Turk J Engin Environ Sci25 (2001) , 299 – 313.c© TÜBİTAK

Anthropogenic Pollution in İzmit Bay: Heavy Metal Concentrationsin Surface Sediments

D. YAŞARInstitute of Marine Sciences and Technology Dokuz Eylül University

Haydar Aliev Bulvarı, İnciraltı, İzmir 35340, TURKEYA. E. AKSU

Department of Earth Sciences, Memorial University of Newfoundland St. John’s,Newfoundland Canada A1B3X5

O. USLUInstitute of Marine Sciences and Technology Dokuz Eylül University

Haydar Aliev Bulvarı, İnciraltı, İzmir 35340, TURKEY

Received 04.09.1999

Abstract

The extent of marine pollution in İzmit Bay is studied using geochemical data in surface sediments. Theconcentrations of 41 elements in 24 samples establish that surface sediments in inner and central İzmit Baydisplay significant enrichments in Ag, As, Cd, Cr, Co, Cu, Hg, Mo, P, Pb, Sb, Ti, V, and Zn associatedwith high concentrations of total organic carbon and sulphur. Geo-accumulation indices indicate that theinner and central İzmit Bay surface sediments are moderately to very strongly polluted with respect toAg, Cd, Hg, Mo and Sb, and unpolluted to moderately polluted with respect to As, Co, Cu, Pb, and Zn.Despite total sedimentary concentrations above their pre-industrial background levels, geo-accumulationindices show that the surface sediments in İzmit Bay are unpolluted with respect to Cr, Ti and V. Exceptfor a localized area offshore Tuzla, the outer İzmit Bay is generally unpolluted with respect to heavy metals.

Key Words: İzmit Bay, Marine Pollution, Heavy Metals

İzmit Körfezi’nde Oluşan Antropojenik Kirlilik: Yüzey sedimanlarındaki AğırMetal Yoğunlukları

Özet

İzmit Körfezindeki deniz kirliliğinin boyutları yüzey sediman örneklerinde yapılan jeokimyasal çalışmasonucu incelenmiştir. Toplam 24 örnek üzerinde belirlenen 41 elementin analizleri sonucu , iç ve orta Körfezyüzey sedimanlarının toplam organik karbon ve sülfür ile gelişen Ag, As, Cd, Cr, Co, Cu, Hg, Mo, P, Pb, Sb,Ti, V, ve Zn gibi ağır metal yoğunluklarında önemli bir artış olduğu saptanmıştır. Yapılan jeo-akümülasyonhesaplamaları ise iç ve orta Körfezin Ag, Cd, Hg, Mo ve Sb yönünden kirli ve çok kirli, As, Co, Cu, Pb, andZn yönünden ise az kirli olduğunu göstermektedir. Yine jeo-akümülasyon hesaplamaları sonucu, sedimandakiyoğunluklarının sanayi öncesi yoğunluklarından fazla bulunmasına karşın İzmit Körfezi sedimanlarının Cr,Ti ve V yönünden kirlenmediği saptanmıştır. Dış körfez ise, Tuzla bölgesinin bazı bölgeleri dışında ağırmetal yönünden kirli değildir.

Anahtar Sözcükler: İzmit Körfezi, Deniz Kirliliği, Ağır Metaller

299

YAŞAR, AKSU, USLU

Introduction

İzmit Bay is a small east-west trending embay-ment situated along the northeastern Marmara Sea(Fig. 1). It is naturally divided into three physio-graphic regions: the inner bay, the central bay andthe outer bay. The shallower (

YAŞAR, AKSU, USLU

tal levels of the following elements in 24 surface and19 core samples: Ag, Al, As, Ba, Be, Ca, Ce, Cd,Cl, Co, Cr, Cu, Dy, Fe, Ga, Hg, K, La, Li, Mg, Mn,Mo, Na, Nb, Ni, P, Pb, Rb, S, Sb, Sc, Si, Sr, Th,Ti, U, V, Y, Zn and Zr. Table 1 gives the full geo-chemical data from İzmit Bay surface samples andthe MAR97-25 reference site, as well as the analyti-cal detection limits for these elements. Total organiccarbon was determined using the technique describedin Aksu et al. (1995).

For major and trace element determinations us-ing ICP-ES or AAS, except Ag, As, Cd, Hg, Rband Sb, sediments were powdered and ∼5 g of pow-der was treated with 30% H2O2 to extract elementsabsorbed by organic matter and amorphous oxides.Samples were then digested in 15 ml of HF, 5 ml ofconcentrated HCl and 5 ml of 1:1 HClO4 at 200◦Cand subsequently in 50 ml of 20% HCl at 100◦C.Measurements were carried out using an ARL ICP-ES instrument. Samples for Cd analyses were pro-cessed as above and Cd was measured using a VarianModel AA1275 AAS. For the determination of Ag,samples were digested for 16 hours in 2 ml of con-centrated HNO3 at room temperature, and then fora further 2 hours in a hot water bath at 90◦C. De-terminations were made using AAS. For As and Sbdeterminations, samples (5 g each) were partially di-gested in concentrated HNO3; subsequently, 1 ml ofconcentrated HCl was added and the samples weredigested for a further 90 minutes. As and Sb weremeasured using ICP-ES. For Hg determinations sam-ples (5 g each) were digested in 10 ml concentratedHNO3 and 1 ml of concentrated HCl for 10 minutes.Samples were then placed in a hot water bath at90◦C and digested for an additional 2 hours. Hg wasmeasured using ICP-ES; 4 on-peak and 4 off-peakmeasurements were made.

For XRF analyses, sediments were powdered and4 g of powder was mixed with 0.7 g of phenolic resinbinder. After homogenization, XRF pellets were pre-pared and baked for 15 minutes at 200◦C. Elementswere determined using an XRF ARL 8420+ sequen-tial wavelength-dispersive spectrometer, with an Rhend-window X-ray tube and an LiF200 crystal, spe-cially treated for enhanced heavy element sensitivity.For energies > 7.471 keV (20◦-50◦ 2Θ), a scintillationdetector was used with power settings at 75 kV and40 mA. For energies < 6.398 keV (50◦-150◦ 2Θ), aflow-proportional detector was used with power set-tings at 30 kV and 100 Ma (Longerich, 1993). Scantimes were 4-6 seconds for the most abundant el-

ements, 20 seconds for the trace elements and 100seconds for Rb, Y and Nb. The intensities werethen matrix-corrected by Compton correction for el-ements with emission energies greater than Fe, andby the LaChance-Traill algorithm (Longerich, 1993)for elements with emission energies less than and in-cluding Fe.

Results

Elemental associations

The inorganic geochemical data consist of mea-surements of 41 variables in 24 surface samples (Ta-ble 1). This large data set does not allow an imme-diate and unambiguous interpretation of elementalassociations and potential sources in the sediments.Factor analysis was used to obtain simple patternsfrom this complex data set. The technique extractsa small number of hypothetical variables (R mode)or samples (Q mode), referred to as “factors”, whichaccount for a given percentage of the total variancein the data set. Both Q and R mode factor analyseswere performed on the geochemical data from İzmitBay.Factor analysis results show that 3 factors ac-

count for a total of 96.5% of the total variance. Theremaining 3.5% of the total variance not accountedfor by the factor analysis is assumed to be random.Factor 1 accounts for 37.6% of the total variance andshows very high factor loadings in Al and Si, withstatistically significant loadings (i.e. >0.3; Klovanand Imbrie, 1971) in Be, Ca, Dy, Fe, Ga, K, La, Li,Mg, Mn, Nb, Ni, Rb, and Y. This factor representsthe aluminosilicates: the most common constituentof siliciclastic marine sediments. Factor 2 accountsfor 35.9% of the total variance and shows very highloadings in C and S, but also shows significant load-ings in Ag, As, Cd, Cr, Co, Cu, Hg, Mo, P, Pb, Sb,Ti, V, and Zn. This factor clearly represents theorganic matter and the enrichments of those metalswhich are most commonly associated with high totalorganic carbon in marine sediments (e.g., Aksu etal., 1998). Factor 3 accounts for 23.0% of the totalvariance and shows high factor loadings in Ca andMg, with significant loadings in Sr and Sc. This fac-tor probably represents the carbonate (biogenic anddetrital) in the study area.

Elemental background concentrations

The average shale values (Turekian and Wede-

301

YAŞAR, AKSU, USLU

pohl, 1961) commonly used for elemental back-ground concentrations (e.g., Ergin et al., 1991) can-not account for the local variations in rock compo-sition and chemistry in the Marmara Sea, and areunlikely to be representative of the İzmit Bay sur-face sediments. Instead, the background concentra-tions of the 40 elements used in this study were es-timated in a ∼200 cm-long core (MAR97-25) recov-ered from the Eastern Marmara Basin, immediatelywest of İzmit Bay. Nineteen samples between 20 cmand 200 cm depth in MAR 97-25 were selected forthe determination of elemental background concen-trations. Radiometric dates in cores from the easternMarmara Sea (e.g., Evans et al., 1989; Ergin et al.,1994) show that the rate of deposition in deep basinsis ∼100 cm per 1000 years, suggesting that the sed-iments used for the calculation of elemental back-ground concentrations were deposited between 2000and 200 yrBP. Because this core site is far removedfrom the regions of immediate anthropogenic efflu-ent input, and the sediments used in the calculationsclearly pre-date the industrial period, the elementalconcentrations in these samples must approximatethe background levels in the eastern Marmara Sea.The background level of each element was calculatedas the average value of that element in 19 samples incore MAR97-25, presented in Table 1.Anthropogenic pollution in surface sedimentsThe level of anthropogenic pollution in İzmit Bay

is evaluated using the technique described by Müller(1979), where the enrichment of an element abovethe background level (geo-accumulation index) is cal-culated using the following equation:

Igeo = log2 Cn / 1.5 x Bn

where Igeo = index of geo-accumulation, Cn = mea-sured concentration of the element in the muddy sed-iment being studied and Bn = geochemical back-ground value (e.g., MAR97-25); the factor 1.5 isused to compensate for possible variations of thebackground data due to post-depositional changes(Müller, 1979). Igeo values were calculated for 40elements (except C) in İzmit Bay surface sedimentsusing the inorganic geochemical data (Cn) and theelemental background concentrations (Bn) from thedeep Marmara Sea core MAR97-25. The data showthat Ag, As, Cd, Cr, Co, Cu, Hg, Mo, P, Pb, Sb, Ti,V, and Zn, all associated with factor 2, consistentlyexhibit total sedimentary concentrations above theirbackground levels in İzmit Bay surface sediments.Enrichments in chalcophile (eg., Cu, As, Ni

and Zn) and siderophile elements (eg., Cr, Ti andFe) associated with sediments rich in total or-ganic carbon and sulphur occur naturally in oxygen-deficient/anoxic basinal settings. In such environ-ments, sulphate reduction takes place at the seafloor,as in the Black Sea (Landing and Lewis, 1991),or immediately below the sediment-water interface,as for the Aegean Sea sapropel S1 (Aksu et al.,1995), leading to the formation of highly insolublesulphides in the presence of H2S, and hence en-richments. However, many of the chalcophile andsiderophile elements are also anthropogenic in ori-gin. Thus, the distinction between natural and an-thropogenic sources is often complicated. The oc-currence on İzmit Bay surface sediments of the rarerbut notorious environmental polluters Ag, Cd, Hgand Mo in quantities well above their backgroundconcentrations, and their clear associations and co-variances with total organic carbon, P and S as wellas the heavy metals As, Cu, Co, Cr, Pb, Sb, Ti, Vand Zn (e.g., Aksu et al., 1998), strongly suggest thatfactor 2 truly represents anthropogenic pollution inthe study area, rather than natural enrichment inpoorly oxygenated sediments.

Distribution of TOC, S and heavy metals insurface sediments

Total organic carbon values are high in İzmit Baysurface sediments, ranging from 1.9-2.6% in the innerbay, to 2.0-3.0% in the central bay and 0.5-0.7% inthe outer bay, except for 1.8% offshore of Tuzla (Fig.3; Table 1). These total organic carbon values areslightly higher than the 0.5-1.7% calculated by Erginet al. (1991) from İzmit Bay. Total sulphur concen-trations show a similar trend to that of total organiccarbon, with the highest values of 0.6-1.0% occurringin inner and central İzmit Bay, and show a notabledecline in outer İzmit Bay (Fig. 3; Table 1). Totalsilver concentrations range between 0.1 and 0.7 ppmin İzmit Bay surface sediments, with moderately highvalues occurring mainly in the central bay, and thehighest value recorded at station 18 offshore of Tuzla(Figs. 2, 4; Table 1). The background value of silverin core MAR97-25 is 0.08 ppm, which suggests that itis moderately enriched in inner and central bay sur-face sediments. Cadmium concentrations are highin central İzmit Bay (0.5-1.0 ppm) and moderate inthe inner bay (0.4-0.6 ppm), but are generally low(

YAŞAR, AKSU, USLU

Table

1.Elementalconcentrations(µg/g=

ppm)in

İzmitBay

surface

sedim

ents.DL=

detectionlimitin

ppm.EBC=

elem

entalbackgroundconcentrations(inppm)ineasternMarm

ara

Sea,calculatedastheaverageconcentrations

in19samplesfrom

core

MAR97-25.

<DL

Ag

Al

AsBa

BeC

Ca

Cd

CeCl

CoCr

CuDyFe

GaHg

KLaLi

DL0.03

302

10.1—

180.05

125

0.11

10.118

10.05

181

0.1

EBC

0.08

7505815

4001.9—

557720.1

741879928

15934

4.95480919

0.07

2115232

63Station 10.2

8725218

3411.819407388600.5

459188

5623677

3.46152020

1.0

2030124

5920.1

9942817

3171.826009364650.4

431932449

13059

3.56020421

1.0

2247622

5630.2

7720012

4331.723985167240.5

524040342

14361

3.85873919

1.5

1838626

4740.2

7738411

4441.626130304590.5

541995545

40652

2.94192013

1.0

2006927

5150.1

7830119

4951.923238451910.1

608898

3527137

3.34292417

1.3

2220229

4860.4

5621120

3221.928418342961.0

669948047

15049

4.35557519

1.5

2012839

6270.4

4663619

3111.931182363501.0

647860860

21546

5.55980717

1.0

1899444

6680.3

4975728

2761.630762357020.9

558164818

15043

3.55598118

0.6

1902030

5390.3

5017127

2921.826165333690.7

539163320

14445

3.35718418

1.0

1876928

5810

0.2

5622113

2841.822902422000.5

518991018

12741

3.25331715

0.8

1842027

5711

0.2

6958515

2971.822909320420.4

524233519

12339

3.55495317

0.7

2045529

6012

0.3

4801019

2881.823458396010.7

567140317

13438

3.85475217

0.5

1882532

5713

0.2

6078830

2911.312051616950.5

442719722

22547

2.54444311

0.4

1816024

3814

0.1

5866510

2921.27101

970460.3

482124026

28226

2.93632211

<DL1599425

3815

0.2

6494018

2791.66024

417090.3

525238235

16441

3.35317816

<DL1777026

5316

0.3

7795921

2901.75178

473990.4

502534418

13739

3.15242117

<DL1884827

5117

0.1

6329825

2591.35001

956990.4

432765715

11025

2.84067813

<DL1732827

4018

0.7

7020010

3021.812014994800.8

722292511

13282

3.63809815

<DL1966240

3619

0.1

4954026

2901.87124

538870.4

544948417

13538

3.35416317

<DL1752529

5620

0.2

7888714

3041.85146

474760.3

532858519

14839

3.15409518

<DL2018028

5521

0.2

4475713

2711.54190

420020.2

559194731

14438

3.05518015

<DL1624928

5222

0.1

7424915

2891.34056

454080.2

441423727

19926

2.84489214

<DL1608624

4123

0.1

559086

2031.03814

883780.3

362152118

20520

2.1347879

<DL1588822

3324

0.3

5178023

2851.76234

435890.3

539355018

12937

3.05536216

<DL1752228

54

303

YAŞAR, AKSU, USLU

Tab

le1.

Continued

Mg

Mn

Mo

Na

NbNi

PPbRb

SSb

ScSi

SrTh

Ti

UV

YZn

ZrDL50

10.1

640.71

11

0.7

9.9

0.03

0.127.8

12.8

13.4

10.61

1.1

EBC

283503325

1.9

1356814

8740927

1194606

0.7

23214626

2508.5

4615

1.5

14825

6399

Station 124969450

4.1

1225411

6660032

958722

1.4

19217210

1346.5

5683

1.0

17923

17177

230329558

4.8

1671212

4959820

917072

1.5

19224613

1244.8

6327

<DL17623

11091

320681427

8.1

2672212

3949827

949048

1.4

19210118

1542.6

5350

0.1

16924

135112

419517411

6.2

1926513

10844429

897707

0.9

12277935

1846.0

4302

2.5

12221

10772

517158999

7.2

1170715

7835522

1068041

1.0

12279826

2268.1

4349

3.3

10723

7580

617089421

8.1

9993712

4652843

1118375

1.8

16172541

1760.9

4872

0.8

18531

15977

713469562

8.5

5701413

6576242

1116254

0.9

16153599

1734.9

4812

4.3

16248

13981

818337893

8.9

6041712

4743141

1089415

0.8

14159822

1910.2

4467

1.2

18422

15872

914285766

5.3

9307312

4740643

1096693

0.8

16164914

1722.0

4795

1.8

17021

16376

1018792591

4.6

7584212

4241031

1036398

0.8

15179249

2032.7

4647

2.1

15919

14778

1124726999

2.8

2636712

4344835

1087611

1.1

15204464

1542.0

4892

1.2

15321

12481

1214501435

3.8

5218713

3640136

1057143

1.8

15159179

1722.0

4699

1.5

15024

13380

1317820507

5.2

1964110

5861943

774926

0.9

9255901

2313.2

3420

2.7

11714

15656

1419481450

2.7

201918

6444721

663878

0.6

9238320

5045.5

3343

0.1

9015

7952

1519371474

1.5

5433012

5145326

903558

0.7

16207449

1805.1

5134

0.4

15020

11283

1625806500

<DL1822912

4653531

953607

0.9

15232452

1973.3

5186

1.9

15320

11473

1720650427

<DL2371510

2452530

782843

0.7

10232238

3135.1

3943

2.4

10517

9954

1815467228

3.3

1653921

2565861

1027417

0.8

10221406

40310.5

5088

3.0

11625

24078

1915087438

<DL3668314

3741939

1022683

1.1

15167256

2185.5

4714

1.0

13420

2576

2028436622

<DL1807713

5852136

1063026

1.1

15228587

2124.5

4924

1.2

15219

11769

2112187468

1.8

9016612

4534828

892137

1.1

15162113

1822.9

5141

1.6

15320

9872

2222033525

<DL1616910

4449227

702214

0.9

13259116

1817.6

4963

0.6

12818

8270

2321065342

1.1

147678

4438821

622785

1.0

8234571

2333.4

3249

0.4

9113

5936

2414757514

<DL8734813

3640428

962219

1.3

16176525

1845.0

5061

0.4

14529

11276

304

YAŞAR, AKSU, USLU

Tab

le1.Continued

Mg

Mn

Mo

Na

NbNi

PPbRb

SSb

ScSi

SrTh

Ti

UV

YZn

ZrDL50

10.1

640.71

11

0.7

9.9

0.03

0.127.8

12.8

13.4

10.61

1.1

EBC

283503325

1.9

1356814

8740927

1194606

0.7

23214626

2508.5

4615

1.5

14825

6399

Station 124969450

4.1

1225411

6660032

958722

1.4

19217210

1346.5

5683

1.0

17923

17177

230329558

4.8

1671212

4959820

917072

1.5

19224613

1244.8

6327

<DL17623

11091

320681427

8.1

2672212

3949827

949048

1.4

19210118

1542.6

5350

0.1

16924

135112

419517411

6.2

1926513

10844429

897707

0.9

12277935

1846.0

4302

2.5

12221

10772

517158999

7.2

1170715

7835522

1068041

1.0

12279826

2268.1

4349

3.3

10723

7580

617089421

8.1

9993712

4652843

1118375

1.8

16172541

1760.9

4872

0.8

18531

15977

713469562

8.5

5701413

6576242

1116254

0.9

16153599

1734.9

4812

4.3

16248

13981

818337893

8.9

6041712

4743141

1089415

0.8

14159822

1910.2

4467

1.2

18422

15872

914285766

5.3

9307312

4740643

1096693

0.8

16164914

1722.0

4795

1.8

17021

16376

1018792591

4.6

7584212

4241031

1036398

0.8

15179249

2032.7

4647

2.1

15919

14778

1124726999

2.8

2636712

4344835

1087611

1.1

15204464

1542.0

4892

1.2

15321

12481

1214501435

3.8

5218713

3640136

1057143

1.8

15159179

1722.0

4699

1.5

15024

13380

1317820507

5.2

1964110

5861943

774926

0.9

9255901

2313.2

3420

2.7

11714

15656

1419481450

2.7

201918

6444721

663878

0.6

9238320

5045.5

3343

0.1

9015

7952

1519371474

1.5

5433012

5145326

903558

0.7

16207449

1805.1

5134

0.4

15020

11283

1625806500

<DL1822912

4653531

953607

0.9

15232452

1973.3

5186

1.9

15320

11473

1720650427

<DL2371510

2452530

782843

0.7

10232238

3135.1

3943

2.4

10517

9954

1815467228

3.3

1653921

2565861

1027417

0.8

10221406

40310.5

5088

3.0

11625

24078

1915087438

<DL3668314

3741939

1022683

1.1

15167256

2185.5

4714

1.0

13420

2576

2028436622

<DL1807713

5852136

1063026

1.1

15228587

2124.5

4924

1.2

15219

11769

2112187468

1.8

9016612

4534828

892137

1.1

15162113

1822.9

5141

1.6

15320

9872

2222033525

<DL1616910

4449227

702214

0.9

13259116

1817.6

4963

0.6

12818

8270

2321065342

1.1

147678

4438821

622785

1.0

8234571

2333.4

3249

0.4

9113

5936

2414757514

<DL8734813

3640428

962219

1.3

16176525

1845.0

5061

0.4

14529

11276

305

YAŞAR, AKSU, USLU

between its detection limit and 0.2 ppm, with an av-erage concentration of 0.1 ppm (Table 1). These datasuggest that the central and inner bay surface sedi-ments are considerably enriched in cadmium. Cobaltvalues are generally low (10-25 ppm) in outer andcentral İzmit Bay, but notably increase to 40-60 ppmin the northeastern outer bay and the inner bay (Fig.5; Table 1). The background level of this metal inthe deep Marmara Sea core is 28 ppm, which indi-cates that only surface sediments from the inner andnortheastern segment of the central İzmit Bay are

slightly enriched in cobalt. Copper values are moder-ate in the study area, with the highest values occur-ring in inner İzmit Bay surface sediments and at sta-tion 18 offshore of Tuzla (Fig. 5). Its abundance inthe uncontaminated Marmara Sea sediments rangesbetween 22 and 59 ppm, with an average of 34 ppm(Table 1), which is nearly identical to the averagecopper concentration of 35 ppm in shales (Adriano,1986). These data show that copper is only slightlyenriched in central and inner İzmit Bay sediments(Fig. 5; Table 1).

(%)

3.00

2.40

1.80

1.20

0.60

0.00

1.0

0.8

0.6

0.4

0.2

0

(%)

Figure 3. Bar graphs showing the distribution and abundance (%) of total sulphur (top) and total organic carbon(bottom) in İzmit Bay surface sediments.

(ppm)

1.00

0.80

0.60

0.40

0.20

0.00

0.70

0.56

0.42

0.28

0.14

0.00

(ppm)

Figure 4. Bar graphs showing the distribution and abundance (ppm) of total concentrations of silver (top) and cadmium(bottom) in İzmit Bay surface sediments. Arrows indicate background concentrations.

(ppm)

80

64

48

32

16

0

60

48

36

24

12

0

(ppm)

Figure 5. Bar graphs showing the distribution and abundance (ppm) of total concentrations of cobalt (top) and copper(bottom) in İzmit Bay surface sediments. Arrows indicate background concentrations.

Mercury concentrations are highest (1.2-1.5 ppm)in inner and eastern central İzmit Bay surface sedi-ments and show a dramatic decline to 0.5 ppm at the

Dil Burnu Entrance (Fig. 2), west of which the Hgvalues are below the detection limit in all samples(Fig. 6; Table 1). The background level of mer-

306

YAŞAR, AKSU, USLU

cury in Marmara Sea core MAR97-25 is calculatedto be 0.07 ppm, which indicates that all samples fromthe inner and central İzmit Bay are considerably en-riched in mercury. The distribution of molybdenumconcentrations in surface sediments is similar to thatof mercury, with the highest values of 7.0-9.0 ppmoccurring in inner and central İzmit Bay (Fig. 6).West of the Dil Burnu Entrance, Mo concentrationsare generally less than 3 ppm; for 6 samples valueswere below the detection limit. The pre-industrialbackground level of 1.9 ppm for this metal in coreMAR97-25 shows that it is enriched by 5-7 ppm inthe inner and central İzmit Bay surface sediments(Table 1).Nickel concentrations range between 30 and 110

ppm in İzmit Bay surface sediments, with the high-est values occurring in the eastern central bay andthe entrance to the inner bay (Table 1). Ergin etal. (1991) reports nickel values ∼15-20% higher thanthose in this study. The background level for nickel is87 ppm in the Marmara Sea core MAR97-25 (Table1), so that except for one sample (station 4) none ofthe surface sediments in İzmit Bay show enrichmentin nickel. Lead concentrations range between 21 and61 ppm in İzmit Bay, with the highest values occur-ring in the central bay (30-50 ppm) and at station 18,

offshore of Tuzla (Fig. 7; Table 1). These values arenearly identical to previously reported lead valuesfrom İzmit Bay (Ergin et al., 1991). A backgroundlevel of 27 ppm suggests that the surface sedimentsthroughout the study area are slightly enriched inlead.

Arsenic concentrations range between 6 and 30ppm in İzmit Bay surface sediments (Fig. 7; Table1). The background level of 15 ppm suggests thatsediments from the inner, central and eastern seg-ments of outer İzmit Bay are moderately enriched inarsenic. Antimony concentrations are variable buthigh in the study area, ranging between 0.5 and 1.8ppm, with the highest values occurring in the in-ner bay and along the northern shores of the centralbay (Fig. 8). Antimony is rare in the uncontam-inated Marmara Sea sediments (Table 1), with anaverage value of 0.7 ppm in core MAR97-25. Thisindicates that most of the inner and central bay sur-face sediments are moderately enriched in this metal(Table 1). Zinc values are moderate in İzmit Baywith slightly higher values in the central bay, andthe highest value at station 18 (Fig. 8; Table 1). Er-gin et al. (1991) reported zinc values 40-50% lowerthan in this study.

(ppm)

9.00

7.20

5.40

3.60

1.80

0.00

1.50

1.20

0.90

0.60

0.30

0.00

(ppm)

Figure 6. Bar graphs showing the distribution and abundance (ppm) of total concentrations of mercury (top) andmolybdenum (bottom) in İzmit Bay surface sediments. Arrows indicate background concentrations.

(ppm)

60

48

36

24

12

0

30

24

18

12

6

0

(ppm)

Figure 7. Bar graphs showing the distribution and abundance (ppm) of total concentrations of arsenic (top) and lead(bottom) in İzmit Bay surface sediments. Arrows indicate background concentrations.

307

YAŞAR, AKSU, USLU

(ppm)

240

192

144

96

48

0

1.80

1.44

1.08

0.72

0.36

0.00

(ppm)

Figure 8. Bar graphs showing the distribution and abundance (ppm) of total concentrations of antimony (top) and zinc(bottom) in İzmit Bay surface sediments. Arrows indicate background concentrations.

Degree of sediment contamination

Figure 9 shows the geographical distribution ofthe normalized varimax components of factor 2 inİzmit Bay, where zones A to F correspond respec-tively to factor loadings of 0.75. The highestconcentrations of heavy metals associated with fac-tor 2 occur in inner İzmit Bay (Zone F), east ofthe İpraş-Gölcük line (Fig.9). The geo-accumulationindices show that this zone is strongly to verystrongly polluted in Hg (25x1.5=48-fold enrichments;Igeo=5), strongly polluted in Cd (24x1.5=24-fold en-richments; Igeo=4), and moderately polluted in Ag,Mo and Sb (23x1.5=6-fold enrichments; Igeo=3).However, the zone is either unpolluted (Igeo=0) orunpolluted to moderately polluted with respect toAs, Cr, Co, Cu, P, Pb, Ti, V and Zn, despite thefact that these elements display concentrations abovetheir background levels (Fig. 9).Zone E is located along the northeast segment of

central İzmit Bay and displays strong to very strongpollution in Cd and Hg (Igeo=5), moderate to strongpollution in Ag (Igeo=3) and moderate pollution inMo and Sb (Igeo=2; Fig. 9). A trend of decreasingIgeo values is clearly visible across central İzmit baytowards the outer bay (Zones D through A). Zone Dextends from Diliskelesi to Hereke along the northernshores of central İzmit Bay, encircling Zone E, and inthe northwestern portion of the study area offshoreTuzla (Fig. 9). It shows strong to very strong pollu-tion in Cd (Igeo=5), strong pollution in Hg (Igeo=4),moderate to strong pollution in Ag (Igeo=3) andmoderate pollution in Mo (Igeo=2). Zone C formsa narrow band at the western margin of Zone D incentral İzmit Bay and encircles the small zone off-shore of Tuzla. The zone displays strong pollutionin Cd (Igeo=4), moderate to strong pollution in Hg(Igeo=3) and moderate pollution in Ag (Igeo=2; Fig.

9). Zone B mainly occurs in shallower shelf depths ofouter İzmit Bay, as well as the southwestern segmentof the central bay. These sediments are largely un-polluted except for moderate to strong pollution inCd (Igeo=3), and moderate pollution in Ag and Sb(Igeo=2; Fig. 9). Zone A is situated along the deeperaxis of outer İzmit Bay and shows only moderate pol-lution in Cd and Sb (Igeo=2; Fig. 9). Despite no-table total sedimentary concentrations of Cr, Ti andV, the geo-accumulation index shows that surfacesediments in İzmit Bay are unpolluted (Igeo=0) withrespect to these elements. Similarly, these sedimentsare unpolluted to moderately polluted (Igeo=1) withrespect to As, Co, Cu, P, Pb and Zn.

Discussion

The following general sources can be identifiedfor the pollution observed in the İzmit Bay surfacesediments:

1) Domestic pollution: untreated liquid and solidwaste produced by the inhabitants of the cities,supplied to the bay by the sewage network andsmall creeks;

2) Industrial pollution: untreated liquid and solidwaste produced by over 120 industrial centres,also supplied to the bay by the sewage networkand small creeks;

3) Atmospheric pollution: particulate emissionproduced by vehicles, trains and vessels; fly ashemission from coal- and fuel-oil-burning plants,by-products of open-air burning of municipaland industrial solid waste and garbage, all ofwhich are eventually transported into the in-ner and central Izmir Bay by precipitation anddrainage.

308

YAŞAR, AKSU, USLU

29°50’N—

—

—

— — — —

— — — —

—

—

—

29°45’N

29°40’N

29°50’E29°40’E29°30’E29°20’E

‹prafl‹zmit

HerekeDiliskelesiDar›ca

Tuzla

Gölcük

KaramürselDil

Burnu

Yalova

A

D

B

B

C

BD

E

F

0 3 6

Scale (km)

Figure 9. Map of the study area showing the degree of sediment contamination by Ag, As, Cd, Co, Cu, Hg, Mo, Pb,Sb and Zn, discussed in text (top) and the index of geo-accumulation, Igeo (bottom) in İzmit Bay surfacesediments.

There are numerous raw sewage outlets in the pe-riphery of İzmit Bay, actively discharging ∼370,000m3 day−1 of waste water into the bay, includ-ing untreated sewage (TÇV, 1998). Approximately∼51.4% (190,000 m3 day−1) of this discharge is do-mestic waste, 79% of which is discharged along thenorthern shores of İzmit Bay. Nearly all indus-trial waste is also untreated and is either dischargedthrough the domestic sewage network or via a fewsmall rivers and creeks. There are no data on thechemical composition of the sewage discharge intoİzmit Bay. However, high total organic carbon in thesurface sediments can be correlated with the organicwaste input into the bay. For example, there arereports of periods of an intense foul smell emanat-ing from the inner bay, particularly during the hotsummer months. During these months, the dissolved

oxygen levels in inner İzmit Bay decrease dramati-cally to values close to zero (TÇV, 1998), reflect-ing partial decomposition of organic waste in shal-low waters, and suggesting that the bottom waters inthis region must periodically become anoxic. Similarfoul odours are reported along the shoreline seawardof the major sewage outfalls of smaller communitiesand industrial centres, particularly along the north-ern shores of the bay.The high concentrations of total organic carbon

in surface sediments (Fig. 3) reflect organic matterinflux via effluent input, as well as input through therelatively high primary productivity in the water col-umn (10-170 g C m2 year−1; TÇV, 1998). High phos-phorus may be related to the use of detergents andthe discharge of phosphate through the sewage sys-tem, particularly in central İzmit Bay (Table 1). Fly

309

YAŞAR, AKSU, USLU

ash from coal- and petroleum-burning plants, as wellas the İpraş Refinery, is probably the major sourceof sulphur in İzmit Bay surface sediments (Fig. 3).Oxidation of the organic waste probably accounts forthe low dissolved oxygen levels observed within theentire İzmit Bay water column.There are over 120 major industrial centres along

İzmit Bay, largely concentrated along the north-ern shores (Fig. 10). Many of these industriesempty their untreated effluent into İzmit Bay; how-ever, there are no data available on the types andrates of heavy metal discharges from these indus-tries. Fifteen major industries discharge their wastewaters into the inner bay, including 3 large pulpmills and chlor-alkali plants, 4 pharmaceutical andchemical plants, 4 petrochemical plants, 3 steel fac-tories/smelters, and a leather factory (TÇV, 1998).SEKA (Fig. 10) is the largest of the pulp mills/chlor-

alkali plants and provides over 50% of the anthro-pogenic organic matter influx into İzmit Bay (TÇV,1998). A major proportion of the industrial effluentis supplied into the inner İzmit Bay via the KumlaRiver and its tributaries, including the dischargesfrom a large slaughter house, a number of leather-processing plants and small smelters. Central İzmitBay receives the effluent from 10 large industrial cen-tres, including 4 chemical plants, 2 petroleum re-fineries, a chlor-alkali plant, a paper mill, a ceramicfactory and two large textile factories (Fig. 10). Ex-cept for one large textile factory in Karamürsel, allthese industries are concentrated along the northernshores of central İzmit Bay (Fig. 10). Three majorindustries discharge their waste into the outer bay,including a chemical plant, an automobile plant, andan appliance factory.

Figure 10. Map of İzmit Bay, showing the distribution of major industrial centres (data from TÇV, 1998).

Fly ash constitutes a major pollutant from coal-burning and is known to contribute notable quan-tities of As, Cd, Cu, Mo, Pb, Sb, and Zn to theenvironment (Natusch et al., 1975). Nearly all thesmelters in the region use coal as their primary fuel.The fly ash emanating from these plants initially set-tles over the greater city of İzmit and it is probablyquickly washed by rain into the inner bay via thesewage network and the Kumla River. Leather tan-ning plants which employ large quantities of arsenicand chromium salts in the tanning procedure dis-charge their untreated waste products directly intoinner İzmit Bay. Nearly all antifouling paints appliedto commercial, naval and recreational vessels at thelarge military shipyard at Gölcük and the civilianshipyards and docks at Tuzla and Dil Iskelesi (Fig.

2) contain significant quantities of copper and lesserquantities of antimony, lead and mercury, probablycontributing the high total sedimentary concentra-tions of these heavy metals in inner İzmit Bay aswell as offshore of Tuzla (Figs. 5-8).

The main source of lead in the study area is prob-ably automobile traffic. Until recently, the great ma-jority of the cars in Turkey had no catalytic convert-ers and burned leaded petrol. Lead sulphide ore hasalso been used since antiquity for the extraction ofsilver impurities (Patterson, 1971). The effluent fromlarge chlor-alkali plants as well as industries whichuse bleach probably supply most of the mercury andmercury salts found in the İzmit Bay surface sedi-ments. The correlation between the highest mercuryvalues in inner İzmit Bay and the SEKA complex

310

YAŞAR, AKSU, USLU

is noteworthy (Fig. 10). Until recently, mercury-bearing compounds have been used in the preventionand control of industrial slime, and these compoundseventually ended up in inner İzmit Bay. Large bat-tery recycling plants are known for their potential fordelivering mercury and lead compounds into the en-vironment. Therefore, some mercury compounds areprobably also supplied by the Kumla River. In theprocess of chromium- and nickel-plating and galva-nizing, the surfaces to be plated are often cleanedusing strong acids. This process leaches a num-ber of heavy metals, which are released into İzmitBay via the sewage network. Furthermore, variouschromium, nickel and zinc pigments and compoundsare used in metal plating, probably contributingsome undeterminable quantities of these heavy met-als into the bay via the Kumla River.

In the greater İzmit Bay region, cadmium is ex-tensively used by various industries in the produc-tion of copper, lead, silver and aluminium alloys,and in various pigments. Some of the finest Turk-ish rugs are woven in Hereke, where cadmium, an-timony, copper, cobalt and zinc salts are commonlyused in the process of wool-dyeing. Cadmium saltsare also widely used by the ceramic industries, andmay also be contributed as wastes from zinc, lead andcopper smelters. High cadmium abundances in cen-tral İzmit Bay, particularly along the northern shore(Fig. 4), appear to reflect the contributions frommainly the large ceramic plants and textile indus-tries (Fig. 10). Silver in İzmit Bay surface sedimentsis probably largely supplied by photographic indus-tries, with lesser contributions from the disposal ofsilver-zinc and silver-cadmium batteries.

Molybdenum is frequently found in municipalsludges at levels that are above the background con-centrations (Adriano, 1986). In industrial regions itis mainly supplied by fly ash resulting from powergeneration using coal combustion, with estimates of15 tonnes of Mo per year for every 1000 MW of powergenerated (Adriano, 1986). Molybdenum salts arealso commonly used in fertilizers. The highest Moconcentrations in inner and central İzmit Bay, wellabove background levels, clearly suggest an indus-trial anthropogenic influx (Figs. 6, 10). The pri-mary sources of vanadium in the study area are prob-ably combustion byproducts of coal-burning plantsand the fly ash from petroleum-fired plants in theform of vanadium pentoxide. Antimony is used inpaints, lacquers, ceramic enamels and glass and pot-tery industries. It is also commonly used, alloyed

with lead, by manufacturing industries. The notableabundance of antimony in the inner and central İzmitBay surface sediments, above background levels, sug-gests contamination mainly from shipyards, and ce-ramic factories.

Ergin et al. (1991) concluded that the surfacesediments in İzmit Bay are basically uncontaminatedby anthropogenic pollution. This study was primar-ily based on the evaluation of the total sedimentaryconcentrations of a limited number of heavy metals:Cu, Co, Cr, Fe, Mn, Ni, Pb and Zn. The inorganicgeochemical data presented in this paper stronglycontradicts the conclusions of Ergin et al. (1991)and clearly show that the surface sediments in İzmitBay are moderately to very strongly polluted withrespect to Ag, Cd, Hg, Mo and Sb, with Igeo valuesbetween 2 and 5, and unpolluted to moderately pol-luted with respect to As, Co, Cu, Pb, and Zn, withIgeo values between 0 and 1. These data indicatethat anthropogenic pollution must be viewed seri-ously within this embayment. The heaviest contam-ination is observed along the northeastern segmentof the central bay offshore of Hereke and İpraş (ZoneE) and the entire inner bay (Zone F; Fig. 9), reflect-ing the high concentrations of chemical, petroleum,chlor-alkali and pharmaceutical industries (Fig. 10).The data conclusively show that the anthropogenicpollution observed in inner and central İzmit Bay isalmost exclusively confined to these regions by theshallow sill at the Dil Burnu Entrance (Fig. 9). Thelow tidal range and restricted water circulation be-tween the central and outer parts of İzmit Bay con-trol this confinement.

Most previous studies along the coastal watersof Turkey focussed on the surficial sedimentary con-centrations of only a limited numbers of elements;thus, the distinction between the natural and anthro-pogenic elemental associations was not always pos-sible (e.g., Ergin et al., 1991). In many instances,the sedimentary concentration of an element is as-sumed to be anthropogenic, because of the involve-ment of that element in polluted sediments else-where. The present study highlights the importanceof a broader elemental data set in the assessmentof anthropogenic pollution in small coastal embay-ments.

Increased concentrations of heavy metals in ma-rine sediments enhance the importance of sedimentsas a direct source of potential toxins in nearshoreenvironments. However, sedimentary geochemistrydata, such as those presented here, provide no ev-

311

YAŞAR, AKSU, USLU

idence for biological damage in a marine environ-ment. Therefore, a comprehensive assessment of ma-rine pollution should ideally include coupled studieslinking the levels of pollutants in the seawater withthose in the surface sediments, and the seawater andsediment with tissues in various biota. Furthermore,the bio-availability of a heavy metal is critical in theassessment of its damage in the ecosystem. Thus,studies determining the chemical speciation of heavymetals are important.

Conclusions

1. Inorganic geochemical data show that innerand central İzmit Bay surface sediments display highabundances of total organic carbon, S and P, asso-ciated with enrichments in Ag, As, Cd, Cr, Co, Cu,Hg, Mo, P, Pb, Sb, Ti, V, and Zn, which are abovetheir pre-industrial background levels. Distributionmaps illustrate that heavy metal concentrations arehighest in central and inner İzmit Bay, dramaticallydecreasing near Dil Burnu Entrance.2. The geo-accumulation indices show that the

surface sediments of central and inner İzmit Bayare moderately to very strongly polluted with re-spect to Ag, Cd, Hg, Mo and Sb, and unpollutedto moderately polluted with respect to As, Co, Cu,Pb, and Zn. Despite total sedimentary concentra-

tions above their pre-industrial background levels,geo-accumulation indices show that the surface sed-iments in İzmit Bay are unpolluted with respect toCr, Ti and V.3. Except for a localized area offshore of Tuzla,

outer İzmit Bay is generally unpolluted with respectto heavy metals.

Acknowledgments

We thank Prof. Dr. Erol Izdar, the Directorof the Piri Reis Foundation for Maritime and Ma-rine Resources Development and Education, for hiscontinual support and encouragement. We extendour special thanks to the officers and crew of theRV KocaP iriReis for their assistance in data ac-quisition, in particular Captain Mehmet Özsaygılıand Chief Engineer Ömer Çubuk. We acknowledgeresearch funds from the Natural Sciences and En-gineering Research Council of Canada (NSERC) toAksu (OGP0042760) and special grants from the PiriReis Foundation for Maritime and Marine ResourcesDevelopment and Education. Hüseyin Avni Benli(Efe) of the Institute of Marine Sciences and Tech-nology and Richard Hiscott of Memorial Universityof Newfoundland critically read the manuscript andprovided many valuable comments.

Reference

Adriano, D.C., Trace elements in the terrestrial en-vironments. Springer- Verlag, New York, Berlin, 5331986.

Aksu, A.E., Yaşar, D. and Mudie, P.J., “Paleo-climatic and paleoceanographic circumstances lead-ing to the development of sapropel layer S1 in theAegean Sea Basins”. Paleogeography Paleoclimatol-ogy Paleoecology, 116: 71-101, 1995.

Aksu, A.E., Yaşar, D. and Uslu, O., “Assessment ofpollution in Izmir Bay - heavy metal and organiccompound concentrations in surficial sediments”.Turkish Journal of Engineering and EnvironmentalSciences, 22: 387-415, 1998

Ergin, M., Saydam, C., Baştürk, Ö., Erdem, E. andYörük, R., “Heavy metal concentraitons in surfacesediments from the two coastal inlets (Golden HornEstuary and İzmit Bay) of the northeastern Sea ofMarmara”. Chemical Geology, 91: 269-285, 1991.

Ergin, M., Bodur, M.N., Yıldız, M., Ediger, D., Edi-ger, V., Yemencioğlu, S. and Yücesoy, F., “Sedimen-tation rates in the Sea of Marmara: a comparison ofresults based on organic carbon-primary productiv-

ity and 210Pb dating”. Continental Shelf Research,14: 1371-1387, 1994.

Evans, G., Erten, H., Alavi, S.N., Von Gunten, H.R.and Ergin, M., “Superficial deep-water sediments ofthe eastern Marmara basin”. Geo-Marine Letters, 9,27-36, 1989.

Klovan, J.E. and Imbrie, J., “An algorithm forFortran-IV program for large-scale Q-mode factoranalysis and calculation of factor scores”. Journalof Mathematical Geology, 3: 61-77, 1971.

Landing, W.M. and Lewis, B.L., “Thermodynamicmodelling of trace metal speciation in the BlackSea”. In: E. İzdar and J.W. Murray (Eds.), BlackSea Oceanography, NATO ASI Series C, KluwerAcademic Publisher, Dordrecht, 351: 125-160, 1991.

Longerich, H.P., “Application of a new treated LiFcrystal to X-ray fluorescence analysis of geologicalsamples”. X-Ray Spectrometry, 22: 114-118, 1993.

Müller, G., Schwermetalle in der Sedimenten desRheins - Vernderungen seit 1971. Umschau 72: 192-193, 1979.

312

YAŞAR, AKSU, USLU

Natusch, D.F.S., Bauer, C.F., Matusiewicz, H.,Evans, C.A., Baker, J., Loh, A, Linton, R.W. andHopke, P.K., Characterization of trace elements infly ash. International Conference on Heavy Metals inthe Environment, Symposium Proceedings, October27-31, 1975, Volume II, Part 2, 553-575. Pathwaysand Cycling, Toronto, Canada.

Patterson, C.C., “Native copper, silver and gold ac-cessible to early metallurgists”. American Antiquity,36: 286-321, 1971.

Taymaz, K., Yiğit, V., Özbal, H., Geritoğlu, A. andMüftügil, N., “Heavy metal concentrations in wa-

ter sediment and fish from İzmit Bay”. Journal ofEnvironmental Analytical Chemistry, 16: 263-265,1983.

Turekian, K.K. and Wedepohl, K.H., “Distributionof elements in some major units of the Earth’scrust”. Geological Society of America Bulletin, 72:175-192, 1961.

TÇV, Environmental Issues in Turkey (TürkiyeÇevre Sorunları), Türkiye Çevre Vakfı Yayını, 131,464, 1988.

313