Environ Monit Assess (2010) 161:217–227DOI 10.1007/s10661-008-0739-y

Assessment of heavy metal contamination in the sedimentsof Nansihu Lake Catchment, China

Enfeng Liu · Ji Shen · Liyuan Yang ·Enlou Zhang · Xianghua Meng · Jianjun Wang

Received: 8 August 2008 / Accepted: 23 December 2008 / Published online: 23 January 2009© Springer Science + Business Media B.V. 2009

Abstract At present, anthropogenic contributionof heavy metals far exceeds natural input in someaquatic sediment, but the proportions are difficultto differentiate due to the changes in sedimentcharacters. In this paper, the metal (Al, Fe, K, Mg,Ca, Cr, Cu, Ni, and Zn) concentrations, grain size,and total organic carbon (TOC) content in the sur-face and core sediments of Nansihu Lake Catch-ment (the open lake and six inflow rivers) weredetermined. The chemical speciations of the met-als (Al, Fe, Cr, Cu, Ni, and Zn) in the surface sed-iments were also analyzed. Approaches of factoranalysis, normalized enrichment factor (EF) andthe new non-residual fractions enrichment factor(KNRF) were used to differentiate the sources ofthe metals in the sediments, from detrital clasticdebris or anthropogenic input, and to quantifythe anthropogenic contamination. The results in-dicate that natural processes were more dominant

E. Liu (B) · J. Shen · E. Zhang · J. WangState Key Laboratory of Lake Scienceand Environment, Nanjing Institute of Geographyand Limnology, Chinese Academy of Sciences,73, East Beijing Road, Nanjing 210008,People’s Republic of Chinae-mail: efliu@niglas.ac.cn

L. Yang · X. MengSchool of City Development, University of Jinan,Jinan, China

in concentrating the metals in the surface and coresediments of the open lake. High concentrationof Ca and deficiency of other metals in the upperlayers of the sediment core were attributed tothe input of carbonate minerals in the catchmentwith increasing human activities since 1980s. HighTOC content magnified the deficiency of the met-als. Nevertheless, the EF and KNRF both revealmoderate to significant anthropogenic contamina-tion of Cr, Cu, Ni, and Zn in the surface sedimentsof Laoyun River and the estuary and Cr in thesurface sediments of Baima River. The proportionof non-residual fractions (acid soluble, reducible,and oxidizable fractions) of Cr, Cu, Ni, and Zn inthe contaminated sediments increased to 37–99%from the background levels less than 30%.

Keywords Heavy metal · Chemical speciation ·EF and KNRF · Contamination · Sediment ·Nansihu Lake

Introduction

Heavy metals in the sediment constitute a specialgroup component of aquatic system and receivespecial attention due to the increasing anthro-pogenic input and potential ecotoxicity (Nriaguand Pacyna 1988; Singh et al. 2005). More than90% of the anthropogenic metals are bound toparticulate matters and deposited on the bed,

218 Environ Monit Assess (2010) 161:217–227

synchronously with the debris from the weatheredmother rock and soil in the catchment (Gomez-Parra et al. 2000; Amin et al. 2009). Due to thevariation in weathering, erosion, transport condi-tions, and aquatic productivity, the accumulatedsediments have different characters in grain size,mineral, organic matters content, etc., which canalso create anomalously high heavy metals con-centration as well as anthropogenic contamination(Liaghati et al. 2003; Chen et al. 2004; Reimannand de Caritat 2005). The distinguishing of anthro-pogenic versus natural contribution of heavy met-als in the sediments and accurate quantification ofanthropogenic contamination are important andindispensable for the effective protecting produceand remedial actions of aquatic environment.

Some statistical approaches, such as the factoranalysis, were used to identify the parameters thatcontrol metal distribution in the sediment andhelp to differentiate the sources (Valdés et al.2005). Besides, geochemical procedure, normaliz-ing the metals concentrations by the inert metalsAl, Fe, etc., can be used to compensate for thetextural and composition variation of sediments(Çelo et al. 1999; Liaghati et al. 2003; Chen et al.2004; Reimann and de Caritat 2005; Kartal et al.2006). This approach and the subsequent normal-ized enrichment factor (EF) were usually appliedto quantify the anthropogenic contamination ofmetals in the sediments (Çelo et al. 1999; Liaghatiet al. 2003; Chen et al. 2004; Reimann and deCaritat 2005; Kartal et al. 2006).

In recent years, much attention was also beenpaid to the chemical speciation of the metals in thesediments. The metals are operationally definedas the fractions that are exchangeable and car-bonate bound, Fe/Mn oxide bound, and organicmatter and sulfide bound, which are usually re-ferred as the extractable fractions (non-residualfractions) in the sequential extraction procedures(Quevauviller et al. 1997; Abd El-Azim and El-Moselhy 2005; Singh et al. 2005). Besides, themetals are mostly bound to the mineral matrix inthe natural sediments, namely, the residual fac-tion (Abd El-Azim and El-Moselhy 2005; Singhet al. 2005). Anthropogenic contamination couldchange the speciation compositions of the metalsin the sediments, mainly contributing to the non-

residual fractions (Abd El-Azim and El-Moselhy2005; Singh et al. 2005). Based on these, the chem-ical speciation compositions of the metals werealso used as the indicator of anthropogenic con-tamination (Abd El-Azim and El-Moselhy 2005;Singh et al. 2005), but how to quantify the an-thropogenic proportion was seldom involved inprevious studies.

In this paper, the concentrations and chemicalspeciations of the metals in the sediments of Nan-sihu Lake catchment were determined. Combin-ing the approaches of factor analysis, normalizedEF, and a new non-residual fractions enrichmentfactor (KNRF), our study was aimed to (1) in-vestigate the spatial and temporal distribution ofmetals in the sediments and (2) differentiate theanthropogenic versus natural contribution of met-als in the sediments and quantify the anthro-pogenic contamination.

Study area

Nansihu Lake is one of the typical freshwaterlakes, located in eastern China, with a surfacearea of 1,266 km2, a mean depth of 1.46 m(Fig. 1). The catchment area of Nansihu Lakeis 31,700 km2. It is mainly composed of alluvialplain, with scattered Cambrian and Ordoviciancalcareous hills and mountains in the east of thecatchment. The average annual precipitation inNansihu Lake catchment is 684 mm, 71% of whichis concentrated in summer (July to September).Nansihu Lake is recharged primarily by the pre-cipitation and drained out from the outlet south ofthe lake. With rapid economy development since1980s, large quantities of wastewater from agricul-tural, industrial, and domestic sources were alsodrained into the lake (Compilation Committeeof Jining Water Conservancy 1997). The concen-trations of nitrogen and phosphorus in water ofNansihu Lake increased 20- to 30-fold in recent20 years. Excessive nutrient load had also inducedthe enrichment of phosphorus in the sedimentsand eutrophication of the lake (Liu et al. 2008).Yang et al. (2004) had performed a preliminaryinvestigation of heavy metal distribution in thesediment of Nansihu Lake, but the proportion ofanthropogenic input was less understanding.

Environ Monit Assess (2010) 161:217–227 219

Fig. 1 Location of Nansihu Lake and sampling sites

Materials and methods

Sampling

Sixteen surface sediments (0–1 cm) were collectedat seven sites of Nansihu Lake and nine sites of sixmain inflow rivers in September, 2005 (Fig. 1). Ashort sediment core (DU-3) was also taken in themiddle of the lake and subsampled at 1-cm resolu-tion (Fig. 1). The surface and core sediments wereall obtained using a gravity corer. The sampleswere transferred to plastic bags and transportedto the laboratory for analysis.

Digestion method

The EPA Method 3052 was referenced for thesamples digestion (USEPA 1996). Briefly, thedried and ground samples (about 0.125 g) were

pressure-digested in the Teflon bombs in a mi-crowave oven (Berghof MWS-3 Digester) with6 ml HNO3, 0.5 ml HCl, and 3.0 ml HF for 15 minat 180 ± 5◦C. Then, the solution and residue weretransferred into a Teflon breaker; 0.5 ml HClO4

and 0.25 ml H2O2 were added and braised nearlydried on a heating block at about 200◦C. Finally,2.5 ml HNO3 (1 mol/l) was added and diluted to25 ml with double-distilled deionized water foranalysis.

Sequential extraction procedure

The sequential extraction procedure recommend-ed by Standard, Measurements, and Testing(formerly the Community Bureau of Reference)Programme was applied to determine the metalchemical speciations in the surface sediments(Quevauviller et al. 1997). It includes four frac-tions: acid-soluble fraction (exchangeable and

220 Environ Monit Assess (2010) 161:217–227

carbonate bound), reducible fraction (Fe/Mn ox-ide bound), oxidizable fraction (organic matterand sulfide bound), and residual fraction (boundto the mineral matrix). The three non-residualfractions (acid-soluble, reducible, oxidizable frac-tions) were all determined in this paper.

Metal concentration determinationand quality control

The metal concentrations in the solutions pre-pared by digestion and extraction procedureswere determined using inductively coupledplasma–atomic emission spectrometry (LeemanLabs, Profile DV). It includes the major rock-forming metals Al, Fe, K, Mg, and Ca and theheavy metals Cr, Cu, Ni, and Zn for the digestionsolution and Al, Fe, Cr, Cu, Ni, and Zn for theextraction solution.

For quality control, a blank and a certified ref-erence material GSD-11, supplied by the ChineseAcademy of Geological Sciences, were includedin each batch of 12 samples during the digestion.The analytical accuracy is better than 93% of thecertified values for the metals.

During the sequential extraction analysis, ac-curacy control was also performed with parallelsamples from sites 1 and 6. The maximum relativestandard deviation is lower than 10%, with the ex-ception of the reducible fraction of Cu (20.6%) inthe sample from site 1. It indicates good precisionand reproducibility for the SM&T procedure usedin this study.

Grain size and TOC analysis

Grain size analysis was carried out using Malvernautomated laser-optical particle size analyzer(Mastersizer-2000) after organic matters destruc-tion with 5% H2O2.

Total organic carbon (TOC) content of thesediment was determined using a CE440 elemen-tal analyzer (EAI Company). The acetylanilinewas used as the reference material for accuracycontrol.

Results and discussions

Grain size and TOC

Figure 2 presents the grain size distributions andTOC contents in the surface and core sediments.The sediments were dominated by fine-grainedfractions (<16 μm). For the surface sediments, thecontents of clay (<4 μm) and fine silt (4–16 μm)were 18–45% and 26–47%, with relatively lowervalues in sites 1–7, which may associate withthe strong hydrodynamic conditions of the rivers(Valdés et al. 2005). Grain size distributions inCore DU-3 were similar to that in surface sed-iments, with 23–41% clay and 24–48% fine silt(Fig. 2). Due to the greater surface area of clayfraction and its scavenging and adsorption capac-ity to metals (Chen et al. 2004; Barbanti and Both-ner 1993), the content of clay fraction is used asthe indictor of sediment texture in the subsequentdiscussion.

Fig. 2 Grain size and TOC distributions in surface andcore (DU-3) sediments

Environ Monit Assess (2010) 161:217–227 221

TOC contents were with a range of 1–24% inthe surface sediments and were 3–16% in CoreDU-3, with higher levels in the upper layers(Fig. 2). The absence of typical correlation (r =0.230) between TOC and clay content indicatesthat the hydrodynamic process was not the mainreason for the concentration of organic mattersin the surface sediments (Valdés et al. 2005).Higher TOC contents in sites 6 and 10 shouldcorrelate with the aquatic productivity or anthro-pogenic input. Organic matters in the sediments ofthe open lake were mainly endogenesis, indicatedby the C/N ratio (Yang et al. 2007). Therefore,higher TOC contents in the surface sedimentsof the open lake and the upper layers of CoreDU-3 were correlated with the lake productivity(Shen et al. 1998).

Distribution of metals in surfaceand core sediments

The spatial and temporal distributions of the an-alyzed metals are present in Fig. 3. In the surface

sediments, no definite regular trends can be ob-served for the rock-forming metals Al, Fe, K, Mg,and Ca. As for Cr, Cu, Ni, and Zn, typical highconcentrations were presented in sites 6 and 10.Cr was also with higher concentrations in sites 1and 2.

In the core sediments, there were two typicalvariation stages for the metals. Mean levels ofAl, Fe, K, Mg, Cr, Cu, Ni, and Zn were lower in0- to 8-cm depth and higher below 8-cm depth.However, Ca took on the opposite variation withother metals along the sediment core, with higherconcentration in 0- to 8-cm depth.

Factor analysis

In order to identify the parameters that controlmetal distribution in the sediment, factor analysiswas performed, and the varimax rotation factormatrix results are shown in Table 1.

For the core sediments, only one major factor isfound, which explains 91.3% of the total variance,with negative loading for Ca and TOC, positive

Fig. 3 Metalsconcentrations in surfaceand core (DU-3)sediments

222 Environ Monit Assess (2010) 161:217–227

Table 1 Rotated factorloadings of the variablesfor the surface and coresediments (only loadingsgreater than 0.4 areshown)

Parameters Surface sediments Core DU-3

Factor Factor

1 2 3 1

Al −0.675 0.675 – 0.992Fe 0.404 0.862 – 0.980K −0.848 0.466 – 0.996Mg −0.527 0.586 0.528 0.956Ca – – 0.914 −0.940Cr 0.805 – – 0.993Cu 0.934 – – 0.972Ni 0.936 – – 0.993Zn 0.841 – – 0.967Clay – 0.784 0.424 0.779TOC 0.912 – – −0.923Variance (%) 50.8 25.0 14.2 91.3

loading for Al, Fe, K, Mg, Cr, Cu, Ni, Zn, and clay.For the surface sediments, three factors explain90.0% of the total variance. Factor 1 accounts for50.8% of the total variance. One group is com-posed of Cr, Cu, Ni, Zn, Fe, and TOC, presentingthe positive loading. Another group is made up ofAl, K, and Mg, presenting the negative loading.Factor 2 explains 25.0% of the total variance andis characterized by positive loading of Al, Fe, K,Mg, and clay and negative loading of Ca. Factor 3accounts for 14.2% of the total variance and showsa positive loading of Ca, Mg, and clay.

Al, Fe, K, Mg, and Ca are rock-forming metalsand scarcely from anthropogenic contaminationsources. Factor analysis results denote that sedi-ment textural and mineral compositions were themain factors controlling the rock-forming metalsconcentrations in the surface and core sediments(Valdés et al. 2005). There had scarcely any au-togenetic carbonate in the freshwater lakes, sohigh concentration of Ca in the upper layers (0–8 cm) of Core DU-3 indicates the accumulationof carbonate minerals from the catchment. Highconcentration of Ca had also resulted in the lowconcentrations of Al, Fe, K, Mg, Cr, Cu, Ni, andZn in the top layers of the sediment core dueto the fact that the concentration of metals wasexpressed as the weight percentage of the driedsediment. High TOC content was another factordiluting the metal concentrations in the upperlayers of the sediment core (Valdés et al. 2005).The factor analysis results also indicate that thesources of Cr, Cu, Ni, Zn, and rock-forming metals

were different in the surface sediments. Partiallyof Cr, Cu, Ni, and Zn in some of the surfacesediments were from anthropogenic input, and theconcentrations of the metals with anthropogeniccontamination were influenced by organic mattersand Fe, namely, a proportion of them were boundby organic matters and cooperated with Fe/Mnoxides (Fig. 5; Reimann and de Caritat 2005).

EF of the metals

In order to compensate the influence of sedi-ment characters on metals concentrations and toquantify the anthropogenic input, the geochemicalnormalization approach is applied, and the nor-malized EF of the metals is calculated accordingto the following equation.

EF = (M

/X

)sample

/(M

/X

)background (1)

where M is the target metal, X is the selectednormalizer (reference metal) and (M/X)sample and(M/X)background are the ratios of target metal andthe normalizer in the interest and backgroundsediments, respectively. A five-category rankingsystem is used in this paper to denote the degreeof anthropogenic contamination. EF < 2 statesdeficiency to minimal contamination, EF = 2–5moderate contamination, EF = 5–20 significantcontamination, EF = 20–40 very high contamina-tion, and EF > 40 extremely high contamination(Sutherland 2000; Kartal et al. 2006).

Environ Monit Assess (2010) 161:217–227 223

To calculate the EF of the metals, the nor-malizer and the background levels of the metalsshould be determined. The inert elements Aland Fe are less anthropogenic contamination inaquatic sediment and were used as the normalizermost frequently (Çelo et al. 1999; Liaghati et al.2003; Chen et al. 2004; Reimann and de Caritat2005). Contrasting with Al, Fe has active chemicalfeatures and may remobilize easily in the sedimentwith geochemical conditions variation (Whiteleyand Pearce 2003). Consequently, Al appears to besuperior as the normalizer.

The metal concentrations in the shale or localpre-industrial sediments were usually used as thebackground levels (Reimann and de Caritat 2005).In view of the regional features, the weatheredmother rock and soil (Reimann and de Caritat2005; Abrahim and Parker 2008), pre-industriallevels of the metals in the sediment are superiorin this paper.

According to 137Cs dating results of the sedi-ment core (Yang et al. 2004; Fig. 1), the averagesedimentation rate was 3.5 mm/a. Moreover, it

was about 4 mm/a for Nansihu Lake calculatedby the average silting amount in recent years(Compilation Committee of Jining Water Conser-vancy 1997). Based on these, the depth of 8 cmof Core DU-3 was corresponding to early 1980s.In Nansihu Lake Catchment, the rapid economydevelopment and water pollution occurred since1980s. Therefore, anthropogenic contamination ofheavy metals in the sediment should be no earlierthan 1980s. There were uniform levels for themetals bellow 8-cm depth of Core DU-3, reflectingthe pre-industrial local baseline. Therefore, theaverage metals concentrations in the bottom threesamples of Core DU-3 are used as the backgroundlevels.

According to the selected background levelsand the normalizer, EF of the metals in the coreand surface sediments of Nansihu Lake Catch-ment are calculated by Eq. 1, and the results areshown in Fig. 4.

Ca presents higher EF in 0–8 cm of Core DU-3. Ca is enriched in the carbonate minerals andscarcely contaminated by human activities. It had

Fig. 4 Enrichment factor(EF) of metals in surfaceand core (DU-3)sediments (dotted line,EF = 1)

224 Environ Monit Assess (2010) 161:217–227

been indicated that, with the rapid developmentof agriculture in Nansihu Lake Catchment since1980s, intensive human activities aggravated thesoil erosion (Liu and Xin 1999). More carbonateminerals from the calcareous hills and mountainswere transported into the lake, which should bethe main reason of enriched Ca in the sediments(Liaghati et al. 2003). The EFs of Al, Fe, K, Mg,Cr, Cu, Ni, and Zn are approximately 1 along thesediment core, indicating that there had scarcelyany anthropogenic contamination. The lower con-centrations of the metals in the upper layers ofCore DU-3 were due to the dilution effect ofaccumulated carbonate minerals and TOC.

In the surface sediments, the EFs of Al, Fe, K,and Mg are approximate 1–2, indicating naturalsource of the metals. The varied EF of Ca shouldcorrelate with the local characters of the naturaldebris. Nevertheless, the EFs of Cr, Cu, Ni, andZn vary largely, with typical high values above 2in sites 6 and 10, indicating the enrichment andanthropogenic contamination.

Chemical speciations and KNRF of the metalsin surface sediments

The speciation compositions of Al, Fe, Cr, Cu, Ni,and Zn in the surface sediments are presented inFig. 5. Less than 10% of Al was in the non-residualfractions (mainly oxidizable fraction). As for Fe,non-residual fractions are relatively higher thanAl. Specifically, Fe was with 4–15% reducible frac-tion and 2–22% oxidizable fraction. Therefore, Aland Fe were mainly from natural debris (Abd El-Azim and El-Moselhy 2005). The lower concen-tration of non-residual factions for Al also verifiesour foregoing decision that Al as the normalizerfor EF calculation is optimum.

The speciation compositions of Cr, Cu, Ni, andZn varied largely in spatial, different with thatof Al and Fe. They were mainly presented inthe non-residual fractions (acid soluble, reducible,oxidizable fraction) in the sediments of sites 6and 10, accounting for 48–99% of the metal con-centrations. Cr was also with higher non-residualfractions content (37–64%) in sites 1 and 2. Morespecifically, Cr and Cu were mainly presented inthe oxidizable fraction; Ni was mainly presentedin the oxidizable and acid soluble fractions; Zn

was commonly found in the acid soluble fraction,as well as the reducible and oxidizable fractions.For other surface sediments, it was with only 10–30% non-residual fractions for Cr, Cu, Ni, andZn. The enriched non-residual fractions of Cr,Cu, Ni, and Zn in sites 6 and 10 and Cr in sites1 and 2 indicates that they were partially fromanthropogenic sources.

In order to quantify the anthropogenic contam-ination of the metals according to the speciationcompositions, the KNRF approach is presented in

Fig. 5 Chemical speciations of the metals in surfacesediments

Environ Monit Assess (2010) 161:217–227 225

this paper. The KNRF of the metals is calculatedusing the following formula.

KNRF = (MNRF(a)

/MTOT(a)

)/MNRF(b)

/MTOT(b)

(2)

where MNRF(a) and MNRF(b) are the non-residualfractions concentrations of the metals in the in-terest and background sediments, MTOT(a) andMTOT(b) are the total concentrations of the metalsin the interest and background sediments.

In the sediments of Sihe River, ZhuzhaoxinRiver, and Jinghang Canal (sites 3, 4, 7–9), Cr,Cu, Ni, and Zn were mainly natural origins indi-cated by the EF results. There had similar speci-ation compositions for Cr, Cu, Ni, and Zn in thefive sites, with less 30% of non-residual fractions,which also indicated the natural origin of the met-als. Therefore, the speciation compositions of Cr,Cu, Ni, and Zn in sites 3, 4, 7–9 are used as thebackground levels for KNRF calculation, and theresults are presented in Fig. 6. There have similarvariations for KNRF and EF in the surface sedi-ments (r > 0.9), generally with the values above2 in sites 6 and 10 for Cr, Cu, Ni, and Zn and insites 1–2 for Cr.

Quantification of anthropogenic contaminationof heavy metals in surface sediments

Anthropogenic contamination of the metals in thesurface sediments indicated by the KNRF and EFresults are discussed subsequently. The same five-category ranking system is also used for KNRF.

The KNRF and EFs of Cr, Cu, Ni, and Zn in thesediment of site 6 are 4.6–10.1 and 4.5–7.2, bothreflecting moderate to significant anthropogeniccontamination. Laoyun River is one drainageriver of Jining City and receives high input ofdomestic and industrial wastewater. There had nocontamination for Cr, Cu, Ni, and Zn in site 7indicated by the KNRF and EF, where LaoyunRiver originates. Therefore, the enrichment of Cr,Cu, Ni, and Zn in Laoyun River sediment was dueto the wastewater discharge of Jining City.

The KNRF and EF of Cr, Cu, Ni, and Zn inthe sediment of site 10 are 3.3–5.9 and 2.0–5.9,showing moderate to significant contamination.The KNRF and EF of Cr, Cu, Ni, and Zn in site

Fig. 6 KNRF and EF for Cr, Cu, Ni, and Zn in surfacesediments

5 are lower than 2, which indicate that GuangfuRiver was not mainly contributed to the enrich-ment of Cr, Cu, Ni, and Zn in site 10. Therefore,the enriched Cr, Cu, Ni, and Zn in site 10 maybe from the settling of anthropogenic metals fromLaoyun River, and this is also consistent with thedecreasing KNRF and EF values in sites 6 and 10.With the deposition of the pollutants near theestuary of Laoyun River, the KNRF and EF of Cr,Cu, Ni, and Zn are below 2 in sites 11–12, whereno typical contamination was found.

The KNRF and EF of Cr in site 1 are 8.3 and 2.5and are 4.3 and 1.6 in site 2. It generally reflectsthe moderate to significant contamination of Cr inthe sediments of Baima River. Though the EF ofNi and Zn in site 1 are below 2, their KNRF are

226 Environ Monit Assess (2010) 161:217–227

2.6 and 2.4, respectively, which indicate a minimalto moderate contamination. There are decreasingKNRF and EF values for Cr, Ni, and Zn alongBaima River, showing that anthropogenic sourcesof the metals were in the upstream area of BaimaRiver, most likely from Zoucheng City.

Though there had been anthropogenic contam-ination for Cr, Cu, Ni, and Zn in the sediments ofLaoyun River and Cr in Baima River, the KNRF

and EF all indicate that there was scarcely anycontamination for Cr, Cu, Ni, and Zn in the sur-face sediments of the open lake, which may be dueto the deposition of the anthropogenic metals inthe river and estuary before they were transportedinto the open lake.

The anthropogenic contamination degree ofthe metals in the surface sediments is similar ac-cording to the KNRF and EF results. It shows thatthe KNRF approach is feasible to be used in quan-tifying the metal contamination in the sedimentsas well as the EF approach.

Conclusions

The following are the conclusions of this study:

1. The spatial and temporal distribution of met-als in the sediments of Nansihu Lake catch-ment was studied. Although all metals showedvaried concentrations, the approaches of fac-tor analysis, normalized EF, and the KNRF

proposed in this paper were effectively usedto differentiate the natural and anthropogenicsources of the metals. Both the EF and KNRF

indicated similar anthropogenic contamina-tion degree of the metals.

2. Spatially, the rock-forming metals (Al, Fe, K,Mg, and Ca) and the heavy metals (Cr, Cu, Ni,and Zn) were mainly from natural debris andthe concentrations of which were influencedby the sediment characters in the open lakeand most of the inflow rivers. Cr, Cu, Ni, andZn in the sediments of Laoyun River and theestuary and Cr in the sediments of BaimaRiver were also partially from anthropogenicinput, with moderate to significant contami-nation. The anthropogenic proportions of the

metals were mainly presented in the non-residual fractions.

3. Temporally, there had no typical anthro-pogenic contamination for the metals in theopen lake. In the upper 0–8 cm of thesediment core, high concentration of Cademonstrated the strengthening erosion of thecalcareous debris in Nansihu Lake Catch-ment with increasing human activities since1980s. The accumulated carbonate mineralsand TOC also induced the deficiency of Al,Fe, K, Mg, Cr, Cu, Ni, and Zn in the sedi-ments.

4. Overall, natural processes were more domi-nant than anthropogenic inputs for the metalsin the surface and core sediments of NansihuLake Catchment, which is different with thestudy results of phosphorus (Liu et al. 2008).

Acknowledgements The work is financially supportedby the Scientific and Technical Supporting Programs ofMinistry of Science & Technology of China (No.2006BAC10B03), National Natural Science Foundation ofChina (No.40702058; 40672076), and Nanjing Institute ofGeography and Limnology, Chinese Academy of Sciences(No.S260015). The authors are thankful to Dr. RichardJones, Department of Geography, University of Exeter, forthe help during the preparation of this paper.

References

Abd El-Azim, H., & El-Moselhy, Kh. M. (2005). De-termination and partitioning of metals in sedimentsalong the Suez Canal by sequential extraction. Jour-nal of Marine Systems, 56, 363–374. doi:10.1016/j.jmarsys.2004.12.001.

Abrahim, G. M. S., & Parker, R. J. (2008). Assess-ment of heavy metal enrichment factors and thedegree of contamination in marine sediments fromTamaki Estuary, Auckland, New Zealand. Environ-mental Monitoring and Assessment, 136, 227–238.doi:10.1007/s10661-007-9678-2.

Amin, B., Ismail, A., Arshad, A., Yap, C. K., & Kamarudin,M. S. (2009). Anthropogenic impacts on heavy metalconcentrations in the coastal sediments of Dumai,Indonesia. Environmental Monitoring and Assess-ment, 148, 291–305. doi:10.1007/s10661-008-0159-z.

Barbanti, A., & Bothner, M. H. (1993). A procedurefor partitioning bulk sediments into distinct grain-size fractions for geochemical analysis. EnvironmentalGeology, 21, 3–13. doi:10.1007/BF00775044.

Çelo, V., Babi, D., Baraj, B., & Çullaj, A. (1999). An assess-ment of heavy metal pollution in the sediments along

Environ Monit Assess (2010) 161:217–227 227

the Albanian Coast. Water, Air, and Soil Pollution,111, 235–250. doi:10.1023/A:1005086208998.

Chen, Z. Y., Saito, Y., Kanai, Y., Wei, T. Y., Li, L. Q., Yao,H. S., et al. (2004). Low concentration of heavy metalsin the Yangtze estuarine sediments, China: A dilutingsetting. Estuarine, Coastal and Shelf Science, 60, 91–100. doi:10.1016/j.ecss.2003.11.021.

Compilation Committee of Jining Water Conservancy(1997). Jining water conservancy records. Jining: Jin-ing Municipal Bureau of Press.

Gomez-Parra, A., Forja, J. M., DelValls, T. A., Saenz,I., & Riba, I. (2000). Early contamination by heavymetals of the Guadalquiver estuary after the Az-nalcollar mining spill (SW Spain). Marine PollutionBulletin, 40, 1115–1123. doi:10.1016/S0025-326X(00)00065-5.

Kartal, S., Aydın, Z., & Tokalıoglu, S. (2006). Fraction-ation of metals in street sediment samples by us-ing the BCR sequential extraction procedure andmultivariate statistical elucidation of the data. Jour-nal of Hazardous Materials, 132, 80–89. doi:10.1016/j.jhazmat.2005.11.091.

Liaghati, T., Preda, M., & Cox, M. (2003). Heavy metaldistribution and controlling factors within coastalplain sediments, Bells Creek Catchment, southeastQueensland, Australia. Environment International, 29,935–948. doi:10.1016/S0160-4120(03)00060-6.

Liu, E. F., Shen, J., Yang, L. Y., Sun, Q. Y., & Wang, J. J.(2008). Phosphorus forms in the sediments of NansihuLake and its main inflow rivers. Geochimica, 37(3),190–296.

Liu, Q. H., & Xin, S. R. (1999). Analysis of soil and waterloss in Shandong Province and the sustainable devel-opmental measures. Chinese Journal of Soil Science,30(2), 49–50.

Nriagu, J. O., & Pacyna, J. M. (1988). Quantitative as-sessment of worldwide contamination of air, waterand soils by trace metals. Nature, 333, 134–139. doi:10.1038/333134a0.

Quevauviller, Ph., Rauret, G., López-Sánchez, J. F., Rubio,R., Ure, A., & Muntau, H. (1997). Certificationof trace metal extractable contents in a sedimentreference material (CRM 601) following a three-step sequential extraction procedure. The Scienceof the Total Environment, 205, 223–234. doi:10.1016/S0048-9697(97)00205-2.

Reimann, C., & de Caritat, P. (2005). Distinguishingbetween natural and anthropogenic sources for el-ements in the environment: Regional geochemicalsurveys versus enrichment factors. The Science ofthe Total Environment, 337, 91–107. doi:10.1016/j.scitotenv.2004.06.011.

Shen, J., Zhang, Z. K., Sun, Q. Y., Pan, H. X., Xia, W. L.,Wu, Y. H., et al. (1998). Character and paleoenviron-mental significance of the pigment and organic δ13C insediments of Nansihu Lake. Journal of Lake Science,10(2), 17–22.

Singh, K. P., Mohan, M., Singh, V. K., & Malik, M. (2005).Studies on distribution and fractionation of heavymetals in Gomati river sediment—A tributary of theGanges, India. Journal of Hydrology (Amsterdam),312, 14–27. doi:10.1016/j.jhydrol.2005.01.021.

Sutherland, R. A. (2000). Bed sediment-associatedtrace metals in an urban stream, Oahu, Hawaii.Environmental Geology, 39, 611–627. doi:10.1007/s002540050473.

USEPA (1996). Microwave assisted acid digestion ofsiliceous and organically based matrices, method 3052,Rev 0. In SW-846: Test methods for evaluating solidwaste, physical/chemical methods. Washington, DC:US Environmental Protection Agency Office of SolidWaste, 3052-1–3052-20.

Valdés, J., Vargas, G., Sifeddine, A., Ortlieb, L., &Guiñez, M. (2005). Distribution and enrichment eval-uation of heavy metals in Mejillones Bay (23◦S),Northern Chile: Geochemical and statistical ap-proach. Marine Pollution Bulletin, 50(12), 1558–1568.doi:10.1016/j.marpolbul.2005.06.024.

Whiteley, J. D., & Pearce, N. J. G. (2003). Metal dis-tribution during diagenesis in the contaminated sed-iments of Dulas Bay, Anglesey, N. Wales, UK.Applied Geochemistry, 18, 901–913. doi:10.1016/S0883-2927(02)00183-X.

Yang, L. Y., Shen, J., Zhang, Z. K., Jin, Z. D., & Zhu, Y.X. (2004). Human influence on heavy metal distribu-tion in the upper lake Nansihu Sediments, ShandongProvince, China. Chinese Journal of Geochemistry, 23,177–185. doi:10.1007/BF02868982.

Yang, L. Y., Shen, J., Liu, E. F., & Ji, J. F. (2007). Char-acteristics of nutrients distribution from recent sedi-ment in Lake Nansihu. Journal of Lake Science, 19(4),390–396.