Mineralogy and Petrology (2002) 75: 101122

Geochemistry and petrography

of gold-quartz-tourmaline veins
of the Okote area, southern Ethiopia:
implications for gold exploration
D. J. Deksissa and C. Koeberl

Institute of Geochemistry, University of Vienna, Austria

With 10 Figures

Received November 17, 1999;

revised version accepted July 23, 2001

Summary
Quartz-tourmaline vein-hosting rocks of the Okote area belong to the Neoproterozoic
Adola Belt. Metasomatic auriferous quartz-tourmaline veins occur in ductile NS
trending, sinistral shear zones. These veins commonly contain quartz, carbonates, and
tourmaline, with minor pyrite, and accessory chalcopyrite, pyrrhotite, and gold. Tour-
maline forms isolated euhedral crystals in the fracture surfaces of quartz carbonate
veins. Many of the tourmaline crystals are optically zoned with a bluish core and a
bluish to brown rim. Electron microprobe analyses show that the tourmalines comprise
an intermediate dravite-schorl solid solution with a mean FeO=(FeO MgO) 0.47.
Abrupt transitions between the colour zones within single tourmaline crystals are ac-
companied by relative variations in the FeO=(FeO MgO) ratios. The tourmaline
separates indicate that the tourmalines contain highly variable average contents of trace
elements. Chondrite-normalized rare earth element (REE) abundances of tourmaline
separates from auriferous veins show LREE-enriched to LREE-depleted patterns with
negative to positive Eu anomalies and a at, near-chondritic HREE pattern. The au-
riferous quartz-tourmaline veins have LREE-enriched patterns without a Eu anomaly
and a at HREE pattern, but tourmaline-free gold-quartz veins have very low REE
contents and LREE-depleted patterns also without Eu anomalies. The FeO=(FeO
MgO) ratios, major and trace element compositions, and the types of wall-rock
alteration are used to suggest that the sources of boron are dominantly metamorphic
(dehydration and devolatilization processes), but do not totally exclude the possibility
of a magmatic source. The occurrences of high-grade gold associated with tourmaline
make tourmaline a valuable prospecting guide for hydrothermal gold mineralization in
the Adola Belt, southern Ethiopia.
102 D. J. Deksissa and C. Koeberl

Introduction
The relationship between tourmaline-rich rocks and gold deposits has been doc-
umented in many metallogenic provinces, such as in the Precambrian basement of
Ethiopia (Augustithis, 1967), Brazil (Fleischer and Routhier, 1973), Australia
(Plimer, 1986), in Caledonian rocks of southern Ireland (McArdle et al., 1989), and
the Archean Barberton greenstone sequence of South Africa (Byerly and Palmer,
1991). King (1988) found that the REEs in tourmalines from the gold vein deposits
of Lake Superior do not correlate with the REE levels in the host rock. He sug-
gested that there must have been REE mobilization in hydrothermal solutions.
Tourmaline is also common in hydrothermal gangue minerals in Archean to
Phanerozoic mesothermal and intrusion-related gold deposits (see, for example,
King, 1988; Groves and Foster, 1991; Nesbitt, 1991; Garba, 1996). Garba (1996)
and King (1988) indicated that tourmaline is a potential pathnder mineral for
source region and vein gold mineralization. The composition of hydrothermal tour-
malines is sensitive to variables, such as uid chemistry, physicochemical con-
ditions, and the nature of uid rock interactions (e.g. uid rock ratios; Garba,
1996). Bone (1988) suggested that tourmalinites (with tourmaline > 15 vol.%) are
somewhat enigmatic rocks and highly signicant as ngerprints of unusual hydro-
geochemical and paleogeographical conditions and should foster exploration
interests.
In this study we discuss the petrography and geochemistry of auriferous quartz-
tourmaline veins and their signicance for gold exploration in the Adola Belt.

Geological setting
The Okote gold mineralization is located in the southern part of the Adola Belt.
The Adola Belt, covering an area of about 5000 km2 (Billay et al., 1997), consists
of two, NS trending volcanosedimentary belts, namely the Kenticha Belt in the
east and the Megado Belt in the west (Fig. 1; Upper Complex). These two belts are
separated and bounded in the east and west by high grade gneisses of the Middle
complex and Lower complex (according to the classication of Kazmin, 1972).
The study area is the southward extension of the Megado volcanosedimentary
belt. The lithologic contacts, foliation surfaces, fold axes, and fault planes of the
Megado volcanosedimentary belt change in strike at the Okote area, from NS to
almost EW, due to a NW striking regional shear zone (Fig. 1). The lithologic
association, structural evolution, metamorphism, and geochemistry of the Megado
volcano-sedimentary belt has been documented in detail already by Woldehaimanot
and Behrmann (1995) and Worku and Schandelmeier (1996).
The geology of the Okote area consists of amphibole- and biotite-quartzo-
feldspathic gneiss, amphibolite, meta-ultramac rocks, meta-gabbro, meta-
volcanic rocks, chlorite-carbonate-amphibole schist, meta-tonalite, meta-granite,
plagiogranite, and meta-sedimentary rocks (graphitic quartzite and graphitic schist)
(Fig. 2). The meta-tonalite intruded the meta-volcanic rocks and meta-gabbro and
contains amphibolite xenoliths. The peraluminous meta-granite intruded the
biotite-quartzofeldspathic gneiss. The chlorite-carbonate-amphibole schist consti-
tutes three hydrothermally altered, NS trending, narrow, sinistral ductile shear
 Gold-quartz-tourmaline veins of southern Ethiopia 103

Fig. 1. Location of the study area on the geological map of southern Ethiopia. Inset shows
the inter-ngering relationships of the Arabian-Nubian Shield (ANS) and the Mozambique
Belt (MB) and the location of Precambrian rocks of southern Ethiopia (modied from
Mohammed, 1999; Kazmin, 1972; Vail, 1987)

zones that cut across metagabbro and consist of chlorite-carbonate-biotite schist,

epidote-amphibole-chlorite schist, and amphibole-carbonate-chlorite schist.
The Okote gold mineralization is restricted to an anastomosing system of
narrow high strain, NS trending, ductile shear zones. Within these shear zones,
gold occurs in quartz veins and quartz-carbonate-tourmaline veins, and their wall-
rocks.

Petrography of the host rock

The investigated quartz-tourmaline veins occur in chlorite-carbonate-amphibole
schist and talc-tremolite schist. The chlorite-carbonate-amphibole schist is an altera-
tion product of meta-gabbro and is sub-classied into chlorite-carbonate-biotite
schist, epidote-amphibole-chlorite schist, and amphibole-carbonate-chlorite schist,
based on their mineralogical assemblages and intensity of deformation (Fig. 3).
104 D. J. Deksissa and C. Koeberl

Fig. 2. Regional geological map of the Okote area, southern Ethiopia

Chlorite-carbonate-biotite schist: Chlorite-carbonate-biotite schist is ne-

grained, light green to grey in colour, intensely sheared, and contains abundant
NS trending, foliation parallel, auriferous, quartz-carbonate- and quartz-carbonate-
tourmaline veins. Chlorite, carbonate (calcite and ankerite), quartz, biotite, epidote,
and albite are the dominant minerals, whereas pyrite, muscovite, sericite, K-feldspar,
apatite, ilmenite, and rutile are accessory. Chlorite and carbonates constitute up to
80 vol.% of the rock. Chlorite, carbonate, and quartz crystals are preferentially
aligned parallel to the shear zone boundary. Biotite crystals appear to have formed by
the break-down of chlorites and muscovite. Quartz grains occur predominantly in
veins and veinlets that always enveloped by hydrothermal alteration minerals.
Calcite is crystallizing at the expense of ankerite.
Epidote-amphibole-chlorite schist: This schist subtype occurs next to chlorite-
carbonate-biotite schist towards fresh meta-gabbro. The contacts of epidote-
amphibole-chlorite schist with chlorite-carbonate-biotite schist and meta-gabbro are
always gradational. The schist is ne- to medium-grained, well foliated, and con-
sists of epidote, zoisite=clinozoisite, albite, actinolite-tremolite, hornblende, relict
plagioclase, chlorite, and accessory pyrite, magnetite, carbonate, apatite, and relict
 Gold-quartz-tourmaline veins of southern Ethiopia 105

Fig. 3. Detailed geological map of the sheared, altered, and mineralized part of the Okote
area, southern Ethiopia (after Deksissa, 2000)

pyroxene. Plagioclase porphyroclasts are almost completely altered to epidote,

clinozoisite and albite, and contain chlorite inclusions.
Amphibole-carbonate-chlorite schist: The amphibole-carbonate-chlorite schist
is dominantly exposed in the eastern alteration zone (Fig. 3). The schist is green in
colour, medium-grained, well foliated, and composed mainly of tremolite-actinolite,
carbonate, chlorite, albite, with accessory epidote zoisite=clinozoisite, iron oxides,
and relict plagioclase. The intensity of shearing is comparable to that observed for
the chlorite-carbonate-biotite schist.

Structural evolution of the Okote area

Three main deformation episodes have been recognized in the eld, based on dif-
fering structural styles and overprinting relationships. An earlier folding and
thrusting event (D1) was followed by NS shearing and folding (D2), and, nally,
folding and NW-striking sinstral brittle-ductile shearing (D3).
D1 structures: The rst progressive deformation D1 was responsible for the
formation of foliation and the oldest quartz veins. This deformation was followed by
thrusting that resulted in the development of recumbent folds and boudinaged folds
in quartz veins in biotite-quartzofeldspathic gneiss.
D2 structures: The second deformation D2 formed a regional synform in the NW
part of the area (Fig. 2) and NS trending shear zones. The D2 mesoscale structural
elements in the mac-ultramac rocks are regional penetrative foliations that strike
N20  E and dip steeply (60 to 80 ) to the west, quartz veins and quartz-carbonate-
tourmaline veins concordant to foliation, asymmetrical folds, S-C fabrics, and
anastomosing shear structures.
106 D. J. Deksissa and C. Koeberl

D3 structures: The mesoscale NWSE trending brittle to ductile shear zones

are very common in the study area. They clearly display sinistral sense of movement
and are younger than the NS trending structures. They contain concordant quartz
veins that cut and displace all of the older auriferous quartz veins.

Quartz veins
Three main groups of quartz veins associated with the above three deformation
phases have been observed. These are EW trending highly deformed quartz veins
(qv1), NS trending quartz-carbonate-tourmaline veins and quartz-carbonate veins
(qv2), and NWSE trending quartz veins (qv3). The oldest qv1 veins are tightly
folded, cut by the younger veins and shear foliations, and are barren of gold. The
intermediate gold-bearing qv2 veins dip subvertically towards west and locally
towards east, show pinch- and swell structure, and are massive to laminated. The
youngest veins (qv3) are very thin, barren of gold, and concordant to brittle shear
zones that strike 30 to 40 NW dip at 70 towards NE. The present work con-
centrates mainly on qv2 veins.
Textures of auriferous quartz-tourmaline veins (qv2): The veins are milky white
to light gray in colour and strike sub-parallel to NS trending chlorite-carbonate-
amphibole units and NS trending shear zones that are illustrated in Fig. 2. Vein
widths range from less than 10 centimeters to 3 meters and commonly vary dra-
matically over short distances along the strike and dip directions. Variation in
thickness is commonly the result of progressive shearing during and after vein
formation. The veins commonly display laminations on the cm-scale. Elongated
tabular clasts of wall-rocks are occasionally included in the veins, generally close to
the wall-rock contact, and the extremities of these clasts commonly display
continuity with the vein laminations. The mineralogical composition of the laminae
is dominated by chlorite, carbonate, sulde, and tourmaline. Tourmaline is a minor
to major gangue mineral in the gold-bearing veins at the Okote area. The major to
accessory gangue minerals are quartz, carbonate (ankerite and calcite), orthoclase,
albite, and white mica=muscovite. The ore minerals are mainly pyrite with sub-
ordinate amounts of chalcopyrite, and accessory pyrrhotite, covellite, chalcocite,
gold, and galena.
Tourmaline occurs in the form of black prismatic crystals visible in hand
specimen and shows a brown to bluish pleochroism in thin section. The basal sec-
tions of most of the studied tourmalines show clear colour zoning from bluish-green
cores to brownish rims (Fig. 4A). The boundary between the core and rim is
commonly sharp. A tourmaline sample from a talc-tremolite host rock shows os-
cillatory zoning, with a wide bluish core and thin alternating bluish and dark blue to
brown layers.
Tourmaline crystals show two modes of occurrences. The rst type is ne-
grained (20300 mm in cross-section), and euhedral acicular tourmaline crystals are
disseminated within the undeformed quartz grains (Fig. 4B). The second type consist
of tourmalines concentrated along the fractures within the quartz veins (Fig. 4C, D).
The latter tourmalines commonly ll cavities by developing comb-textured
tourmaline crystals perpendicular to the wall of the cavity. The tourmalines that
follow cracks and intergranular spaces of the quartz indicate a later paragenetic
 Gold-quartz-tourmaline veins of southern Ethiopia 107

Fig. 4. Photomicrograph (plane polars) showing A tourmaline crystal with sharp boundary
between core and rim, B ne-grained tourmaline crystals disseminated in quartz matrix,
C tourmaline crystallization along fractures in auriferous quartz vein, and D tourmaline
with radial growth pattern in fracture of quartz vein. Width of eld of view is 1.5 mm

sequence. The quartz from the older quartz veins is highly deformed, shows
undulatory extinction and marginal recrystallization. These features generally
indicate that the tourmaline fracture llings are younger than the disseminated
tourmalines. In quartz-carbonate veins, calcite seems to be cogenetic with the second
group of tourmalines.
The crystallization sequence of vein materials can be deduced from patterns of
mineral association. The rst generation of tourmaline is characterized by well-
developed elongated prismatic crystals disseminated in the rst generation quartz
matrix (Fig. 4B). The crystallization of quartz was not restricted to this initial single
stage. Calcite inclusions in the rst generation quartz-tourmaline vein contain
ankerite. The second generation tourmaline with native gold occurs as inclusions in
calcite (Fig. 5). Very small grains of galena, commonly less than 10 mm in size,
occur within grain boundaries of tourmalines. The occurrence of tourmaline in a
calcite matrix indicates that calcite formed syn- to post genetically to tourmaline.
Considering eld, thin section, and electron microprobe studies, we suggest that
tourmaline, calcite, and gold formed cogenetically. Galena is the youngest mineral
observed in the quartz-tourmaline veins.
108 D. J. Deksissa and C. Koeberl

Fig. 5. Back-scattered elec-

tron image showing relation-
ships among tourmaline (T),
gold (Au), calcite (C) and
quartz (Q) within the quartz-
tourmaline vein

Gold mineralization
Gold occurs as native gold inclusions in chalcopyrite, pyrite, calcite (both in veins
and wall-rock), and as free gold in the quartz matrix. The auriferous veins are always
surrounded by less than a meter to tens of meters wide intense hydrothermal
alteration. The fabrics and primary minerals of the wall-rocks have commonly been
replaced by chlorite, carbonate (predominantly calcite and minor ankerite), pyrite,
and biotite with minor chalcopyrite and pyrrhotite. Tourmaline occurs in quartz
veins containing very high concentrations of gold (e.g., sample DKR-75, Au
56.7 ppm, see Table 3). Although the gold concentration displays signicant
temporal and spatial association with the lamination, gold particles tend to occur in
fractures in the quartz associated with calcite and tourmaline.

Geochemistry

Sampling and analytical methods

Representative bulk samples have been collected from auriferous quartz-carbonate-
tourmaline veins in chlorite-carbonate-amphibole schist for the major and trace
element analyses using electron microprobe, Inductively Coupled Plasma Mass
Spectrometer (ICP-MS), and Instrumental Neutron Activation Analysis (INAA).
The sample locations are shown in Figs. 2 and 3.
Electron microprobe analyses: Polished sections have been prepared from
tourmaline-rich samples and were carbon-coated before analysis. Tourmaline
compositions were determined using a CAMECA SX 100 electron microprobe at
the Institute of Petrology, University of Vienna (Table 1). Natural standards were
used for analyses of major and minor elements abundances. Measurement times
per element were 20 s with the emission current of 15 nA and an acceleration
voltage of 15 kV. The content of Na was measured rst for 10 s to avoid errors
resulting from the volatilization.
Table 1. Composition of representative tourmaline crystals from auriferous quartz-tourmaline veins of the Okote area, southern Ethiopia
DKR-31 DKR-36 DKR-43 DKR-52 DKR-56 DKR-58 DKR-75
Rim Core Rim Core Rim Core Rim Core Rim Core Rim Core
SiO2 36.16 37.58 36.77 36.87 36.91 37.41 37.05 37.54 35.80 37.30 36.12 37.29 37.29 37.17
TiO2 0.44 0.04 0.45 0.41 0.52 0.18 0.97 0.12 1.09 0.06 0.42 0.04 0.08 0.21
B2O3 Calc 10.59 10.80 10.78 10.73 10.63 10.85 10.72 10.81 10.38 10.68 10.58 10.85 10.80 10.82
Al2O3 31.86 33.80 33.11 32.84 32.54 33.86 30.58 32.24 30.53 33.02 32.39 34.13 33.68 33.75
FeO 8.08 6.82 5.89 5.69 7.89 4.47 6.71 6.24 9.11 7.33 7.59 6.61 4.40 5.54
MnO 0.02 0.04 0.01 0.03 0.01 0.00 0.03 0.05 0.00 0.00 0.02 0.00 0.00 0.02
MgO 6.69 6.37 7.60 7.62 5.69 7.82 8.29 7.98 5.86 6.19 6.69 6.49 7.74 7.37
CaO 1.01 0.24 0.85 0.69 0.41 0.24 0.74 0.26 0.89 0.23 0.69 0.21 0.55 0.61
Na2O 2.25 1.85 2.01 2.06 2.17 2.28 2.63 2.75 1.87 1.86 1.98 1.70 1.58 1.77
K2O 0.01 0.01 0.00 0.01 0.01 0.01 0.00 0.02 0.01 0.01 0.00 0.01 0.02 0.01
LiO2 Calc 0.16 0.21 0.15 0.16 0.30 0.18 0.13 0.18 0.17 0.25 0.09 0.10 0.19 0.13
Total 97.27 97.76 97.62 97.11 97.08 97.30 97.85 98.19 95.71 96.93 96.57 97.43 96.33 97.4
Structural formulae calculated on the basis of 29 oxygens
Si 5.936 6.047 5.931 5.968 6.035 5.994 6.011 6.034 5.993 6.072 5.933 6.002 6.022 5.972
AlIV 0.064 0.069 0.032 0.006 0.007 0.067 0.028
B 2.999 3.000 3.001 2.999 3.000 3.001 3.002 2.999 2.999 3.001 3.000 3.000 3.011 3.001
Ti 0.054 0.004 0.054 0.050 0.064 0.022 0.118 0.015 0.138 0.007 0.052 0.005 0.010 0.025
Al 5.946 5.996 5.946 5.950 5.936 5.978 5.882 5.985 5.862 5.993 5.948 5.995 5.990 5.975
Al 0.150 0.414 0.279 0.286 0.335 0.409 0.123 0.152 0.342 0.256 0.485 0.420 0.389
Fe 1.108 0.917 0.794 0.170 1.079 0.599 0.910 0.839 1.275 0.997 1.043 0.891 0.594 0.744
Mn 0.003 0.006 0.002 0.004 0.001 0.000 0.003 0.006 0.001 0.000 0.003 0.000 0.000 0.003
Gold-quartz-tourmaline veins of southern Ethiopia

Mg 1.635 1.528 1.827 1.840 1.388 1.868 2.006 1.912 1.462 1.501 1.638 1.558 1.863 1.765
Li 0.105 0.135 0.096 0.102 0.196 0.119 0.081 0.116 0.109 0.159 0.057 0.062 0.120 0.083
Y total 3.001 3.000 2.998 3.001 2.999 2.995 2.997 2.997 3.000 2.999 3.000 3.000 3.000 3.000
Ca 0.117 0.041 0.146 0.120 0.073 0.042 0.128 0.044 0.160 0.040 0.121 0.036 0.095 0.105
Na 0.714 0.578 0.629 0.647 0.687 0.070 0.826 0.856 0.606 0.586 0.631 0.531 0.495 0.551
K 0.001 0.001 0.001 0.002 0.002 0.002 0.001 0.004 0.003 0.001 0.000 0.002 0.004 0.002
X total 0.891 0.619 0.775 0.767 0.760 0.749 0.955 0.900 0.765 0.626 0.752 0.569 0.594 0.658
109

Fe=(Fe Mg) 0.404 0.375 0.303 0.295 0.437 0.243 0.312 0.305 0.466 0.399 0.389 0.364 0.242 0.297
Na=(Na Ca) 0.801 0.934 0.811 0.843 0.905 0.945 0.866 0.951 0.791 0.936 0.839 0.937 0.839 0.840
Oxides in wt %; total Fe as FeO; Calc calculated;  tourmaline sample without zoning
110 D. J. Deksissa and C. Koeberl

The calculation of the atomic proportions was made by assuming that B has a
stoichiometric value of 3 atoms per formula unit (apfu), OH lls the four O(1) O(3)
sites (i.e., normalization of the cations based on 29 oxygen), and Li lls the
deciencies in the Y site, i.e., Li 3-Ysite total. Both the amount of B2O3 necessary to
produce 3 B cations and the amount of Li2O required to ll the deciencies in Y were
calculated from stoichiometry.
After establishing the appropriate normalization scheme, the cations were
allocated among the different sites of the tourmaline. Na, Ca, and K are assigned to
the X site, with any site deciency assumed to represent an X-site vacancy. Si is
assumed to be exclusively in the T site, with any deciency made up by Al. There
is more Al than cations required to ll the Z site. Excess Al from T and Z sites and
Fe, Mn, Mg, and Li were assigned to the Y site. All iron is assumed to be Fe2 , as
total charge calculation gave no charge excess or deciency. It is common practice
to assume that all transition elements are in the divalent state (e.g., all Fe as

Table 2. Trace element contents of tourmaline, Okote area, southern Ethiopia

DKR-31 DKR-36 DKR-43 DKR-52 DKR-56 DKR-58
Sc 38.4 10.9 64.2 9.58 32.6 37.3
Cr 34.3 20.9 177 614 3.32 539
Co 21.4 21.3 55.7 29.9 52.4 21.4
Ni < 114 224 484 19.6 36.5 96
Zn 54 82 195 23 47 72
As 0.61 2.36 0.21 < 0.6 1.36 < 0.6
Rb 1.68 1.15 < 8.3 < 5.2 4.55 < 7.1
Sr 114 96 < 161 2009 138 168
Zr < 131 30 < 136 9.1 < 112 < 102
Sb 0.70 0.74 0.38 0.07 0.3 0.12
Cs 0.70 < 0.1 < 0.5 0.12 0.12 0.06
Ba < 84 < 43 36 18 < 82 < 70
La 3.96 0.13 0.17 0.08 0.13 0.36
Ce 5.88 1.75 < 1.2 0.48 < 1.1 < 1.0
Nd 2.92 1.67 < 2.7 < 1.7 < 2.6 < 2.3
Sm 0.95 0.07 0.09 0.02 0.06 0.40
Eu 0.19 0.03 0.07 0.21 0.07 0.04
Gd < 2.5 < 1.0 < 3.0 < 1.7 < 2.3
Tb 0.14 < 0.1 0.08 0.06 0.09 0.10
Tm < 0.1 < 0.01 0.51 < 0.1 < 0.1 0.10
Yd 0.33 0.18 0.17 0.14 0.31 0.19
Lu 0.05 < 0.01 0.02 0.01 0.05 0.03
Hf 0.42 0.08 0.18 0.07 < 0.3 0.8
Ta 0.29 0.02 1.10 0.13 0.42 0.39
W 0.13 0.2 1.29 0.68 0.54 0.52
Au (ppb) 29 2.7 3.0 2.90 1.7 < 5.2
Th 0.13 0.02 0.11 0.03 < 0.2 < 0.2
U 0.30 0.05 < 0.3 < 0.6 0.61 0.31
All contents are in ppm, except as noted
 Gold-quartz-tourmaline veins of southern Ethiopia 111

FeO), particularly if there is no deciency in the cation charge (Henry and Dutrow,
1996).
Trace element analyses: Bulk quartz vein samples of about 10 kg were collected
from auriferous quartz-carbonate-tourmaline veins and quartz-carbonate veins of
the similar host rock, strike and dip, and, possibly, the same age. The vein samples
were crushed to less than 2 mm diameter in a stainless steel jaw crusher and 200 g
sub-samples were taken by the cone and quartering method. The nal samples were
powdered in an agate mill. Tourmaline-rich parts of the same quartz-tourmaline
veins were sampled separately, coarsely crushed (less than 2 mm), and pure tour-
maline crystals were selected under a binocular microscope. These pure tourmaline
crystals were pulverized in an agate mortar.
The vein samples were analyzed for the contents of Na, Mg, Al, K, Ca, V, Cr,
Mn, Fe, Co, Ni, Zn, Ga, Rb, Sr, Ba, Au, Bi, and Pb using a Perkin-Elmer Sciex
Elan 5000 ICP-MS at the Arsenal Laboratory, Vienna, Austria for (see Table 3).
For the analysis 100 mg samples were dissolved in high purity acids (1:1.5 ml
HNO3 HClO4 [2.5 ml of each] and 10 ml HF). The solution was evaporated to
near-dryness at 120  C. Then, 5 ml ultra pure HNO3 were added and the residue
dissolved and cooled to room temperature. Ultrapure water was added and the
contents were transferred into a 100 ml volumetric ask. Finally, 2 ml of the sample
solution were diluted with 8 ml ultrapure water. The analytical precision has been
calculated based on duplicate analyses and was better than 10 rel.% for most of
these elements.
Rare earth elements (REE) and other trace elements (Sc, Co, Ni, Zn, As, Rb, Sr,
Zr, Sb, Cs, Ba, Hf, Ta, W, Au, Th, and U) in bulk vein samples and tourmaline
separates were analyzed using INAA (Table 2), following procedures described by
Koeberl (1993).

Results

Tourmaline chemistry
The major and minor element compositions of hydrothermal tourmaline grains
from the Okote area are reported in Tables 1, 2. Signicant compositional variations
occur for TiO2, CaO, FeO, MgO, and Na2O. Tourmaline samples from the
talc-tremolite schist host rock have the highest average abundances of TiO2, MgO,
and Na2O, and the lowest abundance of Al2O3. The content of F was not directly
determined, which may explain the low totals of the microprobe analyses.
The average composition of the tourmaline samples varies from 0.01 to
1.10 wt% TiO2, 4.40 to 8.56 wt% FeO, 0.21 to 1.01 CaO, 5.52 to 8.63 wt% MgO,
and 1.48 to 2.75 wt% Na2O; the FeO=(FeO MgO) ratios range from 0.42 to 0.61
(mean 0.47, n 7). The average Na2O=(Na2O CaO) ratios of individual tour-
maline samples vary from 0.68 to 0.91 (mean 0.80, n 7). The FeO=(FeO
MgO) and Na2O=(Na2O CaO) ratios are lower for auriferous veins compared to
barren veins. Tourmaline grains from gold-enriched quartz-carbonate-tourmaline
veins (up to 56.7 ppm Au; DKR-75, Table 3) contain the highest amounts of SiO2,
Al2O3, and the lowest contents of TiO2, FeO, and Na2O. The data indicate a
decrease of total iron content (as FeO) and an increase of the Al2O3 content is
Table 3. Major and trace element contents of auriferous quartz-tourmaline veins of the Okote area, southern Ethiopia
DKR-31 DKR-36 DKR-56 DKR-58 DKR-75 KC-318 KC-328 B8-72 KC-347 KC-353 KC-368 KC-389 KC-390 KC-400 KC-401 KC-466
Al 1.12 0.11 1.29 0.17 0.26 5.01 0.05 0.16 0.05 0.12 0.09 0.05 0.07 1.88 0.74
Fe 0.80 0.60 1.40 0.20 0.60 4.90 0.10 0.77 1.60 0.20 2.10 0.40 0.50 0.10 1.60 1.80
Mn (ppm) 220 290 240 27 153 830 18 35 15 60 31 18 32 330 960
Mg 0.17 0.20 0.28 0.02 0.16 1.38 0.02 0.04 0.01 < 0.01 0.04 < 0.01 0.01 0.26 1.45
Ca 4.90 3.48 1.75 < 0.05 1.93 23.30 < 0.05 < 0.05 < 0.05 < 0.05 0.08 < 0.05 < 0.05 2.08 11.60
Na 0.21 < 0.01 0.16 < 0.01 0.01 0.06 < 0.01 0.13 < 0.01 < 0.01 0.01 < 0.01 < 0.01 < 0.01 0.01 0.04
K 0.08 < 0.01 0.04 0.04 < 0.01 0.14 < 0.01 < 0.01 < 0.01 < 0.01 < 0.01 0.01 < 0.01 < 0.01 0.02
V 17 13 33 5 6 108 4 12 3 8 8 6 4 49 52
Cr 20.0 16.0 19.0 37.0 61.0 15.0 44.0 55.6 17.0 15.0 43.0 43.0 24.0 68.0 28.0 39.0
Co 2.8 2.7 5.1 0.8 3.7 19.0 0.7 3.62 2.7 0.6 1.7 2.9 3.0 0.8 5.4 11
Ni 8 16 10 3 16 17 3 9.58 6 7 4 12 3 3 6 42
Cu 9.1 3.0 15 2.3 42 30 5.6 5.9 1.5 8.2 750 29 2.2 13 57
Zn 6 51 12 <3 6 52 <3 8.2 <3 <3 <3 7 <3 <3 10 17
Ga 2.1 0.3 3.5 0.3 0.5 9.9 0.1 0.4 < 0.1 0.5 0.2 0.2 0.1 4.1 1.5
As 0.3 < 0.5 < 0.10 0.42 < 0.4 0.2 3.3 < 0.3 < 0.6
Rb 2.2 0.6 1.2 0.8 < 0.5 4.7 < 0.5 1.3 < 0.5 < 0.5 < 0.5 < 0.5 < 0.5 < 0.5 < 0.5 0.8
Sr 67 13 46 4 4 200 <1 25 12 1 3 1 <1 2 61 24
Sb < 0.5 < 0.1 0.1 < 0.2 < 0.5
Ba 41 25 29 19 17 131 2 61 6 4 52 3 7 5 11 71
La 1.39 2.79 0.09 0.44 0.37 0.03 0.18 0.04 3.5
Ce 1.8 4.55 0.52 1.11 0.80 0.13 0.3 < 0.16 4.63
Nd 0.98 1.37 0.10 < 0.2 < 3.6 < 0.4 < 0.90 < 1.9 3.2
Sm 0.31 0.68 0.02 0.15 0.21 < 0.01 0.13 0.01 0.66
Eu 0.18 0.25 0.10 0.03 0.07 < 0.01 0.05 0.01 0.25
Gd 0.12 0.77 < 0.04 0.19 0.26 0.05 0.14 0.04 0.20
Tb 0.11 0.15 0.01 0.02 0.05 < 0.01 0.04 0.01 0.11
Tm 0.02 0.01 0.02 0.01 < 0.03 < 0.01 0.03 0.01 0.06
Yd 0.30 0.55 0.09 0.23 0.16 0.04 0.25 0.26 0.34
Lu 0.05 0.09 0.02 0.04 0.02 0.01 0.04 0.01 0.06
W < 0.3 < 0.3 0.16 0.45 0.13 < 0.2 0.08 13.9
Au (ppb) 1600 < 10 2380 80 56700 270 20 11530 180360 2490 8160 350 47450 50 80 580
Pb 2.4 0.8 1.2 1.1 0.9 3.5 0.5 3.2 1.1 2.2 0.8 0.3 0.2 0.8 0.9
Bi 0.4 < 0.1 0.1 < 0.1 0.2 0.2 < 0.1 0.1 2.2 4.6 0.1 0.2 < 0.1 < 0.1 0.2
Measured by ICP-MS and INAA; major element data in wt% and trace element data in ppm, except as noted
 D. J. Deksissa and C. Koeberl: Gold-quartz-tourmaline veins 113

Fig. 6. Interelement and ratio plots, showing compositional variation among tourmalines
from quartz-tourmaline veins (qv2) from different localities of the Okote area. Note that all
tourmaline samples have older cores with higher concentration of Al2O3, MgO, and SiO2,
but have lower FeO and Na2O=(Na2O CaO) than the rim

noticeable from the rims to the cores of the tourmalines. Correlation diagrams
show strong negative correlations between the contents of Al2O3 and FeO (Fig. 6a),
FeO and the SiO2 (Fig. 6b), and FeO and MgO (Fig. 6c). Generally, the cores of
most tourmalines grains show MgO enrichments and FeO depletions, in contrast to
their rims. The oxide ratios FeO=(FeO MgO) and Na2O=(Na2O=CaO) indicate
that there is no signicant compositional difference between the samples (Fig. 6d).
Different cation substitutions have been recognized in the Okote hydrothermal
tourmalines. The Fe and Mg contents of tourmaline show strong negative cor-
relations within each separate zone, with the rim having higher Fe than the core (e.g.,
DKR-52; Fig. 7a). The compositional variation is due to changes in uid chemistry
during tourmaline deposition. The strong negative correlation between Al and
divalent cations (Fe Mg Mn) (e.g., DKR-58; Fig. 7b) can be ascribed to an alkali
decient substitution (NaX (Fe Mg Mn)Y AlY X site vacancy), because of
a strong negative correlation observed between NaX and AlY (e.g., DKR-58; Fig. 7c),
114 D. J. Deksissa and C. Koeberl

Fig. 7. Intercation correlation diagram for 2 representative samples showing clear com-
positional variations between cores and rims of tourmaline crystals within a vein sample
(apfu atoms per formula unit)

but positive correlation observed between AlY and X site vacancy (e.g. DKR-58;
Fig. 7d). Tourmaline samples from auriferous quartz-carbonate-tourmaline veins
(e.g., DKR-31, DKR-56 and DKR-58) generally show Si and total Al enrichments at
the core of the crystals as compared to the rim (Table 1).
Trace element compositions of tourmaline separates are shown in Table 2. The
analytical results indicate low concentrations of most of the elements determined.
Nickel and Cr contents show strong variations. Tourmaline samples from talc-
tremolite host rock shows very high Cr and Sr but low Ni, Sc, and Zn. The gold
contents of the tourmaline crystals vary between 1.7 and 29 ppb. The highest value
is for tourmaline grains from the well-mineralized quartz vein DKR-31. The
 Gold-quartz-tourmaline veins of southern Ethiopia 115

Fig. 8. Chondrite-normalized rare earth element (REE) abundance patterns for tourmaline
separates. Open symbols samples: paragenetically older auriferous quartz-tourmaline veins
in epidote-amphibole-chlorite schist; solid symbols: samples are from paragenetically
younger auriferous quartz-tourmaline; normalization factors from Taylor and McLennan
(1985)

contents of the gold-pathnder elements Zr, La, Hf, Ta, W, As, and Sb are generally
very low. Granites are typically depleted in the elements Zr, Hf, and La due to their
high degrees of fractionation (e.g., Blevin and Chappell, 1995).
The rare earth element (REE) contents of the tourmaline crystals are very low
compared to chondritic abundances (Fig. 8). Chondrite-normalized REE abundances
show two different patterns. The rst group shows chondritic enriched LREEs, a
negative Eu anomaly, and a at and near-chondritic HREE pattern (solid symbols,
Fig. 8). The second group has subchondritic, but at, LREE pattern, a positive Eu
anomaly, and a at to slightly depleted HREE pattern (open symbols, Fig. 8).

Geochemistry of auriferous quartz-tourmaline veins

The bulk vein samples have very high gold contents of 10 ppb to 180 ppm and low
abundances of base metal elements, such as Cu, Co, Zn, Ni, and Pb (Table 3). Most
gold mineralized veins (i.e., Au > 500 ppb) are quartz-carbonate-tourmaline veins.
The copper contents of the veins are low except for a few samples that have
> 10 ppm. The Bi contents of the veins are also low, but  0.1 ppm Bi was detected
in samples containing high concentrations of gold. Generally, the analytical results
imply that gold is not associated with base metals.
The quartz-tourmaline vein samples display two different chondrite-normalized
REE patterns, both without an Eu-anomaly (Fig. 9). The rst group is enriched in
the LREEs, with a at, depleted chondrite-normalized HREE pattern (solid sym-
bols, Fig. 9). These veins contain signicant amounts of tourmaline and carbonates,
116 D. J. Deksissa and C. Koeberl

Fig. 9. Chondrite-normalized rare earth element (REE) abundance patterns for quartz-
tourmaline veins. Solid symbols are tourmaline-rich auriferous quartz veins and open sym-
bols are samples from tourmaline poor quartz veins; normalization factors from Taylor and
McLennan (1985)

except for sample KC-347. The second group has very low total REE abundances,
is LREE depleted and consists of quartz, gold, very low tourmaline, and hematite
and iron oxide (open symbols, Fig. 9). The REE patterns of quartz tourmaline veins
indicate LREE mobility during vein formation (see Hellman et al., 1979).

Discussion
Tourmaline is a complex borosilicate mineral consisting essentially of Mg, Fe, Al,
Na, Ca, B, Si, and O with the general formula (X)(Y3)(Z6)T6O18(BO3)3(W)4. In the
X position, Na may be partially replaced by K or Ca if the valance conditions are
satised. The Y site can be predominantly lled with Li , Mg2 , Fe2 , and Al3 ,
and minor Mn2 , Cr3 , V3 , Fe3 , Ti4 , Zn, Cu, and Ba. The Z position is
occupied mainly by Al3 , Fe3 , Cr3 , V3 , and some Mg2 and Fe2 . Si is as-
sumed to be exclusively in the T-site with any deciency made up by Al. The W
site consists of OH, F, and O (Henry and Dutrow, 1996). The crystal structure of
tourmaline allows the substitution of a variety of ions, depending on the size and
charge, in three lattice sites, leading to a wide variety in compositions.

Tourmaline-gold mineralization
The composition of hydrothermal tourmalinite is sensitive to uid chemistry,
physicochemical conditions, and nature of uid-rock interactions (uid=rock
ratios) (Garba, 1996). Comparisons between the composition of tourmaline in the
veins and their wall-rocks indicate active interaction between uids and wall-rocks.
 Gold-quartz-tourmaline veins of southern Ethiopia 117

A tourmaline sample from the talc-tremolite wall-rock has higher MgO, TiO2, SiO2
and lower Al2O3 contents than those from the chlorite-carbonate-biotite schist and
amphibole-carbonate-chlorite schist. This indicates that at least Mg is partitioned
into the hydrothermal uids in similar proportions to their relative abundances in
the host rocks. The pronounced zoning of tourmaline, with a Mg-rich core and Fe-
rich rim in most of the samples, indicates a variation in the composition of uids
during tourmaline deposition. During the early formation of the Mg-rich core, there
is the possibility of Fe being partitioned into an iron oxide, such as magnetite.
Magnetite is paragenetically older than pyrite in the host rocks of the veins
(Deksissa, 2000). The increase in the Fe contents in the rims of tourmalines may be
due to high concentrations of Fe in the hydrothermal solutions. This enrichment of
Fe is caused by increasing intensity of alteration and formation of chlorite after
ferro-actinolite in the wall-rock, replacement of magnetite by pyrite, and ankerite
by calcite veins (Deksissa, 2000). The occurrence of gold as inclusions in pyrite
and chalcopyrite has been reported by Deksissa (2000). The association of gold,
calcite, and second generation tourmaline in a quartz vein is shown in Fig. 5. These
observations suggest that gold deposition occurred during the growth of the iron-
rich tourmaline rims.
Nature and origin of uids responsible for tourmalinization: The occurrence of
tourmaline in the vein paragenesis in the Okote area indicates that boron is an
important constituent in the hydrothermal systems responsible for gold min-
eralization. The typical comb-like growth structures, in which the long axes of
tourmaline are oriented perpendicular to fracture surfaces in the quartz-tourmaline
veins, indicates that the inux of boron-rich uids is slightly younger than the
uids responsible for deposition of early stage qv2 veins. The laminar textures
of the quartz-tourmaline veins suggest the presence of alternating shearing and
hydrothermal uid inltration events within a major tectonic episode. The
concentration of pyrite, carbonate (calcite and relict ankerite), and tourmaline in
the gold-quartz-tourmaline veins indicates that S-bearing species, such as H2S, as
well as CO2 and B, were signicant volatile components in the mineralizing uids.
Similar mineral associations have been reported from hydrothermal alteration
zones at Bin Yauri, Nigeria (Garba, 1996). Ethier and Campbell (1977) indicate
that tourmaline is stable in weakly acid to moderately alkaline solutions, but that it
is unstable in strongly alkaline solutions. The typical schorl-dravite composition of
tourmaline at the Okote area suggests that they were formed from hydrothermal
uids within the temperature range mentioned below. Temperatures calculated
based on AlIV contents in the tetrahedral site of chlorite, according to the
regression equation of Cathelineau (1988), range from 323 to 411 C (mean
366  37  C, n 80 crystals) (Deksissa, 2000). The metasomatic mineral assem-
blages (for example, chlorite, carbonate, biotite, epidote), which are typical of
greenschist facies, suggest an equilibrium temperature within the stability eld of
biotite ( 400  C). The presence of tourmaline in the gold carbonate paragenesis
suggests a likely pH range of 6 to 8 for the mineralizing uids.
Source of boron and gold: The origin of hydrothermal B in mac-ultramac
rocks is enigmatic, as these rocks have very low B contents. Unaltered mac igneous
rocks (mostly oceanic crust) contain < 10 ppm B (Chaussidon and Jambon, 1994).
Basaltic and ultramac components of oceanic lithosphere are known to contain 80
118 D. J. Deksissa and C. Koeberl

to 110 ppm B, following hydrothermal alteration at mid ocean ridges (Mottl and
Holland, 1978). On the other hand, terrestrial sediments have about 100 ppm B,
whereas values for pelagic (marine) clays are mostly in the range of 100 to
1000 ppm. Marine evaporites represent potential sources of concentrated boron with
representative values of 100 to 2000 ppm (London et al., 1996). Granitic magmas,
especially those derived from meta-sedimentary sources, are essential in the
redistribution of B from deep to shallow levels of the Earth's crust (London et al.,
1996).
The occurrence of various lithologic units and structural complexities, such as
thrusting followed by intensive shearing and hydrothermal alteration, makes the
source of the uid and gold more complex. Fluid inclusion studies of samples
DKR-31 and DKR-36 indicate that the uid is of low salinity (maximum 4.94 wt%
NaCl equivalent), and H2OCO2 uid. The uid inclusions are generally two phase
liquid=vapor inclusions that homogenize into liquid and vapor phase. Trapping
temperatures of the uids range from 218 to 300  C (Deksissa, 2000). Dewatering
of the protolith during progressive metamorphism can effectively remove boron
from the meta-sedimentary rocks (Moran et al., 1992).
The mean FeO=(FeO MgO) ratios of tourmalines from the Okote auriferous
quartz-tourmaline veins range from 0.41 to 0.56 (mean 0.47, n 7 samples)
(Fig. 10). The FeO=(FeO MgO) ratio of tourmaline has been suggested as an
indictor of source environments. Ethier and Campbell (1977) compiled FeO=
(FeO MgO) ratios (0.41 to 0.67) of tourmaline from meta-sedimentary environ-
ments. Granitic tourmaline and tourmalines associated with sediment-hosted
massive sulde deposits are 0.86 to 0.96 and 0.21, respectively (Taylor and Slack,
1984). Pirajino and Smithies (1992) have shown that the FeO=(FeO MgO) ratio
of tourmalines associated with granite-related Sn and Sn-W hydrothermal
mineralization in South Africa, Namibia, and New Zealand vary from < 0.6 to
> 0.8, depending on the distance from the granitic source. They proposed that
FeO=(FeO MgO) ratios < 0.6 indicate deposition environments distal to the
granite source. Tourmalines in metamorphosed and hydrothermally altered mac-
ultramac rocks tend to be dravitic, but can have a signicant schorl component
(Slack, 1996). Comparisons of our data with literature values indicate that the
FeO=(FeO MgO) ratios do not discriminate between distal granitic and meta-
sedimentary source rocks. Based on uid inclusion data, Deksissa (2000) suggested
a metamorphic origin for the B-bearing uids.
The strong association of tourmaline and carbonate in quartz veins, with
intense wall-rock alteration, indicates that tourmaline has crystallized from
hydrothermal uids. However, the sources of the hydrothermal uids are still
unknown. The plot of ternary (FeCaMg and FeAlMg) discriminant diagrams
(Fig. 10) favor meta-sedimentary rocks as the main source. The mean of
carbon (13C) and oxygen (18O) stable isotope ratios of 6 samples from cogenetic
calcite veins are 7.1  2.05 and 8.1  1.19, respectively (Deksissa, 2000),
indicating either magmatic melts or metamorphic dehydration processes as the
most probable sources of the hydrothermal uids. The mean sulfur isotope ratio
(34S) of pyrite from the hydrothermal alteration zones associated with quartz veins
is 2.68  0.98 (n 10), suggesting a signicant contribution of magmatic sulfur
(Deksissa, 2000).
 Gold-quartz-tourmaline veins of southern Ethiopia 119

Fig. 10. Classication of chemical composition of the quartz-tourmaline veins of the Okote
area, southern Ethiopia, using ternary diagrams of Henry and Guidotti (1985). Diagram
a AlFe(tot)Mg diagram. Fields 1 Li-rich granitoids, pegmatites, and aplites; 2 Li-
poor granitoids and their pegmatites and aplites; 3 Fe3 -rich quartz-tourmaline rocks
(hydrothermally altered granites); 4 Metapelites and metapsammites with Al-saturated
phase; 5 Metapelites and meta-psammites without Al-saturated phase; 6 Fe3 -rich
quartz-tourmaline rocks, calc-silicate rocks, and metapelites; 7 Low Ca meta-ultramac
and Cr-, V-rich meta-sediments; 8 Metacarbonates and metapyroxenites. b CaFe(tot)
Mg diagram: Fields 1 Li-rich granitoids, pegmatites, and aplites; 2 Li-poor granitoids
and their pegmatites and aplites; 3 Ca-rich metapelites, metapsammites, and calc-silicate
rocks; 4 Ca-poor metapelites, metapsammites, and calc-silicate rocks; 5 metacarbo-
nates; 6 meta-ultramacs

The occurrence of tourmaline associated with pervasive hydrothermal

alteration (Deksissa, 2000) in mac-ultramac rocks of the study area can be
related to the tectonic evolution of the Adola Belt (Berhe, 1990; Worku and
120 D. J. Deksissa and C. Koeberl

Schandelmeier, 1996). The generation of hydrothermal boron and associated gold

involves several crustal processes, such as ocean oor spreading and deposition of
boron-rich sediments, followed by subduction, greenschist to amphibolite facies
metamorphism, and intrusion of granites (Berhe, 1990; Woldehaimanot and
Behrmann, 1995; Worku and Schandelmeier, 1996). Finally, NS trending shear
zones created favorable conditions for deep-seated hydrothermal uids to leach
boron and gold from wall-rocks and to precipitate them in veins (Deksissa, 2000).

Metallogenic and exploration signicance

The very high gold contents of some quartz-carbonate-tourmaline veins and
textural relationships among gold, calcite, and tourmaline in the veins indicate that
tourmalinization and gold mineralization are genetically linked. Signicant
tourmaline contents are known to be associated with mesothermal, shear zone-
hosted, Archean gold deposits of the Kolar Gold elds of India, and Au-quartz
veins of the Superior Province of Ontario (Canada) and the Kaapvaal Craton of
South Africa (Slack, 1996, and references therein). Tourmaline is also a local
gangue mineral in the Proterozoic Au-quartz vein from Ondonoc, W. Ethiopia
(Augustithis, 1967). The giant Muruntau ore body in Uzbekistan, possibly the
largest hydrothermal gold deposit in the world, consists of Au-bearing quartz veins
and minor to abundant tourmaline (Slack, 1996, and references therein).
Tourmalinites may in some cases be interpreted as facies equivalents of
exhalative sulde deposits and banded iron formations (Plimer, 1988), and their
recognition should stimulate exploration interest in any area in which they occur.
The high Mg-contents of the tourmalines are consistent with hydrothermal
leaching of the mac-ultramac host rock, but they do not exclude the possibility
that some of the hydrothermal uids were derived from granitoid rocks.

Conclusions
The following points are concluded from the present study:
The mineralized zones of the Okote area were produced during intense alteration
of mac-ultramac rocks by interaction with B-, CO2-, and S-bearing hydro-
thermal uids that were channeled along high strain zones.
Petrographic and geochemical data of auriferous quartz-tourmaline veins show
that gold and tourmaline are genetically related.
The strong correlation between the gold contents and the abundance of Mg-rich
tourmalines in carbonate-rich quartz veins makes this mineral assemblage an
important prospecting guide for gold exploration in the Adola Belt of southern
Ethiopia.

Acknowledgments
This work is part of a Ph.D. project by the rst author, which was funded by the Austrian

Academic Exchange Service (Osterreichischer 
Akademischer Austauschdienst, OAD).
Laboratory expenses were covered by the Austrian Science Foundation Y58-GEO (to
C. Koeberl). The authors also thank the National Mining Corporation, Dawa-Digati gold
 Gold-quartz-tourmaline veins of southern Ethiopia 121

exploration project, for admittance to their project area, and for providing support during
the eld work to D. J. Deksissa, who also acknowledges the project geologists in general
and S. Demisse, T. Tadese, Y. Edossa, and A. Adugna in particular, for their stimulating
discussions during the eld work. The authors would also like to thank J. Newton (Univ.
Vienna) for help with stable isotope analyses, and T. Ntaos (Univ. Vienna) for his help
with the electron microprobe analyses. We further thank P. Spindler and H. Froschl
(Arsenal, Vienna) for their help with ICP-MS analyses.

References
Augustithis SS (1967) On the textures and paragenesis of the gold-quartz-tourmaline veins
of Ondonoc, W-Ethiopia. Mineral Deposita 3: 4855
Berhe SM (1990) Ophiolite in northeast and east Africa: implication for Proterozoic crustal
growth. J Geol Soc Lond 147: 4157
Billay AY, Kisters AFM, Meyer FM, Schneider J (1997) The geology of Lega Dembi gold
deposit, southern Ethiopia: implication for Pan-African gold exploration. Mineral
Deposita 32: 491504
Blevin PL, Chappell BW (1995) Chemistry, origin and evolution of mineralized granites in
the Lachlan Fold Belt, Australia: the metallogeny of I- and S-type granites. Econ Geol
90: 16041619
Bone Y (1988) The geological setting of tourmaline at Rum Jungle, NT, Australia genetic
and economic implications. Mineral Deposita 23: 3441
Byerly GR, Palmer MR (1991) Tourmaline mineralization in Barberton greenstone belt,
South Africa: early Archean metamorphism by evaporite-derived boron. Contrib
Mineral Petrol 107: 387402
Cathelineau M (1988) Cation site occupancy in chlorites and illites as a function of
temperature. Clay Mineral 23: 471485
Chaussidon M, Jambon A (1994) Boron content and isotopic composition of oceanic
basalt: geochemical and cosmochemical implications. Earth Planet Sci Lett 121:
277291
Deksissa D (2000) Geology, geochemistry, and genesis of gold mineralization of the Okote
area, southern Ethiopia. Thesis, University of Vienna, 262 pp (unpublished)
Ethier VG, Campbell FA (1977) Tourmaline concentrations in Proterozoic sediments of the
southern Cordillera of Canada and their economic signicance. Can J Earth Sci 14:
23482363
Fleischer R, Routhier P (1973) The consanguinous origin of a tourmaline-bearing gold
deposit: Pasagem de Mariana (Brazil). Econ Geol 68: 1122
Garba I (1996) Tourmalinization related to Late Proterozoic-Early Paleozoic lode gold
mineralization in the Bin Yauri area, Nigeria. Mineral Deposita 31: 201209
Groves DI, Foster RP (1991) Archean lode gold deposits In: Foster RP (ed) Gold
metallogeny and exploration. Blackie, Glasgow London, pp 104132
Hellman PL, Smith RE, Henderson P (1979) The mobility of rare earth elements: evidence
and implication from selected terrains affected by burial metamorphism. Contrib
Mineral Petrol 71: 2344
Henry DJ, Dutrow BL (1996) Metamorphic tourmaline and its petrologic implications. In:
Grew ES, Anovitz LM (eds) Boron: mineralogy, petrology, and geochemistry. Rev
Mineral 33: 503557
Henry DJ, Guidotti CV (1985) Tourmalines as a petrogenetic indicator mineral: an example
from the staurolite-grade metapelites of NW Maine. Am Mineral 70: 115
Kazmin V (1972) Geology of Ethiopia (explanatory note to geological map of Ethiopia
1:2000,000). Geol Surv Ethiopia, Minstry Mines, Addis Ababa, Ethiopia 211 p
122 D. J. Deksissa and C. Koeberl: Gold-quartz-tourmaline veins

King RW (1988) Geochemical characteristics of tourmaline from Superior Province Archean

lode gold deposits: implications for source regions and processes. In: Bicentennial Gold
`88. Geol Soc Austr Abstract Series Vol 2: 445447
Koeberl C (1993) Instrumental neutron activation analysis of geochemical and cosmo-
chemical samples: a fast and reliable method for small sample analysis. J Radioanal Nucl
Chem 168: 4760
London D, Morgan VI GI, Wolf MD (1996) Boron in granitic rocks and their contact aureoles.
In: Grew ES, Anovitz LM (eds) Boron: mineralogy, petrology, and geochemistry. Rev
Mineral 33: 299330
McArdle P, Fitzell M, Oosterom MG, O'Connor PJ, Kennan PS (1989) Tourmaline as a
potential host rock for gold in the Caledonides of southeast Ireland. Mineral Deposita
24: 154159
Moran AE, Sisson VB, Leeman WP (1992) Boron depletion during progressive meta-
morphism: implication for subduction processes. Earth Planet Sci Lett 111: 331349
Mohammed BH (1999) The geology, structure and geochemistry of the crystalline rocks of
the Moyale area, southern Ethiopia: implication for the tectogenesis of the Precambrian
basement. Tubinger Geowiss Arbeiten, Reihe A, 50: 102 pp
Mottl MJ, Holland HD (1978) Chemical exchange during hydrothermal alteration of basalt
by sea water. I. Experimental results for major and minor components of sea water.
Geochim Cosmochim Acta 42: 11031115
Nesbitt BE (1991) Phanerozoic gold deposits in the tectonically active continental margins.
In: Foster RP (ed) Gold metallogeny and exploration. Blackie, Glasgow London,
pp 104132
Pirajino F, Smithies RH (1992) The FeO=(FeO MgO) ratio of tourmaline: a useful
indicator of spatial variations in granite-related hydrothermal mineral deposits. J Geo-
chem Explor 42: 371381
Plimer IR (1986) Tourmaline from Golden Dyke Dome Northern Australia. Mineral
Deposita 21: 263270
Plimer IR (1988) Tourmalinites associated with Australian Proterozoic submarine
exhalative ores. In: Friedrich GH, Herzig PM (eds) Base metal sulde deposits in
sedimentary and volcanic environments. Springer, Berlin Heidelberg New York Tokyo,
pp 255283
Slack JF (1996) Tourmaline association with hydrothermal ore deposits In: Grew ES,
Anovitz LM (eds) Boron: mineralogy, petrology, and geochemistry. Rev Mineral 33:
559643
Taylor BE, Slack JF (1984) Tourmaline from Appalachian-Caledonian massive sulde
deposits: textural, chemical, and isotopic relationships. Econ Geol 79: 17031726
Taylor SR, McLennan SM (1985) The continental crust: its composition and evolution.
Blackwell, Oxford, 312 pp
Vail JR (1987) Late Proterozoic tectonic terrain in the Arabian-Nubian shield and their
characteristic mineralization. Geol J Winter Them Iss 22: 161174
Woldehaimanot B, Behrmann JH (1995) A study of metabasite and meta-granite chemistry
in the Adola region (south Ethiopia): implication for the evolution of the East African
Orogen. J Afr Earth Sci 21: 459476
Worku H, Schandelmeier H (1996) Tectonic evolution of the Neoproterozoic Adola Belt of
southern Ethiopia: evidence for a Wilson Cycle process and implications for oblique
plate collision. Precamb Res 77: 179210

Authors' address: D. J. Deksissa and C. Koeberl (corresponding author), Institute of Geo-

chemistry, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria, e-mail:
[email protected]