1. The Geology of South Central Milos, Greece
Mapping Project 2016
Department of Geology, School of Natural Sciences, Trinity College Dublin.
Loman Begley
Final Year Thesis
Student number: 13328091
Word Count: 6,493 words
2. Front Cover: Photo of Cape Zefyros and Firiplaka in the south central region, taken
from Cape Kalamos in the south east.
Abstract
The geology of south central Milos consists of three major periods- the Mesozoic, the
Neogene and the Pliocene-Pleistocene. The Mesozoic schist basement is the oldest unit
on the island and it relates to the collision of the African and Aegean plates during the
Alpine Orogeny. The uplift resulted in the formation of new islands including Milos.
The Neogene sedimentary deposits overlie the schist basement and are related to the
onset of slab roll-back of the subducting African plate which manifested in large scale
back arc extension. The first unit of this tectonic environment is a fluvial conglomerate
and sandstone sequence. The overlying limestone formation represents a significant
transgressional surface and the facies of each carbonate bed show the interplay between
eustatic sea level fluctuations and the relative rates of subsidence and sedimentation.
The Pliocene-Pleistocene period is when active volcanism prevailed on the island. This
island arc volcanism was initiated by the release of volatiles from the subducting
African plate below the Aegean, causing melting of the slab and the mantle wedge
forming calc-alkaline magma. Adiabatic decompression may have played a minor role
in the generation of melt due to crustal thinning in the Aegean plate. The first volcanic
event on south central milos was a pyroclastic flow, followed by rhyolitic and
rhyodacitic lavas and sub-aqueous rhyodacitic fallout deposits. Hollocene phreatic
deposits is the last volcanic event on the entire island. It coincided with human
habitation of the island and entrained roman pottery allowing for accurate dating of the
deposit. This area remains tectonically and hydrothermally active shown by the
abundant sulphur fumaroles and the minor seismic activity in 1992.
4. 1
1. Introduction
Over the summer of 2016, the geology of south central milos was mapped over an
area of 24km2
. The area was chosen for its broad variety of geology and excellent
exposure.
1.1 Aims
To reconstruct the geological evolution of south central Milos and produce a
geological map at 1:10,000. Correlating the geological history to the regional
tectonic history of the Aegean was another aim of this project.
1.2 Location
Milos is an island located in the Central Aegean Sea. It is part of the Cyclades and is
located 152km south west of Santorini. The area mapped is shown in Fig 1.1.
Fig.1.1 Satellite image of Milos with mapping area marked in the south central region.
5. 2
1.3 Regional Tectonic Setting
Milos is a volcanic island arc situated in the central part of the Southern Aegean
Volcanic Arc (SAVA). Santorini and Nisyros are also part of the SAVA belt which lies
~125km above the Benioff zone of the subducting African plate. The volcanism is the
result of the partial melting associated with the subduction of the African oceanic plate
below the Aegean microplate.
The onset of slab roll back reverted the tectonic system of the SAVA from a collisional
to a back arc extensional tectonic system. This has resulted in crustal shortening in the
region to approximately half its original thickness at ~20-30km (McKenzie, 1978) and
heat flow here is consequently high (Kalogeropoulos & Paritsis, 1990). The present rate
of subduction along the Hellenic trench is 4.0-4.5cm.a-¹. Extension has been
concentrated in two major grabens in the Aegean, the Cretan and Anatolian troughs.
Between these lies the Central Aseismic Plateau (CAP), a stable crustal block
characterized by low seismicity. The volcanic arcs including Milos lie on the faulted
southern boundary of the CAP (Druitt et al., 1999).
Fig.1.2 Milos is marked by a green arrow. It is located along the SAVA belt which is marked by
a dashed blue line. From Friedrich (2000).
6. 3
2. Lithologies
2.1 Stratigraphic Column
Fig.2.1 Correlating the lithostratigraphy from the west, east and central regions of the mapping
area. The key to the geological colours are given in Fig 2.2. (The eastern region has a closer
correlation to the western region than does the central region)
Fig.2.2 Chronostratigraphy based on radiometric dating of the volcanic deposits by Fytickas et
al., 1986.
7. 4
2.2 Mesozoic Schist Basement.
This is the oldest unit exposed on the island. The protolith was a muddy shale
deposited on the Tethys deep ocean floor. It was metamorphosed and highly
deformed during the Mesozoic when the North African and Eurasian plates collided.
As Milos is located ~380km from the Hellenic trench (subduction zone) it was only
exposed to greenschist facies metamorphism (medium temperature and pressure).
No index minerals were found in the schist which constrains the interpretation of the
metamorphic grade. However, the metamorphic grade was seen to slightly vary over
just 200m of exposure, grading from shaley schist, to muscovite schist, to chlorite
schist, to gneissic schist.
Fig.2.3 A: Recumbent fold in basement showing high angle fold limbs of anticlines and
synclines. The folds are cross-cut by a quartz vein ~5cm think. B: Chevron fold in schist. The
schistose texture is seen by the alternating mafic (biotite and chlorite rich) and felsic
(plagioclase and quartz) bands.
8. 5
It is exposed on the south east at Paliochori as a host block and is overlain
unconformably by young phreatic deposits. It has been uplifted by normal faults
roughly striking north east. The schistosity is an obvious compressional stress fabric.
Locally, there is some well-preserved felsic and mafic boudinage which is an
extensional stress fabric. They formed when minerals clumped together as the rock
was being sheared and pulled apart forming elongate tapering structures varying
from 2-13cm in length and 1-7cm in width. They only occur on a roughly south
facing cliff face corresponding to an E-W extensional primary stress orientation.
Fig.2.4 Boudinage; extensional stress fabric overprinting compressional stress fabric
(schistosity). Basement has been sheared with both felsic (quartz and plagioclase) and mafic
(biotite, chlroite and muscovite) material being ‘strung out’ as seen in the picture. The boudins
have a block geometry and are symmetric, drawn and tapering.
9. 6
2.3 Fluvial Basal Conglomerate & Sandstone Sequence.
This unit outcrops at Cape Akrotiri and Provatas, below the limestone formation.
It is ~21m thick and consists of predominantly schist, quartz and marble clasts
eroded from the exposed metamorphic basement by palaeo-rivers. The marble
clasts indicate a secondary provenance as the local metamorphic basement is
entirely schist. It is known however that there is marble basement elsewhere on
Milos. Consequently the marble clasts are more rounded and spherical due to the
greater duration of transport. The schist and quartz are angular-sub angular and
non-spherical.
The conglomerate beds are river channel facies. It is poorly sorted with clast size
ranging from 1-18cm, model 5.4cm. Imbrication records bi-directional
palaeoflow consistent with a meandering fluvial channel. Width of the beds vary
from ~0.7m-2.4m.
Fig.2.5 Imbrication in the conglomerate bed recording an ESE palaeoflow. Palaeoflow
alternates in each conglomerate bed, the next conglomerate bed in this sequence has a NW
palaeoflow. The reddish colour was interpreted as a secondary feature from pore fluids
containing haematite causing the orange-red coloration as the rock is permeable.
The sandstone beds are point bar to floodplain facies. It is moderately sorted,
normally graded, sub-angular and non-spherical schist and quartz clasts with
lesser amounts of sub-rounded marble clasts. The size of the clasts range from
0.4-1.2cm, and the model is ~0.7cm. There is very subtle cross-bedding with bi-
directional palaeoflow. Width of the beds vary from 1.8m-3.2m.
10. 7
The top of this unit is an unconformable surface represented by a 9-12cm think
palaeosol representing a significant hiatus. This was interpreted to represent the
time period when the metamorphic basement was virtually eroded to base level
which terminated sedimentation from the basement. It is also likely that it
represents a change in climate from tropical to arid, thus resulting in an end to
fluvial sedimentation. The arid weather chemically weathered the top of this unit
by pedogenic processes creating an iron oxide rich soil.
Fig.2.6 A: Basal Conglomerate showing normal grading to a coarse sandstone, to a finer
sandstone and then a sharp boundary to repeating conglomerate. The conglomerate is laterally
discontinuous which tapers and pinches out. (The normal fault is strking NE dipping SE)
B: Sandstone taken from the centre of the cliff face seen in picture A. Note the large sub-
angular quartz clast and abundant fine biotite’s weathered from the schist.
It was interpreted to be of a meandering fluvial environment because of the
repeated grading and the alternating palaeoflow interpreted to show the
environment going from a fluvial channel to a point bar to a floodplain.
Eventually the rate of subsidence exceeded the rate of sedimentation and a
transgression occurred resulting in a shallow tropical sea flooding the basal
conglomerate and sandstone sequence.
11. 8
2.4 Neogene Limestone Formation
The onset of carbonate beds overlying the basal conglomerates marks a
transgressive bounding surface. This is due to the rate subsidence outpacing the
rate of sedimentation caused by an increased rate of slab roll back in the
asthenosphere. The limestone formation is only exposed at Cape Akrotiri (Kipos)
and at the Provatas cliffs on the south west coast. It provides a 46m cross section
through the Limestone Formation. The width of the carbonate beds ranges from
~30-120cm, model ~55cm.
The main alternating facies is the presence or absence of silt, which was a useful
parameter for interpreting the changing shallow marine environment.
Grainstone formed in the absence of silt –a crystalline rock predominantly made of
sparry calcite. It is very hard and flakey and usually coloured red due to affiliation
with haematite from pore fluids.
When silt is present the rock is a softer, non-crystalline and usually has a yellow-
white/grey colour. The ratio of silty limestone to grainstone is 78:22. Both rock
types may contain fossils. The presence of silt (or fine sand) was interpreted to be
characteristic of a shallower depositional environment as it is often cross-bedded
and the grainstone beds are not. The grainstone may contain little or no siliciclastic
material which was interpreted to be related to its greater distance from the
shoreline (or else a highly productive marine system which recycled all the silt).
Evidence for this is that a silty rippled limestone containing marine fauna and
ostracodes is the first carbonate bed overlying the basal conglomerate and
sandstone sequence. This depositional environment had to be shallow and close to
the shoreline as it marks the onset of the transgression.
It may also be related to fluctuations in the palaeoclimate due to its control on the
rate of denudation and abundance of rainfall required to transport siliciclastic
material to the shallow marine environment. During periods of abundant rainfall
and higher temperatures, there is increased probability of fine material being
deposited as far as the marine shelf. During arid periods a much cleaner limestone
formed (grainstone) and flora and fauna thrived.
12. 9
Fig.2.7 Kipos Limestone Formation exposing ~46m of alternating grainstones, fossiliferous
grainstones, stromatolites and silty limestones.
The fossil assemblage shows no evolution or systematic change through the
formation. The most abundant fossil are bivalves which are up to 3.6cm in length,
followed by ostracodes, and rarely there are molluscs, gastropods, foraminifera
and echinoderms. The calcium carbonate shells are not preserved in any beds
which may be indicative of large scale dissolution by pore fluids of similar
chemistry circulating throughout the entire formation. Secondary sparry calcite is a
common constituent in these carbonate beds, given that it is also a widespread
phenomenon the two may be related. It was interpreted that the dissolved calcium
carbonate shells later or contemporaneously precipitated the carbonate out as
crystalline subhedral calcite.
Two stromatolitic beds were identified. The first is a thrombolite red grainstone
which occurs at the bottom of the formation (second carbonate bed overlying the
basal conglomerate and sandstone sequence, Log.1-B2). There is no clear dome
structure but it can be seen from the underside of the rock. The rock contains
elongate pores up to 3.1cm in length and 1.3cm in width. Many of these pores are
‘birdseye’ shaped while most of them are more irregular in shape. This observation
(along with the underside doming) led to the interpretation that these are primary
pores ‘birdeye fenestrae’ which formed when respiratory and photosynthetic gases
released by microbes became entrapped by sediment. The rock was thus
interpreted to be a thrombolite.
13. 10
The second stromatolite identified was a domed silty stromatolite which occurs
eleven beds above (7.4m) the thrombolite grainstone. These beds are both
interpreted to be of a hypersaline environment such as an intertidal sabkha. The
grainstone thrombolite is more indicative of an arid sabhka while the silty domed
stromatolite is more likely to have formed in a humid sabhka. This is because algal
material is more likely be preserved in a humid environment. There is no algal
material remaining in the thrombolite grainstone due to the arid depositional
environment and also the diagenetic effects of compaction and dissolution.
Fig.2.8 A: Domed silty stromatolite (see appendix for petrographic description). Note the
stromatolitic layering marked by dense brownish bands and fine silt between these layers.
B: Vuggy thrombolite grainstone. Where the rock has bright red patches there is an
internal vug (tested).
The boundary between the limestone formation and the overlying ignimbrite is
erosive and represents a disconformity. The youngest carbonate bed is not clear as
the beds dip away from the ignimbrite so the youngest bed may not outcrop or it
may have been completely eroded by the basal surge of the pyroclastic flow. The
youngest mapped carbonate is an interbedded sand and silt limestone similar to the
oldest bed representing a shallow marine, high sediment flux environment.
14. 11
2.5 Andesitic Ignimbrite
This represents an early explosive pyroclastic flow which may be the earliest volcanic
event on Milos as it overlies the limestone formation. Each outcrop shows a large
variation in proportion of clasts, size of clasts and flow features due to the chaotic
nature of the deposit and the part of the sequence which is outcropping. The basal
surge has fully preserved metre scale ripples (fig. 2.9) with a wavelength of 60-240cm,
and an amplitude of 41-65cm. It is abundant in lithics of non-juvenile material (~15%)
- schist and limestone rip-up clasts showing the erosive nature of the ignimbrite. It
often shows normal grading and consists of matrix supported ash and lapilli pumice.
Locally in the lower sequence fine blocks in an ash matrix are prevalent, which is
indicative of a highly explosive ignimbrite.
Fig. 2.9 Basal surge of ignimbrite deposit showing undulating surface and locally clast
supported fine blocks of lava and lapilli clasts of schist, limestone and pumice. White bed
where hand is placed in picture is the boundary between basal surge and pumice and ash
sequence.
The lithic clasts of lava constitute ~8% and are angular and vary from rhyodacitic
to andesitic. The rhyolitic and rhyodacitic clasts are usually finer grained than the
andesitic clasts.
Detailed interpretations were difficult as all the outcrops were partially weathered
and/or altered to a white clay. Palaeoflow was interpreted from foliation preserved
in the pumice clasts and by the wave ripples in the basal surge. However these
indicators could not decipher between the source and movement direction but it
was interpreted that it came from the same source as the overlying rhyolitic
magma which is the dome volcanoes in the west including the Halepa dome.
15. 12
2.6 Quartz-Feldspar-Biotite Rhyolitic Porphyry (QFBRP) Lava
This covers the entire western region and includes a series of dome volcanoes. The main
dome of this phase is Halepa ~1.2km west of the limit of the area mapped. The lava is
cryptocrystalline with phenocrysts of quartz, albite, orthoclase and biotite which are
relatively equigranular. Model grain size ~0.4cm in length, ranging from 0.1-1.0cm.
Sub-rounded mafic enclaves are common and are dioritic in composition and consist of
~35% euhedral acicular amphibole phenocrysts which quenched rapidly. Due to it sub-
rounded morphology it may be indicative of magma mixing in the chamber with the two
magmas failing to equilibrate before the eruption. However, because the phenocrysts are
much finer compared to the normal lava phenocryst, it was interpreted that there was a
late injection of melt into the chamber. Obsidian clasts are rare and range from 3-10cm
in length, sub-angular and contain glass inclusions ~5mm and quartz phenocrysts
~10mm. Also indicative of pieces of melt with virtually no pre-existing crystals, this
represents much younger melt which has not had time to equilibrate with the silicic
chamber walls and develop a more felsic composition.
Fig.2.10 A: Porphyritic rhyolite containing subhedral biotite phenocryst and anhedral
phenocrysts of quartz, plagioclase and orthoclase. Fibrous creamy minerals are feldspars in the
early stages of devitrification. B: Dioritic mafic enclave containing ~35% rapidly quenched
amphibole phenocrysts of acicular habit and mostly euhedral.
Mafic enclaves was a characteristic feature of this lava flow and was a characteristic
feature in distinguishing it from the similar rhyodacite lava flows. Gas vesicles in the
lava constitute ~8% of the lava due to the high gas content. Wispy fragments of white-
creamy coloured material oriented parallel to vesicles is where feldspar crystals are
devitrifying.
16. 13
Fig.2.11 Angle of flow banding relative to palaeotopography alternating as the lava flows and
becomes increasingly viscous resulting in vertical flow banding and auto-brecciation.
This lava overlies the andesitic ignimbrite which is capped with a thin palaeosol. Based
on this observation, and the low level of auto-brecciation, it was interpreted to be a sub-
aerial eruption. It is also unlikely to be have been sub-aqueously deposited as it would
be expected that some material would have deposited above it considering it is
relatively old; 0.95±0.08 Ma (Fytickas et al., 1986).
Fig. 2.12 Rhyolite lava showing flow banding and palaeoflow direction.
17. 14
2.7 Quartz-Feldspar Rhyodacite Porphyry (QFRDP) Lava
Very similar rock composition to the rhyolitic lava but it does not have a pinkish
colour due to the relative absence of k-feldspar. It has a porphyritic texture with
phenocrysts of plagioclase, biotite and quartz, with groundmass devitrifying from a
glassy matrix. Biotite phenocrysts are slightly more abundant ~5%. There is also no
mafic enclaves or obsidian in the lava. Locally the lava is highly vesiculated due to
both the high gas content of the rhyodacite magma and due to devitrification forming
fibres that occupy up to 40% less volume than the parent glassy material.
Mapped as a series of small (~80m length, 20m wide-average) domes which were
subsequently covered in lapilli ash fallout. Only the top of these domes became
exposed. The Cape Kalamos dome volcano is probably the most recent active dome. It
also erupted from the Fyriplaka crater and may have destroyed the northwestern part
of the edifice. The small domes in the centre are interpreted to be cryptodomes due to
their varied palaeoflow and low elevation. They may also be coulees due to the
overprinting of lava flows over each dome. This lava is highly autobrecciated and was
interpreted to have erupted sub-aqueously, which was further supported when the
overlying fallout deposit was also interpreted to be a sub-aqueous eruption
Fig.2.13 Viscous flow mechanism preserved at Cape Kalamos dome volcano. The top part is
more sluggish and has slightly flowed over itself due to the effects of gravity. The layer above
this was severely auto-brecciated and has mostly eroded away. The bottom part is not auto-
brecciated because it had more time to conductively cool and due to its very close proximity to
the source.
18. 15
2.8 Rhyodacite Explosive Fallout
2.8.1 Description
A large fallout deposit with a maximum observed thickness of ~70m. This
explosive phase is the last magmatic eruption on the entire island. It erupted from
the large crater of Fyriplaka. The deposit grades from lapilli stone close to the
source to lapilli ash away from the immediate source and ash (often mixed with
diatoms see Chp. 2.9 Ash Diatomite) in the distal region. The south wall of the
crater edifice is covered in lapilli stone while the east wall is covered in lapilli ash.
This may correspond to northerly trade wind direction during this period.
Rhyodacitic lapilli clasts are sub-angular ranging from 0.1-2.0cm. In the lapilli
stone the clast size ranges from 8-70cm. As seen with the lavas, there are also
phenocrysts of biotite, quartz and plagioclase although they are smaller ~1mm.
There is also signs of devitrification from the glassy parent material. Slow cooling
of some crystals in the magma chamber followed by rapid cooling upon eruption.
Fig.2.14 Cross laminated lapilli ash clearly reworked by water. Flow mechanics continuously
alternates shown by the planar bed overlying the cross-beds, and the coarsening of the cross
beds.
2.8.2 Interpretation
Accumulation of volatiles and gases in the magma chamber resulted in an
explosive plinian eruption at Fyriplaka which created the larger crater (Fig.3.1).
Due the deposit being extensively cross-laminated, (and rich in non-juvenile
material ~20%) it was interpreted to be have been water reworked and thus
deposited under sub-aqueous conditions. Stewart and McPhie, 2006 interpreted
19. 16
this to be a sub-aerial deposit and suggested the island evolved from subaqueous to
subaerial conditions by volcanic constructional processes. However, further
evidence for a more recent partial marine (or possibly lacustrine) submergence was
found on the north coast at Achivadolimni where very young beach deposits and a
marine terrace were identied. It is thus interpreted that there may have been a large
relative sea level rise related to eustatic processes or possibly even crustal sinking
due to expulsion of magmatic material which led to this area becoming at least
partially submerged. Van Hinsbergen et al., 2004 suggested a dramatic sea level
rise of ~900m occurred around 5Ma on Milos- based on seismic profiles and
submarine tectonic maps of the Saronic Gulf. This evidence directly supports the
theory proposed that this unit (and the rhyodacite lava) was deposited under
submarine conditions.
Fig.2.15 Lapilli Stone with very subtle bedding (marked by dashed black lines) of interbedded
stone sized fragments and lapilli sized fragments at Fyriplaka south.
20. 17
2.9 Ash Diatomite
This is a homogeneous, low density rock with a chalky, brittle texture and a pure white
colour. The low density is due to the microscopic cavities in the diatoms. It was
interpreted to be of the same age as the lapilli ash deposits. However this rock does not
contain any lapilli fragments which is probably the result of the depositional
environment being a considerable distance from the source resulting in all the heavier
lapilli sized fragments settling out before reaching the seas where the diatom blooms
were occurring. The diagenesis of feldspars in the ash releases soluble Si resulting in the
seawater becoming saturated in silica. This promotes the blooming of diatoms and also
prevents the frustules from dissolving on the seafloor. Ocean currents continuously
mixed the diatoms and ash resulting in a relatively homogeneous mix.
Fig.2.16 A: Fluid escape structure in ash diatomite. This secondary feature was the result of the
overburden weight of sediment overlying the waterlogged ash diatomite layers. The outer
surface is stained from fluids leaching down from overlying phreatic deposits B: Pure white
colour in ash diatomite rock. A normal fault is providing a pathway for the movement of
hydrothermal fluids altering the rock to an orange coloured clay (note purple pen for scale)
The subtle laminations seen in the flame structures are interpreted to be alternating ash
rich layers and diatom rich layers. Therefore this may be representative of the waxing
and waning of the explosive volcanic event and the periodical emergence of new diatom
blooms. It may also represent settling of sediment according to relative density on the
seafloor. Diatom frustules are less dense than ash particles and so these will organize
themselves to the top of seafloor sediment resulting in fine laminations.
21. 17
2.10 Phreatic Schist Conglomerate
This is the most recent volcanic activity on the entire island. There is no juvenile
material in this deposit. It was caused by an underlying hot magma source
superheating the overlying water-table, causing a massive pressure build-up within the
local strata. Eventually the overlying pressure of the rocks was less than that of the
superheated groundwater which resulting in a huge explosion which blasted all the
overlying beds into fragments along with water and silt. This resulted in a debris flow
which plastered the surrounding area. It consists of matrix-clast supported schist
conglomerate also containing limestone and volcanic clasts. Deposit shows subtle flow
features of cross-bedding and imbrication, but is mostly chaotically deposited and very
poorly sorted. Interpreted to have moved at a low velocity over a short period of time.
There are two phases of this deposit, the first deposit occurs only at the Tsigrado cliffs
on the south coast. This deposit is overlain by a 20cm thick palaeosol followed by
lapilli ash fallout. The second deposit is the youngest unit on Milos and is much more
laterally extensive, covering much of the north eastern area.
Fig.2.17 Phreatic deposits disconformably overlying ash diatomite. The lower conglomerate bed
is matric supported and shows subtle imbrication-may represent a basal surge of the deposit.
The upper bed is clast supported and contains blocks of lava and marble up to 65cm in length.
22. 18
Fragments of pottery were found in the younger phreatic conglomerate along roadside
exposures east of Zefira. Traineau and Dalabakis (1989) identified pottery in these
deposits to be of the Roman Era. They carried out ¹⁴C measurements on the pottery
which indicated that the phreatic activity occurred between 200BC and 200AD.
Fig.2.17 Roman pottery found in the phreatic deposits east of Zefira.
23. 19
3. Structure of South Central Milos
Due to the widespread hydrothermal activity in this area, relative orientation of fault
blocks is rarely seen and no slickensides have been preserved due to fault zones being
altered to white clay by the hydrothermal fluids. All significant faults are extensional
although the large faults in the centre and east may be oblique, however there is little
evidence to support this.
The sedimentary deposits are gently folded with fold limbs dipping at approximately
18˚. This may be related to the constructional effects of volcanism because the tectonic
environment has been extensional since the deposition of these beds.
The schist basement was subjected to multiple phase of deformation and consequently
the fold axes of the folds plunge in alternating directions producing recumbent folds and
asymmetrical antiforms and synforms. It was subjected to compressional stresses in the
Mesozoic during the alpine orogeny and extensional stresses since the Neogene period.
The eastern region (where sulphur fumaroles are abundant) was the focus of seismic
activity in March 1992 with focal depths of between 1&4km (Papanikolau et al.1993)
3.1 Horst and Graben
The area mapped has an asymmetric horst and graben structure. The west and east
constitute a horst block uplifted at least 90m above the graben. The SW coast exposes
~80m of Neogene deposits. Directly north of these deposits on the NW coast the sea
level relative to the topography is relatively flat covered by QFRP lava. This is direct
evidence for oblique extensional faulting. The other possibility is that this fault is not
continuous and phases out in the central region, although it is interpreted that the fault is
continuous and may have played a role in the formation of the Gulf of Milos (Fig. 1.1).
The graben structure constitutes flat lowlands called the Zefiria plain in the north central
region. The eastern horst exposes the schist basement. The Fyriplaka crater formed later
at the graben and rose to a similar height of the two horst’s.
The uplifted schist basement in the south east is cross-cut by numerous north trending
faults. These faults were interpreted to be locally oblique as the uplift is restricted to the
south east only.
24. 20
3.2 Crater Edifice
Fig.3.1 Satellite image of the Fyriplaka crater edifice, 1.75km in diameter and only
partially preserved due to both the explosive nature of the deposits and the effects of
lava flowing over the western side of the edifice. A later edifice labelled 2 formed
within the large crater. The blue lines are showing the wavy ridges of lapilli ash &stone
which form around the edifice. These were interpreted to be a possible subaqueous
feature of the crater. The green line is a large normal fault and the yellow box with an S
is a sulphur fumarole. This fault striking NNW is interpreted to have accommodated the
migration of melt and the caused the volcano to erupt here.
25. 21
4. Hydrothermal Activity
4.1 Hydrothermal Processes
Hydrothermal activity is extensive along the major faults in the south east at Cape
Thiares, at Cape Kalamos dome and around Fyriplaka. It is characterized by the
alteration of the local lithologies to white clay, and by sulphur fumaroles which
continuously emit sulphuric gases.
Fig.4.1 Mineralising reverse fault within phreatic deposits. The hydrothermal fluids altered the
rock to white clay and catalysed the oxidation of iron seen by the deep red band overlying the
clay.
The fumaroles result in the precipitation of mineral precipitates (mainly gypsum and
sulphur) on the surrounding lithologies. Botryoidal haematite, chert and amorphous
silica also formed from the hydrothermal fluids which may derived from seawater that
percolated into the crust through the faults, or from calcium rich meteoric water. Gases
coming from the magma source enriched the fluids in sulphuric acids and carbon
dioxide.
Fig.4.2 A: Elongate and euhedral gypsum crystals up to 4cm in length growing within schist,
perpendicular to the cleavage planes. B: Rhyodacite lava almost completely altered to white
clay by hydrothermal fluids.
26. 22
4.2 Hydrothermal Precipitates
Pure white, soft and acicular sulphur crystals are found on the roofs of the fumaroles,
indicating an anoxic composition of the hydrothermal fluids. Over 10metres away from
the fumaroles, gypsum and silica oxide minerals occur including chalcedony, chert and
amorphous silica. This indicates that the anoxic hydrothermal precipitates have oxidised
forming predominantly gypsum and where the sulphur has been leached the silica
oxides formed.
Fig.4.3 A: Botryoidal haematite crystrals 1-2mm in size. Mostly black coloured, locally contain
an array of bright colours. Lustre is vitreous-waxy. B: Elongate euhedral gypsum crystals 3-
4.5cm in length. C: Brown chert containing some quartz inclusions. D: Dark red jasper. Crystal
face beside the coin is shiny and contains an outer band of purple chalcedony. E: Botryoidal
crystals of amorphous silica, light grey-blueish colour may derive from the presence of
manganese. F: banded chalcedony, each band is wavy and crinkled-low temperature precipitate.
Blue-purple colour with a waxy lustre.
27. 23
5. Geological History
5.1 Mesozoic Alpine Orogeny-Schist Basement.
Given the tectonic history of the Aegean, the schist protolith was interpreted to
be a deep marine shale deposited on the Tethy’s ocean bottom. Milos is located
~380km from the subduction zone at the Hellenic trench and was consequently
subjected to low grade - medium pressure, low temperature, greenschist facies
metamorphism during the Alpine Orogeny. The schist was also severely
deformed during this period as can be seen by the recumbent folds, overturned
antiforms, boudins and high angle fold limbs resulting in hinge folds.
Fig.5.1 African and Aegean plates collide resulting in widespread metamorphism and
back arc up-lift forming new islands in the Aegean sea including Milos.
5.2 Neogene Subsidence
Slab roll-back of the subducting African plate became significant ~23Ma (Ring
et al., 2010). This lead to back-arc extension in the southern and central Aegean
which manifested itself as regional subsidence through crustal thinning and
normal faulting. Van Hinsbergen et al., 2004 proposed that ~900m of subsidence
occurred on Milos between ~5 and 4.4Ma. After the schist basement became
sub-aerially exposed during the Alpine Orogeny, rivers began to incise the
basement resulting in a fluvial sedimentary basin. This palaeoenvironment led to
the deposition of a basal conglomerate and sandstone sequence of meandering
river facies.
28. 24
When subsidence began to outpace the rate of sedimentation, a transgression
occurred which submerged the fluvial environment. This marine environment
alternated between shallow tropical, arid and humid sabkhas and high and low
sediment flux.
Fig.5.2 Slab role-back due to the weight and effects of gravity result in tectonic
extension in the overlying Aegean plate resulting in crustal thinning, horst and graben
structures and ultimately an extensive period of subsidence.
29. 25
5.3 Volcanic History
5.3.1 Volcanic Origins on Milos
Island Arc magmatism first began on Milos at 2.66 +/-0.07 Ma (Stewart at al.,
2003). The subducting African plate led to the generation of melt in the
asthenosphere, which was aided by the dehydration of hydrous minerals in the
plate and by the addition of water at the subduction zone both of which resulted
in lowering the melting temperature of the mantle. The location of these island
arcs is governed primarily by the depth of the subducting slab, as the quantity of
melt generated increases with depth, largely due to the melting of hydrous
phases which occurs at depths of 100-150km (Philpotts, 1990). The melt
generated rose up through the mantle and ponded below the overlying crust. The
hot magma caused the broadly andesitic crust to melt differentially with the
felsic minerals melting first. The felsic melt generated then migrated up through
the crust aided by faults and fissures and eventually erupted at the surface,
causing crustal differentiation in the crust by leaving a more mafic rich residue.
Adiabatic decompression associated with back arc extension may have also
contributed to the generation of melt in the underlying mantle. Van Hinsbergen
et al., 2004 used the Mckenzie (1978) model to show that the stretching
associated with rapid early Pliocene subsidence proceeding volcanism did not
lead to significant melting of the underlying mantle, as the stretching factor the
model yielded was insufficiently large.
Fig.5.3 The release of water and carbon dioxide from the subducting melting slab
induces melting in the overlying mantle wedge. The extensive normal faults in the
overlying African plate accommodates the migration of the melt to the surface.
30. 26
5.3.2 Volcanic Events
The first phase of volcanism in south central Milos is an andesitic pyroclastic
flow which was interpreted to be representative of an immature magma which
has not undergone significant fractional crystallization. Thus it may be
compositionally similar to the base of the African plate which the melt derived
from. It is likely this is the oldest volcanic event on Milos as it overlies the
Limestone Formation which precedes all volcanism on the island. However it is
likely that this is a disconformable surface and volcanism may have prevailed
elsewhere on Milos before the onset of this pyroclastic flow. The ignimbrite
came from the west, possibly from the same domes as the rhyolitic lavas. It was
interpreted to have been sub-aerially deposited due the presence of a palaeosol
underlying this unit. There was no sub-aqueous features in the ignimbrite.
The second phase is the rhyolitic (QFBRP) lava which covers the entire western
region. It erupted from a series of dome volcanoes the largest being the Halepa
dome ~1.8km west of the western border of the area mapped. These steep domes
were interpreted to be volcanic domes by the palaeoflow analysis which showed
how the lava flowed away from each dome structure often in tongue shaped
bodies 8-24m wide which followed paths of least resistance. These domes do not
have crater structures and no intrusive rocks were found. Fytickas et al., 1986
used K-Ar radiometric dating of biotite to date this lava to 0.95±0.08 Ma.
The third phase is the rhyodacite (QFRDP) lava which may in fact be phase two
of the volcanic events as there in no stratigraphic relationship between these
units. However Fytikas et al. 1986 dated the rocks to 0.48 ± 0.05 by K-Ar
radiometric dating of the biotite phenocrysts. The lava is severely auto-
brecciated on the south coast which is part of the proximal zone of the lava flow.
The level of auto-brecciation led to the indefinite interpretation that this may be
a subaqueous lava deposit. This interpretation was further supported when the
overlying rhyodacite lapilli ash fallout was also interpreted to be a subaqueous
deposit (due to cross-bedding). However, this is a disconformable surface and
schist phreatic deposits are locally deposited between these two units. This lava
originated from the Cape Kalamos dome volcano in the south east, and a series
of cryptodomes and/or coulees in the centre. The Fyriplaka crater may have been
31. 27
the principle source of rhyodacite lava, but this is unclear as it entirely covered
in lapilli ash.
The fourth phase is a localized phreatic deposit that is only exposed at the
Tsigrado cliffs on the south central coast. It is compositionally very similar to
the recent phreatic deposits.
The fifth phase was an explosive eruption dated by Fytikas et al. 1986 to
l0.09±0.02 Ma by K-Ar radiometric dating of the biotites. The cross-bedding
and abundance of non-juvenile material led to the interpretation that it was water
reworked and so must have been deposited under submarine conditions. The ash
diatomite was interpreted to be the distal regions of this deposit, where diatoms
were blooming and the distance from the source was too great for the lapilli
fragments of the deposit to reach.
The sixth and final volcanic phase which is also the ultimate volcanic phase on
the entire island is the phreatic schist conglomerate deposit. No magma was
erupted, instead it was caused by magma superheating overlying groundwater
and fluids within rocks causing them to become superheated causing an
explosive eruption fragmenting rocks into fragments and causing a debris flow
of this material.
32. 28
Acknowledgements
I would like to thank my supervisor Dr. Chris Nicholas and all the lecturers of the
Geology Department at Trinity College Dublin for all their mentoring and dedication. A
special thank you also to my mapping partner Katie Corrigan.
33. 29
References
Dalabakis et al., 1989. Recent explosive episodes on the island of Kos.
Druitt et al., 1999. Santorini Volcano.
Fytickas et al., 1986. Volcanology and petrology of volcanic products from the
island of Milos and neighbouring islets.
Kalogeropoulos & Paritsis, 1990. Geological and geochemical evolution of the
Santorini volcano.
Kokkalas & Aydin, 2013. Is there a link between faulting and magmatism in the
south-central Aegean Sea.
McKenzie, 1978. Some remarks on the development of sedimentary basins.
Philpotts, 1990. Principles of igneous and metamorphic petrology.
Van Hinsbergen et al., 2004. Vertical motions in the Aegean volcanic arc:
evidence for rapid subsidence preceeding volcanic activity on Milos and Aegina.
Stewart & McPhie, 2004. Facies architecture and late Pliocene - Pleistocene
evolution of a felsic island, Milos, Greece.
35. 32
P25340 Muscovite Chlorite Schist
Fig.1 Photomicrograph in cross polarised light showing definite compositional banding, and
altered muscovite and chlorite bands to a dense brown clay material. (Scale bar is 500µm).
Mineralogy
Quartz 55%
Muscovite 25%
Plagioclase 10%
Chlorite 8%
Epidote 2%
Description
This is a medium grained metapelite. It is compositionally banded in sheets of
polygonal quartz bands and muscovite & chlorite bands. The compositional banding is
poorly defined and often elongate muscovite grains intertwine through the polygonal
quartz. This demonstrates the low grade nature of the metamorphism (greenschist
facies) that formed this schistosity. The width of the compositional bands varies from
0.2-2mm. The quartz grains are 0.05-0.3mm in size and display strong undulose
extinction. They have concavo-convex contacts indicating pressure solution. The release
of fluids from pressure solution may have initiated the hydrous breakdown of the
muscovite grains. However it is more likely that hydrothermal alteration caused the
hydrous breakdown of the muscovites to clays given the hydrothermal environment.
Only ~5% of the muscovites are intact and not partially altered to clays. The epidote
crystals are anhedral and are ~0.1mm in size, occur in the felsic bands.
Interpretation
The protolith was a muddy shale deposited on the Tethys ocean bottom. It was
subjected to greenschist facies metamorphism during the Mesozoic due to compression
and elevated temperatures in the crust associated with the collision of the African and
Aegean plates. The schistose texture and alignment of platy minerals parallel to the
cleavage planes is a product of the deviatoric compressional stress that prevailed during
the Mesozoic. This process did not go to completion resulting in the quartz bands often
containing muscovite grains. The compositional bands are not evenly spaced also due to
the low grade nature of the metamorphism.
36. 33
P25341 Lithic Arenite (Basal Sandstone)
Fig.2 Photomicrograph in cross polarized light showing a range of clasts- large angular quartz
clasts (bottom left), polymineralic quartz bands from schist clast (right side), deformed
muscovite bands from schist (middle) and marble clasts (top central). (Scale bar is 500µm).
Composition
34% Quartz (Monocryst:Polycryst ratio 3:1)
26% Rock fragments
18% Carbonate clasts (Marble)
5% Carbonate cement
15% Muscovite aggregates from schist
2% Biotite
1% Opaques
Description
This is a poorly sorted sandstone rich in lithic fragments and non-spherical angular
grains. The quartz grains are angular to sub-angular and display strong undulose
extinction similar to that seen in the schist sample. The carbonate clasts (marble) are
sub-rounded and partially altered containing abundant disordered opaque inclusions.
The schist basement provenance is clearly seen in the muscovite and chlorite aggregate
grains which are plastically deformed. This indicates a very near source as the
muscovite bands would disintegrate over a relatively short transport distance.
Interpretation
This is an immature sandstone of recycled orogenic metamorphic basement. The
provenance consists of at least two sources, the source further away contains the marble
basement, which results in these clasts being more rounded and slightly spherical. The
carbonate pore fluids is interpreted to have derived from these clasts by pressure
dissolution during mesogenesis. Evidence for this is the concavo-convex contacts
between marble clasts which released carbonate pore fluids into the rock. This is also
seen in the polymineralic quartz grains but this is a pre-existing texture of the schist
parent material.
37. 34
P25342 Moldic Bioclastic Wackestone
Fig.3 Photomicrograph in plane polarized light of a moldic fragment of a bivalve which
contains secondary calcite rhombs. The matric consists of fine grained micrite, fine pores and
sparry cement and metal oxide opaques. (Scale bar is 1mm)
Composition Classification
Micrite 75% Grain Size: Bioclastic calcilutite
Porosity 20% Dunham’s: Bioclastic packstone
Pelloids 5% Folk’s: Biomicrite
Secondary calcite rhombs 4%
Quartz 2%
Iron Oxides 1%
Description
Bivalves constitute the bulk of the moldic fossils present ~80%. They range from 1-
6mm in size, model 3mm. Gastropods are 1-1.8mm in size and the foraminifera are
~0.5mm. The quartz grains are ~0.2mm in size, sub-angular and display a strong
undulose extinction, thus interpreted to be sourced from the metamorphic basement.
The pelloids have a dark brown colour and are very fine grained <0.1mm. The
secondary calcite crystals are euhedral and have a maximum size of 0.05mm. They form
thin borders in the fossil moulds.
Interpretation
This rock formed in a shallow tropical ocean with favourable conditions to support the
survival of organisms. The fossils were moderately preserved indicating a slightly
turbulent environment. During diagenesis, slightly acidic porefluids dissolved the
calcium carbonate shells of the fossils. When a regression occurred the environment
became marine vadose. This lead to the precipitation of pure calcite crystals with within
the fossil grains, forming a very thin film around each mould. Crystallization also
occurred within the matrix. Much of the material present was microbially micritized as
seen in the photomicrograph.
38. 35
P25343 Domed Silty Stromatolite
Fig.4 Photomicrograph in plane polarized light showing dark wavy bands of fine grained algal
material displaying stromatolitic layering. Rounded pores are seen in the top region of the
photomicrograph. Scale bar is 1mm in length. (Scale bar is 1mm)
Composition Classification
Micrite 78% Grain Size: intraclastic calcilutite
Algal material 8% Dunham’s: pelintraclastic mudstone
Pelloids 5% Folk’s: pelintramicrite
Porosity 6%
Quartz 2%
Secondary calcite rhombs 1%
Muscovite 0.5%
Iron oxides 0.5%
Description
Stromatolitic banding is seen by the wavy convex bands of algal material-dense dark
brown colour. These laminae are on the micron scale and repeat regularly at ~2cm
intervals. The pores size ranges from 0.05-0.25mm. They have curved wavy shapes and
were interpreted to be primary-formed by the entrapment of biologic gases released
from the microbes. The interior of these pores also contains a very fine thin film of
secondary euhedral calcite crystals-similar to that seen in 6. Moldic Bioclastic
Wackestone- must have been the same secondary calcite mineralization event. The
matrix consists of micrite with some very fine rounded quartz clasts 0.02-0.1mm in size,
sub-angular with undulose extinction-source is also the metamorphic basement.
Interpretation
The depositional environment was an intertidal sabkha, which was environmentally
extreme in some form, possibly hyper-saline. This prevented any other organisms from
thriving and allowed these microbial mats to thrive and be preserved. Carbonate
precipitated on these microbes which preserved the microbes in domes laminae.
39. 36
P25344 Quartz-Feldspar-Biotite Rhyolitic Porphyry (QFBRP)
Fig.5 Photomicrograph in cross polarized light of a biotite phenocryst with consertal intergroths
of plagioclase and k-feldspar which displays oscillatory zoning. Sub-ophitic texture. Biotite has
minor feldspar inclusions. (Scale bar is 2mm)
Mineralogy
Groundmass 65%
Phenocrysts 20% (Biotite 45%, Albite 35%, K-Feldspar 10%, Quartz 10%)
Gas Vesicles 15%
Opaques 2%
Descriptions
Plagioclase: occur as both oikocrysts
This is a porphoritic cryptocrystalline rock with a seriate grain size. Phenocrysts of
biotite are 0.5-3mm, Albite 0.5-4mm and k-feldspar 1-2mm and quartz 0.2-1mm.
Biotite and albite constitute ~80% of the phenocrysts and have a poikilitic relationship
where biotite most often occurs as the oikocryst (host), but albite grains may also
contain biotite chadocrysts. This is an intrusive texture. The crystal boundaries are often
intergrown as consertal intergrowths- coalescing to achieve a more stable energy state
and minimise the surface area. This was interpreted as an extrusive texture which
occurred while the magma was erupting and flowing. The groundmass is locally
partially crystalline where the feldspars in the matrix have begun to devitrify. The albite
phenocrysts display discontinuous normal zoning and the core is anorthite.
Interpretation
The magma initially cooled slowly in the magma chamber resulting in the growth of
abundant large phenocrysts. When the magma erupted it flowed for ~150m before
solidifying resulting in a glassy matrix with a pilotaxtic texture and the phenocrysts
aligned parrallel to the flow orientation. The discontinuous normal zoning in the albite
phenocrysts shows is indicative of abrupt changes in the magma chamber which is
characteristic of rhyolitic lavas, constituting multiple stages of growth as the melt
evolves in the magma chamber and slowly migrates towards the surface.
40. 37
P25345 Rhyodacite (QFRDP) lava
Fig.6 Photomicrograph in cross polarized light showing simple twinning in a plagioclase
polyhedral phenocryst in a pilotaxtic groundmass. The very fine white feldspars in the glassy
matrix is interpreted to be the glassy matrix devitrifying as a secondary equilibration process.
(Scale bar is 2mm, note also the arrow in the top left illustrating the flow orientation)
Mineralogy
Groundmass 60%
Gas Vesicles 35%
Phenocrysts 5% (Albite 70%, Quartz 18%, Biotite 10%, K-Feldspar 2%)
Description
The groundmass has a strong pilotaxtic fabric and is mostly glass due to rapid
quenching. It is a highly vesiculated rock and may not be characteristic of the overall
lava deposit. The average abundance of gas vesicles is ~20%. This is a finer grained
lava than the rhyolite lava. The phenocrysts are mostly albite which range from 0.2-
2mm mostly subhedral. Biotite phenocrysts are much less abundant and smaller than the
albites ranging from 0.05-1mm. They mostly form acicular euhedral grains and are thus
interpreted to be a quench texture of the rock. They rarely form intergrowths with the
quartz and plagioclase grains. Glomerocrysts of albite grains are also common.
Interpretation
Relatively newly formed melt in the magma chamber erupted by the time ~5% of the
melt had crystallised (mainly albite). This prevented the more mafic phases from
crystallising and resulted in a cryptocrystalline rock of predominantly glass, locally
showing signs of devitrification of the fine grained feldspars in the groundmass. The
albite glomerocrysts was interpreted as an extrusive texture which occurred during the
eruption and flow-albite phenocrysts bunched together to achive a more stable energy
state. It could also be a quench texture during the final stages of slow cooling in the
magma chamber.
41. 38
P25346. Rhyodacite Lapilli Stone Block
Fig.7 Photomicrograph in cross polarized light showing a 1.2mm glomerocryst of plagioclase
phenocrysts displaying lamellar twinning. The crystals are slightly altered with microfractures
filled with brown-grey clays. (Scale bar is 200µm)
Mineralogy
Groundmass 75%
Phenocrysts 15% (Albite 45%, Quartz 30%, Biotite 18%, K-Feldspar 5%, Opaques 2%)
Gas Vesicles 10%
Opaques 1%
Description
The groundmass has a strong pilotaxtic fabric and is mostly glass due to rapid
quenching. The phenocrysts are mostly albite which range from 0.2-1mm mostly
subhedral. The biotite phenocrysts are elongate and perfectly aligned to the flow
orientation of the glassy groundmass. The albite phenocrysts form glomerocrysts similar
to those in the rhyodacite lava sample and are also interpreted to be a late texture which
formed during the eruption by the phenocrysts clotting together to minimise surface
area with the disequilibrium surrounding melt.
Interpretation
This magma like the lavas also underwent a prolonged period of slow cooling in a
magma chamber resulting in the growth of phenocrysts. A large gas build-up in the
chamber caused the magma to erupt explosively resulting in a plinian eruption. This
resulted in the groundmass rapidly cooling resulting in a glassy pilotaxtic groundmass.
Fractures formed in the feldspars due to the explosive expulsion of the material into the
atmosphere and the physical impact of the blocks on the surface. These fractures acted
as planes of weakness along which fluids could permeate. This has resulted in hydrous
reactions partially altering the fractured areas to clays as seen in the photomicrograph.
