Environ Sci Pollut Res (2009) 16:10–24 DOI 10.1007/s11356-008-0038-8
AREA 2 • AQUATIC CHEMISTRY AND BIOLOGY, HEALTH • RESEARCH ARTICLE
The relationships between mercury and selenium in plankton and fish from a tropical food web Helena do A. Kehrig & Tércia G. Seixas & Elisabete A. Palermo & Aida P. Baêta & Christina W. Castelo-Branco & Olaf Malm & Isabel Moreira
Received: 1 February 2008 / Accepted: 8 August 2008 / Published online: 27 August 2008 # Springer-Verlag 2008
Abstract Background, aim, and scope Selenium (Se) has been shown to reduce mercury (Hg) bioavailability and trophic transfer in aquatic ecosystems. The study of methylmercury (MeHg) and Se bioaccumulation by plankton is therefore of great significance in order to obtain a better understanding of the estuarine processes concerning Hg and Se accumulation and biomagnification throughout the food web. In the western South Atlantic, few studies have documented trace element and MeHg in fish tissues. No previous study about trace elements and MeHg in plankton has been conducted concerning tropical marine food webs. Se, Hg, and MeHg were determined in two size classes of plankton, microplankton (70–290 μm) and mesoplankton (≥290 μm), and also in muscle tissues and livers of four fish species of different trophic levels (Mugil liza, a planktivorous fish; Bagre spp., an omnivorous fish; Micropogonias furnieri, a benthic carnivorous fish; and Centropomus undecimalis, a pelagic carnivorous fish) from a polluted estuary in the Brazilian Southeast coast, Guanabara Bay. Biological and Responsible editor: Lee Young H. A. Kehrig (*) : T. G. Seixas : E. A. Palermo : O. Malm Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, 21941-902 Rio de Janeiro, Brazil e-mail:
[email protected] T. G. Seixas : A. P. Baêta : I. Moreira Departamento de Química, Pontifícia Universidade Católica do Rio de Janeiro, 22453-900 Rio de Janeiro, Brazil C. W. Castelo-Branco Departamento de Zoologia, CCBS, Universidade do Rio de Janeiro, 22290-240 Rio de Janeiro, Brazil
ecological factors such as body length, feeding habits, and trophic transfer were considered in order to outline the relationships between these two elements. The differences in trace element levels among the different trophic levels were investigated. Materials and methods Fish were collected from July 2004 to August 2005 at Guanabara Bay. Plankton was collected from six locations within the bay in August 2005. Total mercury (THg) was determined by cold vapor atomic absorption spectrometry (CV-AAS) with sodium borohydride as a reducing agent. MeHg analysis was conducted by digesting samples with an alcoholic potassium hydroxide solution followed by dithizone-toluene extraction. MeHg was then identified and quantified in the toluene layer by gas chromatography with an electron capture detector (GCECD). Se was determined by AAS using graphite tube with Pin platform and Zeeman background correction. Results and discussion Total mercury, MeHg, and Se increased with plankton size class. THg and Se values were below 2.0 and 4.8 μg g−1 dry wt in microplankton and mesoplankton, respectively. A large excess of molar concentrations of Se in relation to THg was observed in both plankton size class and both fish tissues. Plankton presented the lowest concentrations of this element. In fish, the liver showed the highest THg and Se concentrations. THg and Se in muscle were higher in Centropomus undecimalis (3.4 and 25.5 nmol g−1) than in Micropogonias furnieri (2.9 and 15.3 nmol g−1), Bagre spp (1.3 and 3.4 nmol g−1) and Mugil liza (0.3 and 5.1 nmol g−1), respectively. The trophic transfer of THg and Se was observed between trophic levels from prey (considering microplankton and mesoplankton) to top predator (fish). The top predators in this ecosystem, Centropomus undecimalis and Micropogonias furnieri, presented similar MeHg concentrations in muscles and liver. Microplankton pre-
Environ Sci Pollut Res (2009) 16:10–24
sented lower ratios of methylmercury to total mercury concentration (MeHg/THg) (34%) than those found in mesoplankton (69%) and in the muscle of planktivorous fish, Mugil liza (56%). The other fish species presented similar MeHg/THg in muscle tissue (of around 100%). M. liza showed lower MeHg/THg in the liver than C. undecimalis (35%), M. furnieri (31%) and Bagre spp. (22%). Significant positive linear relationships were observed between the molar concentrations of THg and Se in the muscle tissue of M. furnieri and M. liza. These fish species also showed significant inverse linear relationships between hepatic MeHg and Se, suggesting a strong antagonistic effect of Se on MeHg assimilation and accumulation. Conclusions Differences found among the concentrations THg, MeHg, and Se in microplankton, mesozooplankton, and fishes were probably related to the preferred prey and bioavailability of these elements in the marine environment. The increasing concentration of MeHg and Se at successively higher trophic levels of the food web of Guanabara Bay corresponds to a transfer between trophic levels from the lower trophic level to the top-level predator, suggesting that MeHg and Se were biomagnified throughout the food web. Hg and Se were positively correlated with the fish standard length, suggesting that larger and older fish bioaccumulated more of these trace elements. THg, MeHg, and Se were a function of the plankton size. Recommendations and perspectives There is a need to assess the role of selenium in mercury accumulation in tropical ecosystems. Without further studies of the speciation of selenium in livers of fishes from this region, the precise role of this element, if any, cannot be verified in positively affecting mercury accumulation. Further studies of this element in the study of marine species should include liver samples containing relatively high concentrations of mercury. A basin-wide survey of selenium in fishes is also recommended. Keywords Biomagnification . Feeding habits . Fish tissues . Mercury–selenium relationship . Methylmercury . Plankton . Trophic transfer . Tropical estuary
1 Background, aim, and scope Plankton may significantly mediate the transport and cycling of trace elements in coastal waters. Plankton is an integral part of coastal food webs, since it is a food source for many juvenile and adult fish. It could also be a dietary source of mercury and selenium (Pukerson et al. 2003). Trophic transfer of trace elements along marine food webs has been recognized as an important process
11
influencing bioaccumulation and geochemical cycling of many elements (Fisher and Reinfelder 1995). It has been shown that food web accumulation is the major pathway for selenium and mercury, as well as methylmercury bioaccumulation in aquatic biota (Baldwin et al. 1996; Watras et al. 1998; Mason et al. 2000; Barwick and Maher 2003). The trophic level is thus important; suggesting that bioaccumulation of trace elements may be due to the feeding habits of organisms in each level (Turner and Swick 1983). Further, tertiary consumers (e.g., piscivorous fish) will have higher concentrations of pollutants than primary or secondary consumers (e.g., species that feed on benthic diatoms or invertebrates) (Varanasi et al. 1994). The analysis of different trophic levels can inform about the bioaccumulation and biomagnification processes that may occur, even if the indicated species do not prey directly on each other (Costa et al. 2004). The presence and behavior of mercury in aquatic systems is of great interest and importance, since it is the only heavy metal that bioaccumulates and biomagnifies through the aquatic food web (Lindqvist et al. 1991), thus affecting productivity, reproduction, and survival of coastal and marine animals, and which can, eventually, be hazardous to humans (WHO 1989). Methylmercury is largely responsible for the accumulation of mercury in organisms and the transfer of mercury from one trophic level to another. However, only limited information is available on the contaminant loads of tropical biota found in southern hemisphere waters. Almost all the mercury in marine fish muscles is methylated as methylmercury. However, the major part of mercury accumulated in their internal organs, especially the liver, is inorganic mercury, suggesting that demethylation is possible (Joiris et al. 1999; Zhang et al. 2001; Kehrig et al. 2004). Mercury, which accumulates during growth, is exogenous and harmful. Conversely, selenium (Se) is recognized as an essential micronutrient for animals, active in the activities of enzymes. It has been reported that the liver of aquatic organisms may act as an organ for demethylation and/or sequestration of both organic and inorganic forms of mercury (Wagemann et al. 2000), and that selenium is involved in both of these mechanisms (Ikemoto et al. 2004). Literature reports that selenium presented in selenoproteins is involved in both demethylation and immobilization of mercury in aquatic organisms, especially in mammals, suggesting mercury selenide (HgSe) formation (Nigro and Leonzio 1996). According to Joiris et al. (2001), the accumulation of high concentrations of inorganic mercury in the liver is probably due to a slow demethylation process, implied in the formation of HgSe (mercuric selenide).
12
In the livers of some marine mammals and seabirds, the molar ratio between selenium and mercury is close to equimolar (Nigro and Leonzio 1996; Ikemoto et al. 2004), but a large excess of selenium in relation to mercury has been observed in fish (Nigro and Lenzio 1996; Dietz et al. 2000). Less information is available about fish muscle tissues (Chen et al. 2001; Jin et al. 2006). Even more limited information is available about the manner in which mercury and selenium interact within tropical marine food webs (Kehrig et al. 2004; Kojadinovic et al. 2007). Mercury and selenium in fish tissues are expected to vary in a wide range of concentration, reflecting feeding behavior and exposure to environmental levels (Reinfelder et al. 1998). However, mercury and selenium content in fish varies greatly among species from the same location. Tissue mercury and selenium concentrations are instead controlled by the dietary habits and size of an organism, the physical and chemical characteristics of the system they inhabit, and the magnitude of human activities in the catchment (Harris et al. 2008). The study of mercury, mainly as methylmercury, and selenium bioaccumulation by marine plankton is therefore of great significance for reaching a better understanding of the processes concerning Hg and Se accumulation and biomagnification throughout an estuarine food web. The present study evaluated selenium, total mercury and methylmercury concentrations in two size classes of plankton, microplankton and mesoplankton (70–290 μm and ≥290 μm, respectively), and in muscle tissues and livers of four fish species (Mugil liza, Bagre spp., Micropogonias furnieri, and Centropomus undecimalis) from Guanabara Bay. This study also assessed the inter-element relationships in the planktonic community and fish tissues and their possible influence on trace element bioaccumulation throughout the food web. Biological and ecological factors, such as standard length and feeding habits, were considered. The fish species, Lebranche mullet (Mugil liza— planktivorous fish), catfish (Bagre spp.—omnivorous fish), whitemouth croaker (M. furnieri—benthic carnivorous fish), and common snook (C. undecimalis—pelagic carnivorous fish), were chosen based on their various feeding and ecological habits within Guanabara Bay. This estuary has been the object of numerous studies, some of them dealing with mercury and methylmercury (Kehrig et al. 1998, 2001, 2006; Baêta et al. 2006) and selenium (Seixas et al. 2007) in the aquatic biota. However, no previous data on mercury, selenium, and methylmercury concentrations in plankton from tropical marine food webs have yet been reported. This is the first work to approach selenium and its relationships with mercury and methylmercury, comparing their concentrations in plankton and fish of a tropical estuary.
Environ Sci Pollut Res (2009) 16:10–24
2 Materials and methods 2.1 Study area and sampling Guanabara Bay (22°S, 43°W) is a eutrophic bay (384 km2) in the southern Brazilian coast (Fig. 1). However, the halflife of its water turn-over (T50%) is relatively short, being 11.4 days (Kjerfve et al. 1997), and exemplified by the good water quality of the deep central channel, which connects the bay to the ocean in a north–south direction. Tides at Guanabara Bay are mainly semidiurnal. Its hydrobiology exhibits seasonal trends as well as spatial patterns associated with tidal currents, which create typical horizontal and vertical gradients (Kjerfve et al. 1997). The local climate is tropical humid, characterized by a dry season in winter and a rainy one in summer. Despite the intense pollution, habitat degradation, and the reduced number of species available in commercial quantities, Guanabara Bay is the most important estuary for fish production on the Southeastern Brazilian coast. Its ecosystem is impacted by the polluted discharge from the Rio de Janeiro metropolitan area. It receives untreated domestic and industrial sewage from a densely populated area, and from the second largest industrialized region in Brazil, with approximately 10,000 industrial plants, two harbors, shipyards, and oil terminals. In some areas of the bay, the ecosystem is heavily impacted by organic matter, oil and heavy metals, whose main consequences are elevated concentrations of toxic metals and hydrocarbons in sediments and changes in the pelagic and benthic communities (Valentin et al. 1999). Water and planktonic organisms were sampled from approximately 0.5–1.0 m of depth in relation to the water surface of Guanabara Bay, at the same time on 11 August 2005. Sampling was performed at neap tide cycle during the day at five sampling points inside, and one outside, Guanabara Bay (see Fig. 1). Tide variations at Guanabara Bay, obtained on 11 August 2005, are presented in Table 1. Water samples were collected using 5-l VanDorn bottles. Surface temperature, salinity, pH, and dissolved O2 (DO) were measured using an electronic switch gear, type MC5 probe, London. Secchi depth was measured with a 20-cm diameter Secchi disk. For laboratory analysis, samples were collected at each site in plastic bottles, which were previously washed with phosphorusfree detergent and cleaned with diluted hydrochloric acid (10%) in deionized water. Just before filling these bottles, they were washed with local water at each sampling point. Subsamples of water were filtered through a precombusted GF/C Whattman membrane to separate the dissolved and particulate fractions, and stored on ice. In the laboratory,
Environ Sci Pollut Res (2009) 16:10–24
13
Fig. 1 Map of study area, showing plankton sampling sites
22 ˚40’S
Brazil Rio d de Janeiro State
Praia de Mau Mau·á
Guanabara Bay
3
N
4 2
5
1
6
23 ˚05’W 43 ˚40’W
43 ˚00’W
the concentration of dissolved ammonium was determined following Grasshoff et al. (1983). The chlorophyll concentration was determined in the membranes as soon as the samples arrived at the laboratory, following the procedure of Strickland and Parsons (1972). A summary Table 1 Tide variations at Guanabara Bay, Rio de Janeiro State, obtained on 11 August 2005 Tide Variation Time (LST) 06: 10: 13: 14: 18: 22:
11 45 00 17 39 45
Height (m) 1.0 0.5 0.6 0.5 0.8 0.6
of a number of physical and chemical parameters of water from Guanabara Bay on 11 August 2005 is presented in Table 2. The plankton samples were taken at the six sampling points, by horizontal hauls at the water surface, using planktonic conical nets of 70- and 29-µm mesh size, aiming to separate microplankton and mesoplankton, respectively. For identification of the taxonomic groups, subsamples of plankton were preserved in 4% buffered formalin. The quantitative and qualitative analyses of planktonic organisms were performed in a Sedgewick–Rafter counting cell using an electronic microscope (BX-Olympus). Organisms were counted and separated by taxonomic groups. For chemical analysis, the plankton samples were stored alive in pre-cleaned Teflon bottles, identified per site, and separated as microplankton and mesoplankton samples. At the laboratory, these samples were freeze-dried and stored in hermetic vessels until chemical analysis.
14
Environ Sci Pollut Res (2009) 16:10–24
Table 2 Summary of some physical and chemical parameters of the water from Guanabara Bay, on 11 August 2005 Na Secchi depth (m) Temperature (°C) Salinity pH DO (mg l−1) Total Chlorophyll (µg l−1) N–NH4 + (µM) Mean (min–max) Mean (min–max) Mean (min–max) Mean (min–max) Mean (min–max) Mean (min–max) Mean (min–max) 6 a
3.4 (1.45–4.20)
22.4 (22.2–22.7)
32.1 (29–34.7)
8.16 (7.97–8.20)
6.8 (6.2–7.0)
16.35 (6.58–27.18)
10.5 (0.6–46.9)
Number of sampling points
The fish specimens were obtained from the fishing community of Praia de Maua (see Fig. 1), from fishermen within the bay, using a variety of fishing techniques. Following the determination of body weight, standard length, and species identification, a skinless cube of the white dorsal muscle tissue and the whole liver were removed. These tissues were stored in airtight plastic bags at −18°C until freeze-dried. A total of 35 specimens of M. liza (Lebranche mullet—planktivorous fish), 14 specimens of Bagre spp. (catfish—omnivorous fish), 34 specimens of M. furnieri (whitemouth croaker—benthic carnivorous fish), and six specimens of C. undecimalis (common snook—pelagic carnivorous fish) were collected and analyzed. All these fish species are characteristic from the western South Atlantic coastal waters and occur abundantly along the whole Brazilian coast. However, among demersal fish, croakers, mullets, and catfishes comprised the main part of the fish captured in Guanabara Bay (Jablonski et al. 2006). Total mercury (THg), methylmercury (MeHg), and selenium (Se) were determined in the dried samples of microplankton, mesoplankton, and fish (muscle and liver). The concentrations are presented in a dry weight basis (dry wt.). 2.2 Trace elements and methylmercury analysis The determination of total mercury (THg), methylmercury (MeHg), and selenium (Se) concentrations were performed in three aliquots of approximately 100 mg from each organ per fish specimen and also in the microplankton and mesoplankton samples. THg determinations were conducted by cold vapor atomic absorption spectrometry (CV-AAS) (FIMS-system) (Perkin-Elmer, USA) equipped with auto sampler AS90 (Perkin Elmer, USA), using sodium borohydride as a reducing agent (Kehrig et al. 2006, 2008). For MeHg analysis, aliquots of dried samples were digested with an alcoholic potassium hydroxide solution followed by dithizone-toluene extraction. MeHg was then identified and quantified in the toluene layer by gas chromatography with an electron capture detector (GC-ECD), using a Shimadzu gas chromatograph GC 14 (Kehrig et al. 2006, 2008). Aliquots of dried samples were acid mineralized and Se determination was made by graphite furnace atomic absorption spectrometry (GF-AAS), using an Analytic Jena Model ZEEnit 60 spectrometer (Analytik Jena, Germany)
with Zeeman Effect background correction equipped with an MPE-52 auto sampler. Palladium nitrate was used as a chemical modifier (Seixas et al. 2007). 2.3 Analytical quality control The precision and accuracy of the analytical methods were determined and monitored using certified reference materials obtained from the International Atomic Energy Agency (IAEA 350-Tuna fish sample, IAEA MA-1-copepod homogenate), National Research Council-Canada (DORM 2Dogfish muscle sample). Results for THg DORM 2 (N=9) and IAEA MA-1 (N=18) were 4.54±0.13 µg g−1 and 0.29± 0.03 μg g−1, whereas the CRM has a certified THg value of 4.64±0.26 µg g−1 and 0.28±0.02 μg g−1, respectively. Our routine methylmercury result for reference sample IAEA 350 (N=39) was of 3.59±0.38 µg g−1; whereas the CRM MeHg value was of 3.65±0.35 µg g−1. For selenium DORM 2 (N=18), results were of 1.47±0.27 µg g−1 and the CRM has a certified Se value of 1.40±0.09 µg g−1. The coefficient of variation (SD/mean) for the duplicate samples was less than 10%. Our trace elements and methylmercury results of the analysis of the certified reference samples demonstrated the high precision and accuracy of the analytical methods, where Hg, Se, and MeHg quantified in the reference materials were within 97% and 105% of the mean certified values. 2.4 Statistical analysis Statistical analyses were performed using STATISTICA® 6.0 for Windows (StatSoft, 1984–2001, USA). Nonparametric statistic tests were used. The results are presented as median and their range (min–max) values. The analysis of variance was done by Kruskal–Wallis ANOVA followed by a Post hoc test (Mann–Whitney U test) to compare the concentration of selenium, mercury, and methylmercury among the plankton and the fish tissues (muscle and liver), and the different feeding habits among the organisms. A multiple regression statistic (R2) was performed to verify the existing relationship between the standard length and the concentrations in the different tissues and, also, to verify the inter-element relationships (on molar basis) in organs (muscle and liver).
Environ Sci Pollut Res (2009) 16:10–24
3 Results Analyses of plankton revealed that the smaller size class of microplankton (70 μm mesh) from the six sampling locations within Guanabara Bay was mostly composed of phytoplankton, primarily diatoms (>70%), and protozooplankton (tintinnids). At these same six locations within the bay, the mesoplankton (≥290 μm) was composed mainly of mesozooplankton microcrustaceans, primarily copepods (85%). Results concerning the total mercury (THg), methylmercury (MeHg), and selenium (Se) in the two size classes of plankton showed that concentrations increased from the microplankton to the mesoplankton, indicating that THg, MeHg, and Se accumulated as a function of plankton size (Table 3). In fish, THg and Se concentrations varied between muscle and liver. Liver of all fish specimens showed higher molar concentrations of THg and Se than those found in their muscle tissue, presenting significant differences (p
