Pejrup diagram (Fig. 2) shows that most points (P03, P04, P05, P06, P07, and P09) were associated with moderate hydrodynamic conditions. Among them, only P04 and P09 showed a predominance of sand deposited under moderate hydrodynamic conditions (groups B and II). P03, P05, P06, and P07 showed a predominance of fine grains deposited under the same hydrodynamic conditions (Groups C, II, D, and II). P01 was associated with the predominance of sand deposited under low hydrodynamic conditions (groups B and I). P02 and P08 were the only ones indicated under high hydrodynamics (groups C and III and B and III, respectively) (Fig. 2).

Metal concentration in sediments
The concentrations of Zn, Cu, Pb, Cr, Mn, and Fe in the surface sediments at the nine sampled points are summarized in Table 6. Based on the mean values, the concentration pattern followed the order Fe > Mn > Cr > Zn > Cu > Pb at all sampling points except for P07, whose pattern followed the order Fe > Mn > Zn > Cr > Cu > Pb.
The concentrations obtained presented a normal distribution and homogeneity of variances only for Fe (Shapiro-Wilk, p > 0.05; Levene, p > 0.05). Fe was also the only element that showed a significant difference between P05 and P06 (ANOVA, p < 0.05). The other metals did not show significant differences between sampling points.
Metal concentrations at most sampling points exceeded regional background values in both estuarine complexes-SMEC and SJEC (Table 6). The lowest and highest mean values of Zn were recorded in P07 (12.07 ± 6.91 µg/g) and P03 (57.09 ± 55.71 µg/g), respectively. The furthest point from the port area, P07, recorded the lowest mean values for Cu (10.10 ± 1.38 µg/g), Pb (2.52 ± 0.00 µg/g), Cr (11.06 ± 6.17 µg/g) and Mn (19.53 ± 19.49 µg/g). All registered values of the sampling points were lower than those of TEL, PEL, ERL, and ERM (Fig. 3). Cu was an exception, as its concentrations were higher than the TEL in P02 (Cu = 21.05) and P04 (Cu = 19.91). All points registered concentrations below those established by the national legislation contained in resolutions 344/2004 and 454/2012 of the National Council for the Environment (CON-AMA, Brazil) for estuarine and marine sediments, whose guidelines are based on ERL and ERM values.

Metal concentration in macroalgae
The concentrations of Zn, Cu, Pb, Cr, Mn, and Fe in the macroalgae between the sampling points (P1 to P6) are summarized in Table 7. In the macroalgae collected at points P07 to P09, the concentrations of metals were below the detection level; therefore, they were removed from the table.
The highest concentration (5.61 µg/g) of Pb was recorded in P04, and the lowest (1.00 µg/g) was recorded in P06 in the macroalgae C. membranacea. Cr had the lowest concentration detected in the species C. caespitosa (2.06 µg/g) and the highest in the species R. africanum (19.90 µg/g), also recorded in P04 and P06, respectively ( Table 7).
The concentrations of Mn and Fe were higher in all species than those of the other quantified elements (ANOVA, p < 0.05). Mn showed a high variation between the highest (825.86 µg/g) and the lowest (13.56 µg/g) values detected in C. membranacea (P05) and R. africanum (P04). Fe also showed a high rate of variation between the highest (8060.56 µg/g) and lowest (549.02 µg/g) values detected in C. membranacea (P04) and C. caespitosa (P01) ( Table 7).
Bioaccumulation factor in macroalgae Figure 4 shows the bioaccumulation factor (BCF), indicating that almost all macroalgae collected in this study were susceptible to bioaccumulation. B. radicans recorded the highest bioaccumulation value for Mn (2.22) and the lowest for Pb (0.31). C. caespitosa, also Rhodophyceae, showed low bioaccumulation values compared to the other species (Cu = 0.15, Pb = 0.001, Cr = 0.09, Mn = 0.64, and Fe = 0.03).

Geochemical indices
The geoaccumulation index (I geo ) was calculated to quantify metal pollution in the sediment, the values of which are shown in Fig. 5. In general, the sampled points presented an I geo below two, which characterizes the environment as moderately contaminated to noncontaminated.
The points closest to the port areas (P01, P02, P03, and P04) showed moderate contamination (0 < I geo < 1) of two or more metals. The highest I geo values were identified in P03 and P07 for Zn (I geo = 1.51) and Cu (I geo = 1.05). However, Mn was below 0 at all sampling points, characterizing them as noncontaminated for this element. Although the I geo values varied between environments, only Fe showed a significant difference (ANOVA, p < 0.05) between points P05 and P06 (Fig. 5). For CF, only Mn showed values below one at all sampled points, indicating low contamination of this metal in mangroves. In general, based on the CF values, the environment tended to be moderately contaminated. P02 and P03 exhibited considerable contamination with Cu (CF = 3.10) and Zn (CF = 4.26), respectively. The other elements followed the general trend of moderate contamination.
To assess the contribution of natural and anthropogenic sources to mineral enrichment at the sampling points, an enrichment factor was applied, as summarized in Fig. 7. In general, the environment showed a tendency toward minimal mineral enrichment (1 < EF < 3). Zn and Cu presented EF values above five at P03, indicating significant enrichment at this point. The EF values for Cu at all sampling points, except P03 and P05, were between 1.5 and 3 (minimum enrichment), showing anthropogenic influence for this metal. P03 was significantly different at all points for all metals (Kruskal-Wallis, p < 0.05).
The Er values generally showed that all sampling points had a low potential ecological risk for the elements Zn, Cu, Pb and Cr (Fig. 8). The highest values of Er for Zn were 4.26 (P04), Cu 15.48 (P02), Pb 7.62 (P03), and Cr 4.47 (P02). P07 showed the lowest Er values for all metals (Zn = 0.9; Cu = 7.43; Pb = 0.71; Cr = 1.23). The ecological risk index (Ri) ranged from 10.27 to 29.84, confirming the low ecological risk for these metals at all sampled points (Fig. 8).

Discussion
The values of the physical and chemical variables obtained in this study agree with those recorded for both SMEC and SJEC by Cavalcanti and Cutrim (2018) and Serejo et al. (2020).
Although many studies point to salinity, pH, and DO variables as relevant to the mobility of metals in estuarine environments Li et al., 2013;Samani et al., 2014), these parameters showed weak positive correlations, although not significant, with the metal values in the sediments. However, among the parameters measured in this study, the temperature was moderate to strongly related to Zn, Pb, and Cr in the sedimentary compartment (Table S1). This highlights sediment as a potential source of these metals for interstitial and surrounding waters.
It is known that high temperatures provide a high release of metals from the sediment to the liquid medium, which, consequently, are bioavailable for macroalgae Schuhmacher et al., 1995;Zhao et al., 2013) because such organisms only accumulate metallic ions dissolved in water (Luoma, 1983;Luoma et al., 1982).
Only DO showed a strong correlation with metal concentrations in macroalgae. Such correlations were observed between the DO and Cu, Pb, Cr, and Mn (Table S2). This is the same relationship reported by Macedo et al. (2009), who found that the copper content in rhodophyte species is influenced by dissolved oxygen. Studies that address the relationship between the accumulation of metals in macroalgae and DO are scarce, thus compromising our understanding of this interaction.
Some studies have described the influence of DO on metals adsorbed to the sediment (Atkinson et al., 2007;Huang et al., 2017;Kang et al., 2019;Li et al., 2013;Liu et al., 2019). DO values ranging between 7.0 and 9.0 mg/L promote the release of metals from the sediment to the interstitial water , becoming bioavailable for macroalgae, while DO values lower than 7.0 mg/L provide little or no release of metals to water, depending on the metal .
Conversely, DO values lower than 7.0 mg/L provide little or no release of metals to water, depending on the metal ; this primarily occurs if such values are observed due to the intense degradation of organic matter (Gerringa, 1991), which may explain the high concentrations of metals in P09. At the time of collection, domestic sewage was discharged directly into the sampling location at P09.

Metals in sediment
The metal concentrations obtained in the sediment samples were compared with the world mean concentrations of metals in shale (Turekian & Wedepohl, 1961) and the regional average (De Paula Filho et al., 2015) (Table 6). Regarding the global average, the results of this study were well below; however, above the regional average, sometimes on the order of two to three times higher. Mn was the exception, being below global and regional concentrations, corroborating the assertion of non-contamination of the environment by this metal (Table 6).
The global values proposed by Turekian and Wedepohl (1961) are high and therefore may not be representative of different sedimentary basins. Thus, it is necessary to define the use of background values at local or regional levels, according to the content deposited in the sedimentary basin in preindustrial times (De Paula Filho et al., 2015, 2021. Only Fe showed a significant difference among the sampling points (ANOVA, p > 0.05). However, all metals were significantly strongly and positively correlated with each other, indicating common sources and processes (De Paula Filho et al., 2021). Zn, Pb, and Cr were strongly correlated with Mn (Table S1), indicating the probable adsorption of these metals to Mn oxide hydroxides (Chakraborty et al., 2014;Udechukwu et al., 2014). In contrast, Cu was more strongly correlated with Fe (Table S1), indicating the probable adsorption of Cu to Fe oxide hydroxides (Chakraborty et al., 2014).
Mn manifested itself as a metal carrier for the sediment in the estuarine complexes of the Brazilian equatorial margin, probably because of its high oxidation resistance and subsequent flocculation, remaining longer in the water column and thus adsorbing a greater number of metals onto Fe. Mn tends to deposit in the form of oxyhydroxide only when oxygen-rich marine waters enter the estuary (Lacerda et al., 1999). Santos et al. (2019) verified that significant differences were observed between Mn, Fe, and Zn in SMEC, suggesting different origins and carriers, and Fe and Mn showed strong correlations with each other. Such results are contrary to those of this study, which suggests the relevant capacity of the system to behave in different ways depending on the level of contamination to which it is exposed. However, both studies point to similar behaviors for Cu, Pb, and Cr.

Geochemical indices
Several studies have evaluated metal contamination in port areas around the world, highlighting these areas as potential sources of Zn, Cu, Cr, and Pb (Jahan & Strezov, 2019;Sari et al., 2014;Shafie et al., 2013;Buruaem et al., 2012). The results for the port area of the Brazilian equatorial margin were not different, since, in general, the I geo , CF and EF indices pointed to Zn, Cu, and Cr as the main contaminants of the sampled environments, originating from anthropic sources, mainly Cu. The I geo varied between 0 and 2, mainly in the points associated with the adjacencies of the port areas. The FC pointed out points from P02 to P05 as being the most affected by metal contamination. Such sampling points were considered moderately contaminated (Fig. 6), corroborated by the degree of contamination index (DC). Finally, the PLI reinforces that contamination at these sampling points has caused their pollution (Fig. 6).
Although the geochemical indices showed contamination and pollution in the port area (P02) and adjacent areas (P03 and P04), the values of Er and Ri revealed a low potential for ecological risk. In general, Er and Ri recorded low Er values for all nine sampling points (Fig. 8). Notably, the values of Er and Ri are the results of the evaluation of the elements Zn, Cu, Pb, and Cr only. As in this study, Williams and Antoine (2020) revealed a low ecological risk for the metals Zn, Cu, Pb, and Cr in a port region of Jamaica (Kingston Harbor). However, it revealed Er values ranging from 89.8 to 192.9 for Co, showing that this area has a considerable high potential ecological risk caused by this element. In the Ri assessment for all sampled metals, Williams and Antoine (2020) found Ri values from 157.6 to 347.6, evidencing moderate to considerable ecological risk driven by Co values. Therefore, in the Brazilian equatorial margin, one cannot rule out the possibility of a potential ecological risk caused by other metals with high toxicity coefficients (Tr), such as Co, Hg, Cd, and As, not evaluated in this study.
The high values of Pb and Cr on the Brazilian equatorial margin may be associated with the wear of the hulls of ships that dock daily in ports in this region (Santos et al., 2019). In turn, Cu can be released by antifouling paint used on vessels (Santos et al., 2019;Shafie et al., 2013). High Cu values are also related to potential anthropic sources regarding the release of domestic sewage toward the estuary (Costa et al., 2015). Therefore, the contamination levels recorded in the points adjacent to the port area result in a synergistic effect between port activities and domestic sewage discharges in this region. It is more evident when looking at the concentrations of Cu in P03, characterized by being the closest point to the port area and having an extensive community of fishermen who live without adequate sanitation.
Recent studies have shown a similar pattern of behavior for PTMs in estuarine ecosystems where anthropic activities are present. Bakshi et al. (2018) Page 17 of 23 432 Vol.: (0123456789) and Ghosh et al. (2021) found similar behavior to this study. The geochemical indices indicated low to moderate contamination of the sediments for metals such as Zn, Pb, Cr, Fe and Mn. On the other hand, Cu proved to be an element of high contamination, corroborating the findings of this study. Yap and Al-Mutairi (2023) pointed out that, in the Klang mangrove (Malaysia), Cu proved to be, through geochemical indices, an element of moderate to high contamination of the environment, in addition to presenting a considerable potential ecological risk (80 ≤ RE < 160), according to Hakanson. The strong, positive, and significant correlations between all sampled points (Table S3) suggest that the estuarine complexes of the Brazilian equatorial margin, SMEC and SJEC, share similar sources and processes responsible for controlling the mobility and bioavailability of metallic elements. However, although there are similarities, the points located in the SMEC seem to suffer from high anthropic interference in terms of sources and processes. For example, P02, P03, and P04 exceeded two to three times the regional background values (Table 6), demonstrating relevant levels of contamination, as evidenced by the different geochemical indices. The local hydrodynamics seem to contribute to the trapping of metals from P1 to P5, as there was a decrease in the gradient of metal concentrations. P02, P03, and P04 are in places with low depths (< 20 m) and currents with speeds of less than 1.1 m/s (González-Gorbeña et al., 2015), which can favor flocculation and deposition processes at such points. The local hydrodynamics seem to contribute to the trapping of metals from P1 to P5, as there was a decrease in the gradient of metal concentrations. Points P02, P03 and P04 are in places with low depths (< 20 m), currents with velocities below 1.1 m/s (González-Gorbeña et al., 2015) and finer sedimentary grains (Table 5), which may favor flocculation and deposition processes in these areas. Santos et al. (2019) commented that high concentrations of metals in low-energy zones are a consequence of the greater ability that these metals acquire to adsorb to fine sediments and organic material. This effect was evident when comparing the concentrations of metals in the sediment and the results of the Pejrup diagram. P02, P03, and P04 showed low to moderate hydrodynamics and the highest concentrations of metals in the sediment ( Fig. 2; Table 6). TEL, PEL, ERL, and ERM Compared with the TEL, PEL, ERL, and ERM values, the metal concentrations between the points were well below the established limits, except for the Cu concentrations in P02 and P04, which were above the TEL and below the PEL (Fig. 3).
This range between the TEL and PEL draws attention to possible adverse effects on the biological community (De Paula Filho et al., 2021). Therefore, although Cu is an essential element, it can be toxic to aquatic organisms, especially those that live in or are in direct contact with sediments whose Cu values are in this range (CCME, 1999). De Paula  observed this Cu performance in the Parnaíba Delta, a region also located on the Brazilian equatorial margin, whose concentrations exceeded the TEL at 61% of the sampling sites, mainly in areas with more sediment deposition.
In Brazil, ERL and ERM limits were adopted as reference values by resolutions 344/2004 and 454/2012 of the National Council for the Environment (CONAMA, Brazil) for estuarine and marine sediments. Therefore, the sampled points comply with national legislation.
The lack of rocky substrate, low salinity, and high turbidity of coastal waters make the island of São Luís an environment with a low diversity of macroalgae (Oliveira- Filho, 1984). These conditions favor the dominance of species adapted to mangrove ecosystems. Macroalgal species grow epiphytically in pneumatophores, rhizophores, and trunks (Phillips et al., 1994(Phillips et al., , 1996, as is the case with the species found in this study. The correlation values between the concentrations of metals in macroalgae and the concentrations in sediments showed strong and significant correlations and demonstrated that macroalgal bioaccumulation is strongly associated with the concentrations of metals found in sediments (Table S4). Bioaccumulation factor (BCF) All macroalgae bioaccumulated in different orders and magnitudes according to their BCF (Fig. 4). Although macroalgae share the same habitat and conditions, they exhibit particularities in metal accumulation, such as life span, morphology, contact surface area, growth rate, and selective metal affinities (Chakraborty et al., 2014;Schintu et al., 2010). Among the essential metals, absorption followed the order Mn > Zn > Cu > Fe for B. radicans and C. membranacea, whereas C. caespitosa and R. africanum followed the order Zn > Mn > Cu > Fe.
Among the nonessential metals, the absorption process followed the order of Cr > Pb for both groups of macroalgae. Chlorophyceae biosorbed Cr and Pb (nonessential) at higher concentrations than essential metals, such as Cu and Fe.
The biosorption process occurs in macroalgae because cell wall polysaccharides have functional groups (hydroxyl, sulfate, and carboxyl) that act as ion exchangers and bind to metallic cations (Trifan et al., 2015). Once adsorbed to the cell wall, metals can precipitate through the excretion of metabolic products by macroalgae or be transported slowly into the cytoplasm through chemisorption (Veglio & Beolchini, 1997). This last process is the most harmful, as metals reaching the cytoplasm and bioaccumulation can cause adverse effects on macroalgae, such as oxidative stress, inhibited photosynthesis and chlorophyll production, and deficient macroalgal growth (Baumann et al., 2009;Pinto et al., 2003;Gledhill et al., 1997).
The bioaccumulation of metals is not restricted to microalgae as primary producers because, once absorbed, these metals can be transferred along the food chain and reach humans (Santos et al., 2019).
Thick macroalgae tend to have lower metal concentrations than filamentous macroalgae (Trifan et al., 2015), which explains the reduced accumulation values in C. caespitosa compared to other species. Chlorophyceous showed a better accumulation response at the sampled points, with high observed concentrations of metals, most of which were nonessential metals. This affinity of chlorophytes with metals can be attributed to their growth in direct contact with the sediment and the surrounding water, especially C. membranacea, which adheres to the muddy substrate and retains large amounts of sediment owing to its cushion-like morphology.
The results obtained in this study are similar to those of other studies conducted on other mangroves. Chlorophytes are excellent biomonitors owing to their high capacity to reflect environmental contamination. Chakraborty et al. (2014) verified the efficiency of Ulva lactuca in accumulating high levels of Cu, Fe, and Zn, and Akcali and Kucuksezgin (2011) reported high levels of Cu, Pb, Zn, and Fe.
During the absorption process, macroalgae tend to control the composition and concentration of essential metals when they are available in the environment at levels that only satisfy the needs of metabolism and growth. When these levels exceed those required by the organism, biological regulation is impaired owing to external forcing. Consequently, macroalgae accumulate high levels of essential metals and concomitantly accumulate these nonessential metals (Baoli & Congqiang, 2004). The strong positive and significant correlations between metal concentrations in macroalgae and sediment, accompanied by the high values recorded for the bioaccumulation factor, reinforce this statement.
Mn is an essential metal required by macroalgae, and in this study, it was a relevant carrier of metals such as Zn, Pb, and Cr to the sediment. Mn was the most bioaccumulated metal ( Fig. 4), and we realized that it influenced the bioaccumulation of these metals. Hall and Brown (2002) demonstrated that the association between Mn and Cu results in a high absorption of Cd in macroalgae.
The representativeness of C. membranacea in the bioaccumulation of Mn in relation to other species allows it to be chosen as a key species for monitoring this metal and others associated with it. Compared to Chlorophyceae, the two species of Rhodophyceae proved to be better biomonitors for assessing environmental contamination by Fe (Kruskal-Wallis; p < 0.05). Therefore, the use of the four species of macroalgae is complementary and essential for a broader response to contaminants in mangrove ecosystems.

Conclusion
The metal concentrations recorded in the sediments of both estuarine complexes were below the mean values of the global background but above the regional level. This reinforces what past studies have pointed out regarding the need to carry out local and regional surveys of the concentrations of metals deposited in preindustrial times, as the high global mean values may not be representative of the different sedimentary basins.
The geochemical indices indicate moderate contamination and enrichment among the studied environments, showing significant enrichment of Zn and Cu in the regions under the influence of port activity. This result reinforces the idea that although there is contamination by metals in other estuarine areas on the Brazilian equatorial margin, port activity remains a relevant source of contaminants in the estuarine environment.
Local hydrodynamics have a significant influence on the metal deposition process in areas close to harbor regions. Cu concentrations were above the TEL and below the PEL, indicating possible adverse effects on the biological community. The results showed different relationships from those reported in previous studies, indicating different environmental responses to their exposure to contaminants.
The absence of significant differences and strong correlations between metals suggests similar sources and processes for the deposition of these elements in the equatorial margin of Brazil. Mn showed a strong positive correlation with Zn, Pb, and Cr, which proved to be excellent metal carriers in this region.
All macroalgae bioaccumulate metals, demonstrating the biomonitor potential of these organisms. The species showed different patterns of bioaccumulation; however, Mn and Zn were the most bioaccumulated. The divergent behavior in the bioaccumulation pattern among macroalgae suggests that monitoring different species is essential for a more accurate response to environmental contamination by different types of metals.
We suggest complementary studies to verify possible seasonal variations in the bioaccumulation patterns of macroalgae associated with mangroves on the Brazilian equatorial margin, primarily because the dynamics of metal bioaccumulation in the macroalgal species found in this area are poorly investigated on the international scene, which is the main difficulty of this work. In this way, we provide preliminary information about the accumulation of metals regarding the species found in our investigation.