Mercury in Predatory and Non-Predatory Fish in the Madeira River, Brazilian Amazon Region

Abstract

Madeira River (Rondônia State, Brazil) is known to be contaminated with mercury (Hg), which can be biomagnified and bioaccumulated through the food chain. The situation raises concern given there are riverside inhabitants that consume fish. Therefore, 18 fish species with a broad spectrum of eating habits were investigated. Mercury concentrations in fish muscle were evaluated according to the limits established by the Brazilian National Health Surveillance Agency (ANVISA) and the European Union (EU). ANVISA sets maximum limits of 1.0 µg g^-1 for predator species and 0.5 µg g^-1 for non-predator species (Brazil), whereas EU establishes a maximum limit of 0.5 µg g^-1 (EU). Thirteen species have Hg levels above the maximum recommended (ANVISA limits). The Hg concentration ranged from 0.15 to 20.7 and 0.012 to 1.14 µg g^-1 for predatory and non predatory fish, respectively. A flow batch chemical vapor generation-inductively coupled plasma mass spectrometric (CV-ICP-MS) method was used, with limit of detection (LOD) of 0.012 µg g^-1, value below the maximum limit for mercury in fish established by ANVISA and the European Union. Accuracy was evaluated by analyzing certified refence material (CRM), with precision better than 5% (relative standard deviation, n = 5). The results demonstrate the current status of Hg in fish and highlight the seriousness of environmental contamination, as mercury concentrations in several species exceed expected levels. In addition, Hg levels in fish muscle are comparable to or higher than those reported in studies conducted in the same region since 2000.

Keywords:
mercury; fish; mercury bioaccumulation; risk assessment

Introduction

Mercury (Hg) has a strong affinity for sulfur and is primarily found in nature as cinnabar (HgS). The emission of Hg into the environment can occur naturally or anthropogenically, with the main natural source of emission coming from the degassing process of the Earth crust or active volcanoes. Anthropogenic emissions may result from several industrial uses of Hg as in the production of chlorine and soda, and pesticides. In addition, improper electronic waste disposal and waste incineration, among other anthropogenic activities, contribute to Hg emissions. It is estimated that, in 2015, approximately 2,500 tons of Hg were released into the environment by anthropogenic activities, representing about 57% of total mercury released globally.¹,²

In the North region of Brazil, Hg is used in large amounts for gold mining. As reported by Basta³ there are 4,114 illegal gold mining sites in the Amazon biome (Organização do Tratado de Cooperação Amazônica (OTCA)), where it is estimated that more than 221.5 tons of Hg were released into the environment in 2023 alone. The results indicate a close relation between illegality (mining without permission) and great mercury anthropogenic introduction to the environment.⁴ In this process, metallic Hg is used, and the resulting waste is often improperly discarded into rivers, causing huge environmental impacts and primarily contaminating water, sediments, and biota. In addition, deforestation is another critical source of Hg emission contributing to the contamination of the Amazon region.⁵,⁶ Therefore, indiscriminate use of Hg produces harmful environmental contamination, especially in regions closest to disposal.²,⁵-⁸

Mercury, after being released to the environment, can be converted into different inorganic species, including elemental Hg (Hg⁰), mercurous (Hg₂²⁺) and mercuric (Hg²⁺) ions, and organic compounds, where the most common are methylmercury (CH₃Hg⁺) and dimethylmercury ((CH₃)₂Hg).⁹ The conversion of the element between species occurs in most environmental compartments, where it can remain in water, sediment, or soil, or be volatilized into the atmosphere.¹⁰ In water, elemental mercury can be converted to inorganic species, being Hg²⁺ the most common form. This species is highly soluble in water and can be transported over long distances, leading to the contamination of extensive areas. In contrast, Hg₂²⁺ exhibits low solubility in water and therefore limited environmental mobility. In surface waters, most of the inorganic Hg is covalently bound with soluble organic compounds or is biologically converted by bacteria, plankton and fungi to organic species.⁶ Briefly, released Hg⁰ can be converted to Hg²⁺, which can subsequently be methylated by the action of sulfateand iron-reducing bacteria and/or at methanogenic environment and under anoxic conditions in sediments. In addition, methylation occurs in periphyton and wetland areas, and is highest in sediments with high content of organic compounds and sulfate.¹¹

Mercury speciation in aquatic systems is dynamically regulated by the presence of sulfide and dissolved organic matter (DOM). In surface waters, up to 30% of total Hg can be in the form of methylmercury,¹⁰ while the predominant inorganic species vary between Hg-DOM complexes (in the absence of sulfide) and Hg-sulfide complexes (when S^2- concentration is higher than 20 µg L^-1). Intermediate sulfide concentrations (ca. 0.32 ng L^-1) favor the coexistence of both species, demonstrating that CH₃Hg⁺ formation is closely linked to these chemical equilibria, further mediated by factors such as iron and redox conditions.¹¹,¹² Mercury retention in river sediments is high due to complexation with organic matter (humic and fulvic acids), inhibiting the in situ methylation. However, erosive processes may transport Hg to the water column, where subsequent methylation in aquatic sediments leads to the formation of the highly bioaccumulative species, CH₃Hg⁺. This form represents approximately 80% of Hg in fish tissues, constituting the main route of human exposure.¹³,¹⁴ Chronic ingestion is associated with severe neurological, renal, cardiovascular, and immunological toxicities,¹⁵,¹⁶ ranking Hg as the third most toxic element to humans according to Agency for Toxic Substances and Disease Registry (ASTDR).¹⁷

The risk of contamination is even greater for the riverside population, where fish consumption is greater.¹⁸,¹⁹ As previously mentioned, in the Northern region of Brazil, one of the main sources of contamination in aquatic environments is associated with anthropogenic emissions from gold mining activities, in which large amounts of Hg are estimated to be used. However, there is no precise information on the amount of Hg used in this activity. It is estimated that there are more than 4,000 illegal mining sites in the Amazon region and that, on average, approximately 150 tons of mercury are discarded annually in the region.²,²⁰

The amount of Hg present in fish species is mainly related to their feeding habits, as fish at the top of the trophic chain accumulate higher levels of Hg than those at lower trophic levels. In this context, trophic levels can be classified as herbivorous and detritivorous (lower trophic levels), omnivorous (intermediate level), and piscivorous and carnivorous (top level). Accordingly, species may also be classified as non-predatory or predatory, with herbivorous and detritivorous species considered non-predatory, and omnivorous, piscivorous, and carnivorous species classified as predatory. Another important factor is associated to the characteristics of the aquatic environment in which the fish lives, particularly its vertical habitat distribution, whether near the bottom (close to sediments) or near the surface. Fish that inhabit areas close to the sediment are more susceptible to Hg contamination and bioaccumulation, as Hg concentrations are generally higher in sediments than in surface waters.²¹,²²

In Brazil, the National Health Surveillance Agency²³ (ANVISA) establishes maximum limits for the total Hg concentration in fish, according to the trophic level of the species. Therefore, for non-predatory fish species the maximum limit must not exceed 0.5 µg g^-1 and for predatory species the total Hg concentration must not exceed 1.0 µg g^-1. The limits adopted by the European Union²⁴ also depends on the type of fish, where maximum levels of 0.3, 0.5 and 1.0 mg g^-1 is allowed, while the Food and Agriculture Organization of the United Nations (WHO/FAO)²⁵ recommends maximum values of 0.5 and 1.0 µg g^-1 of CH₃Hg⁺ for non-predatory and predatory fish, respectively.

Given these characteristics of Hg, it is necessary to monitor its levels in different environmental matrices, including fish. Mercury in fish can be determined directly from the solid sample or after sample decomposition. Numerous studies report the use of a wide variety of analytical techniques for the determination of Hg based on cold vapor generation (CV) coupled to spectrometric techniques, such as atomic absorption spectrometry (CV AAS), inductively coupled plasma optical emission spectrometry (C ICP OES), and inductively coupled plasma mass spectrometry (CV ICP MS).²⁶-²⁹

Therefore, the main focus of the present study was to evaluate mercury concentrations in 18 fish species collected from the Madeira River, located in the Porto Velho region (Rondônia State, Brazil), a region heavily impacted by the use of mercury in gold mining. The samples were chosen in order to include fish species with different feeding habits, including both predatory and non-predatory species. Mercury was quantified by inductively coupled plasma mass spectrometry (ICP-MS) with sample introduction through a cold vapor generation technique employing a miniaturized flow batch (FB) system, which was in house developed and validated. The results were compared with data reported in studies published since 2000. A literature search was conducted in ScienceDirect, Web of Science, PubMed, SciELO, and Scopus using the keywords “Madeira River,” “mercury,” and “fish”.

Experimental

Samples

Fish samples (muscle) were acquired in five markets located in the municipality of Porto Velho (latitude 08°45’43” S and longitude 63°54’14” W), Rondônia State, Brazil. All samples are from the Madeira River and surrounding areas of Porto Velho city. At the time of collection, the characteristics of the fish were recorded, including the local common name of each species in the region, length and mass. Approximately 100 g of muscle from one fish of each species were removed with the aid of a stainless-steel knife, with a total of 18 species collected. After sampling, the samples were kept and transported at low temperature (frozen at approximately 0 ^oC) until preparation in the laboratory. To better understand mercury concentrations in different fish species, a survey of species characteristics was conducted, including feeding habits, physical characteristics (leather or scaled fish) and habitat (whether the species inhabits surface waters or lives near the river bottom). All relevant sample information is described in Table 1.

Table 1
Main characteristics of the evaluated fish

Trophic level	Fish species	Physical characteristic	Size / cm	Mass / kg
Herbivore (non-predatory)	Semaprochilodus insignis (Jaraqui)	scale	33	0.480
	Psectrogaster amazonica (Branquinha)	scale	36	0.710
	Piaractus brachypomus (Pirapitinga)	scale	25	0.255
Omnivorous (predatory)	Danio rerio (Zebra)	scale	3.8	1.140
	Arapaima gigas (Pirarucu)	scale	160	130
	Seriola dumerili (Lírio)	scale	42	0.470
	Phractocephalus hemeliopterus (Pirarara)	leather	118	15.00
	Astronotus ocellatus (Carauaçu)	scale	29	0.635
	Colossoma macropomum (Tambaqui)	scale	55	3.085
	Piaractus mesopotamicus (Pacu)	scale	17	0.145
Piscivorous (predatory)	Paulicea luetkeni (Jaú)	scale	100	12.00
	Salminus maxillosus (Barbado)	leather	49	0.870
Carnivore (predatory)	Brachyplatystoma platynema (Babão)	leather	99	6.230
	Salminus brasiliensis (Dourado)	scale	72	8.00
	Pseudoplatystoma corruscans (Pintado)	leather	65	1.510
	Osteoglossum bicirhossum (Aruanã)	scale	65	1.685
	Cichla ocellaris (Tucunaré)	scale	36	0.710
	Calophysus macropterus (Piracatinga)	leather	100	12.00

Sample preparation

Reagents used for sample decomposition and solution preparation were of high purity (analytical grade or better). Sample decomposition was carried out as described by Viana et al.³⁰ In short, after thawing, samples were cut into small pieces and a wet mass of 500 mg was accurately weighed and transferred to a quartz flask with a capacity of 80 mL. Then, 3 mL of concentrated nitric acid, 1 mL of concentrated hydrochloric acid and 2 mL of purified water were added. Acids of P.A. grade were purified by sub-boiling and water was distilled and deionized in order to obtain 18.2 MΩ cm of resistivity. The flasks were closed and the samples decomposed by heating in a microwave oven with a 5 min ramp and 10 min of residence time at 250 °C and 80 bar. After cooling, the samples were collected in 50 mL polypropylene flasks and made up to 20 mL with purified water. To check the accuracy of the results, dogfish liver certified reference material (CRM - DOLT-4, NRCC, Canada) was analyzed. This CRM was prepared by using 100 mg and the same conditions as applied to the fish samples. Blanks were prepared under the same conditions as the samples, using only water and nitric and hydrochloric acids.

Mercury determination in fish muscle

Total Hg determination was carried out using a flow batch chemical vapor generation system coupled to an ICP MS instrument (FB-CV-ICP-MS), previously described and validated by Viana et al.³⁰ In short, a 0.005% m v^-1 NaBH₄ in 0.1% m v^-1 NaOH solution was employed as reducing agent, while 1% v v^-1 HCl solution was used to acidify the reaction medium. Both reagents and samples were pumped to the system with a constant flow rate of 1.0 mL min^-1. Argon was used as a carrier with a flow rate of 1.12 L min^-1. Mercury determinations were done by external calibration, processing the calibration curve solutions in the same way as those in the sample solutions. Hg²⁺ reference solutions were prepared at concentrations ranging from 0.15 to 5.0 µ L^-1 in 1% v v^-1 HNO₃ and 0.3% v v^-1 HCl, considering the sample preparation conditions. This ensured similar matrix conditions for all solutions and helped maintain ICP-MS instrument stability.

Health risk assessment

An estimated health risk assessment was made considering the total Hg concentrations in fish muscle found in this study. The estimated daily intake (EDI) was calculated by using equation 1:

(1) $$\text{EDI}\left(\mathrm{mg}\mathrm{Hg}{\mathrm{kg}}^{-1}\text{bw day}{}^{-1}\right)=\frac{\mathrm{C}\times \mathrm{IR}}{\mathrm{BW}}$$

where C is the Hg concentration (µg g^-1), IR the daily intake rate of fish (g), and BW the body weight (kg).

The EDI was calculated for each fish species collected using the total Hg concentration found and assuming a generic average body mass of 70 kg. For the daily intake rate of fish, data from the Brazilian Institute of Geography and Statistics (IBGE)³¹ was used, which reported that average daily intake of fish per person was 45 g in the North region, where the samples were collected. To measure a possible health risk associated with ingestion of these fish, the EDI values obtained were compared to a tolerable weekly intake established by the European Food Security Authority (EFSA)³² following the equation 2:

(2) $$\mathrm{HR}=\frac{\mathrm{EDI}}{\mathrm{RfD}}$$

where HR is the health risk and RfD is the reference dose established by EFSA (1.3 μg methylmercury kg^-1 bw per week). This assumption is based on the fact that most of the mercury present in fish occurs as methylmercury (EFSA).³² Values of HR lower than 1 means no health risks involved in the consumption of the fish considering a daily intake of 45 g. It is also noteworthy to mention that the RfD value was divided by 7 to express the tolerable intake as daily dose like the EDI values.

Results and Discussion

For total Hg determination in fish, the method described by Viana et al.³⁰ was employed, which is based on a FB CV ICP-MS. The proposed method has good linearity (y = 240862x + 1650, R² = 0.9999, where y is in counts and x in μg L^-1 Hg), precision better than 5% (n = 3), limits of detection (LOD) and of quantification (LOQ) of 0.008 and 0.012 µg g^-1, respectively. The LOD and LOQ methods were calculated from 10 measurements of a sample blank, in accordance with the IUPAC recommendation. For this, the mean of blank plus three and ten times the standard deviation was used. In addition, high sample throughput is achieved, allowing up to 90 measurements per h. Method accuracy was evaluated using the certified reference material dogfish muscle (DOLT-4). The obtained value (2.85 ± 0.07 µg g^-1) was in good agreement, at the 95% confidence level and according to the recommendations of the European Commission (Institute for Reference Materials and Measurements (IRMM)) (2.58 ± 0.22 µg g^-1).

The maximum permitted mercury concentrations in fish for human consumption vary depending on the legislation of each country or region. Limits of 1.0 μg g^-1 for predatory species and 0.5 μg g^-1 for non-predatory species are established in Brazil (ANVISA);²³ 0.5 μg g^-1 by the EFSA and EU regulations for most fish species, and 1.0 µg g^-1 for predatory species, alongside setting health-based guidance values like a tolerable weekly intake (TWI) for methylmercury (1.3 µg kg^-1 body mass) to protect vulnerable groups, balancing fish benefits (omega-3) with mercury risks, especially for pregnant women;³³ and the U.S. Environmental Protection Agency (USEPA) uses a safety threshold of 0.46 µg g^-1 to categorize fish for safe consumption.³⁴

As shown in Table 1, the fish species collected have different eating habits. Therefore, to better understand the Hg concentration present in fish, the species were divided into groups according to their eating habits, being carnivorous, omnivorous, herbivorous, piscivorous and scavengers, and classified as predatory or non-predatory. Among the five carnivorous fish species, which feed exclusively on animal protein, four showed Hg concentrations above the limits permitted by Brazilian legislation. The relatively high Hg concentrations observed in these species are associated with their trophic position, since Hg bioaccumulates and undergoes biomagnification along the trophic chain. In addition to eating habits, the environment where these fish species live and their characteristics must also be taken into account. It is reported that up to 30% of the total Hg present in river water can be in the form of CH₃Hg⁺.¹⁰ In fresh and surface water, the most common forms of inorganic Hg species are oxides and sulfides, whose relative occurrence depends on several factors. Accordingly, a study carried out on the distribution of Hg species in humid areas, when sulfide is found at concentrations higher than 20 μg L^-1, inorganic sulfide species are formed and affected by a variety of environmental conditions. At low levels of sulfide, complexes of Hg-DOM are predominant. However, when sulfide is present in concentrations around 0.32 ng L^-1, Hg-DOM and Hg-S complexes are observed. Therefore, these complexes could be related to CH₃Hg⁺ in the environment and are dependent on sulfide concentrations, dissolved organic matter, iron, among others.¹¹,¹² In addition, the Hg distribution in river sediments is related by the concentrations of organic carbon, iron, clay, redox potential, and other factors. Soils and sediments have a high Hg retention capacity mainly related to the high content of organic matter, such as humic and fulvic acids. Due to the stability of Hg²⁺ complexes with humic acids, Hg methylation reactions in soil are infrequent. However, erosion caused by rainwater can carry Hg present in the soil to surface waters, where methylation processes occur in sediments.¹³ Once introduced into the river water, Hg can be bioaccumulated by the biota and higher animals, reaching humans and other animals that consume these fish.¹⁴ In this sense, the Calophysus macropterus and Brachyplatystoma platynema fishes have the highest Hg concentration, probably due to its eating habit (carnivorous), size (and age) and for living close to the bottom of the river. This is in accordance with the results reported by Bastos et al.,²¹ who associated Hg concentrations with the feeding habits of the species and the environment in which the fish live. Mercury concentrations present in fish with a predatory and non-predatory eating habit are described in Table 2. From these predatory species, Cichla ocellaris is the fish species with the lowest Hg concentration. This can be attributed to this species being a carnivorous fish that lives close to the water surface, and in less contact with the higher Hg concentrations usually present in sediments. The results are in accordance with studies done by Lechler et al.,³⁵ who compared Hg levels in fish living near the water surface or sediment.

Table 2
Mercury concentration in fish species from Madeira River (Rondônia State, Brazil). Results represent the mean and standard deviation (SD) of n = 3. Estimated daily intake (EDI) and health risk (HR) were calculated based on the Hg concentrations

Trophic level	Fish species	Hg / (g g-1)	EDI / (mg Hg kg-1 bw day-1)	HR
Herbivorous	Semaprochilodus insignis Piaractus brachypomus Psectrogaster amazonica	1.14 ± 0.01 0.09 ± 0.01 < 0.012	0.73 0.06 -	3.86 0.30 -
Omnivorous	Danio rerio Arapaima gigas Seriola dumerili Phractocephalus hemeliopterus Astronotus ocellatus Colossoma macropomum Piaractus mesopotamicus	5.15 ± 0.09 4.23 ± 0.04 3.63 ± 0.04 2.19 ± 0.01 2.07 ± 0.05 0.74 ± 0.01 0.15 ± 0.01	3.31 2.72 2.33 1.41 1.33 0.48 0.10	17.42 14.31 12.28 7.41 7.00 2.50 0.51
Piscivorous	Paulicea luetkeni Salminus maxillosus	20.7 ± 0.4 8.55 ± 0.07	13.31 5.50	70.04 28.93
Carnivorous	Brachyplatystoma platynema Salminus brasiliensis Pseudoplatystoma corruscans Osteoglossum bicirhossum Cichla ocellaris Calophysus macropterus	4.20 ± 0.01 2.46 ± 0.01 1.99 ± 0.01 1.24 ± 0.05 0.56 ± 0.01 8.37 ± 0.08	2.70 1.58 1.28 0.80 0.36 5.38	14.21 8.32 6.73 4.20 1.89 28.32

Omnivorous fishes have more diverse eating habits, which can be of animal or vegetable origin. Of the seven species studied, five showed Hg concentrations above the safe limits for human consumption, while the other two were below the permitted levels. The Hg concentrations present in these species are also shown in Table 2. As previously mentioned, the Hg concentration profile in carnivorous species is similar as for omnivores, i.e., fish that live close to the sediments have a higher Hg concentration than fish that live closer to the surface.

Species that have herbivorous eating habits have a diet based exclusively on vegetables. Of the three species evaluated, only one has the mercury concentration above that allowed by Brazilian legislation, which is 0.5 µg g^-1 for non-predator species. In relation to the other two species, Psectrogaster amazonica and Piaractus brachypomus, the mercury content is below the maximum permitted limit (Table 2). As can be observed, among all the fish species studied, only Psectrogaster amazonica has a Hg concentration below the LOQ value. Considering that it is a small species that lives close to the surface water and feeds exclusively on plant material, its low Hg concentration is expected.

Mercury concentration in fish species with piscivorous eating habits, that feed exclusively fish, are those with the highest Hg concentration, having average levels of 8.55 and 20.7 μg g^-1 Hg for the species Paulicea luetkeni and Salminus maxillosus, respectively. The two fish species have the same dietary characteristics, both living close to the sediments. In 2000, Maurice-Bourgoin et al.³⁶ found Hg concentrations for fish-eating species four-fold higher than those permitted by the safety limit stipulated by the World Health Organization in 1976. After 23 years, it is observed that Hg concentrations in these species tend to be higher, possibly due to the increase of the Hg concentration in the environment, mainly in consequence of the increase of illicit gold mining activity. According to Agência Brasileira de Inteligência,⁸ gold mining using Hg is concentrated mainly in the North and Central-West Regions of Brazil, particularly near water bodies in southern and southwestern Pará State, northwestern Mato Grosso State, Roraima State (especially in the Yanomami Indigenous Territory), Amapá State, northern Rondônia State (Madeira River basin), and western Amazonas State. These data confirm the reason for the increase in Hg in fish analyzed in the present study. Besides, Lacerda et al.⁵ compiled the data on Hg content in two fish species from the Madeira River, including the Cichla ocellaris species, in the period 1987-2022, and also pointed out that there was a five-fold increase in the concentration of the element.

Table 2 also presents the EDI and HR values calculated based on the Hg concentrations determined in the fish. Only 3 species of the 18 have a HR value below 1, while the other species have values up to 70, indicating a possible health risk of their consumers. As mentioned before, the daily intake rate of fish used to calculate the EDI was obtained from a search done with the whole North region of Brazil.³¹ However, there is a part of the population that lives by the riverside and rely heavily on fish consumption as the main protein source, and so will have a higher intake rate. Oliveira et al.¹⁸ studied riverside populations on the Madeira River, and observed that the mean daily consumption of fish was 406 g per person, clearly showing that these populations are much more exposed to the Hg present in fish. Another aspect related to the consumption is how the fish is processed beforehand. As reported in previous studies,³²,³⁷,³⁸ Hg concentrations may vary under different cooking processes; however, the total Hg content generally remains relatively unchanged. Therefore, it is reasonable to assume that Hg intake is similar to that from raw fish.

Regarding biomagnification, Azevedo et al.³⁹ observed that there is a significant variation in Hg concentration in relation to the trophic levels of four fish species, being Hg concentration of decreasing order in piscivores, detritivores and, lastly, herbivores fish. Vieira et al.⁴⁰ evaluated the concentration levels of Hg in liver and muscle of two piscivorous and non-carnivorous fish species. They found that piscivorous species had higher Hg content in the liver, being 1219 ± 15 and 1044 ± 13.6 μg kg^-1, respectively. In muscle, the determined Hg concentration was 101 ± 1.30 and 87.4 ± 0.9 μg kg^-1, respectively. However, non-predatory species presented relatively lower Hg concentrations in liver (852 ± 11.1 μg kg^-1) and muscle (71.4 ± 0.930 μg kg^-1). These results are in accordance with the results found in this study. Dorea et al.⁴¹ studied eleven fish species with different eating habits from the Negro River, in the Amazonas State (Brazil), and observed that the Hg concentration in these species is associated with their eating habits, size and the period of drought or flood of the river. They also observed that the Hg bioaccumulation is more associated with the feeding habits and size of the species than with the variation in the level of the river water. The variations in Hg concentrations were greater for predatory species than for non-predatory species, which is in agreement with the results obtained in the present study. Another relevant study was conducted by Silva and Lima,⁴² who analyzed seventeen fish species from the Alto Solimões region (Amazonas State) and observed CH₃Hg⁺ concentrations above the safety limit for consumption established by the WHO (0.50 mg kg^-1),²⁵ with higher concentrations found in predatory (carnivorous and piscivorous) species.

Due to the use of mercury in gold mining over the years, Hg contamination in fish is recurrent in the Madeira River.⁸ However, mercury mining occurs on a much larger scale, involving regions of the central-west and throughout the north of Brazil, areas of high socio-environmental sensitivity and affecting vulnerable populations subject to exposure through the consumption of water, fish, and the environment itself. According to data produced by MapBiomas Project,⁴³ there are gold mining concentrated near bodies of water in the south and southwest of Pará Sate, in the northwest of Mato Grosso Sate, in Roraima State (especially in the Yanomami Indigenous Territory), in Amapá Sate, in the north of Rondônia State (Madeira River basin), and in the west of Amazonas State. The populations residing in these areas are at higher risk of mercury contamination, since these locations tend to simultaneously concentrate small-scale gold extraction and riverside communities with high consumption of local fish. In addition to anthropogenic release of Hg, there is also natural release of the element through volatilization in biomass (forest) burning and other natural sources. So, the element is subject to atmospheric transport, deposition, and transformations (primarily into toxic methylmercury by microbes), leading to bioaccumulation in food webs, particularly in fish, and posing health risks. Key steps include atmospheric circulation, photooxidation, microbial methylation (Hg²⁺ to metil-Hg), re-emission as volatile elemental mercury (Hg⁰), and long-term sequestration in sediments, significantly increasing environmental levels. Pfeiffer et al.²⁰ found Hg concentrations in fish around five-fold higher than that permitted by Brazilian legislation. Taking into account the high amount of Hg released to the environment, the biogeochemical cycle of Hg in the aquatic environment and its ability to bioaccumulate and biomagnify, high Hg concentrations present in the analyzed samples are justified.⁴⁴ Therefore, Bastos et al.⁴⁵ evaluated the relationship between Hg concentrations in different fish species, in sediments and in the riverine population. The results indicate that species with carnivorous eating habits have a higher Hg concentration when compared to herbivorous ones. Furthermore, Hg concentrations in carnivorous and omnivorous species are above the limits permitted by current legislations, being also in accordance with the present study. In this sense, to understand more about Hg content in fish over time, a compilation of data from the literature, since the year 2000, of predatory and non-predatory fish species found in the Madeira River was considered (Table 3). It can be observed that Hg content enhanced for Arapaima gigas, Phractocephalus hemeliopterus, Astronotus ocellatus, Colossoma macropomum, Brachyplatystoma platynema, Pseudoplatystoma corruscans, and Calophysus macropterus, and decreased for Osteoglossum bicirhossum and Cichla ocellaris. It cannot be conclusively stated that the increase in Hg concentrations in most of the studied fish species is attributable to intensified gold mining activities, as a more comprehensive evaluation would be necessary, such as a comparison of the age and exact location of the fish capture, as well as their migration habits. Furthermore, it would be necessary to know other aspects, such as the Hg content in the river water and sediment. However, given the Hg contamination in the Madeira River and, consequently, the elevated Hg concentrations observed in fish species in this study, the riverside population (including breastfeeding children) are exposed to Hg. High levels of the element have been detected in hair, blood, urine, and breast milk as reported by Bastos et al.,⁴⁵ Marques et al.⁴⁶ and Ha et al.⁴⁷

Table 3
Data of Hg concentration in predatory and non-predatory fish species from Madeira River (Rondônia State, Brazil) from 2000 to 2025. Results represent the minimum and maximum Hg concentration

Author	Date of publication	Species (No. of samples)	Hg / (g g -1)
Boischio et al.48	2000	predator (31) non-predator (11)	Nd-11.15 Nd-1.21
Gali et al.49	2005	predator (31) non-predator (12)	0.01-1.86 0.07-0.14
Bastos et al.45	2006	predator (6) non-predator (2)	0.046-0.414 0.083-0.095
Bastos et al.50	2007	predator (4) non-predator (1)	0.354-1.053 0.098-0.116
Bastos et al.51	2008	predator (49) non-predator (15)	0.002-1.697 0.001-0.898
Hacon et al.52	2014	predator (37) non-predator (29)	0.06-1.64 0.01-0.71
Braga et al.53	2015	predator (1) non-predator (0)	0.331
Bastos et al.54	2015	predator (37) non-predator (29)	0.010-1.759 0.010-0.109
Bastos et al.21	2015	predator (8) non-predator (0)	0.385-1.986
Vieira et al.55	2015	predator (1) non-predator (0)	0.435
Bastos et al.56	2016	predator (29) non-predator (12)	0.25-0.70 0.10-0.40
Vieira et al.57	2018	predator (0) non-predator (1)	0.132
Soares et al.19	2018	predator (1) non-predator (3)	0.13-0.29 0.04-0.22
Queiroz et al.58	2019	predator (1) non-predator (0)	0.378
Bittarello et al.59	2019	predator (2) non-predator (0)	0.054-0.068
Cavecci-Mendonça et al.60	2020	predator (0) non-predator (2)	0.035-0.085
Bataglioli et al.61	2022	predator (1) non-predator (0)	0.153
Vieira et al.62	2022	predator (1) non-predator (0)	0.060
Costa et al.63	2022	predator (1) non-predator (0)	1.09
Vieira et al.40	2023	predator (2) non-predator (1)	0.0874-0.101 0.0714
Canela et al.64	2024	predator (3) non-predator (6)	0.40-6.54 0.01-0.85

Conclusions

Among the 18 fish species studied from the Madeira River, 13 showed total Hg concentrations above the limits permitted by current Brazilian legislation, which establishes maximum levels of 0.5 and 1.0 µg g^-1 for non-predatory and predatory species, respectively. It is also shown that of the 18 fish species studied, 15 present health risk (HR) due to the presence of Hg.

The results of this study are significant, as they confirm an increasing trend in Hg concentrations in the analyzed fish species and emphasize the worrying environmental situation in the region. This trend is supported by comparisons between Hg concentrations reported since 2000 and those observed in the present study. Therefore, these data collaborate and complement existing data. However, due to the biogeochemical cycle of Hg and the increase in gold extraction in the Madeira River, it is important to investigate Hg levels in sediments, water and in the riverside population. Another important aspect to investigate is related to the presence of Hg species, such as Hg⁰, Hg²⁺, Hg₂²⁺ and CH₃Hg⁺ present in this environment, since the largest source of anthropogenic emissions is probably elemental mercury. In this way, it is possible to evaluate the relationships among the concentration and species of Hg present in the environment compartments, mainly water and sediment, and verify whether bioaccumulation and biomagnification occur. To this end, a periodic and more ample study is necessary.

The FB-CV-ICP-MS method used for total Hg determination presented good performance, since it is accurate and precise, and LOD meets the values required by current national and international legislation.

Acknowledgments

The authors are grateful to CAPES and CNPq (Process No. 310228/2021-2 for V. L. D. and 131372/2021-1 for A. B. V.) for financial support and scholarship for the authors.

Data Availability Statement

Data supporting of this study are available from the corresponding author upon reasonable request.

References

• ¹ Outridge, P. M.; Mason, R. P.; Wang, F.; Guerrero, S.; Heimbürger-Boavida, L. E.; Environ. Sci. Technol. 2018, 52,11466. [Crossref]
  » Crossref
• ² Pestana, I. A.; Rezende, C. E. D.; Almeida, R.; Lacerda, L. D. D.; Bastos, W. R.; Extr. Ind. Soc. 2022, 11, 101122. [Crossref]
  » Crossref
• ³ Basta, P. C.; Cad. Saúde Pública 2023, 39, e00111823. [Crossref]
  » Crossref
• ⁴ Ministério do Meio Ambiente e Mudança do Clima; Relatório final Projeto Avaliação Inicial da Convenção de Minamata sobre Mercúrio, 2023. [Link] accessed in December 2025
  » Link
• ⁵ Lacerda, L. D.; Almeida, R.; Bastos, W. R.; Toxics 2024, 12, 144. [Crossref]
  » Crossref
• ⁶ Silva, H. A. M; Guimarães, J. R. D.; Chemosphere 2024, 350, 141054. [Crossref]
  » Crossref
• ⁷ Bastos, W. R.; Lacerda, L. D.; Geochim. Bras. 2004, 18, 99. [Link]
  » Link
• ⁸ Agência Brasileira de Inteligência, Fórum Brasileiro de Segurança Pública e Ministério do Meio Ambiente e Mudança do Clima; Mercúrio na Amazônia: Redes Criminosas, Vulnerabilidade Socioambiental e Desafios para a Governança, 2025. [Link] accessed in December 2025
  » Link
• ⁹ Bjørklund, G.; Dadar, M.; Mutter. J.; Aaseth, J.; Environ. Res. 2017, 159, 545. [Crossref]
  » Crossref
• ¹⁰ Horvat, M.; Gibičar, D. In Handbook of Elemental, vol. 2; Cornelis, R.; Caruso, J.; Crews, H.; Heumann, K., eds.; John Wiley & Sons: Chichester, 2005, p. 281-304. [Crossref]
  » Crossref
• ¹¹ Zi llioux, E. J.; Environ. Indic. 2015, 743. [Crossref]
  » Crossref
• ¹² Jiang, P.; Liu, G.; Cui, W.; Cai, Y.; Sci. Tot. Environ. 2018, 640 641, 454. [Crossref]
  » Crossref
• ¹³ Bisinoti, M. C.; Jardim, W. F.; Quim Nova 2004, 27, 593. [Crossref]
  » Crossref
• ¹⁴ Araújo, B. F.; Almeida, M. G. D.; Salomão, M. S. M. B.; Gobo, R. R.; Siqueira, V. C.; Ovalle, A. R. C.; de Rezende, C. E.; Química Nova 2010, 33, 501. [Crossref]
  » Crossref
• ¹⁵ Khoury, E. D. T.; Souza, G. D. S.; Silveira, L. C. D. L.; Costa, C. A. D.; Araújo, A. A. D.; Pinheiro, M. D. C. N.; Cad. de Saúde Pública 2013, 29, 2307. [Crossref]
  » Crossref
• ¹⁶ Zahir, F.; Rizwi, S. J.; Haq, S. K.; Khan, R. H.; Environ. Toxicol. Pharmacol. 2005, 20, 351. [Crossref]
  » Crossref
• ¹⁷ Agency for Toxic Substances and Disease Registry (ATSDR), the Public Health Service; Toxicological Profile for Mercury, 2024. [Link] acessed in December 2025
  » Link
• ¹⁸ Oliveira, R. C.; Dórea, J. G.; Bernardi, J. V. E.; Bastos, W. R.; Almeida, R.; Manzatto, Â. G.; Ann. Hum. Biol. 2010, 37, 629. [Crossref]
  » Crossref
• ¹⁹ Soares. J. M.; Gomes, J. M.; Anjos, M. R.; Silveira, J. N.; Custódio, F. B.; Gloria, M. B. A.; Food Res. Int. 2018, 109, 537. [Crossref]
  » Crossref
• ²⁰ Pfeiffer, W. C.; Malm, O.; Souza, C. M. M.; Lacerda, L. D.; Silveira, E. G.; Bastos, W. R.; For. Ecol. Manage. 1991, 38, 239. [Crossref]
  » Crossref
• ²¹ Bastos, W. R.; Dórea, J. G.; Bernardi, J. V.; Lauthartte, L. C.; Mussy, M. H.; Lacerda, L. D.; Malm, O.; Environ. Res. 2015, 140, 191. [Crossref]
  » Crossref
• ²² Paiva, T. D.C.; Dary, E. P.; Pestana, I. A.; Amadio, S. A.; Malm, O.; Kasper, D.; Environ. Res. 2022, 215, 114307. [Crossref]
  » Crossref
• ²³ Agência Nacional de Vigilância Sanitária (ANVISA); Resolução da Diretoria Colegiada (RDC) No. 12, de 2 de janeiro de 2001, aprova o Regulamento Tiéccon sobre Padrões Microbiologicos para Alimentos, 2021. [Link] accessed in November 2025
  » Link
• ²⁴ European Commission, Official Journal of the European Communities. Regulation (EC) No. 1881/2006 of the European Parliament and of the Council of 25 April 2023. [Link] accessed in November 2025
  » Link
• ²⁵ Joint FAO/WHO Expert Committee on Food Additives and Contaminants; Food additives and Food Contaminants FAO Procedural Guidelines for the Joint FAO/WHO Expert Committee on Food Additives, 2003. [Link] accessed in November 2025
  » Link
• ²⁶ Dressler, V. L.; Flores, E. M. M.; Pozebon, D.; Kaercher, L. E.; J. Anal. At. Spectrom. 2002, 17, 790. [Crossref]
  » Crossref
• ²⁷ Zheng, C.; Li, Y.; He, Y.; Qian, M.; Hou, X.; J. Anal. At. Spectrom. 2005, 20, 746. [Crossref]
  » Crossref
• ²⁸ Wu, L.; Zheng, C.; Ma, Q.; Hu, C.; Hou, X.; Appl. Spectrosc. Rev. 2007, 42, 79. [Crossref]
  » Crossref
• ²⁹ Proch, J.; Niedzielski. P.; Talanta 2022, 243, 123372. [Crossref]
  » Crossref
• ³⁰ Viana, A. B.; Pappis. C.; Pereira, C. K.; Dressler, V. L.; Talanta 2020, 279, 126600. [Crossref]
  » Crossref
• ³¹ Instituto Brasileiro de Geografia e Estatística (IBGE); Pesquisa de Orçamentos Familiares: 2017-2018: Análise do Consumo Alimentar Pessoal no Brasil; IBGE: Rio de Janeiro, Brasil, 2020. [Link] accessed in December 2025
  » Link
• ³² EFSA Panel on Contaminants in the Food Chain (CONTAM); EFSA J. 2012, 10, 2985. [Crossref]
  » Crossref
• ³³ European Food Safety Authority (EFSA); EFSA J 2004, 34, 1. [Crossref]
  » Crossref
• ³⁴ Borum, D.; Manibusan, M. K.; Schoeny, R.; Winchester, E. L.; Water Quality Criterion for The Protection of Human Health: Methylmercury; U. S. Environmental Protection Agency (USEPA): USA, 2001. [Link] accessed in November 2025
  » Link
• ³⁵ Lechler, P. J.; Miller, J. R.; Lacerda, L. D.; Vinson, D.; Bonzongo, J. C.; Lyons, W. B.; Warwick, J. J.; Sci. Tot. Environ. 2000, 260, 87. [Crossref]
  » Crossref
• ³⁶ Maurice-Bourgoin, L.; Quiroga. I.; Chincheros, J.; Courau, P.; Sci. Tot. Environ. 2000, 260, 73. [Crossref]
  » Crossref
• ³⁷ Ouédraogo, O.; Amyot, M.; Environ. Res. 2011, 111, 1064. [Crossref]
  » Crossref
• ³⁸ Ribeiro, M.; Douis, L; da Silva, J. A. L.; Castanheira, I.; Leufroy, A.; Jitaru, P.; Foods 2024, 13, 374. [Crossref]
  » Crossref
• ³⁹ Azevedo, L. S.; Pestana, I. A.; Almeida, M. G.; Nery, A. F. C.; Bastos, W. R.; Souza, C. M. M.; Ecotoxicol. Environ. Saf. 2021, 207, 111249. [Crossref]
  » Crossref
• ⁴⁰ Vieira, J. C. S.; Braga, C. P.; Queiroz, J. V.; Cavecci-Mendonça, B.; Oliveira, G.; Freitas, N. G.; Fernandes, A. A. H.; Fernandes, M. S.; Buzalaf, M. A. R.; Adamec, J.; Zara, L. F.; Padilha, P. M.; Chemosphere 2023, 316, 137779. [Crossref]
  » Crossref
• ⁴¹ Dorea, J. G.; Barbosa, A. C.; Silva, G. S.; Comp. Biochem. Physiol. C: Toxicol. Pharmacol. 2006, 143, 275. [Crossref]
  » Crossref
• ⁴² Silva, S. F.; Lima, M. D. O.; Chemosphere 2020, 248, 125989. [Crossref]
  » Crossref
• ⁴⁴ Neto, O. G. N.; Dias, S. R.; Albuquerque, F. E. A.; Miranda, M.; Lopez-Alonso, M.; Oliveira, R. B.; Pinto, D.; Minervino, A. H. H.; Chemosphere 2024, 351, 141146. [Crossref]
  » Crossref
• ⁴³ MapBiomas; Proximidade de Garimpo, Rios e Lagos na Amazônia, 2024. [Link] accessed in February 2026
  » Link
• ⁴⁵ Bastos, W. R.; Gomes, J. P.; Oliveira, R. C.; Almeida, R.; Nascimento, E. L.; Bernardi, J. V.; Lacerda, L. D.; da Silveira, E. G.; Pfeffer, W. C.; Sci. Total Environ. 2006, 368, 344. [Crossref]
  » Crossref
• ⁴⁶ Marques, R.; Dórea, J. G.; Bastos, W. R.; Rebelo, M. D. F.; Fonseca, M. D. F.; Malm, O.; Int. J. Hyg. Environ. Health. 2007, 210, 51. [Crossref]
  » Crossref
• ⁴⁷ Há, E.; Basu, N.; Bose-O’Reilly, S.; Dórea, J. G.; McSorley, E.; Sakamoto, M.; Chan, H. M.; Environ. Res. 2017, 152, 419. [Crossref]
  » Crossref
• ⁴⁸ Boischio, A. A. P.; Henshe, D.; Environ. Res. 2000, 84, 108. [Crossref]
  » Crossref
• ⁴⁹ Gali, P. A. S.; Bonotto, D. M.; Silveira, E. G.; Bastos, W. R.; Mercury in Amazonian Fish from Madeira River Basin, Rondônia State, vol. 9; Environmental Health Risk II, WIT Transactions on Biomedicine and Health: Brazil, 2005.
• ⁵⁰ Bastos, W. R.; Almeida, R.; Dórea, J. G.; Barbosa, A. C.; Ecotox 2007, 16, 341. [Crossref]
  » Crossref
• ⁵¹ Bastos, W. R.; Rebelo, M. F.; Fonseca, M. F.; Almeida, R.; Malm, O.; Acta Amaz. 2008, 38, 431. [Crossref]
  » Crossref
• ⁵² Hacon, S. S.; Dórea, J. G.; Fonseca, M. F.; Oliveira, B. A.; Mourão, D. S.; Ruiz, C. M. V.; Gonçalves, R. A.; Mariani, C. F.; Bastos, W. R.; Int. J. Environ. Res. Public Health 2014, 11, 2437. [Crossref]
  » Crossref
• ⁵³ Braga, C. P.; Bittarello, A. C.; Padilha, C. C. F.; Leite, A. L.; Moraes, P. M.; Buzalaf, M. A. R.; Zara, L. F.; Padilha, P. M.; Talanta 2015, 132, 239. [Crossref]
  » Crossref
• ⁵⁴ Bastos, W. R.; Dórea, J. G.; Bernardi, J. V. E.; Lauthartte, L. C.; Mussy, M. H.; Hauser, M.; Dória, C. R. C.; Malm, O.; Ecotox. Environ. Saf. 2015, 118, 90. [Crossref]
  » Crossref
• ⁵⁵ Vieira, J. C. S.; Cavecci, B.; Queiroz, J. V.; Braga, C. P.; Padilha, C. C. F.; Leite, A. L.; Figueiredo, W. S.; Buzalaf, M. A. R.; Zara, L. F.; Padilha, P. M.; Arch. Environ. Contam. Toxicol. 2015, 69, 422. [Crossref]
  » Crossref
• ⁵⁶ Bastos, W. R.; Dórea, J. G.; Bernardi, J. V. E.; Manzatto, A. G.; Mussy, M. H.; Lauthartte, L. C.; Lacerda, L. D.; Malm, O.; Environ. Res. 2016, 144 Part A, 73. [Crossref]
  » Crossref
• ⁵⁷ Vieira, J. C. S.; Braga, C. P.; Oliveira, G.; Padilha, C. C. F.; Moraes, P. M.; Zara, L. F.; Leite, A. L.; Buzalaf, M. A. R.; Padilha, P. M.; Biol. Trace Elem. Res. 2018, 183, 164. [Crossref]
  » Crossref
• ⁵⁸ Queiroz, J. V.; Vieira, J. C. S.; Oliveira, G.; Braga, C. P.; Bataglioli, I. C.; Silva, J. M.; Araújo, W. L. P.; Padilha, P. M.; Biol. Trace Elem. Res. 2019, 187, 291. [Crossref]
  » Crossref
• ⁵⁹ Bittarello, A. C.; Vieira, J. C. S.; Braga, C.; Araújo, W. L. P.; Bataglioli, I. C.; Silva, J M.; Buzalaf, M. A. R.; Fleuri, L. F.; Padilha, P. M.; Food Chem. 2019, 276, 247. [Crossref]
  » Crossref
• ⁶⁰ Cavecci-Mendonça, B.; Vieira, J. C. S.; Lima, P. M.; Leite, A. L.; Buzalaf, M. A. R.; Zara, L. F.; Padilha, P. M.; Food Chem. 2020, 309, 125460. [Crossref]
  » Crossref
• ⁶¹ Bataglioli, I. C.; Queiroz, J. V.; Vieira, J. C. S.; Cavalline, N. G.; Braga, C. P.; Buzalaf, M. A. R.; Zara, L. F.; Adamec, J.; Padilha, P. M.; Environ. Monit. Assess. 2022, 194, 705. [Crossref]
  » Crossref
• ⁶² Vieira, J. C. S.; Oliveira, G.; Cavallini, N. G.; Braga, C. P.; Adamec, J.; Zara, L. F.; Buzalaf, M. A. R.; Padilha, P. M.; Biol. Trace Elem. Res. 2022, 200, 1872. [Crossref]
  » Crossref
• ⁶³ Costa, I. D.; Nascimento, E. L.; Faccheti, M. S. A.; Nunes, N. T. S.; Gomes, J. P. O.; de Almeida, R.; Bastos, W. R.; Acta Amaz. 2022, 52, 60. [Crossref]
  » Crossref
• ⁶⁴ Canela, T. A.; Monteiro, L. C.; Cabral, C. S.; Ximenes, F. S.; Oliveira, I. A. S.; Bernardi, J. V. E.; de Almeida, R.; Bastos, W. R.; Toxics 2024, 12, 208. [Crossref]
  » Crossref

Edited by

• Editor handled this article:
  Josué Carinhanha Caldas Santos (Associate)

Publication Dates

• Publication in this collection
  18 May 2026
• Date of issue
  2026

History

• Received
  27 Nov 2025
• Accepted
  02 Feb 2026
• Published
  25 Feb 2026