HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTSOF ANYANG RIVER, KOREA: IMPLICATIONS

FOR WATER QUALITY

SANGHOON LEE1,∗, JI-WON MOON2 and HI-SOO MOON3

1Division of Biotechnology, The Catholic University of Korea, Buchon, Gyonggi-Do, 420-743,South Korea; 2Environmental Sciences Division, Building 1505, MS-6038, Oak Ridge NationalLaboratory, Oak Ridge, TN 37831-6038, USA; 3Department of Earth System Sciences, Yonsei

University, 134 Shinchon-dong, Seodaemun-ku, Seoul, 120-749, South Korea (∗author forcorrespodence: fax: +82-32-340-3765; e-mail: slee@catholic.ac.kr)

Received 14 October 2002; accepted in revised form 7 June 2003

Abstract. The objective of this study is to compare Anyang River bed sediments with water chemicalcomposition and to assess the anthropogenic chemical inputs into the river system. Eight samplinglocations were chosen along the river channel. Bed and suspended river sediments and water sampleswere collected, and analyzed for their chemical and physical composition. Data revealed that traceelement concentrations in the river water were generally below world average, except for As, Mn,Ni and Cr. Among the three phases: water, bed and suspended sediment, more than 99% of the traceelements was associated with the bed sediment. Concentrations of trace elements in the sedimentwere a function a particle size distribution and organic content. The calculated degrees of enrichmentbased on the least influenced sample (ASD 1) indicated the river sediments were enriched withrespect to background. The enrichment factors for Pb, Zn and As were relatively lower than forCr, Co, Ni and Zn. The difference in the enrichment seems to reflect the human activities influencein the basin, and specially for Cd. Speciation of the elements in the five different chemical forms inthe sediment by sequential extraction indicated that the reducible fraction was predominant for Fe,Zinc and Cu showed an irregular variation among the different fractions; whereas, Cd and Pb weremore regular. Zinc and Cu highly existed mostly in exchangeable forms. Acid soluble and reducibleforms were also important for most metals. The speciation implies that the metals associated withthe sediment are subject to release into water bodies as goechemical variables (pH and Eh) change.Currently, the introduced metals are deposited near the source area and are mostly associated withthe sediment, implying that the river bed sediment acts mainly as a sink, rather than a pool. Theaccumulated and enriched toxic trace elements can pose a potential pollution of river water.

Key words: Anyang River, heavy metals, sequential extraction, stream sediment, water chemistry

1. Introduction

Bed sediments in river system are repositories for various elements, acting bothas sinks and sources of supply for the elements to overlying water columns. Oncechemicals are introduced into a river, these chemicals undergo interactive reactionswith water bodies through series of processes including dissolution/precipitationand sedimentation/resuspension during migration and then finally they settle down

Environmental Geochemistry and Health 25: 433–452, 2003.© 2003 Kluwer Academic Publishers. Printed in the Netherlands.

434 SANGHOON LEE ET AL.

as bed sediments (Forstner and Wittmann, 1983). Among the various chemicals,trace elements are either associated with organic matter, sorbed onto Fe/Mn oxidesor complexed with hydroxides, sulfides and carbonates (Fornster and Wittmann,1983). The sorbed elements can be remobilized either by physical disturbanceslike floods or changes in geochemical parameters such as pH and redox potential(Dekov et al., 1997; Facetti et al., 1998; Soares et al., 1999). Therefore, when eval-uating river water quality, both sediments and water column should be considered(Zoumis et al., 2001 ).

The Anyang River is one of the main branches of the Han River, emerginginto the lower reaches of the Han River. The river received raw industrial wa-ter and sewage discharged from the basin area prior to 1992 when the sewagetreatment plant in the river began its treatment operation. It is probable that theriver sediments have been contaminated based upon the current measurement ofBOD in the river, which was 13.3 mg L−1 in 1999 (MOE, 2000). Major sourcesof trace elements into the river system include both from natural sources such asweathering of bedrock in the river basin or from those from anthropogenic origin(e.g. direct discharges of urban sewage or industrial wastewater). The introductionof the elements into river would be projected through their content in the sedi-ments. Thus the chemistry of river sediments would reflect the degree of naturaland anthropogenic influences (Forstner and Wittman, 1983; Dekov et al., 1997;Facetti et al., 1998; Soares et al., 1999). Despite the current contamination overthe world average concentrations (Martin and Meybeack, 1979) and potential forfurther contamination, to present no data are yet cited in literature of trace metalcontent in bed and suspended sediments of Anyang River, where knowledge of thedata is inevitable for assessing the potential for the changes in the river quality.Thus, the current work is mainly focussed on the geochemical characterization ofthe sediments and its implications on water quality of the river. This is done by asimultaneous collection of bed and suspended sediments and water (Anyang River)and determining their chemical and physical composition. As such, comprehensiveevaluation of the current water quality is attained through geochemical sedimentcharacterization. In addition, assessing and predicting the mobility and toxicity ofthe trace elements associated with the river bed sediments is also attained throughsequential extraction methods.

2. Materials and methods

2.1. STUDY AREA

The Anyang River originates from the south of Seoul and then joins the Han River,with a total length of 32.5 km. The river basin area covers 286 km2, extending oversix satellite cities bordering Seoul. The basin area is highly populated with 3.3million people and it is intensely industrialized.

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 435

Figure 1. Geological map and sampling sites.

The geology of the river basin is shown in Figure 1. Cambrian granitic gneissand overlying quaternary alluvium predominate the basin area. Limestone andlimesilicates also occur along the western side of the river. The sewage treatmentplant is located between the sampling points 7 and 8 (Figure 1). The pollution ofthe river water is probably due to the rapid urbanization and industrialization fromthe early 1970s, giving rise to a large increase in population. Among the indus-tries, the electronic sector comprises the largest part (44%), followed by machinery(24.5%) and chemicals/textiles (14%).

2.2. SAMPLING

Sampling was made in early June 1998, just before the rainy season in Korea. Bedsediments and river water were collected from eight different points along the riverfrom the upper reaches at intervals of 3–5 km downstream (Figure 1, Table I). Ateach sampling point, about 1 kg of bed sediment was collected using a stainlesssteel shovel. The collected sediments were stored in sealed PE bottles. They werethen air-dried in the laboratory and screened using a 2 mm sieve for further analysis.River water samples were collected in two 500 ml PE bottles; one acidified byHNO3 for cation analysis and the other untreated for anions. The water samples

436 SANGHOON LEE ET AL.

TABLE I

Characteristics of the sampling sites

Location Description of the area near the Remarks

sampling sites

AS 1 Next to textile factory but no discharge The upper most in the stream.

from the factory Width about 5 m

AS 2 Various industries including chemicals River width about 15 m

and textiles

AS 3 Machinery (car manufacturer) industry River width about 25–30 m

near the site

AS 4 Mixed industrial and residential area River width about 30 m

AS 5 Mostly residential area River width about 50 m

AS 6 Similar to AS 5 River width about 60–70 m

AS 7 Large residential area After sewage treatment plant,

similar river width 60–70 m

AS 8 Large residential area Just before the river emerges into

Han river, river width 60–70 m

were filtered in the field using a hand vacuum pump and 0.45 µm membrane filterbefore acidification. The water samples were then transferred to the laboratoryon the same day of collection, stored in an icebox. The water samples were thenstored in a refrigerator until analysis. The suspended sediment samples retained onthe micropore filters were air and oven-dried prior to storage in a dessicator untilfurther treatment and analysis.

2.3. ANALYSIS

2.3.1. Bed sedimentsMajor and trace elements of the river bed sediment samples were determined byenergy-dispersive XRF (X-lab 2000, Spectro Analytical Instruments). The sampleswere ground using an agate mortar and pestle to 200 mesh for the analysis. Totalcarbon contents in the sediments were analyzed using a wide-range carbon de-terminator (Leco WR112 788-600). The maximum error range for the result usinga standard coal sample is ±0.75%.

2.3.2. Suspended sedimentsTrace (metal) elements of the suspended sediments retained on the 0.45 µm mem-brane filter were determined by a pseudo-total dissolution method (modified APHAStandard methods for the examination of water and wastewater, 3030 ◦C). Fiftymilliliters of deionised water was added to the suspended solids in a 100 ml

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 437

Teflon beaker. The water and filter paper in the beaker were treated for 1 min inan ultrasonic bath to displace and loosen the suspended solids from the paper.Five milliliters of concentrated HNO3 was then added to the slurry and heatedfor 15 min at 80 ◦C. The final solution was made up to 100 ml with deionised waterand filtered through a 0.45 µm membrane filter. The solutions were analysed byICP-MS (Plasma Quad II Plus, VG Elemental) at the Seoul Branch of Korea BasicScience Institute.

2.3.3. WaterThe pH was measured in the field using Orion 290A portable pH meter(ThermoOrion). ICP-AES (138 Ultrace, Jovin Yvon), and ICP-MS (Plasma QuadII Plus, VG Elemental) were used for the determination of cations in the river water.Anions were analyzed by Ion chromatography (DX-120, Dionex).

2.3.4. Size fractionation of the sedimentsThe river sediment into clay, silt and sand fractions was fractionated using sed-imentation methods (Moore and Reynolds, 1997). Each fraction was stored forfurther mineralogical analysis.

2.3.5. Mineralogy of the bed and suspended sedimentsThe qualitative and semi-quantitative analysis of the minerals present was deter-mined using an X-ray diffractometer (MXP-3, Mac Science). The bed and sus-pended sediment samples were scanned between 3 and 60◦ 2θ , using Ni-filtered CuKα radiation, 40 kV/30 mA, divergent and scattering slits of 1 mm, a receiving slitof 0.15 mm, with stepping of 0.01◦ and scanning speed of 3◦/min. In addition to thewhole sediment sample, size fractionated samples were also analyzed for select-ed sediment samples, using XRD. Semi-quantitative analysis was performed fromthe calculations using reference intensity ratios (RIRs) of each mineral constituentobtained from PDF files (Chung, 1974).

2.3.6. Sequential extractionRiver sediments were sequentially extracted following the modified method ofTessier et al. (1979). The main purpose of the sequential extraction was to inves-tigate the fractionation of the elements in solids through several different steps:(1) Exhcangeable fraction (Salt-displaceable, agitation with 1 M MgCl2), (2) Car-bonate bound (acid extractable fraction 1 M CH3COONa), (3) Fe/Mn oxide bound(Reducible fraction, NH2OH.HCl), (4) Organic matter and sulphide bound (Ox-idisable fraction, 30% H2O2, 0.02 M HNO3, 3.2 M CH3COONH4 in 20% HNO3),(5) Total extraction (Residual, HNO3, HNO3 + HClO4, 6 M HCl). All the extracted

438 SANGHOON LEE ET AL.

solutions were filtered and acidified at the end of the individual steps and thenanalysed using ICP-AES (138 Ultrace, Jovin Yvon).

3. Results and discussion

3.1. CHEMICAL COMPOSITION OF RIVER WATER

The pH values of the river water were neutral to slightly alkaline, within a narrowrange between 7.3 and 7.7 (Table II). The concentrations of Na, K, Mg, and Ca inthe Anynag River water were much higher than world average values (Martin andMeyback, 1979): Na and K six to seven times, Ca and Mg two times. These majorcations (Na, K, Mg, and Ca) reflect the influence of the basin geology, mostly com-posed of granites and gneisses. Source rock deduction (Hounslow, 1995) based onthe water analysis indicated weathering of granite and silicate minerals (Figure 2).Interaction of the geological media with the river water and the influence on thewater chemistry was calculated by the geochemical modelling program PHREEQC(Parkhurst, 1995) with MINTEQ database (Allison et al., 1991), using water chem-istry data. Most of the water samples were undersaturated with respect to carbonateminerals (Table III). Only some calcite and ferric oxides were within the saturationindex zone (SIZ). Minerals in the saturation index zone are actually taking partin the dissolution/precipitation reaction (Deutsch, 1997) and the calcite may con-trol the water pH, by supplying alkalinity, through dissolution. Chloride exhibitedwide and high concentration range, and this may be attributed to an anthropogenicorigin. The NO3

− concentration was highest for ASW 1 (31.6 mg L−1) due toan agricultural and suburban zone, and then, decreased downstream, showing anaverage value of 1.72 mg L−1. Denitrification processes seem to be going on asmore organic substances are transported to the river, consuming more oxygen. Thedecrease in NO3

− corresponds to the increase in NO2− and NH4

+ concentrationsas a result of denitrification.

Some of the trace elements concentrations, including As, Ni and Cr, wereslightly higher than the world average (Table II). In particular, the average con-centration of the Mn in the river water is much higher than the world averageof 8.2 µg L−1, recording 159 µg L−1. Concentrations of Al, Fe, Cu, Pb and Znwere lower than the world average values. The metal and synthetic detergents arethe main sources of Mn (Fetter, 1993) and the higher Mn concentrations in theAnyang River can be attributed to such anthropogenic inputs. Although, Mn ismore stable than Fe in an acidic environment, it generally precipitates in lowerpH conditions when in contact with the atmosphere such as in the stream water(Hem, 1992). Considering the pH values for the Anyang River were around 7, Mnin the water might be released from Fe–Mn solid solution or Mn-bearing minerals,rather than coprecipitated with other heavy metals. This is supported by the higherconcentrations for Mn in the river water, exceeding the values of world average(Table II). It was noteworthy that Hg was detectable in the river water with the

HE

AV

YM

ETA

LS

INT

HE

BE

DA

ND

SUSPE

ND

ED

SED

IME

NT

SO

FA

NY

AN

GR

IVE

R439

TABLE II

Chemical composition of Anyang River water

Naa Mga Ka Caa Sia Alb Feb Cub Cdb Asb Seb Pbb Mnb Znb Nib Hgb Crb pH F− a Cl− a NO3− a SO4

2− a Alkalinitya,c

ASW 1 10.7 3.96 4.28 21.6 5.32 2.50 18.0 1.44 0.23 0.86 0.32 BDLd 31.8 4.59 1.01 4.09 0.30 7.70 0.36 2.84 31.6 22.5 45.8ASW 2 51.0 5.39 10.2 34.1 3.40 3.50 11.0 2.24 0.23 3.25 0.40 BDL 247 35.9 6.56 2.70 4.03 7.26 1.11 95.5 2.47 87.8 122ASW 3 23.9 4.37 5.54 26.9 3.10 3.50 2.00 1.07 0.22 1.83 0.25 BDL 122 5.37 1.77 2.42 1.06 7.40 0.66 41.9 4.17 44.9 113ASW 4 34.3 5.15 8.32 34.3 3.60 8.20 4.00 2.79 0.22 3.13 0.63 0.20 216 13.8 2.85 2.24 1.89 7.58 0.88 63.8 0.75 52.4 85.3ASW 5 49.3 6.02 11.0 34.8 3.77 6.10 2.00 1.77 0.23 4.00 0.94 0.11 134 8.1 3.71 2.21 3.48 7.29 1.54 119 1.55 101 163ASW 6 46.2 6.00 11.5 27.8 2.87 3.50 12.0 1.13 0.22 2.67 1.12 0.02 174 6.26 3.70 2.19 2.63 7.50 1.21 92.6 1.97 61.2 153ASW 7 42.5 5.64 10.7 32.0 3.30 2.50 4.00 0.08 0.22 3.11 0.81 BDL 301 3.93 3.12 2.18 1.90 7.44 1.11 82.2 0.28 64.4 142ASW 8 38.8 6.11 9.90 30.7 2.73 2.60 2.00 1.33 BDL 2.95 1.32 BDL 47.6 8.11 4.80 2.15 6.48 7.34 0.98 72.4 0.87 55.3 140Average 37.1 5.33 8.92 30.3 3.51 4.05 6.88 1.48 0.22 2.72 0.72 0.11 159 10.8 3.44 2.52 2.72W.A.e 5.1 3.8 1.4 14.6 50 40 10 1.7 1 8.2 30 2.2 1

a Measured in mg L−1.b Measured in µg L−1.c Alkalinity in mg L−1 as CaCO3.d Below detection limit.e World average by Martin and Meyback (1979).

440 SANGHOON LEE ET AL.

Figure 2. Source-rock deduction using parameters based on the concentrations of river water samples.Ions in meq L−1 and SiO2 in mmol L−1.

highest value at ASW 1, 4.09 µg L−1, some unknown sources for the Hg weresuspected.

3.2. PHYSICAL AND MINERALOGICAL CHARACTERISTICS

OF SEDIMENTS

Physical and mineralogical properties of bed sediments are provided in Table IV.The texture of samples ASD 3 and 4 is sandy loam and 8 loam, based on sizefractionation. All other samples were sands. The sum of clay and silt contents ofASD 3, 4, and 8 were about 10 times higher then the other samples (3, 4, and 8;ranging from 35.2 to 56%, 1, 2, 5, 6, and 7; 2.8 to 10.6%). This may be related tothe river drainage pattern, the nature of bank materials, and the flow rate. The sites3 and 4 are located at the meandering points where alluvial sediments are depositedand the river width in sampling point 8 is much wider, slowing down the flow rate.

Quantitative mineralogical analysis of whole sediment samples is shown inTable IV. Predominant mineral compositions of the sediments were quartz (35–86%), feldspar (4.1–24%) and mica (3.2–33%). This seems to be the reflectionof lithology composition of the basin, which is mainly composed of granite andgneiss. The clay fractions in ASD 8 have been separated and analysed to seemineralogical variation, against whole sample. The clay fractions are mainly com-posed of illite and kaolinite and a minor amount of chlorite (Figure 3). Upwardoval curvature between 15◦ and 35◦ in the diffraction pattern indicates the pres-ence of amorphous material which is thought to be mainly organic materials andamorphous iron precipitates.

HE

AV

YM

ETA

LS

INT

HE

BE

DA

ND

SUSPE

ND

ED

SED

IME

NT

SO

FA

NY

AN

GR

IVE

R441

TABLE III

Saturation indices of the Anyang River water

SIZa range ASW 1 ASW 2 ASW 3 ASW 4 ASW 5 ASW 6 ASW 7 ASW 8

Anhydrite 0.218 −2.511 −1.860 −2.181 −2.037 −1.811 −2.085 −2.003 −2.075

Calcite 0.424 −0.692 −0.595 −0.545 −0.396 −0.445 −0.333 −0.364 −0.479

Dolomite 0.855 −1.850 −1.719 −1.610 −1.345 −1.380 −1.061 −1.212 −1.390

Magnesite 0.401 −1.655 −1.620 −1.561 −1.445 −1.432 −1.224 −1.344 −1.406

Gypsum 0.229 −2.300 −1.650 −1.970 −1.826 −1.600 −1.874 −1.792 −1.864

Fe3(OH)8 1.011 0.707 −1.465 −2.941 −1.486 −3.546 −0.284 −1.920 −3.236

Goethite 0.050 6.032 5.162 4.716 5.261 4.478 5.635 5.070 4.598

Hematite 0.200 17.07 15.33 14.44 15.53 13.96 16.28 15.15 14.20

Maghemite 0.319 6.679 4.938 4.047 5.137 3.570 5.885 4.754 3.810

Magnetite 2.523 17.19 15.02 13.54 15.00 12.94 16.20 14.57 13.25

Siderite 0.528 −2.771 −2.347 −3.100 −3.041 −2.966 −2.255 −2.734 −3.010

Manganite 1.267 −4.519 −5.030 −4.876 −4.096 −5.221 −4.452 −4.392 −5.486

Rhodochrosite 0.557 −1.726 −0.941 −1.095 −0.802 −1.069 −0.745 −0.599 −1.497

Fluorite 0.548 −2.287 −1.217 −1.715 −1.377 −0.942 −1.217 −1.229 −1.346

Anorthite 0.986 −5.597 −5.824 −5.855 −4.807 −5.250 −5.894 −6.032 −6.232

Microcline 0.031 −1.514 −1.711 −2.034 −1.255 −1.303 −1.802 −1.813 −2.103

Kaolinite 0.372 2.088 2.618 2.370 2.963 3.137 2.146 2.065 2.076

Pyrolusite 2.069 −8.918 −9.869 −9.575 −8.615 −10.03 −9.051 −9.051 −10.24

Lepidocrocite 0.069 5.161 4.291 3.845 4.390 3.607 4.764 4.199 3.727

Montmorillonite 0.134 2.082 1.543 1.257 2.223 1.976 1.348 1.211 0.867

SiO2(a) 0.136 −1.241 −1.433 −1.474 −1.410 −1.388 −1.508 −1.447 −1.529

SiO2(am) 0.136 −1.549 −1.741 −1.782 −1.718 −1.696 −1.816 −1.755 −1.837

a Saturation index zone (5% of log K of a mineral phase).

442 SANGHOON LEE ET AL.

TABLE IV

Physical and mineralogical characteristics of bed sediments

pH Texture (%) Quantitative analysis (%)

Clay Silt Sand Quartz Microcline Illite Anorthite

ASD 1 7.0 Sand 4.1 6.5 89 78 6.0 4.0 12

ASD 2 5.9 Sand 1.8 4.1 94 64 18 9.3 8.8

ASD 3 5.7 Sandy loam 5.2 30 65 60 9.0 12 19

ASD 4 6.0 Sandy loam 10 35 55 35 8.2 33 24

ASD 5 5.5 Sand 2.1 5.0 93 69 15 7.8 8.1

ASD 6 5.0 Sand 2.1 5.1 93 86 5.6 4.7 4.1

ASD 7 4.0 Sand 0.8 2.0 97 81 8.3 3.2 7.3

ASD 8 6.8 loam 14 42 45 63 7.9 16 13

Figure 3. Representative X-ray diffraction patterns of ASD 8; bulk, clay fractioned orientation,mounting on membrane filter from the bottom.

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 443

3.3. CHEMICAL COMPOSITION OF BED SEDIMENTS

Table V provides heavy metal concentrations in the river bed sediments. ASD3, 4 and 8 samples showed different chemical compositions compard with theother five samples. The ASD 3, 4, and 8 samples have higher values for Al2O3

TABLE V

Chemical composition of bed sediments from the Anyang River

Na2O MgO Al2O3 SiO2 P2O5 K2O CaO TiO2 MnO Fe2O3 T.C.a

Major elements (%)

ASD 1 1.31 0.58 9.92 81.3 0.28 3.62 0.61 0.19 0.03 2.13 0.94

ASD 2 1.43 0.71 10.1 80.3 0.24 3.27 0.85 0.40 0.03 2.70 0.64

ASD 3 2.02 1.73 17.8 65.3 0.62 3.39 1.43 0.72 0.07 6.40 1.47

ASD 4 1.97 2.24 19.6 59.3 0.77 3.76 2.44 0.91 0.10 8.21 2.58

ASD 5 1.36 0.63 8.93 81.9 0.30 3.90 0.69 0.25 0.03 1.97 0.39

ASD 6 1.38 0.51 8.83 83.2 0.25 3.30 0.55 0.23 0.03 1.73 0.51

ASD 7 1.73 0.55 9.90 81.6 0.19 3.34 0.63 0.28 0.02 1.72 0.21

ASD 8 1.62 1.62 18.1 63.3 0.86 3.28 1.45 0.73 0.10 8.09 3.66

Cr Co Ni Cu Zn As Se Cd Pb

Trace elements (mg kg−1)

ASD 1 46 9.4 12 22 101 1.3 0.0 0.2 37

ASD 2 58 7.7 20 54 165 3.8 0.2 0.0 50

ASD 3 113 50 37 99 330 8.2 0.6 0.7 75

ASD 4 121 54 44 138 546 10 0.3 1.2 98

ASD 5 38 4.0 17 21 180 0.0 0.1 0.5 34

ASD 6 44 3.5 19 33 240 1.3 0.0 1.4 38

ASD 7 47 4.4 31 48 386 0.3 0.0 1.7 37

ASD 8 177 43 65 224 798 11 0.3 1.2 199

Cr Co Ni Cu Zn As Se Cd Pb Fe Mn

Enrichment factor based on the data of ASD 1

ASD 1 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0

ASD 2 1.2 0.8 1.6 2.4 1.6 2.9 0.0 1.3 1.2 1.1

ASD 3 1.4 3.0 1.7 2.6 1.8 3.5 2.0 1.1 1.7 1.4

ASD 4 1.3 2.9 1.9 3.3 2.8 3.7 3.0 1.3 2.0 1.7

ASD 5 0.9 0.5 1.6 1.1 2.0 0.0 2.8 1.0 1.0 1.0

ASD 6 1.1 0.4 1.8 1.7 2.7 1.1 7.9 1.2 0.9 1.0

ASD 7 1.0 0.5 2.6 2.2 3.8 0.2 8.5 1.0 0.8 0.8

ASD 8 2.1 2.5 3.0 5.7 4.3 4.7 3.3 2.9 2.1 1.9

a Total carbon.

444 SANGHOON LEE ET AL.

(ranges for 3, 4, and 8; 17.8–19.6%, ranges for 1, 2, 5, 6, and 7; 8.83–10.10%),CaO (1.43–2.44%; 0.55–0.85%), MgO (1.62–2.24%; 0.51–0.71%), Fe2O3 (6.40–8.21%; 1.72–2.70%) and lower for SiO2 (59.3–65.3%; 80.3–83.2%). The chemicalcomposition of sediment was closely related to its size fractionation and mineralo-gical composition. ASD 3, 4, and 8 are finer in their particle sizes, compared withother samples (Table IV).

Concentrations of the trace elements were also higher for the ASD 3, 4, and 8and lower for 1, 2, 5, 6, and 7, similar to those of the major elements (Table V).The trace element concentrations were related to size fractionation, mainly claycontent (Tables IV and V). Trace elements concentrations were compared withDutch standards for soil/sediment (Ministry of Transport, Public Works and Wa-ter Management, 2000). The concentrations of Cr, Co, Ni, Cu, and Zn greatlyexceeded the target value (background level) for ASD 3, 4, and 8 (Figure 4).That of Zn also exceeded the target value for all samples except for ASD 1. Theconcentrations of Cd and Pb exceeded the target values for ASD 4 and 8. Thebed sediment sample, ASD 8, collected from the lowest reach of the river showedthe highest values for most elements, exceeding the target values for all elementsexcept for As. ASD 8 even exceeded the intervention values for Cu and Zn. Incomparison, As concentrations in the sediments were for all sites below targetvalues, showing no evidence of contamination. Particle size plays an importantrole in controlling elemental distribution in soil or sediment (Barbanti and Both-ner, 1993; Song et al., 1999) and this was demonstrated in the high association ofheavy metals including Cr, Cd, Co, Cu, Ni, Pb and Zn with sediment clay content(Tables IV and VI); as well as the association of metals with the total carbonexcept for Cd (Table VI). The sediment carbon is assumed to be an organic carbondue to sewage and wastewater discharges into the river. Al2O3 is assumed to bederived from the dissolution of silicate minerals in the igneous rocks and is used toidentify the association of the trace elements with rock-forming minerals. While,Cd and Se do not show any significant correlation with Al2O3, the other elementsincluding V, Cr, Cu, Ni, Pb and Zn show close correlations (Table VI), except atASD 8.

ASD 1 is the upper sampling site and upstream of the other sampling pointsand it would be reasonable to assume it would be the least contaminated sedi-ment, compared to the seven other samples. Aluminium is immobile at neutralpH ranges (pH 5–9) and was used as an indicator to the mobility of other elementsas follows: enrichment factor =[Msediment/Alsediment:]/[Muncontaminated sediment/

Aluncontaminated sediment] (Figure 5). Most elements in the sediments are enriched withrespect to the unpolluted sample, ASD 1 and the degree of enrichment seems to bea function of clay and organic contents. Co, Cd, Cr, Cu, Ni, Pb and Zn in ASD 3,4, and 8 are enriched with respect to ASD 1, with factors from around 1 up to 8.5times. Arsenic was also enriched reaching a factor more than 5 times at ASD 3, 4and 8. Although enrichment factors for most elements do not vary systematicallywith distance downstream, for Ni, Cd, Cu, and Zn they do increase with distance

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 445

Figure 4. Variations in ratios of trace element concentrations in river sediments in relation to Dutchtarget (background) values and intervention values.

from the upper site. Factors for Cr and Pb are relatively constant for all samplesexcept for ASD 8, where they are much higher.

The sources of heavy metals in the river sediments can be both from natural andanthropogenic origins. the addition of metal content to the natural supply in the bed

446 SANGHOON LEE ET AL.

TABLE VI

Correlation coefficients of major oxides (A) and minor elements (B) with T.C.

(A) Major oxides

Na2O MgO Al2O3 SiO2 P2O5 K2O CaO TiO2 MnO Fe2O3 T.C.a

Na2O 1.000

MgO 0.812 1.000

Al2O3 0.807 0.982 1.000

SiO2 −0.787 −0.989 −0.997 1.000

P2O5 0.628 0.927 0.956 −0.964 1.000

K2O −0.047 0.132 0.006 −0.052 0.049 1.000

CaO 0.765 0.970 0.920 −0.940 0.859 0.225 1.000

TiO2 0.814 0.984 0.976 −0.982 0.919 0.004 0.954 1.000

MnO 0.662 0.947 0.971 −0.979 0.992 0.006 0.897 0.951 1.000

Fe2O3 0.715 0.968 0.987 −0.992 0.984 0.006 0.914 0.971 0.996 1.000

T.C.a 0.443 0.812 0.867 −0.877 0.954 −0.061 0.759 0.821 0.953 0.925 1.000

(B) Minor elements

V Cr Co Ni Cu Zn As Se Cd Pb T.C.

V 1.000

Cr 0.965 1.000

Co 0.966 0.876 1.000

Ni 0.882 0.949 0.772 1.000

Cu 0.925 0.984 0.805 0.975 1.000

Zn 0.821 0.891 0.692 0.980 0.944 1.000

As 0.983 0.962 0.940 0.863 0.926 0.787 1.000

Se 0.767 0.693 0.841 0.557 0.586 0.402 0.787 1.000

Cd 0.220 0.240 0.167 0.477 0.320 0.588 0.129 −0.117 1.000

Pb 0.859 0.955 0.703 0.937 0.977 0.910 0.861 0.492 0.251 1.000

T.C. 0.936 0.962 0.827 0.891 0.958 0.865 0.923 0.541 0.193 0.953 1.000

Total carbon contents (%).Statistically significant at p < 0.05.

sediments from anthropogenic inputs can be observed by the correlations of Al2O3

and enrichment factors. The influence of metal content due to human activities isreflected at certain specific sites; for example ASD 3, 4 are near the industrial zoneand ASD 8 is near the highly populated area. This is supported by the fact that theenrichment factors generally increase from the upper to the lower reaches of theriver. Based on the comparison of increasing tendency of enrichment factor withwater chemistry, non-systematically increasing with distance indicate the sedimentprobably act as a sink rather than a pool for pollutants.

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 447

Figure 5. Variations of enrichment factors (EFs) of sediments according to the distance from sampleASD 1. Ef was calculated by [(m/Al) in sediment]/[(m/Al) in ASD 1 sample].

3.4. CHEMICAL COMPOSITION OF SUSPENDED SEDIMENTS

Heavy metals in the river system are associated with bed sediments as adsorbed orprecipitated phases and are also associated with suspended sediments (Forstnerand Wittmann, 1983). Bed sediments provide longer-term information whereassuspended ones reflect relatively short-term information of physical and chemicalchanges such as the external inputs of elements and/or re-suspension of these ele-ments from sediments within the river system (Zhang, 1995; Bilos et al., 1998).The distribution of heavy metals in the suspended sediments were significantlylower than those of the bed sediments (Table VII). More than 99% of heavy metalsare associated with the bed sediment. Anyang suspended matter was depleted forall elements (except for Fe) compared to the world average for suspended mat-ter (Martin and Meybeck, 1979). The weight of suspended solids retained in themembrane filters increased with distance downstream, as expected. Fe and Mnconcentrations also increase with distance downstream and Cr, Zn, Pb, Ni and Cushow wide fluctuation in concentration (Figure 6). As, Se and Hg concentrationsare under the detection limit. The close correlations between the trace elementconcentration and the clay, organic contents indicate that the mineralogical controlis an important factor influencing the trace element concentrations in the sediment.Heavy metal concentrations in the water and suspended sediments do not correlatewith those in the bed sediment. The BOD of the Anyang River is 13.3 mg L−1

448 SANGHOON LEE ET AL.

TABLE VII

Chemical composition of suspended sediments using partial extraction (mg kg−1)

S.W.a Cr Zn Cd Pb Ni Fe Co Mn Cu As Se Hg

ASS 1 0.1 BDL 9 BDL 2 BDL 690 18 10 5 BDL BDL BDL

ASS 2 9.2 103 765 BDL BDL 2 4111 BDL 68 25 BDL BDL BDL

ASS 3 1.5 4 24 BDL 1 1 1831 BDL 23 BDL BDL BDL BDL

ASS 4 3.7 7 83 BDL BDL BDL 2991 BDL 97 11 BDL BDL BDL

ASS 5 4.8 2 98 BDL 8 BDL 3121 18 364 6 BDL BDL BDL

ASS 6 6.1 7 102 BDL 4 5 2241 18 90 18 BDL BDL BDL

ASS 7 8.2 12 153 2 16 4 4731 17 227 29 BDL BDL BDL

ASS 8 14.2 13 162 2 17 BDL 5341 18 1020 35 BDL BDL BDL

a Weight of suspended solid (mg 500 mL−1).

Figure 6. Variations in concentrations of trace elements in suspended sediments from the AnyangRiver using partial extraction according to the distance from sample ASD 1. Symbols are the sameas Figure 5.

and the presence of organic substances in the water may facilitate complexationwith metals, subsequently settling out as sediments near the source area such asindustrialized and urbanized ones. Also hydrated irons generally act as a limitingfactor for the precipitation of goethite, ferrihydrite and ferric hydroxides, which are

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 449

generally present as amorphous coating materials on the sediments from chemicalweathering of biotite and pyroxens in granitic rocks and are important in metaladsorption/desorption (Moon et al., 2000). Lower Fe concentrations in the watercan be explained that the Fe in the river mainly exist as an amorphous iron hydrox-ides, Fe3(OH)8, in the sediment, participating in water-rock interaction based onthermodynamic calculation (Table III). High Fe concentrations in the suspendedsediment might support this.

3.5. SEQUENTIAL EXTRACTION OF RIVER SEDIMENT

The result from the sequential extraction of Fe, Cd, Cu, Pb and Zn in the sedi-ments is presented in Figure 7. For Fe, the reducible forms are predominant andan occurrence in the exchangeable fractions are present in the sediment. The otherdominant phase of Fe is likely to be a crystalline oxide form considering the highproportion of the residual fraction. ASD 3, 4, and 8 were highly associated withorganic sediment fraction, and this may be due to the high organic contents of thesesamples. Fractionation of the Cu and Zn in the sediments varied tremendouslywith location. For Cu, the residual sediment fraction is very small (comprisingof <1%), except for ASD 8 (about 18%). Thus, it is proposed that Cu mighthave been introduced to the river sediments largely from an anthropogenic source.The exchangeable fraction of the Cu in the sediment is relatively high, rangingfrom 17 to 40%. This fractionation implies that Cu is subject to a release fromthe sediment to the water body, depending on the geochemical conditions. Thespeciation of Zn in exchangeable fraction of sediment at ASD 8 is about 60%. Thehigh proportion of exchangeable Zn and the corresponding high total concentrationin the bed sediments is an indication to Zn waste discharge near ASD 8. Reduciblesediment fraction is also dominant for Zn and this may indicate that Zn is renderedto variation of sediment pH and Eh.

Cd and Pb exhibited similar patterns in their fractionation, showing a relativelyeven distribution among the five different chemical sediment fractions. The occur-rence of Pb in the exchangeable sediment fraction increased with distance. Cd inthe exchangeable sediment fraction also increased with distance but with a widervariation. Organically bound and carbonate bound fractions seem to be larger forthe Cd and Pb than those of Cu and Zn. The difference in the proportion of thefractions present may affect the potential for release of metals and their concentra-tions in the river water, along with other factors such as the total metal contents,and geochemical variables (pH and Eh). Changes in geochemical variables mayalso lead to the release of the acid extractable, carbonate bound and organicallybound fractions and these fractions should also be considered as a potential poolfor supplying metals to the river water. Reductive dissolution of Fe oxides mayinduce the release of elements adsorbed on the Fe oxides, resulting in higher metalconcentrations in the river water.

450 SANGHOON LEE ET AL.

Figure 7. Speciation of Fe, Pb, Cu, Zn, and Cd in bed sediments.

4. Conclusions

This study revealed that of the three phases: (river water, suspended sediments andsediments of the Anyang River system) more than 99% of the elements determinedwere associated with the river bed sediment. Trace element concentrations in theriver water were, with the exception of Mn, Ni, Cr, and As, below the world averagevalues. Denitrification of NO3

− in the river may have been induced by degradationof organic substances (BOD 13.3 mg L−1). The high organic content implies that

HEAVY METALS IN THE BED AND SUSPENDED SEDIMENTS OF ANYANG RIVER 451

trace elements introduced into the river may be easily complexed and settle downin bed sediments, rather than be transported as free ions or colloids. Thus river bedsediment acts as a sink removing trace elements from the water column, rather thanas a pool for supply. Manganese concentrations in the suspended sediments roseabruptly at ASD 8, whereas Fe increased downstream from ASD 2. As well as, Fein the suspended loads was related to change in NO3

− concentrations. As dissolvedoxygen is reduced, Fe may undergo reductive dissolution, which was supported bythe high Fe concentration in the suspended sediments. The suspended sedimentmay act as a bridge between dissolved ions in water and bed fractions in sediment.

The presence of trace elements in river sediments was a function of their min-eralogical properties and size fraction of the sediment, both being controlled byriver morphology. The degree of elemental enrichment of trace elements in thesediments compared with the least polluted sample (ASD 1) indicated that the sed-iment was enriched for specific elements at specific locations. Taking into accountthis information and the correlation with Al2O3 content that is thought to be mainlyfrom rock dissolution, the source of Cd seems to be dominantly from anthropogenicinput and Pb and Ni to a lesser extent. As and Co are also influenced by humanactivities. The sediment at ASD 8 was enriched in all elements and this is thoughtto be from the high population density in this sampling area.

Sequential extraction and fractionation of selected elements (Fe, Cd, Cu, Ni,and Zn) in the river sediment indicated that most elements were speciated in theexchangeable, acid-soluble and reducible sediment fractions. Although the ex-changeable sediment fraction is the most available, yet the river environment isvulnerable to external changes such as flood and accidental input of chemicals.Such physical and chemical changes may cause change in geochemical conditions(pH and Eh) and the release of trace elements associated with the sediment into thewater body.

Although the river water is not abstracted for local drinking water or agriculturalusage, the river is very close to a residential area and the contaminated river waterhas a potential for posing ecotoxicologically harm to humans near the river. Theriver has also been under restoration, mainly led by local people and supported bythe local authority. Due to such efforts, some disappeared wildlife species are nowreported to be returning to the area. Contamination of the bed sediments can alsobe a potential pool for supplying contaminants to the ecological chain in the riverand the significance of the bed sediments regarding the ecology should be studiedfurther.

Acknowledgement

This work was supported by grant No. R01-2000-000-00057-0 from the BasicResearch Program of the Korea Science & Engineering Foundation.

452 SANGHOON LEE ET AL.

References

Allison, J.D., Brown, D.S. and Novo-Grdac, K.J.: 1991, Minteqa2/Proefa2, A GeochemicalAssessment Model for Environmental Systems: Version 3.0 User’s Manual, USEPA Washington,DC.

Barbanti, A. and Bothner, M.H.: 1993, A procedure for partitioning bulk sediments into distinctgrain-size fractions for geochemical analysis, Environmental Geology 21, 3–31.

Bilos, C., Colombo, J.C. and Rodriguez Presa, M.J.: 1998, Trace metals in suspended particles,sediments and Asiatic clams (Corbicula fluminea) of the Rio de la Plata Estuary, Argentina,Environmental Pollution 99, 1–11.

Chung, F.H.: 1974, Quantitative interpretation of X-ray diffraction patterns of mixtures. II. Adia-batic principle of X-ray diffraction analysis of mixture, Journal of Applied Crystallography 7,526–531.

Dekov, V.M., Komy, Z., Araujo, F., van Put, A. and van Grieken, R.: 1997, Chemical composition ofsediments, suspended matter, river water and groundwater of the Nile (Aswan-Sohag traverse),The Science of the Total Environment 20, 195–210.

Deutsch, W.J.: 1997, Water/rock interactions, in: Groundwater Geochemistry – Fundamentls andApplication to Contamination, Lewis Publishers, Boca Raton, pp. 47–75.

Facetti, J., Dekov, V.M. and van Grieken, R.: 1998, Heavy metals in sediments from the paraguayRiver: a preliminary study, The Science of the Total Environment 209, 79–86.

Fetter, C.W.: 1993, Contaminant Hydrogeology, Prentice-Hall, Englewood Cliff, NJ, 458 pp.Forstner, U. and Wittmann, G.T.W.: 1983, Metal Pollution in the Aquatic Environment, 2nd edn,

Springer, Berlin, 486 pp.Hem, J.D.: 1992, Study and interpretation of the chemical characteristics of natural water. U.S. Geol.

Surv. Water-Supply Paper 2252.Hounslow, A.W.: 1995, Water quality interpretation, in: Water Quality data – Analysis and

Interpretation, CRC Press, Boca Raton, pp. 71–128.Martin, J.M. and Meyback, M.: 1979, Elemental mass-balance of material carried by major world

rivers, Marine Chemistry 7, 173–206.Ministry of Environment: 2000, Environmental White Book, Ministry of Environment, Seoul,

686 pp.Ministry of Transport, Public Works and Water Management: 2000, Annexes, Circular on Target

Values and Intervention Values for Soil Remediation, Netherlands, 51 pp.Moon, J.-W., Moon, H.-S., Woo, N.-C., Hahn, J.-S., Won, J.-S., Song, Y., Lin, X. and Zhao, Y.: 2000,

Evaluation of heavy metal contamination and implication of multiple sources from HunchunBasin, northeastern China, Environmental Geology 39, 1039–1052.

Moore, D.M. and Reynolds Jr., R.C.: 1997, X-ray Diffraction and the Identification and Analysis ofClay Minerals, 2nd edn, Oxford University Press, Oxford, 378 pp.

Parkhurst, D.L.: 1995, User’s Guide to Phreeqc–A Computer program for Speciation, Reaction-Path,Advective-Transport, and Inverse Geochemical Calculations. U.S. Geological Survey, 143 pp.

Soares, H.M.V.M., Boaventura, R.A.R., Machado, A.A.S.C. and Esteves da Silva, J.C.G.: 1999, Sed-iments as monitors of heavy metal contamination in the Ave river basin (Portuga): multivariateanalysis of data, Environmental Pollution 105, 311–323.

Song, Y., Wilson, M.J., Moon, H.-S., Bacon, J.R. and Bain, D.C.: 1999, Chemical and mineralogicalforms of lead, zinc and cadmium in particle size fractions of some wastes, sediments and soils inKorea, Applied Geochemistry 14, 621–633.

Tessier, A., Campbell, P.G.C. and Bisson, M.: 1979, Sequential extraction procedure for thespeciation of particulate trace metals, Analytical Chemistry 51, 844–851.

Zhang, J.: 1995, Geochemistry of trace metals from Chinese river/estuary systems: an overview,Estuarine Coastal and Shelf Science 41, 631–658.

Zoumins, T., Schmidt, A., Grigorova, L. and Calmano, W.: 2001, Contaminant in sediments:remobilization and demobilization, The Science of the Total Environment 266, 195–202.