Stream Sediment Geochemical Exploration for Gold in theKazda¤ Dome in the Biga Peninsula, Western Turkey

HÜSEY‹N YILMAZ

Dokuz Eylül Üniversitesi, Mühendislik Fakültesi, Jeoloji Mühendisli¤i Bölümü, Buca, TR–35160 ‹zmir, Turkey(E-mail: huseyin.yilmaz@deu.edu.tr)

Abstract: The Tuztafl› Au-Ag mineralized system was discovered within the Kazda¤ dome using BLEG (bulk leachextractable gold) and 180-µm stream sediment geochemical data collected across the Biga Peninsula in westernTurkey. The deposit is located within the hinge of an antiform consisting mainly of high-grade metamorphic andmélange rocks that include the Alt›noluk Pb-Zn-Cu-Ag-Au and Evciler Fe-Au-Cu deposits on the southern and onthe northern flanks, respectively; no mineralization has been reported prior to this work which was first carriedout in 1996 between Alt›noluk and Evciler. The BLEG Ag/Au ratios at Tuztafl› (max 87) and Alt›noluk (max 43) arevery high to extremely high demonstrating the area to be relatively Ag-rich. This Ag enrichment is furtherdemonstrated by rock chip Ag/Au ratios reaching up to 43 at the Tuztafl› deposit. Silver is a moderate to veryeffective pathfinder for Au at Tuztafl›. Arsenic is useful, being more mobile than Sb. Arsenic also constitutes themost coherent and the most significant geochemical anomaly in the northern half of the AYALE (Ayvac›k-Alt›nova-Evciler) area which is underlain mainly by the mélange rocks intruded by granitoid. The BLEG samplingaccompanied by further follow-up 180-µm stream-sediment sampling is a powerful technique in detecting Au-Agdeposits or occurrences. BLEG appears to be a time- and cost-efficient stream sediment geochemical technique fordiscovering relatively large primary geochemical halos encompassing the precious (Au, Ag) and base metal (Cu, Pb,Zn) deposits.

Key Words: BLEG sampling, geochemistry, gold, mineral exploration, Kazda¤, western Turkey

Biga Yar›madas› Kazda¤ Domundaki (Bat› Anandolu)Dere Sediman› Jeokimyasal Alt›n Aramalar›

Özet: Tuztafl› mineralizasyon sistemi Biga Yar›madas›nda Kazda¤ domunda (Bat› Anadolu) BLEG (alt›n›n hacimselayr›flt›r›lmas›) ve 180-µm dere tortulu jeokimyasal verilerini kullanarak keflfedilmifltir. Yatak ço¤unlukla yüksekdereceli metamorfik ve melanj kayalar›ndan oluflan bir antiform s›rt›nda yer al›r ve s›ras›yla bu antiform güney vekuzey kanatlar›ndaki Alt›noluk Pb-Zn-Cu-Ag-Au ve Evciler Fe-Au-Cu yataklar›n› da kapsar; fakat ilk olarak 1966 dayap›lan bu çal›flma öncesinde literatürde Alt›noluk ve Evciler aras›ndaki bir cevherleflmenin (Tuztafl› cevherleflmesi)varl›¤›ndan söz edilmemifltir. Tuztafl›ndaki BLEG Ag/Au (maksiumum 87) ve Alt›noluk (maksiumum 43) oranlar›n›noldukça yüksek oluflu sahan›n göreceli Ag-zengini oldu¤una iflaret eder. Bu Ag zenginleflmesi 43’e kadar ulaflankayaç Ag/Au oranlar›yla da gösterilir. Gümüfl Tuztafl›ndaki Au’›n aranmas›nda oldukça etkin bir iz bulucudur.Antimondan daha hareketli olan As alt›n aramas›nda keza yararl›d›r. Arsenik granotoid sokulumlu melanjkayalar›nca altlanan çal›flma alan›n›n kuzey yar›s›ndaki en tutarl› ve en önemli jeokimyasal anomaliyi oluflturur. BLEGörneklemesini takiben yap›lan180-µm dere sediman› Au-Ag yatak ve zuhurlar›n›n bulunmas›nda güçlü bir tekniktir.BLEG, k›ymetli (Au-Ag) ve baz metal (Cu, Pb, Zn) yataklar›n› çevreleyen göreceli büyük ilksel jeokimyasal halelerinaranmas›nda zaman ve maliyet aç›s›ndan etkin bir dere tortulu jeokimyas› tekni¤i olarak gözükür.

Anahtar Sözcükler: BLEG örneklemesi, jeokimya, alt›n, maden arama, Kazda¤›, Bat› Anadolu

Turkish Journal of Earth Sciences (Turkish J. Earth Sci.), Vol. 16, 2007, pp. 33-55. Copyright ©TÜB‹TAK

33

Introduction

Systematic collection and analysis of drainage samples hasbeen established as method of mineral exploration at boththe reconnaissance and detailed scales in many parts ofthe earth (Ottesen & Theobald 1994). Obtainingmaximum efficiency from a geochemical explorationprogram necessitates a balance between minimizing the

density of sampling/maximizing the length of thedetectable dispersion trains and significantly reducing thecost/time requirements (Cohen et al. 2005). Even whencare is taken to obtain representative samples, goldexplorations can vary markedly as a result of varyinghydraulic processes in the stream and the differentresponses of high- to low-density minerals during

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

34

transport (Nichols et al. 1994). Furthermore, slightchanges in the slope of a stream bed, coupled with theparticle scarcity effects on sample representativity maylead to the inhomogeneous Au distribution. Therefore,BLEG stream sediment sampling is considered as analternative technique to that of 180-µm stream sedimenttechnique to overcome the problem of erratic golddistribution. BLEG appears to be the best way to doreconnaissance-scale sampling for Au.

Western Turkey has been a focus of exploration forgold deposits in the last two decades. However, althoughgeochemical surveys form a major part of mostcompanies’ strategies, little information is available on thegeochemical dispersion of elements from deposits andprospects in western Turkey by which survey design maybe optimized. The extraction of gold from large samplesis a common approach that improves samplerepresentativity and reduces detection limits; such as thebulk leach extractable gold (BLEG) method, described byElliot & Towsey (1989), Radford (1996) and Y›lmaz(2003).

The study area selected for investigating thedispersion characteristics of various elements is located50 km SE of Çanakkale and is accessible by theÇanakkale-Edremit highway (Figure 1). This area iscontained within the Ayvac›k-Alt›nova-Evciler (AYALE)structural dome where extensional features are cut byseveral NE-trending intrusions. The Kazda¤ dome hostsnumerous mineral occurrences and deposits (Figure 1).Three of these deposits, representing a range of deposittypes and differing levels of emplacement, have beenincluded in the study area. The first, the Alt›noluk(Papazl›k) Pb-Zn-Cu-Ag-Au deeper-polymetallicepithermal (?) vein deposit with a resource of 240 K tonsgrading 5 g/t Au and 25 g/t Ag, is located in the southernpart of the area and has been intermittently exploitedduring the last century (MTA– Mineral Research andExploration Institute of Turkey 1966, 1993). The seconddeposit, re-discovered during this study, is the Tuztafl›epithermal Au-Ag deposit where numerous small ancientworkings were recognized. The third deposit is theEvciler (Ayazma) Fe-Au-Cu proximal-skarn deposit in thenorthern margin of the Kazda¤ structural mountainrange, which was exploited in ancient times (Y›lmaz &Kara 1996). The area between the Evciler and Alt›nolukareas was not known to contain gold and silvermineralization until this study (Y›lmaz & Kara 1996).

K›l›ç et al. (2004) reported that Au values ranging upto 14 ppm in soil and 3 ppm in silicified dacitic toandesitic volcanic lava dome rocks aroundK›rantepe/K›sac›k village 13 km east of Ayvac›k. Aydal etal. (2004) also noted the presence of gold anomalies inthe silicified zones of ultramafic rocks at Alakeçili.

This study evaluates the effectiveness of the BLEG(bulk cyanide leach extractable gold) technique and aquaregia-digested metal contents of the 180-µm (-80 #)fraction of stream sediments as well as the 180-µm (-80#) fraction of soil at AYALE, particularly in the Tuztafl›area, which is underlain by metamorphic rocks andmélange. The purpose of this study is to demonstrate theefficiency of the BLEG technique in exploring for gold asa case study rather than investigating the detailed geneticrelationships among these three deposits with probabledifferent levels of emplacement.

Regional Geology at AYALE

The AYALE is located in the southeastern part of theKazda¤ Mountain range (Figure 1), which forms astructural and topographic dome of high-grademetamorphic rocks (Schuling 1959; Bingöl 1969; Okayet al. 1991, 1996). The Kazda¤ range trends NE–SWand, rising to 1767 m above sea-level, forms atopographic anomaly in the northeastern Aegean, wherethe average elevation is below 500 m. The Kazda¤ Groupforms a doubly plunging, NE–SW-trending anticliniorium(Duru et al. 2004). The Ezine Group, over three-km thickand containing systematic occurrences of carbonate rocksin the greenschist facies, represents a fragment of theRhodopian passive margin, a consequence of Permo–Triassic rifting of the future Maliac/Meliata Ocean, alsoobserved in Greece. The emplacement of the Denizgörenophiolite over the Ezine Group occurred during theBalkanic orogeny, a major compressional event, whichaffected the whole Rhodope area, and was characterizedby northward nappe emplacement during Jurassic–EarlyCretaceous times (Beccaletto & Jenny 2004).

Basement rocks in the Tuztafl› area consist ofPalaeozoic gneiss, marble and amphibolite (Okay & Sat›r2000), tectonically overlain by a highly deformed, LateCretaceous to Palaeocene oceanic accretionary mélangecomprising basalt, clastic sediments, limestone andserpentinites (Figure 2); the sequences are separated bythe Alakeçili-Akp›nar shear zone. The Kazda¤ Massif, theshear zone, and the accretionary mélange are intruded by

H. YILMAZ

35

GÖ

KÇ

EA

DA

MA

RM

AR

AS

EA

LA

PS

EK

İ

BİG

A

YE

NİC

E

İVR

İND

İ

BU

RH

AN

İYE

AY

VA

LIK

BE

RG

AM

A

Ova

cık

Gol

dD

epos

it

KO

ZA

K

SO

MA

BA

YR

AM

İÇ

LE

SV

OS

ED

RE

MİT

KA

RA

CA

BE

Y

GÖ

NE

N

MA

NY

AS

MA

NY

AS

LA

KE

SU

SU

RL

UK

BA

LIK

ES

İR

ÇA

NA

KK

AL

E

EC

EA

BA

T

Pb

Pb-

Zn

Cu,

Zn

Au

Pb-

Zn

Au-

Ag

Cu-

Au

Pb-

Zn-

Ag

Au

Kuş

çayı

rı

4500

00

4470

000

5000

00

Pli

ocen

evo

lcan

ics

rock

Mio

cene

volc

anic

sro

ck

Eoc

ene

volc

anic

sro

cks

Ter

tiar

yin

trus

ives

Cre

tace

ous

intr

usiv

es

non-

mag

mat

icro

cks

tow

n

stud

yar

ea

prec

ious

min

eral

occu

rren

ce

prec

ious

min

eral

depo

sit

othe

rm

etal

occu

rren

ce/d

epos

its

Upp

erO

ligo

cene

volc

anic

sro

ck

AE

GE

AN

SE

A

05

1020

30km

.

N

Pb-

Fe

Kir

azlı

Akb

aba

Au

Mag

+Mo+

py+c

py

Mag

+py

+cp

y

Pb-

Zn-

Au

Cu-

Pb-

Zn

Pb-

Zn

Mn

Sb

Fe

Pb-

Zn-

Cu

++

+++ ++

+S

tudy

Are

a

EV

CİL

ERA

u

Tuz

taşı

Çan

akka

le

Ova

cık

Kaz

dağ

Rang

e

(AY

AL

E)

Are

a

Stu

dyA

rea

Au-A

gS

bS

b

BA

LYA

Pb-

Zn-

Sb-

Au

Au-

Ag,

(Pb-

Zn)

Fe-

Cu

Fe

Au-

Ag

Au-

Ag

BA

u-A

g

KA

PID

AĞ

IP

EN

INS

UL

A

Au

Au,

Cu

Au,

Cu

Au

Efe

mçu

kuru

4370

000

İzm

irK

ışla

dağ

DİK

İLİ

GE

LİB

OL

UŞ

EV

KE

TİY

EK

AR

AB

İGA

ER

DE

KB

AN

DIR

MA KE

PS

UT

BİG

AD

İÇ

SIN

DIR

GI

ALT

INO

LU

K

GÜ

LP

INA

R

AY

VA

CIK

EZ

İNE

KÜ

ÇÜ

KD

ER

E

Big

aP

enin

sula

Figu

re 1

.Si

mpl

ified

geo

logi

cal m

ap o

f th

e Te

rtia

ry v

olca

nic

and

plut

onic

fie

ld o

f no

rthw

est

Turk

ey a

nd lo

catio

n of

the

stu

dy a

rea

and

sign

ifica

nt m

iner

aliz

atio

ns (

volc

anic

and

plu

toni

c fie

lds

are

mod

ified

fro

m E

rcan

et

al.1

984,

199

5; E

rkül

et

al.2

005;

Tok

çaer

et

al.2

005;

Ald

anm

az 2

006)

.

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

36

Alakeçi

Kırçalar

Kısacık

Kızılyar

Tuztaşı

Akpınar

0 2 4 km

N

Pliocene sandstone,conglomerate, limestone

Miocene sandstone,tuff, andesite

Lower Miocene granodiorite

Çetmi Melange (basalt, greywacke,limestone, serpentine)

metaserpentinite

gneiss, marble, amphibolite

Triassic limestoneCretaceous eclogite

Alakeçili shear zone

ultramylonite

Kazdağ Group

strike-slip fault

detachment fault

hydrothermal breccia

stratigraphic contact

quartz veins with Au

town

Serhat

Dalaloba

Akpınar

BaharlarTuztaşı

Kırçalar

Çaltı

KızılyarÇetmibaşı

TuztaşıProspect

+ +

V VV

normal fault

Alak

eçili

shea

r zon

e

Figure 2. Geology (Okay & Sat›r 2000) and mineralization map of Tuztafl› area (this study).

Late Oligocene/Early Miocene granitoids (23.8 Ma:Delaloye & Bingöl 2000). The Alakeçili-Akp›nar shearzone of strongly mylonitized gneiss and serpentinite, 2-km thick, occurs between the Kazda¤ Massif and theaccretionary mélange in the north (Okay et al. 1991). TheAkp›nar fault is interpreted as a low-angle detachmentwhich juxtaposes brittly deformed upper crustal rocksover the ductilely-deformed mid-crustal rocks (Okay &Sat›r 2000). The Evciler granitoid is an elliptical, calc-alkaline pluton situated north of the Kazda¤ range(Figures 1 & 2). The granitoid extends northeast–southwest parallel to the trend of the Kazda¤ dome andthe Akp›nar shear zone. Its mineralogical compositionranges from monzodiorite through quartz diorite togranodiorite and the latter is the predominant faciesconstituting over 70% of the pluton (Öngen 1978, 1994;Genç 1998; Yücel-Öztürk et al. 2005). In the north thegranitoid has intruded the Late Oligocene–Middle Mioceneandesites, dacites and intercalated sedimentary rocks.These volcanic rocks are geochemically similar to theEvciler Pluton and are regarded as its extrusiveequivalents (Genç 1998). Thermal metamorphismdeveloped at the contacts between the Late OligoceneEvciler Granodiorite (Yücel-Öztürk 2005) and marbleand/or metamorphic rocks hosting gold mineralization(Y›lmaz & Kara 1996).

The Evciler, and several other I-type, medium toshallow-seated intrusive rocks of granite/granodiorite tomonzodiorite compositions were emplaced into basementof the Kazda¤ Group. Early Miocene volcanism within theEvciler area is characterized mainly by andesite and dacitelavas, and associated pyroclastic rocks (Genç 1998)whereas dacitic lava domes, lava flows and volcanoclasticsare dominant facies in the Alt›noluk area. The Evciler(Ayazma) Fe-Cu-Au, the Tuztafl› Au-Ag and Alt›noluk(Papazl›k) Pb-Zn-Cu-Ag-Au mineralized systems (Figure1) occur within Kazda¤ mountain range which is regardedas an extensional metamorphic core complex of Oligoceneage, consisting of gneiss, marble and amphibolite at itsfootwall, and Early Tertiary accretionary mélange withexotic Upper Cretaceous eclogite blocks in its hangingwall (Okay & Sat›r 2000; Duman et al. 2004; Beccaletto& Jenny 2004; Beccaletto & Steiner 2005).

Local Geology and Mineralization at AYALE

Alt›noluk-Papazl›k Prospect

Information on the geology and mineralization ofAlt›noluk (Papazl›k) base and precious metal deposit

(Figure 1) is almost nil except for MTA inventories (MTA1966, 1993). The Alt›noluk vein-type Pb-Zn-Ag-Aumineralization is hosted by a sheared marble layerintercalated with amphibolite and gneiss of the Kazda¤metamorphic rocks. The thickness of the veins reaches1.6 m over a strike length of several hundred meters.The ore consists predominantly of coarse crystallinegalena, sphalerite, and pyrite with minor chalcopyrite andcovellite. The Alt›noluk mineralization forms a small-tonnage ore body (240 K tons) with grades of 8.2% Zn,6.7% Pb, 5 ppm Au and 25 ppm Ag (MTA 1966, 1993).

Evciler-Ayazma Prospect

The Au concentration reported first at Evciler isassociated primarily with massive pyrite-marcasite-pyrrhotite-chalcopyrite-quartz-calcite mineralizationwithin marble and biotite amphibole gneiss intruded bythe Evciler Pluton (Y›lmaz & Kara 1996). The Evciler(Ayazma) Fe-Cu-Au mineralization occurs mainly asexoskarn consisting predominantly of prograde garnetand diopside with minor retrograde tremolite, epidote,chlorite, scapolite and sericite. Mineralization appears tobe strongly controlled by the intercalated amphibolegneiss and marble contact which is cut by a porphyriticgranodiorite sill of the Evciler pluton. A central core ofmineralization with a 200 m strike length containinghigh-grade gold concentrations has intermittentlyoutcropping, N50°E-trending lensoidal extensionstowards the east (250 m) and west (500 m). The Au- andassociated pyrite-pyrrhotite mineralization appears to becontrolled by a NE-trending structure, dipping at 50°NW(Y›lmaz & Kara 1996). Best rock chip results to datereturned up to 14 ppm Au at 4 m with several othervalues above 4 ppm.

Based on earlier Au-exploration work (Y›lmaz & Kara1966), a detailed study of the Evciler skarn alterationwas carried out by Yücel-Öztürk et al. (2005). Theysuggested that the geochemical characteristics of theÇavufllu monzodiorite, Karaköy granodiorite andmesocratic Evciler rocks were similar to averages for Au-Cu and Fe-skarn granitoids, whereas the geochemicalcharacteristics of the leucocratic Evciler rocks weresimilar to averages for Sn- and Mo-skarn granitoids. TheEvciler granitoid is also characterized by relatively un-evolved to moderately evolved and oxidized suites, as inmost Au-Cu core metal associations globally (Yücel-Öztürk et al. 2005). They concluded that ‘composition

H. YILMAZ

37

and petrologic evolution of Evciler pluton have had theprimary controls on skarn alteration, mineralization andmetal content such as Cu, Au and Fe’, despite the fact thatthere was either no analysis carried out for Cu and Au orno reference cited on Cu and Au mineralization in theirstudy.

Tuztafl› Prospect

The Tuztafl› prospect and its surrounding area areunderlain mainly by Kazda¤ metamorphic rocks includinggneiss, marble and amphibolite, and Çetmi mélangeconsisting of basalt, greywacke, siltstone, chert,limestone, serpentinite and eclogites, which are cut by theLate Oligocene/Early Miocene Evciler granitoid andandesites (23.8 Ma; Delaloye & Bingöl 2000). TheTuztafl› prospect, underlain by fine- to medium-grainedgneiss and the Evciler pluton, contains threeapproximately N50°E-trending and 20° to 40°N-dippingquartz veins (Figure 3). Several additional Au-bearingquartz veins are recorded over a strike length of 8 kmtoward the southwest. However, east-southeast ofK›rcalar the trends of the quartz veins vary betweenN70°E and N60°W (Figure 2). The drusy crystalline tosaccharoidal quartz vein/breccia with vug-infill, comb andminor crustiform textures at the Tuztafl› prospectcontains native Au and Ag, pyrite and arsenopyrite. Thequartz veins, encompassed by a strong to weak argillic-silicic alteration halo of 600 m by 2500 m, have amapped strike length of 1200 m, an average width ofabout 6 m, and reach a maximum width of 12 m. Theargillic alteration within or around the quartz veins isrepresented by illite/sericite. Preliminary fluid inclusionanalysis (40 Th °C measurements) of two samplesindicates mineralization temperatures of 213 °C and 233°C with salinities of 0.7 and 1.7 wt% NaCl equivalent.Fluid inclusions are typically irregular in shape and rangein size from 10 to 100 µm and averaging 40–50 µm.Quartz textures, clay mineralogy and limited fluidinclusion results may indicate an epithermal style of Au-Ag mineralization. Ancient workings associated withearthy slag piles were also located along the quartz veinin the prospect area.

Exploration Geochemistry

A metamorphic core complex within the Kazda¤metamorphic dome stretches from Ayvac›k to Evciler and

is intruded by several NE-trending granitoids. This areawas selected for the first orientation surveys on the useof exploration geochemistry including BLEG, 180-µm (-80#) stream sediment and 180-µm soil geochemicalsampling for discovering precious occurrences at AYALEin Turkey. The two deposits are 20 km apart and noprecious metal mineralization between the deposits hasbeen reported prior to the BLEG geochemical sampling bythis study.

Climate, Topography and Regolith

AYALE has a semiarid-type climate with dry summers andcold, wet winters. July and August are the hottest monthswith average temperatures around 30 °C whereasJanuary and February are the coldest with temperaturesaround -5 °C. Annual rainfall is 700 mm. The landscapeat AYALE is dominated by ENE–WSW-trending first-orderriver valleys which are the surface expression of the~E–W-trending grabens. Mountains, particularly in theTuztafl› prospect area reach 800 m. Further north andsouth the topography is smoother and gentler with amaximum altitude of 250 m around the Evciler andAlt›noluk areas. Major perennial rivers, such as BayramiçÇay›, begin in narrow valleys and further west flow withina major graben; however, the majority of the streams areephemeral. Hills in the AYALE are covered mainly by pineforest. Cultivation is restricted to olive groves on thesouth and fruit on the north because of lack of largevalleys.

Weathering of the metamorphics and granites isgenerally shallow, with deeper weathering along faultsand other structures. Neutral-pH grassland soils consistof a dark A-horizon overlying lighter-coloured parentmaterial, which in turn underlain by B-horizon soils,include concretions and earthy accumulations of calciumcarbonate. The soil material transported into the drainagesystem consists mainly of illite, quartz and metamorphicdetritals with considerable organic substances and minorFe-Mn oxides.

Sampling and Analysis

Sample points were selected from published 1:100,000scale topographic maps to achieve a sampling density of 1sample per 6–7 km2 (Figure 3). A two-man teamconsisting of one geologist and one sampler collected anaverage of 15 samples per day. Active stream sediment

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

38

H. YILMAZ

39

020

0m

weak

arg

illic

altera

tion

epid

ote

-chlo

rite

altera

tion

quart

zvein

bre

ccia

/banded

quart

zvein

sheare

dE

vcile

rP

luto

n

Evcile

rP

luto

n

un

sp

ecifie

dfa

ult

no

rma

lfa

ult

mediu

m-g

rain

ed

gneis

s4

70

00

0

470

00

0

43

94

000

43

94

000

str

ong

arg

illic

altera

tion

Figu

re 3

.G

eolo

gy,

alte

ratio

n an

d m

iner

aliz

atio

n m

ap o

f Tu

ztafl

› pro

spec

t.

was collected along some 30 to 50 m of the stream andsieved on site to provide 2 kg of

Regional BLEG and 180-µm Survey in AYALE Area

Reconnaissance BLEG sampling carried out in the AYALEarea yielded anomalous values ranging from 0.6 to 15.2ppb Au and 0.1 to 190 ppb Ag (Figure 4). Table 1contains the summary statistics for three variables.Skewness coefficients for raw data indicate that alldistributions are positively skewed. Although BLEG Au atthe 95th percentile level appears to represent the veryanomalous values and is more or less similar to those ofAu-Ag threshold1+2 values (Table 1), log-transformedBLEG Au and Ag threshold1+2 values are slightly higher forAu and 2.6 times lower for Ag. Therefore, the decisionwas made to work with log10-transformed values of thedata instead of the raw values, which reduces theasymmetry of the distribution as indicated by theskewness coefficient (log). The log-transformation hashelped the reduce skewness in all cases but a very smallpositive skewness usually remains.

The BLEG anomalies were followed up using 180-µmstream sediments. These samples yielded weak to verystrong Au, Cu, Pb, Zn, As, and Sb anomalies (Table 2,Figure 5) with peak values of 285 ppb Au, 14 ppm Ag,130 ppm Cu, 1280 ppm Pb, 1250 ppm Zn, 253 ppm Asand 121 ppm Sb. In twenty four of the 585 180-µmstream sediment samples Au, Ag, Cu, Pb, Zn, As and Sbvalues are greater than 23 ppb and, 5, 70, 54, 133, 65and 18 ppm, respectively, for the 95th percentile. Thesevalues are indicative of moderately anomalous metalcontents. The threshold1+2 values derived from the rawdata (Table 2) are 2 to 6 times higher than those derivedfrom the log-transformed data. As indicated by the rowdata, all distributions are slightly to moderately positivelyskewed whereas distributions appear to be symmetric atlog-transformed data, although a small positive ornegative residual skewness usually remains. Results ofthe180-µm stream sediment sampling indicate threecoherent clusters of anomalous Au-Sb-As-Cu values(Figure 5). The clusters of major Au and Sb anomaliesexhibit similar dispersion patterns and appear to beperipheral to the cluster of extensive As anomalies. Rock-chip sampling from quartz veins with minor sulfides,particularly in the Tuztafl› prospect area, shows Au valuesup to 5.4 ppm, with 42 samples greater than 0.5 ppm Auand 30 samples greater than 1 ppm Au (Table 3, Figure6). Negligible Au is detected within samples containingore grade Cu, Pb and Zn. This may indicate that Au issituated within quartz rather than sulfide minerals. An

association of Ag with Au in quartz veins is demonstratedby relatively high values (up to 130 ppm Ag) within As-rich zones. However, no correlations occur between Au-Ag and As (Table 5). Antimony is generally weakly andlocally (south of Evciler) strongly anomalous and has asimilar dispersion pattern to that of Au in 180-µm streamsediment samples at this location. Stream samples exhibitcoherent and very strong As anomalies, overlapping thoseof Au. Lead and Zn anomalies are generally weak but verystrong in two isolated locations. They are overlapped bythe As dispersion patterns.

The 180-µm soils at the Tuztafl› Prospect weresampled over an area of 1 km2 at a line spacing of 100 mwith samples at 25 m along lines (Figure 7). The 180-µmsoils contained values ranging from 25 ppb Au. This correlates well with As. To the east, anarea 400-m long and 150-m wide is defined by >25 ppbAu. Another smaller soil gold anomaly to the southeast,occurring between XL2 and XL12, has dimensions ofabout 50–150 m x 1000 m. The first and the second soilAu anomalies overlap with the As anomaly. A moderate tostrong soil As anomaly in an area 50-200 m by 1800 m(between XL1 and XL17) returned As values generally

H. YILMAZ

41

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

42

Figure 4. BLEG sample location and Au, Ag and Cu assay results from Ayvac›k-Alt›nova-Evciler (AYALE) area.

Table 1. Statistical values for the 1.18 mm BLEG geochemical data set of the AYALE area.

Elements Au Ag Cu Au Ag Cu

Data Raw Log-transformed

Minimum 0.1

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

44

Tabl

e 2.

Stat

istic

al v

alue

s fo

r th

e 18

0-µm

str

eam

sed

imen

t ge

oche

mic

al d

ata

set

of t

he A

YALE

are

a.

Elem

ents

AuAg

CuPb

ZnAs

SbAu

AgCu

PbZn

AsSb

Dat

aR

awLo

g10-

trra

nsfo

rmed

Min

imum

<1

<0,

5<

1<

1<

9<

1<

10.

0-0

.30

0.00

0.00

0.95

0.0

0.0

Max

imum

280

14.0

013

012

8012

5025

312

12.

41.

142.

113.

113.

102.

42.

1

Mea

n5

1.26

2526

7721

100.

3-0

.30

1.16

1.22

1.84

1.1

0.6

Med

ian

10.

5021

2169

138

0.0

-0.3

01.

321.

321.

841.

10.

7

Mod

e1

0.50

11

491

10.

0-0

.30

0.00

0.00

1.69

0.8

0.5

S.D

161.

8622

5660

2510

0.5

0.40

0.57

0.44

0.19

0.5

0.5

Q1

10.

5010

1454

61

0.0

-0.3

01.

001.

151.

730.

80.

0

Q3

20.

5035

2888

289

0.3

-0.3

01.

551.

451.

951.

41.

0

95th

perc

entil

e23

5.00

7054

133

6518

1.4

-0.3

01.

851.

732.

121.

81.

3

Skew

ness

103.

001

1913

37

1.8

0.70

1.00

1.18

0.48

0.5

-0.1

Kur

tosi

s15

410

.00

342

024

820

932.

43.

800.

032.

534.

750.

1-1

.3

N58

558

558

558

558

558

558

558

558

558

558

558

558

558

5

BG1

51.

2625

2677

2110

2.0

0.90

1416

6913

4

Thre

shol

d 137

5.00

6913

819

771

3010

4.30

2222

7119

10

BG2

10.

5021

2169

138

0.0

0.90

2020

6813

5

Thre

shol

d 233

3.72

6513

318

963

286.

04.

3028

2672

198

C3.

21.

50.

92.

10.

825

.21.

02.

02.

800.

30.

20.

00.

30.

8

Ag,

Cu,

Pb,

Zn,

As a

nd S

b da

ta in

ppm

; da

ta f

or A

u in

ppb

, SD

– st

anda

rd d

evia

tion,

C–

coef

ficie

nt o

f va

riat

ion,

Q1

and

Q3–

fir

st a

nd t

hird

qua

rtile

s, N

– nu

mbe

r of

sam

ples

. M

ean:

Bac

kgro

und

(BG

1),

Thre

shol

d 1:

BG1

+ 2

sta

ndar

d de

viat

on.

Med

ian:

Bac

kgro

und

(BG

2),

Thre

shol

d 2:

BG2

+ 2

sta

ndar

d de

viat

ion.

Det

ectio

n lim

its:

Au:

1 pp

b, A

g: 0

.5 p

pm,

Cu:

1 pp

m,

Pb:

1ppm

, Zn

:

1ppm

, As

: 1

ppm

, Sb

: 1p

pm

H. YILMAZ

45

Figure 5. Stream sediment geochemistry of Au, Cu, Pb, Zn, As, Sb at the AYALE area.

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

46

Tabl

e 3.

St

atis

tical

val

ues

for

the

rock

chip

geo

chem

ical

dat

a se

t of

the

AYA

LE a

rea.

Elem

ents

AuAg

CuPb

ZnAs

SbM

oW

BiTl

AuAg

CuPb

ZnAs

SbM

oW

BiTl

Dat

aR

awLo

g10-

trra

nsfo

rmed

Min

imum

<1

<0,

5<

1<

1<

1<

1<

1<

1<

1<

1<

10.

00-0

.30

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Max

imum

5400

130.

0023

000

5700

046

000

1120

027

918

737

303

293.

732.

104.

364.

764.

664.

053.

451.

663.

422.

481.

46

Mea

n31

57.

0039

786

373

732

550

214

165

1.19

0.06

1.6

1.18

1.47

1.87

0.9

0.25

0.37

0.55

0.46

Med

ian

120.

5020

1121

808

11

11

1.08

-0.3

01.

301.

041.

321.

900.

900.

000.

000.

000.

00

Mod

e1

0.50

171

11

11

11

10.

00-0

.30

1.23

0.06

1.04

0.00

0.00

0.00

0.00

0.00

0.00

S.D

754

18.0

019

2757

8950

2988

825

82

8139

61.

190.

670.

720.

910.

710.

800.

760.

400.

640.

600.

41

Q1

10.

5013

612

231

11

11

0.00

-0.3

01.

110.

741.

081.

360.

001.

000.

000.

000.

00

Q3

115

1.00

3940

4428

525

45

113

2.06

0.01

1.60

1.60

1.65

2.45

1.39

0.55

0.69

1.03

0.42

95th

perc

entil

e19

1539

.00

1608

482

1076

1834

124

832

3014

3.28

1.59

3.21

2.68

3.03

3.26

2.09

0.90

1.51

1.47

1.13

Skew

ness

34.

008

88

910

58

72

0.50

1.63

1.64

1.19

1.97

-0.2

20.

461.

432.

321.

011.

40

Kur

tosi

s14

17.0

084

6662

9610

627

6955

18-1

.40

1.23

3.74

2.63

6.12

-0.0

8-0

.31

1.26

6.48

0.02

0.68

N24

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

024

0

BG1

315

7.00

397

863

737

325

502

1416

516

.00.

061.

601.

181.

471.

870.

890.

250.

370.

550.

46

Thre

shol

d 118

2343

.00

4251

1244

110

792

2101

566

617

694

1748

.011

.00

53.0

32.0

41.0

91.0

20.0

8.0

13.0

11.0

9.0

BG2

120.

5020

1121

808

11

11

12.0

0.0

20.0

11.0

21.0

80.0

8.0

1.0

1.0

1.0

1.0

Thre

shol

d 215

2037

.00

3874

1158

910

079

1856

596

516

379

1344

.010

.030

.032

.041

.091

.020

.07.

011

.09.

07.

0

C2.

42.

605

77

35

15.

72.

41

1.0

4.5

0.2

0.2

0.2

0.0

0.9

1.5

1.5

1.0

1.0

Ag,

Cu,

Pb,

Zn,

As,

Sb,

Mo,

Cd,

W,

Bi a

nd S

dat

a in

ppm

; da

ta f

or A

u in

ppb

. SD

– st

anda

rd d

evia

tion,

C–

coef

ficie

nt o

f va

riat

ion,

Q1

and

Q3:

fir

st a

nd t

hird

qua

rtile

sN

– nu

mbe

r of

sam

ples

. M

ean–

bac

kgro

und

(BG

1),

Thre

shol

d 1–

BG1

+ 2

sta

ndar

d

devi

atio

n. M

edia

n: b

ackg

roun

d (B

G1)

, Th

resh

old1

: BG

1+ 2

sta

ndar

d de

viat

ion.

Det

ectio

n lim

its:

Au:

1 pp

b, A

g: 0

,5 p

pm,

Cu.

1ppm

, Pb

: 1

ppm

Zn:

1 p

pm,

As:

1ppm

, Sb

: 1

ppm

, M

o: 1

ppm

, W

: 1p

pm,

Bi:

1ppm

.

H. YILMAZ

47

Figure 6. Rock chip geochemistry of Au, Ag, Cu, As and Sb at the Tuztafl› area.

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

48

Tabl

e 4.

Stat

istic

al v

alue

s fo

r th

e 18

0-µm

soi

l geo

chem

ical

dat

a se

t of

the

AYA

LE a

rea.

Elem

ents

AuAg

CuPb

ZnAs

SbM

oW

BiS

AuAg

CuPb

ZnAs

SbM

oW

BiS

Dat

aR

awLo

g10-

trra

nsfo

rmed

Min

imum

<1

<0,

5<

1<

1<

1<

1<

1<

1<

1<

117

0.00

-0.3

00.

000.

001.

040.

000.

000.

000.

000.

001.

20

Max

imum

2100

2043

594

941

852

120

2910

3413

1617

3.32

1.31

2.64

2.98

2.62

2.72

3.31

1.01

1.53

1.11

3.21

Mea

n42

0.60

5021

9427

203

32

188

0.62

-0.2

91.

621.

151.

921.

130.

940.

390.

210.

142.

21

Med

ian

10.

5041

1783

1511

21

116

00.

00-0

.30

1.61

1.22

1.92

1.18

1.04

0.37

0.00

0.00

2.20

Mod

e1

0.50

371

701

11

11

172

0.00

-0.3

01.

560.

002.

070.

000.

000.

000.

000.

000.

00

S.D

166

1.00

4044

4744

108

0.4

41

133

0.82

0.12

0.26

0.38

0.21

0.54

0.47

0.37

0.37

0.30

0.23

Q1

10.

5030

1065

95

11

111

40.

00-0

.03

1.48

0.99

1.81

0.93

1.00

0.00

0.00

0.00

2.06

Q3

160.

5058

2411

625

186

21

223

1.20

-0.3

01.

761.

372.

061.

392.

390.

770.

380.

002.

35

95th

perc

entil

e17

80.

5010

341

184

104

298

107

369

2.25

-0.3

02.

011.

612.

262.

021.

470.

891.

000.

842.

57

Skew

ness

815

518

25

170.

53.

22.

44.

51.

0911

.50

-0.2

2-0

.85

-0.1

4-0

.58

-0.2

20.

131.

461.

760.

38

Kur

tosi

s75

234.

0035

380

7.2

4328

5-1

.215

5.3

330.

1413

74.

002.

730.

721.

201.

75-1

.77

0.70

1.42

1.62

N53

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

353

3

BG1

420.

6050

2194

2720

33

218

84.

00.

043

.015

.082

.012

.09.

03.

01.

01.

416

5.0

Thre

hold

137

42.

6013

010

918

811

523

63.

811

445

416

.00.

045

.021

.084

.018

.015

.09.

07.

05.

416

7.0

BG2

10.

5041

1783

1511

21

116

01.

000.

543

.017

.083

.016

.09.

02.

01.

01.

016

5.0

Thre

shol

d 233

72.

5012

110

517

710

322

72.

89

342

613

.02.

045

.023

.085

.022

.017

.08.

07.

05.

02.

7

C3.

91.

700.

82.

10.

51.

65.

40.

11.

30.

50.

71.

500.

000.

000.

200.

000.

300.

201.

001.

502.

000.

00

Ag,

Cu,

Pb,

Zn,

As,

Sb,

Mo,

Cd,

W,

Bi a

nd S

dat

a in

ppm

; da

ta f

or A

u in

ppb

. SD

– st

anda

rd d

evia

tion,

C–

coef

ficie

nt o

f va

riat

ion,

Q1

and

Q3:

fir

st a

nd t

hird

qua

rtile

s. N

– nu

mbe

r of

sam

ples

.M

ean–

bac

kgro

und

(BG

1),

Thre

shol

d 1–

BG1

+ 2

sta

ndar

d de

via-

tion.

Med

ian–

bac

kgro

und

(BG

2),

Thre

shol

d 1–

BG2

+ 2

sta

ndar

d de

viat

ion.

Det

ectio

n lim

its:

Au:

1 pp

b, A

g: 0

,5 p

pm,

Cu.

1ppm

, Pb

: 1

ppm

Zn:

1 p

pm,

As:

1ppm

, Sb

: 1

ppm

, M

o: 1

ppm

, W

: 1p

pm,

Bi:

1ppm

.

H. YILMAZ

49

1010

010

00

110100

200

Zn

Cu

(a)

R=

21

n=58

5

100

600

110100

1000

Zn

Pb

(e)

R=

0.8

n=53

3

1010

010

00

110100

1000

Zn

Pb

(b)

R=

0.3

2

n=58

5

110

100

700

110100

1000

4000

As

Au

(c)

R=

0.4

0

n=53

3

1010

080

0

110100

800

Zn

Cu

(d)

R=

0.5

n=53

3

Zn

110

100

1000

1000

070

000

110100

1000

1000

0

5000

0

Cu

(f)

R=

0.4

5

n=24

0

110

100

1000

1000

070

000

110100

1000

1000

0

8000

0

Zn

Pb

(g)

R=

1.0

n=24

0

110

100

1000

3000

0

110100

1000

8000

As

Au

(h)

R=

0.2

5

n=24

0

110

100

200

110100

1000

8000

Ag

Au

(i)

R=

0.2

8

n=24

0

Figu

re 7

.Lo

g-Lo

g co

ncen

trat

ions

of:

(a)

Cu v

ersu

s Zn

and

(b)

Pb v

ersu

s Zn

in 1

80-µ

m s

trea

m s

edim

ent;

(c)

Au v

ersu

s As

, (d)

Cu v

ersu

s Zn

and

(e)

Pb v

ersu

sZn

in 1

80-µ

m s

oil;

(f)

Cu v

ersu

s Zn

, (g

)Pb

ver

sus

Zn,

(h)

Au v

ersu

s As

and

(i)

Au v

ersu

s Ag

in r

ock

chip

at

the

AYAL

E ar

ea.

Discussion of Results

Arsenic exhibits a nearly universal enrichment in mosttypes of hypogene gold deposits (Boyle 1979). Gold-bearing polymetallic deposits and certain skarns areinvariably enriched in arsenic, as are the typical quartzveins at Fortitude Mine in Nevada, USA. In these types ofdeposit the arsenic content ranges from a few tens ofparts per million to percentage amounts in the ore.Because the gold is so intimately associated witharsenopyrite- and arsenic-bearing sulfosalts in somedeposits the element appears to be a lattice constituent ofthese minerals. The disseminated gold deposits of northcentral Nevada are all enriched in arsenic (Berger &Bethke 1984). However, the element is not only bound inAs minerals and As-bearing pyrite, which are commonlyassociated with Au mineralization (Yang & Blum 1999),but also forms widespread geochemical haloes around Audeposits, so that the element is commonly used as apathfinder in noble metal exploration. In thenorthwestern part of Hunan Province (China) manyAu–Sb (or Au–W) deposits occur in low-grademetamorphic rocks of Neoproterozoic age. Here,arsenopyrite is the only As-bearing mineral, occurring inminor quantities. Arsenic concentrations in these slatebelt deposits range from 0.10 to 2.36 wt% (mainlybetween 0.4 and 1.0%). However, As has never beenconsidered as pathfinder for Au-exploration in the earlierwork (Yang & Blum 1999). A similar geochemicalsignature to that of northwestern Hunan is recognized inlow-grade metamorphic rocks of western Turkey (Y›lmaz2003). Mesothermal gold mineralization closelyassociated with As is widespread in Upper Mesozoicmetagreywackes deposited in a collisional setting in theSouthern Alps of New Zealand (Craw 2001). Arsenic ispresent in solid solution in many of the sulfides, and inaccessory As minerals in mesothermal deposits (Craw2001). It is readily mobilized from these deposits atneutral to alkaline pH. In contrast to As behavior, Cu, Pb,Zn and Cd are dissolved and passed into soil under highlyacidic conditions. Arsenic is most stable in an oxidizingenvironment, being a constituent of basic sulfatearsenates. Oxidation of native As, arsenopyrite andvarious other arsenides and sulfarsenides yields a varietyof supergene arsenates and basic arsenate-sulfates, themost common being scorodite, Fe (AsO4).H2O, and Co-,Ni-, Pb-, Zn-, and Cu-arsenates (Boyle 1979). Of these,the most important in gold deposits is scorodite; the

others occur only where there are high contents of Ni,Co, Cu, Pb or Zn. All these minerals, in particularscorodite, tend to concentrate gold (Boyle 1979).Oxidation of primary As minerals yield arsenic acidH3AsO4. In these forms, arsenic is relatively mobile, andunder certain conditions considerable amounts of theelement may be removed from the deposits duringoxidation.

Under oxidizing conditions, and in the presence ofiron, inorganic arsenic species are predominantly retainedin the solid phase through interaction with iron oxy-hydroxide coatings on soil particles (Bose & Sharma2002). These authors described several types ofinteractions, i.e., adsorption on amorphous ironhydroxide and adsorption on ferrihydrite and co-precipitation of arsenic (III) and arsenic (V) with iron oxy-hydroxide occurring in an oxidizing environment.

No linear correlation occurs between Au and Ag, As,Sb, Cu, Pb, Zn in BLEG and 180-µm stream sedimentgeochemical and rock chip data (Table 5a & b; Figure 7)from the AYALE area. The critical value for t with 238degrees of freedom and 10% level of significance is t=1.645. Because most of the test statistics fall into uppercritical region, it is concluded that there are truecorrelations (r> 0.1) between the element variablesparticularly in the rock chip (Table 5b) data set (Davis1986). Nevertheless, weak correlations occur betweenCu-Zn and Pb-Zn in 180-µm (Figure 7). All these may becaused by different dissolution and transportcharacteristics of these metals in secondary environmentas well as their emplacement into the host rock atdifferent levels and phases (Boyle 1979; Craw 2001;Bose & Sharma 2002). On the other hand, very weakcorrelations between Au and Ag and As in rocks (Table 5a& b; Figure 7), as in 180-µm stream-sediment samples,at AYALE suggests that they may be related to differentmineralizing events, thereby indicating possibleintroduction of Au, Ag and As in three different phases ofmineralization within the same structural zone. Althoughnot always, anomalous gold zones are usuallyaccompanied by anomalous As zones (Figure 5).Relatively higher correlations occurring between Cu andPb, Zn; Pb and W; Zn and Sb, W and, Mo and W (Table5a & b; Figure 6) in the same mineralized rock-chipsamples may also indicate two distinct phases ofmineralization at least two different times because theseelements would be expected to be found associated within

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

50

intrusive-centred porphyry or skarn systems. Thepresence of weak to moderate correlations between Cuand Pb, Zn as well as W and Pb, Zn, Sb, Mo in rock-chipsamples may suggest that 180-µm stream-sedimentsamples should also be analyzed for Mo and W duringfurther exploration in the AYALE area. Elementscommonly found enriched in mesothermal lode golddeposits in metamorphic terrains include Au, Ag, As, Sb,Hg, W, Bi and Mo (Pirajno 1992). Less commonly, Pb, Znand Cu may be present. Therefore, the AYALE area, morespecifically the Tuztafl› area, may be a significant targetfor some of the above-mentioned mineralization hostedby metamorphic rocks.

Much of the arsenic in the Tuztafl› area might havebeen co-precipitated and/or adsorbed by hydrous iron andother oxides or have reacted with cations such as Fe andCu to give a variety of insoluble arsenates during varioushydrolytic and colloidal reactions in oxidized zones assuggested by Boyle (1979), Craw (2001) and Bose &Sharma (2002) elsewhere. These reactions take placemainly between pH from 4 to 7 (Bose & Sharma 2002),and this may also be so in the Tuztafl› area where someof the illite and chlorite formed possibly by weathering

indicate weakly acidic to neutral conditions. The wallrocks and gangue also contain abundant carbonateminerals such as calcite to neutralize the downwardmoving solutions. Arsenic concentration is variable andclearly not exclusively related to Au enrichment except fortheir coexistence within the same structure (Table 6a & b;Figures 5–7). A similar situation exists for the silver aswell. Positive correlation coefficients between Au and Agand, Au and As, are weak to moderate (r< 0.5) inthe180-µm in soil data set. The critical value for t with531 (samples) degrees of freedom and 10% level ofsignificance is t= 1.645. It is concluded that there aretrue correlations (r > 0.1) between the significantnumbers of element variables within the soil (Table 6b)data set (Davis 1986). These metal associations can onlybe encountered in epithermal and to a certain extent inmesothermal precious metal deposits. A close associationof Au with As and Ag in soil as well as a very closeassociation of W with Sb, Pb, Zn and Mo in rock chipsamples indicate two distinct geochemical signatures thatare possibly related to the epi- to meso-thermal preciousand base metal skarn mineralization systems in theAYALE area.

H. YILMAZ

51

Table 5. Matrix of correlations between measured variables for Tuztafl› prospect rock chip data set (a) and calculation of t values (b) using thecorresponding correlation coefficiecients in (a).

Var* Au Ag Cu Pb Zn As Sb Mo W Bi

Tl -0.09 0.05 0.36 0.06 0.07 -0.01 -0.02 0.12 0.01 0.22Bi 0.21 0.01 0.20 0.07 0.05 0.07 -0.01 0.05 0.02W -0.04 0.22 0.37 0.97 0.96 0.18 0.61 0.39Mo 0.01 0.33 0.18 0.32 0.34 0.05 0.56Sb -0.02 0.11 0.29 0.57 0.60 0.31As 0.25 0.13 0.07 0.19 0.18 (a)Zn -0.04 0.27 0.45 1.00Pb -0.04 0.27 0.44Cu 0.05 0.18Ag 0.28

Var** Au Ag Cu Pb Zn As Sb Mo W Bi

Tl -1.39 0.77 5.95 0.93 1.08 -0.15 -0.31 1.86 0.15 3.48Bi 3.31 0.15 3.15 1.08 0.77 1.08 -0.15 0.77 0.31W -0.62 3.48 6.14 61.56 52.89 2.82 11.88 6.53Mo 0.15 5.39 2.82 5.21 5.58 0.77 10.43Sb -0.31 1.71 4.67 10.70 11.57 5.03As 3.98 2.02 1.08 2.99 2.82 (b)Zn -0.62 4.33 7.77 1090.79Pb -0.62 4.33 7.56Cu 0.77 2.82Ag 4.50

Var* - variable; high-lighted values refer to relatively higher correlations between elements, n= 238.Var** - variable; critical value t= 1,645 for 238 samples. High-lighted values suggest that there is a real correlation between variables.

Although the Akp›nar-Evciler south area stretchingfrom Tuztafl› to Karaköy hosts gold-Ag-As-rich quartzveins/breccia zones, a major, coherent 180-µm streamsediment As anomaly appears to be confined to astructural divide, where Upper Cretaceous–Palaeocenemélange, Upper Oligocene granitoids and Miocenevolcanic rocks, overlie the exposed metamorphic core.This may be caused by the erosion of possible As-bearinghigh-level mineralization zones during exhumation of themetamorphic rocks from ~14-km to ~7-km along anorthward-dipping ductile shear zone (Figures 1, 2 & 5;Okay et al. 1991; Okay & Sat›r 2000).

BLEG, 180-µm stream-sediment and soil samplingtechniques appear to be an effective reconnaissance toolin the AYALE area in west Turkey. The persistence of180-µm Au anomalies for less than 1000 m downstreamfrom the Tuztafl› Prospect (from 90 ppb to 7 ppb)suggests that 180-µm stream-sediment sampling is notas efficient as BLEG on 212 ppb) anomalies.

At the initial stage of regional reconnaissance, BLEGanomalies indicated that Ag in the Tuztafl› area mightoccur as electrum or sulfosalts (log-transformed Ag/Au:4.6). This was further supported by high Ag contents(130 ppm) and log-transformed Ag/Au ratios (3).Arsenic-Sb and Ag are the most reliable pathfinderelements for Au in the study area.

Conclusions

BLEG stream sediment geochemical sampling is a time-and cost-efficient method in assessing large areas ofrugged terrains (1350 km2 in this case). The re-discovery

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

52

Table 6. Matrix of correlations between measured variables for Tuztafl› prospect soil data set (a) and calculation of t values (b) using thecorresponding correlation coefficiecients in (a).

Var* Au Ag Cu Pb Zn As Sb Mo W Bi

S 0.00 0.05 0.23 0.08 0.44 0.24 -0.04 -0.36 -0.07 -0.20Bi 0.15 -0.03 0.07 -0.03 -0.04 -0.01 -0.01 0.45 0.04W 0.01 -0.04 0.05 0.01 0.06 -0.08 0.05 -0.02Mo 0.15 0.09 0.06 -0.03 -0.13 -0.10 0.13Sb 0.03 0.01 0.02 -0.02 -0.03 -0.01As 0.40 0.11 0.08 0.08 0.07 (a)Zn -0.02 0.01 0.51 0.18Pb 0.05 0.02 0.05Cu 0.08 0.02Ag 0.43

Var** Au Ag Cu Pb Zn As Sb Mo W Bi

S 0.00 1.15 5.44 1.85 11.27 5.69 -0.92 -8.88 -1.61 -4.69Bi 3.49 -0.69 1.61 -0.69 -0.92 -0.23 -0.23 11.59 0.92W 0.23 -0.92 1.15 0.23 1.38 -1.85 1.15 -0.46Mo 3.49 2.08 1.38 -0.69 -3.02 -2.31 3.02Sb 0.69 0.23 0.46 -0.46 -0.69 -0.23As 10.04 2.55 1.85 1.85 1.61 (b)Zn -0.46 0.23 13.64 4.21Pb 1.15 0.46 1.15Cu 1.85 0.46Ag 10.95

Var* - variable; high-lighted value refer to relatively higher correlations between elements, n= 531.Var** - variable; Critical value t= 1,645 for 531 samples. High-lighted values suggest that there is a real correlation between vaiables.

H. YILMAZ

53

XL-1

XL-3

XL-5

XL-9

XL-11

XL-13

XL-15

XL-17

F

F

F

200

50

100

100

F

0-25

25-100100-400400-1600>1600

Au (ppb)quartz veins

road

0 200m

drainage

470

000

43 94 000

XL-1

XL-3

XL-7

XL-9

XL-11

XL-1

3

XL-1

5

XL-1

7

0-25

25-100

100-400

>400

0 200 m

As (ppm)

470

000

43 94 000

XL-7

XL-5

(a)

(b)

Figure 8. Distribution of (a) Au and (b) As in 180-µm soil at the Tuztafl› Prospect.

of the Tuztafl› prospect during the follow up of an alreadydelineated BLEG anomaly is an exploration success using180-µm stream sediment and soil geochemistry.Moreover, the BLEG sampling technique proved to be avery sensitive exploration tool even in discovering small-scale mineralization in areas with well-developed drainageas in the Tuztafl› area. Identification and sampling ofaltered and mineralized rock float in streams was criticalin ranking the regional geochemical results. Soil samplingeffectively delineated Au and As geochemical zonings. TheBLEG geochemical technique should not be used inisolation and should be accompanied by a compilationstudy of recent, old and ancient workings, and knownmineralization since the Tuztafl› deposit had been minedon a small-scale mine during ancient times. Two distinct

180-µm Au and As geochemical signatures in the Tuztafl›area appear to be confined to medium- to high-grademetamorphic rocks forming the footwall and crustalhanging-wall accretionary mélange of a majordetachment fault, respectively.

Acknowledgement

I would like to express my appreciation to EurogoldMadencilik/Normandy Mining Ltd., Turkey for generousfinancial support to the project during my tenure asEurogold Exploration Manager. Reviewers Robert B.Cook and anonymous are thanked for their invaluablecomments in improving the quality of this paper. John A.Winchester helped with English of the final text.

STREAM SEDIMENT GEOCHEMICAL EXPLORATION FOR GOLD

54

References

ALDANMAZ, E. 2006. Mineral-chemical constraints on the Miocene calc-alkaline and shoshonitic volcanic rocks of western Turkey:disequilibrium phenocryst assemblages as indicators of magmastorage and mixing conditions. Turkish Journal of Earth Sciences15, 47–73.

AYDAL, D., VURAL, A. & POLAT, O. 2004. Definition the base metal- andgold-bearing hydrothermally-altered areas in volcanic rocks usingLandsat 7 Tm imagery: case study from Bayramiç (Çanakkale).57th Geological Congress of Turkey, Ankara, Abstracts, p. 90.

BECCALETTO, L. & JENNY, C. 2004. Geology and correlation of the EzineZone: A Rhodope fragment in NW Turkey. Turkish Journal ofEarth Sciences 13, 145–176.

BECCALETTO, L. & STENER C. 2005. Evidences of two-stage extensionaltectonics from the northern edge of the Edremit Graben, NWTurkey. Geodinamica Acta 18/3, 225–239.

BEESON, R. 1995. A drainage sediment geochemical orientation study atBoddington, Western Australia. Journal of GeochemicalExploration 54, 63–71.

BERGER, B.G. & BETHKE, P.M. 1984. Geology and geochemistry ofepithermal systems. Reviews in Economic Geology 2, 169–202.

BEUS, A.A. & GRIGORIAN, S.V. 1977. Geochemical Exploration Methodsfor Mineral Deposits. Applied Publishing Ltd., Illinois, USA.

B‹NGÖL, E. 1969. Geology of the central and southeastern parts of theKazda¤ Massif. Maden Tetkik ve Arama (MTA) Dergisi 72,110–23 [in Turkish with English abstract].

BOSE, P. & SHARMA, A. 2002. Role of iron in controlling speciation andmobilization of arsenic in subsurface environment. WaterResearches 36, 4916–4926.

BOYLE, R.W. 1979. The geochemistry of gold and its deposits. BulletinGeological Survey Canada 280, 1–584.

COHEN, D.R., DUNLOP, A.C. & ROSE, T. 2005. Contrasting dispersionpatterns for gold in stream sediments in Timbarra, NSW,Australia. Journal of Geochemical Exploration 85, 1–16.

CRAW, D. 2001. Tectonic controls on gold deposits and theirenvironmental impact, New Zealand. Journal of GeochemicalExploration 73, 43–56.

DAVIS, J.C. 1986. Statistics and Data Analysis in Geology. John Willey &Sons, New York, USA.

DELALOYE, M. & B‹NGÖL, E. 2000. Granitoids from western and northernAnatolia: Geochemistry and modeling of geodynamic evolution.International Geological Review 42, 241–268.

DURU, M., PEHL‹VAN, fi., fiENTÜRK, Y., YAVAfl, F. & KAR, H. 2004. Newresults on the lithostratigraphy of the Kazda¤ Massif in NorthwestTurkey. Turkish Journal of Earth Sciences 13, 177–186.

ELIOT, S.M. & TOWSY, C.A. 1989. Regional Drainage GeochemicalExploration Techniques Used in Queensland, Australia. NorthQueensland Gold ’89 Conference, Townsville, Australia, IMMPublication, 51–61.

ERCAN, T., TÜRKECAN, A., AKYÜREK, B., GÜNAY, E., CEV‹KBAfl, A., ATEfl, M.,CAN, B., ERKAN, M. & ÖZK‹R‹flÇ‹, E. 1984. The geology of Dikili-Bergama-Çandarl› area (Western Anatolia) and petrology of themagmatic rocks. Jeoloji Mühendisli¤i Dergisi 20, 47–60 [inTurkish with English abstract].

ERCAN, T., SATIR, M., STEINITZ, G., DORA, A., SARIFAKO⁄LU, E., ADIS, C.,WALTER, H.J., YILDIRIM, T. 1995. Biga yar›madas› ile Gökçeada,Bozcaada ve Tavflan adalar›ndaki (KB Anadolu) Tersiyervolkanizmas›n›n özellikleri [The characteristics of Tertiaryvolcanism in Biga Peninsula, Gökçeada, Bozcaada and Tavflanislands (NW Anatolia)]. Maden Tetkik ve Arama (MTA) Dergisi117, 55–86 [in Turkish with English abstract].

ERKÜL, F., HELVACI, C. & SÖZB‹L‹R, H. 2005. Stratigraphy andgeochronology of the Early Miocene volcanic units in the Bigadiçborate basin, western Turkey. Turkish Journal of Earth Sciences14, 227–253.

GENÇ, fi.C. 1998. Evolution of the Bayramiç magmatic complex,northwestern Anatolia. Journal of Volcanology and GeothermalResearch 85, 233–49.

KILIÇ, M., KÜÇÜKEFE, fi., AVflAR, M., SARI, R., VURAL, A. & PEHL‹VAN, N.2004. Preliminary geological and geochemical data on K›sac›k(Ayvac›k-Çanakkale) Au mineralization. 57th Geological Congressof Turkey, Ankara, Abstracts, p. 101.

LEVINSON, A.A. 1974. Introduction to Exploration Geochemistry. AppliedPublishing Ltd., Wilmette, IL.

MAZZUCCHELLI, R.H. 1990. Advanced Exploration Geochemistry.Workshop course 704/90, Australian Mineral Foundation,Publication 550.84, Australia.

MAZZUCCHELLI, R.H. 1994. Drainage geochemistry in gold exploration.In: HALE, M. & PLANT, J.A. (eds), Drainage Geochemistry,Handbook of Exploration Geochemistry 6. Elsevier, Amsterdam,379–414.

MEIER, A.L. 1980. Flameless atomic-absorbtion determination of gold ingeological materials. Journal of Geohemical Exploration 13,77–85.

MTA (Maden Tetkik ve Arama Enstitüsü) 1966. Copper, Lead and ZincDeposits of Turkey. MTA Publications 133, Ankara.

MTA (Maden Tetkik ve Arama Enstitüsü) 1993. Lead and Zinc Depositsof Turkey. MTA Publications 199, Ankara.

NICHOLS, I., HALE, M. & FLETCHER, W.K. 1994. Drainage geochemistry ingold exploration. In: HALE, M. & PLANT, J.A. (eds), DrainageGeochemistry, Handbook of Exploration Geochemistry 6.Elsevier, Amsterdam, 499–557.

OKAY, A.‹, S‹YAKO, M. & BÜRKAN, K.A. 1991. Geology and tectonicevolution of the Biga Peninsula, northwest Turkey. Bulletin of theTechnical University of ‹stanbul 44, 191–256.

OKAY, A.‹., SATIR, M., MALUSKI, H., S‹YAKO, M., MONIE, P., METZGER, R. &AKYÜZ, S. 1996. Paleo- and Neo-Tethyan events in northwestTurkey: geological and geochronological constraints. In: YIN, A. &HARRISON, M. (eds), Tectonics of Asia. Cambridge University Press,420–441.

OKAY, A.‹. & SATIR, M. 2000. Coeval plutonism and metamorphism in alatest Oligocene metamorphic core complex in northwest Turkey.Geological Magazine 137, 495–516.

ÖNGEN, S. 1978. Petrographie und Petrochemie des Çavuslu-KaraköyGranitoid Massivs, ‹stanbul Üniversitesi FenFakültesi Mecmuas›B43, 93–115.

ÖNGEN, S. 1994. Le pluton calco-alcalin d’Evciler (peninsula de Biga,Turquie-NW): age, geochemie et signification geodynamique.Comptes Rendus de l’Académie des Sciences, Paris 319,1033–1039.

OTTESEN, R.T. & THEOBALD, P.K. 1994. Stream sediments in mineralexploration. In: HALE, M. & PLANT, J.A. (eds), DrainageGeochemistry, Handbook of Exploration Geochemistry 6.Elsevier, Amsterdam, 147–184.

PIRAJNO, F. 1992. Hyhdrothermal Mineral Deposits. Springer-Verlag.New York, Berlin Heidelberg.

RADFORD, N. 1996. BLEG sampling in gold exploration: an AustralianView. Explore 92, 8–10.

ROSE, A.W., HAWKS, H.E. & WEBB, J.H. 1979. Geochemistry in MineralExploration. Academic Press, NY, USA.

SCHUILING, R.D. 1959. Über eine prä-herzynische Faltungsphase imKazda¤ Kristallin. Maden Tetkik ve Arama (MTA) Dergisi 53,89–93.

TAYLOR, S.R. 1966. The application of trace element data to problems inpetrology. Physics and Chemistry Earth 6, 133–213.

THOMPSON, M. & HOWARTH, R.J. 1978. A new approach to the estimationof analytical precision. Journal of Geochemical Exploration 9,23–30.

TOKÇAER, M., AGOSTINI, S. & SAVAflÇIN, M.Y. 2005. Geotectonic setting andorigin of the youngest Kula volcanics (western Anatolia), with anew emplacement model. Turkish Journal of Earth Sciences 14,143–166.

XUEQIU, W., XUEJING, X., ZHIZHONG, C. & DAWEN, L. 1999. Delineation ofregional geochemical anomalies penetrating through thick coverin concealed terrains – a case history from the Olympic Damdeposit, Australia. Journal of Geochemical Exploration 66,85–97.

WILHELM, E. & ARTIGNAN, D. 1994. Drainage geochemistry in goldexploration. In: HALE, M. & PLANT, J.A. (eds), DrainageGeochemistry, Handbook of Exploration Geochemistry 6.Elsevier, Amsterdam, 415–444.

WOOD, D.G., PORTER, R.G. & WHITE, N.C. 1990. Geological features ofsome Paleozoic epithermal gold occurrences in northeasternQueensland, Australia. Journal of Geochemical Exploration 36,413–443.

YANG, S.X. & BLUM, N. 1999. A fossil hydrothermal system or a sourcebed in the Madiyi Formation near the Xiangxi Au-Sb-W deposit,NW Hunan, PR of China. Chemical Geology 155, 151–169.

YILMAZ, H. 2003. Geochemical exploration for gold in western Turkey:success and failure. Journal of Geochemical Exploration 80,117–135.

YILMAZ, H. & KARA, Z. 1996. Kazda¤ Project. Eurogold, TurkeyCompany Report, 1–20 [unpublished].

YÜCEL-ÖZTÜRK, Y., HELVACI, C. & SATIR, M. 2005. Genetic relationsbetween skarn mineralization and petrogenesis of the Evcilergranitoid, Kazda¤, Çanakkale, Turkey and comparison with worldskarn granitoids. Turkish Journal of Earth Sciences 14, 255–280.

H. YILMAZ

55

Received 03 August 2005; revised typescript received 29 June 2006; accepted 13 October 2006