RESEARCH ARTICLE

Bioaccumulation and trophic transfer of mercury in a foodweb from a large, shallow, hypereutrophic lake (Lake Taihu)in China

Shaofeng Wang & Biao Li & Mingmei Zhang &Denghua Xing & Yonfeng Jia & Chaoyang Wei

Received: 4 November 2011 /Accepted: 23 January 2012 /Published online: 15 February 2012# Springer-Verlag 2012

AbstractPurpose Due to the fast development of industry and theoveruse of agrichemicals in past decades, Lake Taihu, animportant source of aquatic products for Eastern China, hassimultaneously suffered mercury (Hg) contamination andeutrophication. The objectives of this study are to under-stand Hg transfer in the food web in this eutrophic, shallowlake and to evaluate the exposure risk of Hg through fishconsumption.Methods Biota samples including macrophytes, sestons,benthic animals, and fish were collected from Lake Taihuin the fall of 2009. The total mercury (THg), methyl mer-cury (MeHg), 13C and 15N in the samples were measured.Results and discussion The signature for 15N increasedwith the trophic levels. Along with a diet composed of fish,the significant relationship between the 13C and 15N in-dicated that a pelagic foraging habitat is the dominant path-way for energy transfer in Lake Taihu. The concentrations ofTHg and MeHg in the organisms varied dramatically by 3orders of magnitude from primary producers (macrophytes

and sestons) to piscivorous fish. The highest concentrationsof both THg (100 ng g1) and MeHg (66 ng g1), however,were lower than the guideline of 200 ng g1 of MeHg forvulnerable populations that is recommended by the WorldHealth Organization (WHO). The daily intake of THg andMeHg of 92 and 56 ng day1 kg1 body weight, respective-ly, was generally lower than the tolerable intake of230 ng day1 kg1 body weight for children recommendedby the Joint FAO/WHO Expert Committee on Food Addi-tives. Significant relationships between the 15N and thelogarithm of THg and MeHg showed an obvious biomagni-fication of Hg along the food web. The logarithmic bioac-cumulation factor of MeHg in the fish (up to 5.7) from LakeTaihu, however, was relatively low compared to that ofother aquatic ecosystems.Conclusion Health risk of exposure to Hg by consumptionof fish for local residents is relatively low in the Lake Taihuarea. Dilution of Hg levels in the phytoplankton induced byeutrophication is a possible factor inhibiting accumulationof MeHg in fish in eutrophic Lake Taihu.

Keywords Mercury . 13C and 15N . Trophic level .

Bioaccumulation . Lake Taihu

1 Introduction

As a non-essential element for human beings, mercury (Hg)is a ubiquitous environmental contaminant and a neurotoxinthat accumulates in aquatic habitats and poses a risk tohumans and other aquatic organisms (Clarkson 1993; Ullrichet al. 2001). Once released from natural and/or anthropogenicsources, inorganic Hg can enter aquatic environments andaccumulate in sediments, where it can be transformed tomethyl mercury (MeHg), a more toxic and bioaccumulative

Responsible editor: Vera Slaveykova

S. Wang (*) : B. Li :M. Zhang :D. Xing :Y. JiaKey Laboratory of Pollution Ecology and EnvironmentalEngineering, Institute of Applied Ecology,Chinese Academy of Sciences,Shenyang 110016, Chinae-mail: wangshaofeng@iae.ac.cn

B. Li :M. ZhangGraduate University of Chinese Academy of Sciences,Beijing 100039, China

C. WeiInstitute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing 100101, China

Environ Sci Pollut Res (2012) 19:28202831DOI 10.1007/s11356-012-0787-2

form of Hg, by sulfate-reducing bacteria under anoxic con-ditions (Compeau and Bartha 1985; Gilmour et al. 1992). TheMeHg produced in sediments can be absorbed and accumu-lated in the food web by energy transfer. Large predatorymarine fish tend to contain elevated MeHg concentrations ofup to millions times of that in the water they live in. Fish withMeHg concentrations exceeding the World Health Organiza-tion (WHO) guideline of 0.5 mg kg1 have been widelyfound in the lakes of Northern Europe (Munthe et al.2004) and North America (Hobson et al. 1998; Watras et al.1998; Bowles et al. 2001).

Because most of the Hg that accumulates in speciesoriginates from consumption of organisms at lower trophiclevels, rather than direct aqueous accumulation (Hall et al.1997; Kainz and Mazumder 2005), uptake of MeHg fromthe environment by the lowest organisms of the food chainplays a key role in MeHg bioaccumulation and biomagnifi-cation in biota at higher trophic levels. The pathway forMeHg transfer along the food web can be classified aspelagic or benthic according to the foraging habitat. Onthe one hand, a part of the MeHg in sediments can betaken up by benthic animals directly through gut digestion(Gagnon and Fisher 1997; Eagles-Smith et al. 2008). On theother hand, MeHg is also able to enter the above water abovethrough particulate resuspension and diffusion, where it canbe absorbed by phytoplankton (Watras et al. 1998; Eagles-Smith et al. 2008, 2009; Chen et al. 2009) and then biomag-nified to potentially harmful concentrations in the food web(Knauer and Martin 1972). Previous research found the highbioconcentration factor (104106) of theMeHg in phytoplank-ton related to the water (Kainz and Mazumder 2005; Al-Reasiet al. 2007; Coelho et al. 2009), indicating a great biomagni-fication ability of the MeHg by primary producers.

Eutrophication of freshwater lakes is also a primary en-vironmental concern around the world. In China, approxi-mately 66% of the lakes are at a eutrophic level, whileapproximately 22% of the lakes are at a hypereutrophic level(Liu 1997). The most important characteristic of lake eutro-phication is the rapid increase in primary productivity over ashort time, commonly called algae bloom, that induces asignificant change in the environmental parameters andpopulation structure of aquatic organisms in the lake. Someprevious studies have discovered that Hg methylation can beenhanced significantly by elevated loading of organic mate-rial loading in the sediments induced by eutrophication inlakes and/or reservoirs, leading to high concentrations ofMeHg in the sediment and the water column (Macalady etal. 2000; Gray and Hines 2009). The highest concentrationof Hg in the food web, however, was generally lower than0.5 mg kg1 (wet weight) in eutrophic lakes such as ClearLake (Eagles-Smith et al. 2008), Lake Candia (Fontanella etal. 2009), and even Baihua Reservoir, which is heavilycontaminated with Hg (Liu et al. 2012). This result indicates

that bioaccumulation and biomagnification of Hg in thefood chain may be influenced by the distinct changes inprimary productivity in eutrophic lakes.

Lake Taihu, a large, shallow, freshwater lake in China, is animportant resource for drinking water, fishing, and tourism forthe population living in Eastern China. The Taihu Basin area,with a total area of less than 0.4% of China and a population of36 million, is an important base of agriculture and industry,where approximately 10% of Chinese GDP is accounted forfollowing approximately 30 years of fast development. Due todecades of intensive input of nutrients and contaminants fromindustrial and agricultural sources, however, Lake Taihu hassuffered Hg contamination and heavy eutrophication (Sun andMao 2008). The highest concentration of total mercury (THg)in the sediment and the water column reached 0.99 mg kg1

(Fan and Zhang 2009) and approximately 80 ng L1 (Wang etal. 2011), respectively; however, a low ratio of MeHg/THg(

2.2 Sample collections

Macrophytes, sestons, benthic animals, and fish were col-lected in different regions/sites (Fig. 1) during September 110, 2009. Six species of macrophyte samples were collectedrandomly by hand at five sampling sites in the macrophyticlake area in the eastern part of Taihu (Fig. 1). Sestonsamples were collected at nine sampling sites in thealgal lake area using two conical plankton nets (PTN-SW2 and PTN-SW3). According to the mesh size (112and 64 m), the seston samples at each sampling sitewere divided into two subsamples: a small one (64112 m) and a larger one (>112 m). The benthic animals(Lamellibranchia sp. andMargarya sp.) were collected onlyin the Bogongdao area. The muscle tissues of 11 spe-cies of fish captured from five regions with nylon netswere collected using a stainless steel scalpel (Fig. 1; Table 2).According to the composition of their diet (Liu 2008), the fishfrom Lake Taihu could be divided into plankivores, omni-vores, and piscivores (Table 2). To distinguish the impact ofthe aquaculture on the Hg accumulation in fish, three culturedCarassius auratus Linnaeus were purchased in the market inGuandu. The fish size, including the length and the weight,were measured in the field. All of the biota samples werestored in clean plastic bags or bottles and preserved by

Fig. 1 Sampling locations for the macrophytes, sestons, benthic animals, and fish in Lake Taihu. The isoline of the macrophyte coverage is redrawnaccording to You et al. (2008)

Table 1 Physicochemical data and THg and MeHg levels in the sedi-ments and water from Lake Taihu

Altitude (m asl) 4

Lake area (km2) 2,425

Mean depth (m) 1.9

Volume (m3) 4.4109

Algal bloom dominated area (km2) 1,600a

Macrophyte dominated area (km2) 460a

Suspended solids (mg L1) 90 (222005)a

Chlorophyll a (g L1) 15 (1.867)a

Total P (g L1) 110 (42250)a

Total N (mg L1) 3.1 (1.08.1)a

pH 6.99.0b

DO (mg L1) 5.90.28 (5.16.5)b

THg in the filtered water (ng L1) 159.8 (4.246)b

MeHg in the filtered water (ng L1) 0.120.04 (0.0410)b

THg in the sediment (g kg1) 10070 (12470)b

MeHg in the sediment (g kg1) 0.280.22 (0.0180.96)b

a Data were cited from Zhu et al. (2008)b Data were collected during May and September 2009 (Wang et al.2011)

2822 Environ Sci Pollut Res (2012) 19:28202831

freezing in a sample box with dry ice. In the laboratory,all of the samples were washed using ultrapure water(18.2 M ), freeze dried, and ground to a fine powder usingan agate mortar and pestle for further chemical analysis. All ofthe dried samples were preserved in double plastic bags toavoid cross-contamination.

2.3 Stable isotope (13C and15N) analysis

Duplicate samples of the freeze-dried biota (approximately0.1 mg for 13C and 1.03.0 mg for 15N analysis, respec-tively) were weighed and sealed into tin capsules, followed byanalysis for 13C and 15N using a continuous-flow stable

isotope ratio mass spectrometer (Delta plus XP, USA). Sampleswere transformed to CO2 and N2 by combustion at 1,000C,and the gas was separated by a Carbosieve G column beforeintroduction into the spectrometer. The isotope ratios areexpressed in the per mil notation. For quality control, thereference material (glycine) was analyzed approximately ev-ery 40 samples, and values of 100.6 (n05) of the 15Nand 330.9 (n07) for the 13C were obtained, which werecomparable to the certified values of 10.0 for 15N and33.3 for 13C, respectively. Blanks were also analyzedevery 40 samples, and no signals were detected for themthroughout the analysis. The analytical precision of the instru-ment for 13C and 15N is 90%) 374.6 (n011) 980280(n011)

Bogongdao (3),Chachang (4),Dongshan (12),Guandu (4)

Omnivores

C. carpio Linnaeus Common carp Macrophytes (70%),benthic animals (30%)

314.2 (n013) 660290(n013)

Bogongdao (4),Chachang (4),Dongshan (5),Guandu (4),Xiaoxigang (5)

C. auratus Linnaeus Crucian carp Microcystis (93%),benthic animals (7%)

233.0 330120 Bogongdao (9),Xiaoxigang (7)

C. auratus Linnaeus Crucian carp (Cultured) Artificial diet 230.6 35040 Guandu (3)

Mylopharyngodon piceus Black carp Benthic animals 29 (n01) 450 (n01) Dongshan (3)

Neosalanx tangkahkeiitaihuensis Chen

Whitebai Zooplankton(Copepod, 70%;Cladocera, 30%)

ndb ndb Dongshan (mixed)

Piscivores

Ophiocephalus argus cantor Argus snakehead fish Small fish 326.9 470330 Dongshan (3)

Erythroculter MongolianBasilewsky

Mongolian culter Coilia ectenes taihuensisYen etLin (frequency 59%),Toxahbramis swinhonisGunther (26%),Hyporhamphusintermedius (19%),shrimp (19%)

323.1 430130 Xiaoxigang (5)

E. ilishaeformis Bleeker Topmouth culter Coilia ectenes taihuensisYen et Lin(frequency 95%),Hyporhamphusintermedius (31%),shrimp (21%)

287.3 (n018) 300180(n018)

Bogongdao (5),Dongshan (4),Xiaoxigang (11)

Silurus asotus Catfish Small fish 40 450 Dongshan (1)

a According to Liu (2008)b Not determined because the size of the fish species was too small

Environ Sci Pollut Res (2012) 19:28202831 2823

2.4 THg and MeHg analysis

The concentration of THg in all of the samples was directlymeasured by coupling combustion and cold vapor atomicabsorption spectrometry (Milestone DMA-80). The detec-tion limit of the method is approximately 0.5 ng g1, basedon a sample weight of 20100 mg. For MeHg, approximate-ly 0.200 g of dried sample was weighed in a Teflon centri-fugation tube and digested using 10 mL of a 25% KOH/CH3OH (m/v) solution at 80C for 3 h. After digestion, theliquid was neutralized using approximately 3 mL of 12 MHCl. The mixture was then subjected to CH2Cl2 extraction,centrifugation, and reverse extraction. The MeHg concen-trations were determined by isothermal gas chromatographyand cold vapor atomic fluorescence spectrometry (ModelIII, Brooks Rand, USA) following ethylation, purging bynitrogen gas and concentration using Tenax according toLiang et al. (1996). The detection limit of the method forMeHg is approximately 0.02 ng g1. The inorganic Hg(IHg) concentrations in the biota were calculated by sub-tracting the MeHg from the THg.

Method quality control was achieved by analyzing sim-ilar amounts of the reference material (IAEA-407), themethod blanks and the duplicates. The concentrations ofthe THg and the MeHg in the reference material weredetermined to be 21020 ng g1 (n030) and 19510 ng g1 (n024), respectively, which were comparable tothe certified values of 222 ng g1 (216228 ng g1) for THgand 200 ng g1 (188212 ng g1) for MeHg. The averagerelative standard deviation of the duplicates (n020) was5.6%. All of the data for the Hg concentrations in the biotawere calculated based on the specific wet weight (ww) ofthe muscle tissue of each individual fish.

2.5 Data analysis

All of the data were processed using Excel 2007 and plottedusing Origin 8.0. The statistical analysis and the Pearsoncorrelation among the different data clusters were calculatedusing the software of Origin 8.0. The significant levels ofdifference among the different data arrays were examinedusing one-way ANOVA.

3 Results and discussion

3.1 Isotopic composition in the food web

The 13C signature was in the range of 30 to 14,with a mean value of 233.2, while the highest 15N of20 was found in Erythroculter ilishaeformis Bleeker andthe lowest 15N was 4.2 in Nymphoides peltatum (Table 3).Usually, 15N is regarded as a very useful tool for constructing

the food web of a system because the nitrogen pools ofanimals are enriched in 15N relative to their food, and thetop predators have the highest value of this stable isotope (Fry1988). The relationship between the 15N and the trophic levelis shown in Fig. 2. It can be clearly seen that the pool of 15N inthe different species increased distinctly according to thefollowing order: macrophytes

Table3

Species,THg,

MeH

g,1

3C,1

5N,andnumberof

aquatic

organism

scollected

from

LakeTaihu

Species

Com

mon

name

THg(ng/g,ww)

MeH

g(ng/g,ww)

13C(

)1

5N(

)MeH

g/THg(%

)n

Macrophytes

Myrioph

yllum

verticillatum

Watermifoil

3.1

0.11

14

6.4

3.6

1

Ceratop

hyllu

mdemersum

L.

Hornw

ort

5.40.21

(5.35.6)

0.47

0.04

(0.440.50)

20

1.2

9.10.73

8.70.4(8.49.0)

2

Trap

abispinosaRoxb.

Water

chestnut

1.5

0.066

26

7.3

4.4

1

Hydrilla

verticillata

Mezreb

3.1

0.25

17

8.3

8.0

1

N.pelta

tum

Water

poppy

2.00.2(1.82.3)

0.13

0.02

(0.100.14)

25

0.49

6.72.3

6.21.0(5.27.3)

3

Potam

ogeton

octand

rusPoir.

Pondw

eed

3.31.1(2.54.1)

0.18

0.042(0.150.21)

15

0.73

8.12.3

5.50.6(5.16.0)

2

Sestons

Small(64112m

)

6.62.2(4.19.4)

0.17

0.11

(0.0270.29)

26

2.1

121.9(n04)

3.02.7(0.557.1)

6

Large

(>112m

)

7.63.1(3.212)

0.28

0.25

(0.0730.77)

25

2.5

122.4(n04)

4.65.0(1.114)

9

Benthicanimals

Lam

ellib

ranchiasp.

Anodonta

7.12.1(3.09.6)

3.51.38

(1.45.9)

29

0.33

130.85

(n06)

5017

(3488)

11

Marga

ryasp.

Spiralshell

5.2

2.6

22

1851

Mixed

Plankivores

Cteno

pharyngo

donidellus

Grass

carp

9.30.29

(9.19.5)

4.61.2(3.75.5)

16

0.26

10(n01)

4912

(4157)

2

Hypop

htha

lmichthysmolitrix

Silver

carp

104.9

24

461

A.no

bilis

Terch

115.5(3.023)

6.13.8(2.216)

25

1.3

153.2(n013)

5818

(2193)

26

Omnivores

C.carpio

Linnaeus

Com

mon

carp

8.65.4(2.922)

5.23.3(1.516)

22

1.9

173.6(n011)

6319

(3491)

22

C.au

ratusLinnaeus

Crucian

carp

2016

(6.667)

149.9(3.336)

22

1.5

154.8(n06)

7016

(4297)

16

C.au

ratusLinnaeus

Crucian

carp

(Cultured)

7.10.53

(6.57.5)

4.70.37

(4.35.0)

24

0.26

661.2(6567)

3

Mylop

haryng

odon

piceus

Black

carp

3127

(1562)

1813

(9.833)

19

2.2

121.7(n02)

639.2(5270)

3

Neosalanx

tang

kahkeii

taihuensisChen

Whitebai

1511

25

1573

Mixed

Piscivores

Oph

ioceph

alus

argu

scantor

Argus

snakeheadfish

4827

(2778)

2923

(1155)

19

0.53

140.83

(n02)

5514

(4170)

3

ErythroculterMongolian

Basilewsky

Mongolianculter

2812

(1344)

218.5(9.031)

23

1.1

182.2(n03)

779.5(6991)

5

E.ilishaeform

isBleeker

Topmouth

culter

3620

(16100)

2114

(7.066)

23

1.48

183.3(n09)

5718

(2095)

20

Silurusasotus

Catfish

7145

19

63.0

1

Allvalues

arepresentas

thearith

meticmeanstandard

deviation(range).The

units

oftheconcentrations

ofTHgandMeH

gin

allof

thebiotaarebasedon

thewetweight(w

w)

Environ Sci Pollut Res (2012) 19:28202831 2825

between the isotopes (15N and 13C) and the concentra-tions of THg and MeHg in the macrophytes.

3.2.2 Hg in the sestons

As it contains a mixture of phytoplankton and zoo-plankton, the larger size subsample (>112 m) shouldinclude more zooplankton than the smaller sample (64112 m). The signatures of both the 15N and the 13C(Figs. 2 and 3) indicated that the trophic level of thecoarse seston subsample was slightly higher than that ofthe fine seston sample (Table 3). The concentrations ofHg in the seston from Lake Taihu were lower than thatin some lakes/reservoirs. For instance, the average MeHgconcentration in the seston from Lake Murray was reported tobe 15 ng g1 (ww; Bowles et al. 2001). Liu et al. (2011)reported that THg and MeHg concentrations were 170

100 ng g1 (ww) and 0.80.5 ng g1 (ww), respectively, inthe phytoplankton (water content is assumed to be 80%)from a contaminated reservoir (Baihua Reservoir) in China.However, the average concentration of THg in the seston fromLake Taihu was comparable to that in the phytoplankton(15 ng g1) from another eutrophic lake, Clear Lake(Eagles-Smith et al. 2008). The percent of MeHg accountingfor THg was in the range 0.5514%, with an average of 4.04.2%, indicating that the greatest fraction of Hg in the sestonswas presented as IHg. This may be due to the joint effects ofhigh THg and low MeHg levels in the water of Lake Taihu.Concentrations of THg and MeHg in the sestons collectedfrom the northern part of Taihu (sites T1, T2 and T6), thecentral part of Taihu (T35) and the western part of Taihu(T79) did not show significant differences (p>0.05, one-wayANOVA), although the northern part of Taihu was consideredto be strongly influenced by anthropogenic discharging.

Our results also showed that the average concentra-tions of both THg and MeHg in the larger-size subsam-ple were slightly higher (p>0.05, one-way ANOVA)than those in the smaller-size seston (Table 3); however,no significant relationship between 15N and Hg con-centrations in either the larger or the smaller particlesize subsamples was observed, possibly due to the differentlevels of Hg contamination at each sampling site. Partitioningof MeHg between the phytoplankton and the water plays animportant role in biomagnification of Hg in the food webbecause phytoplankton is the base of the aquatic food weband most of the Hg in the biota in the higher trophic levels isaccumulated mainly by food intake, rather than direct uptakefrom the aqueous phase (Hall et al. 1997; Kainz andMazumder 2005; Pickhardt and Fisher 2007). Therefore, thelow concentrations of THg and MeHg in the sestons may be alimiting factor for Hg bioaccumulation in the food web ofLake Taihu.

Macrophytes

0 5 10 15 20 25

Plankton (Small)

Plankton (Large)

Benthic animals

Plankivores

Omnivores

Piscivores

15N ()

Fig. 2 Relationship between 15N and the different species of biotafrom Lake Taihu. Note: the biota is classified according to Table 3

-32 -28 -24 -20 -16 -124

8

12

16

20

24 Seston (small) Seston (large) Terch Common carp Crucian carp Whitebai Mongolian culter Topmouth culter Benthic animals Black carp Macrophytes Grass carp Argus snakehead fish

15 N

()

13C ()

R2 = 0.55, p = 0.021Pelagic foraging habitat

Littoral foraging habitat

Fig. 3 Plot of 15N and 13C for the specific foraging habitat biota inthe Lake Taihu

0 20 40 60 80 100 120

0

20

40

60

80

MeH

g (ng

/g, w

w)

THg (ng/g, ww)

R2 = 0.84, p < 0.001

[MeHg] = 0.61 [THg] - 0.60

Fig. 4 Relationship between THg and MeHg in all biota samples

2826 Environ Sci Pollut Res (2012) 19:28202831

3.2.3 Hg in the benthic animals

The THg levels in Lamellibranchia and Margarya fromLake Taihu (Table 3) were comparable to those (5.6 ng g1,wet weight, the water content is assumed to be 80%) inmollusks from a eutrophic lake (Lake Candia) in Canada(Fontanella et al. 2009). The concentrations of Hg in thebenthic invertebrates from Lake Taihu were significantlylower than those from other aquatic systems, however.For example, Campbell et al. (2008) reported that THglevels in some invertebrates from Lake Tanganyikaranged from 20 to 50 ng g1. Coelho et al. (2008) found thattwo species of invertebrates (Scrobicularia plana and Hedistediversicolor) could accumulate THg andMeHg levels of up to790 and 150 ng g1 (wet weight), respectively. The concen-trations of THg and MeHg in bivalves and gastropods collect-ed from the coastal sites along the Bohai Sea of China rangedfrom 6.7 to 450 and 48170 ng g1 (wet weight), respectively(Liang et al. 2003).

The proportion of MeHg to THg in the benthic inverte-brates ranged from 34% to 88%, which was comparable tothe range of 2189% in mollusks from the coast of theBohai Sea (Liang et al. 2003), but was higher than thepercentage of MeHg (543%) in S. plana and H. diversi-color reported by Coelho et al. (2008). Because Hg can betaken up through digestion via the intestinal wall (Wang etal. 1998; Coelho et al. 2008), the benthic invertebrates canbe used as an indicator of Hg and MeHg contamination inthe benthic aquatic environment. Therefore, the lower con-centrations of MeHg imply that MeHg contamination in thesediments from Lake Taihu is very limited.

3.2.4 Hg in the fish

The concentrations of THg and MeHg showed a distinctincrease from fish species at low trophic levels to hightrophic levels (Table 3). Compared with other eutrophiclakes worldwide, the concentrations of THg and MeHg inthe fish from Lake Taihu were comparable or higher. Forinstance, Eagles-Smith et al. (2008) reported that the meanconcentration of THg in the fish from the eutrophic ClearLake ranged from 30 to 140 ng g1 (ww), except for thelarge largemouth bass, with a mean concentration of310 ng g1 (ww). Fontanella et al. (2009) reported that theTHg concentration in the catfish in the eutrophic LakeCandia was 30 ng g1 (ww). Compared with Baihua Lake,however, which exhibits eutrophication and Hg contamina-tion, the average THg concentration in the fish from LakeTaihu was lower, but the MeHg concentration was higher(Liu et al. 2012).

The proportion of Hg present as MeHg in the fish speciesranged from 20% to 97%, with an arithmetical mean valueof 6217%, which is lower than that (5494%) in Lake

Murray (with a low THg concentration in the lake water(Bowles et al. 2001)) but higher than that in the morecontaminated Baihu Reservoir (4523%; Liu et al. 2012)and Ya-Er Lake (45%; Xu et al. 2006). Our investigationduring the same period revealed that THg concentrationshave been much elevated in the sediment and the lake waterabove it compared to the background value (typically 15 ng L1 in the water and 5(Table 4). The THg and MeHg concentrations correlatedsignificantly with the mass and length for C. carpio and E.ilishaeformis Bleeker. Despite the high correlation coeffi-cients for E. Mongolian Basilewsky, the relationships be-tween Hg (THg and MeHg) concentration and fish size(mass and length) were not significant (Table 4), possiblydue to an insufficient number of samples. But for A. nobilisand C. auratus, no significant relationship was found be-tween THg and MeHg in the muscle tissue and the fish size.In addition, we did not observe a significant relationshipbetween Hg in fish and the stable isotopes (13C and 15N)in the fish species, except that 13C showed a significantlynegative relationship with MeHg in C. auratus (Table 4). Thismay be due to the fact that the difference in Hg between theindividuals was not enough to indicate a relationship with theisotopes 13C and 15N using our data.

3.3 Bioaccumulation and biomagnification of the Hgin the food web

The level of Hg bioaccumulation in all of the biota specieswas expressed by the bioaccumulation factor (BAF) accord-ing to the following equation:

BAF L kg1

Hg concentration in biota mgkg1

Dissolved Hg concentration in water mgL1

We calculated the BAFs of MeHg (Fig. 5a) and IHg(Fig. 5b) of the biota species using the average concentra-tions of MeHg and THg in the filtered lake water from LakeTaihu (Table 1). The BAFs of both MeHg and IHg increased

Environ Sci Pollut Res (2012) 19:28202831 2827

with the trophic levels. But the logarithmic BAFs ofMeHg (Fig. 5a) for all of the trophic levels varied dramat-ically from 2.0 to 5.7 and was approximately 2 orders ofmagnitude higher than the logarithmic BAFs of IHg(Fig. 5b), which were in the range of 1.23.5, indicating ahigh bioavailability of MeHg. The logarithmic BAFs ofMeHg obtained from Lake Taihu were significantly lowerthan the logarithmic BAF of 6.8 found for Lake Murray(Bowles et al. 2001) and the logarithmic BAF of 6 determinedfor Long Island Sound (Hammerschmidt and Fitzgerald2006).

The results provide solid evidence for the biomagnifica-tion of Hg: the highest concentrations of THg were found inthe piscivorous fish (13100 ng g1), the intermediate con-centrations in the plankivores and the benthic animals (3.067 ng g1) and the lowest concentrations in the sestons andthe macrophytes (1.512 ng g1; Table 3). Biomagnificationof THg and MeHg in the food web is obvious based on thesignificant relationships between the logarithm of THg(Fig. 6a) and MeHg (Fig. 6b) concentrations and the 15Nvalues for all of the species in Lake Taihu. The regressionequation can be expressed as a linear relationship:

Log THg 0:052 d15N 0:20

Log MeHg 0:12 d15N 1:3

The slopes of the regressions (0.052 for THg and 0.12 forMeHg) were almost comparable to those (0.07 for THg and0.14 for MeHg) derived by Al-Reasi et al. (2007) in the Gulfof Oman, but were lower than the slope range (0.150.16)for THg determined by Kidd et al. (1995) in six temperatelakes in Ontario and the 0.28 for MeHg in Lake Murray(Bowles et al. 2001). Because the slope of the regression ofthe log-transformed THg and MeHg concentrations and the15N signature in the biota can represent the biomagnifica-tion power of different Hg species, the smaller regressionslopes of THg and MeHg indicate a lower extent of bio-magnification of MeHg and THg in the Lake Taihu foodweb compared to other lakes or marine ecosystems.

It has been noted that the BAFs of MeHg were separatedinto two populations (macrophytes+seston and benthos+fish; Figs. 5a and 6b). This may be the joint effect ofdifferent accumulation mechanisms for Hg between thesestons and the animals. Due to their short living time, thebioaccumulation of MeHg in phytoplankton should be muchweaker than in animals, and the greatest fraction of Hg ispresent in the form of IHg in the sestons (>95%, Table 2). Inaddition, the high concentration of IHg (THgMeHg) in the

Table 4 Pearson correlation coefficients among THg, MeHg, fish length, fish mass, 13C and 15N of five fish species

Species THg vs mass THg vs length MeHg vs mass MeHg vs length MeHg vs 13C MeHg vs 15N

A. nobilis 0.31 (n011) 0.22 (n011) 0.36 (n011) 0.36 (n011) 0.20 (n026) 0.24 (n013)

C. carpio 0.78b (n013) 0.35b (n013) 0.80b (n013) 0.81b (n013) 0.19 (n022) 0.06 (n011)

C. auratus 0.01 (n016) 0.18 (n016) 0.12 (n016) 0.31 (n016) 0.69b (n016) 0.11 (n06)

Erythroculter Mongolian Basilewsky 0.74 (n05) 0.62 (n05) 0.79 (n05) 0.68 (n05) 0.48 (n05) 0.77 (n03)

E. ilishaeformis Bleeker 0.80b (n018) 0.79b (n018) 0.77b (n018) 0.76b (n018) 0.01 (n020) 0.17 (n09)

a The significance level is 0.05b The significance level is 0.01

102

103

104

105

106

BA

Fs o

f MeH

g

A

Mac

roph

ytes

Sesto

n (sm

all)

Sesto

n (la

rge)

Ben

thic

ani

mal

s

Plan

kivo

res

Om

nivo

res

Pisc

ivor

es

101

102

103

BA

Fs o

f IH

g

B

Fig. 5 Box plot of the bioaccmulation factors for MeHg (a) and IHg(b) at different trophic levels from Lake Taihu. Note: the bioaccmula-tion factor of the benthic animals was calculated based on the averageHg concentration in the sediment, while the factors for the other trophiclevels were determined based on the average MeHg and THg concen-trations in the filtered lake water listed in Table 1

2828 Environ Sci Pollut Res (2012) 19:28202831

lake water from Lake Taihu (Table 1) may induce a highconcentration of IHg in phytoplankton, but IHg could beexcreted from the animals. This may lead to a great differencein the BAF for MeHg between the sestons and the animals(Fig. 6b), rather than for THg (Fig. 6a).

The foraging habitat constructed from the signature of13C in the biota indicates that energy fixation is dominatedby the pelagic pathway, while the food chain based onbenthic foraging cannot be constructed due to a lack of fishspecies in these specific foraging habitats (Liu 2009). There-fore, partitioning of IHg and MeHg between the water andthe phytoplankton should play the most important role inbioaccumulation and transfer of Hg in the animals of highertrophic levels from Lake Taihu because the uptake of Hg byphytoplankton should be the starting point for entry of Hginto the food web (Cabana and Rasmussen 1994; Herrin etal. 1998; Watras et al. 1998; Pickhardt and Fisher 2007).The logarithmic BAFs of MeHg for the small sestons (rang-ing from 3.0 to 3.1, with an average of 2.92.6) were lowcompared to those in other aquatic environments. For in-stance, the logarithmic BAFs of MeHg in the sestons fromLake Murray reached 4.5 (as calculated based on the MeHgconcentration) (Bowles et al. 2001). Hammerschmidt and

Fitzgerald (2006) derived a logarithmic BAF of MeHg of4.2 in the sestons from Long Island Sound. Watras et al.(1998) reported that the logarithmic BAFs were 56 inmicrosestons from 15 small lakes in the USA. Comparedto other eutrophic lakes, such as Clear Lake (Eagles-Smithet al. 2008) and Baihua Reservoir (Liu et al. 2012), MeHgconcentrations in the sestons (0.0140.38 ng g1) were alsomuch lower. Pickhardt et al. (2002) demonstrated that algalblooms could reduce the uptake of MeHg in the freshwaterfood web via a dilution effect. Therefore, the low BAFs ofMeHg in the sestons indicate that bioaccumulation of MeHgin the primary producers was strongly inhibited by eutro-phication, consequently resulting in a low biomagnificationof MeHg along the pelagic food chains in Lake Taihu.

3.4 Risk assessment of Hg exposure

For the population living around Lake Taihu, fish consump-tion is a dominant pathway for Hg exposure. The concen-trations of both THg and MeHg in the muscle tissues in allfish samples were below the WHO guideline of 500 ng g1

for THg for adults and even the guideline of 200 ng g1 forchildren, pregnant women and other vulnerable populations.According to the daily fish meals of 75 g for children of30 kg and 150 g for pregnant women of 60 kg and theaverage concentrations of THg (37 ng g1) and MeHg(22 ng g1) in the piscivores, the daily intake amounts ofTHg and MeHg were 92 ng day1 kg1 body weight and56 ng day1 kg1 body weight, respectively. These levels arefar lower than the tolerable intake of 230 ng day1 kg1 bodyweight for MeHg established by the Joint FAO/WHO ExpertCommittee on Food Additives (JECFA) to protect developingfetuses and children (JECFA 2004). The calculated resultsindicate a low risk of exposure to MeHg for the residents inthe Lake Taihu area.

4 Conclusions

The concentrations of THg and MeHg in the biota fromLake Taihu were in the range of 1.1100 and 0.02766 ng g1, respectively. The MeHg/THg ratio was relativelylow in all of the biota species, possibly resulting from thehigh concentrations of THg and the low concentrations ofMeHg in the water and sediments from Lake Taihu. Theconcentrations of Hg in the fish from Lake Taihu were muchlower than the guideline of 500 ng g1 in fish species that isrecommended by the WHO. The daily intake amounts ofTHg (92 ng day1 kg1 body weight) and MeHg(56 ng day1 kg1 body weight) were significantly lowerthan the tolerable intake of 230 ng day1 kg1 body weightfor MeHg for children and pregnant women (JECFA),

2 4 6 8 10 12 14 16 18 20 22-0.5

0.0

0.5

1.0

1.5

2.0

2.5

2 4 6 8 10 12 14 16 18 20 22-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Macrophytes Sestons Benthic animals Plankivores Omnivores Piscivores

Log

10 [T

Hg, n

g/g, w

w]

15N ()

R2 = 0.23, p < 0.001Log[THg] = 0.05215N - 0.20 A

Log

10 [M

eH

g, ng/

g, w

w]

15N ()

R2 = 0.36, p < 0.001Log[MeHg] = 0.11815N - 1.26B

Fig. 6 Relationship between 15N in different trophic levels and thelogarithm of THg (a) and the logarithm of MeHg (b) concentrations inall of the biota species from Lake Taihu

Environ Sci Pollut Res (2012) 19:28202831 2829

indicating a low risk of exposure to MeHg through theconsumption of aquatic products in the Lake Taihu area.

The results showed distinctly increasing bioaccumulationof MeHg from low to high trophic levels as determined bythe signature of 15N along the food web, and the highestlogarithmic BAF of MeHg in the piscivorous fish reachedapproximately 5.7. The relationship between 13C and 15Nindicated that energy transfer is dominated by a pelagicforaging habitat food chain. The low BAFs of MeHg inthe sestons and the macrophytes resulted in a low biomag-nification of Hg along the food web, possibly due to therapid increase in the primary productivity induced by eutro-phication in Lake Taihu.

Acknowledgments We thank Drs. Jianyang Guo and Nan Zhang fortheir contribution to the field work and the pre-treatment of all of thesamples. We also sincerely appreciate Professor Dingyong Wang andhis students for their help on the total mercury determination. Thiswork was financially supported by the National Basic Research Pro-gram of China on Water Environmental Quality Evolution and WaterQuality Criteria in Lakes (no. 2008CB418201), the National ScienceFund for Distinguished Young Scholars (no. 40925011), the Knowl-edge Innovation Program of the Chinese Academy of Sciences (no.KZCX2-EW-QN405) and the National Natural Science Foundation ofChina (no. 41103072).

References

Al-Reasi HA, Ababneh FA, Lean DR (2007) Evaluating mercurybiomagnification in fish from a tropical marine environment usingstable isotopes (delta C-13 and delta N-15). Environ Toxicol Chem26(8):15721581

Bowles KC, Apte SC, Maher WA, Kawei M, Smith R (2001) Bio-accumulation and biomagnification of mercury in Lake Murray,Papua New Guinea. Can J Fish Aquat Sci 58(5):888897

Cabana G, Rasmussen JB (1994) Modeling food-chain structure andcontaminant bioaccumulation using stable nitrogen isotopes. Nature372(6503):255257

Campbell L, Verburg P, Dixon DG, Hecky RE (2008) Mercury bio-magnification in the food web of Lake Tanganyika (Tanzania,East Africa). Sci Total Environ 402(23):184191

Chen LQ, Liu Y, Yang ZF, Chen Y, Zhou ZL, Liu H (2004) Ecologicalsuccession and sustainable development in Taihu Lake. J EastChina Norm Univ (Nat Sci) 4:99106

Chen CY, Dionne M, Mayes BM, Ward DM, Sturup S, Jackson BP(2009) Mercury bioavailability and bioaccumulation in estuarinefood webs in the Gulf of Maine. Environ Sci Technol 43(6):18041810

Clarkson TW (1993) Mercury: major issues in environmental health.Environ Heal Perspect 100:3138

Coelho JP, Nunes M, Dolbeth M, Pereira ME, Duarte AC, Pardal MA(2008) The role of two sediment-dwelling invertebrates on themercury transfer from sediments to the estuarine trophic web.Estuarine Coastal Shelf Sci 78(3):505512

Coelho JP, Pereira ME, Duarte AC, Pardal MA (2009) Contribution ofprimary producers to mercury trophic transfer in estuarine eco-systems: possible effects of eutrophication. Mar Pollut Bull 58(3):358365

Compeau GC, Bartha R (1985) Sulfate-reducing bacteria: Principalmethylators of mercury in anoxic estuarine sediment. Appliedand Environmental Microbiology 50:498502

Eagles-Smith CA, Suchanek TH, Colwell AE, Anderson NL (2008)Mercury trophic transfer in a eutrophic lake: the importantce ofhabitat-specific foraging. Ecol Appl 18(8):A196A212

Eagles-Smith CA, Ackerman JT, De La Cruz SEW, Takekawa JY(2009) Mercury bioaccumulation and risk to three waterbird for-aging guilds is influenced by foraging ecology and breedingstage. Environ Pollut 157(7):19932002

Fan C, Zhang L (2009) Lake Taihu: Principles of sediment pollutioniand remediation. Beijing, Science Press

Fontanella MC, Ravera O, Beone GM, Riccardi N, Cattani I (2009)Mercury distribution in the main compartments of the eutrophicLake Candia (Northern Italy). J Limnol 68(2):352358

Fry B (1988) Food web structure on Georges Bank from stable C, N,and S isotopic compositions. Limnol Oceanogr 33(5):11821190

Gagnon C, Fisher NS (1997) Bioavailability of sediment-bound methyland inorganic mercury to a marine bivalve. Environ Sci Technol31(4):993998

Gray JE, Hines ME (2009) Biogeochemical mercury methylation in-fluenced by reservoir eutrophication, Salmon Falls Creek Reser-voir, Idaho, USA. Chemical Geology 258(34):157167

Hall BD, Bodaly RA, Fudge RJP, Rudd JWM, Rosenberg DM (1997)Food as the dominant pathway of methylmercury uptake by fish.Water Air Soil Pollut 100(12):1324

Hammerschmidt CR, Fitzgerald WF (2006) Bioaccumulation and tro-phic transfer of methylmercury in Long Island Sound. ArchEnviron Contam Toxicol 51(3):416424

Herrin RT, Lathrop RC, Gorski PR, Andren AW (1998) Hypolimneticmethylmercury and its uptake by plankton during fall destratifi-cation: a key entry point of mercury into lake food chains? LimnolOceanogr 43(7):14761486

Hobson KA, Atwell L, Welch HE (1998) Biomagnification and bio-accumulation of mercury in an arctic marine food web: insightsfrom stable nitrogen isotope analysis. Can J Fish Aquat Sci 55(5):11141121

JECFA (2004) Methylmercury. In: Safety evaluation of certain foodadditives and contaminants. Report of the 61st Joint FAO/WHOExpert Committee on Food Additives. WHO Technical ReportSeries. Geneva, World Health Organization, InternationalProgramme on Chemical Safety pp 132139

Kainz M, Mazumder A (2005) Effect of algal and bacterial diet onmethyl mercury concentrations in zooplankton. Environ Sci Technol39(6):16661672

Kidd KA, Hesslein RH, Fudge RJP, Hallard KA (1995) The influence oftrophic level as measured by delta-N-15 on mercury concentrationsin fresh-water organisms. Water Air Soil Pollut 80(14):10111015

Knauer GA, Martin JH (1972) Mercury in a marine pelagic food chain.Limnol Oceanogr 17(6):868876

Liang L, Horvat M, Cernichiari E, Gelein B, Balogh S (1996) Simplesolvent extraction technique for elimination of matrix interferen-ces in the determination of methylmercury in environmental andbiological samples by ethylation gas chromatography cold vaporatomic fluorescence spectrometry. Talanta 43(11):18831888

Liang LN, Shi JB, He B, Jiang GB, Yuan CG (2003) Investigation ofmethylmercury and total mercury contamination in mollusk samplescollected from coastal sites along the Chinese Bohai Sea. J AgricFood Chem 51(25):73737378

Liu EF (2008) A study on diet composition of dominant fishes in LakeTaihu. J Fish China 32(3):395401

Liu ES (2009) Changes of fish community, themechanism of changes, andthe effects of the changes on environment in Taihu Lake. J Hydroecol2(4):814

Liu B, Yan H, Wang C, Li Q, Gudrona S, Spangenberg JE et al(2012) Insights into low fish mercury bioaccumulation in a

2830 Environ Sci Pollut Res (2012) 19:28202831

mercury-contaminated reservoir, Guizhou, China. Environ Pollut160(1):109117

Liu L (1997) Status quo analysis of lake nutrition abundance in China.Journal of Catastrophology 12(3):6165

Macalady JL, Mack EE, Nelson DC, Scow KM (2000) Sedimentmicrobial community structure and mercury methylation in mer-cury-polluted Clear Lake, California. Applied and EnvironmentalMicrobiology 66(4):14791488

Munthe J, Fjeld E, Meili M, Porvari P, Rognerud S, Verta M (2004)Mercury in nordic freshwater fish: an assessment of spatial vari-ability in relation to atmospheric deposition. 7th InternationalConference on Mercury as a Global Pollutant, Ljubljana

Pickhardt PC, Fisher NS (2007) Accumulation of inorganic and meth-ylmercury by freshwater phytoplankton in two contrasting waterbodies. Environ Sci Technol 41(1):125131

Pickhardt PC, Folt CL, Chen CY, Klaue B, Blum JD (2002) Algalblooms reduce the uptake of toxic methylmercury in freshwaterfood webs. Proc Natl Acad Sci U S A 99(7):44194423

Qin BQ, Xu PZ, Wu QL, Luo LC, Zhang YL (2007) Environmentalissues of Lake Taihu, China. Hydrobiologia 581:314

Qin BQ, Zhu GW, Gao G, Zhang YL, Li W, Paerl HW et al(2010) A drinking water crisis in Lake Taihu, China: linkageto climatic variability and lake management. Environ Manag45(1):105112

Sun SC, Mao R (2008) An introduction to Lake Taihu, China. B Qin(ed). Springer: Netherlands. 87:167

Ullrich SM, Tanton TW, Abdrashitova SA (2001) Mercury in theaquatic environment: a review of factors affecting methylation.Crit Rev Environ Sci Technol 31(3):241293

Wang WX, Stupakoff I, Gagnon C, Fisher NS (1998) Bioavailability ofinorganic and methylmercury to a marine deposit feeding poly-chaete. Environ Sci Technol 32(17):25642571

Wang S, Xing D, Jia Y, Li B, Wang K (2011) The distribution oftotal mercury and methyl mercury in a shallow hypereutrophiclake (Lake Taihu) in two seasons. Appl Geochem. doi:10.1016/j.apgeochem.2011.09.029

Watras CJ, Back RC, Halvorsen S, Hudson RJM, Morrison KA, WenteSP (1998) Bioaccumulation of mercury in pelagic freshwater foodwebs. Sci Total Environ 219(23):183208

Xu Y, Jin LN, Liang LN, Jiang GB (2006) Methylmercury, totalmercury and total selenium in four common freshwater fish spe-cies from Ya-Er Lake, China. Environmental Geochemistry andHealth 28(5):401407

You B, Wang T, Fan CX, Zhong J, Yin H, Li B (2008) Effect ofsediment resuspension on aqueous nutrient loading in grass typezone of Lake Taihu. Environ Sci 29(1):2631

Zhang Y, Feng S, Ma R, Liu M, Qin B (2008) Spatial pattern ofeuphotic depth and estimation of phytoplankton primary produc-tion in Lake Taihu in autumn 2004. J Lake Sci 20(3):380388

Zhu GW, Wang F, Gao G, Zhang YL (2008) Variability of phosphorusconcentration in large, shallow and eutrophic Lake Taihu, China.Water Environ Res 80(9):832839

Environ Sci Pollut Res (2012) 19:28202831 2831

Bioaccumulation and trophic transfer of mercury in a food web from a large, shallow, hypereutrophic lake (Lake Taihu) in ChinaAbstractAbstractAbstractAbstractAbstractIntroductionMethodsArea descriptionSample collectionsStable isotope (13C and15N) analysisTHg and MeHg analysisData analysis

Results and discussionIsotopic composition in the food webHg in the food webHg in the macrophytesHg in the sestonsHg in the benthic animalsHg in the fish

Bioaccumulation and biomagnification of the Hg in the food webRisk assessment of Hg exposure

ConclusionsReferences