evolution of boudins under progressively decreasing pore pressure

EVOLUTION OF BOUDINS UNDER PROGRESSIVELY DECREASINGPORE PRESSURE – A CASE STUDY OF PEGMATITES ENCLOSED IN

MARBLE DEFORMING AT HIGH GRADE METAMORPHICCONDITIONS, NAXOS, GREECE

OLIVER SCHENK*, JANOS L. URAI†, and WOUTER VAN DER ZEE**Geologie—Endogene Dynamik, RWTH Aachen University, Lochnerstrasse 4-20,

52056 Aachen, Germany

ABSTRACT. During ductile deformation of marbles under high grade metamorphicconditions on Naxos, Greece, pegmatites enclosed in the marbles were deformed in abrittle fashion forming blocky boudins with quartz crystallized in the interboudinzones. We studied the three-dimensional geometry of the boudins in the field and intwo large samples. The first sample was serially sectioned to observe the structural andmicrostructural evolution; in the second sample we mapped the surface morphology ofthe deformed pegmatite. In profile, the boudins can be classified as symmetric torntype boudins, evolving towards asymmetric boudins with a later domino boudincomponent. In three dimensions, however, the morphology of the boudinaged pegma-tite is a simple set of normal faults with mode-I fractures in the fault tips and rotationof the fault blocks to accommodate the extension in the marble.

Deformation history was constrained by petrology and microstructures in combi-nation with simple order of magnitude calculations of both the cooling and pore fluidpressure evolution. The dynamically recrystallized, coarse-grained calcite of the marbleprovides clear evidence that after the pegmatite intruded the marble and solidified, itwas deformed at peak conditions of M2b metamorphism (� 670 °C and � 0.6 GPa).Such pegmatitic melts contain � 10 percent H2O which is released during crystalliza-tion.

We infer that after crystallization of the pegmatite the pore fluid pressure in thepegmatite remained close to lithostatic due to the very low permeability of thesurrounding marble, and the pegmatite was deformed at very low effective stress whichled to brittle deformation of feldspar and mode-I fracturing of the pegmatite, formingtorn type boudins. With time the pore fluid pressure slowly decreased, increasing theeffective stress. Ongoing N-S extension thus resulted in slip along the quartz-filledinterboudin zones and in block rotation, producing domino boudins.

introductionA marble sequence inside the high grade core on Naxos, Greece, contains an

exceptionally well-exposed field example of isolated pegmatites that intruded themarble, solidified and were deformed into boudins (fig. 1). The evolution of thesystem is constrained by microstructural and petrological data combined with measure-ments from three-dimensional outcrops of the boudinaged pegmatites.

The aim of this study is to constrain the dynamics of the deformation ofpegmatites at subsolidus conditions, to reconstruct the pore pressure evolution in thepegmatites and to give insight on deformation mechanisms, mechanical propertiesand stress conditions. In addition we document the three-dimensional geometry ofboudins in detail and model the evolution and mechanisms of boudinage.

geological settingNaxos is the largest Cycladic island in the Aegean Sea and belongs to the

Attic-Cycladic Massif (Durr and others, 1978), which forms an arcuate belt of metamor-

*Present address: IES, Integrated Exploration Systems, Aachen, Germany; [email protected]**Present address: GeoMechanics International, Mainz, Germany†Corresponding author: [email protected]

[American Journal of Science, Vol. 307, September, 2007, P. 1009–1033, DOI 10.2475/07.2007.03]

1009

phic rocks following the trend of the Hellenic trench, the current location ofNE-directed subduction of the African plate beneath the Apulian-Anatolian micro-plate (Jansen and Schuiling, 1976; Keay and others, 2001).

Naxos is dominated by a N-S trending elongated structural dome (fig. 2) andconsists of a migmatitic gneiss core, containing rafts of marble, amphibolite, felsicschists and pegmatite intrusions, which is overlain by a series of Mesozoic metasedi-ments (predominantly metacarbonates and metapelites). These Lower Plate rockswere affected by at least two Alpine regional tectono-metamorphic events (Urai andothers, 1990). The early compressional tectonic phase during the Eocene (D1) ended50 to 40 Ma ago and involved subduction of continental margin material, generation ofa nappe pile and regional high-pressure, low-temperature (HP-LT) metamorphism(M1; with T � 400 to 460 °C and p � 0.7 to 0.9 GPa) (Feenstra, ms, 1985; Avigad, 1998).This phase was followed by a period of extensional tectonics (D2) in early Miocene(Urai and others, 1990), associated with the formation of a back-arc, thinned crust,high heat flow, rapid uplift of lower crustal rocks and intrusion of granitoid magmas(Pe-Piper and others, 1997), probably related to the southward retreat of the N-dipping subduction zone south of Crete (Urai and Feenstra, 2001). The D2-phase wasaccompanied by regional greenschist facies metamorphism of Barrovian character(M2a; � 25 Ma) and reached amphibolite facies in the core, associated with partialanatexis (M2b; � 20 to 16 Ma with T � up to � 670 °C and p � 0.6 GPa) (Jansen andSchuiling, 1976; Buick and Holland, 1989; Buick, 1991a, 1991b) and the production ofthermal domes and closely spaced isograds (fig. 2). Partial melting inside the leuco-gneiss core was associated with the intrusion of pegmatite bodies (Andriessen andothers, 1991; Matthews and others, 2002, 2003).

During M2 metamorphism the metamorphic core complex was strongly deformedin a major crustal shear zone (Lister and others, 1984) that is now visible as a ductile

Fig. 1. Pegmatite dike and thin, sinusoidal amphibolite layer inside coarse grained marble from thehigh grade core of Naxos, Greece. The pegmatite is fractured and the boudins are rotated and displaced.Width of image is approximately 30 cm.

1010 O. Schenk and others—Evolution of boudins under progressively decreasing pore pressure—

carapace separating the core from the Mesozoic metasediments (Keay and others,2001), syn-M2 quartz mylonites (Krabbendam and others, 2003) and km-scale isoclinalfolds with fold axes trending N-S. These folds are coaxially refolded by open, uprightfolds showing evidence for E-W shortening (Urai and others, 1990) and truncated by

Fig. 2. Simplified geological map of Naxos, Greece (after Jansen and Schuiling, 1976; Urai and others,1990) with a detailed overview of the quarry ‘LB Naxos Marble’ in the high grade core, near Kinidaros(coordinate system: WGS 84).

1011A case study of pegmatites enclosed in marble deforming at high grade . . .

granites emplaced at, or shortly after, the M2b peak. The folding, especially in themarble horizons, was accompanied by intense boudinage, pointing to a component ofsub-horizontal extension parallel to the fold axes (Buick, 1991a).

With decreasing temperatures during further extension and uplift, the deforma-tion was strongly localized in narrow post-M2b mylonites. They are commonly parallelto the local orientation of the bedding or older high-grade schistosity, characterized byextreme grain size reduction and predominantly indicate Upper Plate-to-the-northmovement (Urai and others, 1990; Buick, 1991a, 1991b; Schenk and others, 2005).

The M2 metamorphic event terminates with the emplacement of the westernNaxos granodiorite body (� 14 to 10 Ma) (Koukouvelas and Kokkalas, 2003).Unmetamorphosed Upper Plate rocks are only exposed in E- and W-Naxos, juxtaposedabove metamorphic Lower Plate rocks or the western granodiorite along detachmentfaults (John and Howard, 1995). The rapid uplift led to the exposure of Lower Platerocks approximately 5 Ma before present (Wijbrans and McDougall, 1988).

observations

Field ObservationsThe field area is located in a marble raft inside the Naxos high grade core, close to

the village of Kinidaros at the mountain Bolibas [551 m; 37°05’12.6“ 025°28’20.0”; mapdatum: WGS84; (fig. 2)]. Here, the marble contains thin layers of amphibolite,pegmatite intrusions and some unmelted rafts of felsic schists. Due to its high purityand commercial grade, the coarse-grained calcitic marble (grain size � 15 mm) ismined by the company “LB Naxos Marble”. Flat walls and rectangular cut-offs insidethe quarry allow excellent three-dimensional observation of both boudinaged pegma-tites and folded and boudinaged amphibolites (fig. 3). Additionally, some cut blocksbroken along the marble-pegmatite contact expose the surface morphology of theboudinaged pegmatites (fig. 3D).

Inside the quarry, most amphibolites have a thickness of less than 5 cm. Theamphibolites are folded isoclinally with (sub-) vertical axial planes (average: 088/83)and fold axes dipping towards the south, up to 45°. In the limbs of the isoclinal foldsthey often show pinch-and-swell structures, overprinted by brittle boudins with similarmorphology and orientation as the boudins in the pegmatites discussed in detailbelow.

Pegmatites have thicknesses ranging from 3 to 20 cm. Most pegmatite-marblecontacts are characterized by skarn formed by metasomatic reactions between theacidic melt and marble. In the central part of the quarry the orientation of pegmatitesis quite variable, with an average orientation of 268/74, similar to that of the steeplimbs of the folds in amphibolites. The pegmatites are boudinaged into blocks with thelength (L) ranging from 10 to 30 cm. The interboudin zones are usually filled withquartz, and boudin axes typically dip towards the north, at approximately 90° to thefold axes in the amphibolites.

The outcrops show some crosscutting relationships between pegmatites andfolded amphibolites (figs. 3A and 3B). The intersections are always at the quartz-filledinter-boudin zones of the pegmatites (fig. 3B). Thin amphibolite layers at a distance ofa few centimeters to the fractures of the pegmatites have a wavy shape concordant withthe boudins (figs. 3B, 3C and 4A).

Macrostructural ObservationsThe boudin structures were studied in detail in two blocks from the quarry.

Sample 1 has a size of 25 x 30 x 90 cm and was used to study microstructures, chemistryand three-dimensional geometry of the fractures (fig. 4A). Sample 2 has a surface areaof 100 x 280 cm and was investigated with a special focus on surface geometry and


Fig. 3. Photographs and schematic drawings from the quarry: in-situ (A) – (C) and from cut blocks (D) -(F) highlighting the structural relationship between the different rock types. (A) The quarry walls offer athree-dimensional view of the pegmatite that crosscuts the folded marble-amphibolite sequence (with foldaxes trending N-S); (B) detail of the floor’s surface of (A) showing pegmatites fractured preferentially atintersections with amphibolites; (C) fractured and boudinaged pegmatite (picture looking SE); note theslight folding of the thin amphibolite layer with synclines at the height of pegmatite’s fractures; (D) exposedsurface of a boudinaged pegmatite with remnants of skarn; (E) marble block showing the boudinage ofamphibolite with early pinch-and-swell structures overprinted by later brittle fractures which are interpretedto be coeval with the boudins in the pegmatites; (F) single block (3 x 3 x 2 m) illustrating that the pegmatiteintruded parallel to the limbs of the amphibolites, and in this case into the hinge of an isoclinally foldedlayer.


Fig. 4. The samples studied in detail: (A) Mosaic of Sample 1 cut into 34 slices: this image sequenceprovides three-dimensional information on the internal evolution of the fractures inside the pegmatite; (B)in Sample 2 (100 x 280 cm) the surface of the fractured pegmatite is exposed and enables detailedmeasurements of displacement, length and dip of the fractures; (C) shows the analyzed fractures of Sample2; for the analysis only the faults filled with dark gray shade were chosen, as the faults shown in bright grayrepresent linked fracturing.


orientation of structures of the fractured pegmatite (figs. 4B and 4C), using thenomenclature of boudins from Goscombe and others (2004). By relating themeasurements to the orientation of the pegmatite occurrences in the quarry, wewere able to approximate the blocks’ original orientations. The structures aresummarized in figure 5.

Sample 1 consists of marble, pegmatite and amphibolite. It was cut into 34 sliceswith a thickness of about 2 cm. The pegmatite is 6 to 8 cm thick and boudinaged, whilethe 2 to 3 mm thick layer of amphibolite has a wavy shape following the envelopingsurface of the boudins. Close (less than 2 cm) to the contacts towards both amphiboliteand pegmatite, the color of the marble changes from white to yellow-brownish (see

Fig. 4 (continued)


also fig. 1). The intrusive contact of pegmatite and marble contains skarn with athickness of � 2 to 5 mm.

The analysis of the single slices provides information on the boudin/fracturestructure in three dimensions (fig. 6). It shows that the pegmatite is segmented intoblocky boudins. Their aspect ratios (L/W) range between 0.4 and 2.3, with an averagevalue of 1.2 (fig. 4A). The evolution of a single fracture (along its length Q) is shown infigure 6C: the tips are mode-I fractures, with the relative motion (sub-) normal to thefracture walls and with quartz veins filling the fracture (fig. 7A). Towards its half-length(Q/2) in Sample 1, shear displacement increases, and the fracture is gradually rotatedbetween the single boudins. The orientation of the fracture with respect to thepegmatite is characterized by the angle (�) (fig. 6C) and ranges from 0 to 5° for themode-I fractures and around 18° for reactivated and rotated mode-I fractures athalf-length (Q/2) (fig. 7A). Several such fractures are found in this sample; they are allsubparallel to each other and characterized by the same evolution (fig. 6A).

Sample 2 exposes the surface structure of a fractured and boudinaged pegmatite(fig. 4B). It was carefully mapped at 1:10 scale, followed by measuring length (Q),displacement (D) and dip of the interboudin surface (�; between interboudin surfaceand boudin exterior) (fig. 6B). The displacements along each fracture were measuredwith respect to fracture length or surface morphology (for example crusts of skarn).The maximum displacement (Dmax) for each fracture was recorded and plottedagainst the fracture length (Dmax/Q) (for example, Peacock and Sanderson, 1991)(fig. 7B). Along Q the fractures have commonly straight to arcuate traces and arebell-shaped.

Fig. 5. Three-dimensional sketch of structural features inside the marble raft. The isoclinally foldedamphibolite is crosscut by pegmatites. The pegmatites have roughly the same orientation as the steep limbsof the amphibolites. Both amphibolites and pegmatites were boudinaged during N-S extension.


In total we studied 75 fractures ranging in length (Q) from 25 to 632 mm. Thesefractures were subdivided into single, isolated fractures (N � 39) and faults withsegment linkage (N � 36) (with branches or relay ramps transferring the displacementto another fracture) (see fig. 4C).

The maximum displacements (Dmax) for single, isolated fractures range from 3 to34 mm with an average of 12.6 mm. The displacement gradients (Dmax/Q) of thesefractures range from 0.039 to 0.159 (average calculated from the best fit line: 0.088).

From the tips towards the fracture’s half-length (Q/2) the boudin blocks areprogressively rotated, which is characterized by the angle (�) (fig. 6B) ranging between2 and 20° (average 8.3 � 4.4°). The asymmetric structure of the single rotated boudins

Fig. 6. (A) Three-dimensional sketch of Sample 1, obtained by analysis of its single slices; the fracturesare subparallel and characterized by a length of 10 to 30 cm; (B) sketch depicting the parameters measuredin Sample 1 and Sample 2; (C) schematic drawing through a single fault along Q with average valuesmeasured at Sample 1; at the tips the fault is subnormal to the pegmatite’s layering pointing to mode-Ifracturing; towards the half-length (Q/2) the mode-I fracture is reactivated and inclined to � 72° withrotation of the single fragments.


is characterized by the angle (�), which—for all measured interboudin surfaceincrements in Sample 2—ranges between 39 and 78° (average: 57.1 � 8.1°) (fig. 7A).

Microscale Observations and GeochemistryMicroscopic observations and geochemical data were collected from slice 20 of

Sample 1 (fig. 8A). Composition of the rocks was analyzed by point counting (500points), XRF on bulk samples and microprobe analysis.

Mineralogy, petrology, geochemistry and microstructure.—MarbleBoth microprobe and XRF data establish the high purity of the calcitic marble

(� 99 to 99.5 % CaCO3). The yellow-brownish color changes close to the contacts toamphibolite and pegmatite are due to an increased amount of Fe, with the highestvalues measured close to the boudin necks. Inside specific calcite grains there areminor changes in chemical composition, but no systematic variation from the coretowards the grain boundary. The calcite grains in the marble of Samples 1 and 2 arecharacterized by a bimodal grain size distribution with the small grains ranging from0.7 to 1 mm and the large ones from 6 to 15 mm (grain size data was obtained bymeasuring the equivalent circular diameter in thin sections). The coarse grains have

Fig. 7. (A) Histogram showing the frequency of � (for all measured values of Sample 1 and Sample 2)indicating the rhombic shape of the boudins; the mode-I fractures are only displayed for Sample 1, as theycould not be measured in Sample 2 due to the absence of displacement; (B) Diagram illustrating thenormalized shape of the faults (both displacement and distance are normalized by the fracture length Q) ofall isolated faults of Sample 2 (see faults with dark gray shade in fig. 4A).


Fig. 8. Microstructures. (A) Slice 20 of Sample 1 with rectangles indicating the details in (B) - (E).Numbers point to three different regions: (1) unfractured region, (2) region at fracture’s half-length and(3) contact region (see table 1 for geochemical data of (1) and (3); (B) micrograph illustrating the slipmovement of the boudins parallel to the basal planes of biotite, partly altered to chlorite; (C) micrographshowing dynamic recrystallization of calcite at the contact to the pegmatite; the coarse calcite grains arecharacterized by lobate grain boundaries indicating grain boundary migration recrystallization, the finegrains have similar sizes to the subgrains, indicating subgrain rotation recrystallization; garnet (gt) andhedenbergite (hdb) belong to the typical assemblage of skarns (crossed polarizers); (D) fracture (thickness:2 mm) characterized by a dominant mode-I component and filled with quartz; this vein was subjected tofurther fracturing indicated by a vein with a thickness of � 200 �m, indicating different pulses of brittlefailure; approaching the fracture, plagioclase is progressively altered to sericite; (E) at this fracture bothalbite and orthoclase grains are brittle deformed and healed with orthoclase of different composition; thechemical differences are also revealed by microprobe mapping (MM) of sodium, backscatter analysis (BS)and cathodoluminescence (CL) (see also table 2).


irregular lobate grain boundaries indicating grain boundary migration recrystalliza-tion (fig. 8C). They contain subgrains with similar sizes as the small calcite grains,characteristic for slow grain boundary migration and subgrain rotation recrystalliza-tion. The co-existence of both subgrain rotation and grain boundary migrationrecrystallization is characteristic of dynamic recrystallization of calcite at high tempera-tures (Urai and others, 1986). The microstructure of the calcite is homogeneous: thereare no discernible differences (for example, grain size reduction) as a function ofposition with respect to the displaced parts of the boudins. Some of the calcite grainsshow weak undulose extinction and occasional mechanical twins indicating that themarble was only weakly affected by later deformation during uplift and cooling,although marble samples mylonitized during post-M2 metamorphism at temperaturesof � 300 °C were found in the same marble raft (Schenk and others, 2005).

Grain sizes of both fine and coarse calcite grains were used for paleopiezometry.With a grain size of the small grains of 700 to 1000 �m in the rotation recrystallizationpiezometer and grain size of the large grains of 6000 to 15000 �m in the migrationrecrystallization piezometer of Rutter (1995), differential stresses of 2.6 to 1.9 MPa and1.2 to 0.5 MPa are calculated, respectively.

AmphiboliteThe amphibolite consists of hornblende (65 %; tschermakite) and feldspar (� 28

%; predominantly andesine) together with minor amounts of sphene and muscovite.All minerals are elongated parallel to the lineation. The chemical composition of theamphibolite points to a magmatic (alkalibasaltic) origin, indicated both by theamounts of Ni and Cr and the Ni/Cr ratio (Frohlich, 1960).

PegmatiteThe mineralogical and geochemical data of the pegmatite was subdivided into (1)

the central, undeformed region, (2) the fractured region and (3) the contact regiontowards the marble (fig. 8A and table 1).

The central, undeformed part (1) is dominated by feldspar (predominantlyandesine with minor orthoclase occurrences) with grain sizes of 2 to 3 mm. Otherminerals are quartz (� 9 %), and minor biotite, sericite, apatite, sphene and rareallanite. Here there is no evidence for deformation, as for example undulose extinc-tion, deformation bands or microfractures. Myrmekites occur at contacts of plagioclaseand K-feldspar. Calculated from the H2O content of amphibole and mica, the centralregion of the pegmatite contains � 0.3 weight percent H2O. These measurementsindicate that the pegmatite has a granodioritic composition.

The fractured region close to the quartz veins (2) is characterized by nearly thesame mineralogy as in the undeformed parts of the pegmatite: predominantly feldsparand quartz. In addition, small amounts of biotite partly replaced by chlorite, sphene,apatite and opaque iron-rich minerals are present. With decreasing distance to thequartz veins, clear mineralogical changes are observed: the feldspar crystals aresuccessively enriched with fluid inclusions; plagioclase is altered to sericite; theAn-content of plagioclase changes from 40 (andesine) towards 5 (albite). Close to thequartz veins, feldspar is microfractured (figs. 8D and 8E). The fractures are sealed withorthoclase of different composition as revealed by microprobe analysis (table 2). Alsosome of the euhedral grains of sphene, apatite and quartz are microfractured.

The fractures which show displacement are also filled with quartz and biotite withoccasional microfolds and kinks, locally replaced by chlorite. Traces of fluid inclusionsare also found in the quartz but are less frequent than inside feldspar crystals. The veinquartz is weakly deformed with large subgrains and local grain boundary migration.Locally highly altered plagioclase crystals are enclosed in the quartz vein. Thesemicrostructures all indicate that after complete crystallization of quartz these veinswere weakly deformed.


Table 1

Geochemical data of two pegmatite specimens derived from Sample 1: XRF analysis of rockforming minerals, calculated CIPW norm (after Best, 1982) and modal composition

XRF analysis (1)1 (3)2

SiO2 64.37 63.83TiO2 0.13 0.55Al2O3 18.84 17.88Fe2O3 0.59 0.75FeO 0.40 0.50MnO 0.02 0.04MgO 0.31 0.47CaO 3.00 5.73Na2O 4.65 5.06K2O 6.79 2.28P2O5 0.35 0.18Cr2O3 < 0.005 < 0.005SO3 < 0.20 < 0.20V2O5 < 0.005 < 0.005LOI 0.38 0.94total 99.83 98.21

CIPW norm3

Q 5.85 14.92Ne 0 0Or 40.15 13.46Ab 39.36 42.83An 10.45 19.34Di 1.68 2.50Hy 0.04 0Wo 0 1.96Mt 0.86 0.14Hm 0 0.62Il 0.25 1.05Ap 0.83 0.42total 99.47 97.24

Modal composition4

quartz 8.8 8.2plagioclase 56.7 32.2kali-feldspar 24.3 11.2biotite 4.8 0.8amphibole 0.0 7.4apatite 1.6 0.2sericite 2.8 1.0zircon 0.6 3.8sphene 0.4 -garnet - 6.8Ca-Al-silicates - 16.8scapolite - 11.6total 100.0 100.0

1(1): specimen from undeformed region (Point 1 in figure 8A)2(3): specimen from contact region towards the marble (Point 3 in figure 8A)3abbreviations: Q: quartz; Ne: nepheline; Or: orthoclase; Ab: albite; An: anorthite; Di: diopside; Hy:

hypersthene; Wo: wollastonite; Mt: magnetite; Hm: hematite; Il: ilmenite; Ap: apatite4based on counting 500 minerals


Some calcite veins crosscut the quartz veins and may have formed by replacementreactions of plagioclase (Troger, 1969), or during a later deformation phase, whichhowever had to be weak, as both marble and pegmatite are not subjected to grain sizereduction.

At the contact of pegmatite and marble (3) — macroscopically visible as dark rim—minerals such as scapolite, andradite (garnet) and hedenbergite are observed (fig. 8C andtable 1). Such an assemblage is characteristic for metasomatic reactions between graniticmagma and carbonates, also known as skarn (Meinert, 1992). The chemical compositionof the garnet and pyroxene point to a tungsten skarn, a type that is often associated withpegmatitic or aplitic dikes in a high grade environment, conditions which existed insidethe high grade core of Naxos during M2b metamorphism.

discussionThe pegmatites exhibit many interesting structural aspects: (1) mode-I fractures

with quartz veins subnormal to the pegmatite and with the fragments progressivelyrotated towards domino-type boudins; (2) shear fractures with extraordinarily highdisplacement gradients; (3) feldspar deformed in a brittle fashion under medium-pressure, high-temperature conditions. On the one hand, mode-I fractures and quartzveins suggest high pore fluid pressures, which reduced the effective stress in thepegmatites, so that the low stress in the slowly deforming, dynamically recrystallizingmarble was sufficient to cause brittle fracturing. On the other hand, the rotatedboudins would not have formed under lithostatic pore pressures, because this wouldhave allowed the fractures to widen progressively without rotation of the fragments.

In what follows we present a model that explains the formation of these boudinsduring progressive deformation of the marble, illustrated by simple order of magni-tude calculations of the rates of the process (figs. 9A and 9B).

Cooling and Crystallization of the Pegmatite, Origin of High Pore Fluid Pressureand Mode-I Fracturing

During peak M2b metamorphism (20 to 16 Ma), the high grade core was atmedium-pressure, high-temperature conditions (� 670 °C and � 0.6 GPa) (Jansenand Schuiling, 1976; Buick and Holland, 1989), causing dynamic grain growth ofcalcite inside the marble towards the migmatite complex (Covey-Crump and Rutter,

Table 2

Microprobe data of feldspars shown in fig. 8E.

Ab1 Or12 Or23

SiO2 61.86 63.96 64.18Al2O3 22.88 17.88 17.87TiO2 0.03 0.00 0.00FeO 0.03 0.00 0.05CaO 4.77 0.00 0.01K2O 0.29 16.11 15.41Na2O 8.45 0.31 0.86Total 98.31 98.26 98.38

1Ab: albite with anorthite component of �20%.2Or1: albite with orthoclase component of �95%.3Or2: note higher Na2O content of ‘Or2’ as the result of the brittle fracturing and subsequent sealing of

‘Ab’ and ‘Or1’


1989; Urai and Feenstra, 2001). Swarms of pegmatites intruded the marbles synkine-matically at these peak M2b conditions, crosscutting B1 and B2 folds in the amphibo-lites, which already had a complex history of deformation.

Considering the composition of the pegmatite (table 1), it is reasonable to assumethe emplacement temperature of the pegmatitic melt to be around 750 °C (Car-michael and others, 1974; Whitney, 1988). Rough approximations based on simplecalculations using a Stefan-model (for example, Turcotte and Schubert, 1982) indicatethat 750 °C hot, pegmatitic melt (with a thickness of 7 cm) can be completely solidifiedin the range of hours to days after its emplacement into a host rock with a temperatureof 670 °C. Further cooling would reduce the temperature difference between pegma-tite and marble until the regional temperature would have been achieved aftermonths. Such a rapid crystallization of the pegmatitic melt under water-saturatedconditions is also proposed by Matthews and others (2003), who observed graphicintergrowth of quartz and feldspar in occurrences at the contact of high grade coreand metasediments of the Lower Series. These authors estimated that also thebuoyancy-driven propagation of the dikes was very rapid inside the high grade core:pegmatites with a thickness of 20 cm are estimated to reach a distance of 500 m in � 70days (Matthews and others, 2003). Another, albeit less likely scenario is that thepegmatite melt intruded at the temperature of the marble and crystallized slowlyduring regional cooling; we return to this later in the discussion.

Most pegmatites crystallize from H2O saturated magmas with the fluids commonlybeing in supercritical state (Jahns, 1982). As a consequence of the crystallization ofpredominantly anhydrous crystals, the amount of fluids in the residual melt wouldincrease, allowing an aqueous vapor phase to exsolve and replacement reactions totake place. This is enhanced by the viscosity contrast between melt and co-existingaqueous fluids, the latter being about 8 orders of magnitude higher at 800 °C underhigh pressures (Jahns, 1982).

The amount of dissolved H2O in peraluminuous granitic melts at conditions ofthe pegmatite’s emplacement is about 10 percent (Behrens and Jantos, 2001). Afterthe pegmatite melt intruded the marble, cooling was accompanied by crystallization ofmainly anhydrous minerals and evolution of a hydrous fluid with metasomatic reac-tions producing skarn at the contact zone between pegmatite and marble.

The mode-I fractures in feldspar at peak M2b conditions and at differential stressesof about 2 MPa in the marble require near lithostatic pore fluid pressures in thepegmatite to reduce the effective stress. This produced the quartz veins, the abundantfluid inclusions towards the fractures, and the fractured and healed feldspar crystals.We interpret these observations to indicate that the hydrous fluids released during thecrystallization of the pegmatite were retained for some time in the pegmatite and porepressures were lithostatic, allowing mode-I fracturing of the pegmatite during slowdeformation of the surrounding marble.

The pegmatite at present contains less than 1 percent H2O. Therefore fluids musthave migrated from the pegmatite into the marble (see also Matthews and others,2002). Marble however, is characterized by very low permeabilities at elevated tempera-tures (Baker and others, 1989; Fisher and Paterson, 1992; Buick and others, 1997;Matthews and others, 2002).

Starting from the simple assumption that the far field stress and kinematics ofductile flow in the marble did not change during pegmatite intrusion, the geometry ofthis system is puzzling. Pegmatite intrusion into an isotropic marble would imply 3normal to the pegmatite, while the fractures in the pegmatite constrain 3 to beparallel to the pegmatite. More work is needed to clarify this point, but one explana-tion may be that under these conditions of close-to isotropic stress in the marble


time

(Ma)

met

amor

phis

m

tem

pera

ture

over

burd

en p

ress

ure

pegm

. int

rusi

on

pore

flui

d pr

essu

re

pegm

. def

orm

atio

n

feld

spar

: brit

tle d

ef.

+ fr

actu

re h

ealin

g

met

asom

atis

m

mar

ble

grai

n gr

owth

fold

gen

erat

ions

tect

onic

reg

ime

50 M1

25 M2a

20-1

6

M2b

13 M3

4

480

° C67

0 ° C

400

° C

0.9

GP

a0.

6 G

Pa

0.3

GP

a

surf

ace

cond

ition

s

B1

& B

2B

1 &

B2

& B

3

b ca

A)

deta

il in

B

Fig.

9.T

imet

able

sto

illus

trat

eth

ere

lati

onsh

ipof

met

amor

phis

man

dde

form

atio

n.N

ote

that

the

tim

eax

esar

en

otsc

aled

.See

text

for

deta

iled

desc

ript

ion

.An

not

atio

nsp

oin

tto

refe

ren

ces:

(a)

Jan

sen

and

Sch

uilin

g,19

76;F

een

stra

,ms,

1985

;Wijb

ran

san

dD

ouga

ll,19

88;B

uick

and

Hol

lan

d,19

89;

(b)

Cov

ey-C

rum

pan

dR

utte

r,19

89;(

c)U

raia

nd

oth

ers,

1990

;Wijb

ran

san

dD

ouga

ll,19

88.


Fig.

9(c

onti

nue

d)


[1 600 MPa, (1 – 3) 2 MPa] the orientation of the pegmatites was controlled bythe anisotropy in the marble caused by the amphibolite layers.

Before the pegmatite was completely solidified, stresses in the melt were (1 �2 � 3 � pf), represented by a point in the origin of the Mohr diagram (state 1 in fig.10). Fluids formed during crystallization remained in the cooling pegmatite due to thevery low permeability of the marble during peak M2b conditions, although the veryhigh pore pressures may have increased permeability of the marble. With a differentialstress of � 2 MPa in the marble, the magma pressure was very close to the vertical stress(3 � 0.6 GPa), so that the pressure of the hydrous fluid (pf) was initially also close to3. Therefore the effective stress (�) in the pegmatite

� � � �pf (1)

Fig. 10. Simplified Mohr diagrams illustrating the failure conditions and fracture types. At regional M2bconditions inside the high grade core (0.6 GPa and � 670 °C); the dynamically recrystallized calcitic marbleindicates a differential stress (1-3) � 1 to 3 MPa; 1) the pegmatitic melt intruded and crystals and meltco-existed at these conditions; 2) simple order of magnitude calculations show that high pore fluid pressureswere able to be generated, shifting the Mohr circle towards the tensile field, such that the reduced effectivestress allowed mode-I fractures and brittle deformation of feldspar; 3) at a later stage fluids were able toescape from the pegmatite system; subsequently the pore fluid pressure is reduced, so that the Mohr circledoes not touch the failure envelope anymore; due to the extensional regime during the M2b event thepreviously formed mode-I fractures evolved towards reactivated and rotated mode-I fractures [to observe atthe fractures’ half-length (Q/2)].


was initially close to zero. Deformation of the surrounding marble caused extension ofthe pegmatites, so that stress became anisotropic with 3’ increasingly tensile (state 2 infig. 10) until it reached the tensile strength T of the pegmatite (we estimate this to havebeen a few MPa but less than 5 MPa). Evidence for this is given by i) the brittledeformation of feldspar and fracture healing by feldspar of different composition, ii)by mode-I fractures filled with quartz and iii) by the presence of euhedral quartz grainsthat grew in the veins (see figs. 4B, 4C and 6A).

Measurements of the dynamically recrystallized marble point to low differential stress(1 – 3) of � 2 MPa. Assuming a flow law of regime 3 (Schmid and others, 1980),

ε � A � exp �Q/R�T� � 1 � 3�n, (2)

and using the parameters (A � 18.4 [lnA[MPa-1s-1]], Q � 427 [kJmol-1], R � 8.314[ JK-1mol-1], T � 943 [K], (1 – 3) � 2 [MPa], and n � 4.2) (Schmid and others, 1980;de Bresser and others, 2002), the resulting strain rate in the marble during thisdeformation event is 4 x 10-15 s-1.

An extension (�) of � 1 percent was estimated for the mode-I fractures (see forexample fig. 8D). With the calculated strain rate in the surrounding marble, this eventlasted for approximately 80,000 years, during which pore pressure must have remainedclose to lithostatic.

If we assume a cooling scenario in which the pegmatite melt and surroundingmarble were at almost identical temperatures, crystallization would have been very slowduring regional cooling and would have taken place in the range of millions of years.To allow mode-I fracturing after solidification, the aqueous fluids would have toremain in the pegmatite during this much longer period, maintaining the high porefluid pressure. Thus, regardless of the exact duration of the crystallization, the brittledeformation of the feldspar and the mode-I fractures inside the pegmatite were theresult of high pore fluid pressures maintained for � 80,000 years or longer, at or closeto peak M2b conditions.

An important question is why the mode-I fracturing is commonly not perpendicularto the pegmatites’ layering and why the initial boudins already have a rhombic shape (seefig. 4A). One explanation for this could be that at the start of the deformation thepegmatites were systematically oriented at an angle to the minimum principle stress in themarble (fig. 11A), so that the fractures developed at less than 90° to the pegmatite.Ongoing deformation resulted in progressive rotation of the pegmatites’ envelopingsurface into parallelism with the principle extension direction (fig. 11B).

Evolution of BoudinageIf the pore fluid pressure had remained lithostatic for longer, the boudins would

have been separated without rotation, with the gaps progressively filled with quartzveins. This is not the case, and also the surrounding marble was not able to flow intothe gaps between the fragments, so that extension of the marble was accommodated byrotation of the fragments and sliding on the preexisting fractures and shearing of thequartz veins, as shown by undulose extinction, large subgrains and grain boundarymigration in quartz. This requires an increasing effective stress to prevent progressivewidening of the mode-I fractures (fig. 10; state 3).

Palinspastic restoration of the rotated boudins at the fracture’s half-length (Q/2)into its original state, points to a bulk extension (�) of � 14 percent, which in turn givesa duration of the extensional event of � 1.1 million years using the calculated strainrate of � 4 x 10-15 s-1 in the surrounding marble. We infer this to have taken place atpeak metamorphic conditions, based on the microstructures in the quartz vein, and onthe homogeneous microstructures in the marble. If this deformation happened later,for example at greenschist facies metamorphic conditions, the marble would show


grain size reduction in the regions close to the sliding fractures (Schenk and others,2005). Along the reactivated mode-I fractures, there is no evidence for skarn (fig. 8B).This again indicates that during this movement the pegmatite was already crystallized.

Chlorite replacing biotite often forms under greenschist facies conditions in thepresence of fluids. This is consistent with a minor deformation of the marble at lowertemperatures (strain of up to 1 %), as indicated by the mechanical twins and slightundulose extinction in the marble. Another mechanism for the formation of chlorite isat high temperatures during the final phase of crystallization and autometasomatism(Troger, 1969). In addition the decreasing An-content of plagioclase with the forma-tion of sericite and calcite is characteristic for autometasomatism (Troger, 1969). Morework is needed to determine the conditions of chlorite formation, but both scenariosdiscussed above are in agreement with the main deformation and rotation of boudinsto have taken place at or close to peak metamorphic conditions.

During continuing M2b extension, deformation in the surrounding ductile marbleproduced shear stresses acting on the predetermined, inclined mode-I fractures inside

Fig. 11. (A) Sketch to illustrate that mode-I fractures are inclined by some degrees, if pegmatitesintruded not perfectly parallel to the E-W shortening (that is the zones of weakness parallel to the steep limbsof the amphibolites). (B) Sketch to illustrate the structural evolution of the deformation of a pegmatite. Thesomewhat inclined fractures result in rhombic boudins that progressively rotate and displace duringextension, without the fragments being separated.


the pegmatites (Ramberg, 1955), resulting in the progressive rotation of singleboudins about their centers (fig. 11B), similar to experiments and theoretical consider-ations of Mandal and Khan (1991) and Kusznir and others (1991).

According to Mandal and Khan (1991) during layer-normal compression of aviscous fluid, rhombic segments undergo rotation with either their separation orinterfacial slip along the interboudin zones as a function of aspect ratio (W/L) andorientation of the fracture (�; orientation of the fracture with respect to the pegmatitelayering) (fig. 12). Plotting our data derived from Sample 1 into this diagram, theypredominantly fall into the field of kr � 1, that is into the field that predicts separationof the single boudins, structures that we did not observe either in the samples or in thequarry. This may be explained by two processes: i) the fractures form a three-dimensional network, so that, even if in a two-dimensional section the pegmatite isfractured, connection in three dimensions provides some coherence; in addition, thefinite length of fractures (Q) and high displacement gradients would shift thefield-separating graph (kr � 1) towards the left for our case; or ii) there was somecohesion between the boudins due to crack-seal processes and precipitation of quartzveins, which inhibited the ductile marble from flowing into the interboudin regions.

The deformational behavior of brittle pegmatite and ductile marble is compa-rable with the flexural cantilever model for lithospheric extension during the forma-tion of sedimentary basins (Kusznir and others, 1991, and fig. 10 therein). This modelof coupled simple and pure shear during continental extension suggests that the

Fig. 12. Orientation of the fractures with respect to the normal of pegmatitic layering of Sample 1plotted into the diagram of Mandal and Khan (1991) showing the relationship between width/length ratioand angle of cut (�) that determines whether the single boudins are separated (k�1) or rotated anddisplaced (k�1). Note that the reciprocal value of the aspect ratio is used in this diagram.


brittle upper crust deforms by faulting along planar faults, while the ductile lower crustand mantle deform by distributed pure shear. It also explains the familiar domino-styleblock-rotation in the upper crust with the hanging- and footwalls being always kept incontact, similar in style to the pegmatite fragments that rotated in the matrix of theductile deforming marble.

In profile plane, the boudinaged pegmatites suggest the development frominitially dilational asymmetric torn boudins (characterized by high-angle quartz-filledinterboudin zones) towards a later domino boudin component documented by blockrotation (fig. 4A). As discussed above, we interpret these structures to be the result ofone single event, during which only the pore fluid pressure decreased progressively.

In recent papers, such asymmetric boudins are proposed as important shear senseindicators (Goscombe and Passchier, 2003; Goscombe and others, 2004). The three-dimensional geometry of the pegmatites of our study reveals that the fractures (Q) are verysimilar to those of normal fault series observed in extensional regimes (for example,Kusznir and others, 1991; Schlische and others, 1996; Ackermann and others, 2003; Cowieand others, 2005). Inside a general regime of N-S extension and coaxial flow of the marble,normal faulting is able to account for the boudin block rotation along predeterminedfractures with an initially developed dominant mode-I component. We infer that there isno need for non-coaxial flow in the marble to produce such asymmetric boudin structures,and that the use of asymmetric boudins in a foliation-parallel boudin train as shear senseindicators needs to be regarded with caution.

The fracturing of the pegmatite inside the ductile marble may be used to constrainthe mechanical properties of pegmatites during medium-pressure faulting at subsolidusconditions. We plotted our data of maximum displacement (Dmax/2) against the length ofthe fracture (Q/2) with displacement gradients being in the range of � 0.04 to � 0.16

1

10

100

10 0 1 0000

0.0882

Q/2 [mm]

Dm

ax/2

[mm

]

pegmatite (this study)

set 1

set 2

set 3

Wibberley

Fig. 13. Graphs of displacement (Dmax/2) versus distance (Q/2) for the fault profiles of Sample 2compared with faults in unconsolidated sand from Idni, Morocco (Wibberley and others, 1999). The veryhigh displacement gradients of the pegmatite point to its weakness at medium-pressure, high-temperatureconditions, comparable to unconsolidated sand (note that in this diagram the pegmatite’s data of bothdisplacement and length are divided by the factor 2 to facilitate the comparison with the published data).


(0.088 calculated from best fit lines) (fig. 13). The comparison with published data ofWibberley and others (1999) shows that the displacement gradients are extraordinarilyhigh, suggesting that under M2b conditions inside the high grade core on Naxos, thepegmatite is quite weak inside the ductile matrix of marble, as compared to other faultedlithified rocks described in the literature (for example, Walsh and Watterson, 1987;Peacock and Sanderson, 1991).

conclusions

The exceptional outcrop conditions of pegmatites in a marble in the high gradecore of Naxos at M2b conditions represent a natural laboratory to study i) the dynamicsand deformation mechanisms at subsolidus conditions and ii) the three-dimensionalstructure of asymmetric boudinage.

Mode-I fractures in the pegmatites developed after solidification, during slowdeformation of the surrounding marble, under lithostatic pore fluid pressures, withpore fluids produced from the crystallizing pegmatite melt, and the high pressuremaintained by the very low permeability of the surrounding marble.

As pore fluid pressure dropped, continuing extension resulted in rotation of theboudins with slip along the interboudin zones. The fractures are characterized byextremely high displacement gradients, suggesting that pegmatites are weak underthese conditions.

acknowledgmentsThe authors would like to thank W. Kraus for preparing thin sections, S. Knipfer

and M. Blumenthal for data collection and processing, A. Wiechowski and R. Kling-hardt for their assistance with the microprobe, R. Neef for carrying out the XRFanalysis, A. Heide and M. Hansen for transport of Sample 1 to Aachen, and the peopleof the Naxos marble company for their hospitality during field work and for sectioningSample 1.

We acknowledge the valuable discussions with C. Hilgers, S. Sindern, H. Behrens,H. Forster and F. M. Meyer, and thank C. Wibberley for providing the data ondisplacement gradients and I. Buick for his comments on the regional geology ofNaxos. N. Mancktelow and an anonymous reviewer are thanked for their critical andconstructive reviews that highly improved the manuscript.

This paper is dedicated to the memory of the late Anne Feenstra with fondmemories of a long cooperation on Naxos geology.

References

Ackermann, R. V., Schlische, R. W., Patino, L. C., and Johnson, L. A., 2003, A Lagerstatte of rift-relatedtectonic structures from the Solite Quarry, Dan River/Danville rift basin, in LeTourneau, P. M., andOlsen, P. E., editors, The Great Rift Valleys of Pangea: New York, Columbia University Press, p. 118–133.

Andriessen, P. A. M., Hebeda, E. H., Simon, O. J., and Verschure, R. H., 1991, Tourmaline K-Ar agescompared to other radiometric dating systems in Alpine anatectic leucosomes and metamorphic rocks(Cyclades and southern Spain): Chemical Geology, v. 91, p. 33–48.

Avigad, D., 1998, High-pressure metamorphism and cooling on SE Naxos (Cyclades, Greece): EuropeanJournal of Mineralogy, v. 10, p. 1309–1319.

Baker, J., Bickle, M. J., Buick, I. S., Holland, T. J. B., and Matthews, A., 1989, Isotopic and petrologicalevidence for the infiltration of water-rich fluids during the Miocene M2 metamorphism on Naxos,Greece: Geochimica et Cosmochimica Acta, v. 53, p. 2037–2050.

Behrens, H., and Jantos, N., 2001, The effect of anhydrous composition on water solubility in granitic melts:American Mineralogist, v. 86, p. 14–20.

Best, M. G., 1982, Igneous and Metamorphic Petrology: San Francisco, W. H. Freeman and Company Ltd.,630 p.

Buick, I. S., 1991a, The Late-Alpine evolution of an extensional shear zone, Naxos, Greece: London, Journalof the Geological Society, v. 148, p. 93–103.

–––––– 1991b, Mylonite fabric development on Naxos, Greece: Journal of Structural Geology, v. 13,p. 643–655.


Buick, I. S., and Holland, T. J. B., 1989, The P-T-t Path associated with crustal extension, Naxos, in Daly, J. S.,Cliff, R. A., and Yardley, B. W. D., editors, Evolution of Metamorphic Belts: London, Geological SocietySpecial Publication, v. 43, p. 365–369

Buick, I. S., Cartwright, I., and Williams, I. S., 1997, High-temperature retrogression of granulite-faciesmarbles from the Reynolds Range Group: phase equilibria, isotopic resetting and fluid fluxes: Journal ofPetrology, v. 38, p. 877–910.

Carmichael, I. S. E., Turner, F. J., and Verhoogen, J., 1974, Igneous Petrology: New York, McGraw-Hill, 739 p.Covey-Crump, S. J., and Rutter, E. H., 1989, Thermally-induced grain growth of calcite marbles on Naxos

island, Greece: Contributions to Mineralogy and Petrology, v. 101, p. 69–86.Cowie, P. A., Underhill, J. R., Behn, M. D., Lin, J., and Gill, C. E., 2005, Spatio-temporal evolution of strain

accumulation derived from multi-scale observations of Late Jurassic rifting in the northern North Sea: Acritical test of models for lithospheric extension: Earth and Planetary Science Letters, v. 234, p. 401–419.

de Bresser, J. H. P., Evans, B., and Renner, J., 2002, On estimating the strength of calcite rocks under naturalconditions, in de Meer, S., Drury, M. R., de Bresser, J. H. P., and Pennock, G. M., editors, DeformationMechanisms, Rheology and Tectonics: Current Status and Future Perspectives: London, GeologicalSociety Special Publication, v. 200, p. 309–329.

Durr, S., Altherr, R., Keller, J., Okrusch, M., and Seidel, E., 1978, The median Aegean crystalline belt:stratigraphy, structure, metamorphism, magmatism, in Cloos, H., Roeder, D., and Schmidt, K., editors,Alps, Apennines, Hellenides: Stuttgart, E. Schweitzerbart’sche Verlagsbuchhandlung, p. 455–476.

Feenstra, A., ms, 1985, Metamorphism of Bauxites on Naxos, Greece: The Netherlands, RijksuniversiteitUtrecht, Ph. D. thesis, Geologica Ultraiectina, 206 p.

Fisher, G. J., and Paterson, M. S., 1992, Measurement of permeability and storage capacity in rocks duringdeformation at high temperature and pressure, in Evans, B., and Wong, T. F., editors, Fault mechanicsand transport properties of rocks: London, Academic Press Ltd., p. 213–253.

Frohlich, F., 1960, Ein Beitrag zur Geochemie des Chroms: Geochimica et Cosmochimica Acta, v. 20,p. 215–240.

Goscombe, B. D., and Passchier, C. W., 2003, Asymmetric boudins as shear sense indicators - an assessmentfrom field data: Journal of Structural Geology, v. 25, p. 575–589.

Goscombe, B. D., Passchier, C. W., and Hand, M., 2004, Boudinage classification: end-member boudin typesand modified boudin structures: Journal of Structural Geology, v. 26, p. 739–763.

Jahns, R. H., 1982, Internal evolution of pegmatite bodies, in Cerny, P., editor, Short course in graniticpegmatites in science and industry: Mineralogical Association of Canada, p. 293–327.

Jansen, J. B. H., and Schuiling, R. D., 1976, Metamorphism on Naxos: petrology and geothermal gradients:American Journal of Science, v. 276, p. 1225–1253.

John, B. E., and Howard, K. A., 1995, Rapid extension recorded by cooling-age patterns and brittledeformation, Naxos, Greece: Journal of Geophysical Research, v. 100, p. 9969–9979.

Keay, S., Lister, G. S., and Buick, I. S., 2001, The Timing of partial melting, barrovian metamorphism andgranite intrusion in the Naxos metamorrphic core complex, Cyclades, Aegean Sea, Greece: Tectonophys-ics, v. 342, p. 275–312.

Koukouvelas, I. K., and Kokkalas, S., 2003, Emplacement of the Miocene west Naxos pluton (Aegean Sea,Greece): a structural study: Geological Magazine, v. 140, p. 45–61.

Krabbendam, M., Urai, J. L., and van Vliet, L. J., 2003, Grain size stabilisation by dispersed graphite in ahigh-grade quartz mylonite: an example from Naxos (Greece): Journal of Structural Geology, v. 25,p. 855–866.

Kusznir, N. J., Marsden, G., and Egan, S. S., 1991, A flexural-cantilever simple-shear/pure-shear model ofcontinental lithosphere extension: applications to the Jeanne d’Arc Basin, Grand Banks and VikingGraben, North Sea, in Roberts, A. M., Yielding, G., and Freeman, B., editors, The Geometry of NormalFaults: London, Geological Society Special Publication, v. 147, p. 41–60.

Lister, G. S., Banga, G., and Feenstra, A., 1984, Metamorphic core complexes of Cordilleran type in theCyclades, Aegean Sea, Greece: Geology, v. 12, p. 221–225.

Mandal, N., and Khan, D., 1991, Rotation, offset and separation of oblique-fracture (rhombic) boudins;theory and experiments under layer-normal compression: Journal of Structural Geology, v. 13, p. 349–356.

Matthews, A., Baker, J., and Mattey, D. P., 2002, High-temperature metamorphism in marbles as aconsequence of volatile release from crystallizing anatectic melts, Naxos, Greece: European Journal ofMineralogy, v. 14, p. 37–47.

Matthews, A., Putlitz, B., Hamiel, Y., and Hervig, R. L., 2003, Volatile transport during the crystallization ofanatectic melts: oxygen, boron and hydrogen stable isotope study on the metamorphic complex ofNaxos, Greece: Geochimica et Cosmochimica Acta, v. 67, p. 3145–3163.

Meinert, L. D., 1992, Skarn and skarn deposits: Geoscience Canada, v. 19, p. 145–162.Peacock, D. C. P., and Sanderson, D. J., 1991, Displacements, segment linkage and relay ramps in normal

fault zones: Journal of Structural Geology, v. 13, p. 721–733.Pe-Piper, G., Kotopouli, C. N., and Piper, D. J. W., 1997, Granitoid rocks of Naxos, Greece: Regional geology

and petrology: Geological Journal, v. 32, p. 153–177.Ramberg, H., 1955, Natural and experimental boudinge and pinch-and-swell structures: Journal of Geology,

v. 63, p. 512–526.Rutter, E. H., 1995, Experimental study of the influence of stress, temperature, and strain on the dynamic

recrystallization of Carrara marble: Journal of Geophysical Research, v. 100, p. 24,651–24,663.


Schenk, O., Urai, J. L., and Evans, B., 2005, The effect of water on recrystallization behavior and grainboundary morphology in calcite - observations of natural marble mylonites: Journal of StructuralGeology, v. 27, p. 1856–1872.

Schlische, R. W., Young, S. S., Ackermann, R. V., and Gupta, A., 1996, Geometry and scaling relations of apopulation of very small rift-related normal faults: Geology, v. 24, p. 683–686.

Schmid, S. M., Paterson, M. S., and Boland, J. N., 1980, High temperature flow and dynamic recrystallizationin Carrara Marble: Tectonophysics, v. 65, p. 245–280.

Troger, W. E., 1969, Optische Bestimmung der gesteinsbildenden Minerale – Teil 2: Stuttgart, Germany, E.Schweitzerbart’sche Verlagsbuchhandlung, second edition, 822 p.

Turcotte, D. L., and Schubert, G., 1982, Geodynamics: Applications of continuum physics to geologicalproblems: New York, John Wiley and Sons, 450 p.

Urai, J. L., and Feenstra, A., 2001, Weakening associated with the diaspore-corundum dehydration reactionin metabauxites; an example from Naxos (Greece): Journal of Structural Geology, v. 23, p. 941–950.

Urai, J. L., Means, W. D., and Lister, G. S., 1986, Dynamic recrystallization of minerals, in Hobbs, B. E., andHeard, H. C., editors, Mineral and Rock Deformation; Laboratory Studies; the Paterson volume:Washington, AGU Geophysical Monograph 36, p. 161–199.

Urai, J. L., Schuiling, R. D., and Jansen, J. B. H., 1990, Alpine deformation on Naxos (Greece), in Knipe, R. J.,and Rutter, E. H., editors, Deformation Mechanisms, Rheology and Tectonics: London, GeologicalSociety Special Publication, v. 54, p. 509–522.

Walsh, J. J., and Watterson, J., 1987, Distribution of cumulative displacement and seismic slip on a singlenormal fault surface: Journal of Structural Geology, v. 9, p. 1039–1046.

Whitney, J. A., 1988, The origin of granite: the role and source of water in the evolution of granitic magmas:Geological Society of America Bulletin, v. 100, p. 1886–1897.

Wibberley, C. A. J., Petit, J. P., and Rives, T., 1999, Mechanics of high displacement gradient faulting prior tolithification: Journal of Structural Geology, v. 21, p. 251–257.

Wijbrans, J. R., and McDougall, I., 1988, Metamorphic evolution of the Attic Cycladic Metamorphic Belt onNaxos (Cyclades, Greece) utilizing 40Ar/39Ar age spectrum measurements: Journal of MetamorphicGeology, v. 6, p. 571–594.


evolution of boudins under progressively decreasing pore pressure

Documents