Oxygen isotopic heterogeneities of

metamorphic rocks: an original

tectonostratigraphic signature, or an

imprint of exotic fluids? A case study of

Sifnos and Tinos islands (Greece)

European Journal of Mineralogy 8 , 719-732. 1996

Author(s):

Jiwchar Ganor

Department of Geological and Environmental Sciences, Ben-Gurion

University of the Negev.

P. O. Box 653

84105 Beer Sheva, Israel

e-mail: ganor@bgumail.bgu.ac.il

Phone: +972-7-6472651

FAX: +972-7-6472997

Alan Matthews

Institute of Earth Sciences, The Hebrew University of Jerusalem

Jerusalem 91904, Israel. E-mail: alan@vms.huji.ac.il

Phone: +972-2-658-4913

FAX : +972-2-566-2581

Manfred Schliestedt

Institut f�r Mineralogie, Universitהt Hannover,

Welfengarten 1,

30167 Hannover, Germany

Zvi Garfunkel

Institute of Earth Sciences, The Hebrew University of Jerusalem

Jerusalem 91904, Israel. E-mail: zvi_garf@cc.huji.ac.il

Phone: +972-2-658-4678

FAX : +972-2-566-2581

Abstract:

Oxygen isotope compositional heterogeneities among Eocene high P/T

rocks and retrograde overprinting assemblages on the Cycladic islands of

Sifnos and Tinos are studied with the aim of determining he extent to which

they reflect an original tectonostratigraphic signature or the effects of fluid

infiltration. Plots of whole-rock 18O values against the chemical index of

Garlick show that it is possible to interpret regional isotopic compositional

variation in terms of either exchange with an 18O-enriched fluid during

retrograde eclogite to greenschist- facies transformations or an originally

heterogeneous pre-metamorphic sequence of rock units. However, a detailed

analysis of isotopic compositional variations on an outcrop scale shows that

there are no identifiable 18O changes accompanying the retrograde

metamorphic transformations, thus favouring the latter alternative. The

isotopic data support the view that outcrop- and layer-scale variations in the

degree of retrograde metamorphic transformation were controlled by

selective infiltration (possibly involving local deformation-enhanced

permeability creation) of small amounts of fluids Simple mass-balance

calculations based on the stoichiometries of model eclogite-blueschist and

blueschist-greenschist reactions suggest that these amounts of fluid were of

the order of several weight percent. The isotopic compositional

heterogeneity of regional metamorphic sequences must be taken into

account in when applying models of fluid infiltration.

Key-words: regional metamorphism, oxygen isotopes, fluid infiltration,

Cyclades.

Introduction

The extent to which fluids are involved in metamorphism is still in debate

and constitutes one of the more important questions in understanding

regional metamorphism. In the last two decades a number of papers have

addressed different aspects of fluid involvement in regional metamorphism.

Variation in isotopic composition of rocks from different metamorphic

grades has been used as means to determine the amount of fluid that

infiltrated into a terrane during metamorphism, and the composition of the

fluid. A basic assumption in several of these studies is that all rocks have the

same pre-metamorphic isotopic composition. However, regional

metamorphic complexes are often made up of sequences of stacked thrust

sheets and it is possible that the initial rock sequence was isotopically

heterogeneous. The detailed study of the stable isotope compositions of

pelites, amphibolites and marbles on Naxos showed that the regional

isotopic pattern largely reflected the presence of a metasedimentary series

overlying an older basement (Baker & Matthews, 1995). This paper

examines in detail these two alternatives of original isotopic signature

versus fluid imprint in the case of retrograde (eclogite to greenschist-facies)

metamorphism of the islands of Sifnos and Tinos in Cycladic complex of

Greece (Fig. 1). The amount of fluid involved in metamorphism can be

expressed in terms of a time integrated fluid / rock ratio which defines the

time-integrated amount of fluid that has reacted with the rock. In many

studies the evaluation of fluid/rock ratios (F/R) in metamorphism has been

done using the mass-balance equations of Taylor (1977). For a closed

system (cs): (1)

,

where f and i are, respectively, the final and initial isotopic composition

for the element of interest, F and R are the atomic fraction of the element of

interest relative to the entire fluid/rock system (i.e. F+R = 1) and =f(fluid) -

f(rock). For an open system (os):

. (2)

,

The final isotopic composition of the rock is the only term in the equation

that is obtained by direct analysis. The value of is usually calculated with

the assumption that the final fluid and rock are in isotopic equilibrium at an

assumed metamorphic temperature. The initial isotopic composition of fluid

and protolith are unknown quantities and therefore must be assumed or

indirectly inferred. The validity of the fluid/rock ratio calculation is clearly

dependent on the accuracy of these assumptions. The critical assumption,

essential for the estimation of the initial isotopic fluid composition and the

fluid/rock ratio, is the composition of the protolith (i.e., the rock before

metamorphism and/or before fluid infiltration). In some studies, the protolith

is assumed to have a "normal" or average isotopic value of a sedimentary,

magmatic or metamorphic rock, for example: average marble (Bebout &

Carlson, 1986), or marine carbonate (Ganor et al., 1989). In many other

cases, the isotopic composition of low-grade metamorphic (or non-

metamorphic) rocks from the same area is assumed to represent the protolith

composition (e.g., Rumble & Spear, 1983; Hoernes & Hoffer, 1985;

Wickham & Taylor, 1985; Wickham & Taylor, 1987; Rye & Bradbury,

1988). Using an approach based on the open-system mass balance equations

of Taylor (1977), Schliestedt & Matthews (1987) calculated water-rock

ratios of 0.2-0.4 for the eclogite/blueschist to greenschist transformation on

the Cycladic island of Sifnos (Fig. 1a). Their basic assumption was that the

greenschists of Central Sifnos (the lowest stratigraphic unit) originally

possessed the same isotopic compositions as unaltered blueschists and

eclogites in Vroulidia Bay, northern Sifnos (higher in the section, Fig. 1a).

Differences in isotopic composition between eclogites, blueschists and

greenschists (minerals and whole-rocks) were considered to be due to

exchange between rocks and an upward infiltrating fluid during overprinting

reactions. Because the greenschists of Central Sifnos showed higher 18O

values than eclogites and blueschists of Vroulidia, Matthews & Schliestedt

(1984) proposed that the infiltrating fluids were enriched in 18O by

exchange with marbles. However, the suggestion of Avigad (1993) that the

Central Sifnos greenschist unit is separated from the overlying units by a

late tectonic contact, means that a continuous stratigraphy with uniform

protolith composition cannot be assumed. Moreover, isotopic analyses of

blueschist and greenschist minerals from the nearby island of Tinos (Fig.

1b) did not reveal the 18O differences found on Sifnos (Brš cker et al 1993)

In this paper we examine the hypothesis that the isotopic compositions of

eclogites, blueschists and greenschists from Sifnos and Tinos largely reflect

protolith values, rather than compositions modified by infiltration. Isotopic

analysis of rocks from Tinos and Sifshow that marbles away from contacts

ware characterized by oxygen and carbon isotopic compositions similar to

that of sedimentary marine carbonates (Ganor et al., 1989; 1991) In contrast,

marbles in contact with schists demonstrate low 18O values due to

diffusional exchange with schists in a narrow margin (less than 1 m wide)

(Ganor et al., 1991). The isotopic composition of the schists may reflect

either the isotopic composition of the protolith (which is regarded here as

isotopic composition of the rock prior to the retrograde metamorphism), or

exchange with infiltrating fluids that invaded the sequence during

exhumation. To avoid difficulties associated with diffusional exchange with

the marbles, we will consider only analyses of schists that were sampled at

least 1 m away from contact with a marble layer.

Analytical methods

Whole rock silicates were reacted with bromine pentafluoride at 550°C for

16 hours (Clayton & Mayeda, 1963) and the released oxygen converted to

carbon dioxide. The CO2 was isotopically analyzed on a VG Micromass ES

mass spectrometer. Analyses of 35 NBS 28, international standard (quartz)

gave 9.46‰ with accuracy (1s) of ±0.3‰. Replicates were made of all analyses. 18O analyses are reported relative to the SMOW scale. Whole-

rock chemical analyses were made using X-ray fluorescence for all elements

except H2O, CO2 and FeO. H2O was determined using Carl- Fischer

titration, CO2 by infrared spectroscopy, and FeO by oxidimetric methods.

Procedures have been described in detail by Schliestedt (1980).

Geological setting

The islands of Sifnos and Tinos are situated in the northwestern part of the

Cycladic crystalline massif (Fig. 1). Both islands feature lithological units

that have undergone Eocene, Alpine, collision-related, high-pressure

metamorphism and subsequent overprinting in the greenschist facies during

exhumation in the Oligocene - Miocene. The geology of Sifnos is outlined

in Fig. 1a (after Avigad, 1990). An approximately 2.5 km thick sequence of

metamorphic rocks is exposed, which has been divided by Avigad (1993)

into two major units separated from each other by a flat-lying tectonic

contact which post-dates metamorphism. The lower unit is the 1000 m thick

Central Sifnos Greenschist unit which mainly consists of marbles and

metapelites in its lower parts and basic metavolcanics in its upper parts.

Metamorphic assemblages in this unit are dominated by the greenschist-

facies overprint. Conditions of greenschist metamorphism have been

estimated between 400- 500°C and 5-7 kbar (Matthews & Schliestedt, 1984;

Schliestedt & Matthews, 1987). The overlying Eclogite - blueschist unit

consists of three sequences. The 800 m thick Main Marble sequence at the

base of the unit is composed of calcite and dolomite marbles with

intercalations of eclogites, blueschists, and acid gneisses overprinted by

greenschist- facies metamorphic assemblages (Schliestedt & Matthews,

1987). Overlying this is a ca. 400 m thick sequence of schists and gneisses

consisting of interlayered eclogites, blueschists, and jadeite gneisses with

minor metasediments and marbles. This is overlain by the 300 m thick

Upper Marble sequence consisting of high-pressure metamorphosed calcite

and dolomite marbles with layers of quartzites, schists, and gneisses.

Temperature and pressure conditions for the eclogite-facies metamorphism

have been estimated at 450° to 500°C and 15±3 kbar (Matthews &

Schliestedt, 1984; Schliestedt & Matthews, 1987). Samples were taken from

the following localities (Fig. 1a): 1) Vroulidia Bay, within the schist-gneiss

sequence of the Eclogite-Blueschist unit; 2) Kamares Bay, within the Main

Marble sequence; 3) various locations within the Central Sifnos Greenschist

unit. The geology of Tinos is outlined in Fig. 1b (after Avigad, 1990;

Melidonis, 1980). The stratigraphy was subdivided by Melidonis (1980) into

two major units: a Lower Metamorphic unit consisting of Eocene eclogite-

facies rocks mostly overprinted by Oligocene- Miocene greenschist facies

metamorphism (Brצcker, 1990; Brצcker, et al., 1993); an Upper Tectonic

unit whose dominant component is a possible Late Cretaceous dismembered

ophiolite sequence metamorphosed under greenschist- to amphibolite- facies

conditions (Katz et al., 1995). This unit has not undergone the high-pressure

metamorphism of the Lower unit and is considered part of an upper plate

tectonically juxtaposed onto the latter (Avigad & Garfunkel, 1991). Both

units are intruded by a Miocene monzogranite, which forms part of a

regional granitoid intrusion phase characterizing the post metamorphic

orogenic evolution of the Cyclades (Altherr et al., 1982; 1988). The Lower

Metamorphic unit is up to 2 km thick and includes three marble horizons.

The lowest marble (M1) is at the base of the unit and has been proposed by

Avigad & Garfunkel (1989) to be a separate tectonic unit (Basal Unit)

overthrust by the Eocene metamorphic sequence. The M2 marble layer is

located in the middle of the Lower metamorphic unit in the region of the

village of Isternia in central Tinos. The M3 marble outcrops near the top of

the unit and is particularly exposed as a series of marble layers interlayered

with well preserved high P/T and overprinted rocks in the area of Karla-

Kionia in the southwest of the island. Sample locations in this study were in

the M2 and M3 marble regions of Isternia and Karla-Kionia, respectively

(Fig. 1b). The terms eclogite, blueschist and greenschist are used in the

literature both as names of metabasic rocks and as names of metamorphic

facies. This multiple usage may result in confusion, especially in areas such

as the Cycladic Massif where the Eocene high P/T blueschists are cofacial

with eclogite rocks (Schliestedt, 1986). The following definitions will be

used in this study. Rock-types are generically classified as metavolcanic or

metasedimentary. Metavolcanics include basic types such as eclogites

(garnet-omphacite), blueschists (glaucophane-garnet- epidote), Ab-

blueschists (glaucophane-albite-epidote), greenschists (albite- actinolite-

chlorite- epidote) and acidic gneiss types (jadeite gneisses, albite gneisses).

Metasedimentary types include marbles, garnet-glaucophane micaschist,

impure marbles and quartzites. Eocene eclogites and blueschists (and all

rocks that coexist with them) are referred to in this paper as eclogite-facies

assemblages. All rocks in which the retrograde overprint is complete are

referred to as greenschist-facies assemblages. There are a variety of rocks

that exhibit varying degrees of retrograde transformation. These include

overprinted eclogites containing relict high P/T minerals, albite epidote

blueschists (Avigad et al., 1992) and interlayered blueschist- greenschist

rocks. These will be referred to as transitional assemblages.

Results

Isotopic and chemical compositions from this study together with data from

Schliestedt & Matthews (1987), Matthews &Schliestedt (1984) and

Schliestedt (1980) are given in Tables 1 and 2. The effects of chemical

compositional variation on the isotopic fractionation behaviour of whole-

rock samples is expressed by the Garlick (1966) chemical index (I), which is

defined as: (3)

,

A higher Garlick index indicates rocks that contain more silica and alumina

(i.e., are more acidic). The use of Garlick index versus 18O plot as a

procedure for determining if a series of whole-rock samples are in isotopic

equilibrium was suggested and described in detail by Matthews &

Schliestedt (1984), Schliestedt & Matthews (1987) and Ganor et al. (1994).

Rocks in isotopic equilibrium with one another should show a simple linear

trend of 18O increasing with I. The slope of this line is proportional to

106/T-2. The 18O values are classified in two ways: 1) in terms of the

metamorphic facies (as defined above) and 2) in terms of the lithological

units in which the samples are located. The isotopic results are plottein Figs.

2 and 3 as a of tGarlick index. The straight line represents the calculated

slope of the 18O vs. I plot at 450°C, a representative temperature for the

retrograde metamorphism on Tinos and Sifnos (Schliestedt & Matthews,

1987). The following general observations are indicated from these plots: 1.

A general trend of increasing 18O with I is observed in most units and for

most facies. 2. Greenschist-facies and transitional assemblage rocks

generally have higher 18O values than eclogite-facies rocks (Figs. 2a, 3a).

3. On Sifnos, rocks from the Main Marble unit (Kamares Bay) have

higher18O values than those from the high P/T Eclogite-Blueschist unit

(Vroulidia Bay). Central Sifnos rocks cover a wide range of 18O values that

are mostly higher or overlap with Kamares Bay rocks (Fig. 2b). 4. Rocks

from the Isternia area on Tinos have higher 18O values than generally

higher-grade rocks from Karla-Kionia (Fig. 3b). 5. The oxygen isotope

composition of rocks from Isternia, Tinos overlap with those of Kamares

Bay, Sifnos, whereas Karla-Kionia results overlap the results from Vroulidia

(Fig. 4). These large-scale observations are more closely examined by

consideration of the 18O vs. I trends within individual sampling sections

(Fig. 5). Sifnos The oxygen isotopic compositions of the rocks from

Vroulidia Bay, Sifnos (Fig. 5a) generally cover a narrow range. 18O of

most eclogites and blueschists are in a range of 9.8-12.7 ‰, whereas the meta-acidic rocks and metasediments range between 12.6 to 14.3 ‰. A

limited 18O range is also observed for rocks at Kamares Bay (Fig. 5b). One

of the interlayered transitional samples at Kamares (Si-87-421) was cut into

two parts; one composed of blueschist minerals and the other of greenschist

minerals. The two parts show similar oxygen isotopic compositions (11.9‰ and 12.5‰ for the blueschist and greenschist parts, respectively). The18O

vs. I trends at both Kamares and Vroulidia plot parallel to the calculated

450°C isotherm. The oxygen isotopic results of samples from the upper

(mostly metavolcanic) parts of the Central Sifnos unit cover an

approximately 5 ‰ range (Fig. 6). There are no observable isotopic

compositional differences between transitional assemblages and

greenschists. The highest 18O values observed in greenschist- facies

samples are for metasediments (Table 1). The results at each of the six

sample localities demonstrate clearly defined range of compositions that

differ from one site to the other. Tinos The rocks from the Karla-Kionia

region (Fig. 5c, Table 2) were sampled from three different outcrops.

Outcrop #1 is composed of eclogite-facies rocks. The rocks at this site have

a limited range of oxygen isotopic compositions (9.3 to 10.7 ‰). The 25 m thick section sampled at outcrop #2 is mostly composed of basic and acid

metavolcanic greenschist-facies rocks. The upper 0.5 m of these section is

composed of vertical and lateral transitions between blueschist- and

greenschist-facies metavolcanics. These samples (dotted field in Fig. 5c)

have lower 18O values than measured for greenschist-facies metavolcanics

lower in the section (densely shaded field). One of the interlayered

blueschist-greenschist samples (Ti- 87-108) was cut into two parts. Similar

oxygen isotopic compositions were observed in the blueschist and

greenschist parts (8.95 ‰ and 8.25 ‰, respectively). The last outcrop (#3) is

composed of eclogite-facies rocks ׀ blueschists, eclogites and acid

gneisses ׀ and shows a broader range of oxygen isotopic compositions than

is consistent with equilibration between coexisting rocks. Fig. 5d plots 18O

values at the Isternia locality. There are no observable differences between

the different types of samples. In an interlayered blueschist- greenschist

sample (DA-659) which was cut into two parts, the blueschist and

greenschist parts have similar oxygen isotopic compositions (Fig. 5d).

Discussion

The general trends of the 18O vs. I plots (Figs. 2, 3) demonstrate that

oxygen isotopic compositions can possibly be correlated either with the

lithological unit from which they are sampled, or with metamorphic facies,

i.e., the degree of metamorphic overprint. Isotopic equilibration within a

particular metamorphic facies is also suggested by the systematic trends

of 18O with I observed at most localities. This possible relation between

lithological site and 18O and between metamorphic facies and 18O raises

two alternative explanations for the association of higher oxygen isotopic

compositions with greenschist-facies rocks. First, the possibility that this

association was determined by equilibration between an isotopically "heavy"

fluid and the rocks during the eclogite to greenschist facies transformation

(Schliestedt & Matthews, 1987) or second, that the transformation reaction

predominantly affected tectonostratigraphic units which originally possessed

"heavier" isotopic compositions. This means that units most strongly

affected by the greenschist- facies overprint originally had higher18O

values than units that were less affected by this overprint. In the following

paragraphs we will show that the latter explanation is more probable. In

comparison with the general picture showing that greenschist- facies rocks

generally have higher 18O than the eclogite facies rocks, an examination of

individual sites shows that oxygen isotopic compositions are independent of

the metamorphic grade. For example, eclogite-facies rocks, transitional

assemblages and greenschist-facies rocks at both Isternia and Kamares have

similar 18O values (Figs. 5a, d). In contrast, the oxygen isotopic

compositions of the greenschist-facies rocks at Karla-Kionia cover a wide

range, that is both "lighter" and "heavier" than for the eclogite-facies rocks

(Fig. 5b). Blueschist and greenschist parts separated from inter- laminated

blueschist-greenschist rocks have similar 18O values. These results

demonstrate that the eclogite to greenschist transformation does not involve

any change in 18O at hand specimen to outcrop scale. This argues against

the possibility that the isotopic compositional differences are due to the

infiltration of an isotopically "heavy" fluid during the eclogite to greenschist

facies transformation. The similarity of the 18O values of rocks from

Isternia with those of Kamares, and from Karla-Kionia with those of

Vroulidia has been noted (Fig. 4). Both Isternia and Kamares Bay sequences

are dominantly composed of greenschist- facies and transitional

assemblages, and unaltered high-pressure rocks are less common. In

contrast, high pressure-assemblages are common in the Karla-Kionia area

and are dominant at Vroulidia. The stratigraphic settings are also similar; on

both islands the eclogite facies units (Karla-Kionia and Vroulidia) are

located higher in the stratigraphy. The correspondence in isotopic

composition, lithology, and stratigraphic position of these units reinforces

the argument that the transformation reactions predominantly affected units

which originally possessed "heavier" isotopic compositions. The data may

also suggest that Isternia unit may be correlative to Kamares unit and Karla-

Kionia is correlative to Vroulidia. The arguments above favor the

interpretation that18O of rocks primarily reflect the original isotopic

composition of tectonic units or sub-units in which they were located. In this

view, the higher 18O values in the more overprinted units simply reflect the

fact that they were originally richer in 18O. The reasons for the higher 18O

values are not yet fully understood, but may be possibly be correlawith

larger amounts of metasedimentary rock components in these units relative

to thelower 18O se. The preservation of an original, pre-metamorphic,

isotopic composition raises the question as to how much fluid was involved

in the eclogite to greenschist transformation. Brצcker (1990) showed that

greenschist and blueschist rocks from Tinos have similar chemical

compositions and concluded that the blueschist-to- greenschist transition

was not related to differences in the bulk chemistry of the parental basic

rocks. Similarly, an examination of the whole-rock geochemical data for

Sifnos, presented in Table 1, also shows that there are no chemical

compositional differences between greenschists and high P/T rocks.

Calculated phase equilibria show that the fluid responsible for the

greenschist overprint must have been water- rich (Schliestedt & Matthews,

1987; Ganor, 1991). Thus, the most reasonable explanation for the

interlayering of blueschists and greenschists lies in selective water supply

during retrograde metamorphism (Brצcker, 1990). However, it is illogical

that some segments of a layer were exposed to large scale fluid infiltration,

whilst adjacent segments with similar composition and texture remained

dry. This suggests that the amount of water involved in the eclogite to

greenschist-facies transformation was minor. A limited amount of fluid

whose transport was controlled by small permeability variations would have

enabled the partial to complete transformation of rocks to have occurred

adjacent to rocks that remained unaltered. Large-scale infiltration of fluid

may also be rejected by the lack of isotopic equilibrium observable on a

scale larger than hand specimen. Infiltration of large amounts of fluid should

equilibrate isotopic compositions on a larger scale. On the other hand, a

minimum amount of water is required to allow the eclogite to greenschist

transformation reactions. The overall transformation can be regarded in

terms of two model net transfer reactions involving the breakdown of

eclogite to blueschist and of blueschist to greenschist. The amounts of water

required for these reactions can be determined by simple mass-balance

calculation. The eclogite to epidote-blueschist transformation in metabasic

rocks can be represented by the idealized end-member reaction (Evans,

1990), (4)

Mass-balance calculation shows that, if the reactant rock contains the

minerals in the exact stoichiometric ratio, 2.2 wt% of water are required for

the reaction to go to completion. For a detailed description of the method of

calculation see Ganor (1991). Similarly, the reaction: (5)

is the classical blueschist-greenschist transformation reaction (Schliestedt &

Matthews, 1987; Evans, 1990). For this reaction mass-balance calculation

shows that, if the reacting assemblage contains the minerals in the exact

stoichiometric ratio, 1.1 wt% of water are required for complete reaction.

These calculated values of 1.1 and 2.2 wt% water, based on the model

transformation reactions, are in good agreement with measured differences

in water content of 1 to 3% determined by whole-rock chemical analyses of

eclogites, blueschist and greenschists (Table 1, Brצcker, 1990; Ganor,

1991). These values are minimum amounts of water that are needed to

develop the greenschist assemblages during the exhumation of the eclogite-

facies rocks. The arguments presented above favouring selective infiltration

of small amounts of water raise the question as to what physical factors

controlled this selectivity. The absence of conspicuous fracture networks in

greeenschist-facies rocks and the general presence of fabrics with mimetic

overgrowths of undeformed albite on earlier high-pressure minerals

(Avigad, 1993) argue against deformation- enhanced permeability creation.

However, Avigad et al (1988) have shown that the true picture may be more

complex. In the interlayered blueschist -greenschist rocks at Isternia,

blueschists and transitional assemblages are dissected by nets of

anastamosing micro shear zones. These networks are absent in greenschists,

whose fabric is static and undeformed. In transitional assemblages, albite

and chlorite are observed to have preferentially grown in the shear zones.

Thus, it would appear that the initiation of the greenschist overprint occurs

under conditions of local deformation-enhanced permeability increase which

allowed fluids to preferentially enter the rocks. The relation between this

mechanism and the general observation that the more heavily overprinted

units are originally 18O-richer lithologies remains to be identified.

Conclusions

The distribution of isotopic values on the scale of the major lithological

units on the islands of Sifnos and Tinos can equally well be explained in

terms of pervasive infiltration of an 'exotic' fluid during the retrograde

(eclogite to greenschist) metamorphism or original (pre-metamorphic)

signatures. In contrast, at the scale of outcrop and hand specimen, we

observe local heterogeneities and the absence of 18O differences between

eclogite- and greenschist-facies rocks that demonstrate the rocks were not

exposed to large-scale fluid infiltration. This duality appears to be common

in regional metamorphic terranes: oxygen isotope analysis on different

scales may lead to different conclusions (see also Baker & Matthews, 1995).

The original isotopic compositional heterogeneity of regional metamorphic

sequences must be taken into account when applying models of fluid

infiltration. In the case of the Cyclades, the outcrop-scale retrograde

transformations were undoubtedly controlled by selective infiltration of

small amounts of fluids. This limited amount of fluid would have enabled

the partial to complete transformation of rocks to have occurred adjacent to

rocks that remained unaltered. Large scale infiltration of fluid may also be

rejected by the lack of isotopic equilibrium observable on a scale larger than

hand specimen. Simple mass- balance calculations based on stoichiometries

of eclogite-blueschist and blueschist- greenschist reactions suggests that

these amounts of fluid were of the order of several weight percent.

Acknowledegments. We wish to express our gratitude to Dr. Dov Avigad

for his advice on the tectonic, structural and petrographical aspects of this

study. Ms. Nadia Teutsch, Ms. Varda Gur, Mr. Ari Matmon, Ms. Ori Gonen

and Ms. Rivka Nissan performed most of the isotope analyses. The research

was supported by a grant from the German-Israel Fund for Scientific

Research and Development, Jerusalem, Israel. J. Ganor would like to thank

the Lady Davis Fellowship Trust for their support during the period of

writing of this manuscript. Permission for field work in Greece was granted

by the director of the I.G.M.E in Athens.

References

Altherr, R., Henjes-Kunst, F.J., Matthews, A., Firedrichsen, H., Hansen,

B.T. (1988): O-Sr isotopic variations in Miocene granitoids from the

Aegean: evidence for an origin by combined assimilation and fractional

crystallization. Contrib. Mineral. Petrol., 100, 528-541.

Altherr, R., Kreuzer, H., Wendt, I., Lenz, H., Wagner, G.A., Keller, J.,

Harre, W., Höhndorf, A. (1982): A late Oligocene / early Miocene high

temperature belt in the Attic-Cycladic crystalline complex (SE Pelagonian,

Greece). Geol. Jahrb., E23, 97-164.

Altherr, R., Schliestedt, M., Okrusch, M., Seidel, E., Kreuzer, H., Harre, W.,

Lenz, H., Wendt, I., Wagner, G.A. (1979): Geochronology of high pressure

rocks on Sifnos (Cyclades, Greece). Contrib. Mineral. Petrol., 70, 245-255.

Avigad, D. (1990): Geodynamic evolution of the Cyclades complex: a

contribution to the study of continental collision. Ph.D. dissertation, the

Hebrew University of Jerusalem.

Avigad, D. (1993): Tectonic juxtaposition of blueschists and greenschists in

Sifnos island (Aegean sea) - implications for the structure of the Cycladic

blueschist belt. J. structural Geology, 15, 1459-1469.

Avigad, D., Evans, B.W., Matthews, A., Garfunkel, Z. (1992): Metamorphic

evolution of blueschist terrains - cooling during exhumation, Sifnos

(Cyclades) Greece. Eur. J. Mineral., 4, 619-634.

Avigad, D. & Garfu, Z. (1989): Low-angle faults above and below a

blueschist belt - Tinos island, Cyclades, Greece. Terra Nova 1, 182-187.

Avigad, D., Matthews, A., Garfunkel, Z. (1988): The Alpine evolution of

the Cyclades: introduction and excursion guide to the geology of the islands

of Sifnos and Tinos. Geol. Soc. Israel. Spec. Publ, Jerusalem, 66p.

Baker, J. & Matthews, A (1995): The stable isotopic evolution of a

metamorphic complex, Naxos, Greece. Contrib. Mineral. Petrol., 120, 391-

403.

Bebout, G.E. & Carlson, W.D. (1986): Fluid evolution and transport during

metamorphism: evidence from the Llano Uplift, Texas. Contrib. Mineral.

Petrol., 92, 518-529.

Bröcker, M. (1990): Blueschist-to-greenschist Transition in metabasites

from Tinos island (Cyclades, Greece): Compositional control or fluid

infiltration? Lithos, 25, 25-39.

Bröcker, M., Kreuzer, H., Matthews, A., Okrusch, M. (1991): 40Ar/39Ar and

oxygen isotope studies of polymetamorphism from Tinos island, Cycladic

blueschist belt, Greece. J. metamorphic Geol., 11,223-240.

Clayton, R.N. & Mayeda, T.K. (1963): The use of bromine pentafluoride in

the extraction of oxygen from oxides and silicates for isotopic

analysis. Geochim. Cosmochim. Acta, 27, 43-52.

Evans, B.W. (1990): Phase relations of epidote-blueschists. Lithos, 25, 3-23.

Ganor, J. (1991): The influence of fluids on metamorphism in a plate

collision region (The Cycladic Massif, Greece). PhD dissertation, the

Hebrew University of Jerusalem.

Ganor, J., Matthews, A., Paldor, N. (1989): Constraints on effective

diffusivity during oxygen isotope exchange at a marble-schist contact,

Sifnos (Cyclades), Greece. Earth Planet. Sci. Lett., 94, 208-216.

Ganor, J., Matthews, A., Paldor, N. (1991): Diffusional isotopic exchange

across an interlayered marble-schist sequence with an application to Tinos,

Cyclades, Greece. J. Geoph. Res. 96, 18,073-18080.

Ganor, J., Matthews, A., Schliestedt, M. (1994): Post metamorphic low 13C

veins in the Cycladic metamorphic complex (Greece) and their implications

for modeling fluid infiltration processes using carbon isotope

compositions. Eur. J. Mineral., 6, 365-379.

Garlick, G.D. (1966): Oxygen isotope fractionation in igneous rocks. Earth

Planet. Sci. Lett., 1, 361-368.

Hoernes, S. & Hoffer, E. (1985): Stable isotope evidence for fluid-present

and fluid-absent metamorphism in metapelites from the Damara Orogen,

Namibia. Contrib. Mineral. Petrol. 90, 322-330.

Katz, Y., Matthews, A., Garfunkel, Z., Schliestedt, M. (1995): The tectono-

metamorphic evolution of a dismembered ophiolite (Tinos, Cyclades,

Greece). Geol. Mag., In submission

Matthews, A. & Schliestedt, M. (1984): Evolution of the blueschist and

greenschist facies rocks on Sifnos, Cyclades, Greece: A stable isotope study

of subduction-related metamorphism. Contrib. Mineral. Petrol., 88, 150-

163.

Melidonis, N.G. (1980): The geological structure and mineral deposits of

Tinos island (Cyclades, Greece). in "The Geology of Greece", vol. 13,

Institute of Geology and Mineral Exploration, Athens, 1-80.

Rumble, D.I. & Spear, F.S. (1983): Oxygen-isotope equilibration and

permeability enhancement during regional metamorphism. J. Geol. Soc.

London, 140, 619-628.

Rye, D.M. & Bradbury, H.J. (1988): Fluid flow in the crust: an example

from a Pyrenean thrust ramp. Am. J. Sci., 288, 197-235.

Schliestedt, M. (1980): Phasengleichgewichte in Hochdruckgesteinen von

Sifnos, Griechenland. Ph D dissertation, Technischen Universität,

Braunschweig.

_____________ (1986): Eclogite-blueschist relationships as evidenced by

mineral equilibria in the high-pressure metabasic rocks of Sifnos (Cycladic

Islands), Greece. J. Petrol., 27, 1437-1459

Schliestedt, M. & Matthews, A. (1987): Transformation of blueschist to

greenschist facies rocks as a consequence of fluid infiltration, Sifnos

(Cyclades), Greece. Contrib. Mineral. Petrol., 97, 237-250.

Taylor, H.P., Jr. (1977): Water/rock interactions and the origin of H2O in

granitic batholiths. J. Geol. Soc. London, 133, 509-558.

Wickham, S.M. & Taylor, H.P., Jr. (1985): Stable isotopic evidence for

large-scale seawater infiltration in a regional metamorphic terrane; the Toris

Seigneurs Massif, Pyrenees, France. Contrib. Mineral. Petrol., 91, 122-137.

Wickham, S.M. & Taylor, H.P., Jr. (1987): Stable isotopic constraints on the

origin and depth of penetration of hydrothermal fluids associated with

Hercynian regional metamorphism and crustal anatexis in the

Pyrenees. Contrib. Mineral. Petrol., 95, 225-268.

Tables

Figures