Economic Geology
Vol. 95, 2000, pp. 429–446
Genesis of Vein Stockwork and Sedimentary Magnesite and Hydromagnesite Deposits
in the Ultramafic Terranes of Southwestern Turkey: A Stable Isotope Study
VEYSEL ZEDEF,* MICHAEL J. RUSSELL,† ANTHONY E. FALLICK,
Isotope Geosciences Unit, Scottish Universities Research and Reactor Centre, East Kilbride, Glasgow G75 OQF, Scotland
AND
ALLAN J. HALL
Department of Archaeology, University of Glasgow, Glasgow G12 8QQ, Scotland
Abstract
Vein stockworks and lacustrine developments of cryptocrystalline magnesium carbonates of Neogene and
Quaternary age occur within the partially serpentinized, discontinuous ultramafic belts of southwestern Turkey.
They are comparable to the Neogene cryptocrystalline magnesite bodies elsewhere in the Alpine orogen to the
northwest and southeast. Our previous work (Fallick et al., 1991) suggested that cool (≤100°C) modified meteoric water was the mineralizer, that ultramafic rock was the source of the magnesium, but that there were
three separate sources of the (bi)carbonate. These sources were distinguishable by their stable isotope composition as follows: (1) low-temperature carbonate with δ18O(SMOW) values of ~36 per mil and δ13C(PDB) values
of ~4 per mil, derived from atmospheric CO2; (2) moderate-temperature carbonate with δ18O(SMOW) values of
+28 per mil and δ13C(PDB) values of –15 per mil, derived by decarboxylation of organic-rich sediments; and (3)
higher temperature carbonate with δ18O(SMOW) values of ~19 per mil and δ13C(PDB) values of ~3 per mil, assumed to have been generated by thermal contact metamorphism of Paleozoic marine limestone at depth. In
general these magnesite deposits were found to fall into two groups, comprising carbonate generated on two
mixing lines. The first group spanned the putative mixing line from the “atmospheric” source (1) to “organically derived” source of CO2 (2). The second group extended between atmospheric source (1) and the “thermal” source (3), although there were concentrations either around the atmospheric end, or precisely at the contact metamorphic end of the line.
In the present study we found that large stockwork deposits at Helvacıbaba and Koyakcı Tepe have δ13C(PDB)
and δ18O(SMOW) values averaging ~–12 and ~+27 per mil, respectively, indicating a derivation mainly by oxidation of organic-rich metasediments perhaps underthrust at depth (end-member 2), with some involvement of
atmospheric carbon dioxide as bicarbonate in the circulating, hot, and modified meteoric water (end-member
1). Calcite veinlets in a meta-argillite of the Cambro-Ordovician Seydisehir Formation, most likely to have been
underthrust beneath the stockworks, yielded δ13C(PDB) values of –20 per mil, consistent with, though not proving, oxidized organic carbon being one of the sources of carbonate. The δ18O(SMOW) values of these same veinlet carbonates are also rather low (22‰), indicating precipitation from heated ground water, though their age
is unknown.
The major stratiform magnesite deposit at Hirsizdere in the center of the Menderes graben has δ13C(PDB)
and δ18O(SMOW) values averaging ~3 and ~25 per mil, respectively, and thus appears to be an example of the hydrothermal-sedimentary (i.e., exhalative) type (Ilich, 1968). In contrast, the hydromagnesite stromatolites
presently growing in Salda Gölü (Lake Salda) are apparently developing at cool ground-water seepages. The
gross morphology of the Salda Gölü stromatolites and the hydromagnesite sediments derived therefrom is reminiscent of that revealed in the Bela Stena magnesite pit in Serbia. These lacustrine deposits have mean
δ13C(PDB) values of ~4 and ~2 per mil and mean δ18O(SMOW) values of ~36 and ~33 per mil, respectively, i.e.,
they both plot broadly over the atmospheric CO2-meteoric water field (end-member 1), consistent with microbially mediated precipitation at cool ground-water seepages in enclosed evaporating lakes.
Introduction
A NUMBER of explanations have been mooted for the genesis
of vein, stockwork, and sedimentary developments of cryptocrystalline magnesite in ultramafic terranes. Although it is
generally agreed that the source of the magnesium is the ultramafic rocks themselves (e.g., Barnes and O’Neil, 1969;
Barnes et al., 1973, 1978; Dabitzias 1980, 1981; Burgath et
al., 1981), the sources of the carbonate are disputed. The following various origins have been suggested: (1) atmospheric
carbon dioxide (O’Neil and Barnes, 1971; Fallick et al., 1991),
(2) decarboxylation of organic-rich sediments (Fallick et al.,
† Corresponding
author: email, [email protected]
* Present address: Selçuk Üniversitesi, Mühendislik-Mimarlık Fakültesi,
Maden Bölümü, Kampüs 42040 Konya, Turkey.
0361-0128/00/3041/429-18 $6.00
1991; Brydie et al., 1993), (3) thermal decarbonation of limestone (Fallick et al., 1991), (4) soil-derived decomposition of
plant material and surface weathering (Petrov, 1967; Lesko,
1972; Zachmann and Johannes, 1989), (5) regional metamorphic reactions at >300°C (Abu-Jaber and Kimberley, 1992),
(6) volcanogenic sources (Ilich, 1968), (7) a deep-seated
source (cf. Kreulen, 1980); or combinations of the above.
In our own previous contributions to the cryptocrystalline
magnesite problem we concluded tentatively that combinations of three dominant sources of carbonate (1–3) could explain the isotopic make-up of the various massive cryptocrystalline magnesite deposits in Serbia, Bosnia, and Cyprus
(Fallick et al., 1991; Brydie et al., 1993).
The present study of magnesite bodies in southwestern
Turkey gives us the opportunity to extend our knowledge of
429
430
ZEDEF ET AL.
the types of cryptocrystalline magnesites characteristic of
Alpine-type ultramafic complexes and to assess our models
for their origin. As in our earlier studies we find that, on a plot
of δ13C(PDB) vs. δ18O(SMOW), each deposit occupies a restricted
field. In general the hypotheses survive the tests, though oxidation of organic carbon in underthrust metasediments may
be an important mechanism of carbon dioxide generation
(Shock, 1988) rather than decarboxylation. Moreover, CO2
volatilization, evaporation, and cyanobacterial depletion of
12C are complicating factors.
Geologic Environment
The Turkish magnesite deposits are hosted by extensive
Alpine-type ultramafic rock outcrops and subcrops that comprise major ophiolite complexes. These Mesozoic peridotites
and serpentinites are associated with pillow lavas, Upper Cretaceous radiolarian cherts and limestones, Cenozoic limestones, and recent alluvium (Sarp, 1976, Ketin, 1983). The ultramafic rocks were extensively disturbed by Alpine orogenic
movements (Gedik, 1988). According to Erdem (1974), Sengör and Yılmaz (1983), and Collins and Robertson (1998), the
ultramafic rocks were episodically emplaced in their presentday juxtapositions during the mid- and Late Cretaceous to the
late Miocene. The obduction was a consequence of the partial
closure of the Tethys ocean. Serpentinization of the ultramafic
rocks in fault zones and along contacts with country rocks may
have taken place before, during, and/or following emplacement by thrusting over Paleozoic and Mesozoic formations.
Extensional stresses have been operating in western Turkey
since the Oligocene (Koçyigit 1984). These stresses resulted
in the development of several rifts, such as the Büyük
Menderes and Gediz grabens (Fig. 1; Cohen et al., 1995), and
encouraged volcanism (Seyitoglu and Scott, 1991). Kissel et
al. (1993) reported that part of western Turkey (the Isparta
region) has been rotated clockwise up to 40° since the Paleocene. Western Turkey is still seismically active (Jackson and
McKenzie, 1984; Koçyigit, 1984; Livermore and Smith, 1985;
Udias, 1985).
Neotectonic activity results in the development of many
springs, both cool and hot. Percolating water in a continent
will generally have a meteoric origin (Garven, 1995). Large
chemically deposited fresh-water carbonate deposits clearly
require a meteoric water recharge (Fallick et al., 1991). The
drive for the aqueous flow can be either artesian or thermal.
In the latter case, heating of meteoric water takes place by interaction with hot rocks at depth, by exothermic reactions
(Haack and Zimmermann 1996), and/or magmatic intrusions.
In such a tectonically active area as western Turkey, the geothermal gradient would be expected to approach 35° to
40°C/km in places (Tezcan, 1979), that is, higher than the
usual 25° to 30°C/km.
Fig. 1. Location map of carbonate deposits in southwestern Turkey, showing main faults and ultramafic bodies. Many present warm to hot springs and their deposits occur on the northern margin of the Büyük Menderes graben. Examples are the
carbonate terraces currently developing over the warm springs at Pamukkale, the older travertines at Kocabas, 20 km to the
southeast (shown in Fig. 4), and the geothermal plants at Kızıldere, Salavatlı, and Germencik. The outline to Figures 2 and
4 are indicated.
0361-0128/98/000/000-00 $6.00
430
431
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
Cryptocrystalline magnesium carbonate deposits are especially concentrated in southwestern Turkey where late Cenozoic extensional tectonic regimes dominate (Fig. 1). They fall
into the following two categories: vein stockwork and lacustrine deposits.
Descriptions
Vein stockwork deposits
The largest of the cryptocrystalline magnesite deposits is
the presently exploited Helvacıbaba stockwork. It occurs in
the serpentinized Çayırbagı ophiolite in the Konya district
(Figs. 2 and 3, Table 1; Ozcan et al., 1988, Güleç, 1991). Several magnesite pebbles were discovered in the conglomerate
at the base of the lacustrine sequence overlying the stockwork.
TABLE 1. Cryptocrystalline Vein Stockworks
Deposit
Host
Style
Alteration
Helvacıbaba
Serpentinite
Stockwork
Koyakcı Tepe
Serpentinite
Arapömer
Deresi
Serpentinite
Vein
stockwork
Vein
stockwork
Carbonate
and silica
Carbonate
~50
~40
~0.1 ?
Above this conglomerate dolomite beds occur in what are
otherwise calcareous Neogene lake sediments in this region
(Fig. 3).
Fig. 2. Geologic map of the Meram-Çayırbagı area, Konya (Altunel, 1963; Zedef, 1994).
0361-0128/98/000/000-00 $6.00
Carbonate
and silica
Tonnage
(Mt)
431
432
ZEDEF ET AL.
Fig. 3. Stratigraphic section for the Konya district.
The Koyakcı Tepe stockwork, 10 km to the south-southwest
of Helvacıbaba, is almost as large (Fig. 2). Although both
stockworks comprise a myriad of centimeter-sized veins and
veinlets, in the latter some individual west-northwest veins
reach a width of 4 m. The host rocks are carbonated around,
and silicified above, both deposits.
Lacustrine deposits
Cryptocrystalline lacustrine deposits comprise the remainder of the magnesite or hydromagnesite resources. The
largest of these is Hırsızdere, which lies in the Büyük
Menderes rift, 7 km southwest of Bozkurt (Ilich 1974; Figs. 1
and 4). It is a working deposit comprising five magnesite beds
in Pliocene lacustrine sediments. The thickest and purest
magnesite bed lies directly upon carbonated serpentinites
and incorporates pebbles of this footwall at its base.
The other large lacustrine deposit is still forming in Salda
Gölü (Sarp, 1976; Çoban et al., 1995). It is composed of hydromagnesite. Although evaporation increases concentration
to supersaturation, precipitation is brought about by microbes
(Russell, 1993; Braithwaite and Zedef, 1994, 1996a, b: pace
Schmid, 1987; Figs. 4 and 5). Fine particles derived from the
erosion of the mounds are deposited between them to comprise fine, water-saturated muds (Fig. 6a).
The submerged portions of the otherwise white microbial
hydromagnesite mounds are covered with light green, slimy
cyanobacterial filaments (see underwater photographs in
Beköz et al., 1997). An extensive outcrop (>200 × 100 × 1 m
thick) of what appears to be beach deposits of hydromagnesite with layers of lizardite pebbles, trenched by the side of
the road in 1997, leads into Salda Village from the southeast.
0361-0128/98/000/000-00 $6.00
This exposure, now filled in, lies approximately 30 m above
the present lake level and about 2.5 km west-southwest of the
living microbialites (Fig. 4). Although best developed at Kocaadalar Burnu, the fossil to subfossil hydromagnesite microbialites occur discontinuously all around the lake (Fig. 5a, e,
f). The largest growths are developed where valleys, dry in
summer months, meet the shore.
As the presently forming hydromagnesite in Salda Gölü
may provide the key to understanding some ancient lacustrine magnesite deposits, we examined the mechanism of
generation and deposition in some detail.
Salda Gölü is in a region of tectonic extension (Sarp, 1976;
Kastens, 1991). This is a suitable environment for the development of both artesian flow and thermally driven hydrothermal convection. Although the area with the largest geothermal energy potential in Turkey is situated on the north
side of the Menderes graben, at Kızıldere, ~90 km northwest
of Salda Gölü, and there is evidence of Plio-Pleistocene hydrothermal activity just 10 km north-northwest as well as 20
km north of Salda Gölü, there are no signs of recent hotspring activity within the immediate surroundings of the lake.
The only veins we found consisted of lizardite. Thus, we favor
the topographic drive of meteoric water through ultramafic
coarse alluvium as an explanation for the Salda Gölü hydromagnesite deposit, as suggested by Braithwaite and Zedef
(1994, 1996b; cf. Lur’ye and Gablina 1972). A geochemical
model is outlined below.
Cool (~10°C) meteoric fluids, presumably charged with atmospheric and soil-generated CO2, flow along the base of the
Salda River (Karakova Dere) and other smaller channels.
There they react with lizardite (our X-ray diffraction analyses;
eq 1) and harzburgite (eq 2) to dissolve Mg++ (Lesko, 1972):
Mg3Si2O5(OH)4 + H2O + 6CO2 →
3Mg2+ + 2SiO2 + 6HCO3– , and
(1)
10(Mg0.8 Fe0.2) 2SiO4 + 16H2O + 32CO2 + O2 →
16Mg2+ + 10SiO2 + 2Fe2O3 + 32HCO3–.
(2)
The springs, bearing an average of 60 ppm Mg, emanate
from just below the waterline. Presumably the springs also
bear nutrients to the cyanobacteria colonizing the mixing
zone. There are no visible outlets and the water level of the
lake drops a meter or so over the summer. It has dropped several meters in living memory (Ahmet Koç, pers. commun.,
1992) and about 30 m since deposition of the first hydromagnesite beach deposits mentioned above (Russell et al., 1999).
As the lake lies at the head of the regional watershed, some
loss through subsurface drainage channels, particularly
through the Upper Cretaceous limestones comprising a sector of the east shore and bottom of the lake, is likely. Nevertheless, the forty-fold enrichment of Na (200 ppm), compared to the contributing cool springs (5 ppm), attests to the
importance of evaporation (Russell et al., 1999). Also, evaporation of the lake water has enriched the magnesium content
beyond the saturation point of both magnesite and hydromagnesite. Yet the Mg in the lake has only increased by a factor of five or so, from 60 to 300 ppm (Russell et al., 1999; fig.
10b). The missing Mg has been precipitated in hydromagnesite. The precipitation of the supersaturated hydromagnesite
432
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
433
Fig. 4. Simplified geologic map of the area around Salda Lake. Karakova Dere (stream) drains into Salda Lake from the
southwest. Surface flow takes place only after rain. Note that the Arapömer Deresi deposit lies in a separate watershed that
drains to the north. Also shown is the small development of hydromagnesite in Akgöl and the present and past (Pamukkale
and Kocabas) travertines developed on the northern margin of the Menderes graben.
is brought about by cyanobacteria (Russell, 1993) and diatoms (Braithwaite and Zedef, 1994, 1996b), perhaps aided
by associated anaerobic heterotrophs deeper in the mounds.
Riding (1992) notes that cyanobacteria prefer alkaline conditions and that they precipitate carbonate in warm freshwater
environments. Deposition is linked with the photosynthetic
fixation of carbon from bicarbonate by the cyanobacteria
(Burne and Moore, 1987) and the emission of hydroxyl
(Thompson and Ferris, 1990):
hv
HCO3– + H2O → CH2O + O2 + OH –,
0361-0128/98/000/000-00 $6.00
(3)
where hv represents a solar photon which energizes the reaction. The magnesium ion is both small and highly charged,
i.e., it has high ionic potential. It is, therefore, strongly hydrated (Christ and Hostetler, 1970; Lippmann, 1973). This
hydration not only allows supersaturation in aqueous bicarbonate fluids, it also leads to the precipitation of hydromagnesite rather than magnesite at low to medium temperatures
(Sayles and Fyfe 1973), unless carbonate far exceeds magnesium concentration. In the latter case the CO2–
3 might displace
the water dipoles at the crystal surface to become directly
bonded to Mg++, effecting the direct precipitation of magnesite (Lippmann, 1973). But at normal atmospheric pressure
433
434
ZEDEF ET AL.
a
b
c
d
e
f
Fig. 5. Hydromagnesite developments in Salda Gölü. a. Northwesterly prospect across the lake showing hydromagnesite
developments on the beaches which otherwise comprise lizardite and harzburgite pebbles. b. View from east-southeast
across Kocaadalar Burnu (peninsular) comprised of earlier hydromagnesite developments, beyond is the Karakova Dere estuary, and then the ultramafic surroundings; to the north of the peninsular are the presently growing microbial mounds. c.
Residual hydromagnesite stromatolites on Kocaadalar Burnu having the appearance of dunes. d. Recently exposed hydromagnesite microbialites on the islands pictured in (b). The cauliflowerlike structures pictured here are rising 3 m above the
lake surface. e. Looking down on recently emergent microbialites (note centimeter scale). f. Close-up view of partially eroded
stromatolites.
0361-0128/98/000/000-00 $6.00
434
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
a
b
c
d
435
Fig. 6. Contrasting textures of the hydromagnesite developments in Salda Gölü. a. A fine hydromagnesite mud comprises
the floor of a small lagoon partly enclosed by microbialites. b. Highlighted contacts between Recent stromatolitic material
and intervening semilithified hydromagnesite mud in the Kocaadalar Burnu cliff. c. Close-up of (b). d. Partially lithified hydromagnesite mud revealing the incipient conchoidal fracture typical of the cryptocrystalline magnesites in the Alpine ophiolitic terrane. Scale bar is in centimeters.
hydromagnesite precipitates from supersaturated solutions,
thus:
–
5Mg2+ + 4CO2–
3 + 2OH + 4H2O →
Mg5 (CO3)4 (OH) 2 .4H2O.
(4)
The carboxylate groups embodied in the polysaccharide
sheaths of the cyanobacteria could provide the initial epitaxial nucleation sites for the precipitation of the magnesium
ions in the waters seeping through the mounds (cf. Pentecost,
1978, 1995; Thompson and Ferris, 1990; Mann et al., 1993;
and see Möller, 1989). In this case Mg2+ acts as the bridging
cation from the carboxyl to the carbonate, the counter anion,
in the mucilaginous exterior of the cyanobacterial mounds:
R.COO– + Mg2+ + HCO–3 → R.COOMgHCO3.
(5)
Following nucleation, continued deposition is encouraged
by the increase in pH in the periplasm (eqs 3 and 4).
The Geochemist’s Workbench program REACT (Bethke,
1996) was used to examine the relationship between magnesium
0361-0128/98/000/000-00 $6.00
concentration and magnesite supersaturation (Fig. 7). When
water is in equilibrium with atmospheric CO2, oversaturated
in magnesite, and precipitating hydromagnesite, the computed pH is about 9 and the concentration of Mg2+ is only
about 80 ppm. Higher concentrations require lower pH values at higher CO2 fugacities unless the solutions are supersaturated. Salda Gölü contains nearly four times this concentration of magnesium, yet has a pH of around 9 (Russell et al.,
1999). Thus the lake water is also supersaturated in hydromagnesite. As we have seen, the reason for this is the polar
bonding between the small, doubly charged Mg2+ ion with
water molecules (Christ and Hostetler, 1970). The Mg/Ca
ratio, at over 100 (Russell et al., 1999), is particularly high, an
added impetus to the precipitation of hydromagnesite rather
than magnesite (Müller et al., 1972).
Given the association between ultramafic rocks, Mg-rich
springs, and lacustrine hydromagnesite demonstrated at
Salda Gölü, we analyzed springs draining into Akgöl, an enclosed dry lake 15 km north-northeast, and of Salda (Fig. 4).
These springs carried up to 170 mg/l Mg2+ and about 10 mg/l
Na+ (Russell et al. 1999). The bed of this lake also proved to
435
436
ZEDEF ET AL.
Fig. 7. The origin of hydromagnesite in lakes such as Salda Lake is explained using diagrams computed for 25°C using the Geochemists Workbench program REACT (Bethke, 1996). A magnesite-saturated fluid, as
would be formed on equilibration of a CO2-rich fluid with magnesite, is allowed to lose CO2. Loss of CO2 is modeled by lowering the fugacity of CO2.
Magnesite precipitation has been suppressed to represent supersaturation.
Magnesite does not precipitate as the small doubly charged magnesium ion
cannot lose its sheath of polar-bonded water molecules at low temperature
(Christ and Hostetler, 1970). There is no redox aspect. (a). As CO2 fugacity
falls, the solution becomes increasingly supersaturated in magnesite. Eventually hydromagnesite becomes saturated and would be expected to precipitate below log fCO2 = –2.3, though it too may become supersaturated to a degree. (b). The pH is shown to increase from 6.5 to 9.0 as the fugacity of CO2
falls to atmospheric level.
comprise hydromagnesite mud mixed with an equal portion
of clay, at least out to a kilometer from the shoreline springs.
Hydromagnesite is not a common mineral and, previous to
these discoveries, unknown in large concentrations. O’Neil
and Barnes (1971) hypothesize that hydromagnesite occurring to depths of about 10 m in the ultramafic mass from New
Idria and Red Mountain, United States, is the weathering
product of serpentinite and brucite.
Travertine
Whereas the Salda Gölü magnesian carbonate deposits
were, and are being, generated over cool seepages (Braithwaite and Zedef, 1994, 1996b), the travertine (calcite) deposits at Pamukkale, north of Denizli, are developed at hot
springs (~60°C, Ford and Pedley, 1996) about 100 m above
the Büyük Menderes rift floor on its northern scarp (Altunel
and Hancock, 1993). We investigated and analyzed these deposits, as well as some Pleistocene travertines at nearby Kocabas (Fig. 4), to allow us to compare and contrast the stable
isotope ratios of carbonates from this well characterized system with Mg carbonates where genesis is not so clear.
Analytical Procedures
One hundred and twenty-two samples, representing five major
carbonate deposits, several small associated developments,
0361-0128/98/000/000-00 $6.00
and six possible carbonate source rocks, were analyzed for
their δ13C(PDB) and δ18O(SMOW) isotope values. Mineralogical
compositions were determined by X-ray diffraction. Powdered samples (100 mesh) were reacted under vacuum with 5
ml 103 percent H3PO4 (specific gravity 1.92) at 100°C, for
magnesite, hydromagnesite, and dolomite, and at 25°C for
calcite. Eighteen hours were allowed for the reaction. The
CO2 was then purified in the extraction line by the standard
procedure of McCrea (1950) and the yield was measured by
an attached digital manometer. The ratios of 18O/16O and
13
C/12C of the samples were analyzed on a mass spectrometer
(ISOGAS, SIRA-10). A 1 σ precision of ±0.1 per mil for both
carbon and oxygen isotopes was realized. During the acid
(H3PO4)-carbonate reaction, the δ18O value of the released
CO2 is not the precise representative of the original carbonate mineral. For this reason δ18O values must be obtained by
using the oxygen isotope fractionation factor (103 ln α) which
is specific for each mineral. These factors are 9.29, 9.12, and
10.20 for magnesite, dolomite, and calcite, respectively
(Rosenbaum and Sheppard, 1986). The hydromagnesite has
been assigned the magnesite factor. Yields of >77 percent
were routinely achieved. The results are reported as delta values (δ) relative to SMOW and PDB standards.
Results
The tabulated results (Table 2), plotted in Figure 8, are
compared with the Yugoslavian trend (Fallick et al., 1991).
The four calcite samples from veinlets in the Seydisehir
argillites have the lowest δ13C(PDB) values, ranging between
–19.3 to –20.0 per mil, and the δ18O(SMOW) values of these calcites are extremely low (~21.7‰). Salda Gölü hydromagnesites represent the heaviest values (3.9-4.7‰ for δ13C(PDB)
and 30.0-36.8‰ for δ18O(SMOW)), with the recently denatured
samples being at the heaviest end of the range. The average
δ13C(PDB) and δ18O(SMOW) values of two hydromagnesite samples from the dry surface of Akgöl are 0.4 and 28.0 per mil,
respectively. The Hırsızdere sedimentary magnesite deposits
have δ13C(PDB) values of –0.3 to +4.4 per mil but unexpectedly
light δ18O(SMOW) values, between 21.6 and 26.5 per mil.
The results of 11 magnesite samples of the Koyakcı Tepe
vein- and stockwork-type magnesite deposits, near Helvacıbaba, are –11.8 to –14.3 and 25.9 to 27.7 per mil for δ13C(PDB)
and δ18O(SMOW) values, respectively. Isotopically, these are the
lightest values of the magnesite samples we analyzed from
western Turkey, though they are similar to ratios found in vein
magnesites from North Evia, Greece, and in Serbia (Gartzos,
1990; Fallick et al., 1991). Magnesite samples from the Helvacıbaba cryptocrystalline stockwork magnesite deposit itself
have δ13C(PDB) values ranging from –10.1 to –13.8 per mil and
δ18O(SMOW) values from 26.4 to 27.6 per mil. Three magnesite
samples of pebbles from the conglomerate lying unconformably above the Helvacıbaba stockwork magnesites have
significantly higher δ13C(PDB) (–7.5 to –8.2‰) and slightly
lower δ18O(SMOW) (24.9–25.3‰) values with respect to Helvacıbaba stockwork magnesites. The isotopic values of the
three samples collected from the bedded dolomites of probable Neogene age, which in turn overlie the conglomerates,
have values similar to the magnesite pebbles (δ13C(PDB) of
–7.0 to –7.9‰ and δ18O(SMOW) of 25.3–27.0‰).
436
437
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
TABLE 2. Isotopic Compositions of Turkish Carbonates
Sample no.
δ13C (PDB)
(‰)
δ18O(SMOW)
(‰)
Description
Salda Gölü hydromagnesite
MV42
4.3
MV49
4.2
MV12
4.5
MV41
4.4
MV47
4.5
MV45
4.1
MV44
4.7
MV43
4.5
MV8
4.2
MV48
4.6
VZ92-18
4.5
VZ92-19
4.2
VZ92-20
4.4
VZ92-21
4.7
VZ92-22A
4.5
VZ92-22B
3.9
VZ92-24
4.3
VZ93-24-1
3.4
VZ93-24-2
4.3
VZ93-24-3
4.8
VZ93-15
4.4
VZ93-2
3.6
VZ93-5
3.7
VZ93-7
4.7
VZ93-8
4.3
VZ93-9
3.8
VZ93-10
0.2
VZ93-30
4.4
VZ92-27
4.3
VZ92-70
4.2
VZ93-21
4.1
VZ93-22
3.9
30.6
36.0
31.1
33.8
35.1
35.1
34.8
33.7
33.8
33.6
36.3
36.1
36.8
35.6
36.0
34.5
36.2
35.8
36.6
38.1
37.3
33.0
36.2
35.3
34.9
36.1
35.3
36.0
36.5
36.2
35.9
35.2
Stream cobble (SW)
Living microbialite
Mudstone (SW)
Kocaadalar Bay terrace
Kocaadalar Bay terrace
Kocaadalar Bay terrace
Kocaadalar Bay terrace
On harzburgite (SW)
15 m terrace (SW)
30 m terrace (SW)
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
Kocaadalar Burnu
W lake shore
W lake shore
W lake shore
Kocaadalar NW lake shore
E lake shore on Cretaceous 1st
E lake shore on Cretaceous 1st
E lake shore on Cretaceous 1st
E lake shore on Cretaceous 1st
E lake shore
SE lake shore
SE lake shore
SE lake shore
Akgöl hydromagnesite
MV19
0.7
MV19
0.1
27.9
28.2
Dry lake sediment
Dry lake sediment
Pammukale
VZ92-56A
VZ92-56
VZ92-58
VZ92-60
VZ93-PA
VZ93-PB
5.8
6.2
5.6
5.3
7.2
6.2
19.7
20.7
19.3
19.9
21.3
23.0
Travertine calcite
Travertine calcite
Travertine calcite
Travertine calcite
Travertine calcite
Travertine calcite
Kocabas
MV22
MV23
MV24
MV25
5.2
3.2
-0.2
4.5
19.4
19.5
24.7
23.0
Travertine calcite
Travertine calcite
Travertine calcite
Travertine calcite
Hirsizdere
VZ30
VZ31
VZ32
VZ33
VZ34
VZ35
VZ36
VZ37
VZ38
VZ39
VZ40
VZ42
VZ43
VZ44
VZ45
0.9
2.0
1.5
1.9
3.1
2.8
4.3
3.8
3.8
2.2
3.7
1.6
4.4
0.9
3.4
25.3
25.2
26.0
22.8
26.5
25.2
24.6
25.2
25.8
25.3
25.3
24.2
26.0
24.9
25.3
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded dolomite
Bedded magnesite
Bedded magnesite
Bedded magnesite
0361-0128/98/000/000-00 $6.00
Sample no.
δ13C (PDB)
(‰)
δ18O(SMOW)
(‰)
Description
Hirsizdere (Cont.)
VZ46
VZ48
VZ49
VZ50
VZ51
VZ52
VZ53
VZ54
0.8
–0.3
3.8
2.8
1.2
3.2
3.9
3.4
25.3
21.6
25.3
25.8
26.1
25.9
26.1
25.5
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Bedded magnesite
Helvacibaba
VZ13A
VZ13B
VZ13C
–7.5
–8.0
–8.2
24.9
25.3
25.2
Detrital magnesite
Detrital magnesite
Detrital magnesite
Helvacibaba
VZ12N
VZ12NB
VZ14C
VZH-12-2
VZH-12-2B
–7.0
–7.6
–7.9
–7.0
–7.6
25.3
25.8
27.0
25.3
25.8
Bedded dolomite
Bedded dolomite
Bedded dolomite
Bedded dolomite
Bedded dolomite
W-Helvacibaba
VZ13W Top
VZ11W
VZ8W
VZ7W
VZ5W
VZ3W
VZ1W Base
–11.9
–11.6
–11.4
–10.1
–11.3
–10.5
–11.2
26.8
26.5
27.3
26.7
27.2
26.8
26.4
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
E-Helvacibaba
VZ11E Top
VZ9E
VZ8E
VZ7E
VZ5E
VZ3E
VZ1E Base
–13.8
–13.2
–11.1
–11.7
–11.9
–10.3
–10.8
27.2
27.3
27.3
26.9
26.6
27.1
27.6
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Helvacibaba
VZ89
VZ142A
VZ143A
VZ145
VZ148A
–10.8
–10.5
–10.3
–10.3
–10.0
27.4
28.2
29.0
27.3
27.4
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Koyaki
VZ52A
VZ58
VZ59B
VZ61B
VZ61C
VZ62C
VZ63
VZ112
VZ113B
VZ117
VZ121B
–13.3
–13.3
–12.6
–13.4
–12.9
–11.8
–14.3
–13.2
–13.2
–12.8
–13.3
26.4
26.4
26.2
26.4
27.4
27.7
26.3
26.6
27.4
26.8
25.9
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
Vein stockwork magnesite
1.4
0.1
–0.4
0.2
27.5
27.0
27.7
27.7
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
Stockwork magnesite
1.6
2.4
30.9
29.8
Cretaceous limestone
Cretaceous limestone
Arapomer deresi
VZ15A
VZ15B
VZ15C
VZ15D
Salda Gölü
VZ26A
VZ 26B
437
438
ZEDEF ET AL.
TABLE 2. (Cont.)
δ C (PDB)
(‰)
13
Sample no.
Midos
VZ92-4
VZ92-5
VZ92-6
VZ92-7
Bozkir
VZ92-3K
VZ92-4K
δ O(SMOW)
(‰)
Interpretation and Models
18
Status of the three hypothetical carbonate sources
In our previous studies of cryptocrystalline magnesite deposits in ophiolites we suggested that three main sources of
carbonate variously contributed to the deposits (Fallick et al.,
1991). These were (1) atmospheric carbon dioxide (O’Neil
and Barnes, 1971; Fallick et al., 1991), (2) carboxyl from organic-rich sediments (Fallick et al., 1991; Brydie et al., 1993),
and (3) carbon dioxide driven off limestone during contact
metamorphism (Fallick et al., 1991). The δ13C(PDB) v.
δ18O(SMOW) values of the three supposed dominant sources or
end members established for the Yugoslavian magnesite deposits were ~4 and ~36 per mil (end-member 1), –15 and +28
per mil (end-member 2), and ~+3 and ~+19 per mil (endmember 3). A regression line between 1 and 2 was given by:
Description
1.5
1.9
1.9
1.1
28.7
29.3
29.1
28.7
Cretaceous limestone
Cretaceous limestone
Cretaceous limestone
Cretaceous limestone
3.6
3.5
24.5
25.1
Jurassic limestone
Jurassic limestone
Ardicli
VZ 41
VZ 173
2.3
1.8
24.7
23.4
Permian-Jurassic limestone
Permian-Jurassic limestone
Loras
VZ24
VZ187
–4.6
–3.9
20.7
20.7
Jurassic-Cretaceous limestone
Jurassic-Cretaceous limestone
Caltepe
VZ10A
VZ11A
VZ11B
0.5
0.7
0.6
20.8
17.7
19.4
Cambrian limestone
Cambrian limestone
Cambrian limestone
Seydisir
VZ11B
VZ10A
VZ11A
VZ11B
–19.3
–19.9
–20.0
–19.4
21.8
21.3
21.6
22.2
Calcite veinlets
Calcite veinlets
Calcite veinlets
Calcite veinlets
δ18O = 0.508δ13C + 32.3,
Arapömer Deresi cryptocrystalline stockwork magnesites
have isotopic values that contrast strongly with the other cryptocrystalline vein stockwork deposits. Here δ13C(PDB) values
are –0.4 to +1.4 per mil and δ18O(SMOW) values are 27.0 to 27.7
per mil. The calcite samples of Pamukkale travertines have
the lowest δ18O(SMOW) (~20.7‰) and highest δ13C(PDB)
(~6.1‰) values. Two calcium carbonate samples of Cretaceous limestone comprising the southeastern shoreline to
Salda Gölü have 1.6 to 2.4 per mil δ13C(PDB) and 29.8 to 30.9
per mil δ18O(SMOW) values. The water samples of Salda Gölü
have δ18O(SMOW) values of 2.8 per mil and δD values of ~3.5
per mil (Fig. 9).
Fluid Temperatures
In order to calculate the isotopic composition of the mineralizing fluids and hence their temperatures, the equation:
103 ln α ~ δ18Om – δ18Ow = A(106T2–) + B
(6)
(Aharon, 1988) is used where 103 ln is the per mil fractionation between mineral (m) and water (w). Temperature (T) is
in Kelvin, and A and B are the constants appropriate to each
particular carbonate (Table 3). Temperatures of the mineralizing solutions are calculated on the assumption that they are
generally derived from meteoric water with a δ18O(SMOW)
value of –5 per mil. On the other hand, the temperature of
deposition of hydromagnesite in Salda Gölü is calculated
using the analyzed value of the (evaporated) lake water of 2
per mil (Table 4).
0361-0128/98/000/000-00 $6.00
(7)
significant at the 99 percent confidence level (Fallick et al.,
1991).
How these putative end members relate or otherwise to the
main hypothetical sources of the Turkish deposits, and the
waters from which the carbonates were precipitated, is discussed below.
Salda Gölü hydromagnesite (end-member 1): The δ13C(PDB)
value of atmospheric CO2 is –7.0 per mil. The δ13C(PDB) value
of carbon in carbonate precipitated in lakes at around 20°C is
expected to be about 4 per mil or a little more, a fractionation
of around 11 per mil (O’Neil and Barnes, 1971). Meteoric
water in the area has a δ18O(SMOW) value of –5 per mil though
the lake water sampled from Salda Gölü is +2 per mil. Therefore, hydromagnesite precipitated at 20°C would be expected
to have a δ18O(SMOW) value of around 35 per mil (Fig. 8a).
(Were magnesite to have been precipitated from the same
fluid then its δ18O(SMOW) value would have been about 40‰).
With a mean δ13C(PDB) value of 4.4 per mil (excluding one
value of 0.2‰) and a mean δ18O(SMOW) value of 35.9 per mil
(excluding values from older developments well above the
present lake) the isotopic values of the Salda Gölü hydromagnesite samples are as expected from our model. The isotopic
values are also broadly comparable with those of cryptocrystalline sedimentary magnesite deposits such as Bela Stena,
Serbia (Fallick et al., 1991), the huntite-hydromagnesitemagnesite deposits of Servia, Macedonia (Kralik et al., 1989),
the evaporitic stromatolitic magnesite and dolomite developments in the Coorong lagoon, South Australia (Walter et al.,
1973; Botz and von der Borch, 1984), and even one cryptocrystalline stockwork, the Miokovachi Beli Kamen deposit
in Serbia (Fallick et al., 1991).
The isotopic results are best interpreted as indicating precipitation from the supersaturated surface waters of Salda
Lake (end-member 1; Fig. 10a). The fact that the range of
δ13C values is so restricted can be taken to indicate minimal
direct cyanobacterial involvement in precipitation (Guo et al.,
1996), though the bacteria do bring about the nucleation of
hydromagnesite (eqs. 3–5). We estimate, given the volume of
the lake from the rate of evaporation, that a million tons or
more of hydromagnesite could be generated well within 5,000
yr (Russell et al., 1999). Our conclusion is supported by the
438
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
Fig. 8. a. The plot of δ13C(PDB) vs. δ18O(SMOW) for the Turkish magnesite and hydromagnesite and associated carbonate deposits in relation to the regression lines from magnesite deposits in the former Yugoslavia, reported by Fallick et al., (1991).
As with the Yugoslavian data, populations of isotopic ratios for each deposit are relatively distinct; thus symbols contrast only
where fields overlap. b. Plot of δ13C(PDB) vs. δ18O(SMOW) for possible carbonate source rocks. Also shown are the putative end
members EMY1 and EMT1 (theoretical low-temperature atmospheric-meteoric magnesite-hydromagnesite for the Yugoslavian and Turkish fields, respectively) as well as EMT2 (theoretical hot, organically derived CO2 in ground water). Abbreviations: A = Ardlici Limestone of Permo-Jurassic age, B = Bozkir Limestone of Jurassic age, C = Caltepe Limestone at Seydisehir of Cambrian age, L = Loras Limestone of Cretaceous-Jurassic age, M = Midos Limestone of Cretaceous age, S =
Cretaceous Limestone from the east shore of Salda Gölü.
0361-0128/98/000/000-00 $6.00
439
439
440
ZEDEF ET AL.
Fig. 9. Plot of δD vs. δ18O(SMOW) of two water samples from Lake Salda.
The Mediterranean meteoric water line is from Sheppard (1986). The average meteoric water for this region is estimated on the basis of rainfall as
δ18O(SMOW) ~–5 and δD ~–20 per mil.
following geologic information: (1) the deltaic sediments and
beach deposits are riddled with the mineral, (2) the microbialites are especially well developed over the delta top and
apron, (3) the isotopic contrast with both the Pamukkale hotspring travertines and the Hırsızdere magnesite deposit
which are of hydrothermal-exhalative origin (see below), (4)
the temperatures of the lake do not increase toward the base
of the microbialites, and (5) Salda Gölü itself appears to be in
an area of low heat flow (Tezcan, 1979).
Carbonate veinlets in the Seydisehir metapelites (end-member 2): In our previous study of the Serbian magnesite veins
we had speculated on an origin of carbonate by decarboxylation of organic-rich sediments beneath ultramafic rocks. In
TABLE 3. Values of Constants A and B Used in Equation 6
Mineral
A
B
References
Hydromagnesite
2.74
0.83
O’Neil and Barnes (1971)
Friedman and O’Neil (1977)
Magnesite
3.53
–3.58
Aharon (1988)
Calcite
2.78
–2.98
O’Neil et al. (1969)
Friedman and O’Neil (1977)
Dolomite
3.23
–3.29
Sheppard and Schwarcz (1970)
TABLE 4. Calculated and Measured Temperatures of Deposition
Deposit
Arapömer Deresi stockwork
Koyakcı Tepe stockwork
Helvacıbaba stockwork
Helvacıbaba sedimentary dolomites
Salda Gölü
Hırsızdere exhalative deposit
Pamukkale and Kocabas travertines
0361-0128/98/000/000-00 $6.00
Calculated
temperature
(°C)
~80
~80
~80
~50
~18
~100
~40
Measured
temperature
(°C)
20
40–60
particular, we suggested the carbonate in the Golesh, Liska,
Draglica, and Brezack deposits in Serbia, for which we calculated temperatures of deposition of around 70°C, was partly
derived from organic-rich sediments by decarboxylation and
partly from an increasing (from Golesh through to Brezack)
concentration of atmospheric CO2 dissolved in artesian waters (Fallick et al., 1991). Extrapolating back to a pure end
member would give δ13C(PDB) and δ18O(SMOW) values of –20
and +22 per mil, respectively (Fig. 8a).
Phyllites with vestigial organic carbon (now graphite and/or
kerogen) comprise members of the Seydisehir Formation
that crops out about 80 km to the southwest of Konya. This
Cambro-Ordovician unit, metamorphosed to lower greenschist facies (Öncel 1995), looks to have been underthrust beneath the serpentinites hosting the giant stockwork deposits
near Konya (Figs. 1–3). It took our interest as a possible contributor of hydrothermal carbonate via oxidation of inorganic
carbon because it does contain veinlets of carbonate. Clearly
in the case of the Seydisehir meta-argillite and other Paleozoic strata, carboxyl groups could not have survived the metamorphism. Thus, if carbonate were to have been partially derived from this and similar sources it could only have been by
oxidation (e.g., Shock, 1988).
Isotopic ratios of calcites from the Seydisehir Formation
(–19.4‰ δ13C(PDB) and +22.2‰ δ18O(SMOW)), whereas much
lower than the δ13C(PDB) and δ18O(SMOW) values of either vein
or stockwork magnesites, do lie on the extrapolation of the
line drawn by Fallick et al. (1991; eq 7 and Fig. 8) for Serbian
magnesite deposits. Thus, they could represent a contributor
to the isotopically light δ13C(PDB) and δ18O(SMOW) end-member
2, i.e., the putative underthrust source of carbonate in the
vein stockwork deposits of the Konya region. Organic carbon
generally incorporates very light carbon (δ13C(PDB) around
–25‰), whereas carboxyl group carbon normally plots
around –20 per mil (Irwin et al., 1977). Again the results plot
approximately as expected (Fig. 8a), though given the metamorphic grade of the Seydisehir Formation, we cannot consider this a definitive result. Indeed, we might have expected
the isotopically lighter values typical of oxidation of organic
carbon. In any case, it may be that these veinlets formed in
the Variscan, and further work is required to substantiate, or
otherwise, this hypothesis.
Limestones (end-member 3): Because contact thermal
metamorphism of limestone could release CO2, to contribute
one possible source of the carbonate in magnesite (Fallick et
al., 1991), all the formations close to magnesite deposits were
analyzed. In the Konya area these are the Midos Limestone
(Upper Cretaceous), the Loras Limestone (Upper Jurassic to
Upper Cretaceous), the Bozkir Limestone (Jurassic), and the
Ardicli Limestone (Permian to Lower Jurassic: Fig. 2). Also,
the Çaltepe Limestone (Cambro-Ordovician), which crops
out 80 km southwest of Konya (mentioned above), was analyzed as a possible deep source. The Cretaceous Limestone,
which crops out on the eastern shore of Salda Gölü, was analyzed because it impinges on the lake itself. Marine carbonates generally have a δ13C(PDB) value of 0 ± 4 per mil (Field
and Fifarek, 1985; Salomons and Mook, 1986; Emery and
Robinson, 1993). The marine limestones that crop out in the
region generally follow the evolutionary oxygen isotope path
systematically backward, period by period, of ever heavier
440
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
441
small pocket of the limestone that has reacted with basin fluids bearing some petroleum.
a
b
c
Fig. 10. Genetic models. a. Generation of the Lake Salda hydromagnesite
microbialites (thickness of hydromagnesite exaggerated). b. Genesis of magnesite deposits in western Anatolia projected onto a north to south cross section. c. Genesis of the stockwork deposits just west of Konya projected onto
a northeast-southwest section. Partly after Ilich (1988), Fallick et al. (1991),
Braithwaite and Zedef (1994, 1996b), Zedef (1994), and an adaptation of
Lur’ye and Gablina (1972).
oxygen, from the Cretaceous to the Cambrian, presumably a
reflection of progressive diagenesis (Fig. 8b; Keith and
Weber, 1964; Faure, 1986). The one exception is the Cretaceous Loras Limestone with low values for both δ18O(SMOW)
and δ13C(PDB), averaging +21 and –4 per mil, respectively. It is
inappropriate for us to make an interpretation on just two
data points, though it is possible that the samples represent a
0361-0128/98/000/000-00 $6.00
Deposits lying on the putative mixing line 1–2
Helvacıbaba and Koyakcı Tepe deposits: The isotopic results from these vein stockwork deposits are in concert with
expectation. We have previously suggested (Fallick et al.,
1991, Brydie et al., 1993) that the δ13C(PDB) values between
–10 and –15 per mil indicate a contribution of carbonate from
decarboxylation of organic-rich sediments, which may be underthrust beneath the ultramafic host and source rocks (Karamata, 1974; Smith and Spray, 1984; Brydie et al., 1993), with
the remainder derived from surface CO2. According to Curtis (1978) and Emery and Robinson (1993), the decarboxylation of organic-rich sediments begins at temperatures of
around 70° to 75°C. This process results in fractionations
such that a range of δ13C(PDB) values between –10 and –20 per
mil for carbon dioxide (as bicarbonate) are realized (Irwin et
al., 1977). However, Shock (1988, 1994) has argued from
thermodynamic considerations that organic acids are oxidized
directly to carbon dioxide. If so, δ13C(PDB) values found in carbonates might be expected to range between –20 and –25 per
mil. This is not the case in our analyses. Nevertheless, Shock
(1988) goes on to speculate that oxidation can be mediated
by bacteria, a complicating factor in isotopic fractionation
(and see Pedersen, 1997). The simplest explanation of the
δ13C(PDB) values ranging from –15 to –10 per mil in the magnesites might be that a proportion of the carbonate is derived
by the decarboxylation of organic-rich sequences underthrust
beneath the ultramafic host rocks (and see Jedrysek and
Halas, 1990; Abu-Jaber and Kimberley, 1992; Brydie et al.,
1993). However, the Seydisehir veinlet carbonates (see above)
are isotopically a little too heavy to be considered as derived
from light carbon from R groups (i.e., about –25‰, as expected by Shock 1988). Certainly the metamorphic grade of
these meta-argillitic phyllites (Öncel 1995) would preclude a
carboxyl source. It could be that the somewhat sparse accumulations of organic carbon in these meta-argillites underthrust beneath the Çayırbagı ophiolite were oxidized and provided a portion of the carbonate, though as stated above, it
could just as well be argued that the veinlets were formed
during an earlier metamorphic event. Alternatively, the source
could be an unknown Mesozoic organic-rich formation underthrust beneath the same ophiolite. Either way, precipitation would be brought about by the sudden release of carbon
dioxide as fluids saturated in magnesite neared the surface
(Ilich 1974).
The alternative explanation of Petrov (1967) and Lesko
(1972) that such vein stockwork deposits are generated by
surface weathering (see Zachmann and Johannes 1989) is rejected on mass-balance grounds. Moreover the “equilibrium
field” of O’Neil and Barnes (1971, fig.1) presumed for their
surface weathering model does not have the positive correlation between δ13C(PDB) and δ18O(SMOW) values revealed in this
study (Fig. 8).
Arapömer Deresi deposit: The isotopic data from the Arapömer Deresi cryptocrystalline stockwork magnesite deposits
reveal high δ13C(PDB) values (ranging from –0.4 to +1.4‰) with
respect to other stockwork deposits. Cryptocrystalline magnesites in ultramafic complexes are more often represented
441
442
ZEDEF ET AL.
by isotopically lighter δ13C(PDB) values (–6 to –18‰; Kralik et
al., 1989), though Miokovachki, in Serbia, is even heavier
(Fallick et al., 1991). The carbon isotope signature of the Arapömer Deresi magnesite can be explained by a marine carbonate source though some contribution from magmatic CO2
is also likely. When ground water convects at depth through
marine limestones or marbles it could pick up CO2 either by
decarbonation or dissolution of limestone (e.g., as at Kızıldere
65 km to the northwest; Ten Dam and Erentöz, 1970). If so,
isotopic reequilibration could be established between mineralizing fluid and the marine limestone source (δ13C(PDB) from
–2 to +2‰). Moreover the oxygen isotope values are comparable with those from Mesozoic limestones. Decarbonation of
such limestones at depth as a source of carbonate becomes
more attractive when the Neogene to Recent andesiticbasaltic volcanics cropping out at centers 10 and 20 km to the
west are taken into account. We previously explained the genesis of the Oshve vein in Bosnia in the same way. Here the
magnesite was interpreted to have precipitated at around
105°C, the highest temperature calculated in all the deposits
analyzed from the former Yugoslavia (Fallick et al., 1991).
Nevertheless, the coincidence of the isotopic ratios with those
of the hydromagnesite lake sediments in Akgöl does permit a
per descendum model to be considered for this small deposit.
Magnesite (originally hydromagnesite) could have been precipitated in an evaporating lake (any evidence for which has
been eroded) as well as in fractures beneath.
Hırsızdere sedimentary deposits: The isotopic results of
carbonate samples from the Hırsızdere (Denizli) sedimentary
magnesite deposits plot between the fields generally occupied by thermal decarbonation of marine limestones through
to atmospheric values (Fig. 8). The δ18O(SMOW) values are
lower than is usual for sedimentary magnesites, although the
δ13C(PDB) values are comparable with the well-defined sedimentary and volcanic-related vein-type deposits. The deposit
itself is associated with Neogene volcanic rocks judging from
the 1:2,000,000 geologic map of Turkey and lies in the center
of the Büyük Menderes graben (Figs. 1 and 4).
High δ13C(PDB) values indicate that the carbon dioxide could
have an atmospheric origin. Alternatively some carbon dioxide could have evolved during contact metamorphism of marine limestone thrust beneath the serpentinite (cf. Arapömer
Deresi). Yet another possible interpretation is that these values resulted from fractionation during evaporation of the
lighter carbon isotope. The unusually low oxygen isotope ratio
(~25‰) could be explained by depositional temperatures, at
or close to the boiling point. Hydrothermal solutions feeding
the deposit could have risen up a boundary-parallel fault in
the middle of Neogene Lake Acigöl, gaining magnesium as
they traversed the northern fringe of the Yesilova ultramafic
complex (Fig. 10b). Thus this huge sedimentary magnesite
deposit would have formed as envisaged by Ilich (1968) and
could be considered a type example of the hydrothermal-sedimentary process (cf. Shilopaj-Nevade in Serbia; Ilich, 1976;
Fallick et al,. 1991). If so, there could be a vein stockwork deposit beneath awaiting discovery.
Helvacıbaba detrital magnesite and bedded dolomites: The
few isotopic analyses from these Neogene developments are
rather puzzling. The isotopic ratios of the pure magnesite
pebbles occupy a δ13C(PDB) field about 3.5 per mil higher than
0361-0128/98/000/000-00 $6.00
the presumed stockwork source, yet the δ18O(SMOW) values are
about 2 per mil lighter. Perhaps the pebbles reflect the true
isotopic compositions of the now-eroded top of the stockwork
where the fluids were degassing and boiling. Alternatively,
weathering or diagenetic overprinting by ground water occupying the permeable Neogene conglomerate could have been
responsible for the isotopic shift, as suggested by Gartzos
(1990) for a similar association in North Evia, Greece.
The oxygen isotope ratios of the sedimentary dolomites
overlying the stockwork are comparable with the magnesite
pebbles, if the different oxygen isotope fractionation of –2 per
mil between magnesite and dolomite is taken into account.
But the carbon isotope ratios are also about 4 per mil higher
than the stockwork. The data permit the interpretation that
the dolomites are sedimentary-exhalative (cf. Ilich, 1968), the
isotope shift being explained by degassing of CO2. If so, it is
likely that the hydrothermal fluids generated the stockwork
before escaping into the coeval Neogene lake.
Pamukkale and Kocabas: Of all the Neogene-Quaternary
carbonates analyzed, the present-day, hot-spring travertine
(calcite) deposits at Pamukkale have the highest δ13C(PDB) values (avg 6.1‰) and the isotopically lightest δ18O(SMOW) values
(~20‰). The Pleistocene travertines at Kocabas are similar
though their δ18O(SMOW) values rise to 24.7 per mil. The
δ13C(PDB) results could be attributed to carbonate derived
from Paleozoic marbles beneath the travertines as suggested
by Altunel and Hancock (1993), although the δ18O(SMOW) values are also typical of heated ground waters. Indeed, if we assume for the moment calcite to have been precipitated from
local meteoric water with a δ18O(SMOW) value of –5 per mil,
then the calculated temperature for a δ18O value of 25 per mil
turns out to give the expected temperature of ~ 40°C (cooled
from the original 60°C). Yet an atmospheric source of carbon
(in CO2) from carbonate at 40°C would be expected to afford
a lighter δ13C(PDB) value (~2‰). Against this, degassing would
enrich the hot water in the heavier isotope. As another or additional explanation of the high δ13C(PDB) value, Guo et al.
(1996) have shown that cyanobacterial growth tends to fix
12CO from the hot-spring waters leaving more 13CO as a
2
2
source for the carbonate precipitates, although more of a
spread of δ13C values might be expected in such a case
(Amundson and Kelly, 1987).
Comparisons
The isotopic values of the Salda Gölü hydromagnesites and
the Koyakcı Tepe magnesites fall into the same heavy and
light isotopic end-member fields as those of the Bosnian and
Serbian magnesites (Fallick et al., 1991). The Salda Gölü association of hydromagnesite microbialites with the derived
hydromagnesite sediments and beach deposits is a good candidate for precursors to massive irregular magnesite deposits
found in lacustrine sediments. The Bela Stena deposit in Serbia is so far the only possible fossil example known to us.
Given the similarity in gross morphology and stable isotope
ratios between the Salda Gölü hydromagnesite microbialite
mounds and the lacustrine Bela Stena magnesite deposit
(Fallick et al., 1991; Russell et al., 1999), it is possible that the
carbonate at Bela Stena was originally precipitated as hydromagnesite. According to Sayles and Fyfe (1973) hydromagnesite can start converting into magnesite about 55°C at an
442
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
enhanced CO2 pressure (about 1 bar). These conditions are
likely to have occurred at Bela Stena during diagenesis:
Mg5(CO3) 4(OH) 2.4H2O + CO2 → 5MgCO3 + 5H2O.
(8)
In some contrast to the Bela Stena deposit, judging from
the isotopic ratios (δ13C(PDB) and δ18O(SMOW) avg 7.2 and
34.2‰, respectively), the huntite-hydromagnesite-magnesite
deposits in the Servia intermontane basin, Macedonia (Wetzenstein, 1975; Kralik et al.,1989; Stamatakis, 1995), could
have developed over methane-rich seepages derived from lignite seams beneath (Kralik et al., 1989; and see Hovland,
1990). Here magnesite occurs in massive bodies close to
faults. Nevertheless, Wetzenstein (1975, p. 129) has suggested that the magnesium in these huge deposits was “received…through fluviatile transport,” whereas Stamatakis
(1995) argues that the magnesium (and minor calcium) was
delivered as bicarbonate from springs emanating from the
basement, final enrichment being caused by evaporation.
Conclusions and Predictions
The following factors favor the generation of large (hydro)
magnesite deposits in obducted ophiolites: an underthrust organic source of carbonate or atmospheric carbon dioxide, extensional stress, igneous intrusion, moderate relief, a longlived low-temperature artesian system, wet winters, and hot
dry summers. Nevertheless, a direct demonstration of an underthrust and oxidizing organic source is still wanting. From
the research reported here and with reference to our models
for the generation of cryptocrystalline magnesite deposits in
ultramafic terranes, we make the following predictions:
1. Giant stockwork and vein complex magnesite deposits
will generally be found in ultramafic bodies only where underthrust organic-rich sediments, capable of sourcing the carbonate, exist at depth. Helvacıbaba and Koyakcı Tepe are the
main examples in western Turkey. In Serbia, the main vein
deposits are Golesh, Brezack, Draglica, and Liska, and the
main stockwork is Mramor-Razhana (Fallick et al., 1991). Values of δ13C(PDB) and δ18O(SMOW) generally plot along a line
from –15 and +25 to –9 and +28 per mil, respectively, implying a deep organic source of CO2 and fluid temperatures
around 80°C. But in one exceptional case, Miokovachki Beli
Kamen, the δ13C(PDB) and δ18O(SMOW) trend extends to 4 and
36 per mil, respectively (Fallick et al., 1991, fig. 6), which is
consistent with a surface-derived origin of CO2 and a low depositional temperature, conditions possibly met beneath a
shallow evaporating lake.
2. More rarely, vein deposits can occur where limestone,
underthrust beneath the ultramafic body, has supplied carbon dioxide to ground waters through decarbonation by an
igneous intrusion. The type example is Oshve, in Bosnia,
where δ13C(PDB) and δ18O(SMOW) values approximate ~3.1
and 19 per mil, respectively. The Arapömer Deresi cryptocrystalline stockwork magnesite deposit in western
Turkey could also have had such an origin (δ13C(PDB) and
δ18O(SMOW) avg 0.3 and 27.5‰, respectively). The temperature of the mineralizing fluids would have been between
70° and 100°C.
3. Massive strata-bound hydromagnesite stromatolites
comprising cyanobacteria and diatoms (Salda Gölü, western
0361-0128/98/000/000-00 $6.00
443
Turkey), as well as their diagenetically equivalent magnesite
bodies (Bela Stena, Serbia), will be found developed in sediments in enclosed lacustrine basins above river delta slopes
in ultramafic terranes. We should think of them as microbially mediated (nucleated) evaporite deposits. Their
δ13C(PDB) and δ18O(SMOW) values will be those expected of
fractionations from atmospheric-meteoric values at ambient
temperatures. In the southern Alpine terranes these values
will approach 4 and 36 per mil, respectively. This type of deposit, not requiring a source of carbonate at depth, could
develop (perhaps in association with huntite; Stamatakis,
1995) in lakes or lagoons subject to strong evaporation, in
any ultramafic terrane, not only on Earth but also Mars
(Russell et al., 1999). Such a deposit could be awaiting discovery beneath the surface hydromagnesite crusts near the
southwestern margin of Akgöl, 15 km north-northeast of
Salda Gölü.
4. Massive stratiform magnesite beds of hydrothermal exhalative origin will be found to occur above ultramafic bodies
near contemporaneous faults in extensional terranes marked
by local volcanism as predicted by Ilich (1988; and see Ilich,
1952). An example is Hırsızdere, which lies within Pliocene
lacustrine sediments of what was a more extensive Lake
Acigöl, in the middle of the Büyük Menderes graben.
Shilopaj-Nevade is the main example of this type in Serbia
(Ilich, 1976; Fallick et al,. 1991). Carbonate (augmented by
atmospheric carbon dioxide) can be derived either from limestones or carbonaceous rocks traversed by convecting thermal
waters, perhaps augmented by magmatic carbon dioxide. The
oxygen isotope values will reflect the relatively high temperatures of the mineralizing fluids (≤100°C). In Serbia, Bosnia,
Greece, and Turkey, the δ18O(SMOW) values are expected to
range between 20 and 30 per mil. The δ13C(PDB) values can
range across the whole spectrum from about –10 to +7 per
mil.
5. Further isotopic analyses of cryptocrystalline magnesite
deposits in Alpine terranes will plot on the two mixing lines
hypothesized to explain the genesis of the various deposits in
the former Yugoslavia. The δ13C(PDB) v δ18O(SMOW) plot first
revealing these lines (Fallick et al,. 1991) is generally reinforced by analyses of the Turkish deposits.
Acknowledgments
We thank Ahmet Ayhan, Principal of Gebze Yüksek
Teknoloji Enstitüsü, for introducing us to the Anotolian magnesite deposits. We appreciate the help of Dugald Turner,
Murdoch MacLeod, Bill Higgison, M. Salim Öncel, Güler
Göçmez, Jim Bischoff, and Stuart Haszeldine, Milo Ilich,
Greta Orris, Colin Pillinger, and an anonymous reviewer.
Ahmet Koç was our guide to Salda Gölü. Financial support
was maintained by the Principal of Selçuk University, Konya,
Halil Cin, and the Turkish Higher Education Committee
(YOK). Gresham College financed fieldwork for MJR.
August 21, 1998; August 24, 1999
REFERENCES
Abu-Jaber, N.S., and Kimberley, M.M., 1992, Origin of ultramafic-hosted
magnesite on Margarita Island, Venezuela: Mineralium Deposita, v. 27, p.
234–241.
443
444
ZEDEF ET AL.
Aharon, P., 1988, A stable-isotope study of magnesites from the Rum Jungle
uranium field, Australia: Implications for the origin of strata-bound massive
magnesites: Chemical Geology, v. 69, p. 127–145.
Altunel, E., 1963, Ilgin-Konya sheet: Institute of Mineral Research and Exploration, 1:500,000.
Altunel, E., and Hancock, P.L., 1993, Active fissuring and faulting in Quaternary travertines at Pamukkale, western Turkey: Zeitschrifte für Geomorphologie, suppl. v. 94, p. 285–302.
Amundson, R., and Kelly, E., 1987, The chemistry and mineralogy of a CO2rich travertine depositing spring in the California Coast Range: Geochimica et Cosmochimica Acta, v. 51, p. 2883–2890.
Barnes, I., and O’Neil, J.R., 1969, The relationship between fluids in some
fresh Alpine-type ultramafics and possible modern serpentinisation, Western United States: Geological Society America Bulletin, v. 80, p. 1947–1960.
Barnes, I., O’Neil, J.R., Rapp, J.B., and White, D.E., 1973, Silica-carbonate
alteration of serpentine: Wall-rock alteration in mercury deposits of the
California Coast Ranges: ECONOMIC GEOLOGY, v. 68, p. 388–398.
Barnes, I., O’Neil, J.R., and Trescases, J.J., 1978, Present day serpentinization in New Caledonia, Oman and Yugoslavia: Geochimica et Cosmochimica Acta, v. 42, p. 144–145.
Beköz, Ü., Baclavaci, Ö., and Ceylan, T., 1997, Kaynayan Soguk: Sualti
Dünyasi, March, p. 29–32.
Bethke, C., 1996, Geochemical reaction modeling: Concepts and applications: New York, Oxford University Press, 397 p.
Botz, R.W., and von der Borch, C.C., 1984, Stable isotope study of carbonate
sediments from the Coorong area, South Australia: Sedimentology, v. 31, p.
837–849.
Braithwaite, C.J.R., and Zedef, V., 1994, Living hydromagnesite stromatolites
from Turkey: Sedimentary Geology, v. 92, p. 1–5.
——1996a Corrigendum to Braithwaite, C.J.R. and Zedef, V., 1994: Sedimentary Geology, v. 106, p. 309.
——1996b, Hydromagnesite stromatolites and sediments in an alkaline lake,
Salda Gölü, Turkey: Journal of Sedimentary Research, v. 66, p. 991–1002.
Brydie, J.R., Fallick, A.E., Ilich, M., Maliotis, G., and Russell, M.J., 1993, A
stable isotopic study of magnesite deposits in the Akamas area, N.W.
Cyprus: Institution of Mining and Metallurgy Transactions, v. 102, sec. B,
p. B50–B53.
Burgath, K., Grissemann, C., Johannes, W., Jung, D., Kockel, F., Makris, J.,
Mohr, M., Mussalem, K., and Raschka, H., 1981, On the genesis of magnesite deposits in northern Greece: UNESCO International Symposium on
Metallogeny of Mafic and Ultramafic Complexes, Athens, Greece, 1980,
Proceedings, v. 3, p. 226–239.
Burne, R.V., and Moore, L.S., 1987, Microbialites: Organosedimentary deposits of benthic microbial communities: Society of Economic Palaeontologists and Mineralogists, v. 2, p. 241–254.
Christ, C.L., and Hostetler, P.B., 1970, Studies in the system MgO-SiO2CO2-H2O (II): The activity-product constant of magnesite: American Journal of Science, v. 268, p. 439–453.
Çoban, F., Maktav, D., and Kapadisli, S., 1995, Recent hydromagnesite occurrences, the Salda Lake, southern Turkey, in Pasava, J., Kribek, B., and
Zak, K., eds., Mineral deposits: From their origin to their environmental
impact: Rotterdam, A.A. Balkema, p. 561–564.
Cohen, H.A., Dart, C.J., Akyüz, H.S., and Barka, A., 1995, Synsedimentary
sedimentation and structural development of the Gediz and the Büyük
Menderes graben, western Turkey: Geological Society of London Journal,
v. 152, p. 629–638.
Collins, A.S., and Robertson, A.H.F., 1998, Processes of Late Cretaceous to
Late Miocene episodic thrust-sheet translation in the Lycian Taurides, SW
Turkey: Geological Society of London Journal, v. 155, p. 759–772.
Curtis, C.D., 1978, Possible links between sandstone diagenesis and depthrelated geochemical reactions occurring in enclosing mudstones: Geological Society of London Journal, v. 135, p. 107–117.
Dabitzias, G.S., 1980, Petrology and genesis of Vavdos cryptocrystalline magnesite deposits, Chalkidiki Peninsula, northern Greece: ECONOMIC GEOLOGY, v. 75, p. 1138–1151.
——1981, Additional evidence and a synopsis on the origin of the magnesite
deposits in the Vavdos district, northern Greece: UNESCO International
Symposium on Metallogeny of Mafic and Ultramafic Complexes, Athens,
Greece, 1980, Proceedings, v. 3, p. 239–283.
Emery, D., and Robinson, A., 1993, Inorganic geochemistry: Applications to
petroleum geology: London, Blackwell, 254 p.
Erdem, N.P., 1974, The ophiolitic series of Turkey: Maden Tektik ve Arama
(M.T.A) Enstitusu Dergisi (Turkey), v. 83, p. 131–144.
0361-0128/98/000/000-00 $6.00
Fallick, A.E., Ilich, M., and Russell, M.J., 1991, A stable isotope study of the
magnesite deposits associated with the Alpine-type ultramafic rocks of Yugoslavia: ECONOMIC GEOLOGY, v. 86, p. 847–861.
Faure, G., 1986, Principles of isotope geology: New York, NY, John Wiley and
Sons, 589 p.
Field, C.W., and Fifarek, R.H., 1985, Light stable-isotope systematics in the
epithermal environment: Society of Economic Geologists Reviews in Geology, v. 2, p. 99–128.
Ford, T.D., and Pedley, H.M., 1996, A review of tufa and travertine deposits
of the world: Earth-Science Reviews, v. 41, p. 117–175.
Friedman, I., and O’Neil, J.R., 1977, Compilation of stable isotope fractionation factors of geochemical interest: U.S. Geological Survey Professional
Paper No. 440-KK, 12 p.
Gartzos, E., 1990, Carbon and oxygen isotope constraints on the origin of
magnesite deposits, North Evia (Greece): Schweizerische Mineralogische
und Petrographische Mitteilungen, v. 70, p. 67–72.
Garven, G., 1995, Continental-scale groundwater flow and geologic
processes: Annual Review of Earth and Planetary Sciences, v. 23, p.
89–117.
Gedik, I., 1988, A paleogeographic approach to the Devonian of Turkey, in
MacMillan, N.J., Embry, A.F., and Glass, D.J., eds., Devonian of the world:
Calgary, D.W. Friesen and Sons, p. 557–567.
Güleç, N., 1991, Crust-mantle interaction in western Turkey: Implications
from Sr and Nd isotope geochemistry of Tertiary and Quaternary volcanics:
Geological Magazine, v. 128, p. 417–485.
Guo, L., Andrews, J., Riding, R., Dennis, P., and Dresser, Q., 1996, Possible
microbial effects on stable carbon isotopes in hot-spring travertines: Journal of Sedimentary Research, v. 66, p. 468–473.
Haack, U.K., and Zimmermann, H.D., 1996, Retrograde mineral reactions:
A heat source in continental crust?: Geologische Rundshau, v. 85, p.
130–137.
Hovland, M., 1990, Do carbonate reefs form due to fluid seepage? Terra
Nova, v. 2, p. 8–18.
Ilich, M., 1952, Magnezitsko leziste Bela Stena: Zbornik Ravoda RudarskoGeoloskog Fakulteta [Beograd] v. 1, p. 15–52.
——1968, Problems of the genesis and genetic classification of magnesite
deposits: Geologicky Zbornik-Geologica Carpathica XIX, Bratislava, v. 1, p.
149–160.
——1974, Hydrothermal sedimentary dolomite: The missing link: American
Association of Petroleum Geologists Bulletin, v. 58, p. 1331–1347.
——1976, Hydrothermal-sedimentary magnesite deposit of Nevade, Gornji
Milanovac (western Serbia): Zbornik radova Rudarsko-geoloskog fakulteta,
University of Beograd, Faculty of Mining and Geology Transactions, v. 19,
307–329.
——1988, Genetic types of magnesite deposits in the freshwater Miocene
basins of Yugoslavia: Des Annales Géologiques de la Péninsule Balkanique,
v. 51, p. 555–574 (in Russian).
Irwin, H., Curtis, C., and Coleman, M., 1977, Isotopic evidence for source of
diagenetic carbonates formed during burial of organic-rich sediments: Nature, v. 269, p. 209–213.
Jackson, J., and McKenzie, D., 1984, Active tectonics of the Alpine-Himalayan belt between western Turkey and Pakistan: Royal Astronomical
Society Journal, v. 77, p. 185–264.
Jedrysek, M.O., and Halas, S., 1990, The origin of magnesite deposits from
the Polish Foresudetic Block ophiolites: Preliminary δ13C and δ18O investigations: Terra Nova, v. 2, p. 154–159.
Karamata, S., 1974, The geological evolution of the area of Yugoslavia: The
nature and movement of plates and the importance of these properties and
processes for metallogeny, in Jankovic, S., ed., Metallogeny and concepts of
the geotectonic development of Yugoslavia: Belgrade University, Faculty of
Mining and Geology, p. 109–118.
Kastens, K.A., 1991, Responses to impending continent-continent collision in
the Mediterranean: Palisades, NY, Lamont-Doherty Geological Observatory, 1990–1991, Unpublished report, p. 24–32.
Keith, M.L., and Weber, J.N., 1964, Isotopic composition and environmental
classification of selected limestones and fossils: Geochimica et Cosmochimica Acta, v. 28, p. 1787–1816.
Ketin, I., 1983, Turkiye Jeolojisine genel bir bakis: Istanbul, Istanbul Teknik
Universitesi, 595 p.
Kissel, C., Averbuch, O., de Lamotte, D.F., Monod, O., and Allerton, S.,
1993, First paleomagnetic evidence for a post-Eocene clockwise rotation of
the western Taurides thrust belt east of the Isparta reentrant (southwestern
Turkey): Earth and Planetary Science Letters, v. 117, p. 1–14.
444
ULTRAMAFIC TERRANES, SW TURKEY: A STABLE ISOTOPE STUDY
Koçyigit, A., 1984, Güneybati Tütkiye ve yakin dolayinda levha içi yeni tektonik gelisim: Geological Society of Turkey Bulletin, v. 27, p. 1–16.
Kralik, M., Aharon, P., Schroll, E., and Zachmann, D., 1989, Carbon and oxygen isotope systematics of magnesites: A review, in Möller, P., ed., On the
formation of magnesite: Berlin, Gebrüder Borntraeger, Monograph Series
on Mineral Deposits, v. 28, p. 197–223.
Kreulen, R., 1980, CO2-rich fluids during regional metamorphism on Naxos
(Greece): Carbon isotopes and fluid inclusions: American Journal of Science, v. 280, p. 745–771.
Lesko, I., 1972, Über die Bildung von Magnesitlagerstätten: Mineralium Deposita, v. 7, p. 61–72.
Lippmann, F., 1973, Sedimentary carbonate minerals: New York, SpringerVerlag, 228 p.
Livermore, R.A., and Smith, A.G., 1985, Some boundary conditions for the
evolution of the Mediterranean region, in Stanley, D.J., and Wezel, F.C.,
eds., Geological evolution of the Mediterranean basin: New York, SpringerVerlag, p. 83–100.
Lur’ye, A.M., and Gablina, I.F., 1972, The copper source in the production
of Mansfeld-type deposits in the West Ural foreland: Geochemistry International, v. 9, p. 56–67.
Mann, S., Archibald, D.D., Didymus, J.M., Douglas, T., Heywood, B.R., Meldrum, F.C., and Reeves, N.J., 1993, Crystallization at inorganic-organic interfaces: Biominerals and biometric synthesis: Science, v. 261, p. 1286–1292.
McCrea, J.M., 1950, The isotopic chemistry of carbonates and a palaeotemperature scale: Journal of Chemical Physics, v. 18, p. 849–857.
Möller, P., 1989, Nucleation processes of magnesite: A review, in Möller, P.,
ed., On the formation of magnesite: Berlin, Gebrüder Borntraeger, v. 28, p.
287–292.
Müller, G., Irion, G., and Förstner, U., 1972, Formation and diagenesis of inorganic Ca-Mg carbonates in the lacustrine environment: Naturwissenschaften, v. 59, p. 158–164.
Öncel, M.S., 1995, Sarkikaraagaç-Yalvaç (Isparta) araslnln jeolojisi ve boksit
zuhurlarlnln mineralojik, petrografik, jeokimyasal incelemesi: Unpublished
Ph.D. thesis, Fen Bilileri Enststüsü Konya, Turkey, Selcuk University, 147 p.
O’Neil, J.R., and Barnes, I., 1971, C13 and O18 compositions in some freshwater carbonates associated with ultramafic rocks: Western United States:
Geochimica et Cosmochimica Acta, v. 35, p. 687–697.
O’Neil, J.R., Clayton, R.N., and Mayeda, T.K., 1969, Oxygen isotope fractionation in divalent metal carbonates: Journal of Chemical Physics, v. 51,
p. 5547–5558.
Ozcan, A., Goncuoglu, M.C., Turan, N., Uysal, S., Senturk, K., and Isik, A.,
1988, Late Palaeozoic evolution of the Kütahya-Bolkardag belt: Turkey,
Middle East Technical University, v. 21, p. 211–220.
Pedersen, K., 1997, Microbial life in deep granitic rock: Episodes, v. 20, p.
7–9.
Pentecost, A., 1978, Association of cyanobacteria with tufa deposits: Identity,
enumeration and the nature of the sheath material revealed by histochemistry: Geomicrobiology Journal, v. 4, p. 285–298.
——1995, Quaternary travertine deposits of Europe and Asia Minor: Quaternary Science Review, v. 14, p. 1005–1028.
Petrov, V.P., 1967, Osnovy ucenija o drenih korah vyvetrivanija: Nedra,
Moskva (in Russian).
Riding, R., 1992, Temporal variation in calcification in marine cyanobacteria:
Geological Society of London Journal, v. 149, p. 979–989.
Rosenbaum, J., and Sheppard, S.M.F., 1986, An isotopic study of siderites,
dolomites and ankerites at high temperatures: Geochimica et Cosmochimica Acta, v. 50, p. 1147–1150.
Russell, M.J., 1993, The origin of minerals and life at hot springs: Irish Association for Economic Geology Annual Review, p. 20–23.
Russell, M.J., Ingham, J.K., Zedef, V., Maktav, D., Sunar, F., Hall, A.J., and
Fallick, A.E., 1999, Search for signs of ancient life on Mars: Expectations
from hydromagnesite microbialites, Salda Lake, Turkey: Geological Society
of London Journal, v. 156, p. 869-888.
0361-0128/98/000/000-00 $6.00
445
Salomons, W., and Mook, W.G., 1986, Isotopic geochemistry of carbonates in
the weathering zone, in Fritz, P., and Fontes, J.C., eds., Handbook of environmental isotope geochemistry: New York, Elsevier, p. 239–270.
Sarp, H., 1976, Etude geologique et mineralogique du cortege ophiolitique
de la region situee au nord-ouest de Yeilova (Burdur-Turqie): Unpublished
Ph.D. thesis, Geneve, Universite de Geneve 143 p.
Sayles, W., and Fyfe, W.S., 1973, The crystallisation of magnesite from aqueous solution: Geochimica et Cosmochimica Acta, v. 37, p. 87–99.
Schmid, I.H., 1987, Turkey’s Salda Lake, a genetic model for Australia’s
newly discovered magnesite deposits: Industrial Minerals, August, p.
19–31.
Sengör, A.M.C., and Yılmaz, Y., 1983, The evolution of Tethys in Turkey: An
implication of plate tectonics: Ankara, Geological Survey of Turkey, 75 p.
Seyitoglu, G., and Scott, B., 1991, Late Cenozoic crustal extension and basin
formation in west Turkey: Geological Magazine, v. 128, p. 155–166.
Sheppard, S.M.F., 1986, Characterization and isotopic variations in natural
waters, in Valley, J.W., Taylor, H.P., and O’Neil, J.R., eds., Stable isotopes
in high temperature geological processes: Washington, DC, Mineralogical
Society of America, p. 165–185.
Sheppard, S.M.F., and Schwarcz, H.P., 1970, Fractionation of carbon and
oxygen isotopes and magnesium between co-existing metamorphic calcite
and dolomite: Contributions to Mineralogy and Petrology, v. 26, p.
161–198.
Shock, E.L., 1988, Organic acid metastability in sedimentary basins: Geology,
v. 16, p. 886–890.
——1994, Application of thermodynamic calculations to geochemical
processes involving organic acids, in Pittman, E.D., and Lewan, M.D., eds.,
Organic acids in geological processes: Berlin, Springer-Verlag, p. 270–318.
Smith, A.G., and Spray, J.G., 1984, A half-ridge transform model for the Hellenic-Dinaric ophiolites: Geological Society of London Special Publication,
v. 17, p. 629–644.
Stamatakis, M.G., 1995, Occurrence and genesis of huntite—hydromagnesite assemblages, Kozani, Greece—important new white fillers and extenders: Institution of Mining and Metallurgy Transactions, v. 104, sec. B, p.
B179–B186.
Ten Dam, A., and Erentöz, C., 1970, Kızıldere geothermal field-western Anotolia: U.N. Symposium on the Development and Utilization of Geothermal Reources, Pisa, Special Issue, v. 2, pt. 1, p. 124–129.
Tezcan, A.K. 1979, Geothermal studies, their present status and contribution
to heat flow contouring in Turkey, in Cermák, V., and Rybach, L., eds., Terrestrial heat flow in Europe: Berlin, Springer-Verlag, p. 283–292.
Thompson, J.B., and Ferris, F.G., 1990, Cyanobacterial precipitation of gypsum, calcite, and magnesite from natural alkaline lake water: Geology, v. 18,
p. 995–998.
Udias, A., 1985, Seismicity of the Mediterranean basin, in Stanley, D.J., and
Wezel, F.C., eds., Geological evolution of the Mediterranean basin: Berlin,
Springer-Verlag, p. 55–64.
Walter, M.R., Golubic, S., and Preiss, W.V., 1973, Recent stromatolites from
hydromagnesite and aragonite depositing near the Coorong lagoon, South
Australia: Journal of Sedimentary Petrology, v. 43, p. 1021–1030.
Wetzenstein, W., 1975, Limnische huntit-hydromagnesit-magnesit-lagerstätten in Mazedonien/Nordgriechenland: Mineralium Deposita, v. 10, p.
129–140.
Zachmann, D.W., and Johannes, W., 1989, Cryptocrystalline magnesite: A review, in Möller, P., ed., On the formation of magnesite: Berlin, Gebruder
Borntraeger, v. 28, p. 15–28.
Zedef, V., 1994, The origin of magnesite in Turkey, a stable isotope study: Unpublished Ph.D. thesis, Glasgow, UK, Glasgow University, 159 p.
445
0361-0128/98/000/000-00 $6.00
446