Article pubs.acs.org/est
Mercury in Wild Fish from High-Altitude Aquatic Ecosystems in the Tibetan Plateau Qianggong Zhang,†,‡ Ke Pan,†,§ Shichang Kang,‡,∥ Aijia Zhu,⊥ and Wen-Xiong Wang*,§ ‡
Key Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences (CAS), Building 3, Courtyard 16, Lin Cui Road, Chaoyang District, Beijing 100101, People’s Republic of China § Division of Life Science, The Hong Kong University of Science and Technology (HKUST), Clear Water Bay, Kowloon, Hong Kong ∥ State Key Laboratory of Cryospheric Sciences, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences (CAS), Lanzhou 730000, People’s Republic of China ⊥ South China Sea Environmental Monitoring Center, State Oceanic Administration, Guangzhou 510300, People’s Republic of China S Supporting Information *
ABSTRACT: Our understanding of the biogeochemistry of mercury (Hg) in high-altitude aquatic environments remains limited. The Tibetan Plateau (TP) is one of the Earth’s most significant continentalscale high lands, yet much remains unknown about the Hg bioaccumulation and biomagnification in these pristine ecosystems. In this study, 166 wild fish samples of 13 species were collected from 13 rivers and lakes across the southern TP. Total Hg (THg) and methylHg (MeHg) concentrations in the axial muscle of fish ranged from 25.1 to 1218 ng g−1 of wet weight (median ± average deviation of 100.5 ± 149.2 ng g−1) and from 24.9 to 1196 ng g−1 of wet weight (median ± average deviation of 90.7 ± 137.0 ng g−1), respectively. Hg concentrations varied greatly within and between species. The fish Hg concentrations were then linked to the limited available environmental Hg data and special geochemical characteristics in the region, such as Hg loading, pH, low temperature, and high ultraviolet (UV). The long lifespan and slow growth of the fish under the low-productivity environments may be the major biological factors that help to build up the fish Hg levels comparable to those observed in wild fish growing in human-impacted areas. δ13C signals suggested that pelagic fish had higher Hg concentrations, but no relationship was found between the Hg concentrations and the trophic levels. Zooplankton and benthic amphipods had typically higher percentages of MeHg compared to the previously reported values, suggesting the efficient transfer of MeHg from the base of the aquatic food web. This study sheds some light on the geochemical and biological controls of Hg bioaccumulation in fish and biomagnification in the aquatic food web in arid high-altitude environments.
■
INTRODUCTION
m above sea level (a.s.l.), the TP is one of the Earth’s most significant continental-scale high lands. Besides serial magnificent mountain ranges and considerable glacial aggregates, the TP is endowed with sources of major Asian rivers, such as the Yangtze River and Yellow River, and large area of natural lakes. Its unique geographical characteristics make it as unique habitat for hundreds of wildlife endemic to this place. There has been increasing interest in Hg contamination in the TP environments and the role of TP glaciers in global Hg cycling. Recent studies have indicated that Hg levels in the atmosphere, soils, sediments, and glacier snow of TP are close to the natural background.10−13 Much remains unknown about the Hg levels in the wildlife living in the numerous TP rivers and lakes long isolated from human disturbance. More studies on Hg bioaccumulation in the TP
Mercury (Hg) is a widely spread pollutant, owing to its highly mobile nature in both environmental matrixes and food webs. Human activities have accelerated the Hg pollution on the global scale.1−3 Lines of evidence suggested that atmospheric deposition of Hg has been greatly raised by anthropogenic Hg emissions since the Industrial Revolution.4 The rising input of Hg into the atmosphere has resulted in enhanced Hg pollution in both terrestrial and aquatic ecosystems, including those in remote areas.5−7 Previous studies have reported high levels of Hg in fish from remote lakes that have no known history of human disturbance.3,8 Much remains to be explored about the Hg cycling in remote aquatic ecosystems, particularly those at high elevations. Krabbenhoft et al.9 referred to the high-altitude aquatic ecosystems as “typically low pH, low alkalinity and oligotrophic ecosystems” that deserve additional research on Hg cycling. The Tibetan Plateau (TP) is considered as one of those Hgsensitive ecosystems. With its average elevation well above 4000 © 2014 American Chemical Society
Received: Revised: Accepted: Published: 5220
September 25, 2013 April 4, 2014 April 7, 2014 April 7, 2014 dx.doi.org/10.1021/es404275v | Environ. Sci. Technol. 2014, 48, 5220−5228
Environmental Science & Technology
Article
Figure 1. Map showing the study sites.
the southern TP, including 7 river sites and 6 lakes (Figure 1 and see Table S1 of the Supporting Information). The tributaries all belong to the drainage system of the Yarlung Tsangbo River (the upper reach of the Brahmaputra River), and the lakes distribute across the plateau with contrasting elevations (3484−4725 a.s.l.). Fish were caught with the help of local fishermen or by angling. After collection, fish samples were kept in clean polyethylene bags and stored in an ice box, which was transported back to the laboratory immediately. The samples were stored at −20 °C before further laboratory analysis described below. Special sampling effort was made to collect sediments, zooplankton, and amphipods in the Nam Co Lake. The lake has an elevation of 4718 m a.s.l., covering an area of 2015 km2, which is located in the eastern ranges of the Nyainqêntanglha Mountains. Sediments were collected at the littoral areas of the lake. Zooplankton (cladocerans and copepods) and epibenthic macroinvertebrates (amphipods) were sampled using a non-metallic plankton net (mesh size = 100 μm) deployed from a wooden boat. Powder-free clean gloves were worn during sampling and handling. Each sample was collected by a vertical net haul from the bottom to the surface. The invertebrates were only roughly divided into three groups, cladocerans (Daphnia like), copepods, and amphipods. Species were not identified in our study because of the limited available information. Each group of animals was picked out from the samples using clean pipets, transferred to BD Falcon vials, and frozen at −20 °C. Sample Processing and Analysis. The standard lengths and weights of fish samples were measured in the laboratory. The dorsal muscle of fish was dissected, freeze-dried, and ground into a fine powder. The invertebrate samples were also freeze-dried. Three subsamples, each containing 10 amphipods or 50 copepods, were withdrawn from the collection of animals for THg or MeHg analysis. Only two composite samples, each with 10 individuals, were made for cladocerans because of the limited capture. For analysis of THg, approximately 0.1 g of dried fish tissues (or subsample of the invertebrates) was weighted and digested at 190 °C with aqua regia (2 mL of HNO3/6 mL of HCl) in a microwave digestion system. An aliquot of digested samples was taken and diluted appropriately. Bromine monochloride (0.5%, v/v) was added to the diluted sample until a stable
ecosystems are necessary to understand the Hg biogeochemical cycling in the pristine region. Hg concentrations in fish are useful indicators of relative organic Hg (MeHg) levels among different ecosystems, and the information is most relevant to the public and policy community. Recent studies found very low Hg levels in wild freshwater fish in inland areas or marine fish from coastal areas of China. For example, Liu et al.14 found that Hg concentrations were as low as 4−254 ng g−1 [n = 128, wet weight (wwt)] in the fish collected from a Hg-contaminated reservoir. Li et al.15 only detected a range of 0.1−317 ng g−1 of wwt Hg in 327 fish collected from coastal areas of south China. Similarly low Hg levels have been reported in two large-scale studies on marine fish, ranging from 7 to 53 ng g−1 of wwt (n = 75) and from 4 to 254 ng g−1 of wwt (n = 628).16,17 Interestingly, comparable Hg levels (61−596 ng g−1 of wwt) have also been reported in TP fish that grow in pristine high-elevation areas.18 Hg accumulation in Tibetan fish remains poorly understood, and there has long been a lack of information on Hg biomagnification in these ecologically critical environments. Information on Hg concentrations in other trophic levels, such as zooplankton, can provide a more detailed assessment of the Hg concentrations in Tibetan fish, but these data are still lacking to date. The bioaccumulation and trophic transfer of Hg in TP aquatic ecosystems are a matter of interest because the TP aquatic ecosystems are different from many lakes in Europe and North America, where acidification is prevalent,19−21 whereas the TP waters are generally high in pH and alkalinity because of the progressive evaporation or alkaline precipitation. This study is the first to investigate the Hg biomagnification in Tibetan aquatic environments. Total Hg (THg) and methylHg (MeHg) were measured for 166 wild fish collected from different lakes and rivers across the southern TP, and stable isotope signals were applied to seek the relationships between the Hg concentrations and food-web characteristics. In one of the studied lakes, namely, the Nam Co Lake, we also measured Hg in sediments, zooplankton, and amphipods to study the Hg biomagnification in this lake.
■
MATERIALS AND METHODS Study Areas and Sample Collection. During the summer of 2009 and 2010, wild fish were collected from 13 sites across 5221
dx.doi.org/10.1021/es404275v | Environ. Sci. Technol. 2014, 48, 5220−5228
2 4 3 6 40
3 24
18
12 2
Oxygymnocypris stewartii Pareuchiloglanis kamengensis Pseudecheneis sulcatus Ptychobarbus dipogon Schizopygopsis younghusbandi
Schizothorax labiata Schizothorax macropogon
Schizothorax o’connori
Schizothorax waltoni Silurus asotus linnaeus
N
27 1 24
species
Gymnocypris namensis Gymnocypris scleracanthus Gymnocypris waddelli
4 2 1, 3, 6, 7, 9, 11, 12 8, 11 12 11 11, 13 6, 9, 11, 13 1 6, 8, 9, 10, 11, 13 8, 9, 10, 13 8, 9, 11 11
sampling site
25.5−52.0 33.0−42.0
24.3−53.0
30.5−37.1 19.0−48.0
32.8−35.8 11.3−12.9 21.5−26.5 16.2−32.5 12.0−40.0
21.2−46.0 24.0−24.0 11.2−45.0
range
483−612 16−22 29−74 31−150 26−350 381−1105 100−950
100−1388 100−1050 100−300
2.1 0.7 2.5 5.9 6.9
± ± ± ± ±
34.8 ± 3.8 31.6 ± 7.1 38.0 ± 8.5 34.7 ± 8.1 37.5 ± 6.4
34.3 12.0 24.2 22.9 31.2
range 100−750 200−200 21−650
average
± ± ± ± ± 91 3 23 48 105
433 ± 293 200 ± 141
687 ± 422
757 ± 362 344 ± 230
548 19 50 66 206
326 ± 175 200 ± 0 229 ± 185
average
weight (g)
32.8 ± 6.3 24.0 ± 0 25.8 ± 9.7
length (cm)
5.8−23.3
6.0−30.0 10.8
15
10.4
4.7 14.2
2.7−8.1 4.0−21.6
2.8−31.8
5.4
average
5.1−5.7
range
estimated age (year)
131−466 147−349
38.8−138
212−245 73.2−441
149−156 76.5−106 662−1218 51.9−216 33.7−198
33.0−681
25.1−506
range
± ± ± ± ± 5.0 14.3 305 62.4 38.9
315 ± 92.3 248 ± 143
69.6 ± 26.3
232 ± 17.8 206 ± 89.8
153 89.4 868 109 94.3
137 ± 124 184 ± 0 120 ± 135
average
THg (ng g−1 of wwt)
120−402 109−261
36.3−139
218−251 79.4−307
141−164 62.6−96.0 525−1196 37.8−154 32.6−165
27.0- 766
24.9−437
range
278 ± 83.9 185 ± 108
75.0−112 74.0−74.9
72.8−101
101−103 67.7−109
237 ± 16.8 169 ± 57.6 59.5 ± 25.5
90.1−110 74 0.0−90.5 79.3−98.2 68.6−84.2 69.1−107
± ± ± ± ±
71.8−110 71.6−71.6 69.9−112
range
± ± ± ± ±
13.9 7.5 10.7 6.4 9.8
88.4 ± 10.1 74.4 ± 0.6
85.4 ± 8.8
102 ± 1.0 85.9 ± 12.9
100 82.9 85.9 76.0 85.6
87.8 ± 10.6 71.6 ± 0 88.7 ± 11.8
average
MeHg/THg ratio
16.3 15.0 373 44.5 32.7
152 74.3 767 82.3 79.7
118 ± 106 131.6 ± 0 114 ± 150
average
MHg (ng g−1 of wwt)
Table 1. Summary of Fish Species, Sample Size, Length, Weight, Concentrations of THg and MeHg, and Ratio of MeHg/THg [Mean ± Standard Deviation (SD)] in the Samples
Environmental Science & Technology Article
5222
dx.doi.org/10.1021/es404275v | Environ. Sci. Technol. 2014, 48, 5220−5228
Environmental Science & Technology
Article
Figure 2. Relationships between the THg and MeHg concentrations and between the MeHg concentrations and body size in 6 TP fish species. G. namesis, log[MeHg] = 2.5log[length] − 1.8; G. waddelli, log[MeHg] = 1.2log[length] + 0.3; and S. waltoni, log[MeHg] = log[length] + 0.9.
indefinitely, k is the growth coefficient, t is the fish age, and t0 is the age at the length 0. The growth functions of six species of fish collected in our study (Schizothorax o’connori, Schizothorax waltoni, Ptychobarbus dipogon, Schizopygopsis younghusbandi, Oxygymnocypris stewartii, and Schizothorax macropogon) are cited from previous studies, and the age of an individual fish sample was estimated accordingly (see the Supporting Information). Data Treatment and Statistical Analysis. Comparisons of Hg concentrations among the fish (G. waddelli and S. o’connori) collected from different sites were performed using one-way analysis of variation (ANOVA) followed by Tukey’s multiple comparison tests. Linear regression analysis was performed to explore the relationship between fish Hg concentrations and elevations. A positive or negative relationship between fish Hg concentrations and body size or stable isotope values was detected by fitting the data to a designated function model. Comparisons and relationships were considered statistically significant at p < 0.05. All statistical analyses were performed using SPSS 16.0 or Sigmaplot 10.0.
yellow color was obtained, after which the samples were prereduced by the addition of NH2OH·HCl. THg was quantified using the single gold trap amalgamation technique by cold vapor atomic fluorescence (CVAFS, QuickTrace 8000, CETAC Technologies, Omaha, NE). For analysis of MeHg, approximately 40 mg of tissue was digested with 25% KOH in methanol at 60 °C for 3 h. MeHg in the extract was measured with an automated MeHg analytical system (MERX, Brooks Rand). Briefly, 20−50 μL of extract was buffered with sodium acetate at pH 4.9 and ethylated by sodium tetraethylborate in a 40 mL borate glass bottle with a Teflon-lined cap. The quantification of MeHg was automatically performed by the MeHg analyzer with gas chromatographic separation and pyrolysis, following atomic fluorescence detection. Quality controls included measuring analytical blanks and certified reference material (CRM) between sample runs. The analytical blanks were 0.05). While the water chemistry appears not to favor the MeHg production, Hg concentrations in the Tibetan fish may be more likely attributed to the biological traits of the fish adapted to the arid high-altitude conditions. Hg concentrations in fish are controlled by the balance of uptake and excretion of Hg,47 and the latter is found to be temperature-dependent.48 Although the uptake of Hg in the fish is considered low under the clean TP aquatic environments, the low-temperature climate in highaltitude areas may greatly limit the MeHg elimination in the fish. A long lifespan is one of the typical features that make TP fish different from the freshwater fish living in human-impact areas. The estimated age of the fish S. o’connori, S. waltoni, P. dipogon, S. younghusbandi, O. stewartii, and S. macropogon is estimated to be 6.0−30.0, 5.8−23.3, 2.7−8.1, 4.0−21.6, 5.1− 5224
dx.doi.org/10.1021/es404275v | Environ. Sci. Technol. 2014, 48, 5220−5228
Environmental Science & Technology
Article
Figure 3. Relationship between THg concentrations and stable isotope signatures (δ13C and δ15N) in the fish from the Yarlung Tsangbo River (Bayi, site 11).
Table 2. Summary of MeHg, THg, and MeHg/THg in the Aquatic Organisms from the Nam Co Lake (Assuming 80% Water Content), with n = 3 (Mean ± SD) for Copepods and Amphipods and n = 2 (Mean ± SD) for Cladoceransa organism copepods cladocerans amphipods fish G. namensis a
MeHg (ng g−1 of wwt) 6.7 10.7 9.9 117.6
± ± ± ±
THg (ng g−1 of wwt)
0.2 1.6 1.7 105.9
12.9 16.1 16.1 121.1
± ± ± ±
2.4 1.8 1.0 106.7
MeHg/THg (%) 53.1 66.4 61.7 87.8
± ± ± ±
10.4 2.7 9.7 10.6
BAFMeHg (L kg−1) 2.3 3.6 3.4 4.1
× × × ×
105 105 105 106
The BAFs of MeHg are calculated by assuming that 5% THg in the lake (0.59 ng L−1) is MeHg.33
consumers feeding on periphyton than those relying on terrestrial carbon in acidic streams. Riva-Murray et al.56 found that the great accumulation of Hg in secondary consumers was associated with a more depleted δ13C dietary carbon at multiple spatial scales of streams in New York’s Adirondack Mountains and South Carolina’s coastal plain. These results indicated that the higher Hg levels in more δ13C-depleted TP fish, such as P. sulcatus, may be due to their reliance on the Hg-enriched dietary carbon source. No significant relationship was found between the THg or MeHg concentrations and the δ15N signatures (Figure 3). This is not surprising because most Tibetan fish mainly feed on benthic algae, zooplankton, or small benthic invertebrates, such as amphipods.57 The δ15N detected in our fish samples ranged narrowly from 6.4 to 10.8‰. Considering an enrichment of 3.4 per trophic level for δ15N,58 the sampled 13 fish species only cover 1.3 trophic levels, which is consistent with the fact that the food chain is typically short in TP aquatic ecosystems. Enrichment of Hg by a long food chain is thus not expected to be a significant factor accounting for the high Hg levels in the Tibetan fish. The Hg concentrations in sediments, zooplankton, and fish in the Nam Co Lake may also shed some light on the Hg biomagnification in the Tibetan aquatic food web. The Nam Co Lake, like most of the other Tibetan lakes and rivers, is a typical oligotrophic lake with low productivity; e.g., chlorophyll a concentrations were as low as 0.07−1.73 μg L−1.45 A number of studies suggested that fish collected from remote or oligotrophic lakes often contained higher Hg concentrations than those collected from eutrophicated waters, and biomass dilution of the primary producers and fast growth rate of zooplankton are consider as the major causes.14,59,60 In the Nam Co Lake, the average THg concentration detected in littoral sediments was only 5.8 ± 0.8 ng g−1 of dwt, suggesting extremely low Hg contamination in the lake areas. This is in agreement with a previous observation that the dissolved THg concentration in the lake water was only 0.59 ng L−1.33 The MeHg concentrations measured in copepods, amphipods, and
5.7, and 2.8−31.8 years old, respectively (Table 1), which are rarely observed in earlier studies. Tibetan fish generally grow very slowly, probably because of the freezing temperature or poor food quality and availability in the oligotrophic highaltitude environments. For example, naked carp Gymnocypris przewalskii only grow 0.5 kg in 10 years;49 thus, a little growth dilution effect is expected for the TP fish. The long lifespan and slow growth rate may be the major factors that help to build up the Hg concentrations in the TP fish comparable to those observed in human-impact areas. Significant positive relationships were observed between the THg (or MeHg) and body length for G. namensis, G. waddelli, and S. waltoni (Figure 2), suggesting that body size is an important factor controlling the within-species Hg variations. However, the Hg−body length relationship varies between fish species, which reflects the species-dependent growth pattern. The reason remains unclear why the small-size fish P. sulcatus contained such high Hg concentrations up to 1218 ng g−1 of wwt. The species-specific Hg concentrations may be dependent upon other ecological traits, such as feeding habit or trophic level. Basal-carbon sources and trophic levels may also influence the Hg bioaccumulation in the TP fish. A significant negative relationship was found between the THg or MeHg concentrations and the δ13C signatures in fish collected from the river in Bayi (Figure 3 and see Figure S1 of the Supporting Information), indicating that dietary carbon sources provided additional controls of the variable Hg concentrations in the TP fish. Numerous studies have found that food segregation (benthic versus pelagic) has significant effects on fish Hg concentrations in lakes.50,51 The link was however less obvious because of the highly dynamic nature of δ13C for the autochthonous organic matter in streams or rivers.52,53 However, recent studies provided evidence that carbon source potentially influenced the Hg bioaccumulation in consumers in the lotic ecosystems. For example, Tsui et al.54 found a much higher concentration of MeHg in the biofilm than in the terrestrial detritus. Jardine et al.55 observed higher levels of Hg in 5225
dx.doi.org/10.1021/es404275v | Environ. Sci. Technol. 2014, 48, 5220−5228
Environmental Science & Technology
■
cladocerans were 6.7 ± 0.2, 9.9 ± 1.7, and 10.7 ± 1.6 ng g−1 of wwt, respectively, and the THg concentrations were 12.9 ± 2.4, 16.1 ± 1.0, and 16.1 ± 1.8 ng g−1 of wwt, respectively, assuming 80% water content for invertebrates (Table 2). The average MeHg and THg concentrations detected in Nam Co naked carp G. namensis were 140 and 117 ng g−1 of wwt, respectively, showing a 10-fold increase of MeHg from primary consumers to the fish. The calculated bioaccumulation factor (BAF) for MeHg ranges from 2.3 × 105 to 4.1 × 106 for the invertebrates and fish (Table 2). Enhanced MeHg bioaccumulation was observed in copepods, amphipods, and cladocerans, in which the MeHg/THg was 53.1 ± 10.4, 61.8 ± 9.8, and 66.3 ± 2.7%, respectively (Table 2). The range of MeHg/THg ratios for zooplankton in low-altitude lakes (∼500 m a.s.l.) varied greatly (11−97%).61 Interestingly, the ratios in the alkaline TP lakes were apparently higher than those in non-acidified lakes (90%).62,63 Although the data do not permit a rigorous statistical comparison to earlier studies, the enrichment of MeHg in the TP invertebrates is evident. The underlying mechanisms for the high percentage of MeHg as THg in these invertebrates remain unclear. However, the data implied the efficient transfer of MeHg from the base of the aquatic food web in the Nam Co Lake, which partly explained the high Hg levels in the fish. The MeHg/THg ratios in high-altitude lakes were generally low (
